ANALELE UNIVERSITĂŢII “DUNĂREA DE JOS” DIN GALAŢI UDJG MSM 2-2003.pdfMINISTERUL EDUCAŢIEI ŞI CERCETĂRII ANALELE UNIVERSITĂŢII “DUNĂREA DE JOS” DIN GALAŢI Fascicula

MINISTERUL EDUCAŢIEI ŞI CERCETĂRII

ANALELE UNIVERSITĂŢII “DUNĂREA DE JOS” DIN GALAŢI

Fascicula IX

FACULTATEA DE METALURGIE ŞI ŞTIINŢA MATERIALELOR

ANUL XXI (XXVI), nov. 2003, nr.2

ISSN 1453-083X

MINISTRY OF EDUCATION AND RESEARCH

THE ANNALS OF “DUNAREA DE JOS” UNIVERSITY OF GALATI

Fascicle IX

FACULTY OF METALLURGY AND MATERIALS SCIENCE

YEAR XXI (XXVI), nov. 2003, no.2

ISSN 1453-083X

EDITING MANAGEMENT

RESPONSIBLE EDITOR: Prof. Dr. Eng. Alexandru EPUREANU

ASSISTANT EDITORS: Prof. Dr. Eng. Emil CONSTANTIN

Prof. Dr. Eng. Viorel MINZU Prof. Dr. Eng. Mircea BULANCEA Conf. Dr. Ec. Daniela ŞARPE Conf. Dr. Anca GÂŢĂ

SECRETARY: Assoc. Prof. Dr. Eng. Ion ALEXANDRU

EDITING BOARD

Fascicle IX

METALLURGY AND MATERIALS SCIENCE

EDITOR IN CHIEF: Prof. Dr. Chim. Olga Mitoşeriu

SECRETARY: Prof. Dr. Eng. Marian Bordei MEMBERS: Acad. Prof. Dr. Hab. Iurie Nicolaevich Shevcenko–Director of the Termoplasticity Department, National Academy of Science of Ukraine Acad. Prof. Dr. Hab. Valeriu Kantser–Coordinator of the Technical and Scientific Section of the Academy of Moldova Republic Prof. Dr. Rodrigo Martins – President of the Department of Materials Science, Faculty of Science and Technology, NOVA University of Lisbon, Portugal Prof.Dr.Hab. Vasile Marina–Director of Department, State Technical University of Moldova, Kishinau, Moldova Republic Prof. Dr. Eng. Elena Drugescu Prof. Dr. Eng. Nicolae Cănănău Prof. Dr. Eng. Anisoara Ciocan Prof. Dr. Eng. Maria Vlad Prof. Dr. Eng. Petre Stelian Niţă Prof. Dr. Eng. Alexandru Ivănescu Asoc.Prof. Dr. Eng. Sanda Levcovici AFFILIATED WITH: - ROMANIAN SOCIETY FOR METALLURGY - ROMANIAN SOCIETY FOR CHEMISTRY - ROMANIAN SOCIETY FOR BIOMATERIALS - ROMANIAN TECHNICAL FOUNDRY SOCIETY - THE MATERIALS INFORMATION SOCIETY (ASM INTERNATIONAL)


FASCICLE IX METALLURGY AND MATERIALS SCIENCE, ISSN 1453 – 083X NR II – 2003

Table of content

1.Olga MITOŞERIU, Cătălina ITICESCU - Behaviour at the corrosion of Cu/P composite coating obtained by electroless plating ………………………………………………………………………… 2.Lidia BENEA, Magda LAKATOS-VARSANYI, George MAURIN - The electrolytic co-deposition of zirconium oxide particles with nickel…………………………………………………… 3.Pier LUIGI BONORA, Lidia BENEA, François WENGER, Alberto BORELLO, Stefano MARTELLI, Pierre PONTHIAUX - Tribocorrosion Aspects of Nano - Structured SIC – Nickel Composite Coatings …………………………………………………………………………………… 4.Ion DINESCU, Mihaela OPRESCU - Technologies for future precision strike missile system….. 5.Pingan XIAO, Xuanhui QU - Influence of yttrium addition on the oxidation behaviors of hypereutectoid ti-cr alloys with laves phase TiCr2……………………………………………………... 6.Adriana NEAG, Traian CANTA - Some consideration regarding thixoforming of metal alloys…. 7.Stela CONSTANTINESCU, Viorica MUŞAT, F.Braz FERNANDES - Nitride coatings on widia substrate for mechanical applications………………………………………………………………….. 8.Alina-Adriana MINEA, Ovidiu MINEA, Adrian DIMA - Experimental and theoretical contributions on hardness profile of an AlCu2Mg1,5Ni alloy………………………………………… 9.P. MOLDOVAN, M. BUTU, I. APOSTOLESCU, V. ILIESCU - Gas distribution modelling in gas injection of aluminum melts………………………………………………………………………... 10.Marian BORDEI, Ştefan DRAGOMIR, Aurel CIUREA - Considerations regarding the working of the double charging machines for slabs in the continuous pusher type furnace ………….. 11.Dumitru DIMA - Technology of obtaining composite material samples – polyester matrix ranforte glass fibre and ferrite particles………………………………………………………………… 12.Dumitru DIMA - Iosipescu test applied in order to characterize the composite material pe – GFR – Fe3O4…………………………………………………………………………………………………. 13.Aurel CIUREA, Marian BORDEI, Adrian VASILIU - The influence of metallic mass upon cast iron hot plastic deformation……………………………………………………………………….. 14.Ovidiu DIMA - Research and study regarding the influence of the cold plastic deformation on the structure and the properties of some austenitic stainless steels……………………………………….. 15.Lucica BALINT, Simion BALINT, Tamara Radu, Emil STRATULAT - Analysis of diffusion processes at interface cobalt - steel sheets………………………………………………… 16.Marian BORDEI, Ştefan DRAGOMIR, Marian NEACŞU, Beatrice TUDOR - Improving constructive and exploitation parameters of cast turbo blower rotors at ISPAT SIDEX Galati by studying their dynamic behaviour………………………………………………………………………. 17.Viorel DRĂGAN, Alina CIUBOTARIU, Ioan AXINTE - Influence of the cocsochemics subproducts for corrosion in exchanges heats ………………………………………………………… 18.Gheorghe CADAR, Gina NĂSTASE - Preparation possibilities and utilization of different coal tar pitch types…………………………………………………………………………………………… 19.Adrian VASILIU, Marian BORDEI, Alexandru CHIRIAC - Crushed coke content in the sintering machine charge is considered based on the correlation existing between it and the quantity of recycled material……………………………………………………………………………………. 20.Maria VLAD, Varga BELA, Emil STRATULAT - Recycling aluminium west ……………… 21.Carmen – Penelopi PAPADATU - Some aspects regarding the operating conditions and the temperature determination for the rolls on the contact surface with the rolling slab, during their work. 22.Petre Stelian NIŢĂ, Alexe Corneliu GOARZĂ – An improved model of the surface energy for iron group transitions metals…………………………………………………………………………….

5

10

18 24

28 33

37

44

47

50

54

64

70

74

80

84

88

94

98 101

105

100



BEHAVIOUR AT THE CORROSION OF CU/P COMPOSITE COATING OBTAINED BY ELECTROLESS PLATING

Olga MITOŞERIU, Cătălina ITICESCU

„Dunărea de Jos” University of Galaţi 47 Domneasca Street, Galati, Romậnia

email:@ [email protected]

ABSTRACT

This paper presents a study of copper deposition obtained by electroless plating from alkaline electrolyte, using HCHO as reducer. There were obtained copper deposits from the electrolyte by introducing red P particles. Carbon steel was used as support for deposition. Corrosion behaviour of copper layers was tested by salt spray test. The structural aspects of deposition showed different behaviour depending of the nature and concentration of particles from the electrolyte, and also of HCHO content.

KEYWORDS: copper deposition, corrosion, metallizing

1. Introduction

Copper can be deposited by wet chemical plating techniques such as electroplating and electroless plating. These techniques have the advantage of low cost of tools and materials, low processing temperature, high quality material and high throughput of the process. They are thus suitable methods for low-cost processing. Electroless deposition technique is considered to be more attractive than electroplating. It has a very high selectivity, it uses a very thin seed layer, it has excellent step coverage and good trench filling capability this does not need any electrical contacting of wafers during deposition. The main disadvantages of electroless deposition are: lower deposition rate and a larger amount of produced waste comparing to electroplating.

Considerations on copper electroless

Electroless plating occurs simply by immersion of the samples in a plating bath. No rectifiers, batteries or anodes are involved. The essential elements of the solution are soluble metal salt, and reducing agent, additives such as complexing agents, buffers, bath stabilisers and rate promoters. Electroless plating is an autocatalytic process; that means that metal deposition serves to catalyse the reaction [1,2]. Therefore, the plating bath is capable of plating receptive surface, including plating tank and equipment surfaces.To maintain a constant plating rate, high control of pH and temperature are necessary. Metal and reducing agent concentration

must also be kept at optimal specified levels. Advantages of electroless plating include excellent uniformity, bulk processing capability and ability to produce unique catalytic coatings. Aiming at compactly selective metallizing, copper can be deposited from aqueous solutions under the action of several reducers such as formic aldehyde, boron hydride, hydrazine, potassium hypophosphite or of several redox pairs such as: Fe (II) / Fe (III), Ti (III) / Ti (IV), Cr (II) / Cr (III), V (II) / V (V). For practical applications, formic aldehyde and boron hydride are used with the best results. In this study we used formic aldehyde as reducer. The copper chemical deposition can be performed from alkaline solutions only, and reacts according to the chemical equation below: Cu2+ + 2HCHO + 40H-→ Cu + 2HCOO- + H2 + 2H2O

Molecular hydrogen occurs from the reaction, the Cu / H2 ratio being 1.

2. Experimental

The Copper electroless plating has been prepared

using an electrolyte with the chemical composition: CuSO4·5H2O-30g/l; Na2CO3 anh.-12g/l; Tartrat de Na şi K-150g/l, NaOH-50g/l, EDTA - 6g/l [3]. The formic aldehyde (37%) was used as reducing agent and its amount was varied between 15-25 g/l HCHO. The pH of the electrolyte was 13,50 at the temperature of 20-250C. Carbon steel with the composition presented in the Table 1 was used as metallic support for copper deposition.

5

mailto:[email protected]



The preparation of the samples. The samples have the dimensions 10x5x1mm.

The samples were mechanically cleaned by polishing with abrasive papers, organically (with acetone) and chemical degreased (at 70-800C for 5 minutes in alkaline solution). After degreasing, the samples have been washed with hot distilled water until obtaining a continuous film on surfaces and in the end the activation with HCl 15% (2 minutes). In the electrolyte there were introduced particles of red phosphorus (red P) with a concentration between 2-10g/l. There were also used mixtures from red P and CeO2 particles, of 3g/l and respectively 5g/l each. The particles of CeO2 were kept in electrolyte suspension by magnetic stirring. The electrolyte stirring with red P particles was not necessary, the particles remained in suspension. The modification of the electrolyte pH was not observed when introducing the particles. The structure of the layers was analysed by optical with the microscope Olympus PME 3. The corrosion behaviour of samples was studied by salt spray test.

3. Results and discussions

1. The optimisation of copper electrolyte

It was studied the influence of HCHO content for the same stabiliser (6g/l EDTA) in electrolyte. The alkaline electrolyte is more stable when copper is reduced in less time. The content of reducing agent influences the electrolyte stability and the quality of

the deposits. Good results were obtained using a reducing agent of 22,5 g/l HCHO. The results obtained are presented in the Table 2. Experiments using the electrolyte with particles in suspension were also made. It was observed that red P particles have catalytic actions upon the copper reducing reaction (Table 3). The periods of induction were smaller in these cases. The deposits from the electrolyte with particles are more uniform and resistant if speed of deposition is lower. On the other hand a low speed is not appropriate.

The control of reduction speed is necessary. The copper reduction is partial in the case of using of reducing agent in a bigger amount. The reactions are:

Cu2++2e-→ Cu0

Cu2++e-→ Cu1+

The deposits presented different colures in function of copper compounds, depending of the working condition. Structural aspects

Copper chemical deposits from electrolyte with and without particles in suspension were tested by optical microscopy in cross section (Fig.1). There were observed thin layers of copper deposition between 3-6 μm. The aspect of layers is not completely uniform but this can be explained by destruction of metallic support during the preparation of samples for optical analyses. Samples were tested also by electronic microscopy (SEM).

Fig.1. Optical microstructure by cross section of composites Cu-P layer

Fig. 2. SEM micrography for red phosphorus powder

6



Table 1. Chemical composition of metal support

Chemical element C Si Mn P S Al As Cu Cr Ni

% 0.03 0.40 0.30 0.21 0.01 0.001 0.05 0.20 0.10 0.10

Table 2. Period of induction in copper electrolyte

CH2O, g/L 25.0 22.5 20.0 17.5 15.0 100 Period of induction.

min. 9 17 36 48 48 48

Table 3. Period of induction in electrolyte with particles

CH2O, g/L 25,0 22.5 20.0 17,5 15,0 10,0

with 10g/l P red, min 2 4.0 6.0 7.0 8.0 8.0

2. The corrosion behaviour

Studies of corrosion behaviour of the copper deposits from electrolyte with and without particles in suspension in the salt spray test were made. The samples were put in a circular support and were recto-verso photographed after one hour and five hours. The summary results are shown in the Table 4 and in Fig. 3,4 and 5.

The copper deposits from electrolyte without particles showed an advanced corrosion after only 1-hour of exposure to salt spray (Fig.3-left).

The copper deposits obtained in the same conditions from the electrolyte with red P particles (2g/l) in suspension (Fig. 3-right) showed a lower corrosion percent after one hour. The corrosion process was accelerated in time and after five hours, the corrosion aspect of these samples were similar to the as first samples. Presence red P particles in the electrolyte modifies the nucleation process of deposited copper and then layers have new properties. The reducing agent amount (15 g/l) was not enough and there were partial copper oxidisation compounds. Were obtained these

compounds present been low stability in salt spray corrosion. By modifying the amount of the reducing agent to 22,5 g/l it was observed an improvement of corrosion resistance. There were also obtained copper deposits from electrolytes containing red P particles. The corrosion behaviour of these samples are not significantly improved. The presence of particles in the electrolyte in copper electroless plating influences the formation of the deposits and the stability of the formed films. An important observation is that the deposits had different colours depending on the working conditions. There are necessary more studies. For understanding the phenomena, which occur in the electroless plating in order to find the optimal parameters for deposits with better corrosion resistance. The particle effect in the electroless plating can be explained by studying of polarisation curves. This is possible for deposits with more than 10 μm thickness. For less thickness the curves of polarization and Tafel curves gave uncertain information (Fig. 6,7).

Table 4. The corrosion behaviour of the copper deposits in salt spray from different working conditions

Aspects of corrosion, %

after on hour after five hour Particles

Concentration g/L

HCHO

g/L

Time of deposition

h recto verso recto verso

- 15.0 24 50 10 >50 20 2 15.0 24 30 on edges >50 20

6 15.0 24 50 less

corrosion; spots

90 lateral

intensive corrosion

3g/L P + 3g/L CeO2

175 2 40 less corrosion

90 on edges

7



1 2 1 2 (a) (b)

Fig.3. Aspects layers before treatment of salt spray of copper from electrolyte a) Recto; b) Verso

(1) 5g/L P red (2) and with 10g/L P red

1 2 1 2

(a) (b)

Fig.4 Aspects of corrosion layers after one hour in salt spray

1 2 1 2

(a) (b) Fig.5 Aspects of corrosion layers after five hours in salt spray

8



Stromdichte-Potential-Kurven von Cu-Schichten Elektrolyt: 3,5% NaCl, Vorschubgeschwindigkeit: 0,5 mV/s

-5

-4

-3

-2

-1

0

1

2

3

4

-600 -400 -200 0 200 400 600 800 1000 1200

Spannung [mV]

Stro

mdi

chte

[mA

/cm

2]lo

g 6266-Ru16249-Ru16249-2Ru1

Überschreitung des Meßbereiches

Fig. 5. Voltamograms obtained by copper composite coatings

Cu-P(5g/L)-6266, Cu-CeO2(3g/L)-6249

Fig.6. Tafel diagrams for Cu/P coatings

4. Conclusion

• there were obtained copper deposits by electroless plating using HCHO as reducing agent with and without of red P and CeO2; • corrosion behaviour in salt spray test showed

that the presence of particles in the electrolyte influence the deposits;

• the better corrosion behaviour was observed in the presence of CeO2 (3g/l);

• the presence of red P in electrolyte induced an the advanced corrosion in the formed deposit. observed in the presence of CeO2 (3g/l). References

[1].Mitoşeriu, O., Mitoşeriu,L., Iticescu,C., Preda,A., Chemically Obtained Copper Matrix Composite Coatings, Euromat 2001, 7th European Conference on advanced Materials and Processes, Rimini-Italy 10-14 June, 2001.

[2].Iticescu, C., Mitoseriu, L., Mitoseriu, O., Obtainig and Characterision of Copper Matrix Composite Coatings, Euromat Junior, Lausanne, 2-5 septembre 2002. [3].V.V. Suimidov, T.N. Vorobieva, Chemical Deposition of Metals in Watery Solutions, University Publishing House, Minsk, (1987). [4].O. Mitoseriu ,C. Iticescu, F. Potecasu, L. Mitoseriu, Cu-P Composite Coatings, 3rd International Conference of the Chemical Societies of the South-Eastern European Countries on Chemistry in the New Millenium an Endless Frontier, September 22-25, Bucharest, Romania. [5]Concise Encyclopedia of COMPOSITE MATERIALS, Editor Anthony Kelly, Churchill College Cambridge, U.K., 1994. [6].Baudrand, D., Electroless Processes in AESF Surface Finishing Shop Guide. Orlando:American, Electoplaters and Surface Finishers Society, 1995 [7].Davidoff, C., Metallizing Nonconductors, Metal Finishing Guidebook and Directory,Issue.Vol.94, No.1A. New York: Elsevier Science Publishing Co., Inc., 1996. [8].Shimizu, K., Brown, G.M., and coll., Glow Discharge Optical Emission Spectroscopy –a Powerful Tool for the Study of Compositional Nonuniformity in Elsctrodeposited Films, Corrosion Science, 43, 199-205, 2001 [9].Zhongchun, C., and coll., The influence of Powder Particle Size on Microstructural Evolution of Metal-Ceramic Composites, Scripta Materialia, 43, 1103-1109, 2000

9


FASCICLE IX METALLURGY AND MATERIALS SCIENCE, ISSN 1453 – 083X NR. II – 2003

10

THE ELECTROLYTIC CO-DEPOSITION OF ZIRCONIUM OXIDE PARTICLES WITH NICKEL

Lidia BENEA1, Magda LAKATOS-VARSANYI2, George MAURIN3

1Department of Metallurgy and Materials Science, “Dunarea de Jos”

University, 47 Domneasca St., RO-6200 GALATI, Romania; e-mail: [email protected] 2Department of Physical Chemistry, Eötvös Lorand University,

Budapest 112 P.O. Box 32, H-1518 Hungary 3UPR15 du CNRS, Physique des Liquides et Electrochimie” Associé a

l’Université Pierre et Marie Curie, 75252 Paris Cedex 05, France

ABSTRACT

Composite electroplating is a method of co-depositing insoluble dispersed particles of metallic or non-metallic compounds with a metal or alloy in the plating bath to improve material coating properties such as corrosion resistance, lubrication or wear-resistance. With the aim of producing new electro - catalysts for the hydrogen evolution reaction (HER), and possibly other electrochemical reduction processes, we have investigated the cathodic deposition of composites in which an electrodeposited metal is the matrix and a transition – metal oxide is the dispersed phase. This paper describes the electrolytic codeposition of zirconium oxide particles with nickel on a cylindrical electrode. This system was selected because nickel is an industrially important coating material on steel as cathode in many electrocatalytical processes such as water electrolysis for hydrogen production. The cathodic polarization curves in electrolytes have been plotted both in the presence and absence of the insoluble dispersed phase. The electrochemical impedance spectroscopy method was used to obtain additional information on the early steps of nickel and nickel matrix composite electrodeposition. Measurements of impedance data were performed with Solartron type electrochemical interface and frequency response analyzer. A schematic co-deposition mechanism is proposed considering the experimental observations. The influence of zirconium oxide on the nickel electrodeposition steps is discussed.

KEY WORDS: Composite Coating,, Zirconium Oxide, Nickel, Electrodeposition, Electrochemical, Impedance.

1. Introduction Several methods have been developed to produce metal matrix composites. The most widely used is a mechanical mixing method followed by sintering of metallic and non-metallic powders such as oxides or carbides. Another very frequently used method consists of mixing ceramic powders directly with the fused metals or alloys. Composite electroplating is a method of co-depositing insoluble dispersed particles of metallic or non-metallic compounds with metals or alloy in a plating bath, to improve the material coating properties such as corrosion resistance, lubrication or wear-resistance. Such a coating features the properties of both the metal and the dispersed particles. These coatings can be considered to be

metal-matrix composites (MMC), obtained through electroplating. Composite coatings obtained by metal co-deposition of various dispersed phases during electrocrystallisation have been given special attention in recent years [1-10]. Nickel, copper, chromium, iron, cobalt, silver, gold were mainly used as a metal matrix. Metals, metal oxides, carbides, borides and polymers as co-depositing dispersed particles were used. It is known that carbon steel protected with nickel coating is used as electrode for the hydrogen evolution reaction (HER). In order to improve the properties of such materials we studied the possibility of obtaining composite coating by electrodeposition of nickel with dispersed particles. The composite system considered in this study have a nickel matrix and zirconium oxide as dispersed phase. Our aim here is:




11

-To show the effects of dispersed particles on the mechanism of metal electrocrystallisation and on surface morphology of coating obtained; -To describe the influence of zirconium oxide dispersed phase on the intermediate steps of metal (nickel) electrodeposition; -To study the influence of dispersed phases on the intermediate steps of metal electrocrystallisation and on surface structure of the composite coating; -To invesigate the influence of the operating parameters (current density, concentration of the dispersed phase in the electrolyte) on the concentration of the dispersed phase in the composite coating by means of mathematical equations. -To obtain a new material as electrode for the hydrogen evolution reaction (HER) and possibly other electrochemical reactions. Cathodic polarization curves and electrochemical impedance spectroscopy diagrams have been used to study the influence of zirconium oxide as dispersed particles on the mechanism of nickel electrocristallysation. Microstructural and surface morphology information were used to compare the ZrO2-Ni composite coating with pure nickel deposit. Comparisons of both coatings are also presented for HER in acid solution.

2. Experimental Set -Up

2.1. Electrolytic Cell

Fig. 1. Scheme of the experimental set-up for the

electrodeposition of composite coatings: (1) electrolytic cell; (2) electrolyte and dispersed

phase suspension; (3) anode; (3') anode surface cross-section; (4) cathode; (5) copper electrical contact with taper inside the cathode; (6) reference electrode; (7)

insulating system for maintaining the cathode; (8) stirring device

To ensure an uniform current distribution during the electrodeposition process, a 300 cm3 cylindrical electrochemical cell was used having a cylindrically- shaped anode at 20 mm from the cathode surface, Figure 1 The cell provides good thickness uniformity of the metallic and composite coating. The dispersed particles have uniform access to the cathode surface thus the effects of the different current densities and therefore different coating compositions at the edges of a planar electrodes are eliminated. Holes of 5 mm diameter are drilled in the anode to ensure o good suspension uniformity in the electrolyte and to reduce the effective anode active surface. This electrochemical cell also excludes any forced co-deposition by sedimentation. Electrodeposition was made on carbon steel, stainless steel, copper and nickel as supports (cathodes) having a cylindrical shape and an effective area of about 10 cm2. In order to investigate the dispersed phase influence on metal electrodeposition, the cathodic polarization curves with and without dispersed phase were plotted. Cathodic polarization experiments were performed on a potentiostat-galvanostat type EG &G PAR system by means of a computer program for cyclic voltametry and current voltage diagrams plotting. The value of the metal reducing voltage (electrode potential) is given with respect to the saturated calomel electrode (ESC) as reference. Electrochemical impedance diagrams were plotted during electrodeposition process in order to have additional information on co-deposition mechanism.

2.2. Preparation of nickel and nickel composite coatings

Pure nickel and nickel zirconium oxide co-depositions were made in common nickel plating electrolytes (sulfate and chloride). The electrolyte was prepared from p.a. chemicals and distilled water, which provided the required purity for the potentiodynamic investigations and characterizations of the coatings obtained. Pure dispersed zirconium oxide (ZrO2) at different concentrations (50 - 100 g/l) was suspended in the electrolysis bath. The average particle size was 10 µm. Thickness of metal and composite deposits were obtained between 50 and 150 µm and were verified by measuring the weight before and after deposition and also by light microscopy on cross section. The particles were kept in suspension by mechanical or magnetic stirring at a rotation rate between 100 and 1500 r.p.m. A saturated calomel electrode was used as reference electrode (SCE) in order to determine the influence of the dispersed phase on nickel electrodeposition.



12

2.3. Structural and chemical analyses Weighing the electrodes on a microbalance and stripping the deposit we could determine the amount of dispersed phase in the composite coating. The particles were filtered, dried and weighted. The weight percentage of the particles in the composite deposition was calculated by the formula:

100mm

%ps

p= (1)

Where: p% - represents the weight percentage of the particles in the composite coating; mp is the particles mass, in grams; ms is the total composite coating mass, in grams. To determine the influence of the working parameters (electrolyte concentration, dispersed phase concentration, current density) on the inclusion of the dispersed phase in the composite layer, electrodeposition was performed on a platinum cathode, thus eliminating the possibility that the support could dissolve introducing errors in the dispersed phase content. The uniformity of dispersed phase distribution in the composite coatings was examined by light microscopy in cross section. Scanning electron microscopy (SEM) revealed the comparative surface morphology of coatings and the uniformity of zirconium oxide particles in the composite deposit.

3. Results and Discussion

3.1. Cathodic polarization curves The disperse phase influence on the nickel electrodeposition, was observed on cathodic polarization curves plotted in presence and absence of dispersed phase. No such measurements could be found in the literature for purposes of comparison; thus the interpretations of our polarization curves has been based on existing theories of metal and alloy electrocrystallisation [9-13]. From the analysis of cathodic polarization curves in the presence and absence of dispersed zirconium oxide the following remarks can be made: -Dispersed zirconium oxide phase does not alter the shape of the cathodic polarization curves during nickel electroplating. These curves are only slightly shifted to lower values of the metal reducing voltage, see Figure 2. -Zirconium oxide as dispersed phase in the nickel electrolyte change the electrocrystallization mechanism of nickel by increasing the metal deposition rate. For example at the same current density metal reduction over-voltage is lower in the

presence of zirconium oxide in the electrolyte, as it is shown on diagram (b) from Figure 2. -The larger amount of metal being deposited in the presence of zirconium oxide as dispersed phase can be due to the electrically active species which are absorbed or formed on the particle surfaces to be subsequently reduced on the cathode surface.

Fig. 2. Cathodic polarization curves for obtaining

composite layers in a nickel matrix; (a) nickel-plating electrolyte; b) nickel-plating

electrolyte + 100g/l zirconium oxide -On the growing coating of metal, which is the result of a competition between the nucleation steps and crystal growth, the zirconium oxide acts as a catalyst of metal reduction leading to an increase in the number of active nucleation sites. As it is well known, an increase in the current density leads to an increase in the nucleation sites. The crystal size gets smaller within the electrodeposited metal structure. At the same current density the effect of the oxide involved in electrochemical nickel deposition could induces the same growth of the active nucleation sites. This conclusion further suggests that the oxide involved in the metal electrodeposition will change their structure. The modification in the size of crystallites during composite oxide electrodeposition was checked by investigating the surfaces with respect to the pure nickel electrodeposited at the same current density by means of the scanning electronic microscope. The decrease in the crystallite size could have a different influence on the reactivity of composite surface compared with pure nickel surface. 3.2. Impedance diagrams during electrodepositions

The electrochemical impedance spectroscopy (EIS) method was used to obtain additional information on the early steps of nickel deposition and influence of dispersed particles on them. The electrodepositions of pure nickel and nickel in the presence of zirconium oxide have taken place at the same over-voltage of -1150 mV (ESC). For the same cathode surface, the current was 50 mA in the



13

case of pure nickel electrodeposition and it subsequently increased to 87 mA in the presence of zirconium oxide in electrolyte. There is a good agreement with the cathodic polarization diagrams showing the increase in the deposition current at the same over-voltage. The complex plane impedance diagrams obtained for pure nickel electrodeposition and nickel with zirconium oxide co-deposition composite are illustrated in Fig. 3. The shape of diagram in presence of oxide particles is changed on the low frequency range only and as well as the charge transfer resistance is lower.

Fig. 3. Complex plane representation of the impedance diagrams at the same reduced overvoltage:

(B)-solid up-triangles: composite coating co-deposition with ZrO2 particles in a nickel matrix;

(A)-solid cercles: pure nickel depositions On Figures 4(A) and 4(B) the capacitance and conductance curves versus frequency plotted in the same conditions are presented in logarithmic scale.

Fig. 4(A). Impedance data for the composite coating

electrodeposition: [b] (Δ) 1/Rp against frequency; [c] (o) Cp. against frequency.

Solid circles are the negative values

Solid circles or squares on the capacitance against frequency curves represent its negative values. As it can be seen the impedance diagrams for the pure nickel and composite electrodeposition have the same shape, this means that the time constants for the intermediate steps take the appropriate values.

This confirms that dispersed particles do not modify the metal electrocrystallisation steps they only activate the charge transfer process. In Table 1 the values of charge transfer resistance and double layer capacitance calculated from impedance diagrams are shown.

Fig. 4(B). Impedance data for pure Ni

electrodeposition; [b]-(down triangle): 1/Rp versus frequency; [c]-(square) Cp versus frequency.

Solid squares represent the negative values

Table 1. Impedance data calculated from the diagrams ploted for pure nickel and zirconium oxide

composite coatings.

Type of coating⇒ Calculated data

⇓

Pure Ni

Ni with ZrO2

Double layer capacitance Cdl [ µFcm-2.]

59 53

Charge transfer resistance ]cm[R 2

TS Ω

8

5 The charge transfer resistance value is smaller in the case of zirconium oxide co-deposition with nickel, having the value of:

2CT cm5R Ω=

The charge transfer resistance is higher for pure

nickel electroplating, RCT =8Ωcm2 respectively: 2

CT cm8R Ω= .

3.3. Structural aspects of zirconia nickel matrix composite coatings

Micrographies presented allow comparison between a pure nickel coating (Figure 5a) and nickel with zirconium oxide composite coating (Figures 5b, 5c, 5d at different magnification), carried out with 100 gl-1 of ZrO2 in the plating bath.



14

Fig. 5 a. SEM morphology of pure nickel coating

Fig. 5b. SEM Surface morphology of ZrO2 + nickel composite coating

Fig. 5c. SEM Surface morphology of ZrO2 + nickel composite coating

Fig. 5c. SEM Surface morphology of ZrO2 + nickel composite coating

The pure nickel deposit has a rather regular surface, whereas the composite coating develops in a nodular disturbed surface structure. The zirconium oxide particles are clearly visible on the surface with a homogeneous distribution (white particles). The X-ray disperse energy analysis spot on a white particle have revealed a spectra with zirconium but also with nickel (Figure 6). This can suggest that particles are not free on the coating surface but they have a thin layer of nickel. The nickel layer can be done from the reduced nickel ions at the cathode.

Fig. 6. X-ray spectra analysis on a zirconium oxide (white) particle embedded in a nickel matrix

The surface structure morphology of composite coating obtained by two scanning electronic microscopy techniques is presented on the Figure 7. The surface analysis of Ni + ZrO2 composite coatings is not the most suitable method to determine the presence of zirconium oxide in the nickel matrix, but it is appropriate to observe the morphology and surface structure changes.



15

(a) (b)

Fig. 7. (a) Back - scattering SEM image;

(b) Diffuse electron image One of the most reliable methods to observe the distribution of zirconia particles in the nickel matrix consists of studying the cross section of a deposit. The microstructure performed shows the presence and distribution of zirconium oxide particles (8 mass %) in the nickel matrix see Figure 8.

Fig. 8. Optic microstructure of nickel with zirconium oxide composite with 8% ZrO2, cross section of the

coating, magnification 500x On optic microscopy the zirconium oxide appear as black points and are uniform distributed inside of the composite coating.

3. 4. Co-deposition mechanism and steps In the absence of an electric field on the cathode surface immersed in the electrolyte, there is an

adsorbed layer of ionic species, which includes hydrated metal ions and ions from the particles adsorption area. Particles are absorbed on the cathode surface due to the ions absorbed on their surfaces. Bivalent metal transition from ion state to metal state takes places in two steps of electron transfer which call for a lower energy barrier than only one step of two ion changing. In weakly acid or neutral solutions they form chemical active species with the hydroxyl ions (OH--) of the type (MOH+) and M(OH)2 which exist in sufficiently high concentration on the cathode surface to compete with the free metal ion during the mass transfer. Nickel electodeposition is strongly influenced by hydrogen discharge and can take place according to the following intermediate steps, presented in the literature [14, 15]:

++ ⎯→⎯+ ads2 NieNi (2 a)

NieNiads ⎯→⎯++ (2 b) +++ +⎯→⎯++ adsads

2 NiNie2NiNi (2 c) ( ) 2HeH2 ⎯→⎯++ (2 d)

The surface-active species are the adions +adsNi ,

more or less solvated or complexed. is proposed as intermediate specie in the nickel dissolution and deposition mechanism [16].

adsNiOH

Hydrogen adsorption and inclusion into the deposit inhibits hydrogen development according to the reaction (2d). The influence of zirconium oxide particles on the cathodic polarization curves (Fig. 2) and impedance diagrams (Fig.3) reveals the activation effect of nickel reducing reaction, the steps 2a and 2b. Also the reduction of hydrogen can be taking into account as an activated step (reaction 2d) . During the electrochemical co-deposition the zirconium oxide particles catalyse the adsorbed intermediate step according to reaction (2a), thus decreasing the influence of charge transfer for the final reaction (2b). This can be easely observed on the impedance diagrams plotted in the complex plane (Figs. 3) and in Table 1. Metal electrodeposition rate and crystallite growth is a function of the over-voltage (deposition potential), just like other electrode processes. The basic elements of their relation were found by Erdey- Gruz and Volmer [11]. They determined the way in which the rate of the entire process of incorporating the hydrated metal ion into the metal network affects the shape of the current-potential curves. Generally, far enough from the equilibrium potential where metal electrocrystallisation usually takes place the essential step of metal electrodeposition rate is the neutralization of the



16

metal ion by transferring the electron from the cathode to the metal ion. The influence of zirconium oxide dispersed particles on the cathodic polarization curves and impedance diagrams have suggested that composite surface will have a higher surface activity compared with pure nickel surface.

3.5. Mathematical correlation of deposition parameters

Many factors can determine the percentage of dispersed phase in the nickel matrix composite coating. The most important are the concentration of dispersed zirconium oxide particles in the electrolyte and the current density. Considering that the problems related to electrolyte composition and stirring have been solved co-deposition of zirconium oxide particles (weight percentage in the composite layer) for different composite systems may be described mathematically by an equation, which depends on the factor being considered. The experimental results of zirconium oxide content in the deposition versus current density are shown in Fig. 9, by the solid dots. The best-fitting equation over the experimental results is the following:

2DxCx1

BxAy++

+= (3)

Were: y = particle concentration in the composite coating (mass %); A, B, C and D = constants specific to the composite system being electrodeposited; x = current density, i in Acm-2. The constants from equation (3) were calculated from the experimental data by means of a regression computer program (Table 2). The equations used to calculate zirconium oxide concentrations in the composite layers are in good agreement with the experimental data, Figure 9. The line describes the dependence of the dispersed phase on the current density while solid dots represent the experimental results.

Fig. 9. Checking the equations for calculating the dispersed phase content in the composite coating versus current density: solid dots represent the

experimental results and solid line the fitting curve The very good agreement between the equations and the experimental data is shown by the regression factor value, column 5 in Table 2.

Table 2. Constants A, B, C, D and the correlation coefficient from the equation (3) of the dispersed

phase concentration in the nickel composite coating versus current density, after correlating the

experimental results.

A B C D Correlation Coefficient

0.25 3.13 -0.92

0.32

0.99

A different equation was obtained for the zirconium oxide content, mass %, in the composite coating depending on the disperse phase concentration in the electrolyte see Figure 10.

Fig. 10. Checking the equations for calculating the dispersed phase content in the composite coating

versus their concentration in the electrolyte. Solid dots represent the experimental results and

solid line the fitting curve The solid dots represent the experimental results. The best equation fitting the experimental results has the form of a saturation model. The concentration of zirconium oxide in the deposit is linear depending on its concentration in the electrolyte up to about 200 gl-

1. Over this value the increasing of zirconium oxide particles in the electrolyte have small effect on their concentration in the deposit. The fitted curve (solid line in Figure 10) is described by the following equation. (4) DCxBeAy −−=



17

Were: y = particle concentration in the composite coating (mass %); A, B, C and D = constants specific to the composite system being electrodeposited; x = particle concentration in the electrolyte Cp in gl-1. The equation parameters are presented in Table 3. There is a good correlation coefficient with a value of 0.99.

Table 3. Constants A, B, C, D and the correlation coefficient of the equation (4) describing the

dispersed phase content in the composite coating versus its concentrations in the electrolyte, after

correlating the experimental results

A B

C

D Correlation coefficient

18.39 17.00

0.00 1.57 0.99

The results presented in Fig. 10 are obtained at the optimum current density of 0.02 Acm-2. The equations are applicable to the concentration ranges studied for zirconium oxide co-deposition according to the diagrams presented. Because the mathematical models presented do not include all parameters in a simple equation, we believe them to be much more useful and easy to check for practical control applications involving co-deposition of a composite coating.

4. Conclusions Experimental data demonstrated a strong dependence of zirconium oxide particle co-deposition on solution concentration of dispersed phase, current density and other factors of metals electrocrystallisation. The experiments have shown the possibility of obtaining composite layers having a nickel matrix by metal electrodeposition with inert particles of zirconium oxide, featuring different effects in the intermediate steps of electrocrystallisation, which cause different structures of coatings obtained. The pure nickel deposit has a regular surface structure whereas surface of composite coating presents some nodules and a disturbed structure. The surface morphology of composite coating is uniform and the zirconium oxide particles are definitely visible on the surface but with a thin layer of nickel on them. They are disturbed in a homogenous way on the surface as well as inside of the coating, evidenced by electronic microscope images and optic cross section images. The zirconium oxide acts as a catalyst for nickel reduction, resulting in more nucleation sites. It can therefore be concluded that metal electrodeposition

from electrolyte solutions is activated by means of the electrically active ions adsorbed on the oxide surface. Zirconium oxide dispersed in the nickel deposition electrolyte does not change the electrocrystallisation steps of nickel, but it does take part in the process by increasing the metal deposition rate. For the same current density, metal reducing overvoltage is lower in the presence of oxide particles in the electrolyte. The rate of particles inclusion increases with their concentration in the electrolyte, and has a saturation in the vicinity of 200 gl-1 . The current density have a different influence on the rate of inclusion. After an increasing and a maximum the further higher current densities have the effect of decreasing the particle content in the composite coating. Two equations were established between current density, concentration of particles in the plating bath and the dispersed phase content inside of the composite coating. Equations were derived to describe dispersed phase content which were easy to check and apply in practice, for a given electrolyte composition and stirring rate.

References [1] N. Guglielmi, J. Electrochem. Soc. 8, 119 (1972) 1009-1012. [2] Lidia BENEA, Pier Luigi BONORA, Alberto BORELLO, Stefano MARTELLI, François WENGER , Pierre PONTHIAUX, Jacques GALLAND; J. Electrochem. Soc.., 148 (7) C461-C465 (2001) [3] BENEA L., BONORA P.L., BORELLO A., MARTELLI S.; Materials and Corrosion, 53, 23-29 (2002) [4] LIDIA BENEA, PIER LUIGI BONORA, FRANÇOIS WENGER, PIERRE PONTHIAUX, JACQUES GALLAND "Processing and Properties of Electrodeposited Composite Coatings: Results and Perspectives" KEY NOTE LECTURE, CD-ROM PROCEEDING: 15TH International Corrosion Congress -Frontiers in Corrosion Science and Technology, Granada, Spania, 22-27 Septembrie 2002 [5] L. Benea, Composite Electrodeposition -Theory and Practice, Ed: PORTO FRANCO, Romania, ISBN 973 557 490 X, (1998). [6] S. W. Watson; J. Electrochem Soc., 140, 2235 (1993). [7]. G. Maurin and A. Lavanant; J. Appl. Electrochem., 25, 1113-1121 (1995) [8] J. Fransaer, J. P. Celis and J.R.Roos; J. Electrochem. Soc., 139, 413-425 (1992). [9] J. Fransaer, J. P. Celis and J. R. Roos; Metal Finishing, 91, 97-100 (1993). [10] L. Benea and G. Carac, Metallurgy and New Materials Researches V No 2, 1-19(1997). [11] T. Erdey-Grúz ; Kinetics of Electrode Processes, Ed. Akadémiai Kiadó, Budapest, English Ed. Published by A. Hilger Ltd. London, 350-430 (1972). [12] O. Radovici; Tratat de Chimie Fizicã, Vol 4, Electrochimie, Ed. Academy, Bucuresti, (1986). [13] J. O 'M. Bockris; Fundamental Aspects of Electrocrystallization", Plenum Press, New-York, (1967). [14] A.J. Bard, Encyclopedia of Electrochemistry of the Elements Vol. III, Dekker, Bard (1973). [15]. V.V. Skorchelleti, in Theory of Metal Corrosion, Leningrad 1973, Translated from Russian by R.Kondov , 108 (1976). [16] R. Wiart; Electrochimica Acta, 35, 1587-1593 (1990).



18

TRIBOCORROSION ASPECTS OF NANO - STRUCTURED SiC – NICKEL COMPOSITE COATINGS

Pier LUIGI BONORA1, Lidia BENEA2,1,3, François WENGER3, Alberto BORELLO4, Stefano MARTELLI5, Pierre PONTHIAUX3

1 Dept. of Materials Engineering, Trento University, 38050 Via Mesiano 77, Italy

2 Dept. of Metallurgy and Materials Engineering, Dunarea de Jos" University of Galati, 47 Domneasaca St., 6200, Romania, e-mail: [email protected] .

3 Laboratoire C.F.H., Ecole Centrale Paris, F-92290 Châtenay-Malabry, France. 4 ENEA, C.R.Casaccia, Divisione Nuovi Materiali, Rome, Italy 5 ENEA, C.R.Frascati, Divisione Fisica Applicata, Rome, Italy

ABSTRACT

Advances in materials performance often require the development of composite

systems. Coated materials could be one form to use. The abrasion and corrosion resistance of components can be greatly increased by protective coatings and this is a growing industry of considerable economic importance. These paper aims with a tribocorrosion study of Ni-SiC nano - structured composite coating obtained by electrodeposition. A plan (alumina counterface) on cylinder (coating system deposited on steel) tribocorrosion apparatus was used. All the experiments concerning the effects of the rotation speed and of the applied load on the corrosion behaviour and friction coefficient were carried out using the 0.5M Na2SO4 solution. The Ecorr values measured at different rotation speed of the cylindrical specimen and different applied loads show some differences between the types of coatings studied. Both measurements of electrochemical corrosion and friction coefficient show a better resistance of nano - structured composite coating compared with pure nickel coating. The nanocomposite coating show a bigger polarisation resistance and reduced corrosion current density compared with pure nickel coating in hydrodinamic conditions. The friction coefficient of nano composite coating is smaller compared with pure nickel coating. The results could be considered as the effect of nanoparticles embedded and the fine nano-structured coatings resulted.

KEYWORDS: Wear-Corrosion, Composite, Coating, Nickel, Silicon Carbide, Nano-Particles.

1. Introduction

Coatings are used in both aqueous and high temperature applications. Electric power generation, and waste incineration involve severe conditions and thick coatings have proved effective. Diesel and gas turbine engines are subjects of the high temperature corrosion and highly beneficial coatings have been developed. Some nuclear power systems also rely on coatings. Factors that must be taken into account include substrate compatibility, adhesion, porosity, the possibility of repair or recoating, interdiffusion, the effect of thermal cycling, resistance to wear and corrosion, and not at last the cost. Our paper aims with a comparative tribocorrosion study of Ni-SiC nano - structured composite and pure nickel coatings obtained by electrodeposition.

Deposition of electrochemical composite coatings [ECC] is not a newly developed technique [1], but has been in continuous development since the 1970’s [2-9]. The steady interest is explained by easy maintainability and low cost of ECC manufacture as well as by a possibility of changing the properties and adapting them to many applications. The fabrication of nano - structured composite films can be achieved through electrochemical deposition of the matrix material (e.g. metal, alloy, semiconductor, oxide, conducting polymer) from a solution containing suspension of particles such as: oxides, carbides, nitrides, metal powder and s.a. This technique has been used for the fabrication of our nano - structured composite coating tested, using nanometer size SiC (20 nm mean diameter) dispersed particles.




19

Due to their high wear resistance and the low cost of ceramic powder, Ni-SiC composites have been investigated to the greatest extent and successfully commercialised for the protection of friction parts system. The most recent works on the Ni-SiC system are developed on the modulated current in order to obtain SiC gradient distribution [8] or by-layers with different SiC content [9]. The dimensions of dispersed SiC powder are in the range of micrometers and the shapes from round to acicular or a mixture of them are reported. The present work was aimed at extending these researches by using (for the first time) nano-crystals of silicon carbide (mean diam. 20nm) synthesised by ENEA Italy. Prior to the wear corrosion tests, the composition and surface morphology of coatings were investigated in the scanning electron microscope with the aid of EDS [10]. The presence of particles inside of composite coating was detected by X-ray diffraction and directly observed by TEM. The wear corrosion study (tribocorrosion) was performed to compare the pure nickel and Ni-SiC nano - structured composite coatings. All the experiments concerning the effects of the rotation speed and of the applied load on the corrosion behaviour and friction coefficient were carried out using the 0.5M Na2SO4 neutral solutions. The Ecorr values measured at different rotation speed of the cylinder and different applied force differences between the investigated coatings.

2. Experimental set-up For tribocorrosion tests a plan (alumina counterface) on cylinder (coating system deposited on steel) tribocorrosion apparatus was used. A PMMA cell contained the test solution and a transmission shaft with the tested sample, in the shape of a cylinder (diameter 40 mm and height 18 mm), connected to an electrical motor. A moving rod held an alumina parallelepiped counterface and the imposing load system thus obtaining a sliding type wear system. The schematic set-up of wear-corrosion test is shown in Fig. 1.

Fig. 1. The schematic presentation of wear-corrosion

system

The samples were of two types: i) Ni+ SiC nano structured composite coating and ii) pure nickel as coating electrodeposited on carbon steel. The support material was a cylinder of carbon-steel with a diameter of 40 mm and 18 mm high. Coating thickness were 100 μm for each type of coating. The rotation speed of the cylinder samples was varied between 0 and 250 rpm. The normal applied force was varied between 5 to 40 N. As electrolyte was used 0.5M Na2SO4 solution (pH=5.7) open cell. Electrochemical experiments were performed using a potentiostat EG&G PAR 273A connected with a frequency response analyser SOLARTRON 1255. A platinum counter electrode and an Ag/AgCl (207 vs. SHE) reference electrode were also used. The wear parameter was registered as torsion force function versus rpm on computer software at the same time with electrochemical measurements.

3. Results and Discussion

3.1. Electrochemical measures The Ecorr values measured at different rotation speed of the cylinder specimen and different normal loads show some differences between the two types of coatings. The starting value of corrosion potential is more positive for nano - structured composite Ni+SiC, Ecorr= -198 mV (Ag/AgCl), than that for pure nickel coating, Ecorr= -260 mV (Ag/AgCl), as you can see on Figure 2.

Fig. 2. Corrosion potential evolution of nano -

structured composite (up line) and pure Ni (down line) coatings in 0.5M Na2SO4 at different rotation

speed, without normal friction load The corrosion potential trend of nano - structured composite coating is to be shifted to more positive value by increasing the rotation speed of the sample, Fig.2 (up line). Oppositely, the corrosion potential of pure nickel coating is shifted to more negative values by increasing the rotation speed of the cylinder sample, Fig.2 (down line). This can suggest that even without an applied normal load, the



20

increasing of rotation speed activate the surface of nickel coating by removing easier the corrosion products from the surface. The shift to positive values for nano - composite coating could be explained by the formation of corrosion products, which protect the surface. In our case we consider that during rotation of the cylinder the solution is active enough to form silicon oxide or silicon hydro-oxide on composite surface. These oxides or hydro-oxides protect the composite coating surface. The impedance diagrams performed without a normal load but at 150 rpm, constant rotation of the sample are shown in Figure 3. Polarisation resistance of nanocomposite coating is 80 kΩcm2, higher that than for pure nickel layer of 26 kΩcm2.

Fig. 3. Impedance diagrams performed in

0.5M Na2SO4 at 150 rpm, before applying a normal friction load: (solid circle) nanocomposite coating, Ec =-117 mV (Ag/AgCl) , (solid square) pure Ni

coating, Ec= -220 mV (Ag/AgCl) The differences between the two types of coatings was observed also in potentiodynamic diagrams performed from –600 mV to 0 mV at a sweep rate of 0.5 mV/s. The cylinder sample was rotated at constant rotation speed of 150 rpm. During cathodic polarisation the oxide layer was reduced from the coating surface. The corrosion potential, corrosion rates and Tafel slopes calculated from diagrams are presented in Table 1. Table 1. Corrosion potential, corrosion rate and Tafel slopes calculated from potentyidinamic diagrams for

Ni+SiC and pure Ni coatings at 150 rpm. Type of coating

EcorrmV (Ag/

AgCl)

icorr

μAcm-2BBa

mV/dec BBc

mV/dec

Ni+SiC -212.5 1.9 217 236 Pure Ni -281.6 4.1 76 374

By constant normal load of 30 N both corrosion potentials are shifted to more negative values, which shows an activation of the coating surfaces, Fig. 4. The nano - structured composite coating’s corrosion potential have more positive values than that of pure nickel layer and the rotation speed of sample with constant normal load has not an

important effect. From the potentiodynamic diagrams performed during a normal load of 30 N and constant rotation speed of cylinder (150 rpm) the values of corrosion potential, corrosion rate and Tafel slops have been changed for both type of coatings, see Table 2.

Fig. 4. Corrosion potential evolution of

nanocomposite (a) and pure nickel (b) coatings in 0.5M Na2SO4 at different rotation speed,

with 3kg applied sliding load.

Table 2. Ecorr, icorr and Tafel slops calculated from potentiodynamic diagrams for Ni+SiC and pure Ni

coatings at 150 rpm, after sliding load of 3kg Type of coating

EcorrmV (Ag/

AgCl)

icorr

μAcm-2BBa

mV/dec BBc

mV/dec

Ni+SiC -246.9 0.74 54 80 Pure Ni -276.3 7.08 91 408

The sliding applied load activates the corrosion rate for pure nickel coating by increasing the corrosion current density. For nano - structured composite coatings the corrosion rate seems to go slowly down, even with the shift of corrosion potential to more negative values.

Fig. 5. Impedance diagrams performed in

0.5M Na2SO4 at 150 rpm, after 3kg normal load: (solid circle) nanocomposite coating,

Ec =-224 mV (Ag/AgCl) , (solid square) pure Ni coating,

Ec= -285 mV (Ag/AgCl



21

The performed impedance diagrams during the sliding normal load of 30 N, show the same trend. The polarisation resistance of nanocomposite coating is higher than that of the pure nickel coating, Fig. 5.

3.2. Wear measurements

The microhardness of the plated nano-structured composite coatings have been determined by optic microscopy on the samples, made on a cross section of the coatings. Nickel silicon carbide nanocomposite coatings revealed a microhardness between 400 and 450 HV(0.025) compared with pure nickel coating which allows a value of 148 HV(0.025). Friction coefficient values are different for the two types of coatings. Nanocomposite coating have a lower friction coefficient compared with pure nickel coating, as it was calculated from diagrams after experiments. From Figure 6 we can see that at a constant normal sliding load, the friction coefficient is slowly dependent on rotation speed of the sample for nano - structured composite coating. For pure nickel coating it is very high especially at low rotation speed of 50 rpm, layer (A) from Fig. 6.

Fig. 6. Friction coefficient of nano - structured

composite (a) and pure nickel (b) coatings in 0.5M Na2SO4 solution, at different rotation speed and

constant normal load (2kg), layer (A). On Layer (B) zoom of final part of layer (A).

Layer (C) the rotation speed of samples corresponding to friction coefficient versus time.

The differences in friction coefficient values for the two types of coatings are bigger for slower rotation speed. This can be explained by the different surface morphology of two types of coatings [10] and suggest that in the case of nickel coating the adhesive wear plays an important role in the complex

mechanism of wear-corrosion. On the layer (B) of Fig. 6 it is the zoom of final part of layer (A) while the layer (C) represents the rotation speed of samples corresponding to friction coefficient versus time. When a tribo-element is made of a ductile element such as Al, Cu, Ni, Fe or an alloy with a combination of them, material in the contact region plastically deforms severely under the combined stresses of compression and shear. Large plastic deformation generally introduces large wear rate since wear surface tends to become rough and protective surface layers are easily destroyed [11]. Scar surface profiles of pure Ni coating showed higher debris and higher roughness parameters. Surface roughness parameters in the middle of the scar (mean value) is presented in Fig. 7 (Ra= 8.19 μm; Rq= 9.67 μm; Rp= 28.50 μm; Rv= 18.70 μm).

Fig. 7. Mean roughness profile in the middle of the wear scar surface and corresponding roughness parameters for pure nickel coating after constant

normal sliding load (20 N) at different rotation speed of the sample (test experiment presented in Fig. 6).

The introduction of a harder reinforcing phase in the ductile matrix by a certain volume fraction can reduce ductility of the matrix material in the contact region and wear of the matrix can be reduced as a result [11]. Scar surface roughness parameters are about three times smaller for nano - structured composite coating than that for pure Ni coating. The mean scar profile of Ni + SiC coating is shown in Fig. 8 (Ra= 3.09 μm; Rq= 3.07 μm; Rp= 11.50 μm; Rv= 10.50 μm).

Fig. 8. Mean roughness profile in the middle of the

wear scar surface and corresponding roughness parameters for for Ni+ SiC nano composite coating after constant normal sliding load (20 N) at different



22

rotation speed of the sample (test experiment presented in Fig. 6).

Both profiles of the scar surface were measured after the test experiment presented in Fig. 6. By increasing the sliding normal load from 10 to 30 N the potential changes to more negative value and the torque force to higher values, Figures 9 and 10.

Fig. 9. Corrosion potential evolution function of

normal load at constant rotation speed of samples of 150 rpm: (a) Nano - structured Ni + SiC coating,

(b) Pure nickel coating.

Fig. 10. Torque force evolution function of normal

load at constant rotation speed of samples of 150 rpm: (a) Nano-structured Ni + SiC coating,

(b) Pure nickel coating. The potential evolution is mainly dependent on the wear parameters and the increase of the applied normal load (at the same rotation speed) induces a clear decrease in the corrosion potential values for pure nickel coating and not so significant for composite coating. The mechanical damage of the sample surfaces induces an activation of the metal structure, which is well described by the corrosion potential. However the trend with time shows a further increase in the corrosion potential towards more noble values. The equilibrium between the mechanical damage and the electrochemical processes at the metal surface needs some time to be reached. At the beginning, the mechanical wear induces a rapid deterioration of the protective surface layers, which causes a drastic decrease of the corrosion potential. The friction coefficient of nano -

structured composite coating is lower than that of pure nickel coating as is shown in Fig. 11. It is also slowly dependent on applied load.

Fig. 11. Friction coefficient as a function of normal sliding load: (dash line) nano - structured composite

coating and (solid line) pure nickel coating at the same rotation speed (150 rpm).

The wear – corrosion diagram was obtained by weighing the samples after different intervals using an analytical balance with a precision of 10-4 g. Wear corrosion rates for both type of coatings are presented in Fig. 12.

Fig. 12. Wear – corrosion rate of nano - structured

SiC -Ni composite and pure nickel deposits in 0.5M Na2SO4 solution at different applied sliding load.

It must be noticed that this figure represent the overall mass loss due to the corrosion and wear not only the mechanical wear. If we compare the better results in corrosion and wear resistance of nano - structured composite coating with their surface morphology, discussed in the paper [10] we can conclude that the surface change induced by nano - silicon carbide particles codeposition in the nickel matrix have a good effect on the coating performance in the tribocorrosion systems.

4. Conclusions The tribocorrosion aspects of nano-structured SiC –nickel composite coatings compared with pure nickel coating were studied. The sliding wear



23

corrosion system was formed from the samples as a cylinder against an alumina parallelepiped counterface in 0.5M Na2SO4 solution. Both measurements of electrochemical corrosion and wear corrosion show a better resistance of nano - structured composite coating compared with pure nickel coating. The general corrosion resistance in hydrodynamic conditions (150 rotation speed) is higher for nano structured composite coating as it resulted from the impedance and the potentiodynamic diagrams performed. The nano - structured composite coating show a higher polarisation resistance and a reduced corrosion current density compared with pure nickel coating. The friction coefficient of nano -structured composite coating is smaller compared with pure nickel coating as a result of nanoparticles embedded and the fine nano - structured coating resulted. The average values of roughness parameters of the scar surface are three times higher in the case of pure nickel coating, after the same conditions of normal sliding load. Also the wear corrosion rate of nano structured composite coating is smaller than that of nickel deposit. The better behaviour of nano-structured composite coating in tribocorrosion systems make it as a suitable candidate in modern techniques to improve the surface properties of materials for special application.

Our work will continue with the modelling of corrosion – wear of nano - structured composite coatings in order to appreciate the possible synergism between corrosion and wear and to explain the differences in behaviour between a pure metal and a composite coating reinforced with nano crystals of dispersed particles.

References [1] N. Guglielmi, J. Electrochem. Soc. 8, 119 (1972) 1009-1012. [2] L. Benea, Composite Electrodeposition -Theory and Practice, Ed: PORTO FRANCO, Romania, ISBN 973 557 490 X, (1998) 173 p. [3] S. W. Watson; J. Electrochem Soc., 140, (1993) 2235. [4] G. Maurin and A. Lavanant; J. Appl. Electrochem., 25, (1995) 1113-1121. [5] L. Benea and G. Carac, Metallurgy and New Materials Researches V No 2, (1997) 1-19. [6] L.Benea, Proceeding volume "Passivity and Its Breakdown" P.M. Natishan, H.S. Isaacs, M. Janik-Czachor, V.A. Macagno, P. Marcus, and M. Seo, Sept. 1997, ISBN 1-56677-179-X. [7] L. Benea, Materials and Manufacturing Processes, Vol. 14, N°. 2, 1999 231-242. [8] L. Orlowskaja, N. Pereine, M. Kurtinaitiene, S. Surviliene; Surface and Coating Technology 111 (1999) 234-239. [9] Sun Kyu Kim, Hong Jae Yoo; Surface and Coating Technology 108-109 (1998) 564-569. [10] Lidia Benea, Pier Luigi Bonora, Alberto Borello, Stefano Martelli, François Wenger , Pierre Ponthiaux, Jacques Galland; Journal of The Electrochemical Society Inc., 148 (7) C461-C465 (2001) [11] Koji Kato, Wear 241, (200) 151-157.



24

TECHNOLOGIES FOR FUTURE PRECISION STRIKE MISSILE SYSTEM

Ion DINESCU, Mihaela OPRESCU

AIR FORCE ACADEMY „HENRI COANDA” BRASOV

E-mail: [email protected]

ABSTRACT

This paper provides an assessment of the state of the art of future aeromechanical technologies of the precision of striking sistems. The aeromechanical technologies are grouped into specific discussion areas of aerodynamics, propulsion and airframe technological materials.

KEYWORDS: missile, aerodynamic, composites materials, propulsion.

1. Introduction

Missile aeromechanical technologies have benefits that include enhanced flight performance, reduced weight, increased Mach’s number

⎟⎟⎠

⎞⎜⎜⎝

⎛=

soundofspeedthespeedflyings'airshipthenumbers'Mach , reduced

costs, higher reliability, and reduced observables. Figure no. 1 summarizes new aeromechanical technologies for precision striking missiles. Most of the technologies in the figure are covered in this paper [7; 8; 9].

Fig. 1 New aeromechanical technologies for precision striking missiles

2. Missile aerodinamical technologies

This assessment of missile aerodynamical

technologies addresses six new enabling technologies. These are aerodynamical configuration shaping, lattice tail control, split canard control, forward swept surfaces, bank to turn maneuvering, and flight trajectory shaping [3; 4; 6; 7].

An advantage of a tailored lifting body missile is a higher aerodynamical efficiency (lift-to-drag- ratio) for extended range cruise performance and

enhanced maneuverability. Tailored missiles are also synergistic with ramjets for areas such as inlet integration and liquid hydrocarbon fuel packaging.

Disadvantages of tailores missiles include their relative inefficiency for solid subsystems packaging and an adverse impact on launching platform integration, due to a larger span.

Improved methods and tests are required for the reduction of the aerodynamics and the structural loads of non-axisymmetric weapons.

This includes more extensive wind tunnel tests, computational fluid dynamical predictions, and finite element modeling of structural integrity.

Lattice tail control Aerothermal Insulation Lattice tail control is an example of new

aeromechanical technology. Lattice fins have advantages of lower hinge moment and higher control effectiveness at supersonic Mach’s number. Figure no. 2 shows a comparison between a lattice tail control of two conventional approaches and a tail control, all movable and flap control. Except for a radar crossing section, lattice tail has a very good control of superior performance for supersonic missiles [9]. Also shown in the figure, there are exemples of supersonic missiles with tail control alternatives of lattice tail control (a), all movable tail control (b), and flap tail control (c). The smaller chord length of the lattice has less variation in the center of pressure, resulting in lowerhinge moment for lattice tail control.

Airframe

Propulsion

Lattice fins are the most appropriate for either subsonic or high supersonic missiles.

At subsonical Mach’s number, the drag of lattice fins is comparable to that of traditional flight control.



25

At transonical Mach’s number, lattice fins have higher drag and lower control effectiveness than traditional flight control.

The lattice remains choked until the supersonical Mach’s number is high enough to allow the lattice to swallow the shock. An oblique shock is then formed on the leading edge of each surface of the lattice.

At low supersonical Mach’s number, the oblique shock angle is large. Each oblique shock is reflected downstream, at an adjacent lattice surface, resulting in increased drag.

At higher Mach’s numbers the oblique shock is smaller, passing through the lattice without intersecting a lattice surface.

In conclusion, lattice fins have their best application at low subsonical and high supersonical Mach’s numbers, where they have low drag and high control effectiveness.

a

b

c

Fig. 2 Lattice tail control

3. Missile propulsion technologies

The assessment of missile propulsion technologies addresses eight enabling technologies. These are supersonical air breathing propulsion, high temperature combustors, low drag ramjet inlets, ramjet inlet/airframe integration, hingher density fuels, rocket motor thrust magnitude control, high thrust motor, and reaction jet control [3; 7; 9].

Ramjets, scramjets and ducted rockets have high pay offs concerning the precision of striking missiles, operating at supersonical/hypersonical Mach’s number.

Turbojet and turbofan propulsion are a relatively mature technology for the precision striking missiles.

Turbojets/turbofans are most suited for subsonical cruise missiles, providing high efficiency to deliver a warhead at long range against non time critical targets. The operating regime is made with Mach 3.

However, beyond Mach 2, increasingly complex inlet systems are required to match delivered inlet airflow to the compressor’s capacity and

extensive cooling is required to avoid exceeding the limit temperature of the material at the turbine’s inlet.

Solid rockets are capable of providing thrust across the entire Mach’s number range.

Althought the specific impulse of tactical rockets is relatively low, at the order of 250 seconds, rockets have an advantage of much higher acceleration capability than air-breathing propulsion.

Solid rocket boosters are used to boost ramjets to their take-over Mach’s number at about 2,5 for transition to air-breathing propulsion.

The maximum specific impulse of liquid hydrocarbon fuel ramjet is about 1500 seconds, much higher than the specific impulse of a solid rocket.

An efficient cruise condition for a ramjet is about Mach 4 at 80 K feet altitude.

Over Mach 5, the combusting material reaches the maximum limit temperature, the achievable exit velocity and thrust. Also, the deceleration of the inlet airflow to subsonic velocity results in chemical dissociation of the air, which absorbs heat and negates a portion of the energy output by the combustor.

Liquid fuel ramjets are synergistical with noncircular, lifting body airframes because ramjet fuel can be stored in noncircular tanks. Liquid fuel ramjets can be throttled, for efficient matching of the fuel with the inlet airflow.

Throttling provides higher thrust and specific impulse over a broader flight envelope of Mach’s number and altitude. A rocket booster is required to boost the ramjet up to a speed where the ramjet thrust is greater than the drag of the missile. The ramjet’s taking over speed is about Mach 2,5.

Over Mach 6, a supersonical combustion ramjet (scramjet) provides higher performance than a ramjet.

The minimum sustained flight Mach’s number of a scranjet, based on providing sufficient thrust to overcome missile draging, is greater than Mach 4.

The maximum Mach’s number, is based on an engine where the material’s temperature limit, is reached at about Mach 8 to 9.

An efficient cruise condition for a scramjet is about Mach 6 at 100 K feet altitude.

A key of technical challenges is fuel mixing for efficient supersonical combustion.

An enabling technology to enhance supersonical combustion are endothermical fuels. Endothermical fuels decompose, at high temperature, into lighter weight molecular products that burn more readily, providing higher specific impulse and permiting shoter combostor lenght. An endothermical fuel also acts as a heat sink, cooling the adjacent structure. Like the ramjet, the scramjet is rocket boosted to a supersonic takeover speed.

Takeover speed of a scramjet is higher than a ramjet, about Mach 4,5 requiring a larger booster. For a weight limited system, a hypersonical scramjet



26

missile will have less available fuel than a supersonical ramjet missile.

Low Drag Ramjet Inlets Examples of drag inlet alternatives for ramjets

are shown in Figure no. 3. Current operational ramjets have either a nose inlet in different countries (a) or aft axisymmetric inlets (b). A nose inlet has an advantage of lower drag, while aft axisymmetrical inlets have advantages of lighter weight, lower volume, and they do not shroud/degrade warhead effectiveness.

a

b

Fig. 3 Current ramjet inlets are either nose inlet or aft axisymmetric inlets

Because ramjet combustion is subsonic, there

must be a normal shock in the inlet to provide subsonical flow into the combustor. Small oblique shocks prior to the normal shock alleviate the problem of total (stagnation) pressure loss across the normal shock.

This stagnation pressure recovery is much higher than that of an oblique shock prior to the normal shock or for the case of a single normal shock.

Ramjet inlet/airframe integration through forebody compresion (such as a chin inlet) and optimized inlet cowl lip angle provides a higher specific impulse and a higher thrust.

4. Missile airframe materials technologies

The assessment of missile airframe materials technologies addresses five new enabling technologies. These are hypersonical structure materials, composite structure maretials, hypersonical insulation materials, multi-spectral domes, and reduced parts count structure [1; 2; 5; 6].

Composite materials are a new technology that will find increased use in new missile airframe structure. High temperature composites have particular benefits for hypersonic missiles, providing weight reduction. Titanium alloy technology also enables lighter wight missiles in a hypersonic, high temperature flight environment.

At subsonical and low supersonical Mach’s number, graphite epoxy and aluminium or aluminium alloys are attractive choices for lighter weight structure. Graphite epoxy and aluminium alloys have hight strength to weight ratio, are easily fabricated, have a good corrosion resistance, and are low in cost.

For higher Mach’s numbers, graphite polyimide composite structure has an advantage of high structure efficiency at higher temperature for short duration flight Mach’s numbers to about Mach 4.

For flight at about Mach 4,5, without external insulation, the titanium structure and its alloys are prefered. A disadvantage of a titanium structure is higher material and machining cost. However, the cost to cast a part made of titanium is comparable to the cost to cast an aluminium part.

At Mach 5, although it is heavier, a steel structure would probably be used.

Up to Mach 5,7 without external insulation, at about 2000 degrees Fahrenheit, super nickel alloys such as Inconel, rene Hastelloy and haynes must be used. Above Mach 5,7 the super alloys require either external insulations or active cooling [9].

The Mach’s number and temperature application relationships are somehow dependent upon the temperature recovery factor.

At a stagnation region, such as the nose or leading edges, the recovery factor is about 1, resulting in the highest stagnation temperature. A turbulent or laminar boundary layer downstream of the nose or leading edge will have temperature recovery factors of about 0,9 and 0,8 respectively, with local temperatures less than stagnation.

5.Composite Structure Materials The strength to weight capability of advanced

composites is very high. For example, as shown in Figure no. 4, the unidirectional tensile strength of a small diameter graphite (carbon) fiber is more than 400000 pounds per square inch. In addition to small diameter fibers, advanced composite structures have long, continuous fibers and a fiber/matrix ratio that is greater than 50% fibers by volume [9].

Fibers can be: carbon (graphite), kevlar, boron, ceramic, silicon carbide quartz, glass,polyethylene and others.

As an example of strength at the structure level, 50% of the volume of the graphite composite structure can have a strength in a tailored laminate which is above 200000 pounds per square inch, much greater than that of aluminium or even steel [9].

Also the low density of composites further reduces the weight compared to metals.

Graphite fiber composite materials have extermely high modulus of elasticity resulting in low strain and deflection compared to metals.



27

Fig. 4 Composites materials have high strength

However, a note of caution, unlike metals that generally yield gracefully before ultimate failure, composite fibers generally fail suddenly without yield.

For short duration temperatures up to 400 Fahrenheit degrees, graphite epoxy is a good candidate material based on its characteristics of high strength and low density. Graphite polymide can be used at even higher temperatures, up to 1100 Fahrenheit degrees, short duration temperature [3; 4; 9]. Over 1100 Fahrenheit degrees titanium and steel are the best materials based on strength to weight ratio. An area of enabling capability hypersonical precision of striking missiles is short duration insulation technology. Because hypersonical precision of striking missiles have stringent volume and weight constraints, higher density, external airframe and internal insulation, materials are in development. Higher density insulation materials permit more fuel resulting in longer range.

Thermal insulators are used to provide short duration protection for structural materials from either the aerodynamical heating of a hypersonical free stream or from propulsion heating of the combustion chamber and exhaust gases of the nozzle.

Ceramic refractory materials and graphite materials are also candidate insulators for high speed airframes, engines and motor cases.

Although ceramic refractory materials and graphites have high temperature capability, the insulation efficiency for a given weight of a material is not as good as that of plastic composite materials.

At high temperature, the resin melts providing cooling for the structure. Example of bulk ceramics are zirconium ceramic and hafnium ceramic. Bulk ceramics are capable of withstanding height temperatures but like porous ceramics they have relatively poor insulation efficiency.

Finally, graphite insulators provide the highest temperature capability. However, graphites have relatively poor insulation efficiency.

6.Conclusion

The advances in informational technology open new possibilities of new aeromechanical technologies for the striking precision of the future missile sistems. The next millenium will be the age of intelligent missiles and the weapons will begin to suit the capability of modern warfigter.

Note that composite materials are good candidates for lighter weight insulation. For a higher speed precission of striking missiles, medium density plastic composites should be used, such as fiberglass reinforced phenolic resins containing hylon, silica, graphite or carbon.

These have good resistance to erosion, allow high surface temperatures and exhibit good insulation performance.

Medium density plastic composite materials char at high temperature but generally maintain their thickness and aerodynamical shape. They are usually fabricated by wrapping fiberglass tape over a metal form mandrel, so that the grain of the finished unit is oriented for minimum erosion.

Cross flow orientation or other grain directional orientation is optimized to minimize the amount of the material that is required.

After winding, the tape is cured, machined as necessary and assembled with other components using adhesives and sealants.

Airframe structure insulation trades include hot structure internal insulation versus external insulation, “cold” structure versus one piece of self insulating composite structure.

A consideration for a volume limited missile is the total thickness of the airframe insulation.

Large thickness means less volume for fuel resulting in less range.

References [1]. DINESCU, I. Tehnologia materialelor, Tehnologii de bază, Editura Lux Libris, Braşov, 1997. [2]. DINESCU, I. Tehnologia materialelor, Procedee tehnologice, Editura Academiei Aviaţiei şi Apărării Antiaeriene “HENRI COANDĂ”, Braşov, 1999. [3]. DINESCU, I. Tehnologia materialelor, Materiale tehnologice, Editura Academiei Aviaţiei şi Apărării Antiaeriene “HENRI COANDĂ”, Braşov, 2000. [4]. DINESCU, I., ş.a. Studiu privind utilizarea materialelor compozite în construcţiile aerospaţiale, Buletinul ştiinţific al Sesiunii naţionale de comunicări ştiinţifice 1-2 noiembrie 2002, Academia Forţelor Aeriene “HENRI COANDĂ”, Braşov, pag. 41-50. [5]. EFTIMIE, L., DINESCU, I. ş.a. Tehnologia materialelor, Tehnologii secundare, Editura Lux Libris, Braşov, 1998. [6]. ISPAS, ŞT. Materiale compozite, Editura Tehnică, Bucureşti, 1987. [7]. ISPAS, ŞT., LAZĂR, I. Motorul turboreactor, Editura Tehnică, Bucureşti, 1991. [8]. MARINESCU, AL., ANGHEL, V. Aerodinamica si dinamica elicopterului, Editura Academiei Române, Bucureşti, 1992. [9]. *** Research and technology organisation, June, 2001.



INFLUENCE OF YTTRIUM ADDITION ON THE OXIDATION BEHAVIORS OF HYPEREUTECTOID TI-Cr ALLOYS WITH LAVES

PHASE TICr2

Pingan XIAO1, Xuanhui QU2

1 Automobile and Electromechanical Engineering Department of Changsha Communications University,

Changsha, P. R. China; 2 Material Science and Engineering School of Beijing University of Science and Technology, Beijing, P. R. China

Tel.: 86-731-231,8266, email: [email protected]

ABSTRACT

The influence of yttrium on the oxidation behaviors of Ti-Cr alloys with Laves phase TiCr2 and chromium content of 18wt%, 21wt% and 26wt% respectively has been investigated at 6500C and 7000C in air. It is found that tiny addition of yttrium in the alloys with chromium content lower than 26wt% can fine the oxide particles in the scale and promote the formation of chromium and multiple titanium-chromium oxides, such as Cr2TiO5 and Cr0.284Ti0.714O1.857, in the internal layer, which effectively blocks the diffusion of oxygen atoms inward to the matrix and therefore improves the alloys’ oxidation resistance. For Ti-18wt%Cr, the more the yttrium addition in the test range, the lower the alloy’s oxidation rate and both an excellent oxidation resistance and a comprehensively high performance are potentially obtained at the same time. For Ti-21wt%Cr, there always is an optimal yttrium addition, but for Ti-26wt%Cr, yttrium addition put rarely good influence on its oxidation resistance.

1. Introduction

In paper [1] the authors reported the oxidation behaviors of hypereutectoid Ti-Cr alloys with Laves phase in the temperature range of 6500C to 7800C and it was found that when the chromium content in the alloys was 26wt% or more, their scaling rate became 1~2 times lower than that of the alloys with less than 26wt%Cr under the same oxidation conditions, which was mainly contributed to the formation of both chromium oxides and multiple titanium-chromium oxides in the internal oxidation layer that all could effectively block the diffusion of oxygen atoms inward to the matrix. Hence, hypereutectoid Ti-Cr alloys with excellent high temperature oxidation resistance can be fabricated by suitable choice of their chromium content. But, on the other hand, the higher chromium content in a hypereutectoid Ti-Cr alloy, the more Laves phase TiCr2 there is in it, which will deteriorate the alloy’s toughness because of the intrinsic brittleness of intermetallic compound TiCr2. It is well known that tiny addition of reactive rare earth elements is quite helpful in improving the oxidation resistance of high temperature applied alloys and literatures [2] and [3] reported that yttrium addition in alloys with chromium clearly improved the compactness of the Cr2O3-containing scale and

effectively lowered their scaling rate. So in this paper the influence of yttrium addition on the oxidation behaviors of hypereutectoid Ti-Cr alloys with Laves phase TiCr2 was evaluated in order to manufacture Ti-Cr alloys with excellent high temperature oxidation resistance but lower chromium content, which could better guarantee the alloys with a comprehensively high performance, by means of optimized yttrium addition.

2. Experiment procedures

The composition of the tested Ti-Cr alloys with tiny yttrium addition was listed in table 1. The fabrication process of the alloys was almost the same as that reported in an earlier communication [4]. The specimens were cut from the as-annealed buttons to the size of 10×10×10mm by electric discharge machining and then ground with 600 grit Al2O3 abrasive papers before oxidation test. Isothermal oxidation experiments in air were performed in a well resistance furnace at 6500C and 7000C, respectively. The specimens were firstly heated to the tested temperature at 100C/min and then soaked. A specimen in every different composition was withdrawn from the furnace and cooled down in air to room temperature after the following desired

28



exposure times: 25 hours, 50 hours, 77 hours and 102 hours, respectively. Both the weight and the surface area of every specimen were measured before the tests and the oxidized ones were weighed again to evaluate their mass gain from oxidation after the tests. The surface and cross-section micrograph observation and element linear analysis of the scales on oxidized specimens were performed on a JSM-5600LV electron scanning microscope [ESM] with an x-ray energy dispersive system [EDS]. To inspect the phase constitution in the thickness direction of the scales a layer of layer grinding plus x-ray diffraction method was introduced, in which the scale of an as-oxidized alloy was initially x-ray diffracted and then was ground layer by layer plus X-ray diffraction inspection to the left scale after each grinding operation. X-ray diffraction analysis was conducted with a 3014Z x-ray diffraction instrument.

Table 1. The tested alloys and their yttrium content

Alloy

Ti-26wt%Cr

Ti-21wt%Cr Ti-

18wt%Cr

a b c d a b c d a b Yttrium content, ×102wt

% 1.252.503.755,001.252.503.755.00 2.0 4.0

29

3. Experiment results and discussion

3.1 The influence of yttrium on oxidation resistance

of the alloys

The oxidation mass-gain of the tested Ti-Cr alloys as a function of their yttrium content at both 6500C and 7000C for 102 hours are plotted in figure 1 (a) and (b), respectively, and for comparison the mass-gain data of Ti-Cr alloys with the same chromium content but no yttrium from literature [1] are also used. The curves in figure1 reveal that among the three different chromium content Ti-Cr alloys, tiny yttrium addition put the most remarkable positive effect on the one with the lowest chromium content, Ti-18wt%Cr; The higher the yttrium addition in the test range, the lower the oxidation mass-gain and especially at 6500C for Ti-18wt%Cr excellent oxidation resistance, even as good as that of Ti-26wt%, can potentially be obtained by an further increased yttrium addition. For Ti-21wt%Cr yttrium improves its oxidation resistance, but there is always an optimized yttrium addition in the test range and it is of 0.015wt% and 0.025wt% accordingly at 6500C and 7000C. However for Ti-26wt%Cr yttrium addition shows no positive influence on its oxidation property but a little deteriorates it in the test range.

The useful effect of tiny addition of such rare earth elements as yttrium in improving the high temperature oxidation resistance of alloys, especially

to those chromium- containing ones, has be confirmed, but so far on their action mechanics no commonly accepted agreement has be reached because of their extremely high activity and very low content.

Nevertheless, many hypotheses have been proposed and according to them the effectiveness of rare elements can contribute to (1) changing the alloy’s oxidation mechanics, (2) improving the structure at the interface between oxide scale and metal matrix and hence increasing the scale’s adherence to the matrix, (3) strengthening the interfacial chemical bond between the scale and the metal matrix. According to literature [1] Ti-Cr alloys with 26wt% or more chromium hold an excellent oxidation resistance, in this situation it is likely that tiny yttrium addition into them only plays a role as impurity, which explains why yttrium addition in Ti-26wt%Cr shows no good effect in figure 1.

Ti-26wt% Cr Ti-21wt% Cr

mas

s gai

n, m

g/cm

2

Y, wt% (b)

0.00 0.01 0.02 0.03 0.04 0.051.01.52.02.53.03.54.04.55.05.56.0

Ti-18wt% Cr

Ti-26wt% Cr Ti-21wt% Cr

mas

s gai

n, m

g/cm

2

Y, wt% (a)

0.00 0.01 0.02 0.03 0.04 0.050.60.81.01.21.41.61.82.02.22.4

Ti-18wt% Cr

Fig 1. Oxidation mass-gain as a function of yttrium addition in Ti-Cr alloys

at (a) 6500C and (b) 7000C for 102 hours.



3.2 The isothermal oxidation kinetics curves of the tested alloys

The isothermal oxidation kinetics curves of the

tested alloys at both 6500C and 7000C for 102 hous are plotted in figure 2 and the mass-gain data of Ti-Cr alloys with the same chromium content but no yttrium from literature [1] are also used for comparison purpose. After analysis of the results in figure 2, it is found that all the oxidation kinetics curves of the tested alloys follow a parabolic law.

30

According to the function of mass-gain rate

versus time in figure 2 the whole oxidation process could be divided into two phases: the initial rapid mass-gain period and the late slow mass-gain one. For Ti-18wt%Cr it can be seen that its first period is significantly reduced by near a half because of yttrium addition, which not only mainly contributes to the evident improvement of the alloy’s oxidation resistance, but also implies that some changes in oxidation mechanics, phase constitution and microstructures of the alloy have probably taken place in the scale simultaneously.

3.3 The micrographs and phases of the scales The micrographs of oxide particles on the surface of the scales of Ti-18wt%Cr -0.04wt%Y and Ti-26wt%Cr-0.0125wt%Y after oxidation at 650°C for 102 hours are shown in figure 3. From figure 3 and compared with the oxide particles on the surface of Ti-Cr alloys with the same chromium content but no yttrium under the same oxidation conditions in literature [5], it is found that the mean size of TiO2 particles is obviously reduced from 4μm to 2.5μm because of yttrium addition into Ti-18wt%Cr, which

0 20 40 60 80 1000.0

0.5

1.0

1.5

2.0

2.5 Ti-26wt%Cr-0.0125wt%Y Ti-26wt%Cr-0.0250wt%Y Ti-26wt%Cr-0.0375wt%Y Ti-26wt%Cr-0.0500wt%Y Ti-26wt%Cr

m

ass

gain

£¬m

g/cm

2

time, h(f)

Fig. 2. Mass-gain versus time kinetic curves of Ti-Cr alloys oxidized in air at 650°C for (a), (c)

and (e), and 700°C for (b), (d) and (f).

0 20 40 60 80 1000.0

0.5

1.0

1.5

2.0

2.5

3.0 Ti-21wt%Cr-0.0125wt%Y Ti-21wt%Cr-0.0250wt%Y Ti-21wt%Cr-0.0375wt%Y Ti-21wt%Cr-0.0500wt%Y Ti-21wt% Cr

m

ass

gain

£¬m

g/cm

2

time, h(c)

0 20 40 60 80 1000

1

2

3

4

5

6 Ti-18wt%Cr-0.020wt%Y Ti-18wt%Cr-0.040wt%Y Ti-18wt%Cr

m

ass

gain

£¬m

g/cm

2

time, h(b)

0 20 40 60 80 1000.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4 Ti-26wt%Cr-0.0125wt%Y Ti-26wt%Cr-0.0250wt%Y Ti-26wt%Cr-0.0375wt%Y Ti-26wt%Cr-0.0500wt%Y Ti-26wt%Cr

m

ass

gain

£¬m

g/cm

2

time, h(e)

0 20 40 60 80 1000

1

2

3

4

5

6

7 Ti-21wt%Cr-0.0125wt%Y Ti-21wt%Cr-0.0250wt%Y Ti-21wt%Cr-0.0375wt%Y Ti-21wt%Cr-0.0500wt%Y Ti-21wt%Cr

m

ass

gain

£¬m

g/cm

2

time, h(d)

0 20 40 60 80 1000.0

0.5

1.0

1.5

2.0

2.5 Ti-18wt%Cr-0.020wt%Y Ti-18wt%Cr-0.040wt%Y Ti-18wt%Cr

m

ass

gain

£¬m

g/cm

2

time, h(a)



confirms that the nucleation and growth conditions of the oxide particles in the scale has been changed accordingly, but for the yttrium-containing Ti-26wt%Cr alloy rare change can be observed, which once again confirms the weak role of yttrium in Ti-Cr alloys with 26wt% or more chromium content.

31

The cross section of the scale on Ti-18wt%Cr-

0.04wt%Y oxidized at 650°C for 102h exposure was examined by SEM and EDS, and its micrographs and analysis results are shown in figure 4. Compared with those of the scale of Ti-18wt%Cr in literature [1], the microstructures in figure 4 showed little difference, but there do be some difference between their element linear analysis results. From the EDS analysis results in figure 4 it can be found that there is a narrow interleaf with high chromium and oxygen content and low titanium content in the internal layer, more exactly near the interface between the external and internal layers, of the scale, which means that the phase constitution in the scale has be changed, which is vital for the oxidation behaviors of Ti-Cr alloys [4].

After a layer of layer grinding plus x-ray

diffraction inspection, the phase constitution in both internal and external sections of the scale of Ti-18wt%Cr-0.04wt%Y after being oxidized at 650°C for 102 hours is summarized in table 2 and the same sort of results of Ti-18wt%Cr from literature [4] is also listed for comparison. The results in table 2 well designate that it is because of the tiny yttrium addition that the main phases in the internal section has been changed from titanium oxides to multiple titanium-chromium oxides, which also shows that tiny yttrium addition is in favor of the formation of

(a)

(a)

(b)

(b)

Fig. 3. SEM micrographs of oxide particles on the surface of (a) Ti-18wt%Cr-

0.04wt%Y and (b) Ti-26wt%Cr-0.0125wt%Y after oxidation at 650°C for 102h.

(c)

Fig. 4. SEM micrographs and element linear analyses of the cross section of the scale of Ti-18wt%Cr -0.04wt%Y at 650°C for 102 hours

exposure, (a) Ti-Kα, (b) Cr-Kα and (c) O-Kα.



32

chromium oxides in Ti-18wt%Cr. According to literature [4] a high volume fraction of chromium or multiple titanium-chromium oxides and a multiple oxide structure composed of them in the scale can effectively slower the diffusion of oxygen atoms inward to the matrix. Therefore, the more yttrium is added in the tested range, the more chromium oxides are formed in the scale and the better the oxidation resistance of Ti-18wt%Cr.

Table 2 The phase constitution in the scales after oxidation at 650°C for 102 hours

The position in the scale Alloy

External section

Internal section

Ti-18wt%Cr-0.04wt%Y TiO2

Cr2TiO5 and Cr0.286Ti0.714O1.857 as

majority plus some TiO2Ti-18wt%Cr TiO2 TiO2 and TiO as majority

plus some CrO2

4. Conclusions

Tiny addition of yttrium can improve the oxidation resistance of hypereutectoid Ti-Cr alloys with less than 26wt% chromium at 6500C to 7000C. The lower the chromium content in the alloys, such as Ti-18wt%Cr, and the higher yttrium addition in the test range, the better the improvement and hence hypereutectoid Ti-Cr alloys with Laves phase TiCr2, an excellent oxidation resistance and a comprehensively high performance is potentially fabricated; For hypereutectoid Ti-Cr alloys with a

little higher chromium content such as Ti-21wt%Cr, there is always an optimized yttrium addition.

Yttrium addition shows no positive influence on the oxidation resistance of hypereutectoid Ti-Cr alloys with 26wt% or more chromium, but a little deteriorates the property in the test range.

The oxidation resistance improvement of hypereutectoid Ti-Cr alloys with Laves phase TiCr2 for yttrium addition mainly contributes to fining of oxide particles and the formation of chromium and multiple titanium-chromium oxides in the internal section of the scale.

5. Acknowledgement

The support from National Natural science Foundation of China under the contract 59871064 is gratefully acknowledged.

References

[1]Xiao Pingan, Qu Xuanhui, etc. 2002, High temperature oxidation behaviors of Ti-Cr alloys with Laves Phase TiCr2. Trans Nonferrous Met Soc China, , 12(2): 200~2003 [2]Zheng X G and Young D J. 1998, Influence of yttrium on the high temperature corrosion of chromium and Fe-28Cr in CO-CO2-N2(-SO2) atmospheres., Corrosion Science, 40(4/5):: 741~756 [3]Qi Huibin, Yuan Haisun, He Yedong, et al. 1996, Influence of yttrium alloying and multiple Y2O3-Al2O3 coatings on the high temperature oxidation of M38 alloy (in Chinese)., Acta Metallurgica Sinica,, 32(4): 397~403 [4]Xiao Pingan, Qu Xuanhui, et al. 2002, The fabrication of a hypereutectoid Ti-Cr alloy with laves phase TiCr2 (in Chinese). Trans. Nonferrous Met. Soc. China, 12(2):: 236~240. [5]Xiao Pingan, Qu Xuanhui. Investigation of high temperature oxidation mechanics of Ti-Cr alloys with Laves phase TiCr2. Journal of Advanced Materials, in press.



33

SOME CONSIDERATION REGARDING THIXOFORMING OF METAL ALLOYS

Adriana NEAG, Traian CANTA

Department of Materials Processing, Technical University of Cluj-Napoca,

E-mail: adi_n_2000@ yahoo.com;

ABSTRACT

The aim of this paper is to present some general problems about THIXOFORMING-a semi-solid process- and a few methods to obtain the thixotropic structure. The effective use of the semi-solid process presumes good knowledge of the material behavior. Even the process implicate a series of economical investment, the THIXOFORMING technology proposes itself as one of the best to produce components with the mechanical properties required by their use, with costs industrially acceptable.

KEYWORDS: Thixofo rg ing , th ixo t ropy , semi - so l id me ta l a l loy , mic r os t ruc tu re

1.Introduction

Forming process of materials with a liquid phase present is of increasing interest for scientist.

Thixoforging is the forging of metals in the temperature range between their solid and liquid

states. In this range, there exist both, fluid and solid components in the structure (Fig.1). In the literature these processes are referred as forming in mushy state or forming in semi-liquid state or thixoforming.

Fig.1 Phase diagram with temperature range of AlSi7Mg

Fig. 1 shows the temperature range for the cast alloy AlSi7Mg (A356) die-casting at temperatures above 660ºC and thixoforging at temperatures in the range of 577ºC to 586ºC.

A fundamental research in this field was carried out by Flemings, at Massachusetts Institute of Technology during late the 1960s [1].

The Thixoforging consists of four modules. Starting from the production of the raw material by casting, this material is reheated to semi-solid state in a specially designed thermal pre-processing line. After semi-solid forming in a thixoforging press, the properties of the final parts are adjusted by thermal treatment during post-processing. Normally, the cast



34

billet is forged when 30 to 40% is liquid; in the slurry process, 60 to 70% of the material is liquid.

The semi-solid metal forming process (SSM) are developed like an eficient alternativ of making (manufacturing) parts to forms and dimensions more and more near to the final functional form, as result of the thixotropic behavior, obtained by using special melting processes.

Semi-solid forming can be applied for all alloy types that have a solidifying range in which solid and liquid phase can coexist. A lot of metallic alloys were studied, for example, zinc, aluminum, magnesium, copper, but also iron alloys were tested [1]. The results were significant commercial progresses especially in non-ferrous alloy domain in particular aluminum and magnesium alloy.

As a result of the coexistence liquid + solid in slurry state, the metal has total different properties as in solid state.

The forming stress of the metal in slurry state is very low. The forming stress depends on the liquid amount present in the metal, and by reason of its appearance on the grain edges, the binding forces between grains is very low or even zero. In conclusion, the relative movement is easy between the grain in slurry state. So, forming and flow of metal is produced under low stress.

The solid state content fs defined in percent, is an important parameter, which express the state of slurry material. When fs>80% the slurry is consistent and easy to form and stirring. When fs<60 the slurry flow under gravitation force [3].

The slurry with a reduced solid content, can be agitated and mixed with other materials like: different metal powders, ceramic particles, graphite powder and so we obtain diverse mixtures.

The metal in slurry state can be easy separated because of the low intergrain binding forces. Two slurry state metals can be easy combined, due to liquid diffusion phase of the two metals each in the other and then by their solidification together.

2. Microstructure evolution

THIXOTROPY is a physical state in which a solid

material gets more fluid (modifies its viscosity), when constrained on shearing stress in a specific direction. Viscosity increases at the moment the shearing stress stops. After reaching a critical shearing stress, the material starts flowing.

In certain processes that work with low solid phase content, (semi-liquid state), thixotropic phenomenon is less present then in semi-solid state. This property is very important for semi-solid or semi-liquid metal forming process and it permits to avoid turbulences (which can be source of internal

porosity) during liquid casting state, using classical casting processes. This turbulences stays to the origin of porosity that influence mechanical properties of classical casts or weldability.

To obtain this thixotropic effect, it was developed and patented some methods that allow to obtain the mixture of a solid phase formed by degenerate dendrite with round shape, dispersed in a liquid state (inter-grain eutectic). This sllury mixture is determined by the content of solid phase fs.

In the case of most non ferrous alloy types the solidifying structure remains the same, so that the phase and structure changes through thermal treatment and plastic forming are produced in the environment of primary structure and the effects of this technologic processing are strongly influenced by the primary structure properties.

Referring to the fact that non-ferrous metals have tendency to crystallize during solidification with large grains, the question to influence crystallization processes to obtain fine grains and an advanced rate of dispersion decides the properties of the product. The reduction of the grains dimensions increase mechanical characteristics and ensure their value uniformity, emphasizing the quasiisotropy of the material.

Forming in the semisolid state requires that a metal or alloy have a roughly spherical and fine grain microstructure in the moment when it enters in the forming die.

This microstructure is a necessary condition for revealing all advantages of deformation of alloys in a semi-liquid state.

2.1 Methods to obtain a thixotropic structure

To obtain this nondendritic structure (or the

thixotropic structure), there are used different stirring methods during the casting process, to break the dendritical structure and obtain a poly-crystalline structure with solid rounded grains α and intergranular eutectic.

mechanical, electromagnetical, magnetohidro-dynamical or ultrasonic stirring of alloys during solidification

heating to temperatures above solidus of a stock metal which was previously subjected to cold plastic deformation (SIMA method – strain induced melt activated )

Electromagnetic stirring has been recognized as the most effective way to produce a suitable feed stock material with thixotropic behaviour.

Figure 2 shows two types of solidification microstructures for Al-7Si-0,5Mg(357).

It was noticed that a reheated structure of an alloy obtained by classic casting, without refining grains, it is not poly-crystalline and in conclusion it can not be thixoformed [4].



35

When the shearing intensity is large enough, the particle size of the primary α phase is mainly dependent on the cooling rate during solidification.

As long as the intensity of stirring is increased, there are more grains and the liquid between grains will decrease. Also, the increasing of the casting speed leads to decreasing of the grain size.

Fig.2 Microstructures of aluminum alloy Al-7Si-0,5Mg (357) (a) dendritical conventionally cast and (b) )

nondendritical semi-solid formed

3. Advantages of the process for production cycle and products, related with the traditional die casting and forging are: ♦ less volumetrically contraction during

solidification; ♦ less risk of microshrinkage ♦ less heat cracking; ♦ the solidified microstructure is homogene and

uniforme in the whole part, even different dimensions, resulting a good mechanical properties;

♦ it can be obtain parts in final form, with complex geometry and good dimensional precision, reducing the costs of the parts;

♦ heat treatment possibility for higher mechanical characteristics and welding [2];

♦ the evacuated heat is reduced, related to die casting, for a part with the same geometry, results a decrease by 20-25% of the part manufacturing and increases the life expectancy for the die;

♦ energy economy (40%) to the final user; ♦ an better environment for workers without

excessive heat because the material is reheated and not melted;

♦ avoiding the risk of ignition caused by magnesium in liquid states, if not working in protective gases (expensive, toxic or ozone layer-

threatening protective gases), by decreasing the injecting temperature corresponding to a mushy consistency.

Anyway the main advantage is the high and

controllable viscosity of the semi-solid material, which determine a laminar flow, avoiding turbulent flow, sours of internal porosity.

Table 1 presents the mechanical properties resulted by thixoforming for some aluminum alloys, wich the specific heat treatment used.

Table 2 compares the characteristic of aluminum automobile wheels produced by thixoforming and permanent mold casting [4].

4. Aplicability

The main aplicability fiels are : ♦ parts with high mecanical characteristic; ♦ parts that need high tightness, wich actually

aren’t produced through die casting; ♦ parts of hipereutectic alloys needed to resist to

friction ; ♦ some parts made with die casting, but with high

loss caused by the local porosity, etc.

X200



36

5.Conclusion

In conclusion, it can be said that the THIXOFORMING became an industrial reality and the fields of application of the process seems to become more useful. The remarkable properties of the semi-solid gel (slurry) permit to obtain finite parts with thin wall thickness and without defects.

The THIXOFORMING process should focus on applications which really take the advantages of its specificities and the rapid growth of the use of aluminium in the automobile, military and aerospace industry. The fundamentals of this new technology require further investigation wich should made even by Romanian researchers.

Table 1

Aluminum alloy Temper Ultimate tensile

strength (MPa)

Tensile yield strength (MPa)

A

% HB

AlSi7Mg0,3 (356)

F T5 T6

230 260 300

105 170 230

18 15 14

60 80 95

AlSi7Mg0,6 (357)

F T5 T6

240 275 345

110 205 290

17 10 10

65 90

110

AlSi5CuMg F

T5 T6

270 320 405

130 225 320

7 7 5

80 100 130

AlSi6Cu4 F T6

270 405

130 320

7 5

80 130

AlSi17Cu4Mg

F T5 T6

220 270 350

185 270 350

1 <0,2 <0,2

115 140 165

Table 2

Process Al alloy

Weight direct from die or mold

Kg

Finished part

weight

Kg

Production rate per die

or mold,

Pieces/h

Heat treatment

Ultimate tensile

strength

MPa

Yield strength

MPa

A

% Thixoforming 357 7,5 6,1 90 T5 290 214 10

Permanent mold casting

356 11,1 8,6 12 T6 221 152 8

6.References [1].Ken Young, P.Eisen - SSM Technological alternatives for different applications, Semi-solid processing of alloys and composites, 6th International Conference, Turin, 2000, p.97-102 [2].P.Giordao – New forming techologies for light alloys. A winnig challenge, New developments in forging technology, Werkstoff-informations-gesellschaft, june 2003, pg.149

[3].Manabu Kiuchi – Metal forming in mashy state, Plasticity and modern metal-forming technology, Edited by T.Z.Blazynski, p.289 [4]J.Collot - Thixoforming of Magnesium and Aluminum alloys in semi-solid or semi-lichid, that is the question?, 64th World Foundry Congress, Paris, sept.2000



NITRIDE COATINGS ON WIDIA SUBSTRATE FOR MECHANICAL APPLICATIONS

Stela CONSTANTINESCU*, Viorica MUSAT*,

F.BRAZ FERNANDES**

*“Dunarea de Jos “University of Galati, email:[email protected]

** Faculty of Science and Technology, CENIMAT, New University of Lisbon

ABSTRACT

During chemical vapor deposition (CVD) process the surface properties of the metallic or non-metallic materials are modified by the generation of a new layer with special chemical and physical prosperities. This modification in the surface prosperities is produced by subjecting the substrate materials to a single gas or a combination of gases at elevated temperatures. One type of C.V.D. process is carried out in a heated retort and the chemical reactions that occur are initiated in the space around or on the surface of the substrate.

In this paper, titanium nitride layers on widia substrate have been obtained by an original C.V.D. method, in a heat treatment chamber. Thin layers of 1,5-10 μm using ferrotitanium as raw materials were obtained as a function of exposure time at 1050oC. The microhardness of the coated materials ranges between 25000 and 25520 MPa.. Although widia is well know as a very good material for cutting devices, cutting experiments on these coated widia plates show that the endurance increases by 3 to 5 times as compared with the uncoated plates endurance.

KEYWORDS: CVD, microhardness, TiN thin layer, widia substrate.

1. Introduction

The vapor chemical deposition (C.V.D.) method is a widely spread method of making thin layers. It has been gaining ground lately as opposed to the conventional physical vapor deposition, wet chemical deposition or other special deposition procedures.

The process of the chemical deposition from vaporous implies the adsorption of the mobile atoms (monomers) on the substrate surface, their migration with embryos and stable nucleus formation, followed up by further growth through the adsorption of new atoms on the surface and also by the nucleus coalescence. The final structure of the deposed layer is given by various effects such as, the adsorption of impurities, the incorporation of gaseous, the co-deposition of another elements, the crystallization, etc. The ultimate properties of the coatings are further dependent on the nature and composition on the substrate. Therefore, in theory a vast number of substrate-coating combinations is possible, with its own set of physical and chemical characteristics.

Compared with the other methods, the vapor chemical deposition features the following advantages: highly pure thin layers obtained by a suitable choice of the initial materials and reactions; perfectly crystalline layers due to growth under appropriate equilibrium conditions, relatively high temperatures possibility to cover the samples with thin layers of complex shape, the process is and can be automated and adapted to full scale processing of a large number of sub-layers.

If the vapor chemical deposition takes place within a tubular continuous reactor, a gas carrying the reacting species is passed over the substrate. At the substrate surface, the reacting elements undergo a number of chemical reactions leading to the product formation. Part of the products are deposited on the substrate and part of it goes back to the gas stream [1].

Before examining the vapor chemical deposition reactions it must be determined if the reaction is possible thermodynamically, if the calculated concentrations (partial pressures) of the reactants under equilibrium conditions are less than their initial values.

The calculation of the equilibrium concentrations from the equilibrium constant involves

37



a good choice of the number of gas spaces which can be higher than two and the number of the independent relations A relation implies the equilibrium expression depending on the free standard reaction energy and temperature. The other relation consists in that the system pressure is the sum of the partial pressures. If some reactants possess more than one valence state, the reaction should contain the reactant under its most stable valence state.

After making all the calculations the entire range of deposition parameters (temperature, pressure, gas initial composition) is obtained.

The phases coming from the vapor which contain the diffusion element and the carrying gas pass through three main stages; vapor formation, transportation and deposition. These stages differ in

terms of chronology and make up a whole process. According to its essence, such a process is a chemical transport reaction expressed by the solid substance interaction (A) with the gas or vapors (B). In this case only gas products (C) are obtained. The reaction is the following: A(solid) + B (gas or vap.) = C ( gas or vap.) Upon conveying the gas, two processes are possible:

- isothermal , without forced flow; - non-isothermal , with forced flow.

With the latter process, saturation with gas diffusion by contact-free procedure is reported. The system scheme by this method is illustrated in figure 1.

1 2 3 4 5 6 7

Fig. 1. Gas transport system: feeding saturation and regeneration gas (1),

saturation room/chamber (2), heater (3), working room (4), sample (5), inductor (6) and waste gas (7).

The thermodynamics of the TiN vapor chemical deposition is studied taking into account two basic aspects. One is the thermodynamics of TiCl4 by means of chloride acid and the other one is the chemical reaction to form TiN [2-6].

The calculation of the partial equilibrium pressures of the vapors in the Ti-Cl-H system refers to a 3-component system where the gas phase is in equilibrium with the solid state. According to the Gibbs rule, the system has three degrees of freedom: pressure, temperature and a composition variable Cl/H. This ratio Cl//H remains constant during the entire deposition process because the atoms of Cl and H are not added to or removed from the system during the Ti deposition. Its value is determined by the partial gas pressure in the initial mixture. In the system Ti-Cl-H the following eight species prevail:H2, HCl, TiH4, TiH3Cl, TiH2Cl2, TiHCl3, TiCl4, TiCl2. To calculate the partial pressures of the eight mentioned species, a set of eight independent equations is chosen to express the relations among these species:

Ti (s) + 4HCl (g) ↔ TiCl4 (g) + 2H2 (g) (1)

4HClTi

22H4TiCl)1(p

Pa

PPk

⋅

⋅=

Ti (s) + 3 HCl (g) ↔ TiCl3H (g) + H2 (g) (2)

3HClTi

2H3TiCl)2(pPa

PPk⋅⋅

=

Ti (s) + HC l (g) + H2 (g ) ↔ TiClH3 (g) (3)

2HHClTi

3TiClH)3(p

PPa

Pk

⋅⋅=

Ti (s) + 2HCl (g ) ↔ TiCl2 (g) + H2 (g) (4)

2HClTi

2H2TiCl)4(p

Pa

PPk

⋅

⋅=

Ti (s) + 2H2 (g ) ↔ TiH4 (g) (5)

38


FASCICLE IX METALLURGY AND MATERIALS SCIENCE, ISSN 1453 – 083X VOL II – 2003

2HTi

4TiH)s(p

Pa

Pk

⋅=

aTi = 1.The remaining three equations provide the ration Cl/H and the fact that the total pressure of the system is 1atm.

4PTiCl 4 + 3PTiCl 3 + 2PTiCl 2H 2 + 2PTiCl 2 + PTiClH 3 + PHCl

Cl/H = ----------------------------------------------------------------------------------------- 2PH2 + PTiCl 3H + 2PTiCl 2H 2 + 3PTiClH 3 + PHCl + 4PTiH 4

If, for example, Ti concentration is increased

in a mixture TiH2Cl2 - H2, the partial pressures of all the other species are initially null. After determining the values Kp, the partial equilibrium pressures are computed for the eight species, considering the ration Cl / H values of 10-3, 10-2, 10-1 and 10 within the temperature range 800-1600 K [7].

The fFΔ values as functions of temperature are given by Elingham diagrams (Figure 2).

Fig. 2.Variation of gas species formation free energy. 1 7 2 6 3 4 5

substrate

Fig 3. The steps of the C.V.D. process.

Formation of TiN is the result of some heterogeneous reactions where TiCl4 fed as vapors

into the reaction room reacts with N2 and H2 according to the following equations: TiCl 4 + 1/2 H2 → TiCl 3 + HCl TiCl 3 + N 2 → TiN + 3HCl + 1/2 H 2

TiCl 4 + 1/2 N 2 +2H2 → TiN + 4 HCl (6)

Reaction (6) can be regarded as starting reaction and it is crucial for the overall process velocity. The steps of the vapor chemical deposition process are illustrated in figure 3.

For the reaction to take place, the reactants are transferred from the main gas stream to the surface layer by diffusion or convection, are adsorbed onto the surface and then the reaction involving the absorbed molecules takes place with the generation of the thin layer products. The seven stages of the CVD process take place consecutively [3] .

Any stage slower than the others will determine the overall process velocity, however, under the stationary conditions, all the stages will occur at the same speed as they develop serially. Out of the seven stages, steps 1, 2, 6 and 7 are the substance-conveying stages to which reactants are added in the deposition area while the by- products are removed. Stages 2 and 6 represent the species transfer between the main gas stream and the sub-layer surface. These stages occur through physical processes such as diffusion and convection. The deposition processes whose speed is limited by these stages are determined either by mass transport or by diffusion The absorption stages, the surfaces and desorption reactions (steps 3, 4, 5) represent the chemical reactions that involve the sub-layer surface or take place on it .

The processes whose speed is limited by the

one of these steps are chemically, cinematically or surface determined. Finally, it is possible that the speed of the transfer between the gas stream and surface substrate as well as the surface process take place faster than the deposition surface entering sped (step 1). In this case, the gas stream has a residence time long enough in the vicinity of the substrate to get into equilibrium with this. The process is controlled by the reactant supply speed. Like steps 2 and 6, step 1 is a mass transport process

fFΔ

39



. To distinguish between them, the

equilibrium process is controlled by the mass transport type I while steps 2 and 6 is mass transport type II.

Table 1 Rate determinative step Kinetic regime

• Reactants feeding in the reaction zone.

• Mass transport control-led by reactants feeding

rate ( TYPE I ). • Reactants or products

transfer between gas stream and the substrate surface.

• Mass transport control-led by diffusion or con-vection rate (TYPE II).

• -Reactants adsorption; -Products desorption; -Surface reaction; -Atoms incorporation in

reticular positions.

• Kinetic regime control-led by the rate of the surface reaction.

2. Experimental

To obtain thin TiN layers, plates of type S.N.U.N.15.04.08. K20 and T.P.U.N. 22.04.08. P30 were made.

The thickness of the TiN layer was determined by the microscope and by the Kalotest device; for some of the samples, surface profilmeter method has been used. For the first method, a Neophot microscope was used. After previous calibration, a reading on the micrometer shows 0,287 μm /sample: noa ⋅ x = nob ⋅ 0,01 mm and x = 0,287 μm /division. The use of Kalotest device is based on a housing which cuts the deposited TiN layer. The housing is obtained by rotating a steel ball which slides over the plate surface after this has been previously lubricated with diamond paste of 10μm granulation. After five minutes the steel ball makes, by means of the diamond paste, a sphere shell. The

steel ball diameter is 12 mm,fig5. Since the shell diameter is much less than that of the ball, the layer thickness can be calculated by the values x and y from the relations below:

S = R

yx ⋅ ,

where R represents the steel ball radius . The thickness of the deposited layer

increases with the time of exposure at the working temperature. Microhardness was determined by Neophot microscope. Cutting plates of thin TiN layers of 6,8 and 10 μm depths were also tested.

Microhardness values (Table 2) were calculated using the relation:

HV0,05 = 2d2

136sinF2 ⋅ = 1,854 4 ⋅

2dF ,

where F is expressed in kgf/mm2 and the average trace diagonal, d, is equal with (d1 +d2)/2 and is expressed in mm. The microhardness test is carried out in compliance with STAS 7057 – 78 and with loads of 0.0098 – 9,8 N ( 0.001.. 1,0 Kgf).

3. Results and discussions

Figure 4 shows the XRD pattern of nitride thin layer deposed on widia substrate. The thin layer is not TiN pure phase, diffraction peaks of WN are also present, indicating the chemical interaction between the substrate and the atmosphere during chemical vapor deposition process. High intensity of the diffraction peaks of WC crystalline phase from widia substrate indicates that the TiN coating consists of very thin layer [8]. The values of the TiN layers thickness as measured by the Kalotest (Table 2) device are in good agreement with the values measured by microscopic analysis (Table 4).

WC

WN

TiN

WC

WC

WC

WC

TiTi

2-Theta-Scale30 40 50 60

Fig. 4. XRD pattern of nitride thin layer for 5.5 h exposure time.

40



Fig. 5. The sphere shell of TiN coating.

. Table 2

Sample (cutting plates)

Time ball maintenance

[h]

Steel ball radius, R

[mm]

x [μm]

y [μm]

Layer thickness

[μm] 1. 6 6 98 735 12 2. 5 6 83 580 8 3. 4 6 62 434 5,5

Table 3 Test

number Divisions number

d [μm] F [N]

Microhardness, HV0,05 [ MPa ]

1. 2,78 0,80 0,001 28520 2. 3,97 1,14 0,002 28310 3. 6,30 1,81 0,005 28260 4 8,88 2,56 0,010 28180 5. 14,11 4,05 0,025 28090 6. 20,03 5,75 0,050 28000

Measurements were also carried out on TiN coated plates of various thickness of 6,8, 10μm. Micro-hardness is not a constant, as Vickers hardness, in spite of the prints geometry similarity, but it decreases with higher test load depending on the print size. The microhardness HVo,o5 value of 28000 MPa (about 28000 N/mm2 ) are in good agreement with the literature data [4] and confirms the presence of TiN thin layer,table 3.

The plates are all covered with a continuous nitride layer. This is uniform over the entire depth as also shown by the metallographic analysis (figure 6).

The thickness of the thin layer increases with the exposure time at the working temperature, as illustrated in table 4.

Fig. 6. TiN layer thickness.

10 μn

41



Table 4

Material used as Ti source

TiN layer thickness [ μm ]

Exposure time [ h ]

Temperature [ 0C ]

Ferrotitanium

1,5 2,8 3,5 5 6

7,5 8,3 10

3

3,5 4

4,5 5

5,5 6 7

1050 1050 1050 1050 1050 1050 1050 1050

4. Conclusions

Thin layers of TiN on widia substrate have

been obtained by an original C.V.D. method, in a heat treatment chamber by adding chloride acid vapors passed over the incandescent pure titanium or ferrotitanium.

Lab-scale systems have been designed with the possibility of use at industry scale for small production. The support temperature was established at about 1050o C so that the TiN can provide a suitable deposition of the thin TiN layer. The optimum flow rate of 1 l/h on the plate area of 8500 mm2 was established to ensure a sufficient amount of gas in the reactor to allow for a suitable TiN deposition.

The thickness of the deposit layer increases with the time of exposure to the working temperature. Very good results were obtained for five hours

exposure times, leading to an optimum layer thickness of aprox. 7 μm.

The TiN coated widia plates feature higher endurance capabilities than those uncoated for the same cutting speed. Although widia is well know as a very good material for cutting devices, the cutting experiments on the coated widia plates show that the endurance increases by 3 to 5 times as compared with the uncoated plates endurance [9].

The widia plates coating with thin TiN layers entirely suppresses the inconveniences of a relatively rough topography of the common sinterized coatings while preserving the adequate material mechanical strength. The layer begins loosing its tenacity if its thickness increases considerably exceeding the thickness of 10μm, mainly due to the lower strength characteristics. Loosing tenacity and increasing in the inner tensions results in cracks and breakings in the layers.

8 6 4 2 0 -2

DEKTAR 3 Version 2.13 DROG FILE NAME: GUILAV – MP SCAN ROUTINE #: 1 TIME OF SCAN: 01:45 :20 DATA FILE NAME : .001 Scan ID 0 Scan Length 1500 μm Scan Speed Medium (18 sec) Data Points 750 Resolution 2.000 μm/sample Mean .Range 655 KA Profile Hills&Valleys R.Cursor 176674A 115.22 μm M.Cursor -1900A 493.48μm Vert.Delta -71574A Horiz.Delta 378.26 μm ANALYTIC FUNCTIONS: R: μm M: μm Film thickness = 7,1E+4 Å = 7,1 μm.

Laye

r thi

ckne

ss, μ

m

Fig 7 . Surface profilmeter measurement of the thickness of TiN

film for 5,5 h exposure time at 1050oC,

42



References [1] C. Ciocardia, C. and others – Hard Alloys Sintered from Metallic Carbides, Bucharest, Editura Tehnica (1984), p. 103. [2] Stela Constantinescu – Studies on Thin Carbide and Nitride Layers Deposition on Metal Basis, Based on Chemical Reaction at High Temperatures: Academic Thesis, Galati (1998), p. 84 . [3] Stela Constantinescu , Elena Drugescu - Studies on Mechanism and Kinetics of Phase Transformation in Superficial Layers Using Unconventional Procedures. Research Contract no. 5005, Galati (1995), p. 53. [4] Stela Constantinescu, Elena Drugescu,Tamara Radu - Practical application of AE of the different grades of steel. “ Proceedings of the 25thEuropean Conference on Acoustion Emission Testing “ EWGAE 2002 “ Prague Czech Republic , 11 –13 september, 2002 , p. 135.

[5] Stela Constantinescu - Deposition of titanium carbide layers on widia plates using C.V.D. Method. “A 7th European Conference on Advanced Materials and Processes “ EUROMAT 2001”, 10 -14 june 2001, Rimini, Italy , “ Surface Engineering Processes and Application “ . [6] Stela Constantinescu - Influence of manufacturing process on chemical and structural homogeneity of welded pipe sheets for tanks and vessels working under pressure. “ Proceeding of the International Conference on Advances in Materials and Processing Technologies, september 18 – 21 , 2001, Leganes, Madrid, Spain, p.57 . [7] Delman, B., Introduction à la cinetique heterogège, Cap.V, VI, VII, VIII, IX, Paris, (1989) [8] David D., Caplain, R., Methodes usuelles de caracterisation des surfaces,Eyrolles, (1988), p.298-308. [9] Archer, N.J., Proc. 5th Intl. Conference on C.V.D., The Electrochemical Society, (1991), p.722

43



44

EXPERIMENTAL AND THEORETICAL CONTRIBUTIONS ON HARDNESS PROFILE OF AN

AlCu2Mg1,5Ni ALLOY

Alina-Adriana MINEA, Ovidiu MINEA, Adrian DIMA

Materials Science and Engineering Faculty, Technical University “Gh. Asachi” Iasi, Romania;

e-mail: [email protected]

ABSTRACT

This paper presents experimental and theoretical studies regarding the behavior after heat treatment of an AlCu2Mg1,5Ni aluminum alloy. So, we selected a typical charge for studying the variation of hardness. The studies that we have done can reveal the advantage of controlling the heating process at heat treatment of aluminum alloys. Doing the studies that are described in the paper, we have obtained theoretical equations and nomograms that can express the importance of the process.

KEYWORDS: aluminum, heat treatment, microhardness, experimental model, nomograms

1.Introduction This paper presents a wrought aluminum alloy,

AlCu2Mg1,5Ni that is studied in order to establish hardness profile, after diverse heat treatments.

This alloy is heat treated in order to establish the optimum mechanical properties, and we are referring especially at HV microhardness.

So, we took 10 parts, of identical measures, from this alloy and we apply the final heat treatment. At the end we were studying the mechanical properties, in order to establish the optimum technology. The applied technology is a final treatment, which consists in quenching and artificial aging. Every part had a different technology, keeping the same initial work conditions (furnace, quenching medium etc).

We have made experiments in the same conditions of preheating for the furnace, and we used the same equipment for the hardness determinations.

2.Experimental results

In order to obtain a mathematical model for the final heat treating treatment for AlCu2Mg1,5Ni alloy, we will present the experimental results in table 1. This table is the base for obtaining the regression equation that will describe the process of final heat treatment. In comparison with the test part, it can be observed the opportunity to apply these heat treatments (table 1). Also, it can be observed the increasing of the hardness for the heated parts.

Table 1. Experimental results for hardness of AlCu2Mg1,5Ni alloy

Experiment Code Quenching temperature

Aging temperature Hardness

0. Test part 92,5 1. 211 525 190 156,5 2. 212D 525 200 146,57 3. 222 525 210 154 4. 311 530 190 153,29 5. 221 530 200 162,93 6. 222D 530 210 154,2 7. 312 535 190 149,78 8. 321 535 200 161,86 9. 322 535 210 141,07



45

Table 2.The experimental matrix for k=2 Exp. Quenching tempTc,°C Variation,x1 Aging temp, T`, °C Variation,x21. 525 -1 190 -1 2. 525 -1 200 0 3. 525 -1 210 +1 4. 530 0 190 -1 5. 530 0 200 0 6. 530 0 210 +1 7. 535 +1 190 -1 8. 535 +1 200 0 9. 535 +1 210 +1

In these cases, the regression is a polynom with

m degree and the variable is hardness(y). With the results, using 3k factorial experiment

model and with a specific computer program we

obtained the regression equation that describes the process:

22

212 x647,5x177,5x717,1571,160y −−−=

.

140

145

150

155

160

165

185 190 195 200 205 210 215

temperatura de imbatranire, °C

durit

atea

, HV

Tcalire=530°C

Tcalire=525°CTcalire=535°C

Fig 1. Graphical determination for the aging temperature for AlCu2Mg1,5Ni alloy

130

135

140

145

150

155

160

165

520 525 530 535 540temperatura de calire,°C

duritatea,HV

Ti=190°C

Ti=210°C

Ti=200°C

Fig 2. Graphical determination for the quenching temperature for AlCu2Mg1,5Ni alloy



46

On the basis of these experiments and the

regression obtained we made a theoretical study regarding the establish of the heating parameters for quenching and aging in order to obtain a certain stress, needed in service for this alloy.

Also, we have to mention that the determined temperatures must be in the limits that are recommended in the material standards.

3. Theoretical contributions regarding variation of hardness with the heat treatment, for AlCu2Mg1,5Ni alloy

So, we choused the equation, referring to the

hardness that is a major characteristic for an aluminum alloy.

Replacing the variation parameters with the temperature from table 2, we are obtaining:

2`

2c

`c

0,-056T0,207T

T4163,22T504,219661,60232HV

−−

−++−=

For this theoretical study we consider two cases: - stress and quenching temperature are fixed and we determine aging temperature; - stress and aging temperature are fixed and we determine quenching temperature

Case I: quenching temperature Tc is known and we obtain the variation of the aging temperature T` with the hardness HV, 1. x1= -1, Tc = 525°C

HV7,179,27535,198T` −+=

This equation has a restriction: HV < 155,59.

2. x1= 0, Tc = 530°C HV7,175,28455,198T` −+=

This equation has a restriction: HV < 160,77. 3. x1= 1, Tc = 535°C

HV7,179,27535,198T` −+= This equation has a restriction: HV< 155,59.

The relations are represented in figure 1. And

represents the variation of the aging temperature of AlCu2Mg1,5Ni alloy with stress.

Case II: aging temperature Tt is known and we

obtain the variation of the quenching temperature Tc with the hardness HV, 4. x2= -1, T` = 190°C

HV825,4425,756530Tc ⋅−+= This equation has a restriction: HV < 156,641.

5. x2= 0, T` = 200°C HV825,4405,775530Tc ⋅−+=

This equation has a restriction HV < 160,705. 6. x2= 1, T` = 210°C

HV825,4844,7539530Tc ⋅−+= This equation has a restriction: HV <

153,335.

The relations are represented in figure 2. And represents the variation of the quenching temperature of AlCu2Mg1,5Ni alloy with stress

4. Conclusions

With these experimental results, using the specified factorial experiment model we obtained the regression equation that describes the profile of the hardness variation of treated parts. These equations helps to determine certain mechanical characteristics that are needed in the service of the parts and also to determine exactly the optimum final heat treatment.

Also, studying these diagrams it can be observed a maximum limit of 162,93 HV for AlCu2Mg1,5Ni hardness.

As a conclusion, the paper presents the algorithm for applying the optimum heat treatment in order to obtain the necessary properties for the working parts.

5.References

[1]. Golubev, A.I. , 1966, Rolul compusilor intermetalici, I. D. T. Bucuresti [2]. Mocanu, D.R., 1982, Incercarea materialelor, vol.1, E.T. Bucuresti [3]. Gidea, S., 1965, Aliaje neferoase, E. T. Bucuresti



47

GAS DISTRIBUTION MODELING IN GAS INJECTION OF ALUMINUM MELTS

P. MOLDOVAN, M. BUTU, I. APOSTOLESCU, V. ILIESCU Polytechnic University Bucharest

[email protected], [email protected], [email protected]

ABSTRACT

Is very important to know the gas bubbles distribution in gas injection of the metallic melts due to the fact that this provides a considerable reduction in time and materials. This paper presents a modeling of the active zone from a laboratory device, function of the gas flow and stirrer rate.

KEYWORDS: gas, bubbles, distribution, gas injection

1. Introduction

To obtain materials from aluminum and

aluminum alloys with a high deformability one must eliminate the metallic alkaline and alkaline earth impurities resulted from the electrolysis of alumina in melted salts.

Their removal can be done through several methods, among the most efficient ones being the fluxing and bubbling with gases containing chlorides and fluorides.

Another method combines the effects of the two, namely the inoculation of fluxes in the melt by means of a supporting gas.

Thus, the flux in solid state is transported within the melt where, at the working temperature, it passes into gaseous state and, together with the supporting agent, it bubbles the melt.

This method leads to an increasing probability of interaction between the chlorides (fluorides) and the metallic impurities.

In order to increase the efficiency of this method it is necessary to know the distribution of the gas bubbles inside the pot.

This paper presents the studies made on a lab installation regarding the distribution of the gas bubbles depending on the rotational speed of the stirrer and the discharge of the supporting gas.

2. Experimental results

Refining fluxes are introduced in the melt

inside the stirrer with the aid of a supporting gas. Parameters varying during experimental studies were: the stirrer rotation speed and gas ingoing flow (4 Nm3/h, 4.5 Nm3/h, 5 Nm3/h)

To underline the gas bubbles distribution some experiments were realized using transparent liquids (water and oil) and a recipient of equal shape and dimensions with the laboratory device crucible. The research results are shown in the next pictures. (fig 1 – 3)

a) b)






48

c) d)

Fig. 1. Gas bubbles distribution for water blowing with a flow of 4 Nm3/h, at different rotation ratios: a) 400 rot/min; b) 430 rot/min; c) 445 rot/min;d) 460 rot/min.

a) b)

c) d)

Fig. 2. Gas bubbles distribution for water blowing with a flow of 4,5 Nm3/h, at different rotation ratios: a) 415 rot/min; b) 430 rot/min; c) 445 rot/min; d) 490 rot/min



49

a) b)

c) d)

Fig. 3.Gas bubbles distribution for water blowing with a flow of 5 Nm3/h, at different rotation ratios: a) 400 rot/min; b) 430 rot/min; c) 475 rot/min; d) 490 rot/min

3. Conclusions

Following the physical modelling of the gas

distribution bubbles, the following conclusions were noticed: ⇒ for every rotation speed there is a certain gas

flow which is optimum for an efficient and rapid gas blowing. Thus for 4 Nm3/h gas flow an optimum gas bubbles distribution is obtained for a rotation speed of 460 rot/min, for the gas flow of 4.5 Nm3/h the rotation speed value is 445 rot/min and for the gas flow of 5 Nm3/h the rotation speed was of 430 rot/min. ⇒ the rotation speed increasing does not have a

positive influence on the gas bubbles distribution on the research interval. We can see from the picture shown in the paper that at the maximum rotation speed the distribution zones are reduced, bubbles being placed near the rotor.

References [1] P. MOLDOVAN, N. PANAIT, ŞT. MĂRGINEAN, Bazele tratării topiturilor metalice neferoase, Editura INTACT, Bucureşti, 1998; [2] P. MOLDOVAN, N. PANAIT, M. BUZATU, GABRIELA POPESCU şi alţii, Aluminiul de la materia primă la produse finite, Editura Tehnică, Bucureşti, 2000; [3] E. TORRES-CASTILLO, A FLORES-VALDES, F. A. ACOSTA-GONZALES, J. D. CASTREJON-VALDES, A. H. CASTILLEJOS-ESCOBAR, A kinetic study of the removal of magnesium from molten aluminium alloys by submerged powder injection, Light Metals, 1995; [4] S. T. JOHANSEN, R. ANVAR, B. RASCH, Bubble size and removal rate of sodium in impeller stirred refining reactors, Light Metals, 1999; [5] M. MANIRUZZAMAN, M. MAKHLOUF, Mathematical modeling and computer simulation of the rotating impeller particle flotation process: Part I. Fluid flow; [6] M. MANIRUZZAMAN, M. MAKHLOUF, Mathematical modeling and computer simulation of the rotating impeller particle flotation process: Part II. Particle agglomeration and flotation; [7] A. SILNY, T. A. UTIGARD, Interfacial tension between aluminium, aluminium based alloys and chloride-fluoride melts, Light Metals, 1997.



50

CONSIDERATIONS REGARDING THE WORKING OF THE DOUBLE CHARGING MACHINES FOR SLABS IN THE

CONTINUOUS PUSHER TYPE FURNACE

Marian Bordei, Dragomir Ştefan, Ciurea Aurel “Dunarea de Jos” University,


ABSTRACT

In the metallurgical industrial units from Romania there are thousands tones of outfits and equipment – some of them having a special technico-economical importance. It is necessary to know exactly the loads resulted in exploitation, their reduction by an adequate design, and the variations between the nominal value and the overloads to be registered in determined limits. The continuous pusher type furnaces for the slabs heating are great units, consumer of energetic fluids, having productivity of 185-200 t/h. At SC ISPAT SIDEX Galati, there are five units like these, in LBC and three units in LSF.

KEYWORDS: equipment, continuous pusher type furnaces, double charging machine

1.Design, construction and working elements

Each continuous pusher type furnace is served –

for charging – by one double charging machine, a

complex equipment of great dimensions, which has particularly design and exploitation problems (figure 1) and the main characteristics in table 1.

1

2

3

4 2 3

1

5

6

7

8

10

11

9

Fig.1. The cinematic diagram of a double charging machine, afferent to a continuous pusher type furnace LBC ISPAT Sidex: 1-lever; 2-gear; 3-superior roll; 4-inferior roll; 5-rack; 6-gearing; 7-engine; 8-gear; 9-comando-



51

apparatus; 10-coupling; 11-arrester.

Table 1. The main characteristics of the of a double charging machine

The constructive element UM Values ensemble weight kg 633 pushing forces kN 2x200 maximum working run mm 4300 pushing speed m/s 0,16 retreating speed m/s 0,33

kW 2x260 driving: electric motor rot/min 240/500

The machine has two levers which can be: -separately driven – for the charging of slabs having maximum 4700mm length – on two rows; -simultaneous driving – for the charging of slabs of 6000-12000mm length – on one row. The slabs are displaced in the furnace, on slide bars, cooled with water inside. The variations of the furnace (and slabs) temperature, along the five functional zones, determines variations of the friction coefficient between slabs and slide bars (figure 2).

CICS

SI

SM

SS

furnace

semi-finishedproduct

5 10 15 20 25 30Length of furnace, [m]

Tem

pera

tur e

, [C

]0

Fig.2. The temperature variation both of the furnace and the slabs along a continuous pusher type furnace for the semi-finished products heating: furnace temperature: CS-upper part; CI-bottom part; semi-finishing product

temperature: SS-superior side; SI-inferior side; SM-in the middle The weight of the slabs “carpet” (about

650x104N at LBC and about 850x104N at LSF) requires great pushing forces: 2x200 kN at LBC and 2x400 kN at LSF. For the design and exploitation of both the continuous pusher type furnace and the charging machines, a lot of problems are to be solved:

-the pushing forces and the speeds – according to the semi-finished mass on the furnace hearth and real friction coefficients between these and the slide bars; -the rigidity of the metallic construction of the furnaces taking into account the variable forces which appear during the sliding process of the semi-finished products with discontinuities because of the friction conditions, of the couples, … Taking into account the above – mentioned aspects and the phenomena which appeared during some furnace working – which led to their pulling out

of working – we studied the furnaces and the double charging machines from LBC and LSF – ISPAT Sidex Galati.

2. Aspects regarding the working of the charging machine and the continuous

pusher type furnace ensemble During the pushing of the semi-finished products through the furnace, variable forces result, which act over the furnace, the machine levers and the foundation.

The charging machine – the continuous pusher type furnace – the semi-finished products as an ensemble is an elastic system:

-about 70% of elasticity is due to the slabs lot; -about 12% is due to the slide bars; -about 9% is due to the pushing machine.



52

Because of the elasticity of the system and of the continuous and non-uniform modifications of the friction forces, present between the semi-finished products and the slide bars, the semi-finished lot slides non-uniformly, with shocks. This leads to the appearance of the auto-oscillations of the charge, which produce vibrations of the furnace and its afferent zones ensemble. In certain conditions (great weight, variable speed, great friction and with shocks) the auto-oscillation frequency can approach to the own frequency of the some furnace carcass areas (2-5Hz); these lead to the amplification of the vibrations by resonance and, consequently, to the pulling down of the furnace arch. 3. Aspects regarding the deficiency in the working of the slabs charging machines

Although, in conformity with the project, the lot of the semi-finished products has to begin with the first semi-finished products, placed in front of the furnace door, it sometimes begins from the roller truck. As a result, during the pushing process, the levers with racks have vertically oscillatory

movements, according to the variable resistance during the displacement of the semi-finishing lot, at values of over 20 times bigger, in comparison with the permitted functional ones. The vertical displacements (with a great frequency) are turned into horizontal oscillations due to the gearing teeth and these are transferred to the furnace. The vibrations produced cannot be eliminated by rigidity because ample and uncontrollable forces appear.

In the diagram from figure 3 it is emphasized the defectuous way of using some charging machines from LBC; the lever raises from the charging table at the same time with the slab; the pushing head stamps in the lateral side of the slab and cannot keep the pushing direction; when the lever withdraws, it falls on the A roll, producing strong vibrations. Because of the clearance created by elongation deformation or breaking of the screws, at the next pushing, the oscillations grow, by clearance of the pushing levers in a vertical plane. This leads to the breaking of some screws or bolts of the couplings, or even an inclination of the foundation in front of the furnace door.

y=30-60mm

x

4000mm

1 5

2 5 6 5 8

4 7 9 10

M

Fig.3. Defectuous function diagram of the charging machine lever: 1-lever of the charging machine; 2-pinion for gearing-rack; 3-frame of the roller track; 4-mobile lever of the pushing bar; 5-guidance rolls; 6-roller track, 7-slab; 8-charging table; 9-furnace door; 10-furnace.

4.Veryfying the power of the charging slabs machines driving motor

We calculated the power of the driving motors according to the pushing forces: [N] (1) wbG2totGîF ⋅+μ⋅=

t

310dvîF

dKPη

⋅= [kW] (2)

The engine power can be, also, calculated according to the total resistant coupling, reduced to the engine shaft, to the pushing of the semi-finishing lot in furnace. Taking into account a friction coefficient,

35,0=μ (according to the specialized manuals [1], [2]) the following values of the engine power resulted: -the charging machine LBC: 312 kW comparative with 540 kW - nominal power;



53

-the charging machine LSF: 845 kW comparative with 1040 kW – nominal power.

.

Table 2. Vibration measurements for the nr.1 LBC continuous pusher type furnace

Vibration amplitude

Rotation speed from the engine shaft

The furnace loading No. The mode (position) of measurement

[mm] [rot/min] [kN] 1 Furnace working; the charging machine in

repose 2-3 - -

2 Charging machine displacement (without slabs) - avans 7-8 250* -

3 Pushing slab – from the roller track – up to the contact with the slab – in front of the furnace door

10-12 - -

4 Pushing slab - on the table between the roller track and the furnace – up to the contact with the other slab

8 - -

5 Pushing of slabs lot in the furnace 12-30 200*/415 4000-4500

*The values are for the pushing direction of the levers. Table 3.The absorbed power of the engine to pushing, in charging

Current Tension The absorbed power The registered value

No. of registration

[mm] [A] [mm] [V] [kW]

1 25 416 27 311 129 2 26 434 28 323 140 3 26 434 27 311 135 4 27 450 26 300 125

The average of the four consecutive

pushings, in charging 434 312 132

5.Conclusions -Measurement of vibrations were made both at the continuous pusher type machine with vibrometers; measurements were made by directly reading the vibrations frequency and amplitude in two perpendicular plans: horizontally, respectively perpendicular on the pushing direction, in different points (on the lateral sides of the furnaces and on the pushing machines levels).

-Measurements of tensions and intensity of current and of power absorbed begins were made, thus confirming the conclusions drawn by calculus: a supra-dimensioned of the installed power. References [1]. Oprescu I., Vircolacu I. – Utilaje specifice sectoarelor de prelucrări metalurgice, Ed.Did. şi Pedagogică, Bucureşti, 1981. [2]. Moldovan, V. Maniu A. – Utilaje pentru deformări plastice, Ed.Did. şi Pedagogică, Bucureşti, 1982.

TECHNOLOGY OF OBTAINING COMPOSITE MATERIAL SAMPLES – POLYESTER MATRIX RANFORTE GLASS FIBRE AND FERRITE

PARTICLES

Dumitru DIMA

„Dunarea de Jos” University - Galati, Romania E-mail: [email protected]

ABSTRACT This work refers to composite materials having magnetic properties and to the use of some external magnetic fields in order to improve the quality of ranforte matrix in interface. There has been made a specific elaboration technology of test tubes after various recipes, concentration of the magnetic particles for which there was applied a vibrating magnetic field being varied, and his influence was examined especially through SEM investigations. KEYWORDS: Composite Materials, Polyester Matrix, Glass Fibre, Ferrites.

1. Introduction

There has been investigated the influence of a vibrating magnetic field over a composite material with a matrix very much utilized in industry in order to have a direct use such as: polyester matrix ENDUR G903 (NESTE CHEMICALS POLYESTERS); ranforte from glass fibre (150g/m and 10 Tex) made by VORTEX; magnetic particles Fe3O4 were obtained through a specific technology having high purity and submicronic dimensions.

The elaboration technology of composite material is original and the results interpretation through SEM gives us an eloquent image of the processes that meantime have place.

2. Obtaining the composite material

The first process consists of obtaining a polyester resin additivated with ferrite (Fe3O4) necessary to obtain the sample containing magnetic particles spreaded in the matrix of the composite material.

There were prepared small quantities of additivated resin with magnetic particles (usually 500g) because for concentrations bigger than 2,5%, the stability of the suspension without spreaders is for about 72 hours. That is why the experiment was done in a short time, not to happen the sedimentation phenomenon of magnetic particles because of the "clusters" that appear due to the interaction of the fields of the particles being in suspension.

For a better spread and grinding of the magnetic particles, there was proceeded to a dry grinding and after to a wet grinding, fact pointed out through the electronic microscopy SEM, too. In order to get as more samples as possible - with a view to eliminate that ones that present defects as holes or gaseous inclusions - they used parallelepiped moulds having 1 - 2 mm addition in comparison with the standards cotes for the resistance tests regarding the traction and the scissory.

Addition of the ranforte as bidimensional texture of glass fiber was made through immersion in polyester resin, which because of a big initial fluidity (jellifying time 15 - 20 minutes) allows this thing.

The moulds made from glass and p.m.a.m. of methyl have been introduced in a vibrating magnetic field realized using a permanent magnet which turns round with a speed of 120 rot/min. (the work was done in the same conditions for all the systems). In this way, magnetic particles from the polyester resin mass realize a small change of place (while jellifying and after) close to the matrix interface - ranforte and can interact with the microholes or with the gaseous inclusions having the possibility to transfer them in the resin mass. In this case we could consider that we have under control the possible defects that could appear in the composite material interface.

The mould where the composite material is obtained must be covered with a developing coat of paraffin or polyvinyl alcohol and the free surfaces must be covered with wax or chalk paper in order not to stick with the plates that compound under pressure

54


the up and down surfaces of the parallelepipedic test tubes.

The developing takes place after about 15 minutes after the jellifying of the composite material and the test tubes are kept under pressure other 24 hours, because through contraction (about 0,5 – 1%) the material could be deformed (it is one of the most known defect when moulding). After 24 hours, the test tubes obtained in this way can be adapted to the standards cotes for the mechanical resistance tests. All this operations are presented in Fig. no. 2

3. The electronic microscopy analyze (sem) of the structure and chemical

Composition of the composite Materials glass fibre – polyester additivated with ferrite (Fe3O4)

To characterize the structure and the

composition of the composite materials glass fibre – polyester additivated with ferrite (Fe3O4) there have been analyzed the most 6 representative samples of the lot of the 15 samples. The electronic microscopy tests (SEM) were made to the C.F.H. laboratories belonging of the “Ecole Centrale Paris”.

They have the following codes like in the analyze reports (written in the brakes): A0(P0) – Sample representing the matrix of the composite material. F01(P6) - Sample representing composite polyester with a coat of glass fibre. BB0(P4) - Sample representing the polyester matrix additivated with ferrite (5% Fe3O4) C0(P2) - Sample representing the polyester matrix additivated with ferrite (5% Fe3O4) and under the influence of a external vibrating magnetic field. D0(P1) - Sample representing the polyester matrix additivated with ferrite (2.5% Fe3O4) I01(P3) - Sample representing the polyester matrix additivated with ferrite (2.5% Fe3O4 and ranforte with a coat of glass fibre).

There have been made cartographies through electronic microscopy both of the surface of the six types of samples and of their transversal sections. The results of these analyses are presented in the Fig. no.1(a ÷ c). Beside the study of the materials surface and section (P0, P1, P2, P3, P4, P6) there were analyzed three chemical elements (O,Si Fe), too – that can characterize the structure and especially the distribution of the elements that compound the composite material (ranforte, particles).

A0(P0) x 1000 S A0(P0) x 1000 S

A0(P0) x 1000 S A0(P0) x 5000 S A0(P0) x 1000 T S = for surface T = transversal

Fig. 1 The SEM analyze of the matrix of the composite material A0(P0) (a ÷ c) The SEM analyze of the matrix of the composite material in transversal section A0(P0) (e)

55

56

Polymerization

Ranforte preparation

Ranforte cut out

Introduction in ranforte mould

Dry grinding

Wet grinding

Disgassing T = 300 K

p = 10-2 N/m2

Dispersion of the particles in the

resin

Stocking additivated resin

Dimensional

stabilization

Developing sample

Magnetic field Covering with

developing agent

Polymer pouring in the mould

Polymer

ranfortifying

Mechanical processing

Test tube

Fig. 2. Elaboration technology of the test tubes from the experimented composite material

Particles

FeB3O4

Polyester resin Dispersion agents

Catalizator +

Initiator Devel

Moul

Ranfort

oping agent

d

e

THE ANNALS OF “DUNAREA DE JOS” UNIVERSITY OF GALATI FASCICLE IX METALLURGY AND MATERIALS SCIENCE, ISSN 1453 – 083X NR II – 2003

D0(P1) SEMx1000 S OKa,38 SiKa,66 FeKa, 153

Titre: D0(P1)T CARTO X1000 (11 Apr 01 14:06:43)

D0(P1) x 5000 T D0(P1) x 5000 T

Titre: C0(P2) T CARTO X1000 (11 Apr 01 11:04:25)

58

C0(P2) x 1000 T C0(P2) x 1000 T I01(P3) x1000S O

Si Fe

C0(P2) x 1000 T

I01(P3) x 5000S O

Si Fe Titre: I01(P3)T CARTO X1000 (10 Apr 01 15:36:02)

I01(P3) x 1000 T I01(P3) x 1000 T

59

I01(P3) x 1000 T B0(P4)x 1000 S B0(P4)x 1000 S

Titre: B0(P4)T CARTO X1000 (12 Apr 01 09:53:12)

B0(P4) x 1000 T

60

B0(P4) x 1000 T

F01(P6) x 1000S F01(P6) x 1000S F01(P6) x 1000T F01(P6) x 7500T

F01(P6) x 1000 T

61

F01(P6) x 1000 T Fig..3 The SEM analyze of cartography of the composite material surface F01(P6) (a ÷ d ) The SEM analyze of cartography of the composite material in transversal section F01(P6) (e ÷ h)

4. Conclusions

Taking in consideration all the information provided by the analyze through the electronic microscopy (SEM), we can draw the following conclusions:

- Analyzing the matrix surface (P0 and P6 in the SEM reports), in comparison with (P1, P2, P3, P4) having in their structure ferrite for about 2,5 and 5% Fe3O4, there can be noticed a bigger porosity, fact that is confirmed also by the bigger resistance to the water and acid solutions of the composite materials with particles as ferrites.

- The size of the ferrite particles (Fe3O4) is micronic (0,5 – 5µ m) and has a uniform distribution in the mass of the composite material (for that ones without ranforte and glass fibre (P1, P2, P3)).

- In case of ferrite concentrations bigger than 5% Fe3O4 (coded sample SEM with P4), in default of the vibrating magnetic field there can be noticed an uniform distribution of the particles and also a smooth tendency of agglomeration of small particles (due to the interaction of the own magnetic fields of the particles introduced in the matrix).

- Application of an external vibrating magnetic field leads to a “mosaic” of the composite material surface; there appear so-called “clusters” presented also in the specialty literature, especially of the ferrofluids and of the magnetic fluids, that represents chains of magnetic particles that close in their links the matrix of the composite material.

- Materials that contain “clusters” in their composition could generate two effects: first of all they can lead to a decrease of the resistance regarding the traction through anisotropy of the material, and second, to a increase of the resistance regarding the scissory due to the increase of the interactions number at microscopic level etc. (This thing is confirmed through the mechanical tests of the tests tubes).

- At the samples (P3 and P6 SEM codes) containing ranfortifying as glass fibre texture, we can notice in the transversal section a distributions of the ferrite particles near the interface in the interfaze zone.

This thing may have very favorable effects while the size of magnetic particles decreases.

- There cannot be seen imperfections as holes, inclusions, etc. at the interface level; this shows that the elaboration technological process of the composite material is satisfactory.

- Elaboration technology of the composite material is limited but the magnetic particles and the application of the vibrating magnetic field help it a lot (in order to avoid the appearance of the holes and inclusions).

- The homogenization of the matrix material through vibration of ferrite particles introduced in external magnetic field at interface level has a better behavior to mechanical solicitation in time.

5.Bibliography [1]Bode M, Cope B., Fox M. Process and plant technology, de Montfort University Leicester LE 19BH UK. 2001 [2]Chrétien G. Matériaux composites à matrice organique, Technique et Documentation (Lavoisier), 1986. [3]Gay D. Materiaux composites, "Hermes" Publishing House, Paris, 1991. [4]Geier M. H. Manuel Qualité des Composites, Technicque & Documentation - Lavoisier 11, True Lavoisier - F 75384 Paris, Cedes 08 - 1993. [5]Kashevsky B. E. Computer simulation of Ferro suspension structuring in steady and rotating fields, Journal of Magnetism and Magnetic Materials 122 (1993) 34 - 36, North - Holland. [6]Katz H. S. Haudbook of fillers and reinforcements for plastics, vol. 4, Van Nostrand Reinvold Publishing House , 1978. [7]Muşat V. D. Cermica avansată, E.T. Bucharest 2001 [8]Otaigbe J. U. Variation of Interfacial Shear in SFR composites MSE 383,

62

Unit, 3-7 Iowa State University, Materials Science & Engineering Dept. 2001. [9]Parker R. J. Advances in Permanent Magnetism, John Wiley and Sons, Inc., New York, 1990. [10]Parker R. J., Studders R. J Permanent Magnets and Their Application, John Wiley and Sons, Inc., New York, 1962 [11]Pelletier S., Gélinas C., Chagnon F. Effect of compaction temperature on magnetic properties of iron - resin composites, Industrial Materials Institute - Québec Canada J3R4R4 - 2000. [12]Pont R. P., Krishna R. M., Nagi P.S. XRD, SEM, EPR and microwave investigations of ferrofluid - PVA composite fillers,

Elsevier - Journal of Magnetism and Magnetic Materials 149 (1995), [13]Saini D. R., Nadkarni V. M., Grover P. D., Nigam K.D.P Dynamic mechanical, electrical and magnetic properties of ferrite filled styrene - isoprene - styrene, Journal of Materials Science, vol. 21, No. 10/1986 [14]taufer D. Introduction to Percolation Theory, Taylor Publishing House – Francis, London 1985

Acknowledgements The author wish to thank the Ecole Centrale, Paris for their technical support offered for SEM analisys composite materials studied in this material.

63

THE ANNALS OF “DUNAREA DE JOS” UNIVERSITY OF GALATI FASCICLE IX METALLURGY AND MATERIALS SCIENCE, ISSN 1453 – 083X NR. II – 2003

IOSIPESCU TEST APPLIED IN ORDER TO CHARACTERIZE THE COMPOSITE MATERIAL PE – GFR – Fe3O4

Dumitru DIMA

„Dunarea de Jos” University - Galati, Romania

E-mail: [email protected]

ABSTRACT The Iosipescu test has more and more applications in order to characterize the composite type PE – GFR – Fe3O4, which represents the polyester matrix ranforted with glass fibre and additivated with ferrite particles (Fe3O4). Iosipescu test has been applied successfully, showing eloquently the influence of a vibrating magnetic field to improve the quality of the interface. This thing is proved by the comparative study of the same composite material in default of and with a vibrating magnetic field introduced while its elaboration. KEYWORDS: Composite Materials, Polyester Matrix, Glass Fibre, Ferrites.

1. Introduction

The influence of the vibrating magnetic field

and the role of the additivated magnetic microparticles (Fe3O4) has been studied on a composite material type PE – GFR – Fe3O4. There has been used Iosipescu test in studying the influence of the vibrating magnetic field over the interface, proving its high quality. 2. Elaboration technology of the test tubes for the iosipescu test of the composite pe –

GFR – Fe3O4

Test tubes

Test tubes for the scissory have been moulded as parallelipipedic bars, their dimensions containing also the processing additions in comparison with the final dimensions.

All the test tubes were processed in order to be solicitated in the scissory device made on the basis of the Idaho model and Iosipescu testing principle.

The processing of the grasping / setting surfaces and of the slashes has been made at once in the device conceived by dr. B. Leitoiu from the Technique University “Gheorghr Asachi”, Iasi.

For the slashes processing there has been used a diamonted disc having diametre of 180 mm and the thickness of 3 mm, involving a rotation speed of 4000 rot / min.

The slashes symmetry was assured by the turning around the longitudinal axe of the device, each having two test tubes grasped. The turning round

of the test tubes was carried out after the complete processing of a side and of the slash.

After that, setting surfaces were processed on a machine for rectifying the plan, in order to eliminate the material layer with broken / pulled fibres (disc milling machine used for processing cannot cut the glass fibre, extremely taugh, but the tools of high speed could carry out the cut). Rectification of the scissory test tubes has been made on the lateral surfaces, too, in order to assure a constant thickness and the perpendicularity on the sitting surfaces. Due to the rigorous way of elaboration of the scissory test tubes, and to their selection in order to avoid the solicitation of some defected test tubes, the results of scissory trial have proved a great charging capacity of the material (which has a composite slight reinforced, the matrix from polymer being prevalent).

The breaking way of the test tubes and also the characteristic issued by the trial machine, representing the graphic relation between the power and movement of the mobile part of the device, shows a material relatively fragile. This characteristic (fragility) must be related with the leaking speed. The movement speed of the mobile traverse of the machine had values that decreased very much the leak effects. The breaking way of the test tubes is typical for the fragile materials. The fissures start on the edges (where the most powerful crushing action has place) and continue with inclinations to 45° opposite the scissory section. (Fig. nr.1).

64


Fig.1 Fissures propagation in the case of the test tubes subject to scissory

2.Determination of the scissory way g

The trial device The device has been assembled on a support and the charge was made with the same weights (in the same order) like for the traction trials. Generally, the big weights have produced leaking, that is why they were not used.

Tensometric ter devices

TER devices were resistant type, conceited especially for the scissory trials on the Isopescu test tubes, made by MicroMeasurements (USA).

Their type is: N2A-00-C032-500, with:

-the specific factor k=2,03 ± 1,0 %; -transversal sensibility factor kt = (1,2 ± 0,2) %;

-resistance R = 500 Ω.

Fig.2 The device for the scissory trial

Fig.3 Steaking way of TER on the test tubes

65


Materials

The test tubes used for these determinations

have been selected in order to eliminate these ones with defects in the researched area. The materials that have been selected for determinations are: G03 H03 I03 J03

In this way, there could be noticed, as comparison, the influence of the ferrite (ferrite concentration) and the influence of the vibrating magnetic field especially. Moreover, there could be noticed the influence of the previous solicitations over the determined constants (ecruisage). There has been seen a slight increase of the scissory way in the same time with the repetition of the solicitations on the same test tube, after a short time (it is a very interesting phenomenon that could be studied separately).

The characteristic curves

The characteristic curves were made through the unification of the coordinate points (γ, τ) with segments of straight.

There could be noticed for all the curves an accumulation of energy of plastic deformation, proportional with the area of the surface included between the charge curve and the discharge one (specific energy).

Also, there was noticed the decrease of the scissory energy absorbed through rigidity of the polymer structure following repeated solicitations for the materials H03, experiments II, IV) and J03 (in the experiments III, IV, V) in comparison with the previous experiments.

On the basis of the characteristic curves, there was calculated the medium way of scissory on the final zone of charge, obtained results being presented in Fig. nr. 4.

Modulul mediu de forfecare pe zona finala de incarcare

1.97

E+03

1.91

E+03

1.88

E+03

1.85

E+03

1.69

E+03

1.71

E+03

1.87

E+03

1.84

E+03

1.78

E+03

1.91

E+03

1.82

E+03

1.55

E+03

1.53

E+03

1.56

E+03

0.00E+00

5.00E+02

1.00E+03

1.50E+03

2.00E+03

2.50E+03

E0/1

E0/2

F03/

1

F03/

2

F03/

3

G03/

1

G03/

2

I03/

1

I03/

2

J03/

1

J03/

2

J03/

3

J03/

4

J03/

5

Materialul / Numar experiment

G (M

Pa)

Fig. 4 The medium way of scissory for the following materials: E0, F03, G03, I03 and J03

3. The analyze of the experimental

resuls

Due to the big number of systems of composite materials taken in consideration, we can draw many conclusions regarding the reciprocal influences realized by their constituents.

There was followed especially the influence of the ferrite particles over the mechanical properties in relation with the structure of the realized composite materials.

In this way, the rigorous selection of the test tubes has allowed the realisation of a comparative study from which one can draw the conclusion that the ferrite particles could positively influence the quality of the matrix – ranforte interface of a

composite material, leading them under the influence of a external magnetic field.

This thing is possible only in certain conditions regarding the size of the particles and their orientation in space.

The experiments have proved this thing and underlined that the elaboration technology of composite materials represents first of all an outstanding link in the quality “chain” of the composite material.

One could notice a good correlation of the experimental data with the theoretical premises from which the experiment has started, especially for the Iosipescu test for pure scissory trial.

The resistance tests for traction show very clear the limits of the elaboration technology of test

66

tubes, meaning that the unoriented distribution of the ranforte glass fibre or its imperfections lead to great tensions in the composite materials.

So, the most credible conclusions could be drawn analyzing the results of the scissory test trials.

This conclusion is pointed out from the fact that the material J03 presents small values of the elasticity modulus E, expression of a small rigidity, of an increased mobility, of the macromolecular constituents and because of the fitting that do not take over efficiently the external efforts of the composite material from the matrix.

Instead, for the material J03, to which there were assembled the tensiometric factors in the zone with small rigidity (close to the ranforte glass fibre), the values of the elasticity constants were bigger than those ones of the material I03, confirming that the scissory test offers the perspective of a more efficient comparative analyze.

This thing can be due to the test tubes realisation under traction too, that implies a complete revision of it.

Through this experiment, there was followed especially the influence of the vibrating magnetic field over the breaking characteristics. In this way, the following comparisons have been made:

A. The influence of the vibrating magnetic

field over the breaking characteristics - Comparison between B0 and C0 (both of

them containing 5% ferrite):

B0 C0 Variation (%)

σR,max (MPa) 32,903 37,813 +14,9

τR,max (MPa) 58,761 64,103 +9,0

There could be noticed an improvement of the breaking resistance regarding the scissory and traction.

Remark: There has been taken in consideration the best

performancy recorded for each material, because: - the trials were made on a relatively small

number of test tubes; - the breaking when small charges is

provoked by defects: *seen: superficial holes, geometry distorted because of the solidification tensions, or *hiden: internal holes, fissures or anizotropies provoked by sedimentations of the ferrite or by the disposal of the fitting.

- Comparison between D0 and E0 (both of them containing 2,5% ferrite):

D0 E0 Variation (%)

σR,max (MPa) 69,058 59,236 -14

τR,max (MPa) 57,540 69,444 +20

There can be noticed the increase of the resistance for scissory. Analyzing the results obtained to the traction

and scissory resistance in a comparative way B0 – C0 and D0 – E0, one can draw the conclusions that a vibrating magnetic field realizes a bigger homogeneity of the respective material of the polymer matrix.

The immediate result of this influence can be interpretated through an increase of the resistance under traction and through the scissory under bigger concentrations of ferrite (5%).

When the concentration is smaller (2,5% ferrite) the increase of scissory resistance is maintained but the resistance for traction decreases. One can consider that in a vibrating magnetic field there appear “clusters” that do not allow interactions of the influence zones of the magnetic particles, in this way the particles could represents tension factors (thing confirmed by the SEM analyses too).

We could say also that the elaboration technology of the composite material has an influence in this way, fact that confirms that the resistance test for pure scissory is more credible.

The role of a vibrating magnetic field over the interface of a composite material polyester – glass fibre – ferrite particles (Fe3O4) can be underlined by the following comparative studies:

- Comparison between G03 and H03 (both of them containing 5% ferrite):

G03 H03 Variation (%)

σR,max (MPa) 53,161 47,205 -11,2

τR,max (MPa) 56,664 65,312 +15,2

There can be noticed an increase of the resistance for scissory, that confirms the above theories.

- Comparison between I03 and J03 (both of them containing 2,5% ferrite):

I03 J03 Variation

(%)

σR,max (MPa) 60,163 33,126 -44,90

τR,max (MPa) 68,054 60,214 -11,52

67

B.The influence of the vibrating magnetic field over the elasticity characteristics E, υ şi GAlso, through the comparative analyze of the elasticity characteristics of composite materials (E, υ şi G), there could be pointed out the influence of a vibrating magnetic field over the interface of the matrix – ranforte.

In order to determinate these characteristics

there have been used test tubes (carrefully selected); from this reason the values calculated as average of results of the efectuated experiments for a material are compared.

- Comparison between G03 and H03 (both of them containing 5% ferrite):

G03 H03 Variation (%)

Emed. (MPa) 4980 5454 +9,50

υmed. 0,304 0,3254 +7,00

Gmed. (MPa) 1855 2112,5 +13,88

There can be noticed an increase of the rigidity (meaning of the propriety of taking over bigger charges with smaller deformations) under traction and scissory for H03 in comparison with G03.

- Comparison between I03 and J03 (both of

them containing 2,5% ferrite):

I03 J03 Variation (%)

Emed. (MPa) 5945 2198 -63,00

υmed. 0,207 0,1935 -6,52

Gmed. (MPa) 1620 1860 +14,80

The great increase of the medium scissory

modulus (Gmed) especially – regarding the composites refortified with glass fibre and with particles (5% ferrite) under the influence of the vibrating magnetic influence – can be a result of realisation of a matrix structure more homogeneous at interface level, through the displace of microholes from the interface in the matrix through chemisorptions of the moving magnetic particles.

The SEM analyze point out this thing, the ferrite micronic and submicronic particles being very close to the interface and the porosity of the test tubes surface subject to cartography being by far smaller.

These quality studies realized through electronic microscopy SEM and also the uniform distribution of the ferrite in the matrix in samples subject to Iosipescu test confirm the fact that these trials are much more credible in comparison with the resistance test for traction. These are very important to realize dynamic test of resistance under bending, fact that may constitute the subject of a separated study in correlation with the structure of the composite material and the possible modifications in the time of trials.

- Comparison between I03 and J03 (both of

them containing 2,5% ferrite):

I03 J03b Variaţia (%)

Emed. (MPa) 5945 5503 -7,4

υmed. 0,207 0,189 -8,6 There could be noticed the increase tendency

of the values of the elasticity characteristics for the materials having 5% ferrite in comparison with the decrease tendency recorded for the materials having 2,5% ferrite.

The biggest increase could be noticed for the medium values of the scissory modulus (+13, 88% for the materials containing 5% ferrite, for H03 in comparison with G03).

The scissory modulus has also an increase (with 14,8%) for tha materials containing 2,5% ferrite (for J03 in comparison with I03).

Bibliography

[1]Chrétien G. Matériaux composites à matrice organique, Technique et Documentation (Lavoisier), 1986. [2]Cognard Ph. Les applications industrielles des matériaux composites, "Moniteur" Publishing House, Paris, 1989, vol. I şi vol. II. [3]Derrien K., Fitoussi J., Baptiste D Approche statistique multi-échelles de l’endommagement dans les composites, Revue des composites et des matériaux avancés, vol. 8, nr. Hors série/1998, [4]Dima D. Obţinerea unor materiale compozite cu matrice polimeră şi particule de carbon, Conferinţa Naţ. de Chimie - Fizică Secţiune 8 [5]Dima D. Metode de analiză fizico - chimice şi structurale a materialelor composite, Referatul nr. 2 în cadrul minimului de pregătire a tezei de doctorat - Februarie 1998 [6]Doniga A., Dima D. Rapid investigation method of silicon steel magnetically domains, Proceedings of the 39th International Seminar on Modeling and Optimization of Composites - MOC’39 Sensible

68

experiment in materials Science 26 - 27 April 2000 Odessa [7]Fitoussi J., Guo G., Gineste B., Baptiste D. Détermination d'un critère tridimensionnel de rupture a l'interface fibre matrice d'un composite organique à renforts discontinues, Comptes .....des JNC 9, vol. 1, p. 213 - 222, 22 - 24 nov. 1994. [8]Gay D. Materiaux composites, "Hermes" Publishing House, Paris, 1991.

[9]Geier M. H. Manuel Qualité des Composites, Technicque & Documentation - Lavoisier 11, True Lavoisier - F 75384 Paris, Cedes 08 - 1993. [10]Leiţoiu B. Teză de doctorat - Contribuţii la optimizarea formelor epruvetelor din metal şi din materiale compozite utilizate la studierea caracteristicilor mecanice, Univ. Tehnică „Gheorghe Asachi” Iaşi. 1998

69


FASCICLE IX METALLURGY AND MATERIALS SCIENCE, ISSN 1453 – 083X NR.II – 2003

THE INFLUENCE OF METALLIC MASS UPON CAST IRON HOT PLASTIC DEFORMATION

Aurel CIUREA, Marian BORDEI, Adrian VASILIU

“Dunarea de Jos” University, e-mail: [email protected]

ABSTRACT

Cast iron hot plastic deformation is a special process, generated, first of all, by

specific composition and structure of cast iron, but also by the features of plastic deformation. The present work is aiming to present the way chemical composition of the metallic mass influences the process of cast iron plastic deformation.

1.The influence of cast iron chemical composition upon the ability of hot plastic deformation

The chemical composition may have a direct influence upon hot plastic deformation process, by

limitating the maximum deformation degree or it may influence the mechanical properties of deformed cast iron. Thus, the carbon has the tendency to lower the values for hardness, traction strength, elongation and tearing strength.

2,00

1

2

3%

30

40

40

50

50

60

60

HRB

HRB

70

70

80

80

90

90

100

100

110

110

2,4 2,8

% C

Traction

Elongation

Strength

Strength

3,2 3,6

δ

rt iş σσ

2mm/kgfdaN/mm2

Fig. 1 The influence of carbon upon deformed cast iron regarding mechanical properties

As a result, for plastic deformation, those cast irons will be chosen having the carbon content according to the requirements of mechanical properties. Together with carbon, the silicon, an element present in cast iron chemical composition, has an influence upon the ability of plastic deformation, both alone and combined with alloying elements contained by the cast

iron. The influence of the silicon has is showed by an important reduction of compression strength and maximum repressing, in exceeding the content necessary to graphitization.

70



a b

0 1 2 3 4% Si

5 6

10

20 Resturide perlit�

Rezisten�a lacompresiune

Refulare maxim�

30

30

40

50

60

70 kgf/mm2

Încerc�ri preliminareÎncerc�ri principale

Refu

lare

ma

xim

�

Rezi

sten

�a

laco

mp

resi

une

cd e f gdaN/mm2

Fig. 2 The influence of silicon upon cast iron compression strength and maximum repressing

Where a, b, c, d, e, f, g represent various cast irons compositions, as in the table below:

Table 1. The chemical composition of heat numbers a, b, c, d, e, f, g

a b c d e f g

% C 3,08 3,04 3,02 2,70 3,01 2,90 2,80

% Si 0,92 1,72 2,50 2,64 3,70 4,40 5,80

% Mn 0,05 0,05 0,06 - - - 0,13

% P 0,02 0,02 0,02 - - - 0,02

An extrapolation to smaller silicon contents is not possible because, by carbides stabilizing, big quantities of pearlite and ledeburite appear. When nickel is present in cast iron chemical composition, in order is to be observed that up to 3%, it has a favorable influence upon compression strength and maximum

repressing, with values ranging depending on silicon content. Fig. 3 shows the influence nickel has upon compression strength and maximum repressing of cast iron with variable silicon content ranging between 2,5 – 3% for six representative heat numbers.

a

0 1 2 3Nickel

The silicon decreasingfrom 2,5 to ~0,3 %

Max

imum

repr

essin

g

4 5 6 %

10

40

50

60

70kgf/mm2

20

30%

b c d

Compression strength

Maximum repressing

e f daN/mm2

Fig. 3 The influence nickel has upon compression strength and maximum repressing of cast iron with variable

silicon content

71



Where a, b, c, d, e, f represent various compositions of cast irons, according to the table

bellow:

Table 2. The chemical compositions of heat numbers a, b, c, d, e, f

a b c d e f % C 2,84 2,92 3,04 3,08 3,08 3,06 % Si 1,76 1,26 0,70 0,72 0,14 0,30 % Ni 0,90 1,85 2,70 3,70 3,95 5,63

In order to establish the influence of silicon, here’s below, fig. 4, the influence it has upon compression strength and maximum repressing of cast

iron with a constant silicon content, for three main heat numbers.

a b c

10

0 3 52Nickel

Compressionstrenght

Maximum repression

Preliminary testsMain tests

4 %1

Ma

xim

um re

pre

ssio

n

Com

pre

ssio

nst

reng

th

20

30

40

50

60

70kgf/mm2Constant Silicon ~0,8%

%

daN/mm2%

Fig. 4 The influence of nickel upon compression strength and maximum repressing of cast iron

with a constant silicon content Where a, b, c represent various cast iron compositions according to the following table:

Table 3. Chemical compositions of heat numbers a, b, c

a b c % C 2,92 2,79 2,82 % Si 0,80 0,80 0,79 % Ni 1,60 3,37 4,43

The accompanying element that is not wished, the phosphor, influences the compression strength and

maximum repressing as in the fig. 5.

daN/mm2

0 0,5 1,0Phosphor

ÎPreliminary testsMain tests

1,5 2,0 %

10

30

40

50

60cba

20

30%

Ma

xim

um re

pre

ssin

g

Maximum repression

Compressionstrength

Co

mp

ress

ion

stre

ngth

Fig. 5 The influence of phosphor upon cast iron compression strength and maximum repressing.

72



Where a, b, c represent chemical compositions as in the table below:

Table 4. Chemical compositions of heat numbers a, b, c

a b c % C 3,42 3,30 3,20 % Si 1,76 1,76 1,82 % P 0,50 0,88 1,64

As it fig. 5 shows, the increase of phosphor content from 0,5 to 1,7% leads to the decrease of compression strength, while maximum repressing has a decrease of up to 1% and then it is about constant. Phosphor has also influence upon temperatures plastic

deformation can take place and cast iron mechanical characteristics after deformation. In fig. 6 we can see the traction strength has a maximum level of temperatures depending on phosphor content and they as lower as the phosphor content increases.

60030

40

50

60

70

80

90

100

110

700

0,1 % P

0,25 % P

0,5 % P

0,8 % P

1,3 % P

Rolling temperature

Tra

ctio

n st

reng

th

800 900 1000 1100 Co

2mm/kgf

Fig. 6 The way traction strength depends on rolling temperature at phosphor contents ranging between 0,1 and 0,3%

The rolling deformation ability depending on temperature and phosphor content is represented by fig. 7 where we can see that, once the phosphor content increases, the temperature range deformations decreases

below 950˚C. We can say that it is the result of phosphor eutectic formation with a melting temperature of 950˚C.

0500

600

700

800

900

1000

1100

1200

0,2 0,4 0,6 0,8Phosphor

Rolling limits

Mechanical properties range

Rolli

ng te

mp

era

ture

1,0 1,2 %

Co

Fig. 7 Deformation abilities by rolling depending on temperature and phosphor content

2.Conclusions

The chemical composition is one of the factors in hot plastic deformation of cast iron, both regarding the parameters of deformation process and mechanical properties obtained by cast iron further to deformation process.

3.References [1].Ciurea Aurel – Teza de doctorat ”Cercetări privind deformarea plastică a fontelor” – Galaţi – 2003 [2].Piwoworski Eugen – “Fonte de înaltă calitate” – Ed. Tehnică, Bucureşti – 1967 [3].Sofroni L.,Stefanescu D.M., Vincenz Conrad – “Fonta cu grafit nodular” - Ed. Tehnică, Bucureşti – 1978.

73



RESEARCH AND STUDY REGARDING THE INFLUENCE OF THE COLD PLASTIC DEFORMATION ON THE STRUCTURE AND THE

PROPERTIES OF SOME AUSTENITIC STAINLESS STEELS

Ovidiu DIMA


ABSTRACT

The paper shows that, besides the classic curve resulted after the tensile test,

other indicators may be used for assessing the behaviour in cold plastic deformation, such as: the modulus of strain hardening n, the stability index of austenite S, and the martensitic transformation point Ms. The study shows that austenitic stainless steels have, in general, a good behaviour in cold plastic deformation, but variations, sometimes quite small, of the chemistry may determine major changes. Therefore, a rigorous control on the elements ENi, ECr is mandatory for obtaining a stable austenitic structure in the steel. There was highlighted that an unstable austenitic structure transforms into a martensitic-type structure as a result of the cold plastic deformation. This has been magnetically highlighted and emphasizes the hardening phenomenon.

KEYWORDS: tensile test, in cold plastic deformation, austenitic structure

1.Theory

The assessment of the cold plastic deformation property of the austenitic stainless steels may be done taking into account the classical curve for tensile test and the hardening parameter represented by the modulus of strain hardening. The classical tensile test curve figure 1 shows that the austenitic stainless steels have a great deformability as compared to the other types of martensitic, even ferritic stainless steels. Consequently, austenitic stainless steels are better fit for cold plastic deformation than the latter.

Fig.1 The characteristics tensil test curve of the austenitic stainless steels.

Cold deformation is accompanied by the hardening phenomenon. The modulus of strain hardening indicates the intensity of the phenomenon and consequently, the cold processing ability. The modulus of strain hardening can be determined with the following formula:

n = σg εg / σg- σc; σg = Fmax/Ag true tensile strength before test bar’s necking (striction) σc = Fc/Ac yield stress εg = lnAc/Ag logarithmic degree of deformation in vicinity of the necking A0 = initial section of the test bar Ag = test bar’s section in vicinity of the necking The formula can be applied in case of the steels with high yield point, and stainless steels may be matched in this category. Austenitic stainless steels have the modulus of strain hardening of 0.4 – 0.5 if their structure is stable and of 0.5 – 0.95 if their structure is unstable. The stability of the structure in cold plastic deformation is assessed with the stability coefficient S = ENi+0,4ECr where: ENi = equivalent in gamma-type elements ENi =30%C + 30%N + 0.5%Mn + %Ni ECr = equivalent in alpha-type elements ECr = %Cr + % Mo + % Si

Rr [daN/mm2

]

80

60

40

20

20 30 40 A%

Rr=94 Martensitic steel

Rr=53 Feritic steel Rr=60 Austenitic steel

74



The austenite becomes unstable if S< 22.5,

which can be seen through the partial transformation of austenite into martensite-type structure when cold plastic deformation is applied. More intense the martensitic transformation is, stronger the steel hardens.

Figure 2 shows the influence of the cold plastic deformation degree on the martensite quantity. The quantity of martensite increases with the deformation degree applied and the transformation intensity increases once the deformation temperature is reduced to negative values.

-188 oC -70oC -30oC 0 oC

10 oC

22 oC

50 oC80 oC

0,2% 0,4% 0, 6 lnh0/h

%M s

80%

60%

40%

20%

Fig.2 Influence of the cold plastic deformation degree

on the martensite quantity

Every steel is characterized by a temperature

of transformation of austenite in martensite, named point of martensitic transformation Ms. The calculation of Ms point temperature may be done using the empiric formula set up by Eihelman and Hell.

In case of austenitic stainless steels, this point has strongly negative values from –800 to –1500C, as the austenite at ambient temperature is stable or meta-stable. Small changes of the chemistry have a great influence over the temperature of starting of martensitic transformation, i.e. one steel grade CrNi18.8 has Ms = - 1500C and another steel grade CrNi17.7 has Ms = -200C.

The temperature in deformation process influences the hardening of austenitic stainless steels. It was found that in case the deformation process is applied at temperatures higher by 150-2000C than the martensitic transformation temperature, the transformation does not take place irrespective of the deformation degree applied and the hardening trend is more reduced.

In conclusion, the assessment of the cold processing ability of an austenitic stainless steel may be made having in mind the modulus of strain hardening n, the degree of stability of austenite S and the martensitic transformation point Ms.

2.Experimental research

In purpose of the research, plates made of

four austenitic stainless steel grades have been used. They have the chemistry as per the table 1.

Table 1

Chemical composition [%] Steel code

Plates tickness

mm. C Mn Si S P Cr Ni Cu Mo Ti V N Nb

1. 0,8 0,04 1,2 0,5 0,005 0,03 21,5 7,2 0,12 0,1 - 0,12 0,04 - 2. 3,0 0,03 1,2 0,4 0,025 0,02 18,9 8,5 0,19 0,1 0,01 0,09 0,02 - 3. 4,0 0,02 1,2 0,5 0,001 0,03 17,9 13,5 0,07 2,6 0,01 0,05 0,03 0,03 4. 6,0 0,12 1,7 1,8 0,007 0,03 19 14,1 0,2 0,3 0,02 0,05 0,03 0,03 Simbol steel

1. 5 Ni Cr 180 – W4301.

2. 2 Ni Cr 185 – W4306.

3. 2 Mo Ni Cr 175 – W4435.

4. 10 Ni Cr 180 – W4300.

75



Samples rolled microstructures are showed in figure 3.

Fig.3 Samples rolled microstructure x100

a) 5NiCr180, A+C, grain size 3; b) 2NiCr185 A+I+1,5%F; c) 2NiCr185 rolled; d) 2MoNiCr175, A+I, grain size 3-4; e) 2MoNiCr175, rolled;

f) 10NiCr180, A+C grain size 6; g) 10NiCr180, rolled, A+C+0,8%F A-austenite, C-carbides, F-ferite, I-inclusion.

a)

b) c)

d) e)

f) g)

76



Table 2 Steel 1 Steel 2 Steel 3 Steel 4

Code tickness mm

strain Code tickness mm



strain

1.0. 0,8 0 2.0. 3 0 3.0. 4,0 0,133 4.0. 6,0 0 1.1. 0,54 0,393 2.1. 2,5 0,265 3.1. 3,5 0,287 4.1. 5,5 0,087 1.2. 0,38 0,744 2.2. 2,0 0,405 3.2. 3,0 0,470 4.2. 5,0 0,182 1.3. 0,33 0,885 2.3. 1,5 0,693 3.3. 2,5 0,693 4.3. 4,5 0,287

- - - 2.4. 0,75 1,386 3.4. 2,0 1,66 4.4. 4,0 0,405 - - - - - - 3.5. 0,75 - 4.5. 1,1 1,69

Table 3

Steel 1 tickness 0,8 mm Sample code 1.0. 1.1. 1.2. 1.3. - -

g[mm]. 0,8 0,54 0,38 0,33 - - d[mm]. 0,211 0,150 0,145 0,143 - - HV5. 208 412 441 454 - -

ln hi-1/hi 0 0,393 0,744 0,885 - - Steel 2 tickness 3,0 mm.

Sample code 2.0. 2.1. 2.2. 2.3. 2.4. - g[mm]. 3,0 2,5 2 1,5 0,75 - d[mm]. 0,212 0,185 0,165 0,153 0,147 - HV5. 227 271 341 396 429 -

ln hi-1/hi 0 0,205 0,405 0,693 1,386 - Steel 3 tickness 4 mm.

Sample code 3.0. 3.1. 3.2. 3.3. 3.4. 3.5. g[mm]. 4,0 3,5 3,0 2,5 2,0 0,75 d[mm]. 0,222 0,197 0,180 0,167 0,160 0,147 HV5. 188 299 286 332 362 429

ln hi-1/hi 0 0,133 0,287 0,470 0,693 1,66 Steel 4 tickness 6 mm.

Sample code 4.0. 4.1. 4.2. 4.3. 4.4. 4.5. g[mm]. 6,0 5,5 5,0 4,5 4,0 1,1

d[mm]. 0,204 0,188 0,170 0,168 0,159 0,137

HV5. 233 293 317 329 367 494

ln hi-1/hi 0 0,087 0,187 0,287 0,405 1,69

Test bars have been made from the materials to be analyzed. The shape of the test bar was a strip, having the width of 10 mm and the length of 70 mm, at the thickness of the initial plate.

The test bars have been submitted to cold plastic deformation through a rolling process using different deformation degrees and different gauges, being coded with two digits, first representing the steel grade 1,2,3,4 and the second representing the number of the test bar as per the deformation degree applied 0,1,2,3,4,5.

Table 2 shows the cold rolling program achieved. Having in view the high forces of deformation required to cold deformation process and the low rigidity of the rolling stand that has been

used, the rolling process has been conducted using very small degrees of partial deformation, cumulated.

Hardness samples, metallography samples for structure analysis and traction samples have been taken from the test bars obtained through the deformation program applied.

Vickers test HV5 has been used for the assessment of the hardness. The results of hardness test are shown in table 3.

The degree of deformation in cold rolling process has been expressed through the logarithmic deformation degree ε = lnho/h, h0 and h representing the thickness of the plate before and after rolling process.

77



Having in view the reduced sizes of the strips, the test bars have been conceived in non-standardized sizes, as per figure 4.

70

20

g

4

r

20

Fig.4 The shape anal sizes of the traction strips.

The graphic shows that the hardening trend is more accentuated in case of the steels coded with 1, 2 as compared to the steels coded with 3, 4.

Figure 5 shows the variation of the hardness with the degree of deformation and highlights the hardening trend of the steels with the cold plastic deformation degree.

HV5

400

300

200

0 0,5 1 ln h0/h

1 32

4

Fig. 5 The variation of the hardness steel with the

deformation degree

This hardening trend is also highlighted on the basis of the tensile test results from the table 4.

Table 4 Steel 1 Steel 2

Sample code

Strain ε

Rm daN/mm2

A %

ψ %

Sample code

Strain ε

Rm daN/mm2

A %

ψ %

1.0 2.0 0 55,86 50 57,5 1.1 2.1 0,205 72,81 32,5 49.7 1.2 2.2 0,405 91,82 15 43,7 1.3 2.3 0,693 118,2 7,5 20,9 1.4 2.4 1,386 139,6 2 10,3

Steel 3 Steel 4 Sample

code Strain ε

Rm daN/mm2

A %

ψ %

Sample code

Strain ε

Rm daN/mm2

A %

ψ %

3.0 0 56,91 40 54,5 4.0 0 66,16 45 48,6 3.1 0,133 68,18 25 46,4 4.1 0,087 78,92 35 47,4 3.2 0,287 85,18 17,5 39,37 4.2 0,187 92,46 22,5 42,36 3.3 0,470 97,62 12,5 30,7 4.3 0,287 104,8 17,5 34,9 3.4 0,693 98,41 10 29,7 4.4 0,405 111,5 12,5 31,6 3.5 1,66 125,4 5 5,3 4.5 1,69 121,7 5 10,46

We may notice that the steels coded with 2

show a more pronounced trend of increasing the tensile strength and decreasing the elongation and the reduction of section at fracture, as compared to steels coded with 3 and 4. This may be explained based on the modulus of strain hardening and stability index of austenite.

Table 5 shows the calculated values for the modulus of strain hardening n and the stability index of austenite S for the studied steels.

Table 5

Symbol steel. Code. S. n. Ms. 5 Ni Cr 180 – W4301.

1. 17,86 - -120

2 Ni Cr 185 – W4306.

2. 17,8 0,8 -77,5

2 Mo Ni Cr 175 – W4435.

3. 23,19 0,48 -237

10 Ni Cr 180 – W4300.

4. 25,04 0,478 -389.

Steels coded with 1 and 2 have the stability coefficient S < 22.5 and the modulus of strain hardening in the range of values of 0.5 – 0.9. These show that the steels have an unstable austenitic structure. In cold plastic deformation process they undergo a partial transformation of the austenite into martensite-type structures, which emphasizes the hardening trend of the steel.

Steels coded with 3 and 4 have the stability coefficient S > 22.5 and the modulus of strain hardening lower than 0.5. These steels present an austenitic structure with high degree of stability. During cold deformation process they do not undergo martensite-type structure transformations and the hardening trend is more reduced. In this case, the plastic deformation ability is greater, strain level being more reduced.

The assessment of the transformation into martensitic-type structure in cold deforming process

10

78


FASCICLE IX METALLURGY AND MATERIALS SCIENCE, ISSN 1453 – 083X VOL II - 2003

may be also done on the basis of the

martensitic transformation point temperature Ms. We may notice that in case of the steels

coded with 1 and 2, Ms is closer to the ambient temperature. For the steels coded with 3 and 4, Ms has very low values.

This indicator shows that martensitic-type transformations happen in case of the steels coded with 1, 2. Martensitic-type structures present magnetic properties. Therefore, the transformation may be highlighted by measuring the magnetic properties of the samples.

As the metallography analysis showed, all the initial materials had an austenitic structure, some of them with fine carbides or with tracks of * ferrite. None of the initial materials presented magnetic properties. After the deformation, the steels coded with 1 and 2 presented magnetic properties, which confirms the fact they had an unstable austenitic structure which has transformed into a martensitic-type structure as a result of cold plastic deformation. Higher the deformation degree, more intense is the transformation. The qualitative assessment of the transformation may be done with a permanent magnet. Materials coded with 3 and 4 did not presented magnetic properties after the deformation process, which shows that they had a stable austenitic structure.

The quantitative assessment of the martensitic-type transformations in case of stainless steels with unstable austenitic structure may be done by measuring the magnetic properties with specific devices.

3.Conclusions

The study and the research showed that the austenitic stainless steels have a great deformability, this depending upon the chemistry of the steel and the stability degree of the austenite. A variation, though not so great, of the concentration of alpha-type and gamma-type elements may lower the stability index of the austenite and increase the temperature of the martensitic transformation point, making possible the occurrence of some martensitic-type structure in the steel. The intensity of these transformations depends on the deformation degree applied, but also on the deformation temperature. Therefore, the characterization of the behaviour of the austenitic stainless steels in cold plastic deformation may be done based on the modulus of strain hardening n, the stability index of austenite S, and the martensitic transformation point Ms.

References

[1] M. David Le travail a froid des aciers inoxydables. Les toles minces et feuillards. 1969 [2] M. Truşculescu, A Ieremia Oţeluri inoxidabile şi refractare Ed. FACLA Timişoara 1983 [3] V. Micloşi, I. Lupescu Sudarea prin topire a oţelurilor aliate Ed. Tehnică Bucureşti 1970 [4] N. Geru Tehnologia structurală a proprietăţilor E.D.P. Bucureşti 1980 [5] C. Atanasiu s.a. Încercarea materialelor Ed. Tehnică Bucureşti 1982 [6] N. Cănănău, V. Petrescu Tehnologia deformării plastice Ed. MACARIE Târgovişte 2002 [7] N. Cănănău Teoria deformării plastice Universitatea Dunărea de Jos Galaţi 1995

79



ANALYSIS OF DIFFUSION PROCESSES AT INTERFACE COBALT - STEEL SHEETS

Lucica BALINT, Simion BALINT, Tamara RADU, Emil STRATULAT

University “Dunărea de Jos” of Galati, Romania, E-mail: [email protected]

ABSTRACT

In the present study, diffusion processes after heat treatments on cobalt

covered steel sheets was analysed. Cobalt layers on steel sheets were made by electrodeposition, having in view the layers adherence on sheets for optimal conditions of diffusion processes.

Heat treatments were made in protection atmosphere, at temperature above 1000 0C. The work makes out diffusion zone, which results at the interface between cobalt layer and steel sheet, after opposite migration of cobalt atoms and iron atoms. The result of treatments has been studied by optical microscope analysis and electronic microscope examination.

Improvements of conformance properties of electrochemical deposition coated sheets can be made by optimal choice of electrodepositing parameters and heat treatment which have helpful effect on diffusion process.

The base metal is a steel plate of low carbon concentration having the following chemical composition: 0,05%C; 0,26%Mn; 0,02%Si; 0,012%P; 0,015%S; 0,039%Al and iron.

The cobalt coat was electro-chemically deposited in an electrolyte solution of 300g/l CoSO4 ⋅ 7H2O, 50 g/l CoCl2 ⋅ 7H2O, 30 g/l H3BO3, cu pH = 5, and was exposed for 60 minutes, while the current density was 600 A/m2 .

The cobalt coat adherence to the base is an important property as it facilitates the diffusion activation by heat processing.

1. Introduction

The temperature of treatment is analyzing of phase diagrams. Iron and cobalt form a complete series of disordered f.c.c. solid solution γ, at elevated temperatures. At low temperatures the b.c.c. solid solution α, exist above 7300C up to ~ 75 %Co. The α b.c.c. solid solution near equiatomic composition undergo below 7300C an atomic ordering to the CsCl structure tip α/, and this order-disorder transformation plays an important role in determining the magnetic properties of these materials. This materials tend to be brittle because of the ordering of the α phase and the presence of trace amounts of carbon, hydrogen and oxygen. Consequently, the magnetic properties attainable are sensitive to heat treatment and purity of the alloys.

Heat treatments were made in argon protection atmosphere, at 1000 0C, 1150 0C, 1250 0C . The work makes out diffusion zone, which results at the interface between cobalt layer and steel sheet, after opposite migration of cobalt atoms and iron atoms. Various coating procedures, based on cobalt, oriented to physico - chemical and magnetically characteristics modified, have emerged lately. Co-Fe alloy protective coatings are considered among favourites due to their high weld ability, resistance to corrosion, excellent varnish ability, lower specific weight of layer, and thus of the whole product, resulting in zinc saving. Zinc alloying with iron is put into operation using diffusion procedures during heating of galvanized products.

.

80



Photo 1a. Composition and concentration various of Co and

Fe, 10000C, x600

Photo 1b. Co distribution, 10000C x600

Photo 1c. Fe distribution,

10000C x600

Photo 2a. Composition and concentration various of Co and Fe,

11500C, x600

Photo 2b. Co distribution, 11500C x600

Photo 2c. Fe distribution, 11500C x600

Photo 3a. Composition and concentration various of Co

and F, 12500C, x600

Photo 3b. Co distribution,

12500C x600

Photo 3c. Fe distribution, 12500C x600

81



2. RESULTS AND CONCLUSIONS As a result of microscopic examination (both optical and electronic) and X-ray diffraction testing in coating layer were detected phase ratio changes, according to heat treatment parameters. Considering that the physico-chemical and magnetical properties differ for the two phases and phase ratio is to decide the properties of the product; these structural changes are of great importance in practice. The desirable coating would be a Co-Fe coating with 51,3 % Co, with a structure of order-disorder transformation consisting of α/ phase. Therefore it is necessary an accurate mathematical correlation between layer phenomena, depending on heat treatment parameters and iron content of coating, both defining a certain structure. It was found that structural changes, due to rise temperature occur in some distinct stages: a).ε phase disappearance and α/ phase growing (430-450 0C); b).α/ phase gradual disappearance and (αCo, αFe )phase growing (450 - 1000 0C); c).the growths (αCo , γFe ) phase (1000 – 1250 0C). These changes include transformations characteristic to a reactive diffusion process. In this type of diffusion, the rate depends on reaction constant ratio, resulting three possible processes . In the case specified, a) under 450 0C, when the diffusion of iron atoms is negligible, the process is considered as diffusion process; b) at 450÷1000 0C, when the mobility of iron atoms is rising is considered as intermediate process; c) at 1000÷1250oC is considered a kinetic process, emphasized by the rapid growth of Γ phase. Let: x1, x2, x3 – instantaneous points of ε/α/, α/ /(αCo , αFe ), (αCo , αFe )/ (αCo , γFe ) interfaces able to change their position with v1, v2, v3 rates and JI – the iron quantity related to the moving interface and representing the iron given by “i”

“phase to “i+1“phase (where i =1; 2; 3). Phase growing ratio is calculated using the relations:

( ) ;1vdt

,d FeCo =γα

( ) ;1vvdt,d

2FeCo −=

αα

;2vvdt

d3 −=

α′

The content of iron at interface boundaries was considered according to the Fe-Co equilibrium diagram. Considering J0 as total quantity of iron related to the interfaces in the underlayers of coating the law of iron conservation results in the following:

( ) ( ) ;,615,0

2, 10

10 dtdcc

dtdJJ FeCoFeCo γαγα

=+

=−

( ) ( );dt,d17.0

2cc

dt,dJJ FeCo12FeCo

21

αα=

+⋅

αα=−

;dt

d075.0

2cc

dtdJJ 23Co

32

α′=

+⋅

α′=−

Total input quantity of iron is in balance with the four of phase growth (local diffusion of iron atoms is moving at the same rate as interface 3), therefore:

( ) ( ) ;

dtd

075.0dt,d17.0

dt,d615.0J FeCoFeCo

0

α′+

αα+

γα=

where c0=99% Fe, c1=24,5% Fe, c2=10% Fe, c3=5% Fe. The quantity of iron on various interfaces is related to difference of chemical potential of iron through the chemical constant Ki (i=0; 1; 2; 3)

);(KJ bFe

aFeii μ−μ=

where: ;eKK RT

Q

i)T(i

ai−

=

where =μaFe

chemical potential of iron at interface left

side; =μb

Fe chemical potential of iron at interface

right side; Qae = interface activating energy.

82



The chemical potential of iron can be approximated with its concentration, resulting J0, J1, J2, J3 using relation (8)

J0 = K0(cFe-c0) = 0.01K0J1 = K1(c0-c1) = 0.75K1

J2 = K2(c1-c2) = 0.14K2

J3 = K3(c2-c3) = 0.05K3

Replacing these values in relations is obtained:

Are obtained:

v1 = 0,0135K0 - 0,681K1

v2 = 0,0135K0 + 0.702K1-0,679K2

v3 = 0,0135K0 + 0.702K1 + 1,024K2 - 1.31K3

References

( )10

10

a

FeCo K21,1K01,0615,0

JJdt

,d−=

−=⎥⎦

⎤⎢⎣⎡ γα

[1]. Tamara Radu: Euromat ’98, Lisboa, vol. 1, p. 501 – 509. [2]. Balint l., Drugescu E., Potecaşu F., Balint , Temperature influence over electrodeposed layers, 2002 Istanbul [3]. L. Balint, E. Drugescu s.a. –„Temperature influence on electrodeposited layers”-11th Interantional Metallurgy and Materials Congress, Istanbul 2002, p.18. 32

32

a

K66,0K86,0075,0

JJdtd

−=−

=⎟⎠⎞

⎜⎝⎛ α′

[4]. L. Balint, O. Mitoseriu, T.Radu, s.a –« Research Regarding the Obtaining of Hard Magnetic Materials Formed from Composite Material Layer in Co Matrix”-The 11th Romanian International Conference on Chemistry and Chemical Engineering, 1999, Bucuresti. [5]. Radu,L. Balint s.a – „Cinetica procesului de difuzie la acoperirea cu zinc transformat in aliaj Fe-Zn ”-Analele Universitatii “ Dunarea de jos” Galati, fascicula IX, p. 16, ISSN 1453 – 083X, 2001.

( )21

21

a

FeCo K82,0K41,4615,0

JJdt,d

−=−

=⎥⎦⎤

⎢⎣⎡ αα

83



84

IMPROVING CONSTRUCTIVE AND EXPLOITATION PARAMETERS OF CAST TURBO BLOWER ROTORS AT ISPAT-SIDEX GALATI BY

STUDYING THEIR DYNAMIC BEHAVIOUR

Marian BORDEI,Stefan DRAGOMIR, Marian NEACSU, Beatrice TUDOR

“Dunarea de Jos “University of Galati,


ABSTRACT

The turbo blowers integrated in the conglobate sintering systems at ISPAT-SIDEX GALATI have to reach a high performance level (steady rotation speed of the rotor, appropriate pressure and speed of the air in installation under the circumstances of its laminar flow and others.) It is necessary that turbo blower rotor dynamic behaviour should be studied in order to ascertain the critical revolution values, the amplitudes of critical revolution areas and of normal working revolution , possibilities of decreasing these vibration-caused amplitudes by means of an appropriate rotor balancing which should take into account their flexibility during working.

Keywords: turbo blower rotor, vibration level, oscillation

1. Introduction

In order to function at optimal parameters, turbo blower rotors need using some effective dynamic balancing methods in their bearings, checking their dynamic behaviour at different rotation speed rates, a checking and an appropriate finishing of the geometrical configuration in the 3D system.

During its exploitation the rotor undergoes unbalancing due to different causes (blade corrosion, deterioration of some constituents etc.). In order to avoid serious accidents with considerable damages , a permanent controlling of the vibration level of the rotor is necessary during its working. When this level goes over the admissible limits, the respective aggregate must be turned off and the rotor must be checked and rebalanced. An effective dynamic rotor balancing methodology in its bearings succeeds in reducing the non-functioning period of an installation, (the Agglomerating exhausters) , followed by the increasing of their economic efficiency.

All the above mentioned aspects can be solved with satisfying results by simulating the rotor

dynamic behaviour and by studying some dynamic rotor balancing methods in their own bearings.

In order to analyse the dynamic behaviour of the rotor, it will be represented under the form of some concentrated masses connected by bearing sections having a constant crossing section (fig. 1). The mass sections are chosen so that they may coincide with any of the concentrated masses of the rotor, with the position of the bearings and with the both endings of the rotor. Each section will be given a mass (m), a crossing inertia moment (IT), a polar inertia moment (IP), and an unbalancing one.

For each constant crossing section we know : the inertia moment (I), the surface (A), the shape factor for deformation (α), the characteristics of the bearing material ( the longitude coefficient of elasticity E, the crossing compression modulus G, the mass density ρ of the material that the rotor was made of , etc.).

By applying a force and a balancing moment , the cutting forces and the curving moments on both sides of the section can be expressed function of the rotor amplitude and of its leaning near the section, respectively (x, y, θ, ϕ ) by the matrix relation




85

( )100

0000

0000

2

2

2

2

22

22

'

'

,

'

⎪⎪⎭

⎪⎪⎬

⎫

⎪⎪⎩

⎪⎪⎨

⎧

−

+

⎪⎪

⎭

⎪⎪

⎬

⎫

⎪⎪

⎩

⎪⎪

⎨

⎧

⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢

⎣

⎡

−−−−

−−−

+

⎪⎪

⎭

⎪⎪

⎬

⎫

⎪⎪

⎩

⎪⎪

⎨

⎧

=

⎪⎪

⎭

⎪⎪

⎬

⎫

⎪⎪

⎩

⎪⎪

⎨

⎧

uiu

y

x

ZmZZZm

IIIiI

VVMM

V

V

M

M

yyyx

xyxx

Tp

pT

y

x

y

x

y

x

y

x

ω

ωϕθ

ωω

ωωωω

or under a concentrated form : {Y`}={Y}+[A]{x}+{B}(2)

a)

Section n-1 Section n Section n+1

b) Fig.1. Torque representation and stress in the sections of rottor

In order to avoid a further complication of

the analysis , the hydrodynamic forces of the lubricant film in the bearing are considered under linear form (1,2). If the amplitudes of the centre of the neck are X and Y measured from the static balance position , and the corresponding dynamic forces are Fx and Fy , the first term of a developing of these forces in the Taylor series can be written :

••

••

−−−−=

−−−−=

yByKxBxKF

yByKxBxKF

yyyyyxyxy

xyxyxxxxx (3)

The rigidity coefficients Kx , Kxy , Kyx , Kyy and the damping coefficientsBxx , BBxy , Bvx , B BBvy can be directly determined from the partial derivatives of Fx , Fy evaluated around the balance position of the neck corresponding to a certain speed of the rotor. However, they can also be obtained by solving the lubrication equation (Reynolds) for the particular configuration of the respective bearing (5,6). The dynamic coefficients of the bearings can be determined by experiment , too (8,9).



86

n=600rot/minn=0 before equilibration after equilibration

0

Sect.1. Inferior endSect.3. Axial-radial bearingSect.10. Middle of the ventilatorSect.15.Superior end

Amplitude

[ ]mmmm dezequilibrum in sect 0

Fig.2 Simulation of beam (rotor) oscillation before

and after equilibration Fig.3. Inregistration of vibrations diagrams for rotor

If the bearing elasticity is also considered , it

can be taken into account by changing the forces in the bearings (2):

( ) ( ) ( ) ( )( ) ( ) ( ) ( pyypyxpyypyxy

pxypxxpxypxxx

yyBxxByyKxxKF

yyBxxByyKxxKF

−−−−−−−−= )−−−−−−−−=

(4)

which leads to the changing of some terms in the matrix [A]. The balance of a length unit in a bearing section can be written taking into account the bending moment , the cutting forces , and the rotor dip and shifting as it follows :

yy

pr

xz

pr

Vz

Me

Vz

Me

−∂

∂=ℑω−ϕℑ

−∂∂

=ϕℑω+ℑ

&&&

&&&&

(5)

By separating the shiftings and dips from the

bending moments and from the cutting forces we can get a pair of matrix equations that intermediate crossing from one sector of the rotor bending section to another. These equations can be condensed under the form : { } [ ] { } [ ] { }n'

nnn1n YFXEX +=+ (6) for shiftings and dips and : { } [ ] { } [ ] { }n'

nnn1n YQXGY +=+ (7)

for bending moments and cutting forces , [E] , [F] , [G] and [Q] being matrixes whose terms comprise the geometrical and resistance features of the analysed section.

Equation (2) together with (6) and (7) represent a set of recurrence equations by means of which the rotor amplitudes at each section can be calculated. There can be established a calculation algorithm and a programme for calculating the dynamic response of a rotor [1,2,4] which allows knowing the dynamic amplitudes of the centres of the concentrated mass sections , the position of the geometrical axis of the rotor during its rotation and the critical revolutions in a certain functioning area , calculated either taking into account only rigidity or both rigidity and bearing damping.

Conclusions

One of the examples of applying the results

of the dynamic study of a rotor is the analysis made on the exhauster rotor at Aglomerare 2-3.



87

A first conclusion derived from the effected study was the influence of the radial rigidity of the two bearings that support the rotor. As one can see in table 1, there resulted a significant influence of the dynamic characteristics of the radial bearing on the vibration mode as compared to the much lesser influence on the ball bearing. Besides, the most disadvantageous situation would be around the values closed to the designing conditions (106 daN/m for the bearing and for the ball bearing).

From the study of the vibration amplitudes in different calculation sections for the situation corresponding to the designing data , an example is given in fig. 3 concerning the values obtained for the maximum amplitudes in the four calculation sections according to the rotor revolution ; as a conclusion, the critical revolution speed in this case being around the value of 300 rotation/minute.

Another problem was establishing the balancing methodology on the basis of the remnant

unbalancing influence in different sections on the vibration amplitude along the rotor. As a conclusion concerning this study (presented in fig. 4) it is the obvious influence of balancing in the calculation section (1) on the rotor dynamic behaviour at a revolution of 600 rotation / minute.

From the effected analysis there resulted a series of data concerning the construction of the radial-axial bearing in the inferior part of the rotor in order to achieve the desired dynamic coefficients as well as the balancing methodology so that the non-functioning time can be reduced as much as possible.

At the same time , from the conclusions of the dynamic response study of this rotor there also resulted some possibilities for a substantial improving of its behaviour by modifying the rotor building redistributing the masses along the rotor in an improved variant – solution that can be considered in the case of redesigning it.

Table 1

Rigid radial bearing daN/m

Idem Ball bearing

104

105

106

107

108

104

973

983

489 671 1103

723

750 105 128

975

984

493 676 1104

723

753

106 329

988

157 339

997

286 527 711 1111

296

756

298

781

107 501

1016

173 508

1024

375 621 769 1125

428

821

432

843

108 536

1026

173 571

1033

385 650 784 1130

450

843

456

864

References [1] J.W.Lund şi F.K.Orcutt : Calculation and Experiments on the Unbalance Response of a Flexible Rotor, Journal of Engineering for Industry, noiembrie 1997

[2] Cercetări privind simularea pe calculatoare digitale a funcţionării lagarelor şi ansamblului rotor-lagăre, Faza 1 - Program pentru calculul răspunsului dinamic al unui rotor- Memoriu INCREST 1973 cod IX-2-20, [3] Fl.Dimofte şi A.Bănescu, Echilibrarea dinamică a maşinilor - Sinteză documentară - IDT-1992. [4] V.N. Constantinescu şi alţii - Lagăre cu alunecare. Editura tehnică, Bucureşti 1980 [5] J.W.Lund - Evaluation of stiffness and damping coefficients for fluid film beavings - Shock and vibration digest 1998



INFLUENCE OF THE COCSOCHEMICS SUBPRODUCTS FOR CORROSION IN EXCHANGES HEATS

Viorel DRĂGAN, Ciubotariu ALINA, Ioan AXINTE

University “Dunărea de Jos “ Galaţi Email: [email protected]

ABSTRACT

This paper presents corrosive process in heat- exchanges in coke chemistry.

Corrosive process inside the heat exchangers goes on two plans : chemical plan - electricity non conductible liquids and organic substances being in benzenat and debenzenat oil, electrochemical plan -because of salts and gasses from existent water dissolve in to benzenat oil, creating electrolyte solutions. We proposed to appreciate degree of corrosion with measure loss in weigh and assign the corrosion speed. Corrosion effect takes three plans: uniform(that caused an 0,3 mm increasing of the inner diameter of the pipes), local(corrosion started that creates craters and increases compositions heterogeneous. Thickness of the pipes wall is between 0,7-2,0 mm), intrecrystalline (corrosion shows clearly by fissure propagation around the crystals, that can be crossed and may break the material).

Keywords: benzene oil, heat-exchanges, metallographic analyses, corrosive speed, non-metallic

inclusions, uniform corrosion, local corrosion, intrecrystalline corrosion.

1.Introduction Metallic corrosion occurs when metal atoms are oxidized and subsequently leave the metal lattice as ions. Valence electrons associated with metal ions (previously atoms) are left behind in the metal, creating an excess of electrons at the metal surface. The oxidation of metal atoms to ions is referred to as an electrochemical reaction because it is a chemical reaction that involves generation and transfer of electrons to electrochemically active species (dissolved) in the electrolyte. The transfer of electrons enables electronic measurement and study of metallic corrosion. The oxidation half reaction of the metal is referred to the anodic reaction and areas on a metal surface where oxidation occurs are referred to as cathodes reaction and areas where reduction occurs are referred to as cathodes. Anodes and cathodes can have atomic dimension or may be large enough to be observed with the unaided eye. Anodes and cathodes can be separated by finite distances as long as negative and positive ions can move in the electrolyte toward the anodes and cathodes, respectively, to maintain the electrical charge neutrality of the metal and electrolyte. Both anodic and catholic reactions must be present to initiate and sustain metallic corrosion. Electrochemical corrosion reaction equations contain symbols for electrons, reacting elements

(and/or ions and molecules ) and ions or molecules produced by the corrosion reactions. For example, the corrosion of iron is represented by the following anodic electrochem equation: →ical

n

Fe0 Fe+2 + 2e- [1.1] Fe0 represents the ionic atoms at the metal surface, Fe+2 represents iron ions and 2e- represents the two electrons produced by the anodic reactions. Equation 1.1 is referred to as an anodic half reaction of these electrons with an electrochemically active species is missing. It was previously started that both an anodic and cathodic reaction must be present to initiate and sustain metallic corrosion. Consequently, a cathodic half reaction must be written to account for reaction (reduction) of some electrochemically active species with excess electrons. An example of a cathodic half reaction is the reduction of hydrogen ions:

2H+ +2e- H→ 2 [1.2] The overall corrosion reaction is written as the summatio of both half reactions:

Fe0→ Fe+2 + 2e- (anodic) [1.1] 2H+ +2e- →H2 (cathodic) [1.2]

___________________________

Fe0 + 2H+ Fe→ +2 + H2 (overall) [1.3] Equation 1.3 reads: iron atoms are oxidized to iron ions; producing electrons that reduce hydrogen

88



ions (originating from the electrolyte) at the metal surface, forming hydrogen molecules. Examination of the overall corrosion equation 1.3, and the corresponding anodic and cathodic equations (equations 1.1 and 1.2), leads to the hypothesis that corrosion can be halted by preventing one of the half reactions from occurring, and/or removing electrochemically active species from the electrolyte. Unfortunately it is difficult (but not impossible) to prevent half reactions from occurring, and there is a long list of electrochemically active species that cause corrosion. Water, for example, is electrochemically active and can cause corrosion of many commercially important metals such as mild steel and commercially pure aluminum. A partial list of other electrochemically active species included oxygen (O2), carbon dioxide(CO2) dissolved in water, inorganic acids such as hydrochloric acid(HCl), and hydrogen sulfide(H2S), and strong organic acids.

The coke chemistry industry is one of the biggest consumer of metal. Because the big aggressively, the atmosphere and the technological environment by the coke chemistry industry , corrosive speed of the steel carbon is 1 mm/year , but is some technological environment, chromium - nickel steel are corrode with speed the most big 1,5-2,0 mm/year. Depending on concrete condition by work is necessary to detect the corrosive process which affecting installation and equipment which work in different corrosive environment and measured the

corrosive speed into consideration elaborate the optimal method by protection. In one section of the coke chemistry plant in which the corrosive process are very intensely is chemistry plant.

2. Experimental researches The agent corrosive benzenat oil used in laboratory was take by specialist of the chemistry plant on distilled installation of benzenat oil. The test was cut on tubes in the shape of rings with 36,8 mm in exterior diameter, 15 mm width and 8 mm thickness. Chose the shape of rings, not door plate to avoid tension and fissure phenomenon’s of the material processed. After polished and washed in solvents, the tests was dry, then weighing with a precision by 0,1 mg. The tests was immersed in benzenat oil and preserved at temperature by 1300 C below agitate continue of the liquid. Volume of the vessels is 600 cm 3 . After ever tried oil was be substitute. Criterion to appreciate degree of corrosion adopted loss in weight. In this sense after periods of time , good determined the tests was take out, wash, dry and reweigh. Was be calculated loss in weigh and was be assign the corrosive speed of graph.

Characteristic composition for benzenat oil was determinate with chromatographic analyses (table 1.

Table 1: Chromatographic analyses of absorption oil

Nr. Determination New absorption

oil

Benzenat absorption oil

Debenzenat absorption oil

1. ∑ X 1,98 2,05 1,36

2. Naphthalene 22,72 15,08 13,59 3. ∑ X 1,11 1,43 1,20

4. β 2-methylnaphtalene 13,65 9,00 8,65

5. α 1-methylnaphtalene 67,48 5,43 5,67 6. ∑ X 5,66 7,13 7,78

7. 2,6-dimethylnaphtalene 2,61 2,58 2,80 8. 1,3-dimethylnaphtalene 3,62 4,14 4,77 9. 1,5+2,3-dimethylnaphtalene 1,09 0,74 1,28

10. 1,2-dimethylnaphtalene 0,74 0,24 0,53 11. Acenaphtene 13,96 19,88 18,38 12. ∑ X 0,49 1,07 0,60

13. Diphenilen oxide 7,37 10,39 10,67 14. ∑ X 1,17 0,73 1,59

15. Fluorene 7,48 9,08 8,45

89



16. ∑ X 4,90 4,61 5,69

17. Anthracene+phenantrene 4,44 5,71 6,11

18. ∑ X 0,28 0,71 0,88

19. 2-methylantracene 0,13 - - 20. ∑ X 0,12 - -

Density: 1,048g/cm3 (200C) ;Body of oil: 1,60 E; Water : 10-16 % The chemistry analyses of used tests for the corrosive study show that steel used at heat exchangers

tubes in characteristic trade mark OLT-35 indicated for tub by general use without soldering.

Core mark Chemical composition

C Mn Si Pmax Smax

OLT-35 0,09-0,16 0,40-0,80 0,17-0,37 0,04 0,045

Test explorer 0,12 0,50 0,19 0,038 0,040

Both for the probes by corrosive tubes take out heat changes and for tests used at the study corrosive in laboratory was be affect metallographic analyses. Set in evidence of the electrochemistry corrosive which produce in produce in heat changes was measured the current produce by microcell formed with water solution separated by absorption oil.

With this end in view taken two electrode cut out heat changes tubes and after they was be wash and clean put in corrosive water solution come from adsorbed oil. The electrode was connected at terminals milliampermeter and measured current of the corrosive cell at 10 minutes interval one hour. pH solution was 6. Represent graphic variation current corrosive cell function time (figure 1). On right line obtain the anode dissolve current for iron in this microcell.

Fig. 1. Variation current corrosive cell function time

On analyze results study of corrosive for steel tests used in heat changes from distillate installation of benzenat oil find that benzenat oil have corrosive action very intense. The causes are present in benzenat oil of the water, H2S, NH3, salt of cianhidric acid, base pyridine

etching trace quantitative and qualitative determinates which was be effectuated found that benzenat oil used by us continue between 10-16 % water; 1,12-1,28 g/l H 2 S; 2,24-2,40 g/l N H 3, compound with nitrogen (CN-, pyridine).Pas this substances on coke gas in adsorbed oil provoke an intense corrosive at the benzene columns and a heat changes. In specialty working not found notes about influence on this compounds about on properties absorb oil. On analysis tubes used in heat changes results that after debenzolation oil contain great quantity from corrosive compound. In figure 2 are reproduce corrosive speed curves for five tests studies in laboratory. On graphics and table 2 observed that corrosive speed is very intense in the first days, showing down in time.

Fig. 2. Variation corrosive speed in benzenat oil

In heat changes corrosive speed can be considered approximately the same thanks to flow which oil is give circulated and permanently contact of corrosive substances with surface of the metal through continue far off the results products.

90



Table 2: Corrosion speed with loss in weigh

Probe

Probes mass

Corrosion speed ( mg/dm2.day)

Time ( days)

1 5 10 15 20 25 30

1. 11,8508 168 123 76,5 59,8 48,0 42 36,8

2. 10,8616 258 112 70,0 53,2 45,5 39,1 35,5 3. 11,5724 220 103 62,4 48,6 40,7 36,0 31,6 4. 10,6125 178 90,0 50,8 41,6 34,5 30,4 26,7

5. 12,0818 161 84,0 47,7 37,3 29,9 26,3 23,5

In heat changes corrosive speed can be

considered approximately the same thanks to flow which oil is give circulated and permanently contact of corrosive substances with surface of the metal through continue far off the results products. Heat-exchangers from benzenat oil distillery plant are corroded in presence of benzenat oil and water benzenat oil , H2S( 1,12-1,28g/l), NH3(2,24-2,40 g/l) and HCN salts ,organic sulphure and nitrogen etc. These substances migrates from coke gas in absorption oil and its created a very corrosive environment at the usual work temperature. A temperature higher than 130 0C accelerates corrosive process, by more intense reactions at the metal surface.

Corrosive process inside the heat exchangers goes on two plans : - chemical plan because of H2S and NH3 gasses, electricity non conductible liquids and organic substances being in benzenat and Debenzenat oil; -electrochemical plan because of salts and gasses from existent water dissolve into benzenat oil, creating electrolyte solutions.

Metallographic analyses heat exchangers on pipe probes revealed :the presence of non-metallic inclusions have a negative effect on mechanical properties. They also affect corrosion resistance(figure3,4).

Fig. 3. Non-metallic inclusion in transverse section from tried probe in laboratory

Fig. 4. Non-metallic inclusion in corroded

probes in laboratory without attack. Longitudinal section.

Corrosion effects take 3 plans :

- uniform corrosion: takes place on the inner side of the pipes as an effect of continuous distant of the reaction products, caused by the high pressure of the benzenat oil. That caused an 0,3 mm increasing of the inner diameter of the pipes (figure 5).

Fig. 5. Transverse section of the pipe. - local corrosion: takes place on the exterior side of the pipes, caused by very close positive areas that generates unstable passivity. They are corrosion started that creates craters and increases compositions heterogeneous. ( fig 6,7,8,9) Thickness of the pipes wall is between 0,7-2,0 mm.

91



Fig. 6. Portion corroded tubes

Fig. 7. Portion of pipe with intense corrosion

- intrecrystalline corrosion: is caused by the anodic behavior of the separators between grain rated to material inside the grain volume, because of supplementary potential energy associate with the crystalline not disorder confine at grain. Intrecrystalline corrosion shows clearly by fissure propagation around the crystals, that can be crossed and may break the material. ( figure8,9,10)

Fig. 8. Micrographic of the probe from installation. Intrecrystalline and local corrode

Fig. 9. Micrographic of the probe corroded in laboratory. Intrecrystalline and local corrode

Fig. 10. Micrographic of the probe tried in laboratory with non uniformity corrosion in transverse section

3.Conclusions The heat exchangers pipe work environment

shows that corrosion process takes place under pressure. This is caused by the oil pressure inside the pipe and the aggressive wet environment : H2S, NH3 etc. , that accelerates diffusion and chemical reaction. X rays study showed changes of the structure on level crystalline net of corrosion. The metal is not homogenous , internal pressure and density changes creates local different potential that generates microcells. About speed of corrosion variable between 23,5 mg/dm2 by day and 258 mg/dm2 by day ; computed in laboratory. It determined by some factors as: metal composition and structure , absorption oil composition , coke gas characteristics and technical parameters of benzenes hydrocarbons absorption. In normal work conditions of the heat exchangers , corrosion speed is a lot faster , increased by the oil flew and continuous removal of the reaction products and permanent contact of the corrosive substances on metallic surface. Corrosion test revealed that the steel used in distillery changers heat exchangers OLT-35 does not fit the extremely corrosive work environment.

92



References [1] .Constantinescu, A – “Detectarea şi măsurarea coroziunii”, Ed. Tehnica, Bucureşti, 1979 [2].Nedelcu, N. – “Protecţii anticorosive în construcţii industriale şi civile”, Ed. Tehnică, Bucureşti, 1986 [3].Marinescu, A., Andoniant,Gh. Bay E. – “Tehnologii electrochimice şi chimice de protecţie a materialelor metalice”, Ed. Tehnica, Bucureşti, 1984 [4].Oniciu, L. – “Coroziunea metalelor”, Ed. St. si enciclopedica, Bucureşti, 1986

[5].Ronald A, Mc. Cauley – “Corrosion of Ceramics”, Marcel Dekker, INC., New York, 1995 [6].S. Zamfir, R. Vidu, V. Brinzoi – “Coroziunea materialelor metalice”, E.D.P.,Bucureşti, 1994 [7]. W.Stephen Tait- “An Introduction to Electrochemical corrosion testing for practicing engineers and scientist” Pair O Docs Professionals L.L.C. , Madison WI,1994 [8]. Viorel Drăgan, Ioan Axinte, Alina Ciubotariu- “Studiul şi cercetarea procesului de coroziune în instalaţia de distilare a uleiului benzenat” , A XII-a Conferinţă Stiinţifică cu participare internaţională- TEHNOMUS XII , Suceava 2003

93



94

PREPARATION POSSIBILITIES AND UTILIZATION OF DIFFERENT COAL TAR PITCH TYPES

Gheorghe CADAR, Gina NĂSTASE

“Dunarea de Jos” University of Galati [email protected]

ABSTRACT

In certain periods, in our coke plant results a coal tar with high content of quinolin insolubles, which makes impossible the manufacturing of coal tar pitch binder for electrodes. In this case, there are necessary experiments to find other possibilities to prepare and use normal coal tar pitch as follows: -coal tar pitch with high softening point, can be used as leaning material in blends with high percentage of weakly coking coals; -coal tar pitch binder with low softening point for roads building and repairing, which is a substitute of petrleum asphalt. The pilot scale experiments show that the coal tar pitch binder with high softening point R&B 125-1500C was obtained in the heat treatament with bubbling air flow of normal coal tar pitch. After preparation, coal tar pitch with high softening point is passed into a granulation tower. Sferical grains are introduced into various coking blend. Thus it resulted an icreasing of mechanical strength of coke, also being possible to increase the proportion of weakly and non- coking coals in blends. Other pilot and industrial scale experiments for obtaining a coal tar pitch binder with law softening point R&B 35-550C, emphasized that this binder can be prepared by heat treatament of a normal coal tar pitch mixture with different oil types and waste. This binder was used at county road building and repairing with good results, succesfully replacing petrqleum asphalts.

Keywords: coal tar pitch, coking blend, binder.

1. Introduction

Coking charges are made, in general, by some main components as: weakly coking, plastic coking, degreasing. Because of their coking ability, the plastic-coking components offer the possibility of getting more percents of weakly coking coals in the charges. Concerning the diminishing of hard coal reserves and, not least, their price, the addition of coal or oil-derivative pitch instead of hard coals, has proven to be an effective way of improving the carbonisatin properties of weakly coking coal blends, into a coke suitable for blast furnaces. In this respect, in our country because of the normal coal tar pitch surplus some laboratory and pilot scale investigations have been carried out, in order to obtain a high softening point pitch. Coal tar pitch can be used also, to the preparation as binder for roads building and repairing. In this case, usually the petroleum asphalt is used, which

can be substituted by a binder. This binder can be prepared from coal tar pitch and different oil types and waste.

2. Experimental and results

2.1. Coal tar pitch with high softening point

In the experiments was established the

influence of the initial coal tar pitch quality and processing parameters on the high softening point pitch. The pilot scale experiments have been carried out in a 20 kg. reactor equipped with: electric heater, mechanic mixer and air blower.

As raw material, were used: -normal coal tar pitch with: softening point 900C (R&B); quinolin insoluble (QI) 10.25%; toluol insoluble (TI) 24.05%; coke yield (Cfix) 45,5%; ash (Adb) 0,35%;



95

-waste pitch used as binder with: R&B 1000C; QI 14.90%; TI 39.42%; Cfix 52.07%.

Some experiments in the following processing parameters were done: final temperature 3500C; soaking time minimum 60 min.; variable airflow 100-500 l/kg.h. A coal tar pitch with a high softening point, which gave better results taking into account its caking ability in the coking charge, has been obtained with the characteristics: R&B 125-1500C; QI 14.50-20,5 %; TI 35-50 %; Adb 0.40-0.75 %; Cfix 55.25- 65,5 %.

After preparation, coal tar pitch with high softening point passed into a granulation tower,

which have a centrifugal granulation thaler and the possibility to cool the grains with air. Experiments

were carried out on pilot level, figure 1. The liquid coal tar pitch, which entered in the centrifugal distributor, is

dispersed as drops. The drops of coal tar pitch are solidified as spherical grains.

The process is flexible, showing the possibility to obtain grains of various sizes, depending on necessities. In the case of coal tar pitch for using in the coking charges, has been obtained the spherical grains with: Ømax 1mm, bulk density 750-800 kg/m3; slope angle 32-35 degrees. The obtained grains can be send to the customers by: pneumatically transport, in containers, in poliethylen bags, etc. During prolonged storage, in order to avoid the adherence of the pitch grains, they are coated, during granulation, which different powders.

Fig. 1. The pilot installation centrifugal granulating of coal tar pitch with high softening point.

1. Pipe for liquid coal tar pitch input; 2. Automatic input blocking device; 3. Tap; 4. Input funnel; 5.

Multistage centrifugal distributor; 6. Granulating tower; 7 & 8. Circular opening for cold air. 9 & 10. Fans; 11. Aperture for wast gas exhausting;12 & 14. Pipes for gas; 13. Scrubber; 15 & 16. Tapes; 17. Pump; 18. Waste vessel; 19. Level index; 20. Aperture for grains removal; 21. Filling sacks stand; 22. Polyethylene bag; 23. Running belt; 24. Vibrators.

A proportion of 2-5% granulated coal tar pitch with high softening point was introduced in different coking charges using a 0.35 m3 oven. The results show: -the increasing of mechanical strength M40 with 1-2%;

-the possibility to increase the proportion of weakly and non-coking coals with 2-5 %; -the diminishing of prime coking coals with 2-5%.



96

2.2. Coal tar pitch with low softening point

Other direction of the experiments concerning the utilization coal tar pitch with high content of QI,

was the preparation a coal tar pitch binder with low softening point for roads building and repairing, as substitute of petroleum asphalt, figure 2

Fig. 2. The pilot installation for preparation a coal tar pitch binder with low softening point.

1. Reactor; 2 –7. Reservoirs for row materials; 8 & 9. Neutral acid coal tar waste;

10. Mesurement quantity; 11. Scrubber; 12. Reservoir for coal tar pitch binder with low softening point.

The obtaining of a coal tar pitch with low softening point was experimented at pilot and industrial level. As row materials were used: 43-67% normal coal tar pitch; 21-25% anthracene fraction 2; 5-10% absorption fraction; 0-10% creosote; 0-2% light fraction; 0-10%pitch from acid coal tar waste. The characteristics of normal coal tar pitch was: R&B 70-800C; QI 10-14%; TI 20-25%; Adb 0.3-0.35%.

The processing parameters for the heat treatment, of a normal coal tar pitch mixture with different oil types and waste, were used at a temperature of minimum 1000C with a continuos mixing of minimum 120min.

The main characteristics of the coal tar pitch binder obtained by experiments can be compared with those of petroleum asphalt: R&B 35- 550C; penetration 33-85 (1/10mm at 25oC); ductibility 92-100mm; Frass breaking point (-15) - (-200C).

The binder was used experimentally at county roads and repairing with good results, successfully replacing petroleum asphalt.

3. Conclusions

3.1. Obtaining of coal tar pitch with high softening point for utilization and rising the coking ability of coking charges is possible in the following conditions: final temperature 3500C; air flow 100-500 l/kg.h.; soaking time minimum 60 min., having the main characteristics: R&B 125-150 0C; QI 14,5-20,5 %; TI 35-50 %; C fix 55-65 %; Adb 0,3-0,8 %; Yield 85-95 %.

3.2. High softening point coal tar pitch can be granulated to 1 mm maximum diameter by centrifugal pulverization and air cooling.

3.3. High softening point granulated coal tar pitch can be used as a coking charge component involving the rising of weakly and non-coking coals and the incresing of mechanical strength.

3.4. Obtaining of coal tar pitch with low softening point as binder for roads building and repairing, is possible under following conditions: temperature minimum 100 0C; continuous mixing time minimum 120 min., having the main



97

characteristics: R&B 35-55 0C; penetration 33-85 (1/10 mm la 25 0C); ductibility 92-100 mm; Frass breaking point (-15) – (-20) 0C. The binder was used experimentally at county road and repairing with good results, succesfully replacing petroleum asphalts.

4. Symbols R&B – softening point Ring and Ball; QI – quinoline insolubles; TI – toluol insolubles;

Cfix – fixed carbon; A – ash.

References

[1]R. Sakurova, L. I. Lynch, Fuel, nr.6/1993, p. 743; [2]B. Bujnowska, G. Collin, 2th International Cokemaking Congress, London, 1992, vol. 2, p. 93; [3]S. A. Aipshtein, 9th International Conference on Coal Science, Essen, 1997, vol. 2, p. 813; [4]V. Prada, M. Grand. 9th International Conference on Coal Science, Essen, 1997, vol. 2, p. 893;

THE ANNALS OF “DUNAREA DE JOS” UNIVERSITY OF GALATI FASCICLE IX METALLURGY AND MATERIALS SCIENCE, ISSN 1453 – 083X VOL II – 2003

98

CRUSHED COKE CONTENT IN THE SINTERING MACHINE CHARGE IS CONSIDERED BASED ON THE

CORRELATION EXISTING BETWEEN IT AND THE QUANTITY OF RECYCLED MATERIAL

Adrian VASILIU, Marian BORDEI, Alexandru CHIRIAC

Email: [email protected]

Abstracts

Crushed coke content in the sintering machine charge is considered based on the correlation existing between it and the quantity of recycled material the regression line, expressing the interrelationship between the two parameters is given as well as the mathematical model, the calculation algorithm and the calculation procedure of recycled material. Keywords: crushed coke,sintering machine, recycled material

1 Introduction

Optimizing the amount of coke is realized by regulating the amount of return produced at the optimal reference value (ropt), by changing load's amount of carbon. From “ropt” a technological point of view, value corresponds, on the one hand, to the optimal compromise between the specific coke consumption and agglomerate production, and the strengh and reductability of agglomerate on the other hand. Because nor the strengh and reductability of agglomerate are constantly determined, we're going to limit at determining the optimal amount of return only for regulating the specific coke consumption of the load. The real amount of return resulted from the agglomerate process is not weighed but determined by using the balance sheet of materials. Thus, a constant amount of return is obtained, corresponding to a given agglomerate production, and for a statistical working, it is reported to an amount of homogenized ore in the load (Qm), resulting the adimensional ratio: r = Qrp/Qm (1) Balancing the circuit of return is one of the basic condition for assuring a stable stationary service of equipment. By realising the equality:

Qrp=Qropt [t/day] (2)

The minimal value of produced return amount is conditioned. By stabilizing the phisical and chemical characteristics of the load, keeping equaluty (2) is possible because of operating upon Qrp factor by changing the coke consumption of the load. Considering that ratios are regulated themselves in the load: r= Qr/Qm and c= Cs/Qm (3) we can see that the refference value is established: r0 = (Qr/Qm)opt (4) and the amount of return in load is calculated: Qropt=r0 × Qm [t/day] (5) The dependence between the amount of carbon and return in load is given by this tipe of linear equation: c= a-b×r where the constans a and b are determined by statistical analysis used for each equipment. Thus, the


99

calculation relation of the carbon content value of the load results for equilibrating the retun circulation: c=c0 - b(r0 - r) (6) or the necessary corection Δc= -b(r0 - r) (7) where: c0 - the corresponding value of r0. c c1 co c2 -Δr +Δr r2 r0 r1 r Fig 1. Determinatoin of carbon content correction of the load at the amount change of produced return Unbalancing the stationary operation in which r≠r0 is caused either by aleator change under the action of disturbings, the amount of produced return, or the change of homogeneous ore making in the load, imposed by its chemical characteristics (fig.1). The relation (7) is the second equation of the model for coke consumption reducing by reload the return.

2. A software for return

recirculation at the agglomerate band

100 cls 'Model for return-band recirculation Print Print Print ''CALCULATION FOR RETURN RECIRCULATION'' Print Print Input ''Name the number of piles on which the application is based'' For i=1 to ns

Input ''Circulated ore amount'';QmInput ''Circulated return amount'';QrInput ''Statistical constant a'';a

Input ''Statistical constant b'';b Input ''Active level of coke in charge'';co

'make the calculations ro=Qr/Qm c=a-b*ro Qopt=ro*Qm Co=c+b*(ro-r) Δc=-b*(ro-r) 'display data Print Print ''Optimal amount charge Qopt='';Qopt; ''t/day Print ''Necessary coke in charge co='';co; ''t/day'' Print ''Difference coke in charge Δc='';Δc; ''t/day'' End ***Observation: This software proposal was made for checking up the model on each computer. Regarding the industrial applications, this software is written in C++, WinSis If In Touch etc. depending on the system structure used in the real application.

3.Particulars of the model

Analysing the correlation coefficient value in dependence c=f(r) for the three piles, we can see that the best value (0.5411) is obtained as a result of statistical process of the second pile (table 1), and thus the parameters of this pile will be used in checking this model. Table 1

Pile II Coefficients of Dependence

Corel Determ. (%) Qab=f(qk) 0.7373 54.3552 Qab=f(r) 0.9181 46.6576

Vm=f(Qab) 0.9736 94.7890 c=f(r) 0.5411 29.2780

Qm=f(r) 0.7941 63.0602 c2=f(c1) 0.9301 86.5090

Constants a and b have the following

statistical medium values:

a=0.04614 and b=0.06206 and the equation (6) becomes: c=0.04614 – 0.06206 × r The adjust line of the carbon content of the same pile results as the one in the figure 2 (Y11), point 1 (coordinates:0.236; 0.061) representing the optimal functioning point.

The correction equation becomes:


100

Δc=-0.06206(r0 - r) where: r0=ropt=0.236

Fig 2. The regression lines of the carbon content in the load

Value of ''r'' results from the real ratio Qr/Qm, calculated with relation (3): Qr/Qm= 1437.38/3620.44=0.397

So the real point of functioning is point 1 which is found onY11. The correction coefficient of combustible ''Cc'' becomes:

C=0.061 + 0.01=0.071 The real functioning line of the band Y11 is

a parallel line placed under the regression line Y11. Maintaining the return in optimal limits, the

correction calculation of combustible Δc was made and point 2 (coordinations: 0.0236; 0.071) was

obtained on the Y11 line, above the regression line and parallel to this one.

In this situation, the corrected amount of coke becomes: Qkc=cc×Qm/0,89= 0,071×3620,44/0,89=288,7(t/day)

4.Conclusions: Regarding the second pile, the return

regulation at the optimal value based on the carbon correction leaded to an increasing of the amount of combustible from 277,36 to 288,72 t/day, which means with 11,3 t/day, and to the real value of 250,7 t/day, with 37,9 t/day.

From an economic point of view, we can see an upward continuous evolution of indices as an effect of the carbon content regulation in the load which leads to an increased economic efficiency of the return amount.

References

[1]Vasiliu A. -PhD “Modelarea matematică a procesului de aglomerare în vederea conducerii complexe a acestora.1998, Bucuresti [2]Biclineru N.: PhD “Studii privind interdependenţa dintre factorii care influenţează procesul de aglomerare a minereurilor de fier în vederea optimizării acestuia”, 1981 [3]Vasiliu A.- Studii şi cercetări privind influenţa bazicităţii aglomeratului asupra proprietăţilor fizicio-chimice ale acestora, contract “SIDEX” S.A., 1993 [4]Boucrat M., Rochas R. – L’agglomeration sur grille des minerais de fer. Revue de Metallurgie, nr.12, 1994, p.835 – 844


RECYCLING ALUMINIUM WASTE

Maria Vlad, Varga Bela, Emil Stratulat


ABSTRACT

The paper analyses the efficiency of aluminium recycling from residues resulting from the production process of aluminium powder, a product designated to the manufacturing of auxiliary materials used in foundries. We consider that in some circumstances the recycling of residues resulted during manufacturing processes is bound to take place on their formation site, thus with less performed technologies. In the paper we present the efficiency of the recycling obtained by the re-melting of these residues (reject powder, slag, leaks) as compared to that obtained by the re-melting of secondary alloy ingots and residues of old metals with big specific surfaces. The paper also presents a theoretic analysis of metal losses due to the oxidation of the load depending on its shape, or characteristic surface. The obtained results are in a good concordance with the similar data which is to be found in the specialty literature.

KEYWORDS: recycling, re-melt, aluminium, waste

I

. Introduction

The present day trends to create on antientropic society include material recover and recycle as antientropic processes. The search for antientropic effects has become a vital demand of the future society. Waste recovery is acting to diminish environmental entropy increment. Metallic materials are very important among recyclable materials. Ore resources available and exploited today under specific economic and technological conditions are not enough on long term. As for tungsten, tin, mercury, platinum and silver the situation is critical. [1]. For well known reasons, aluminim is a major priority. At the time, the grade of casting secondary aluminium alloys was not satisfactory, primarily because primary aluminium alloys were preferred and available and secondly due to low level requirements of customers, mainly foundries.

After 1990, due to political and economic changes in Romania society, the output of casting secondary aluminium alloys has increased, beyond any control. The increase is due to limited liability companies, which, in hope for nice quick profits with no big investments have come to smelt any silvery metal. This situation gave way to several aluminium smelteries, manufacturing low grade secondary aluminium ingots.

By compulsory scrap metal collection, large amounts of unproper alloys, considering their wear degree have been smelted. Large amounts of “scrap metal “ such as plastic aluminium-base alloys were wasted to manufacture secondary silumin ingots for shaped parts castings. Failure in secondary silumin ingots use is due to the following four reason, least to be mentioned:

- the traditional foundries producing shaped parts used to work with primary alloy ingots, involving consistent and scientific- based refining technologies (degasification and solid non-metallic inclusions removal);

- absence of quantitative and qualitative control methods of gasification degree and solid non-metallic inclusions’ content;

- secondary alloy ingots are also structurally deficient due to lack of melting units fit for silicon alloying in foundries;

- absence of proper laws, first of all of a set of standards and absence of a research program in the field and of a research center, as well.

The structural problems occurring from inadequate silicon alloying technologies have revealed the importance of inherited structure phenomenal and the connections metallic charge - smelting - castings. Consequently, inadequate control of connections between collection, preparation and smelting of scrap

101


metal and aluminium - based waste is a major obstacle in octaining high grade secondary aluminium alloys. New investments would make possible adequate technologies to manufacture casting aluminium alloys from recovered materials. Meanwhile secondary aluminium alloy manufacturing industry continuously expanded worldwide.

2. Analysis of recovery performance The data obtained during lab tests on aluminium recovery performance are given further. It was used waste from aluminium powder manufacturing process. Such aluminium powders are used to produce

accessory materials (exothermic dusts) for foundries. In the re-melting process of highly oxidized and large specific surface material, recovery performance is very important with a view to economic efficiency. In theory, recovery performance mainly depends on the specific surface of the material that is to be re-melted. The variation of metal-air contact specific surface as a function of metallic charge grain-size is shown in Figure 1. It was assumed that the charge contains spherical grains. Consequently, real solid charges of crushed recovered materials shall have higher specific surface values than those appearing in Figure 1 diagram

.

Fig.1. Contact specific surface (Ssp) variation as a function

of charge grain-size (dg) when γMe= 2.7 g/cm3.

The metal-air contact surface is the factor

with the highest influence upon oxide amount, occurring as a result of re-melting of crushed recovered materials. Consequently, this factor also affects metal recovery performance of charge.

The influence of material grade on recovery performance was experimentally tested in lab conditions. Re-melting took place in a resistance furnace with graphite crucible of φ150×180 mm, see Fig. 2

Subsequent to mechanical preparation of material, the following grades were obtained: - waste powder (granules) - tapping (furnace bear) - large-size picked slag, consisting in metal pieces

- small-size picked slag (refuse), consisting both in metal pieces and non-metallic parts.

With an aim to extend the area of influence of metal charge grade, particularly that of specific surface, on recovery performance two other types of material were considered secondary ingots, supplied by the specialized company in metal and

Fig.2. Sketch of furnace used to re-melt aluminium

waste 1- furnace body 2- silit bars 3- graphite crucible 4- furnace cover

non-ferrous alloys processing and beer and soft drinks cans. Approximately 1.5 kg of each type of material were used for the charge. Subsequent to charge preparation and crucible preheating at about 800°C, the crucible was loaded and the charge was covered with a layer of flux I2. The chemical composition of flux I2 is: 40 ÷ 45% NaCl, 20 ÷ 25% CaF2, 15 ÷ 20% Na2SO4 5 ÷ 10% NaF, 1 ÷ 5% Na2SiF6, 1 ÷ 5% Na3AlF6 and SiO2 for the rest. As the charge had a

102



high specific volume, the crucible had to be progressively loaded. Each charge portion loaded was covered with a new layer of flux. Thus, the flux consumption was about 111÷150g per charge, corresponding to a specific consumption of 6 ÷ 10% .

At the end of the melting process, subsequent to flux mixture with the liquid alloy, the slag was removed using a punched skimming spoon. The resulted ingots were weighed and alloy recovery performance was calculated for each grade of recovered material, based on the data. The recovery performance is given by the relation:

100mm

c

arec ⋅=η (1)

where: ηrec - recovery performance [%] ma - ingot weight [kg]

mc - charge weight [kg]

The examination of slag from re-melting process has revealed metallic aluminium, mechanically conveyed in drops, representing 1% of ingot weight. The quantitative data obtained for the six grades of recovered material are shown in Table 1. The chemical composition of resulted ingots was made. The results of chemical analyses are shown in Table 2

Table 1: Average experimental values of recovery efficiency

Sample No. Grade of material ηrec (%) Notes 1 Waste powder (granules) 91 2 Tapping (furnace bear) 60 3 Large-size picked slag 88 4 Small-size picked slag (refuse) - Freezing of furnace 5 Secondary alloy ingots 95 6 Beer cans 65

Table 2: Chemical composition of secondary ingots

Chemical composition (%) Type alloys SampleNO. Al Si Cu Mn Fe Ti Ni Zn 1 91.69 3.4 2.05 0.12 >1.2 0.039 0.15 1.35 2 91.40 3.5 2.02 0.16 >1.2 0.035 0.19 1.45 3 97.38 2.1 0.05 0.04 0.46 0.027 - -

ATO STAS 201/1

4 96.63 0.1 0.15 0.9 >>1.2 0.020 - - AlMn1 STAS 7680

The chemical composition of sample number 5, consisting in secondary aluminium ingot was not established, as ingot was supplied according to STAS 201/1. In the case of sample number 4 (refuse), subsequent to failed re-melting, it was performed a thorough macroscopic inspection of components. It was revealed a structure consisting mainly in non-metallic materials (slag, flux, fireproof clay), flux-covered silicon grains and a small amount of metallic aluminium. The above structure has generated the furnace freezing phenomenon during re-melting. In the case of sample number 1 of waste powder (granules), recovery performance is surprisingly high, which is explainable by low oxidation ratio of this grade of material, even during its collection and preservation. The large-size picked slag, due to its formation conditions has a higher oxidation ratio and as a consequence, a lower recovery performance value. The tapping has a higher oxidation ratio and, meanwhile, non-metallic inclusions, also due to its

formation conditions, highly decreasing recovery performance value. At this grade, as it is impossible to part the metal from non-metal (pieces of refractory material), the fact has a serious influence on recovery performance values. In the case of beer cans, the low values of recovery performance are due to melting specific conditions, particularly to high specific surface of solid charge. The specific surface of beer cans is given by point 6 in Figure 1 diagram. Certain estimations could be made based on measurements performed during re-melting of beer cans, since the shape and size of beer cans were known. Each beer can was pressed previous to charge loading. The data alloyed estimation of oxidation rate to about 10-2 g/cm2 hour. This value is higher than the data found experimentally, which are (0,5…6)10-5 and shown in diagrams [4,5]. The estimate data also emphasize on oxide layer thickness, reaching values of 30 ÷ 40μm during melting. These values come to support size distribution of solid non-metallic inclusions in aluminium-based melting’s, as a result of re-melting. Meanwhile, the thickness of oxide layer at charge

103



surface during heating process explains the low level of fire lasses (0.5 ÷ 0.7%) for aluminium ingots, in case of “average” grain size /specific surface charges, see point 5 in Figure 1. The experimental tests on influence of charge specific surface and manufacturing method on fire losses were extended out of lab, in industrial conditions. It was used AlMg 2 strip waste 0.27 ÷ 0.39 mm thick.

Re-melting was carried out in two specific conditions. In the first case, gas-fired hearth furnaces of 10t were used. Strip and coil waste was loaded into the melting left after discharge. The remained liquid

alloy, about 3 ÷ 3.5t was about 350 ÷ 400 mm deep. The melting process was carried out at abut 760 ÷ 780°C, under flux layer. The chemical composition of flux is: 33% KCl, 64% NaCl, and 3%(Na3AlF6). The amount of flux was about 2.5% of charge weight. In this case, recovery performance was 66% . In the second case, higher capacity (20t) hearth furnaces were used. Thus, the amount of melting left after discharge was higher, about 10t, was 450-500 mm deep, meaning better conditions for a quick re-melting of high specific surface charge. The experimental data are given in Table 3.

Table 3. Experimental data on foil’s re-melting

Charge NO.

Rest of alloy [t]

Making time

[hour]

Charge weight

[t]

Output [t/hour]

Flux quantity

[kg]

Weight of obtained alloy

[t]

Recovery efficiency

[%] 1 15.0 4.25 5.440 1.28 600 3.868 71.1 2 10.0 5.50 8.360 1.52 210 5.643 67.5 3 10.5 4.75 9.310 1.96 600 6.843 73.5 4 10.5 4.40 8.008 1.82 600 5.766 72.0

3.Conclusions The values obtained of recovery efficiency at

high specific surface waste re-melting in crucible furnaces are quite satisfactory. The data obtained in the two specific conditions of work have revealed the advantages of solid metallic charges with high specific surfaces under liquid alloy layer, in other words the melting principle “ quick and hot” was observed. The data obtained are quite encouraging, considering the literature data, indicating 100% fire losses for AlMg2 strips 0.2 mm thick, maintained for 2 hours without flux or liquid alloy cover.

References

[1]Gabor, D., s.a. Să ieşim din epoca risipei. Bucureşti, Editura Politică, 1993. [2]Varga, B., Geaman, V., Some Aspects About the Aluminium Waste Recycling in Closed Route, in Second International Conference on the Recycling of Metals, 19 ÷ 21 October 1994, Amsterdam, pp 330 ÷ 340. [3]Ersov, G. S., s.a. Vîsokoprocnîe aliuminievîe splavî na osnove vtoricinogo sîria, Moscva, Metallurgia, 1979. [4]Radin, A., Ia. Issledovanie kinetiki okislenia jidkovo aliuminia, în Voprosî tehnologii liteinogo proizvodstva, Moscva, Oborongiz, 1

961.

104



SOME ASPECTS REGARDING THE OPERATING CONDITIONS AND THE TEMPERATURE DETERMINATION FOR THE ROLLS

ON THE CONTACT SURFACE WITH THE ROLLING SLAB, DURING THEIR WORK

Carmen – Penelopi PAPADATU

„Dunarea de Jos” University from Galati – Romania

ABSTRACT

Every day, the rolls of the steel continuous casting machine supports a different stresses, especially due to the high temperature of the steel strand. The temperature is continuously modifying in rolls while their rotation due the unstationary processes of the heat transfer in a three-dimensional plan. It’s important to study the rolls stresses, the field of the temperature inside or outside the rolls, during their functioning.

Keywords: rolls stresses, continuous casting, thermo-mechanic tensions, heat transfer

1. Introduction

In TC-1 station case from ISPAT-SIDEX S.A. Galati, there are many continuous casting machines. Each of them has two casting threads. The segments (the groups) with rolls are the proper elements of directing and maintaining for the slab billet. The thread guide path has two segments with five pairs of rolls (245 mm in diameter) each of them and five segments with four pairs of rolls (310 mm in diameter) each of them. The drawing and straightening stand has five segments with three pairs of rolls –each of them, by means of which the guide path connecting to straightening zone is guaranted. It’s possible to discern: central rolls (430 mm-in diameter), lateral rolls (390 mm-in diameter) and supporting rolls (750 mm-in diameter). The upper rolls are not activated and the extreme rolls too. The active rolls are grouped together –in pairs-on the stand. This stand is supported by four telescope cylinders. The rolls have inside cooling, except for the rolls –245 mm in diameter. Studying the outs of use rolls from thread guiding and the straigtening stand, the result is that the greates part of them are distroyed because of breaking off or bending. It’s important to note the reasons: a). The abnormal stresses; b). The inadequate properties of rolls material; c). The stagnation without cooling on hot thread leads to damage the properties of material; The long contact between rolls and slab leads to overheating at rolls surface and, in the last, the rolls are distroyed.

The thread perforation leads to the following situations: a). The distroying of connections at lubricating system; b). The distroying of the cooling system; c). The overheating of the rolls, at the contact with hot plate slab, when the cooling system was distroyed; d). The damage to the tight rings on the sleeve bearings. Finally, there can be the following results: a). The rolling friction bearings are go into stoppage; b). The overheating at rolls surface – more than Ac1and Ac3, which results in the diminishing of the strenght of materials (σr). Because the non-homogeneous heating, some of the rolls can go through a bending process. The stoppage of the bearings leads to the modification of the friction thread-rolls breed: from rolling friction until pure friction. Thus, it’s necessary to have a greater resistance, than the drawing and straightening stand capacity. In this situation, the thread is blocked.After thread stoppage, the rolls are overheating on the contact zone with slab and undergo deformations. The breacking of the rolls which are not affected by the temperature, are produced at the exit from the curve zone, after the extract operations of the blocked plate slab extremity.

2.Determination of the rolls temperature on the contact surface with steel plate

The present study was realysed for the lateral rolls-390 mm in diameter, from drawing stand. There

105



The strand (slab) speed being petty, the free convection was considered – for lateral surfaces. The TIME parameter was approximated as being the necessary time for the steel strand to go through the lenght of the rolls segment.

The balance sheet has the following components: Qconv.+rad, Qcond, Qrad.

doesn’t exist the secondary cooling systems. The curve section of the casting thread has the secondary cooling system. There was selected 41MoCr11 steel grade for rolls, with sulphur and phosphorus – 0,035% max. Table 1: The chemical composition of the 41MoCr11 steel grade:

Steel grade Chemical composition (%)

C Mn Si Cr Mo Cr 42MoCr11 0,38

-0,45

0,40-0,80

0,17-0,37

0,90-1,30

0,15-0,30

0,90-1,30

For the calculation of temperature of the rolls there was considered a theoretical outline which corresponds to one slab segment at t temperature This segment (L=1350 mm, l=1900 mm,g=300 mm) covers N (N=3) rolls, at τ moment. The thermic stresses are determinated especially by the strand steel temperature, the form of the rolls and material, the rotation speed of the rolls, cooling systems,… The approximations don’t change the real situation. In calculations there were considered all the possibilities of heat:losses: thermal conductivity (λ=22), heat exchange by convection and heat exchange by radiation.

Fig.1: One segment with rolls Qconv.+rad = The heat exchange by convection and radiation; Qcond = The heat exchange by conduction toward the roll Qrad = The heat exchange by radiation toward the roll

2.1. The heat passage by means of conduction from the slab to the rolls

According to the Fourier theory, the heat flow through a surface in time unit is directly proportional with area and the thermal gradient. QCOND.=dq / dτ = λ A (dt /dx) (1), where: (dt /dx) = The thermal gradient; λ = The thermal conductivity; (dt /dx) =lim (Δt / Δx) (2), where: ΔN→0 Δx = The distance between two izotherme surfaces. Δx→0 (There was considered the temperature at the roll surface), thus, (dt /dx) →1. The area of contact between slab and rolls has the relation: A= a l N (3), where: a = The contact width with rolls (a=0,015); l = The slab (strand) width (l=1900 mm); There fore, QCOND. = λ A l N(t2 - t1) (4), where: t1 = the temperature of the superficial layer at the contact with roll; t2 = the temperature of the layer at 1 mm depth; Δt = t2 - t1 (5) The proper heat slab is: Qb = mb Cb(t) Δtb (v /L) (6), Where: mb = the slab amount; Cb(t) = the slab heat capacity (850ºC). Considering the cast steel density (ρ=m/V) (7) and the slab volume

(V = L l g ) (8), → mb = ρ V = ρ L l g (9) Qb = ρ L l g Cb(t) Δtb (v /L) (10) 2.2. The ecuation which describes the heat transport fenomenon by radiation from slab to rolls The heat quantity which was transmitted from slab to roll, by radiation, for one segment of slab, is:

QRAD = 2C12 S L N l [ R1 – R2] (11), where: R1 = (t2m +273,15)4 100-4; R2 = (t1m +273,15)4 100-4. C12 = ε1 ε2 C0 (12), where:

ε1= The emissivity coefficient, for slab;

ε2= The emissivity coefficient, for roll; Co =The radiation coefficient for black body; t1m= The average temperature of the roll;

t2m= The average temperature of the steel strand.

106



2.3. The ecuation which describes the heat transport fenomenon by free heat convection and heat radiation to the environment

The lateral surfaces for the slab, come in contact just with the environment. The slab speed (v), being very small, there can be considered that there exists a fenomenon by free heat convection and radiation between the slab and the environment. The heat quantity which was transmitted by lateral slab surfaces by free heat convection and heat radiation to the environment, is defined by the following relation:

QCONV.+RAD= αt A1 (Ts –Ta) (13), unde: αt = The global coefficient for heat exchange by convection and radiation; A1= The heat exchange surface. At the rolls from drawing stand zone, which has not the secondary cooling system, the surface of heat exchange by convection and radiation increases with the upper surface and the bottom surface at slab segment. αt=1,57(Ts –Ta)1/4L-1/4+m(Ts –Ta)1/4+C(Ts –Ta)-1[R3-R4](14), where: R3= (Ts+ 273,15)4100-4; R4=(Ta+ 273,15)4100-4; m= The coefficient which depends on surface direction; m=2,8 – for horisontal plain surfaces toward up transmition; m=1,8 – for horisontal plain surfaces toward down transmition. In this case, m=(2,8+1,8 0,75) /2 (15), Ts= The slab temperature; Ta= The environment temperature; C = The coefficient for heat exchange by radiation at slab to the environment; A1= 2 L g +4 x l (16). 2.4 Heat calculation Be the relation: ΔQb =V ρ Cp(t) ΔT (17), Where: ΔQb = The heat abstraction quantity by plate slab. Because, the slab speed is v=30 m/h ,then ΔQb =V ρ Cp(t) ΔT (v/L) (18), ΔT= The slab temperature variation/segment; Thus, the energy balance sheet ecuation is: ΔQb = QCONV.+RAD + QRAD+ QCOND (19) On the ground that: QCOND (per unit of roll area) << QRAD (20)

We can find the average temperature of the roll at contact surface with slab emerging from relation (19) In table 2, there are presented the results for this study. Table 2: Features Simbol U.M. Value

Slab form l g m2 1,90x 0,30

Number of rolls/segment

N Buc. 3

Slab lenght/segment

L m 1,35

Rolls step/segment X m 0,450 Slab volume /seg V m3 0,769

Steel density ρ Kg/m3 7850 Slab Temperature t1 ˚C 850

Slab speed v m/h 30 Heat capacity of

plate slab Cb(t) Kcal/kg grd 0,210

Heat abstraction quantity by

slab/segment

Qb Kcal/h 792545,07

Slab temperature variation/ segment

Δt ˚C 28

Contact width with rolls

a m 0,015

The hot surface temp. for roll

t1 ˚C 850

Could surface temp.for roll

t2 ˚C 20

Thermal conductivity

λ Kcal/mhgrd 22

QCOND Kcal/h 1561,23 Coefficient from the stand point of

surface

m 2,075

Heat exchanging by conv.and rad.

surface

A1 m2 4,23

Global coefficient for heat exchange

αt Kcal/m2hgrd 93,36

Average temp. of the environment

ta ˚C 50

Abstracted heat toward the

environment

QCONV.+RAD Kcal/h 315930,24

C Kcal/m2h grd4 4,1 Average

coefficient by iradiation at 1 m

from the roll

Φ12 0,99

Radiation area at slab to roll

S12= Φ12 X m 0,445

Thermal emissivity coeff.of slab

ε1 0,28

Thermal emissivity coeff.of roll

ε2 0,79

C0 Kcal/m2h grd4 4,88 Roll average

temperature on the contact surface

t2m ˚C 376

Slab average temperature

t1m ˚C 860

Thermal capacity for roll

Cmr(t) Kcal/kg grd 0,145

Slab mass/segment mb kg 6040,5 Radiation slab

surface S m2 4,488

107



3.Conclusions The present paper demonstrated that, during the

continuous casting machine operating, there exists continuous heat exchange: rolls ↔ slab.

The superficial layer rolls temperature on the contact surface with slab, is roughly less than 400˚C. This low temperature doesn’t modify the layer mechanical characteristics, in the extent to which, initially, the rolls underwent the thermochemical treatment (Nitriding process).

4.Bibliography

[1]STAHL nr.7/1990 – „Exploatarea rolelor bandajate al masinilor de turnare continua a sleburilor”; [2]STAHL UND EISEN nr.12/1971-„Temperatura si contractiile termice in role si cilindri” [3]ZUBAC,V.-„Utilaje pentru turnatorie”-Editura Didactica si Pedagogica, Bucuresti, 1982.

108



CONSIDERENTE PRIVIND FORMALISMUL MECANO-MATEMATIC

CE INTERVIN ÎN FENOMENE SUPERFICIALE CU APLICAŢII LA

METALE DE PURITATE ATOMICĂ DIN GRUPA 3D

Petre Stelian Niţă “Dunarea de Jos” University,

e-mail: @ugal.ro

1.Introducere

Fie un volum V în care se află N particule aflate la temperatura T cu iq , ip coordonatele şi impulsurile generalizate şi hamiltonianul sistemului ce-l descrie:

( ) ijij

2ij

ii um2p

pqH Σ+Σ= v (1)

şi probabilitatea termodinamică de ocupare a stărilor energetice,

( ) TBKH

1ji epq

−−Ξ=ω (2)

în care:

dVeNN TBK

Hi ⋅∫=Ξ

∞+

∞−

− (3)

este funcţia de partiţie a particulelor pe nivele energetice, iar integrarea se face pe tot spaţiul fazelor unde am considerat elementul de volum dV = d iq ·d ip .

Considerând variabilele separabile adică :

iTBKiju

iTBK

im2/2ip

i qdepdeNN

∫⋅∫=Ξ∞+

∞−

Σ−∞+

∞−

Σ−

(4)

şi rezolvând ecuaţia (4) în cazul sistemelor cu număr variabil de particule [1], obţinem distribuţia macrocanonică (grand partition function Ξ) [2], [3], [4]

( ) ( )[ ] iBE

ijiiiNi TKENTq

i

i /exp∑ −⋅Σ=Ξ μ (5)

în care suma se face pentru toţi atomii Ni de pe nivelele Ei cu qi funcţie de partiţie atomică, μi potenţialul chimic al elementului i şi Eij energia perechii de atomi (i, j). Introducând prin definiţie potenţialul chimic al elementului i în funcţie de concentraţia ci şi activitatea chimică γi, iiB

0ii clnTK γ+μ=μ , numărul de particule

NI

,ln

TKNi

'i

Bi μ∂Ξ∂

= cu (6)

∑=ΞΣ−

iE

TBK/iE'i e , cu

iiiiiii PE εΔ⋅+ε= în care avem energia perechii ii şi probabilitatea de formare Pii. In cadrul modelului laticeal [4], [5], [6], [7], [8], [9]

j,ij,ij,ij,i

νφ⋅ζ=Ξ cu (7)

ζi – funcţia de partiţie atomică adică ( ) TbK/iii eTq μ⋅=ζ şi φ i,j – funcţia de distribuţie după energii în latice în funcţie de numărul de coordinaţie z cu densitatea de ocupare: ( ) TK/Pexp Bijijj,ij,i εΔ⋅+ε−=ρ , funcţiile de partiţie exprimându-se

109



zjj

ziij,i φζ+φζ=Ξ cu raportul

j,i

j

i

j

i

j

iν

⎟⎟⎠

⎞⎜⎜⎝

⎛

φφ

ζζ

=ΞΞ

cu (8)

νi,j – coeficienţii stoichiometrici ai perechii i,j şi considerând în acelaşi model probabilitatea de ocupare Xij a laticei cu atomi i şi j adică )1z(2

ijijijX −φζ= cu proprietatea

1X

XX2ij

jjii =⋅

, şi condiţia de normare (de

ocupare completă a laticei),

1XX2X jjijii =++ (9) Scriem funcţia de partiţie (5) pentru perechea ij, ca raportul

( )( ) TK/ENexpq

TK/ENexpq

Bjjij

Biiii

j

i−μΣ−μΣ

=ΞΞ

(10)

şi folosind (6), (7), (8), (9) în (10) obţinem soluţia ecuaţiei (10) ca raport între coeficienţii de activitate chimică ai perechii (ij),

TK

PlnPWPWlnln

B

ijijijijijij

ijj

iΣ+ΔΣ+∑

=γ=γγ (11)

Introducem prin definiţie energia liberă Gibbs, Gij şi energia liberă de exces Gex cu legătura, ijam

ex GGG −= în care am introdus energia liberă de amestec şi

dclnTKNGc

0iBijij ∫ γ= din (12)

V,P,Tij

ex

ijB NGlnTK ⎟

⎟⎠

⎞⎜⎜⎝

⎛

∂∂

=γ cu N

NC ij

ij =

şi folosind (11) în (12) obţinem energia liberă Gibbs:

⎥⎥⎦

⎤

⎢⎢⎣

⎡+φ

Δ+−= ijijij

B

ijijij

B

ijBijij ClnC

TKW

)C1(CTK

WTKNG

(13) energie ce poate fi scrisă pe volum şi suprafaţă scriind în prealabil funcţia de partiţie VΞ , SΞ , adică V

ijG şi SijG .

După [10], [11] şi [12] avem: ijijijij SG σ+ν= cu (14)

P,V,T

ijij N

G⎟⎟⎠

⎞⎜⎜⎝

⎛

∂

∂=μ

ν – numărul de moli; σij – coeficientul de tensiune superficială; Sij – suprafaţa liberă, relaţie în care sumarea se face după indici. Inlocuind (13) în (14) obţinem

[ ]ij

ijijBijijijijijij S

ClnTCKW)C1(CWN +φΔ+−=σ

(15) în care NSij ⋅= α cu α – coeficientul de corecţie a numărului de particule de pe suprafaţă faţă de numărul de particule din volum, ∑ α=α

iiiC cu

αi – coeficientul de corecţie pentru atomul i aflat în celula elementară i cu volumul liber diferit de 0. Relaţia (15) este relaţia de definiţie a coeficientului de tensiune superficială folosind [13], [14], [15], [16], [17], [18] şi panta coeficientului în funcţie de temperatură este

[ ]

⎪⎭

⎪⎬⎫

⎪⎩

⎪⎨⎧

α

+φΔ+−=⎟

⎟⎠

⎞⎜⎜⎝

⎛ σ

ijij

ijijBijijijijij

C,N

ij

CClnTCKW)C1(CW

dTd

dTd

(16)

în care ijφ sunt funcţiile ce definesc concentraţia elementului i în j adică

N/NC,N/NC

,,C/CC

jjii

Vi

Siij

Vj

Siij

==

φ−φ=φ=

110



rapoarte de concentraţie obţinute din raportul potenţialelor chimice ce dau şi măsura presiunii parţiale a elementului i faţă de j aflat pe suprafaţa sau în volumul amestecului (ij).

Din punct de vedere matematic, mărimile σij se pot scrie sub formă matriceal-tensorială. In acest caz, potenţialul chimic devine:

⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢

⎣

⎡

⋅+μ=μ

PP

PP

PP

PP

PP

PP

PPP

PP

P

TK

333

222

113

321

221

112

311

211

111

ijkB0ijkijk

(17) în care Pijk sunt presiunile parţiale iar P este presiunea totală şi matricea de probabilitate,

⎥⎥⎥

⎦

⎤

⎢⎢⎢

⎣

⎡=

333

222

113

321

221

112

311

211

111

ijkPPP

PPP

PPP

P

iar coeficientul de tensiune superficială din (15) se scrie sub formă tensorială

[ ][ ] [ ][ ] [ ][ ]ijk

ijkijkBijkijkijkijkijk S

ClnCTKWCW +φΔ+=σ

(18) în care am considerat în matricele Cijk, ijkφ şi Sijk scăderea elementelor corespondente ale concentraţiei de pe suprafaţă şi de pe volum şi referindu-ne la distribuţii de nanoscală (cluster sau chiar reţea) în Sijk am considerat matricea cosinuşilor directori, [cosαijk] scrisă în funcţie de indicii Miller (h, k, l). Tensorul σijk este antisimetric, σijk = - σkij definind astfel posibilitatea existenţei substanţelor tensoactive pozitive şi negative, tensorul poate fi adus la forma diagonală, adică

⎥⎥⎥

⎦

⎤

⎢⎢⎢

⎣

⎡=

z

y

x

ijk

σσ

σσ 0

0

0

0

00 (19)

şi funcţia de distribuţie pe volum şi pe suprafaţă (5) se scrie sub formă tensorială,

( )[ ]∑ −⋅=Ξijk

ijkBijkijkijksV

ijk TKEN /exp, μ

(20) în care am considerat partea de translaţie, vibraţie, rotaţie atât pentru electron cât şi pentru ioni.

Chiar dacă unele mărimi din relaţia (18) sunt scalare şi vectoriale pentru omogenitatea ecuaţiei şi nu numai le-am considerat mărimi tensoriale, fapt ce poate fi admis într-o distribuţie macrocanonică scrisă din punct de vedere cuantic în care un scalar fiind o mărime discretă (şi de axă) poate fi scris sub formă tensorială. Din punct de vedere a distribuţiei de electroni în reţeaua ionică, stratul superficial poate fi considerat cu o precizie destul de bună o plasmă în care atât electronii cât şi ionii aflaţi în câmp electrostatic Vr oscilează cu lungimi de undă Debye asociate [19], [20].

2/12

0

2 −

±

±

± ⎟⎟⎠

⎞⎜⎜⎝

⎛ ⋅= ∑

Z B

SD TK

ZNeε

λ (21)

în câmpul

±−± = DrZr e

reV λ

πε/

04 (22)

determinând o barieră de potenţial,

⎥⎦

⎤⎢⎣

⎡+

+= −+ −−−+ DD rr

S

SS

r er

eer

eZN

NZNW λλ

πεπε/

0

/

0 44 (23)

în care SSS NNN =+ −+ reprezintă numărul de ioni şi electroni de pe un monostrat putându-se considera chiar mărimi egale cu parametrul de reţea. Din (23) se constată că în strat superficial avem unde de electroni şi de ioni care în cazul cel mai general pot fi considerate şi ele sub formă tensorială de pulsaţie relativă,

2/1

TT

⎟⎟⎠

⎞⎜⎜⎝

⎛=

ωω

+

−

−

+ (24)

în care semnul + l-am considerat pentru ioni şi – pentru electroni, deci am considerat temperaturi diferite pentru cele două stări, ionică şi electronică.

111



Pentru metale monocomponente de puritate atomică i = j = k şi funcţia de distribuţie pe suprafaţă devine: ( )∑ −⋅μ=Ξ

EBiii

S TK/ENexp

( )∑ ⋅σ=Ξi

BS TK/Sexp

λ=μ=μ lnTiKB0ii

cu 2/1

B )TmK2(h

π=λ (25)

i

Sii WZNE ⋅⋅=Σ , i este numărul gradelor

de libertate şi i = 3 S

ii NS α= cu

1)TK/hexp(

hWB

i −γγ

= dată de oscilatorul

armonic independent (Einstein) cu corecţia pentru oscilatorii cuplaţi după Debye,

2B

B2

B22

i)1TK/h(exp

TK/hexpTK/hW

−γ

γγ=Δ

2/1

22B

ma4

TK3⎟⎟⎠

⎞⎜⎜⎝

⎛

π=ν

iar coeficientul de tensiune superficială se calculează după

i

iii

i

0i

iWZW

αΔ+

−α

μ=σ (26)

căci Ci =1 şi 1=iφ 3/23/2

MiAi VN ⋅= −α şi panta de variaţie a tensiunii superficiale cu temperatura este

2B

B2

B

22

]1)TK/h[exp()TK/hexp(

TKZh

dTd

−ν

νν−=

σ (27)

Elem. A

(uam) μ0

(J/atom) W (J)

α Z

σ (J/m2)

dσ/dT (mJ/m2K)

σ exp (J/m2)

dσ/dTexp (mJ/m2K)

Cr 52 32,260 0,490 5,772 8 1,627 0,313 1,630 0,305 Mn 55 31,054 0,363 5,836 11 1,219 0,217 1,200 0,203 Fe 56 31,064 0,440 5,698 8 1,931 0,455 1,930 0,435 Co 58,9 31,095 0,442 5,471 8 1,970 0,266 1,958 0,287 Ni 58,7 31,094 0,434 5,432 9 1,889 0,360 1,890 0,385

Valorile μ0, W, α se calculează după cele din tabel înmulţite cu 10-20 iar pentru coeficientul de corecţie al volumului liber în care se află atomul i din celula i se ia αi pentru Cr, Mn, Fe, Co, Ni după cum urmează: αCr = 1,130; CVC; 1s2; 2s22p6; 3s23p63d5; 4s1

αMn = 1,120; CS; 1s2; 2s22p6; 3s23p63d5; 4s2 αFe = 1,102; CVC; 1s2; 2s22p6; 3s23p63d6; 4s2 αCo = 1,120; CFC; 1s2; 2s22p6; 3s23p63d7; 4s2 αNi = 1,060; CFC; 1s2; 2s22p6; 3s23p63d8; 4s2 Din punct de vedere experimental, coeficientul de tensiune superficială în funcţie de temperatură şi concentraţie se poate determina trimiţând pe stratul superficial un fascicol colimat de lumină naturală care în urma interacţiunii cu interfaţa lichid-gaz sau lichid-vid devine parţial şi apoi liniar polarizat sub un unghi de reflexie α funcţie de σ, T, C. Precizia de măsură poate ajunge până la grosimea unui singur strat de atomi pentru că diferenţa de drum dintre fascicolul reflectat şi cel refractat poate fi adusă până la dimensiunea constantei de reţea. Analiza

fascicolului liniar polarizat se face după colectarea lui cu un sistem optic în care puterea rotatorilor specifică a materialului este:

)T,c,m()a(o

r σ⋅⋅α=α

în care )a(o

α este o constantă de aparat dependentă de puterea de rezoluţie în fascicol polarizat. De asemenea, se mai poate determina distribuţia de energie liberă Gibbs în funcţie de energia oscilatorului şi numărul de oscilatori presupunând indirect că sunt proporţionali cu suprafaţa liberă, după WNGS Δ⋅Δ=Δ cu W din oscilatorul armonic liniar independent cu corecţia Debye şi S

iN din distribuţia de atomi pe suprafaţă în funcţie de energia suprafeţei după Maxwell - Boltzmann ( )[ ]iB

Si

SSi TK/Wexp1NN −−=

BIBLIOGRAFIE

[1] R. Kubo – Statistikal mecanics, Amsterdam, 1965

112



[2] C.A.Croxton – Introduction to liquid state physics John Willy Sons 1974 [3] T.E.Hill – Statistical mecanics, Mc.Gram-Hill Book, 1956 [4] A. Isihara – Statistical physics, Academic Press, New York, 1971 [5] N.N.Bogoliubov – Problems of dynamical theory in statistical physics, Amsterdam, 1962 [6] A.B.Bhatia – Phys Rev. 1974 [7] A.B.Bhatia – Phys Rev. 1984 [8] R.N. Singh – J. Phys. Chem, 1998 [9] A.B.Bhatia, R.N. Singh – Phys. Chem. Liq., 1982 [10] T. Utigard – Departament of Metalurgy and Mat Sci, Toronto, Canada, 1991 [11] K.Mori – Kyushu University, Japan, 1996 [12] R.H. Fowler and E.A.Guggenheim – Statistical Termodynamics, Cambrige University Press, 1939 [13] A.Akhiezer, S. Peletminski – Les methodes de la physique statistique, Edition de Moscou, 1980

[14] P.W.Atkins – Physical Chemistry, Oxford University Press, 1994 [15] D.Boa, I.Ansara – Thermochim, Acta, 1998 [16] L.C.Prasad, A.Mikula - J.Alloys compd, 1999 [17] L.D.Landau, E.M.Lifşit – Mecanique quantique, Ed. du Moscou, 1980 [18] L.D.Landau, E.M.Lifşit – Physique statistique, Ed. du Moscou, 1980 [19] I.Jr.Spitzer – Physics of fully ionized gazes, New York, 1955 [20] J.G.Linhard – Plasma Physics, Amsterdam, 1960

113

MANUSCRISELE, CARŢILE ŞI REVISTELE PENTRU SCHIMB, PRECUM SI ORICE CORESPONDENTE SE VOR TRIMITE PE ADRESA:

MANUSCRIPTS, REVIEWS AND BOOKS FOR EXCHANGE COOPERATION, AS WELL

AS ANY CORRESPONDANCE WILL BE MAILED TO:

LES MANUSCRIPTS, LES REVUES ET LES LIVRES POUR L’ECHANGE, TOUT AUSSI QUE LA CORRESPONDANCE SERONT ENVOYES A L’ADRESSE:

MANUSCKRIPTEN, ZIETSCHRITFTEN UND BUCHER FUR AUSTAUCH SOWIE DIE

KORRESPONDENZ SID AN FOLGENDE ANSCHRIFT ZU SEDEN:

UNIVERSITATEA “DUNĂREA DE JOS” DIN GALAŢI REDACŢIA ANALELOR

Str. Domnească nr. 47 – 6200 Galaţi, ROMÂNIA E-mail: [email protected]

Revistă bianuală acretată CNCIS

editată sub egida: Facultăţii de Metalurgie şi Ştiinţa Materialelor

Centrului de Cercetare Calitatea Materialelor şi a Mediului,

acreditat CNCSIS

Annual subscription (2 issues per year) accredited CNCSIS

Edited under the care: Faculty of

Metallurgy and Materials Science Research Center

Quality of Materials and Environment accredited CNCSIS

Data editării: 15.11.2003

Tiraj: 200 exemplare Tiparul executat la

Fundaţia “Universităţii Dunărea de Jos” din Galaţi, editură acreditată CNCSIS

Str. Domnească nr. 47. Galati 800201 Romania

Edited date: 15.11.2003 Issues number: 200

Printed by “Dunărea de Jos” University Fundation

accredited CNCSIS 47 Domnească Street, 800201 Galati Romania

Documents

ANALELE UNIVERSITĂŢII “DUNĂREA DE JOS” DIN GALAŢI UDJG MSM 2-2003.pdfMINISTERUL EDUCAŢIEI ŞI CERCETĂRII ANALELE UNIVERSITĂŢII “DUNĂREA DE JOS” DIN GALAŢI Fascicula