EFFECT OF NANOPOWDERED ADDITIONS OF TUNGSTEN … · 2011-11-09 · Hard alloy BK8 ГОСТ 3882-74 was used as an initial material, tungsten monocarbide nanopowders of 5 mass.%, 10

Fiabilitate si Durabilitate - Fiability & Durability no 1(7)/ 2011 Editura “Academica Brâncuşi” , Târgu Jiu, ISSN 1844 – 640X

1

EFFECT OF NANOPOWDERED ADDITIONS OF TUNGSTEN MONOCARBIDE ON PROPERTIES

OF HARD-ALLOY CUTTING MATERIAL

Edwin GEVORKYAN1, Yuriy GUTSALENKO2, Nikolay PROKOPIV3 1Professor, Ukrainian State Academy of Railway Transport, Ukraine

2Senior Staff Scientist, National Technical University "Kharkov Polytechnic Institute", Ukraine

3Senior Staff Scientist, Bakul Institute for Superhard Materials of the National Academy of Sciences of Ukraine, Ukraine

Abstract: Development of hard-alloy composition with nanopowdered additions produced by means of electrical sintering under pressure, test results, EM and X-rays research are presented. Basic component of improved mix for sintering of hard alloys is considered as bifractional combination of powders of standard sizing and nanosizing features. The trend to increasing of tungsten semicarbide presence is revealed in the compositions sintered in accordance with the proposed approach. At the same time under comparative test conditions the hard-alloy samples with nanopowdered additions have improved characteristics versus conventional alternatives, their hardness appreciably exceeds ones obtained by means of standard technologies that are of special importance under cutting conditions. Key words: hard alloy, tungsten monocarbide, nanopowdered addition, electrosintering 1. Introduction

Tungsten monocarbide (WC) is one of the most requested ceramic materials. It is depend on a set of unique properties of this composition. First of all, it is high melting temperature (2600-2850 °C), and high wear resistance, resistibility to thermal shock (temperatures jump) and good oxidation stability.

Therefore materials based on tungsten monocarbide are widely used for creation of wear-resistant parts, tool materials, nozzles, molds. Tungsten carbide is the hardest binary carbide keeping the properties at elevated temperatures up to 1000 °C. Corrosion resistance of WC makes it possible to use it for coatings in space engineering. Now the main application of a tungsten carbide powder is tool production where a scale of cermets on the basis of WC powder is used. The properties of WC-Co alloy such as hardness, density, resistibility, abrasivity and thermal conductivity depend on the size of WC grains. Among a large number of refractory compounds the tungsten monocarbide is distinguished by its low coefficient of thermal expansion and the exceptional hardness in a wide temperature range.

In recent years the forced investigations on obtaining of nanosized tungsten carbide and materials on its basis with lower sintering temperature and improved thermomechanical characteristics are carried out [1-3]. It is known that tungsten monocarbide has high electrocatalytic activity in oxidation reactions of hydrogen and organic connections at rather low temperatures. Therefore this compound is considered as a potential substitute for platinum and an electrode material in various electrochemical processes. From this point of view production of tungsten carbide with nanosized grains is of special complementary interest now.


2

2. Field of research tasks It is known that in the W-C system along with the monocarbides WC, which has a low-

temperature hexagonal modification and high-temperature cubic modification, there is a compound W2C, which also has a number of polymorphic modifications. Positive influence of W2C additions on the operational properties of cermets WC-Co has been established in previous practical experience [4].

Operational properties of given material to a considerable degree are stipulated by their structural state, which by-turn are the result of various technological operations of production and processing. The structure is determinative parameter when linking production technology with material properties, so it is one of the main objects of the cheking in the production and processing of materials with prescribed properties. To establish a correlation between the structure of the material and its properties, expressed in terms of quantity, it is necessary to describe a structure also in terms of quantity (by means of geometrical and other values). Quantitative characteristics of the spatial structure cannot be determined by any of the well-known "direct" of structure observation methods. These characteristics and their correlation with the properties are still studied insufficiently. Tungsten monocarbide (WC) and cobalt are the main components for production of hard alloys of BK group. The alloys under consideration are produced by the method of powder metallurgy – sintering of W and Co monocarbide powder mix at the temperature of 1320 - 1480 °C (depending on composition) in the presence of the liquid phase. At high concentrations, there is only one phase in the form of a solid solution based on the Co. WC is a rather solid compound with the high module of elasticity and high decomposition temperature of about 2600 °C.

The bonding phase of the hard alloy is a solid solution of W and C in Co, formed as replacement – implantation. The content of dissolved elements in Co varies from 1 to 10% (by mass) in the two-phase alloys.

The carbon content in the alloys influences on the chemical composition of the bonding phase: at the low content of carbon, the content of W rises up to 12-18 %, varying depending on the cooling rate of the alloys. The cobalt phase is less solid than tungsten monocarbide (by 4-5 times), it has lower modulus of elasticity and relatively low melting point (~1350 °C). Physical-mechanical and operational properties of two-phase WC-Co alloys depend on many factors, the most important of which are: technology of forming, modes of sintering, bonding phase content, WC grain size, bonding phase content, cooling rate during sintering. Properties of two-phase alloys can be varied by using different technological methods: intensified desintegration of mixes, the use of fine-dispersed powders W and WC and etc. In many cases, these methods lead to fine changes in the structure of alloys [4,5]. Change in the concentration of the bonding phase is one of the main ways to control the strength and wear resistance of the alloys. At the same content of cobalt the physical-mechanical and operational properties of the alloys are to a large degree determined by the graininess of the carbide phase, mainly the average size of WC grains. Change in the size of WC (and accordingly thickness of cobalt sphere) leads to significant changes in strength, ductility and viscosity of the alloys. Strength, viscosity and wear resistance of tungsten hard alloys will also to a great extent depend on the alloy structure nature. Existing now hypothesis about the structure of hard alloys are based on the idea of a solid carbide frame with the inclusion of cobalt (up to a specified high content of the bonding phases in the alloys) or cobalt matrix with the included WC grains, with large or a lower degree of contact. A model in the form of two interpenetrating frames is also considered as probabe one [5].


3

3. Experiments In our opinion it would be interesting to study the effect of nanopowder additions WC

on the properties of hard alloys of BK group. Tungsten monocarbide nanopowders with an average size of 70 µm was used as such additions (Fig.1).

Diameter, μm

100

90

40

30

50

80

70

60

20

10

0,05 0,1

Gra

in s

ize,

% fr

om w

eigh

t

00,01

a)

×30000

b)

Fig.1. Granulometric composition (a) and general appearance (b) of WC powder

Hard alloy BK8 ГОСТ 3882-74 was used as an initial material, tungsten monocarbide nanopowders of 5 mass.%, 10 mass.%, 15 mass.% was added to the original WC and a change in the structure and hardness of sintered hard alloys was observed.

Mixing of powders was carried out in a special ceramic drum, where air is fed under the pressure of 15 MPa. This allowed to enough evenly mix the particles with the size of 1-2 µm of mixture WC-Co with tungsten monocarbide nanopowders. It was found that this mixing method is more efficient than mixing with the use of special grinding bodies. Petrolatum, the material which is the product of oil refining and has a high plasticity was used as a plasticizer during pressing. After pressing the samples were sintered in a vacuum furnace under pressure of 100 MPa at temperature of 1400 °C and exposure of 1.5 hours. Hardness of the obtained samples were measured using the microhardness measuring device PMT-3.


4

4. Results and discussion When measuring hardness of obtained samples, it was found that the content of

nanopowder in blends of upto 15 wt.% leads to jump in material hardness upto 18-19 GPa, while hardness of regular BK8 is 16-17 GPa. The structure of the obtained samples are shown in Fig.2.

a)

b)

Fig.2. Structure of alloys BK8 (a) and BK8 with nanopowdered additions (b)

After computer processing the statistical characteristics were obtained: a number of carbide grain, a number of binding-sites, connectedness of carbide particles and others. One can see from the obtained data that the addition of tungsten monocarbide nanopowders does not lead to an increase in the size of carbides and binding-sites, but leads to an increase in the number of the grains and binding-sites.

The average conditional size of carbide zone has not been increased like in most cases for the average conditional size of the bond. It is obvious that the introduction of tungsten monocarbide nanopowders promotes the best packing of grains during the compressing process and the activation of sintering process. Small grains of nanopowders block in the growth of carbide grains, they partly diffuse in cobalt bond and that leads to hardening of the carbide-metal bonding. It seems to be explained such a significant increase in hardness.


5

Fig.3. X-ray pattern of samples BK8 with additions of 15 mass.% of tungsten monocarbide nanopowders

X-ray diffraction study (Fig. 3) has shown that in case of addition of tungsten

monocarbide, which has hexagonal crystal lattice, in the presence of nanopowdered component in blends of up to 15 mass.%, the content of W2C phase having the cubic crystal lattice increases. Content of the W2C in the BK8 alloy samples with additions of tungsten monocarbide nanopowders becomes 5-8 mass.% while it does not exceed 2-3 mass. % in common BK8. This is apparently another factor which influences on transformation of operational properties after introduction of WC nanoadditions. 5. Conclusion

Thus the carried out investigations show that nanopowder additions to ceramic mixes with the sizes of grains from 1to 5 µm can essentially improve properties of these materials obtained by sintering or hot pressing of materials. References [1] E.S. Gevorkyan; Yu.G. Gutsalenko; V.A. Chishkala; A.P. Krishtal.: Sintering of

Al2O3 and WC powders activated by electric field. In: Proceedings of 5th International Conference "Research and Development in Mechanical Industry – RaDMI 2005", Vrnjancka Banja, Serbia and Montenegro, 04-07 September 2005. Edited by Predreg Dasic. Trstenik: High Technical Mechanical School, 2005, pp. 694-695.

[2] E.S. Gevorkjan; L.A. Timofeeva; V.A Chishkala; P.S. Kislyi.: Hot Compaction of Tungsten Monocarbide Nanopowders by Electric Heating, Nanostructural Material Science – No 2-2006, pp. 46-51. (in Russian).


6

[3] E. Gevorkyan; Yu. Gutsalenko.: Activated hot Pressing Behaviour of WC nanopowders. In: Proceedings of 3nd Symposium with international participation "Durability and Reliability of Mechanical Systems": Targu-Jiu, May, 20-21, 2010, pp. 218-221.

[4] I.N. Chaporova; K.S. Chernyavskiy.: Structure of Sintered Hard Alloys: Publishing Company "Metallurgy", Moscow, 1975, 248 p. (in Russian).

[5] V.I. Tret’yakov.: Basic Foundation of Physical Metallurgy and Production technology of Sintered Hard Alloys: Publishing Company "Metallurgy", Moscow, 1976, 527 p. (in Russian).


7

ANALYSIS OF FATIGUE STRESS IN A HERTZIAN FORM

Prof.dr.eng. Stefan GHIMIŞI, „Constantin Brâncuşi” University of Târgu Jiu

Abstract.Analysis by finite element method is a modern method for studying various contacts, allowing the determination of important parameters for the study of various contacts. We consider the sphere –plan contacts which represent the materialization of the point contacts from “Experimental Stand for the fretting study” which is in the Machine Elements Laboratory

Keywords: finite element method, hertzian contact, tensions Von Mises and Tresca

1. Introduction To better study the stress state at the contact level, in the professional literature a number of criteria are introduced that allow a better assessment of the state of stress, being able to make predictions about the probability of occurrence of cracks in the contact. A first criterion can be deduced from the Von Mises criterion as the square root of second invariant of stress tensor. In rectangular form that can be written:

( ) ( )[ ]222222 6

1yyzzzzyyxyxzzy pppppppJ −+−+++= (1)

the limit is reached when 2/12 )(J attains in a point the simple tangential forces value a

simple tangential force. It can also be used as criteria of fatigue: plasticity criterion, Tresca tension stress criterion, Dang Van criterion.

Existing fatigue state in a hertzian contact can be investigated through the finite element method, thus achieving results which can not be achievable by conventional methods.

The knowledge of fatigue state in a hertzian contact is very important, it provides information regarding the appearance of contact cracks , this fact contributing to to predict the appearance of the cracks, and therefore the disposal of respective contact.

Solutie exacta

Numarulnodurilor(elementelor)

No

Solutie prinelemente finite

( - )

0 (

+ )

Eroa

re %

Fig.1. Digitization error based on the number of elements


8

2. Stages of the finite element method analysis Using specialized software for finite element analysis (Cosmos, Ansys, Nice, Pro /

driver, etc..) can proceed to study the structures comprising the steps: preprocessing and postprocessing.

Preprocessing - initial stage for the structure where we can define 1 - the system of units 2 - reference system; 3 - the geometry of the structure; 4 – the material that will make the structure; 5 - item type used to structure the digitization 6 - the analysis will be performed; 7 - structural loads; 8 - the contour conditions; 9 - software type used.

For realising the geometry of the system the was used Pro / ENGINEER , program version 2000i and the analyse by finite element method was done using Pro / Mechanica version 21 - which working in an integrated way with Pro / ENGINEER, thus reducing transfer errors.

For digitization of the structure we used the "p-element method which at the circular surfece digitized use the “block” method. The sensitivity of the digitization depends of the interpolation degree of the polynom.

3. Tension analysis in a sfere –plan contact We consider the sphere –plan contacts which represent the materialization of the point

contacts from “Experimental Standfor the fretting study” which is in the Machine Elements Laboratory [1]. Thus was considered the geometrical model shown in fig.2, the digitization of the elements being represented in fig.3.

Fig.2. Theoretical sphere-plan model

The assumed model is represented by an elastic lamella assembled through 8 bearings ball φ 19 mm diameter, thus achieving eight point contacts. Lamella receives at the free end, an upward diplacement by 20 mm. The blade recess area is based on a fix lower part, top part being loaded by 4 screws with an uniform distributed force by 200 N on each screw.


9

Fig.3. Model digitization

Using finite element method we investigated point contacts in terms of stress Taking into consideration the stress tension which appear at the contact level , causing tensions Von Mises and Tresca fatigue[2]. Von Mises tensions for upper and lower balls are given in fig.4 and fig.5. It can be observed high an different values depending by position and ball contact. Thus, for higher balls the tensions oscillate between a minimum and a maximum of 272.2 MPa 19,533 MPa, and for inferior balls the tension arise values between 436,6 MPa and 1833.9 MPa.

Fig.4. Von Mises tensions for superior balls


10

Fig.5. Von Mises tensions for inferior balls

In fig 6 and 7 are given the representations of fatigue Tresca stress for contact between the upper and lower balls ,respectively the inferior balls and lamella. It can be obseved that the tension values are 479.1 MPa to 1017.9 MPa at superior balls, respectively 1035.7 MPa to 245.4 MPa at inferior balls.

Fig.6.Tresca tensions at the superior balls level


11

Fig.7. Tresca tensions at the inferior balls level

The second model examined involved a down displacement of the lamella at the open end with 20 mm. The blade recess area is based on a fix lower part, top part being loaded by 4 screws with an uniform distributed force by 200 N on each screw.

Tresca fatigue tension for the inferior balls are given in fig.8 and the values are between 478.1 MPa and 998.2 MPa. For the superior Tresca balls the tensions are between 229.9 MPa and 942.5 MPa (Fig. 9). It can be obseved a displacement of the maximum tension stress position depending by the lamella displacements.

Fig.8. Tresca tensions at the inferior Fig.9. Tresca tensions at the superior balls level balls level

An other tension fatigue stress is Von Mises tension representation of this tension is

made in Fig.10 for inferior balls an in Fig 11 for superior balls. The observed values are between 871 MPa and 19,875 MPa for the inferior balls and between 264.9 MPa and 1621 MPa for superior balls.


12

Fig.10. Von Mises tensions pentru inferior balls

Fig.11. Von Mises tensions pentru superior balls

4. Conclusions Analysis by finite element method is a modern method for studying various

contacts, allowing the determination of important parameters for the study of various contacts. In case of hertzian contact are obtained precise information regarding the state of deformation at the contact level, it can be obtained information about the state of the existing tension at the contact level. Analysis of the point contact using the classical method (Hertz theory) does not permit to to achieve the results of such precision .The determination of Von Mises and Tresca tensions permits a proper analysis of the point contacts.

References 1.Ghimisi Stefan, Contribution regarding the wear by fretting with applications at the elastic assembling. Doctoral thesis, Politehnica University of Bucharest, 2000. 2.Hertz, H. On the contact of elastic solids. J reine une angewandth, Mathematik no. 92, p.156-171, London: Macmillan, 1896, p.88-114.


13

A SPECIAL SPIRAL: THE COCHLEOID

Prof. PhD. Liliana LUCA, Constantin Brancusi University of Targu-Jiu, Prof. PhD. Iulian POPESCU, University of Craiova, [email protected]

Abstract. They are shown geometrical considerations and properties of cochleoid, which is a flat curve from the spiral’s family. Are established various mathematical relations, which are useful when drawing the curves. They are given examples of different drawn cochleoid.

Keywords: cochleoid, flat curve, curve tracing

1. Introduction

The cochleoid (in English cochleoid, in Italian cocleoide, in Greek Kochlias) draws its name from the snail "shell". It is a flat curve, similar to the vertical projection of the spatial curve representing the line describing the snail shell (Fig. 1).

Fig. 1. The snail shells The cochleoid was researched by J. Peck in 1700, by Bernoulli in 1726 and by

Benthan and Falkenburg in 1884. The cochleoid is part of the spiral’s family.

2. Geometrical aspects The BD cochleoid from Fig. 2 is described by B point on the AP radius, when P point

goes on DP circle of radius AP = a = constant, with AB = ρ , the condition being that the PQ segment is equal to the BC arc, meaning:

ρ φ = a sinφ (1)

For an arc of r radius and length of EF =a (Fig. 3), which is positioned symmetrically regarding y-axis, it is determined the ordinate of center of gravity G, with the relationship [1]:

'' FEarOG •= (2)

where E'F 'is the projection of EF arc on the x-axis


14

Having: a = r. α , it is obtained:

2sin2 α

αrOG = (3)

Fig. 2. Geometry of cochleoid Fig. 3. The gravity center of the arc of a circle

If we consider an arc of constant length, but varying radius of the circle, then the geometrical locus of gravity center of the arcs with the same length but with variable rays is a cochleoid (Fig. 4) [1], with the equations:

- in polar coordinates:

2

2sin

. θ

θ

ar = (4)

- in cartesian coordinates:

0)( 22 =−+ ayyxarctgyx . (5)

ABi arc is shown in the figure and the center of gravity is Ci. In [2] it is shown that (Fig. 5): - the tangent in P to cochleoid passes through the A' symmetrical to OP, the one for A belonging to the curve, at its intersection with the x-axis; - the curve points where the tangent is parallel to the x axis is on a circle with A center and radius a = arc OP.


15

Fig. 4 .Variable circle radius Fig. 5. Geometric properties

3. The cochleoid drawing Fig. 6 shows the possibility of tracing the cochleoid: are plotted circles with OAi rays and

OPi arcs are taken on this, with the same length. Pi points are on the cochleoid. Each of these points is also on the one ortocycloid which runs the length OM = constant.

Fig. 6. Geometrical drawing of cochleoid


16

4. The obtained results Considering the arc OP = a = constant (Fig. 7), it shows that OAP is an isosceles triangle.

It results: arc OP = r.φ = a = constant, (6)

φ = 2.θ. (7)

Fig. 7. Polar coordinates Radius r can be cycled, resulting in polar coordinates of cochleoid. It is obtain φ = 2.θ = a / r , θ = a / (2.r), ρ= 2. r cos (90 – θ) = 2. r. sin θ. It was developed a calculation program which drew more cochleoids. Was taken a= 20

mm and r was cycled from 0.01 to 600 mm In Fig. 8 is shown the obtained cochleoid for these data.

Fig. 8. Cochleoid with r> 0 If the r radius is negative, it is obtained the cochleoid from Fig. 9. In Fig. 10 is shown the complete cochleoid.


17

Fig. 9. Cochleoid with r <0 Fig. 10. The complete cochleoid

If it is changed the arc length (a = 10, a = 25), they are obtained two cochleoids, which

are given in Fig. 11 (the inner one corresponding to a= 10). r was cycled between 0.01 and 600, with step of 0.005.

If the ray size is limited, they are obtained incomplete cochleoids (Fig. 10).

Fig. 11. Two cochleoids Fig. 12. Incomplete cochleoids

5. Conclusions - The cochleoid is a flat curve of spiral’s type. - This curve has interesting geometric properties. - It has applications in the center of gravity of circle arcs. - They drew different variants of cochleoids.


18

References

[1]. Dragnea, O. – Geometria maselor, Editura Didactică şi Pedagogică, Bucureşti, 1972.

[2]. Teodorescu, I.D., Teodorescu, Şt.D. – Culegere de probleme de geometrie superioară, Editura Didactică şi pedagogică, Bucureşti, 1975.

[3]. http://www.2dcurves.com/spiral/spiralc.html [4]. http://xtsunxet.usc.es/cordero/curvasplanas/cocleoidesurf.htm [5]. http://spazioinwind.libero.it/corradobrogi%20/V/V-333.htm


19

INFLUENCE OF THERMAL BARRIER TYPE CERAMIC LAYER’S CHEMICAL COMPOSITION AND QUICK THERMAL SCHOCK ON

THEIR POROSITY AND ADHERENCE

Lecturer dr.ing. Florin Ciofu, University "Constantin Brâncuşi" of Târgu-Jiu

Abstract: The paper presents several types of thermal barrier type ceramic layers and the variations concerning their porosity and adherence related with their chemical composition. The structural and compositional characterization of the obtained layers was performed by electron microscopy, one of the most relevant investigation methods in order to choose the optimum composition for the external layer deposited on copper devices. The effect of quick thermal shock on the samples porosity and on chemical elements distribution along the layer’s thickness was also taken into account.

Keywords: ceramic protection, shock resistance, copper support; 1. Introduction

The ceramic protection coatings, also named TBC (thermal barrier coatings), aim the prevention of rapid deterioration of metallic components following to corrosion and erosion processes by reduction or even elimination of the chemical attack of the surfaces due to working environment’s chemical agents, especially at high temperatures (in some cases over 1000°C).

The thermal barrier coating system is made of a bond coat, an oxide coat - TGO (thermally grown oxide) and an external thermally insulating coat. In most cases, the bond coat is MeCrAlY (Me = Ni or NiCo) or modified platinum aluminate coat.

The bond coat must be as compact as possible in order to ensure the oxidation resistance and the corrosion resistance at high temperatures; these properties are guaranteed by the formation of an adherent oxide coat – TGO. The former appears following to oxidation of the aluminium existing in the bond coat. It is also necessary to ensure a good adherence between the metallic bond coat and the ceramic external coat.

At the present day, the composition of external ceramic layers is limited by phasic modifications, with direct implications on the thermal stability at extreme temperatures, that can lead to coat fragmentation due to high temperature erosion.

Thus, it is very important to find new solutions regarding the composition of the thermal barrier layer system, in order to increase their adherence, the corrosion resistance and thermal shock resistance.

On this line, multilayered coatings, having a compositional gradient between the metallic support and the ceramic layer, in order to restrict the coating exfoliation, represent a new solution for metallic parts protection. Such multilayered protections, obtained by arc plasma spray, are superior to duplex layers thanks to the improvement of thermal and mechanical properties in thermal cycling and high temperature conditions.

2. Coatings selection

Some samples have been investigated in order to compare their structures and compositions, having the following bond coatings and external coatings (table 1):


20

Table 1 No. Sample No. Bond coating External coating 1 C 1 NiCoCrAlY ZrO2 + 8%Y2O3 2 C 13 NiCoCrAlY ZrO2 + 7%Y2O3 3 C 59 NiCoCrAlY ZrO2 + 24%MgO 4* C 61 NiCoCrAlY ZrO2 + 24%MgO

*Sample no.4 – C61 was submitted to quick thermal shock, at 800ºC

3. Experimental results Sample C 1

Fig.1. Secondary electrons image of copper support, bond coat and external coat, x 400, sample C1

The figure 1 presents a secondary electrons image of the whole sample with bond coat NiCoCrAlY + external coat ZrO2 + 8%Y2O3 .

In the image’s left side is the copper support, than we can observe the grey thin bond coat and, to the right, the external coat.

The bond coat does not present microfissures, discontinuities, the adherence at the support being satisfactory. On the contrary, the ceramic external layer has an important porosity, discontinuous fissures but also a continuous one. Thus, the bond coat/ceramic layer adherence is weak, the latter being in danger to come off.

It can see an abrupt decrease in copper concentration when passing from the support to the coatings and, as expected, a raise of the nickel and chromium content in the bond coat. Zirconium and the other elements presents a slight variation in the external layer.


21

Sample C 13

Fig.2. Secondary electrons image of copper support, bond coat and external coat, x 200, sample C13.

The figure 2 presents a secondary electrons image of the whole sample with bond coat

NiCoCrAlY + external coat ZrO2 + 7%Y2O3 . In the image’s left side is the copper support, than we can observe the grey thin bond coat and, to the right, the external coat. The bond coat does not present microfissures, discontinuities, the adherence at the support being satisfactory excepting some small zones.

The external ceramic layer presents a small number of porosities and discontinuous fissures. Thus the adherence is superior to that of the bond coat/external layer in sample C1.

We can observe a raise of the copper content, by diffusion, and a significant raise of the nickel and chromium content in the bond coat. Zirconium and other elements presents a quite uniform distribution in the external ceramic layer.

Sample C 59



22

The figure 3 presents a secondary electrons image of the whole sample with bond coat NiCoCrAlY + external coat ZrO2 + 24%MgO. In the image’s left side is the copper support, than we can observe the grey thin bond coat and, to the right, the external coat.

The bond coat presents dimensional nonuniformities, a small volume of porosities, but nevertheless the adherence at the support is satisfactory. The external ceramic coat presents porosities, but no microfissures.

The quality of the ceramic layer is satisfactory. It is to be noticed the diminution of the copper content when passing from the support to the coats and an uniform distribution of the zirconium along the external coat. Magnesium and oxygene also have uniform distributions.

Sample C 61


The figure 4.a presents a secondary electrons image of the whole sample with bond

coat NiCoCrAlY + external coat ZrO2 + 24%MgO. The sample was submitted to the quick thermal shock, at 800ºC. In the image’s left side is the copper support, than we can observe the grey thin bond coat and, to the right, the external coat.

The bond coat presents dimensional nonuniformities, a small volume of porosities, but nevertheless the adherence at the support is satisfactory. The external ceramic layer has some porosities but no fissures, its quality being satisfactory. We can notice, comparing to the sample with the same materials, non-treated, a reduction of the porosity, meaning a raise of the compactity of the coating material.

It is to be noticed the diminution of the copper content when passing from the support to the coats and a less uniform distribution of the zirconium along the bond and external coat.

Magnesium and oxygene also have less uniform distributions, apparently suffering a segregation process along the external layer.


23

4. Conclusions Following to the investigations and interpretation of the obtained results, we can make

the further assessments: 1. The bond coat generally have a good adherence, even in the absence of the

compositional similitude with the support material, copper. In isolated zones there is the possibility for the bond coat to come off; nevertheless, this possibility is low, beacause there are only isolated pores and not fissures.

2. The sample containing 7% Y2O3 in the external layer (C13) presents higher characteristics of the external layer as against the sample containing 8% Y2O3. It is nevertheless hard to estimate that these differences are due exclusively to the layers composition, taking into account the much to small variation between the two, and not also to a change in their deposition conditions.

3. Ceramic layers ZrO2Y2O3 have smaller dimensions pores as against ceramic layers ZrO2MgO.

4. Porosity lowers after quick thermal shock for the samples with ZrO2MgO external layer.

5. After the quick thermal shock, the distribution of the chemical elements in the external layer changes, the layer becoming non-uniform, following to the temperature raise and the migration by diffusion of the elements.

6. The ceramic layer ZrO2Y2O3 has prove itself to be superior to the ceramic layer ZrO2 MgO, thus confirming the theoretical data that indicates it the most resistant ceramic layer. References [1]Ciofu Florin, Nioaţă Alin – Densification of Al2O3 powder using spark plasma sintering (SPS) - Annual session of scientific papers “IMT ORADEA – 2006” – Oradea, 18 – 19 Mai, 2006, ISSN 1583-0691, pag 817-822. [2]Ciofu Florin, Nioaţă Alin - Technological bases of the preparation, structure and properties of powder steels sintered by electric contact heating of annular specimens - A XI-a ediţie a Conferinţei ştiinţifice a Facultăţii de inginerie, cu participare internaţională, Tg-Jiu 3-4 Noiembrie 2006, ISSN 1842-4856, pag. 87-92. [3]Ciofu Florin, Nioaţă Alin - Electric rolling of powder materials with a dielectric phase - A XI-a ediţie a Conferinţei ştiinţifice a Facultăţii de inginerie, cu participare internaţională, Tg-Jiu 3-4 Noiembrie 2006, ISSN 1842-4856, pag. 93-96. [4]Ciofu Florin, Nioaţă Alin - Properties of various textured fabricated by spark plasma sintering - The 13thInternational Conference of Nonconventional Technologies 17-18 Mai 2007(ICNcT2007), Iaşi , România, publicată în Revista de Tehnologii Neconvenţionale, nr.3, Editura PIM, ISSN – 1454-3087, pag. 11-14. [5]Ciofu Florin, Nioaţă Alin - Model for the consolidation of ultrafine refractori metal powder - The 13thInternational Conference of Nonconventional Technologies (ICNcT2007), Iaşi , ISSN – 1454-3087. [6]Ciofu Florin, Nioaţă Alin – Phase composition of electrospark tungsten carbide-based coatings after heating and isothermal soaking - „IMT ORADEA 2007” ISSN 1583 – 0691.


24

[7]Ciofu Florin, Nioaţă Alin - Mechanics of powder material compaction - „IMT ORADEA 2007” ISSN 1583 – 0691, pag. 1274-1279. [8] Gurrappa, S.,Sambasiva Rao, A., Thermal barrier coatings for enhanced efficiency of gas turbine engines, Surface and Coatings Technology, vol. 201, Issue 6, 4 December 2006, pp. 223-227; [9] Manoliu, V., Coşmeleaţă, G., Stefan, A., Ionescu, G., Multilayered ceramics within the cogenerative system in the power industry, 6th International Symposium of Croatian Mettalurgical Society SHMD’2004, Šibenik 2004, June 20-24, Solaris Holiday Resort, CROATIA; [10] Manoliu, V., lonescu, G., Ştefan, A., Principles regarding the thermal tests of the malerials in extreme conditions, 7th International Symposium of Croatian Metalurgical Society, SMHD 2006; [11] Kwon, J.Y., Lee, J.H., Jung, Y.G., Paik, U., Effect of bond coat nature and thickness on mechanical characteristic and contact damage of zirconia-based thermal barrier coatings, Surface and Coatings Technology, vol. 201, Issue 6, 4 December 2006. [12]Vida – Simiti, I., - Materiale poroase permeabile sinterizate, Editura OID-ICM, Bucureşti, 1992 [13]Vida-Simiti, I. ş.a. Prelucrabilitatea materialelor metalice, Editura Dacia, Cluj-Napoca, 1996. [14]Vida-Simiti, I., -Proprietăţi tehnologice şi metalurgia pulberilor, Editura Enciclopedică, Bucureşti, 1999. [15]Voicu,M., Amza, Gh, Tehnologia Materialelor, Editura Didactică şi Pedagogică, Bucureşti, 1981. [16]Wyszynski N. M., The FEA Approach to Temperature Fields and Heat Transfer in Powders (SLS) by ANSYS Environment, Solid Freeform Fabrication Symposium, August 2-4, Austin-Texas, SUA 2004; [17]Yamamoto K., Kujime S., Takahara K., Properties of nano-multilayered hard coatings deposited by a new aip-ubms hybrid coating system, 9-th International Conference on Plasma Surface Engineering, September 13 – 17, Garmisch-Partenkirchen, Germany, 2004.


25

MACRO MECHANICAL STUDY OF HOMOGENEOUS AND ISOTROPIC PROPERTIES OF A COMPOSITE WITH A

FIBER-MATRIX INTERFACE

Assistant dr. eng. Cătălina IANĂŞI, “Constantin Brâncuşi” University, Tg-Jiu Ph. D. Lecturer Minodora PASĂRE, “Constantin Brâncusi” University, Tg-Jiu

Abstract: Conventional materials used in their natural state can not simultaneously achieve a satisfactory level of complex requirements so that recourse to the completion of their combinations, generically called composite materials. They optimize the technical design of various structures, primarily based on high diversity, practically inexhaustible, of combinations that can be implemented. It must be added the possibility (if no use ordinary materials) to predict and even to "steer" a composite properties by suitable choice of nature, form and presentation of the weight of its constituents, or through application of appropriate technological steps.

Keywords: composite materials, macro mechanic, isotropic, fiber-matrix interface

1. Introduction

Machinery and equipment used in modern technology must meet, among others, two fundamental requirements: to enable high speed of work and to use energy rationally, whatever concrete form it is consumed in considered operations. These principles are concerned, with specific weights, in various practical fields, the construction of airplanes, ships, tanks and missiles or to manufacture consumer goods, sports equipment or recreational. In this regard, it should be on the market and be as easily accessible materials ever better performance, but with as low density (to allow the manufacture of parts characterized by small forces of gravity, friction and inertia) in terms of production and operating costs acceptable.

It can say that this work falls in this framework. It proposes a summary of the fundamental principles of analysis of composite materials and the design and use. Study consolidation of composites revealed that they are made by bonding fibrous material impregnated with resin on the surface of various elements, to restore or increase the load carrying capacity (bending, cutting, compression and / or torque) without significant damage of their rigidity [1, 2]. Fibers used in building applications can be fiberglass, aramid or carbon. Items that can be strengthened are concrete, brick, wood, steel and stone, and in terms of structural beams, walls, columns and floors, applying lately and beam-column nodes. Composite materials are carried out in a metallic or nonmetallic matrix that is reinforced by the dispersion of particles, fiber or gas. Those achieved in a non-matrix, more precisely epoxy are strengthened by the dispersion of fiber glass, aramid, boron, carbon. Fibers are able to withstand elastic applications but to resist and other requests for compression, cutting, bending. Examples of composites are: CFRP (carbon fiber reinforced plastic), GFRP (glass fiber reinforced plastic), AFRP (aramid fiber reinforced plastic), etc. Unions, they must be combined with a matrix to form a compound. Matrix serves only a number of functions, most of which are: ensuring a support for fiber, stabilizing the fibers against the forces of disruption and cutting, from the resins system used epoxy resins offer a practical and versatile field of properties including adhesion and volatility negligible during the treatment also have good mechanical properties in the form of resin treated.


26

2. Composites − isotropic materials It is done an analysis of the calculation and design of composite materials. It seeks to

determine the constitutive relations which can be estimated using physical and mechanical characteristics of a composite material [6, 7]. To arrive at these constituents relations in final form, are considered three separate issues: A. composite material has a matrix structure consisting of fibers parallel embedded which is drive along the axial fiber required; B. composite material has a matrix structure consisting of fibers parallel embedded which is drive along the axial fiber required in a perpendicular direction from the arrangement of fibers direction; C. the material is a composite aggregate applied uniaxial tensile. It was found that if a composite material consisting of matrix (M) parallel fibers embedded (f) is applied along the uniaxial tensile strength fibers Fc, which produces elastic deformation of composite materials and its components, its behavior can be described as: composite specific deformations (elongation) matrix and fibers on direction of force application are equal. If EM, EF and EC are the longitudinal elastic modules of matrix, fiber and composite and σM, σf, σc are the normal stresses (in the direction of the force Fc) generated by mechanical stress in the matrix, fiber and composite it can apply Hooke's law and follows the relationship (1):

M

MM E

σε = f

ff E

σε =

C

CC E

σε = (1)

Also, the normal stress (on Fc force direction of application) generated by σM σf and σc are equal. If the assumption is made that the effects of these applications on wood samples overlap without mutual influence may be inferred that the total values of the three components of specific strains will be calculated depending on the components of the state of global tension, with the following relationship (2):

⎪⎪⎪⎪

⎩

⎪⎪⎪⎪

⎨

⎧

++=

++−=

+−=

xy

xy

y

yyx

x

xxsxy

xy

xysy

y

y

x

xxyy

xy

xysx

y

yyx

x

xx

GEE

GEE

GEE

τση

σηγ

τη

σσνε

τη

σν

σε

(2)

As we seen, this expression connects the parameters of a state of tension with the flat plane strain state. We reach to offset the constitutive equations of a given material. Description of this type of response can only be achieved by introducing additional constant challenge to the material, making tests be more complex than for isotropic materials. Constitutive equations are corresponding to the generalized Hooke's law, which is studied in the concepts of elasticity theory, with reference to homogeneous and isotropic materials. Looking closely, we find that these relations may be written in the format matrix as (3):


27

=⎪⎭

⎪⎬

⎫

⎪⎩

⎪⎨

⎧

xy

y

x

γεε

⎪⎭

⎪⎬

⎫

⎪⎩

⎪⎨

⎧

×

⎪⎪⎪⎪

⎭

⎪⎪⎪⎪

⎬

⎫

⎪⎪⎪⎪

⎩

⎪⎪⎪⎪

⎨

⎧

−

−

xy

y

x

xyy

ys

x

xs

xy

sy

yx

xy

xy

sx

y

yx

x

GEE

GEE

GEE

τσσ

ηη

ην

ην

1

1

1

(3)

The constitutive relations can be written, for example, in a similar way as above, with components denoted by { iε }, specific strains denoted by {ε } i = { }{ }jijS σ , where {S ij } is the flexibility matrix of the material. Moreover, the same constituent relations can be written {σ } i = { }{ }jijC ε , where {Cij} is called rigidities material matrix.

As a conclusion, considering expressions result that the matrices {S ij } and {C ij } contain the coefficients of stress state parameters and the deformation for constitutive relations are written. Also, the highest degree of anisotropy is recorded for long continuous fiber reinforced composites, especially if reinforcement is unidirectional. In addition, it is clear that the most pronounced differences will be between the longitudinal characteristics values and transverse respectively.

So, in this paper, is presented a deepening macro mechanic study of composite materials in the sense that it develops the idea of composite and its constituents, namely the matrix material and reinforcing material for. It brings in discussion the material that will be strengthened, namely wood [3]. During this paper we work with two wood species namely Abies (fir) and Fagus sylvatica (beech) including profile type metal material INP 100. Composite material used to strengthen wood samples is the type of carbon fiber cloth (Megawrap-200) and carbon fiber blades (Megaplate), supplied by Sika Romania [4, 5].

Fig. 2. Wood sample carbon fiber reinforced

Reinforcing material is bonded to the surface of wood with epoxy resin Epomax PL provided by the same firm (fig.1 and fig.2).

Fig. 1. Wood sample carbon sheet reinforced


28

Fig.3.Wood sample (beech) carbon fiber reinforced I Fig.4.Wood sample (beech) carbon fiber reinforced II

Fig.5. Wood sample (fir) carbon fiber reinforced Fig.6. INP 100 metal sample carbon fiber reinforced

To control with high precision disposition, share and their percentage volume, were made in the laboratory several types of laminated products, namely: -Beech-laminated beams, with dimensions of 25mm x 50mm x 500mm, reinforced with sheet or carbon fiber with an epoxy resin; -Fir beams, with dimensions of 50mm x 50mm x 500mm, reinforced with pretension carbon plates; -Beech-beams, with dimensions 100mm x 100mm x 1000mm, reinforced pretension carbon plates; -Iron beams, profile INP 100, 1000mm length, reinforced pretension carbon plates 25mm × 50mm × beams were reinforced with 500mm carbon fiber sheet, 1 mm thick, glued with epoxy resin EPOMAX-PL, or carbon fiber plates, 1,2 mm thick, glued in the same way with epoxy resin. Wooden beams and iron beams 50mm x 50mm x 500mm and 100mm x 100mm x 1000mm have been reinforced with carbon plates pretension. Tensioning has been achieved through two special design devices (fig.3, fig.4, fig.5). Experimental studied cases are: Case 1: reinforced composite beam: -1.a: softwood with Megaplate reinforced composite plates (fig.5); -1.b: beech with Megaplate reinforced composite plates (fig. 3 and fig.4); -1.c: metal beam (INP type 100) with Megaplate reinforced composite plates (fig.6). Case 2: beam reinforced with Megawrap-200 sheet and composite plates Megaplate bonded with epoxy resin Epomax-PL type (fig.1 and fig.2).


29

In 1.a case the beam is reinforced bottom up high strength plates composite material. Composite plates axially tensioned on the beams with a screw and are leaning against the head and driven by transverse fracture occurring transverse force, axial force (the tension), maximum displacement. Work equipment and instrumentation used in this case are: universal machine for mechanical tests, data acquisition system Spider 8, 12 bit resolution linear WA300 race inductive transducer, force transducer S9 50kN, signal conditioning NEXUS 2692 - A-014, 4391 type piezoelectric accelerometer, IBM ThinkPad R51 notebook. Parameters recorded after the bending tests are: F (kN)-compressive strength of hydraulic press, Ft (kN -transverse compressive strength, CRS (mm)-race piston, Acc (m/s2) -acceleration of beam vibration (sensor break).

In 1.b case we have beech beam strengthened up and down with high strength composite material, composite strips into the modified axially tensioned where tensioning screw is replaced with a special hydraulic; beam is leaning against the head and driven across to the breaking strength recorded maximum cross and displacement (no longer measured axial tensioning force). In 1.c case we have rolled steel beam (rolling type INP100), reinforced high strength composite material below, composite plate bottom attached with straps and screws.

Functional dependency graphs are represented in Excel. Analyzing of vibration acceleration can be observed moments of failure of the first beam of wood fibers, resulting in gradual reduction of prestressing force. This decrease has required manual intervention through the screw-nut system to increase the tensioning force, resulting in graphic form as in fig.7.

Fig. 7. Determined characteristics for a Ft = 25 kN force

3. Conclusions

Due to proposed in this thesis, two devices were made to study the breaking beams phenomenon, carbon fibers reinforced, due to bending forces imposed on them [8]. Along with the main objective of the paper was intended achieving a particular computer program, Presa.txt for experimental determination of bending strength for the four representative samples. This was done with Spider 8 data acquisition equipment connected to fixtures and fittings of the samples on the machine table.


30

Following conclusions can highlight: wood is a material with a certain degree of heterogeneity, which makes its mechanical properties vary in a range too wide, so it is especially necessary to improve resistance with composite reinforcement; the wood properties are depending on wood fibers that showing heterogeneity so in experiments it is required a large number of samples to make a statistical analysis and to determine safe levels of resistance and rigidity; it is necessary to work with very good quality samples, no prospective concentrators power to distort test results; beech wood subjected to bending tests gave better results than fir; beech relative humidity was around 12% which improved mechanical resistance to bending of both consolidated and unconsolidated beams; strengthening the composite material are more effective if they are away from the section neutral axis; variants with reinforcements placed in the middle section of the beam did not increase strength in some cases even having negative effect; composite material used must be of quality, a proper matrix being crucial; where the resin impregnated sheet was used directly on the sample did not lead to expected increases in composite strength due to poor quality product; metallic material used (profile INP 100) was subjected to successive loading and unloading without exceeding the elastic so that testing can provide conclusive data, the phenomenon has resulted in a hysterezis curve.the large number of parameters and variables involved in breaking problems, we achieved an idealization of actual models. Developed mathematical models based on these assumptions, will be solved analytically or numerically, and finally verified by experimental testing. The main purpose of this paper is to analyze bending phenomenon fracture at the wood-composite samples. To achieve the objective

References [1] Abdel-Magid B, Dagher HJ, and Kimball T. The effect of composite reinforcement on

structural wood. Proceedings–ASCE 1994 materials engineering conference. Infrastructure: new materials and methods for repair, San Diego, CA, November 14–16; 1994.

[2] Borri A, Corradi M, Grazini A, A method for flexural reinforcement of old wood beams with CFRP materials, Department of Civil and Environmental Engineering, School of Engineering, University of Perugia, Composites: Part B 36 (2005) 143–153.

[3] Ianăşi C., PhD Thesis, University of Craiova, Craiova, 2008. [4] Ianăşi C., Pasăre M., Mechanical properties study of composite materials in beech beams

strengthening case, Revista de Materiale Plastice, 48, 1/2011, pag 78-82, included in ISI/SCI Web of Science and Web of Knowledge.

[5] Ianăşi C. Improving the mechanical properties of wood beams reinforced with CFRP composites plates, Annals of the Oradea University, Fascicle of Management and Technological engineering, CD-Rom Edition, , Universitatea din Oradea, 28-29 Mai, 2010, ISSN1583-0691, cotaţie CNCSIS B+, cod 665.

[6] Mareş M, Materiale compozite, proprietăţi şi modelare, Ed. Tehnopress, Iaşi, 2007. [7] Rosca V., Ilincioiu D., M. Marin, Aştefanei I., Resistance Testing of Materials.

Publishing HouseTOP Universitaria, Craiova, 2007.


31

GRAPHENE: A NEW MATERIAL

Prof.PhD.eng. Cătălin IANCU, Engineering Faculty,”C-tin Brâncuşi” Univ. of Tg-Jiu Lecturer.PhD.eng . Alin STĂNCIOIU, Engineering Faculty,”C-tin Brâncuşi”

Univ. of Tg-Jiu

Abstract:The paper presents the properties of a new but allready known material – graphene. Graphene is a 2-dimensional network of carbon atoms. Are presented the estonished characteristics of this form of carbon, along with some interesting field of use. Keywords: graphene, carbon layer, mechanical properties. 1. Introduction Graphene is a 2-dimensional network of carbon atoms. These carbon atoms are bound within the plane by strong bonds into a honeycomb array comprised of six-membered rings. By stacking of these layers on top of each other, the well known 3-dimensional graphite crystal is formed. Thus, graphen is nothing else than a single graphite layer.

The quasi-1-dimensional carbon nanotubes are formed by rolling up graphene sheets. By adding five-membered rings it is possible to form the quasi-0-dimensional fullerenes. The most prominent one is the Buckminsterfullerene C60 (Bucky ball), which looks like a football ball (figure 1).

Fig.1. Honeybomb lattice of graphene.

Graphene layers can be stacked into graphite or rolled up into carbon nanotubes. The formation of fullerenes requires the incorporation of five-membered rings.


32

2. Electronic properties of graphene The electronic properties of graphene are most interesting. The electronic structure of graphene has been known from theory for a long time and was frequently used to describe the properties of carbon nanotubes. Recent progress in the preparation of graphene and thin stacks of graphene (so-called few layer graphene, FLG) has put graphene into the focus of many research activities worldwide. Thereby different methods are currently used to obtain graphene. Among them are mechanical or chemical exfoliation from graphite and epitaxial growth on silicon carbide or transition metal surfaces. Graphene differs extremely from conventional 2D electron gas systems created in semiconductor heterostructures. The reason is the linear dispersion relation E(k) of the charge carriers in the vicinity of the K-point of the hexagonal Brillouin zone, where the bonding π-band meets the anti-bonding π*-band. The band structure thus has the form of a double cone, which formally is equivalent to the dispersion relation of rest-mass-less particles. The symmetry of the lattice requires a two-component wave funktion, similar to particles known from relativistic quantum mechanics. The charge carriers and their behavior are described by Diracs equation for mass-less fermions. This has several interesting consequences. An example is the unusual Landau level spectrum when the system is subject to a magnetic field, and the resulting new half-integer quantum Hall effect. Other interesting properties of charge carriers in graphene are their scattering and interference phenomena. Graphene thus provides new and interesting physics to study both experimentally and theoretically. Another interesting property of graphene is its high charge carrier mobility, for which values of 10.000 Vs/cm2, in some cases even 200.000 Vs/cm2 were reported. Graphene is thus a candidate for high frequency electronic devices. But other applications such as ultrasensitive gas detectors, spintronics, or quantum computing were proposed as well. Several groups in Erlangen (Germany) are now working on this interesting material. The research carried out concerns different aspects such as growth of graphene, its electronic properties and its modification (e.g. doping, band gap engineering), and processing of graphene into electronic devices.

Fig.2. Band structure of graphene Band structure of graphene (figure 2) showing the π-bands which are responsible for charge carrier transport. In contrast to semiconductors, which possess a parabolic dispersion relation, graphene exhibits a linear dependence of the electron energy on the wave vector.


33

3. “AR” atmosphere improves growth of epitaxial graphene on SiC A recent publication of the group of Th. Seyller in Nature Materials shows the beneficial influence of an atmosphere of Argon gas on the morphology of graphene layers grown on SiC (figure 3). The Ar atmosphere slows down Si evaporation, thus enabling a higher growth temperature which leads to a smoother surface morphology. The paper has been hilighted by a News and Views article written by P. Sutter. This technique can lead to industrial obtaining of grapheme layers, which can be used in different technologies and applications

Fig.3. graphene layers grown on SiC 4. Mechanical properties of graphene As of 2009, graphene appears to be one of the strongest materials ever tested. Measurements have shown that graphene has a breaking strength 200 times greater than steel, with a tensile strength of 130GPa (19,000,000 psi).[4] However, the process of separating it from graphite, where it occurs naturally, will require some technological development before it is economical enough to be used in industrial processes, though this may be changing soon. Graphene paper or GP has recently been developed by a research department from the University of Technology Sydney by Guoxiu Wang, that can be processed, reshaped and reformed from its original raw material state. Researchers have successfully milled the raw graphite by purifying and filtering it with chemicals to reshape and reform it into nano-structured configurations which are then processed into sheets as thin as paper, according to a university statement. Lead researcher Ali Reza Ranjbartoreh said: 'Not only is it lighter, stronger, harder and more flexible than steel, it is also a recyclable and sustainably manufacturable product that is eco-friendly and cost effective in its use.' Ranjbartoreh said the results would allow the development of lighter and stronger cars and planes that use less fuel, generate less pollution, are cheaper to run and ecologically sustainable. He said large aerospace companies such as Boeing have already started to replace metals with carbon fibres and carbon-based materials, and graphene paper with its incomparable mechanical properties would be the next material for them to explore. Using an atomic force microscope (AFM), the spring constant of suspended graphene sheets has been measured. Graphene sheets, held together by van der Waals forces, were suspended over SiO2 cavities where an AFM tip was probed to test its mechanical properties. Its spring constant was in the range 1–5 N/m and the Young's modulus was 0.5 TPa, which


34

differs from that of the bulk graphite. These high values make graphene very strong and rigid. These intrinsic properties could lead to using graphene for NEMS

(nanoelectromechanical systems) applications such as pressure sensors and resonators. As is true of all materials, regions of graphene are subject to thermal and quantum fluctuations in relative displacement. Although the amplitude of these fluctuations is bounded in 3D structures (even in the limit of infinite size), the Mermin-Wagner theorem shows that the amplitude of long-wavelength fluctuations will grow logarithmically with the scale of a 2D structure, and would therefore be unbounded in structures of infinite size. Local deformation and elastic strain are negligibly affected by this long-range divergence in relative displacement. It is believed that a sufficiently large 2D structure, in the absence of applied lateral tension, will bend and crumple to form a fluctuating 3D structure. Researchers have observed ripples in suspended layers of graphene, and it has been proposed that the ripples are caused by thermal fluctuations in the material. As a consequence of these dynamical deformations, it is debatable whether graphene is truly a 2D structure. 5. Conclusions Until now graphene of sufficient quality has only been produced in the form of small flakes of tiny fractions of a millimeter, using painstaking methods such as peeling layers off graphite crystals with sticky tape. Producing useable electronics requires much larger areas of material to be grown. This project saw researchers, for the first time, produce and successfully operate a large number of electronic devices from a sizable area of graphene layers (approximately 50 mm2). The graphene sample, was produced epitaxially - a process of growing one crystal layer on another - on silicon carbide. Having such a significant sample not only proves that it can be done in a practical, scalable way, but also allowed the scientists to better understand important properties. Graphene also has potential for exciting new innovations such as touchscreen technology, LCD displays and solar cells. Its unparalleled strength and transparency make it perfect for these applications, and its conductivity would offers a dramatic increase in efficiency on existing materials. References [1] K.V. Emtsev, A. Bostwick, K. Horn, J. Jobst, G.L. Kellogg, L. Ley, J.L. McChesney, T. Ohta, S.A. Reshanov, J. Röhrl, E. Rotenberg, A.K. Schmid, D. Waldmann, H.B. Weber, Th. Seyller, Towards wafersize graphene layers by atmospheric pressure graphitization of SiC, Nature Materials 8, 2009 [2] Th. Seyller, A. Bostwick, K. V. Emtsev, K. Horn, L. Ley, J. L. McChesney, T. Ohta, J. D. Riley, E. Rotenberg, F. Speck, Epitaxial Graphene: A new Material, Phys. Stat. Sol. B 245, 2008 [3] http://www.graphene.nat.uni-erlangen.de/indexeng.htm [4] Lee, C. et al., Measurement of the Elastic Properties and Intrinsic Strength of Monolayer Graphene, Science, 2008 [5] Breakthrough in Developing Super-Material Graphene, ScienceDaily. 2010-01-20. http://www.sciencedaily.com/releases/2010/01/. [6] Tzalenchuk et al. Towards a quantum resistance standard based on epitaxial graphene, Nature Nanotechnology, 2010


35

MECHANICAL TESTING OF SINTERED MATERIALS

Ş.l. dr.ing. Cristina IONICI, Universitatea ,,Constantin Brâncuşi”,

Conf. dr. ing. Camelia CĂPĂŢÂNĂ, Universitatea ,,Constantin Brâncuşi” Abstract: Alloying elements have an important affect upon the mechanical properties of the Sintered Materials. Compression tests show that the cupper based material obtains high results in cryogenic environments.Tests reveal the weak resistance of materials that contain 0,8% C, due to a poor classification and a inhomogenous metallographycal structure. Keywords: sintered material, mechanical properties, compresion test. 1.Introduction

The metallic materials sintered are the materials obtained through the new technologies and exactly the metallurgy of dusts. The metallurgy development was determined by the scientific and technical progress that needs products with special properties that cannot be elaborated through other procedures. The metallurgy of dusts may enter successfully among the classical technologies by the varieties of pieces obtained through this procedure. In the last years the major of countries strongly industrialized MP concur successfully the conventional technologies. As it may be observed, through a classical method of deformation, the sintered material does not behave at a mechanic request as the compact material. In the structure of a MS the presence of pores influence the behavior at the mechanic request, after it ahs been concluded also in the case of the mechanic trial of compression. The trial to compression is a static request and consists in the appliance of some axial charges of compression of a cylindrical test piece until breakage or to the apparition of a fissure. These testing are made with low speed of deformation, applying the charge progressively and without shocks that is why deformation is realized slowly. The material characteristics submitted to testing are mentioned in table 1. Table 1. Characteristics of material testing

Material

Compoziţie Densitate g/cm3

min.

Alungire %

min.

Rezistenţa la tracţiune daN/mm2

Fără tratament termic HB

Duritate după TT

C %

Fe% Cu% Altele %

FC 40 0,4 Rest - Max.2 6,95 3 18 50 FC 80 0,8 Rest - Max.2 6,95 2 F50U3 0,5 Rest 3 Max.2 6,95 2 35 105 320 F80U3 0,8 Rest 3 Max.2 6,95 0,5 35 105 HF 320

The direction of application in the case of compression test can be represented as in Figure 2. Track parameters for testing are: • initial diameter, do, before the test piece diameter, [mm] • final diameter, du, diameter at appearance cracking; [mm] • Initial width (initial thickness) of the specimen, h 0; [mm] • width (thickness) of the specimen end, hu, [mm]


36

• shortening of the specimen, Sc,;[%]

100×⋅

=u

uoc h

hhS [%] (1)

• Crush resistance achieved, K. [10]. [10 MPa3 ]

( )2fh

fDFku

m

⋅−⋅

= [MPa] (2)

Fig. 2. Test of compression

Test environment influences structure.

2.The experimental part The test of compression at low temperatures is made between 130-150 K and are

respected the conditions foreseen by the technical norms of execution of the performed trials at low temperatures.

The test at compression at low temperatures consists, as in the case of the environment, in applying the compression charge until the apparition of the first fissure.

During the compression trial it is followed the shortage and crowning of the test piece. The first operation before performing the test of compression is the visual control of

the cylindrical test pieces that follow to be submitted to the trial, after which the test piece is introduced in the cooling environment, liquid azotes. Choosing the liquid azotes as a cooling agent is due to the critical point inferior to the trial temperature, the trial temperature minimum corresponds to the value of 77K. The cooling installation is a shaft with liquid azotes due to the small dimensions of the compression tests.

The test pieces were introduced in the shaft with liquid azotes and after five minutes from the violent boiling of the azotes, at the temperature of 77K, are taken out the test pieces with the help of pliers and are positioned in the trial machine.

The positioning is made correctly, so that the test piece to be in the centre of the trial zone, not to create in the test piece another tensions but the compression one.


37

After the positioning, it is measured, with the help of a thermometer, the temperature directly on the test to have correct data upon the measurements 3.Results and discussions

There are visible the material deformations through the test pieces tightness and flattening, figure 3. The test pieces structures are submitted to the compression strictures at low temperatures.

Fig. 3 The cylindrical specimens The test pieces structures submitted to the compression trials at low temperatures are

pointed out through the micrographics 3.a, b, c, and result from the tests processed and their submission to a metallographic attack with Nital.

Fig.3.a. The structural analyze of the material FC 40

The structural analyze of the material FC 40 submitted to compression at low temperatures. After requesting the compression of the test pieces from FC 40 at low temperatures it is observed the same ferito-perlitical structure with a large dimensional heterogeneity. So, it is also accentuated a large amount of freckles. The structural constituents loose their aspect of deformation observed in the structure of the requested test at the environment temperature.


38

Fig.3.b. The structural analyze of the material FC 80

The requested tests at compressions at low temperatures from the material with 0, 8% carbon, FC 80, presents the same constituents, protected and pearled with a non-homogenate repartition in the field. The constitutuents granulation is getting smaller and looses its aspect of practical deformation. The structure is much more homogeneousness as a result of the request in the cryogenically environment, and the freckles are better accentuated.

Fig.3.c. The structural analyze of the material FC 50U3

The compression test at low temperatures presents a homogeneousness structure towards other tests submitted to the cryogenic environment and also towards the tests submitted to the request in the environment. The constituent’s granulation is more delicate, the deformation aspect is less favorable and the freckles network is more evident. If there is made a comparison of the compression results at low temperatures with the achieved results from the environment, we can say that the resulted values at low temperatures are superior, I illustrate this comparison through the representation of the compression force with the materials of cylindrical test pieces in figure 4.


39

Fig.4. The influence of materials upon the compression force

Figure 4 shows that the best behavior but also the weakest, respectively the material with 0,5% carbon and 3% copper, FC 50U3 and the material with 0,8% carbon. The material with 0, 4 % carbon keeps the proportion had in the case of the environment. Another important parameter of the composition influence of the material in the answer to the mechanic request of compression is the visualization between the two averages in figure 5.

Fig. 5. Mechanics request of compression

Analyzes show that an intensification of the material densification in the case of low

temperatures that leads to an increased resistance in the case of mechanic requests, large deformation tensions being accepted. For the materials sintered by the chosen composition for the type of test pieces submitted to the trial at environmental temperature, the corresponding breakage the obtained porosity may be a ductile breakage.

050

100150200250300350400

0 50 100 150 200 250 300

Forta [KN]

Scurtarea [%]

FC-40 FC-80 FC-50U3

0

500

1000

Mediu ambiant

Mediu criogenic

FC40237

FC40327

FC80214

FC80293

FC50U3289

FC50U3353Forţa de

compresiune [KN]

Influenţa materialelor

FC40 FC80 FC50U3


40

4.Conclusions Among the experimental data corresponding to compression from the calculated data we have considered the following observations upon the behavior of materials at request:

• The compression of the sintered materials is different from the compact materials ones because plasticity is not based on the law of constant volume;

• The biggest value of the compression force is accepted by the material with 0,5% carbon and 3% carbon, FC50U3; at the same time we may say that the test piece from the FC50U3 material at a force value of 60N;

• Also, the material FC50U3, is behaving good at deforming because the test piece flattening is realized at bigger charges than for the other two materials and the value of the agreed force is much bigger than the other two materials, the agreed force is 353 KN;

• The weakest behavior is presented by FC 80 through the acceptance of a small resistance;

• The breakage is not preceded as in the case of compact materials by the test splitting but only for its deformation through flattening and crowning, and for the maximum value of the test piece deforming force is transformed in slag;

References [1]. A.P.M.I. Int, Powder Metallurgy 1977 Facts, www.epma.com.; [2]. P.Beiss, Steam treatment of sintered parts, PM 1990, World Conf.London, England,

1990,p.488-491; [3].C.Ionici, Studii asupra comportării oţelurilor sinterizate supuse solicitărilor mecanice la

temperaturi scăzute, Craiova 2004; [4]. M.Mangra, Materiale fabricate prin metalurgia pulberilor, Ed.Universitaria Craiova, 1997 [5]. M.Mangra, Tehnologii şi aplicaţii in metalurgia pulberilor, Ed.Universitaria, Craiova,

2002; [6]. A.Salak, Powder Metallurgy 1977 Facts, .epma.com.; [7]. Catalog de firmă, Sinterom Cluj-Napoca, 2004 ; [8]. STAS/SREN-10045-1-1993, Încercări de rezistenţă, Produse din pulberi metalice

sinterizate, Colecţie de stasuri ; [9]. ISO 5740, Încercări de rezistenţă, Materiale metalice sinterizate, Colecţie de

standarde; [10]. SR, ISO 5755/2, Încercări la temperatură scăzută, Colecţie de standarde; [11]. SR, EN ISO 10002, Forma epruvetelor la tracţiune şi compresiune, Colecţie de standarde.


41

PROCESS AND UNCONVENTIONAL METHOD OF MARKING OR STAMPING OF HARD METAL

Prof. Dr. Eng. Gheorghe Popescu, - “Constantin. Brâncusi” University of Târgu Jiu Lecturer dr. eng. Alin Stăncioiu, - “Constantin. Brâncusi” University of Târgu Jiu

Abstract: The authors present in this work a new unconventional technological way of marking or stamping of very hard metals from tempered steel with help of explosives. Even this method is in experimental stage, it can became an object for study in another related domains. Keywords: marking, stamping, explosives, explosive’s force, excavations, dynamic wave of shock, hard metal. 1. Introduction

The operations of marking or stamping are the technological operations of pressing which help to execution of a convex-concave relief on the pieces’ surface through change the thickness of material in different sections.

To make excavations or reliefs on surface of a very hard metal can be used the force of a explosive load.

The explosive’s force f is defined like mechanic work which can be made by bases which are resulted from explosive transformation of a mole or of a mass unity of explosive if these are let to relax at a normal atmosphere pressure and temperature To [2]. It is express by relation:

eTRnf ⋅⋅= [J/Kg], (1) where: n - moles number of gaseous products which are resulted from decomposition of

a mole or of a mass unity explosive; R - the universale constant of gases; Te - temperature of explosion.

2.The experimental pattern proposed and materials used To make excavations on the plane surface of a small plate by tempered steel 4, we

made a stall of trials presented in figure1, [1,2] . The detonators 1 were made in accordance with STAS 8136-88 having a suplementary speed of detonation between 8100 m/s and 8200 m/s.

The protection tube of detonators was mode by electrolytical copper through the cool extrusion with and outside diameter by 8 mm. The bottom of tube was made in two ways:


42

- With plate bottom and a thickness between 0,3 and 1 mm;

- Whith a bottom by 0,8 mm marked by engraving/poansoning in a depth by 0,5 mm, in accordance with figure 2.

In figure 3 is presented a detonator completely equiped. The experiences was performed for many types of metals:

- small plates by 5 mm thickness by tempered steel OSC 10, returned to 62 HRC and rectified after tempere;

- small plates by 5 mm thickness by untreated steel OL 37; - small plates by 5 mm thickness by brass Cu-Zn 63; - small plates by 5 mm thickness by lead with a purity by 98%.

The experiments conditions was the same for all types of plates.

2. The result of experiments For each type of plates was performed ten

experiments using the same type of detonators. The bottom of detonators was set straight on the surface of metal plates like in figure 1.

After all these experiments was obtained the next results [1]:

a. In casse of lead plates. All detonators with the bottom’s thickness of protector tube less than 0,8 mm and no matter of brisant charge they bored completely the plate of lead, and diameter of hole created was biqqer than 8 mm (figure 4).

Fig.1. The test stand.

Fig.3. The detonator completely equiped.

Fig. 2. Marking on the

bottom of the tube.


43

Fig.4. Hole made.

Fig. 5. Excavation done.

b. In case of steel OL 37 plates . For all type of detonators used, was obtained a gently stamp in shape of on irregular excavation with a depth which is variable between 0,1 and 0,3 mm (figure 5).

c. In case of Cu – Zn 63 plates. For all types of detonators used was obtained a pronounced stamp with finished aspect with a variable depth between 1 and 3 mm, in inverte proportion to depth of tube’s bottom (figure 6).

Fig.6. Excavation done.

Fig.7. Excavation done.

d. In case of plates by steel OSC 10 tempered-returned to 62 HRC. For all types of detonators with plate bottom . wasn’t obtained any stamp or excavation.

For thase detonators with bottom marked by engraving some numbers, was obtained on the plate by tempered steel on excavation in shape and dimensions of numbers which was printed on the bottom of tube with a depth between 0,2 and 0,6 mm, (figure 7). The last result we wanted to know with such a perfect border.

In this sense, we extended researches to direction of propagation of dynamic wave of shock through the explosive string of detonators with bottom plate (unmarked through engraving) and detonators with bottom marked by engraving. Through engraving we realised that the material is leaking axially in cavity of engraver (poinson) and the thickness of bottom is making thin in the aria of number about 0,3 to 0,2 mm. This is possible because the engraving is made with against-engraver.

We detonated such as detonators in a dark room, taking photos in the same time to this phenomena. After the photos was processing it comes out that:


44

Fig.8. The shape of dynamic wave of shock.

- at detonators with bottom marked by engraving the shape of dynamic wave of shock is mainly axially (figure 8);

- at unmarked detonators (with plate bottom) the shape of dynamic wave of shock is mainly lateral function of thickness of bottom of tube, like in figure 9.

3. Conclusions The experimental researches performed lead us to the next conclusions:

- The dynamic wave of shock can be led to obtain some marks or stams on a hard metal through perform some marks or different outlines on the bottom of a detonator;

- The surface on which the marking or stamping can be made is determinated by diameter of tube of detonated charge;

- The cality of processed surface by marking and stamping using explosion depends by the properties of material used;

- This process can’t be used for metalic material with low harduess; References [1]. Popescu Gh; Stancioiu Alin. Nonconventional technology regarding marking or stamping through explosion of heavy metal. Revista de tehnologii neconvenţionale, Nr 2/2007, Editura PIM –Iaşi, ISSN-1454-3087, pag. 95. [2]. Popescu Gh. The elaboration of a new types of explosives with more security and means of initiate these for inflamable environements from mine with oil and sulph, Thesis, University of Petrosani, 1991.

Fig.9. The shape of dynamic wave of shock.


45

SYSTEMS FOR THE ADJUSTMENT OF THE MAIN KINEMATIC CHAIN OF THE VERTICAL TURNING LATHES

Prof.dr.ing. Dan PRODAN, conf.dr.ing. Anca BUCUREŞTEANU,

Conf.dr.ing. Emilia BĂLAN, Universitatea „Politehnica” din Bucureşti

Abstract: This paperwork presents some systems for the adjustment of the main kinematic chains of the Vertical Turning Lathes that have been used within the last 20 years. There are presented adjustment systems used when having the machine equipped with older type motors, with only one speed, and also modern systems for which A.C. motors with the speed adjustable by frequency variation are needed. It insists on the present role of the “gearboxes” as torque multipliers and not as speed reducer. Keywords: vertical turning lathes, gearboxes, adjustment mechanisms.

1. Introduction

The Vertical Turning Lathes are Heavy Machine Tools and initially they were intended for cutting circular workpieces for which the ratio between length and diameter is less than 1, between 0.5 and 0.9. At the beginning they were able to perform only turning. Then, drilling, milling, grinding and gear cutting became possible, especially after the implementation of the CNC equipment on the Vertical Turning and Boring Machines. Because of the vertical position of the main spindle, these machines have rigid kinematic chains that allow heavy machining with a high accuracy and efficiency [1].

The main characteristic of the Vertical Lathes is the diameter of the table or workpiece. Usually, the table diameter is between 1000 mm and 6000 mm. However, there are Vertical Lathes with the table up two 20000 mm [2]. If the Vertical Lathe is for turning only, then the main kinematic chain ensures exclusively the table rotation. If millings, drillings are needed, then the table is driven mostly by the main kinematic chain or by a circular feed kinematic chain [3]. 2. Electric motors used for driving the main kinematic chains of the Vertical Lathes

The asynchronous motors were the first motors used for driving the main kinematic chains of the Vertical Lathes [2]. The characteristic curve of the asynchronous motors is shown in Figure 1. The figure indicates: A – motor idle running starting point, B – motor idle running starting point, when connected to the main kinematic chain, C – maximum torque and speed point. Useful are can be defined as BCED.

For these motors the adjustment mode of the speed results from the following relation:

( )sp

fn −⋅⋅

= 160 [rpm] (1)

The notation in the relation (1) represents: n – speed [rpm]; f – frequency of the mains [Hz]; p – number of pole pairs of the stator coil [-]; s – slip [-]. The speed adjustment by slip variation [2] implies unwanted speed variations and poor efficiency.

The number of pole pairs is usually p = 1 or p = 2, and thus, it can be considered practically that the motor has two speeds. Usually, the power corresponding to the upper speed is by 20% smaller than the one corresponding to the lower speed.

This method does not satisfy the operational requirements, imposing complicated gearboxes as shown further. Initially, the speed adjustment by frequency variation (using frequency converters) represented a complicated and very expensive solution.


46

The reasons presented above have led to the driving by using DC motors. That had been

the solution preferred by the manufacturers until 90’s. Meanwhile, the development of the frequency converters and the decrease of their price have brought back the attention on asynchronous electric motors, this time with electronic adjustment of the speed by actuating continuously the frequency. The characteristic curves of the DC motors and asynchronous motor with frequency adjustments look like those in Figure 2.

An adjustment at constant maximum torque is made until reaching the rated speed [3], then, above this value, the adjustment is made at constant power. For the Vertical Lathes, especially for rough machining, the machining at constant torque is usually preferred, i.e. with speeds smaller than the rated speed. For fine cutting and depending on each machine type, machining at constant power is possible.

Between the rated torque MNOM, rated speed nNOM and rated power PNOM there is the following relation:

NOM

NOMNOM n

PM ⋅= 9550 [Nm] (2)

For machining on a Vertical Lathe, the kinematic chain must ensure at its output an established speed and sufficient torque for the required cutting. Usually, the rated speed and the maximum speed are much higher then the required value and the required torque is much bigger than the one developed by the motor. This is the reason of using the gearboxes. In the beginning they were used just for reducing the speed mechanically. After the development of the adjustment with speed converters, the gearboxes have been used as torque multipliers, after having been modified and simplified by the technical evolution.

3. Mechanical systems for the adjustment of the speed and torque on Heavy Vertical Lathes

These systems that are usually named gearboxes vary on machine-tool size and field of application and on the manufacturer, as well. They can have horizontal or vertical shafts. The horizontal shafts gearboxes are older and simpler solutions which do not affect the working

Fig. 1. The characteristic curve of the asynchronous motors

Fig. 2. T - torque, P - power, n - speed, ABHGD – torque range, DEFG – power range, DC – range of the speeds smaller than the rated speed, nNOM – rated speed, nMAX – maximum speed


47

area. Because of the horizontal shafts, a group of bevel gears is needed for driving the main spindle and that induces uncontrollable backlashes and specific noise.

The gearboxes with vertical shafts are more complicated and imply the usage of some systems for securing the vertical sliding gears, and they are close to the working area. However, they do not need bevel gears and therefore the backlash and the noise are removed.

Figure 3 represents the kinematic diagram of an older Vertical Lathe, with horizontal gearbox. The electric AC motor EM has a single speed. For the adjustment the gearbox has a sliding gear with three positions and two sliding gears with two positions, ensuring seven speed steps. The kinematic chain includes a belt reducer (235/385) and the final mechanism – pinion crown gear (18/113), as well. There are five shafts in the gearbox.

Figure 4 represents the kinematic diagram of the gearbox of a Vertical Lathe with 6000-mm table diameter, manufactured before 1960.

The electric motor is mounted vertically, without belt transmission. The gearbox has four steps ensured by means of two double clutches, to the shafts II-III and IV-V. The solution with clutches is simpler from constructive point of view.

For larger size machine-tools, with table diameter over 6000 mm, some manufacturers of Vertical lathes use two motors – each with its own gearbox – to develop the required power. The synchronization of the motion is made electronically. Figure 5 shows the kinematic diagram of a vertical gearbox manufactured at present by an European company.

The electric motor is an AC motor with frequency converter, rated power 100 kW, rated torque 636 Nm at a rated speed of 1500 rpm. Maximum speed is 4500 rpm. The speeds ensured at the main spindle are nMS = 0.27 -33.65 rpm, while those for the adjustment of the motor are nM = 50 – 1500 rpm. These are adjustments in the range of constant torque. In the range of constant power nMS = 100 rpm can be reached. This machine has milling capabilities as well, the main kinematic chain becoming a circular feed kinematic chain [4]. The Vertical Lathes with table diameter within 1200 mm – 4000 mm made in Romania for the last 20 years have been equipped very often with gearboxes with three speed ranges (1:1, 1:3 and 1:9). The related kinematic diagram is presented in Figure 6. It shows the variant used for SC 17, with a pinion – crown gear mechanism with the transmission ratio 20:217. It is a compact box, with only three shafts, out of which two are coaxial.

Fig. 4. EM – driving electric motor, TM, nM – motor torque and speed, TMS, nMS – main spindle

torque and speed

Fig. 3. EM – driving electric motor, TM, nM – motor torque and speed, TMS, nMS – main spindle torque and speed


48

The actuation of the 3-position sliding gear is hydraulically achieved. The motor is

connected directly to the gearbox, the output being on the same direction. The 21-tooth bevel pinion transmits the motion to crown gear pinion through a shaft named secondary shaft. The gearbox and the secondary shaft just above equip at present Conventional and CNC Vertical Lathes. They are characterized by a high reliability on condition of rational exploitation. Some of the disadvantages are: specific noise – especially when operating in the speed range 1:3 and high backlashes. Because of the backlash, it cannot be used within the circular feed kinematic chains required by the machines provided with milling head. Such cases require a separate kinematic chain [4].

A modern solution for the brand new or refabricated Vertical Lathes is represented by two-step gearboxes made by specialized companies for machine-tools [5], [6]. They have the following advantages: reduced backlashes allowing their usage on Vertical Lathes able to perform milling, eliminating the circular feed kinematic chain; separate lubrication line, which reduces heat transmission to the main spindle; connection to the main spindle through belt transmission, preventing vibration; reduced noise; simple and compact design with direct connection to the main driving motor; high efficiency, over 95%; speed change is made through a device integrated in the gearbox, electrically, pneumatically or hydraulically actuated.

The kinematic diagram of such gearbox, with the transmission ratio 1-1 and 1-4 is shown in Figure 7. The electric motor 1 – EM is connected directly to the gearbox 2 – GB. The motor shaft drives the gear 3. If the sliding clutch 4 is put on step I, through the holes a, the motion reaches the planetary gears 6 and then the output shaft 7. The gear 9 is locked. This is the step that ensures the ratio 1-4 [6]. If the sliding clutch is on position II, in holes b, the planetary gears 6 and the gear 9 ensures the ratio 1-1. 5 and 8 are bearings systems. The motion from the gearbox output shaft is transferred to the main spindle trough pinion-crown gear or belt mechanisms. This solution removes heat transfer and vibration. Because of their design, these gearboxes look like an extension of the electric motors. They could operate in any position provided that the lubrication is ensured according to the gearbox manufacturer recommendation.

Fig. 5. EM – driving electric motor, TM, nM – motor torque and speed, TMS, nMS – main spindle torque and speed

Fig. 6. EM – electric driving motor, TM, nM – motor torque and speed, TMS, nMS – main spindle torque and speed


49

These gearboxes have reduced noise and backlash and as a consequence they can be

incorporated within the feed kinematic chains. There have been manufactured Vertical Lathes with two-speed gearboxes instead of three-speed gearboxes as those presented in Figures 7-10. For the same main driving motor, asynchronous with frequency converter, there is no difference with respect to speed adjustment; however the two-speed gearbox is to be preferred. Problems can occur regarding the developed torque: the maximum reducing ratio is 1:4 for the two-speed gearbox, while for the three-speed gearbox is 1:9.

For example, it is considered the driving by a motor with the following parameters: nNOM = 1500 rpm, PNOM = 37 KW, TNOM = 236 Nm [7]. The characteristics in Figure 8 are obtained at the output of the three-speed gearbox. The maximum torque developed is 2124 Nm up to a speed of 166.6 rpm, 708 Nm up to 500 rpm and 236 Nm up to 1500 rpm. When using a two-speed gearbox for the same motor, the characteristic curve in Figure 9 is obtained. The maximum developed torque is only 944 Nm for a speed up to 375 rpm and 236 Nm for a speed up to 1500 rpm.

Depending on the machine-tool kinematic features and machining needs, the following variants result: developed torque is sufficient (no rough machining on hard materials); other reducing mechanisms are incorporated [4]; selection of a more powerful motor (bigger torque) [6].

Two-speed gearboxes have become a commonly used solution for the Vertical Lathes SC14 and SC17, when keeping the secondary shaft – see the kinematic diagram in Figure 10.

The electric motor is mounted vertically, on the same axis with the gearbox. A special timing belt (usually made of carbon fiber) takes the motion from the gearbox and transfers it to the secondary shaft, without needing bevel transmission. Only the pinion driving the crown gear is kept from the secondary shaft. The motion transferred from the electric motor to the secondary shaft is backlash free, but the classic pinion- crown gear mechanism presents backlash values that are not acceptable for the feed kinematic chain. If the machine performs milling operation, then pinions with backlash compensation [4] or a kinematic diagram as the one in Figure 11 can be used.

Fig. 7. 1. – EM - motor , 2 – GB - gearbox, 3 - gear, 4 – sliding clutch, 5 -

bearing, 6 – planetary pinions, 7 – output shaft, 8 - bearing, 9 – inner

toothing gear


50

4. Conclusions

The development of the electronic adjustment systems with frequency converters has mechanically simplified the mechanisms and adjustment systems in the main kinematic chains of the Vertical Lathes.

The adjustment of the main kinematic chain is made for technological processing reason: speed and cutting torque, as well. The gearboxes are theoretically no longer necessary for the adjustment of the speed. However, if the gearboxes were not be used, then the driving motors would be oversized and very expensive.

Technically and economically, the two-speed gearboxes, with electric shifting of the speed range, made by specialized companies, represent an advantageous solution for the manufacturing or refabrication of the Heavy Vertical Lathes, mostly for the CNC ones. References

1. Vasile Moraru, Victor Tabara, Dumitru Catrina, Nicolae Predincea, Exploatarea masinilor-unelte, Institutul Politehnic Bucuresti, 1978.

2. Emil Botez, Masini-unelte, Editura Tehnica, Bucuresti, 1977. 3. Bozina Perovic, Handbuch Werkzeug-Maschinen, Carl Hanser, Munchen, 2006.

Fig. 8. The characteristics at the output of the three-speed gearbox

Fig. 9. The characteristics at the output of the two-speed gearbox

Fig. 10. Two-speed gearboxes for the Vertical Lathes SC14

and SC17

Fig. 11. The kinetic diagram for a Vertical Lathe with milling

operation


51

STUDIES REGARDING THE WEAR OF THE TOOLS USED IN RUBBER REFINEMENT

Prof. dr. ing. Dan Dobrotă, Ş.l. dr. ing. Alin Nioaţă,

„Constantin Brâncuşi” University of Târgu-Jiu Abstract: The paper presents the results obtained in the study of tools wear that occurs when used in the refining of rubber and solutions to reduce the level of the wear. Thus the wear that occurs in this type of tool is a dimensional one, but also one of corrosion cracking which results in a rapid removal of this type of tool from use. Key words: tools, wear, rubber, refinement

1. Introduction Most machine parts that have a functional role are taken out of service due to wear of the contact surfaces in relative motion (friction). For this reason, knowledge and choice of materials resistant to wear, respectively the design of numerous items of machinery and plant construction based on the wear resistance condition has great significance [1,2]. In order to analyze the wear of the tools used in grinding scrap rubber we must first of all establish the reasons why it occurs. Wear which appears in tools used in the refining process of scrap rubber is due to the dry rolling phenomena that occur between a metal and a nonmetal [3]. The evolution in time of wear for any friction hitch comprises three distinct stages (fig. 1). This development is also valid for a friction coupling which consists of a metal and rubber.

Fig. 1. The evolution of wear in time.

From the previous figure we can distinguish three areas: 1. 0 – t1 – the burn-in period, in which the wears grow rapidly until they are

stabilized at normal work values. 2. t1 – t2 - the curve has an almost linear character, the development of wear is

proportional to operating time. 3. t2 – tc - the rapid growth tendency of wear.

Wear

A B

C

U1

U2

U

O t1 t2 tc

t


52

U - the maximum value of wear allowed in service; U2 - the value of wear corresponding to the burn-in period; U1 - the value of normal wear in service. Because the curve U is uncomfortable, in the calculations we use the line 0C, defined

by the equation : U=k • t

where: k is an angular coefficient dependent on the type of construction of the part, the type

of usage, maintenance, etc. tgα – wear intensity, specific to every part; t2 – normal operating time of the part; t1 – burn-in period;

tc – maximum operating time (not advisable to be achieved for machines which must meet special operating conditions).

To conclude, if α is low, t will be higher, and if t2 is high, the wear after the burn-in U2, has a low value, and the normal operating time will increase.

Burn-in wear is necessary in order to achieve the functional games between the parts and is controlled during the burn-in-period. Normal operating wear is almost constant for a long period of time and determines the lifetime of the tool. Accelerated wear prevents the normal operation of the subassembly and determines the disposal of tools. Rolling friction between the tools (rolls) and rubber causes the following types of wear: dimensional wear and the wear type cracking corrosion. Dimensional wear is characterized by the fact that during the grinding process of the rubber, some of the material mass of the tool is lost resulting in a reduction in size. This dimensional wear process can be compensated by modifying the distance between the cylinders. In most cases dimensional wear also determines a geometry change in the cylinders [4,5]. Because the geometry of the cylinders changes the distance between the tools will be adjusted only after having first restored their geometry. Along with the recovery of the cylinder geometry, a considerable amount of material will be removed, which causes a reduction in the size of the cylinder with a much higher value than the size of their wear. This process of recovering the geometry can be done several times, but bear in mind that it should not exceed a dimension reduction of the tools in the radial direction more than 20 mm. Preponderant influence on the size of the dimensional wear have the efforts that accompany the process of refining rubber.

The process of normal dimensional wear may be replaced by a process of a very violent nature, but only for a short period of time, when an abrasive material is present between moving elements (metal inserts, rigid bodies of various materials). The characteristics of the material the tools are made of have a great deal of influence on the size of dimensional wear, but also of the material submitted to refinement. Appropriate choice of chemical composition, degree of alloying and heat treatment for the materials the tools are made of can have a favorable effect on increasing the wear resistance.

Wear of the type corrosion cracking is of great importance because its occurrence leads to cracking of the tools and taking them out of use before the scheduled date.

1

1

2

2 ;tt

UUtgkt

Utgkc −

−==== αα


53

For a practitioner, knowledge of the susceptibility of materials to corrosion cracking has a special significance in terms of optimal design, ie the ratio of the mechanical charge, material consumption, investment costs, operation and maintenance, other influences (environment, human factors etc.). Traditionally in most cases, as a parameter for the susceptibility to corrosion cracking still serves the so-called durability of the appointed material, in an aggressive environment, in circumstances of ongoing mechanical stress.

2. Experimental research With technological exploitation, tear or rupture disposals are due, in their vast

majority, to the extension of the size of a crack type defect as a result of corrosion, respectively the action of an aggressive environment, the effect of cyclical variability in the intensity of mechanical stress (fatigue, etc..). This phenomenon is specific especially for lacerations which occur for low intensity mechanical stresses, without prior global deformation of the element affected (cracked). The tear by cracking corrosion goes through the following three stages:

- stage 1 – the formation of primers, like light sticking points (fig. 2, a) on the metal surface;

- stage 2 – the primer becomes a crack (fig. 2, b) whose size continues to slowly extend from a macroscopic point of view;

- stage 3 – tear (fig. 2, c) when the crack, by expanding, reaches a certain size (length), big enough, called critical.

The evolution of a crack and the way in which it expands, respectively the way in which it is propagated, depends heavily on the state of the existing efforts in the area where the unit is placed.

a b c

Fig. 2. Stages of corrosive tear: a – formation of primer (stage 1); b – cracking, respectively expansion of the crack, (stage 2); c – tear (stage 3).

The tools used in the refining of waste rubber are submitted cyclically to fatigue, but

they also work well in a chemically aggressive environment. The aggressiveness of the environment in which these tools work can be explained by the presence in the chemical composition of the waste rubber of various chemicals (particularly sulfur). These chemical elements, under certain conditions of pressure and temperature, may form some compounds that can be corrosive on metallic materials.

If we take into account the fact that during the shred of the waste there is a random distribution of the efforts, all possibilities of application must be taken into account. Thus, after the relative motion of the fracture surfaces, located on either side of the plane in which the crack extends, change, and hence its propagation can be done in accordance with the following basic modes of travel (Fig. 3).


54

a b c Fig. 3. The fundamental ways of travel, respectively expansion (propagation) of the cracks.

a – opening the crack (means I); b – the right sliding (means II); c – the curve sliding (means III). 1 – surface attacked by corrosion; 2 – surfaces (flanks) of the crack; 3 – front of the crack; 4 – root (tip of the crack).

means I – the crack is extended by opening, the points belonging to the crack surface shifting normally in the plane of the crack (fig. 3, a); means II – the crack is extended by right sliding, frontal, the shifting of the points in the cracking surface being done in the plane of the crack, perpendicular to its edge and in the direction of its advancement (fig. 3, b); means III – the crack is extended by curve sliding, lateral or spiral, the shifting of the points in the cracking surface also being done in the plane of the crack, but parallel to its front. (fig. 3, c).

Obviously, all other possible failure modes can be described by superposition, respectively by the proper combination of the three fundamental ways presented. With the advent of cracks, the unitary efforts σ change in value. The intensity factor of the mechanical tension, named K, is the measure of increase of unified efforts σ , generated by the presence of cracks, compared with the same nominal unit efforts, existing in an item in the absence of cracks. K’s values are always over the unit, depending on the geometry of the structure or element or of the system studied and the crack length at one time.

Slow cracking in a corrosive environment can be fully described only by the intensity factor of the voltage, the method is, therefore, based on the K factor. For higher levels of intensity of mechanical stress, there isn’t an unequivocal correlation between the K factor and the development speed of corrosion cracking. This correlation does not allow prediction of its future expansion for the cracks detected at one time. This time, the maximum utility only shows the results of the measurements under constant load of the crack extension speed.

For tools used to refine scrap rubber it is very important to avoid from the outset of the execution period the occasional hot spots or inclusions in these materials. The presence of these defects in the material of the tools can still be triggered as the main reasons for crack initiation and therefore corrosion cracking.

Regarding the wear of the type cracking corrosion, it results in the removal of the tools used to refine rubber and in that respect this type of wear should be avoided (Figure 4). The main cause of the occurrence of wear crack is represented by the working environment and conditions for applying the tools.


55

a. b.

Fig. 4. The refining cylinder and the way it breaks due to cracking corrosion a – presentation of the refining cylinder, b – cracking wear which occurs for the refining cylinders

3. Conclusions Following the analysis of the process of rubber refining, a series of changes occur in the material of the used tools and the following conclusions can be drawn:

- the process of refining waste rubber causes both a heat stress and a and a very high mechanical one for the material of the rolling cylinders;

- the wear that appears to tools used in the refining of waste rubber is both a dimensional one and one of cracking corrosion (CORFIS);

- following the use of tools, in their material a series structural changes could be seen, due to thermal effects accompanying the process of refining rubber;

- in the material of spent cylinders a diffusion of sulfur in rubber was observed, but also a reduction of carbon concentration;

- after using cylinders, their material also suffers a loss of hardness and especially in the surface layer in direct contact with waste rubber.

To avoid diffusion phenomena, the structural changes, respectively the loss of toughness, that occur in the material of used cylinders the following measures are proposed:

- performing suitable heat treatments for the material of cylinders; - using new materials for the cylinders so that they have a better behavior during

operation; - structural changes of the cylinders so that they may be required to make lesser efforts

during use; - optimization of the operating process parameters depending on the characteristics of

the materials subjected to refining; - ultrasonic activation of cylinders at different stages of the refining process.

References [1] Amza Gh., Dobrotă D., Researches concerning the ultrasonic energy’s influence over the resistance at extraction of the metallic insertion from the rubber matrix, Revue Plastic Materials, 45, ISSUE 4, page 377-380, no 4/2008; [2] Dobrotă, D., Experimental researches regarding processing rubber wastes with metallic insertion, Revue Plastic Materials, page 65-68, volume 43, no. 1/2006; [3] Dobrotă, D., Considerations on constituent equations used in study of mincing rubber waste reinforced with metallic insertion, Revue Plastic Materials, ISSN 0025/5289, page 4, volume 43, no. 3/2006;


56

RESEARCH ON THE POSSIBILITIES OF RECYCLING MATERIALS USED IN VEHICLES USING HAMMER CONSTRUCTION

TECHNOLOGY

Gheorghe AMZA, Zlatko GARAC, Zoia APOSTOLESCU, Maria GROZA DRAGOMIR, Sanda Liana PAISE,

Polytechnic University of Bucharest Abstract. The recycling of automobiles, has been part of the steel economy. Steel production requires less raw iron if more scrap metal is available as a feed material. The paper presents techniques used in recycling materials used in vehicles.

Keywords: recycling materials, vehicles, hammer mill technology

1.Introduction

Recycling techniques, even in their most simple forms, are at least 5.000 years old, taking into consideration for example the Biblical text “from swords to plough shares”. But scientific interest in recycling remained limited otherwise literature on the subject would not be so exceedingly scarce, even today. In the light of the rising public interest, which recycling has enjoyed in recent years, this thesis is giving a comprehensive in-depth understanding to the use of+ recovery techniques in one special area, namely the automobile recycling.

But from the early days of mass motoring, the recycling of automobiles, has been part of the steel economy. Ferrous scrap, as recovered from end-of-life vehicles, has a long tradition in this area as a secondary raw material substituting raw iron. Scrap is a feed material for steel production that can be brought straight into steel mills, while raw iron must first be smelted from iron ore in an energy-consuming process. Steel production requires less raw iron if more scrap metal is available as a feed material. Ferrous scrap can reduce the energy consumption of steel smelting by as much as 70%, resulting in substantially lower production costs.

2.Recycling techniques for materials used in automobile construction

When we say "recycling" we differentiate between re-using complete components (product recycling) and recycling of materials such as steel, aluminium, copper, glass, plastics, rubber and suchlike (material recycling). Product recycling has always been the domain of the dismantling companies and is already well developed. Even so, a great deal of material derived from car wrecks remains to be disposed of even after product recycling. Hence the emphasis needs to be placed on the material recycling, for which there is still a great deal of unexploited potential.

Modern end-of-life vehicle recycling not only is supposed to help protect the environment and natural resources, but is also expected to work economically, keeping costs at bay for its clients. Costs economy requires rationalized procedures.

An improved eco-balance, on the other hand, needs individual decisions about the extent of dismantling each vehicle will sensibly have to undergo. A two-step processing system accounts for both these requirements.


57

Local car recycling businesses dismantle end-of-life vehicles according to their make, model, condition and the demand in the spare parts market, deciding individually what to do with each car. As a compulsory first step, though, all fluids are extracted from vehicles to avoid environmental risks from spillage. Only then spare parts and those components bound for separate recycling are removed (product recycling). Central shredding plants process the stripped car bodies on an industrial scale acting as production plants for metallic secondary raw materials (material recycling) .

3. Processes for recycling of dismantled materials

A modern passenger car contains up to 40 types of plastics. Apart from 20-30 basic synthetic materials, there are compounds, or “blends”, and plastics containing addition materials. Their weight in modern cars amounts to an average 145-160 kilograms and shows a tendency towards further increase. Interior fittings account for more than half of this weight, exterior body parts add up to another quarter. Compared with metal, the production costs of most plastics are rather low. This makes reusing the material hardly economical since, like metals, the different types of plastics must be sorted prior to reprocessing in order to avoid low-quality results. There are no sorting systems for separating different types of synthetic material from end-of-life vehicles after shredding. Plastic components bound for recycling must therefore be removed beforehand 3.1. Hammer mills Hammer mills are not so common in comparison to shredders, which is due to the fact that from the big shredder producers only Thyssen-Henschel Recycling Technik GmbH (in short HRT) from Kassel, Germany, is offering this type of “shredder”. The mills can be compared to very large coffee grinders, with the axle rotating vertically. The scrap to be shredded drops from the top to the bottom, and is then ejected. There is no grate in these mills. The rotor is slightly conical. There are three or four discs on the vertical shaft to which the axles of the “hammers” are attached. But, as with turnings crushers, these “hammers” are alloy steel toothed impact rings with an internal diameter larger than the axle around which they can rotate freely. These impact rings therefore leave space for passing scrap. The further the scrap drops, the smaller the space between the ring hammers and the liners or wear-resistant walls becomes. A number of fixed “wings” are attached to the rotor shaft at the top of the rotor housing, smaller in diameter than the discs with rings. The raw material is torn apart by these wings to prepare it for the grinding action of the rings below. Dedusting takes place at the bottom of the mill, its kick-out door is situated at the top. A secondary dust collection unit is nowadays placed on the zigzag air classification unit through which the scarp falls. 3.2 Hammer mill installation

The process flow chart of the Henschel hammer mill located in Germany is shown in figure 1 . This plant is housed in a warehouse, where the scrap to be processed is stored as well. The zig-zag separator of the plant was intentionally operated with less air (pendulum flap at the feed, fixed in open state), to ensure that the non-ferrous metal losses in the waste product are on a low level. It should be noted that a worker was employed for sorting the non-ferrous metal pre-concentrate by removing large rubber and aluminum pieces from the non-ferrous metal precon-centrate. During the entire sampling period, only automobile scrap was put through. The spot samples (in total 100) were packed into bags.


58

The samples taken during the sampling procedure as well as the sampling parameters are shown in table 1.

Table 1: Sampling procedure Sample material

Sample designation

No. of spot samples

Sampling time (sec.)

Sampling interval

(sec.)

Sample mass total (kg)

Use of samples

Non-ferrous metal pre-concentrate

B1/1-40 40 10 0 130 for examinations regarding sampling

B2/1-40 40 20 300 420 for characterization of material composition

Composite sample S31)

– 10 300 approx. 200 For small-scale analyses, if required

Waste product

B3/1-20 20 3 600 70 for characterization of material composition

Composite sample S4

– 3 600 approx. 100 For small-scale analyses, if required

4 Results of the analysis

As already mentioned before, the analyses of the samples from the non-ferrous fraction in regard to the distribution by piece size and the extent of disintegration of each piece did reveal that the halved values of the piece sizes are at the installation

• shredder about 30 mm and at the installation • hammer mill about 45 mm.

This fact, connected with the reduced disintegration extent of the samples from the non-ferrous fraction from the hammer mill installation indicates that the hammer mill has a smaller shredding effect than the shredder installation. The composition analyses is showing that the magnetic sorting of the hammer mill installation is not operating sufficient enough because in the large-sized fractions are a lot of magnetic pieces, which are “spoiling” the non-ferrous metal fraction. In the waste fractions of both installations, i.e. shredder and mill, are over 50% of all pieces with a size smaller than 4 mm. For a metallurgical utilization only the iron content could be taken into consideration, but the enriching of the iron content to metallurgically interesting values seems at least problematic from the economical point of view.

4.1 Pre-concentrate of non-ferrous metals

The 80 samples taken with their total mass of more than 850 kg were sufficient for the analysis of the non-ferrous metal pre-concentrate from the shredder and from the hammer mill, as a basis for the establishment of metal balances and the development of process suggestions.

Figure 2 show the lump size distribution functions determined for the two non-ferrous metal pre-concentrates , which show the half-value lump sizes for the shredder amounting to 30 mm and for the hammer mill of 45 mm.


59

Fig. 1. Process flow chart of the hammer mill installation

In connection with the higher degree of disintegration of ferromagnetic steel, this

indicates that in the hammer mill, the separation efficiency of the magnetic separator is not high. A reduction of the ferrous content in the hammer mill system, by way of calculation, results besides the increase in the content of non-ferrous metals, in an approx. 10% increase in the non-metal content in the non-ferrous metal pre-concentrate of the hammer mill to approx. 50%. This extremely high non-metal content can be explained by the high content of organic non-metals, as compared with the shredder, in the lump size classes between 4 and 32 mm, which indicates that the air velocity has been set rather low in the air separator of the hammer mill (fig.3.). The ratio of the non-ferrous metal contents between each other is largely the same.

What is essential for the generation of a light-metal concentrate is the ratio between the aluminum content and the magnesium content. The non-ferrous metal pre-concentrate from the shredder has a content of magnesium and magnesium alloys of approx. 11% of the aluminum content as compared with the hammer mill, where the magnesium content is only just under 4% of the aluminum content. In contrast, the non-ferrous metal pre-concentrate from the hammer mill shows a relatively high content of zinc and zinc alloys (fig.4) .


60

Lump Size Distribution Function of Hammer Mill

0,00

20,00

40,00

60,00

80,00

100,00

-2 mm

2-4 mm

4-8 mm

8-16 mm

16-32 mm

32-63mm

63-125mm

125+ mm

lump size classes

perc

enta

ge mean value

lower limit ofconfidence levelupper limit ofconfidence level

Contents of non-ferrous metal pre-concentrate from hammer mill

17,9%

1,5%

2,0%

0,7%

3,4%

1,4%

33,3%

39,8%

aluminium in total

copper in total

copper alloys

magnesium and Mg-alloys

zinc and Zn-alloys

lead and Pb-alloys

iron and steel in total

non-metals in total

Fig. 2. Lump size distribution function of the non-ferrous metal pre- concentrate from hammer mill.

Fig. 3. Contents of essential components of the non-ferrous metal pre- concentrate from the hammer mill in total.

An enrichment of individual non-ferrous metals in individual lump size classes, which

might be exploited technically, cannot be seen, although the maximum contents of the individual non-ferrous metals occur in different lump size classes:

- copper: < 4 mm - lead and lead alloys: < 8 mm - copper alloys: 4 mm to 16 mm - cast aluminum alloys: 16 mm to 32 mm - zinc and zinc alloys: 16 mm to 32 mm - magnesium and magnesium alloys: 32 mm to 63 mm - malleable aluminum alloys: > 32 mm.


61

Contents of non-ferrous pre-concentrate from hammer mill

0,00

10,00

20,00

30,00

40,00

50,00

60,00

70,00

80,00

90,00

-2 mm

2-4 mm

4-8 mm

8-16 mm

16-32 mm

32-63mm

63-125mm

125+ mm

lump size classes

perc

entage

aluminium in total

copper in total

copper alloys

magnesium and Mg-alloyszinc and Zn-alloys

lead and Pb-alloys


non-metals in total

Fig. 4. Contents of essential components of the non-ferrous metal pre-

concentrate from the hammer mill in relation to the size class.

The same results are visable in the course of recovery of the individual components as seen in Figures 3 and 4 and. In addition, there are the basic differences mentioned above. Furthermore, ferromagnetic steel must be excluded, because its far higher degree of disintegration in the non-ferrous metal pre-concentrate from the hammer mill is not due to shredding advantages to be attributed to the hammer mill, but, as explained above, to the low efficiency of its magnet separator (fig.5.). Ideal magnet separation should not leave any disintegrated ferromagnetic ferrous materials in the non-ferrous metal pre-concentrate of such shredding plants. The course of the degree of disintegration shows the expected tendency, which means, the bigger the lump size, the lower the degree of disintegration. The components zinc, zinc alloys, and non-ferromagnetic steel are in contradiction with this tendency. With respect to these components, the decrease in the degree of disintegration can only be seen up to the medium lump size classes; the highest lump size classes, instead, show a significant increase in the degree of disintegration. When reflecting about increasing the degree of disintegration, and thus the quality of non-ferrous metal concentrates to be produced from the non-ferrous metal pre-concentrates, from a process engineering point of view, one must consider that, except for lead and lead alloys, more than 50%, in the case of magnesium and magnesium alloys even 100%, of the metal contents present in compounds are compounds with ferromagnetic ferrous materials . This can be explained by the predominance of ferromagnetic (plain) steel as the material employed for standard parts. This content of ferromagnetic steel in the compounds can be used for re-circulation of the compounds concerned to the shredding machine through magnet separation, which will presumably result in an increase in the degree of disintegration.

Owing to their specific use in starter batteries, lead and lead alloys are almost exclusively found as compounds with the non-metal textile (NM paper) and plastic (NM plast) materials in the medium and higher lump size classes.


62

Recovery of essential components from hammer mill

0

10

20

30

40

50

60

70

80

90

100

-2 mm

2-4 mm

4-8 mm

8-16 mm

16-32 mm

32-63mm

63-125mm

125+ mm

lump size classes

reco

very

in p

erce

ntaluminium in total

copper in total

copper alloys

zinc and Zn-alloys

magnesium and Mg-alloyslead and Pb-alloys


non-metals in total

Fig. 5. Recovery of the essential components from the non-ferrous metal pre-concentrate of the hammer mill.

Degrees of Disintegration for Hammer Mill

0

10

20

30

40

50

60

70

80

90

100

-2 mm

2-4 mm

4-8 mm

8-16 mm

16-32 mm

32-63mm

63-125mm

125+ mm

lump size classes

perc

enta

ge

aluminium in total

copper in total

copper alloys

magnesium and Mg-alloyszinc and Zn-alloys

lead and Pb-alloys

iron and steel intotalnon-metals in total

Fig. 6. Degrees of disintegration of the essential components of the non-ferrous metal pre-concentrate from hammer mill.

The less steep course of the lump size distribution function of the waste from the hammer mill, together with the contents of the individual components shown below (fig. 6 ), indicates that the separator in the hammer mill is operated at a very low air speed. The metal content of the waste is made up of fragments of brittle alloys in the fine lump size range (e.g. isolated fragments of cast iron in the lump size range 4 to 8 mm) and thin-walled pieces


63

Lump size distribution function of the waste from hammer mill

0102030405060708090

100

-2 mm

2-4 mm

4-8 mm

8-16 mm

16-32 mm

32-63mm

63-125mm

125+ mm

lump size classes

perc

entage

mean

lower limit ofconfidence level

upper limit ofconfidence level

(wires, thin sheets) in the high lump size range. In addition, it must be considered that the content of lead and lead alloys increases

with the decrease in the lump size, which is due to grinding or abrasion of this soft metal (fig.7.). These globally described tendencies more significantly show in the waste from the shredder than in that from the hammer mill owing to the higher air speed in the separator of the former.

Fig. 7. Lump size distribution function of the waste from hammer mill. 5. Conclusions 10 Based on the assumption that the shredder yields the product masses per year are as

follows: 100.000 tons/ year – steel finished product, 5480 tons/year – non-ferrous metal pre-concentrated and 31500 tons/year – waste product;

20 Based on the assumption that the hammer mill yields, the product masses per year as follows: 25000 tons/year steel finished product; 3080 tons/year non-ferrous metal pre-concentrated and 14.400 tons/year – waste product;

30 The analysis of the samples from the non-ferrous fraction in regard to the distribution by piece size and the extend of disintegration of each piece shown that non-ferrous fraction from the hammer mill installation has a smaller shredding effect than the shredder installation;

40 Contents of essential components of the non-ferrous metal pre-concentrate from the shredder are: 38% aluminum; 13.2% copper; 1.9% copper alloys; 2.5% Mg and Mg-alloys; 3.7% Zn and Zn alloys, 4.6%Pb and Pb alloys; 32,3% steel and steel alloys and 3.8% non-metallic compounds;

50 Contents of essential components of the non-ferrous metal pre-concentrate from the hammer mill are: 39.8% aluminum, 33.3% copper; 1.4% copper alloys; 3.4%Mg and Mg alloys; 0.7% Zn and Zn alloys; 2,0% Pb and Pb alloys; 17.9% iron and steel and 1.5% non-metallic compounds; 60 The metal content of the waste is made up of fragments of brittle alloys in the fine lump size range and thin-walled pieces in the high lump size range.


64

References [1] Amza, Gh., Garac Zlatko - Recycling techniques for materials used in automobile industry, International Conference TQSD -08, May 2008, Bucharest, Romania. [2] Ando, G.; Kaoru, S.; Automobile shredder residue treatment in Japan, Proceedings of International Automobile Recycling Congress, Geneva, Switzerland, March 2002. [3] Andres, U.; Method of Preparation Of heavy media for mineral separation by specific gravity, Russian Patent No. 177365, 1964. [4] EUWID; Europäischer Wirtschaftsdienst GmbH, Recycling und Entsorgung, Deponiepreise für Shredderabfälle, no. 21, Gernsbach, Germany, 26.05.1999, p. 14. [5] Feillard, P.; ELV Directive implementation in Europe, Proceedings of International Automobile Recycling Congress, Geneva, Switzerland, March 2002. [6] Feistner: Rückgewinnung von NE-Metallen mit Wirbelstromseparatoren, Metall 37 (1989, 2, S. 18.

[7] Schenk, M.; Altautomobilrecycling, Technisch-ökonomische Zusammenhänge und wirtschaftspolitische Implikationen, Gabler Verlag, Wiesbaden, Germany, 1998. [8] Schmidt C.F.; NE Separation am Beispiel einer Schwimm-Sink-Anlage, IKH, Hannover, Germany, 1993. [9] Schmidt, J.; Altautoverwertung und –entsorgung. Stand und Aktivitäten beim Altautomobilrecycling, Expert Verlag, Renningen, Germany. 1995. [10] Schmidt, J.; Leithner, R.; Automobilrecycling. Stoffliche, rohstoffliche und thermische Verwertung bei Automobilproduktion und Altautorecycling, Springer Verlag, Berlin, Germany, 1995.


65

SOME ASPECTS REGARDING COMPUTER AIDED DESIGN OF CAM

MECHANISMS WITH FLAT FACE OSCILLATING FOLLOWER

Associate Professor Stelian ALACI, Suceava University, [email protected] Lecturer Florina Carmen CIORNEI, Suceava University, [email protected] Professor Dumitru AMARANDEI, Suceava University, [email protected]

Associate Professor Constantin FILOTE, Suceava University, [email protected] Assistant Simion PĂTRAŞ-CICEU, Suceava University, [email protected]

Lecturer Dorel PRODAN, Suceava University, [email protected]

Abstract: The paper presents some aspects concerning the design of cam mechanisms with oscillating flat faced follower, using computer softwares. There are deduced the analytical equations for the profile of the cam, thus enabling the study of mechanism’s reliability. There are emphasised the effects due to an inadequate choice of law of motion, mainly shocks during mechanism operation. Keywords: cam mechanism, oscillating flat face follower, FEM analysis 1. Introduction

Cam mechanisms are machines widely used in technical applications due to reduced constructive complexity, [1], [2], consisting of only two mobile elements, the cam as mechanically driver element and the follower as driven element. The main advantage of these mechanisms is that, by an appropriate cam design, the driven element can carry out any imposed motion. The major disadvantage of these mechanisms consists in the presence of the cam-follower higher pair. This kinematical joint is characterised by occurrence of contact pressures that subjects the materials to important stresses in the immediate vicinity of the contact. In addition, the vast majority of mechanisms have a periodic motion, conducting to periodic loading for both cam and follower and therefore to fatigue phenomenon.

In the case of mechanisms with flat face follower, the cam-follower interaction presents also a tangential component. This component is directly influenced by the contact pressure from the higher pair and can produce adverse effects such as wear of elements and from this, cam profile modification with direct consequence, the alteration of the law of motion. A key parameter for these mechanisms is the minimum radius of curvature allowed in the cam. This parameter presents a direct influence upon both contact pressure and friction force from the higher pair. The value of the minimum radius of curvature is directly ordered by the follower’s law of motion and by the main geometrical parameters. The main geometrical parameters are constant characteristics that together with the law of motion control the cam’s size and shape. For the cam mechanism with translating flat-faced follower the relation is straightforward and is given in all monographies on theory of mechanisms, [3] and for the knife-edge follower, the authors found these relations, [4], [5].

For the mechanisms with oscillating flat faced follower and revolving cam, this approach is not possible as a very complicated expression results and therefore, it is more convenient to find the minimum radius of curvature after setting the cam profile.


66

2. Tracing the profile of the mechanism with rotating cam and oscillating flat-faced follower The mechanism is presented in Figure 1 and the geometrical and kinematical parameters of the mechanism are shown in Figure 2. For the lowest position, the follower is tangent to the circle with minimum radius of curvature, in the point 0T . The main geometrical parameters of the mechanism are:

− the distance d between the two joints, of the cam and of the follower; − the angle 000TOA Ψ= , produced by the follower direction in the lowest position and the axis of the joints. The cam rotates with a constant angular velocity, ω . At a certain moment, the cam’s

position is characterised by the angle of rotation, ϕ while the position of the follower is given by the angle )(ϕψ=ψ .

Fig.1. Mechanism with rotating cam and

oscillating flat-faced follower

Fig.2. Geometrical and kinematical parameters of the

mechanism

For a kinematical study of the motion, it is more convenient to perceive the motion of the entire mechanism to the plane of the cam. Consequently, the cam becomes fixed and the follower, besides its own motion, gets a rotation transport motion around the centre of the cam with the angular speed ( ω− ). The same method is used in finding the profile of the cam. The follower is supposed to perform two motions: rotation around the centre of the cam and relative oscillation against the frame. The cam’s profile results as an envelope of the successive positions of the follower. From Figure 3 it can be noticed that for the same geometrical parameters and law of motion, two cam profiles can be obtained. The two positions corresponding to the follower,

1Δ and 2Δ , can be regarded as two variable curves, having the position depending on the parameter ϕ . In the cam’s co-ordinate system, each of these takes the form of general equation:

0),y,x(f =ϕ (1)

ω −ω

ϕ

)(ϕψ

0T

T

0A

O

A

d0ψ


67

According to Ionescu, [6], in order to obtain the envelope of the curve family (1), the parameter ϕ should be eliminated from the system:

⎩⎨⎧

=∂∂=

0/),y,x(f0),y,x(fϕϕ

ϕ.

(2)

With the above system solved concerning variables x and y , these would represent factual the parametric co-ordinates of the envelope. The equations of the two straight lines can be written in a concentrated manner:

0)]}([()tan{(sindy 0 =+±−−+ ϕψψϕπϕ (3) where the sign ""+ corresponds to the straight line 1Δ and „-“ to 2Δ , respectively.

Fig.3. The two possible cam profile generations

For the particular case of system (2) for the equation (3), solving the system about x and y , the equations of the cam profiles generated by the two straight lines are obtained as:

⎪⎪⎪

⎩

⎪⎪⎪

⎨

⎧

+±−+−−=

+±−+−=

ϕϕψ

ϕψψϕϕψψϕ

ϕϕϕψ

ϕψψϕϕψψϕ

ϕ

d)(d1

))]((sin[)](cos[sin

d)(y

d)(d1

))]((cos[)](cos[cos

d)(x

002,1

002,1

m

m

; (4)

Using the MATHCAD software, a programme was written with the relations (3) and (4)

to obtain the profiles of the cams as envelopes of the straight lines (3). The result given by the programme is presented in Figures 4 and 5. From Figure 5 it results that the undercut occurs on the cam profile and therefore it cannot be used.

ϕ−

0ψ

0ψ )(ϕψ

)(ϕψ

O

A

x

y)( 2Δ

)( 1Δ

d


68

Fig.4. Tracing profile of the cam, "1"

Fig.5. Tracing the profile of the cam, "2"

3. Finding the radius of curvature

In order to use the cam in a mechanism with flat-faced follower, the obtained cam profile should look similar to the one obtained in Figure 4. To characterize unequivocal the concavity of the profile of the cam, the definition of concavity reported to the centre of the cam is requested, using as pole the rotation centre of the cam. In a point, the profile is concave if the pole and the curve are situated on the same side of the tangent in the considered point. As the profile of the cam can present straight line regions having the radius of curvature ∞=cρ , this aspect can be avoided by introducing the curvature, defined as the inverse of the radius of curvature, c/1k ρ= . In polar co-ordinates the relation for the curvature is given by:

2 3 2

2 2 2 3/ 2

' " ' 2 ' ' ' "[ ' ' ]

t t t t t t t t t

t t t

r r r r r rk

r rϕ ϕ ϕ ϕ

ϕ+ + +

=+

, (5)

where tr and tϕ are the polar co-ordinates of the contact point between the profile of the cam and the follower and the symbols (‘) and (“) represent the first and the second derivative, correspondingly, as regards angle ϕ .

The complexity of the relation makes impossible to express the curvature as function of geometrical and kinematical parameters of the mechanism. For this reason, the equations (4) for the cam’s profile are used and next, by applying the relation (5), the curvature is found.


69

As an example, it is considered a mechanism with continuous acceleration, on regions, for ascending phase and with sinusoidal acceleration for descending. The variation of the follower rotation, )(ϕψ , of reduced angular velocity 12 /)( ωϕω , and of reduced angular

acceleration 212 /)( ωϕε are presented in Figure 6. In Figure 7 there are presented the

variations of the curvatures for the two profiles, considered with changed sign.

Fig.6. Kinematical parameters of the follower for a complete cam rotation

Fig.7. Variation of radius of curvature for the two profiles

From Figure 7 it can be observed that for finite discontinuities of the follower’s

acceleration, the radius of curvature will also present consequent discontinuities. In addition, the minimum radius of the cam represents the minimum radius of curvature. The presence of acceleration discontinuities will induce shocks in mechanism’s operation and finite variations of the radius of curvature for the cam. This fact adds more complexity to the calculus for contact strength of the mechanism. In the theory of line Hertz contact, the contact pressure is calculated based on the assumption of continuous variation of the radius of curvature, [7]. Further more, for the deformation calculus, there are not available general analytical relations. The calculus of the deformations is made either via empirical relations or by relations applicable for narrow domains of the variables.

Fig.8. Contact normal stress variation from FE analysis a) cam – follower contact; b) extreme stresses in cam; c) extreme stresses in follower

With the purpose of illustrating the effect of finite variation of the curvature radius of the cam, it was considered the contact between an elastic half-plane and an indenter with radii

mm50R1 = and mm200R2 = , loaded by a normal force N1000Q = .

a b c


70

The thickness of the two elements is the same, mm4.0 .

Fig.9. a) deformed shape of the follower; b) local deformation of follower; c) local deformation of cam 4. Conclusions

The present work presents the methodology of obtaining the profile of the cam from mechanisms with flat-faced oscillating follower. The profile of the cam results as the envelope of the positions of the follower performing relative motion with respect to the cam. The parametric Cartesian equations of the cam are obtained. The relation for calculus of radius of curvature is presented, expressed in polar co-ordinates.

The relation is applied for a given case, when for rising phase, the acceleration presents finite discontinuities. This fact is reflected into geometry as finite discontinuities for the radius of curvature of the cam and by dynamic point of view, into shocks generated during mechanism’s operation. The effect of finite variation of the contact radius upon deformations and stresses from the cam-follower contact region represents a special case, analyzed via FEM. The analysis is especially useful for Hertzian line contact since there are no analytical relations for the characterization of contact deformations. AKNOWLEDGEMENT: This paper was supported by the project "Progress and development through post-doctoral research and innovation in engineering and applied sciences – PRiDE - Contract no. POSDRU/89/1.5/S/57083", project co-funded from European Social Fund through Sectorial Operational Program Human Resources 2007-2013. References [1].Gonzales-Palacios M.A., Angeles, J., Cam Synthesys (Solid Mechanics and Its Applications), Springer 1993; [2].Angeles, J., C.S. López-Cajún, Optimization of Cam Mechanisms, Springer 1991; [3].Uicker, J. , Pennock, G., Shigley, J., Theory of Machines and Mechanisms, Oxford University Press, 2010; [4].Alaci S., Computer Aided Design of Cam Mechanisms with Translating Roller Follower, PRASIC 2006, 9-10 noiembrie, 2006, Braşov, Vol. II, Mecanisme şi Tribologie, ISBN (10)973-635-824-0, (13) 978-973-635-821-1; [5].Alaci, S., Rusu, O., Computer Aided Design of Radial Cams and Oscillating Roller Follower, Annals of the Oradea University, Fascicle of Management and Technological Engineering, Vol.V(XV) 2006, ISSN 1583-0691; [6].Ionescu , Gh, D., Teoria diferenţială a curbelor şi suprafeţelor cu aplicaţii în tehnică, Ed. Dacia, Cluj-Napoca, 1984;

a b c


71

SIMULATION MECHANISMS ECCENTRIC PRESS AND GEAR USING SOFTWARE ADAMS FOR REDUCING MATHEMATICAL

CALCULATIONS AND THE TIME ASSOCIATED WITH THESE

Ş.l.dr.ing. Alin STĂNCIOIU, Universitatea Constantin Brâncuşi Tg-Jiu Prof.univ.dr.ing. Gheorghe POPESCU, Universitatea Constantin Brâncuşi Tg-Jiu

Prof.univ.dr.ing. Cătălin IANCU, Universitatea Constantin Brâncuşi Tg-Jiu

Abstract: The development of the approximate blank size should be done to determine the size of a blank to produce the shell to the required depth and to determine how many draws will be necessary to produce the shell. This is determined by the ratio of the blank size to the shell size. Various methods have been developed to determine the size of blanks for drawn shells. These methods are based on algebraic calculations; the use of graphical layouts; a combination of graphical layouts and mathematics.

Key words: MSC ADAMS, presses, eccentric

1 Introduction

Presses can be evaluated with regard to their function as: • Energy-producing machines, with the application of this energy being abrupt and

instantaneous. All the machine’s energy storage is depleted at the end of its work cycle. Such machines are hammers, which use a “free fall” principle as a basis of their function. However, sometimes their efficiency is increased by an addition of steam or pressurized air. The hammers’ energy source is temporarily disconnected at the time the ram is released for an operating cycle, which consists of its falling down on the workpiece.

• Force-producing machines, which operate by generating a considerable amount of force, independent of the position of the ram. Hydraulic machinery is the main representative of this category.

• Stroke-controlled machines, or mechanical presses, the function of which depends on the movement and location of the ram. During the work cycle, the ram is always in contact with the source of its power. These presses can be further divided as:

• Presses driven by a crank or eccentric drive. Simple crank drives are the most commonly used types, with extended crank drives in knuckle-joint presses.

• Presses driven by a cam. Cam-driven systems are used in presses with less tonnage. Each of these groups has certain advantages and disadvantages closely connected with the type of process they represent. For example, hydraulic presses have fewer operating parts, which brings the cost of their repairs down. However, should these machines be in need of any repair, such a procedure will be very demanding. With mechanical presses their breakdowns are visually detectable, but a complete knowledge of the circuit is required to find a problem within a hydraulic system. Also the tolerance range of hydraulics is not as impressive as that of mechanical presses, with the latter group also being faster. Fortunately, each type’s usefulness is given by the type of its applications, and that’s where their function, along with its advantages and drawbacks, fits (fig. 1.).


72

Fig.1. Various types of press drives

2. Simulation with ADAMS trial Adams is the most widely used multibody dynamics and motion analysis software in

the world. Adams helps engineers to study the dynamics of moving parts, how loads and forces are distributed throughout mechanical systems, and to improve and optimize the performance of their products.

Traditional “build and test” design methods are now too expensive, too time consuming, and sometimes even impossible to do. CAD-based tools help to evaluate things like interference between parts, and basic kinematic motion, but neglect the true physics-based dynamics of complex mechanical systems. FEA is perfect for studying linear vibration and transient dynamics, but way too inefficient to analyze the large rotations and other highly nonlinear motion of full mechanical systems.

Adams multibody dynamics software enables engineers to easily create and test virtual prototypes of mechanical systems in a fraction of the time and cost required for physical build and test. Unlike most CAD embedded tools, Adams incorporates real physics by simultaneously solving equations for kinematics, statics, quasi-statics, and dynamics. Utilizing multibody dynamics solution technology, Adams also runs nonlinear dynamics in a tiny fraction of the time required by FEA solutions. Loads and forces computed by Adams simulations improve the accuracy of FEA by providing better assessment of how they vary throughout a full range of motion and operating environments.

In the study were simulated using ADAMS software trial presses mechanical components of two mechanisms, namely: eccentric mechanism and gear.

3D images ofthe two mechanisms are illustrated in Figure 2 and graphs speed,acceleration, potential and kinetic energies are shown in Figures 3 ... 6.


73

a). Eccentric mechanism b). Gear mechanism Fig.2.The spatial modeling

a). Eccentric mechanism b). Gear mechanism

Fig. 3. Velocity graphics


Fig.4. Acceleration graphics


74


Fig.5. Potential energy graphics

a). Eccentric mechanism b). Gear mechanism Fig.6. Kinetic energy graphics

3. Concluzions:

Analyzing the graphs of the two mechanisms can be seen that : - the ram speed press with eccentric mechanism is much larger than the press ram

with gear ; - cycles of chart velocity are more differentiated gear; - at the bottom of the ram stroke is noticed a decrease in speed which is desirable in achieving deep drawing operations;

- the potential-energy acceleration and the press ram gear are lower than those with mechanical eccentric press ram..

By the method of simulation mechanisms mechanical presses cold compressing them using ADAMS software greatly simplifies the mathematical calculations and also reduces time associated with them.

This method is part of new methods for investigating the behavior of different bodies are in motion. References [1]. J. Matej, Tracked mechanism simulation of mobile machine, Department of Forest and Mobile Technology, Faculty of Environmental and Manufacturing Technology, Technical University in Zvolen, Zvolen, Slovak Republic, Vol. 56, 2010, No. 1: 1–7; [2]. Ivana Suchy, Handbook of Die design, Michigan, 1990; [3]. Alin Stăncioiu, Gheorghe Popescu, Proiectarea Moderna A Dispozitivelor De Presare La Rece, 3nd Symposium“Durability and Reliability of Mechanical Systems” Târgu-Jiu, Romania, 20, 21 may 2010;


75

SOFTWARE APPLICATIONS FOR FIELD RELIABILITY DATA

Dr.ing. Adrian Stere PARIS, Univ. Politehnica Bucharest Abstract The paper details the peculiarities of processing the field reliability data to find the reliability function or the cumulative distribution function assisted by the regression analysis. An overview of the application software for statistical and reliability calculi, particularly for field data, is presented with the main characteristics.

Key words: field data reliability, software applications 1.Introduction The ultimate test of a manufactured product is how well it performs in the field, in the hands of the customer. Accordingly, the collection and analysis of data on the field performance or reliability of products is important to manufacturers and consumers alike [13]. Such data can be used in many ways by a manufacturer, including (i) to asses field reliability and make comparisons with engineering predictions, (ii) to provide information for product modification and improvement, (iii) to asses the effects of design changes, (iv) to estimate and explain warranty costs, and (v) to aid in the design of warranty, maintenance and parts replacement programs[14]. 2. Field reliability data Nevertheless, many manufacturers pay insufficient attention to the collection of field performance data. One reason is that comprehensive data are often seen as expensiv to obtain; another may be a lack of familiarity with methods for response-selective observational schemes and for combining information from different sources[13]. Foe some problems methodology indeed needs to be developed. This paper deals with reliability, which is an important component of field performance [14]. Reliability prediction models depend on observed lifetime data. There are several means of acquiring lifetime data for a system such as artificially stressing a system, initial laboratory tests and the like. Most data sources have some sort of bias, and the historic military preference has been for field data: information acquired by observing the lifetime of components in their normal use [12]. Prediction models depend on this data for several reasons:

- the prediction formula are themselves derived from field data, without some idea of the natural life spans of the components.

- engineers building new components plug existing field data into their derived prediction models to make an estimate. Field data is one of a group of data types that can be used, but it is usually the most thorough [2].

Acquiring field data is a relatively onerous task: by definition, field data is gathered by observing parts fail in situ. A well designed part is less likely to have a long life time, leading to extended waiting time for any useful information. Because the task is so time consuming, there are relatively few sources, usually from the manufacturers themselves [2]. Figure 1 illustrates the problem of matching a reliability or risk prediction to the eventual field performance. In practice, prediction adresses the component-based ‘design reliability’ and is necessary to take account of the additional factors when assessing the integrity of a system.


76

1.Failure Data types A Reliability data set can contain complete data, right censored data, interval censored data and left censored data. An overview of this is presented in fig. 2. The common case is obvious for the first variant (fig. 2), respectively non-grouped data, all failed, exact time to failure. Even this is not the present case; it is preferred for the facilities of

Fig.1 The complete chain of reliability data processing (only the packages with

specialized reliability applications enable censured data), the result being a distorted information. 3. Processing field reliability data and software packages Even the simplest computer aided processing of field reliability data becomes difficult without specialized applications: frequently it starts with times to failures (TTF – for non- repairable products) or times between failures (TBF-for repairable products), generally ti. For the regression analysis data couples (xi, yi) are necessary. As such it should be determined t

Fig.2. Classification of reliability data types the yi values.

The reliability specialized software provide this automatically, but in the case of general statistical packages, the allocations of yi values should be manually accomplished using ranking (mean or median rank), based on the ordered range ti (empirical cumulative distribution frequency is approximated).


77

There are two approaches to fit reliability distribution to the failure data: • derive directly from the data an empirical reliability function • a theoretical distribution, such as exponential, Weibull, normal, etc.[1;15], the most

preferred approach An overview of statistical packages with a wider application area for field data is given in Table 1[4], (here excerpts for a few packages). Price note [1] indicates a promotional one (higher prices are for current purchases); note [2] indicates that lower/penetration pricing is offered to academic purchasers. (Tab.1Examples of statistical packages)

Product

Example(s) Developer Latest

versionCost

(USD) Software license

Interface

Written in

Scripting languages

Minitab Minitab Inc.May 18,

2010

$895–$1395[2],

$542 annual, $30/….

academic

Proprietary CLI/GUI

Origin OriginLab $699 Proprietary GUI

R R Foundation

2011 Free GNU GPL CLI/GUI

[6] C Perl by Statistics Rmodule

SAS

SAS Institute

March 2008

Ac+stud free Com. ~$6000

Proprietary CLI/GUI

SPSS IBM 2007 $1599[2] Proprietary CLI/GUI Java Python

STATISTICA StatSoft 2010 >$695 Proprietary GUI

StatPlus AnalystSoft 2007 $150[1][2] Proprietary GUI

XLSTAT Addinsoft 2009 $395[2] Proprietary Excel


78

A detailed analysis of software special applications is developed in Table 2. It is visible the connection with reliability data in the column “Survival analysis”. An example of the utilization of such programs is presented in fig. 3 (for Minitab 15) [16]. (Tab.2 Examples of statistical packages with reliability connection).

Product s/w type

[18]

Descriptive statistics Nonparametric statistics

Qualitycontrol

Survivalanalysis

Data processing

Base stat.[19]

Normality tests[20] CTA[21] Nonparametric

comparison, ANOVAClusteranalysis

Discriminant analysis BDP[22] Ext.[23]

Mathemathica S + + + + + + + + + +

Minitab S + + + + + + + + + +

Origin S + + ‐ ‐ + +/‐ ‐ ‐ + +

R St + + + + + + + + + +

STATISTICA S + + + + + + + + + +

StatPlus S + + + + + + ‐ ‐ + +

SPSS S + + + + + + + + + +

XLSTAT X + + + + ‐ + + + N/A +

where: 18 means S = Standalone executive; St = Standalone executive, primitive textual (DOS or terminal) interface; A = Access Add-in; X = Excel Plug-In ;19- Base Statistics (such as t-test, f-test, etc.); 20- Normality Tests, data exploring; 21-Contingency Tables Analysis; 22-Base Data Processing, f.ex. sorting; 23- Extended (data sampling, transformation) 4. Reliability software packages The extended importance of reliability and the diversification of applications has determinates the development of reliability software packages, with main role to give an extended application of reliability techniques. A few will be mentioned below: 1.Weibull++7, offered by ReliaSoft Corporation [8] (Reliasoft Single user 1165 $) provides a toolset available for reliability life data analysis; other packages are for Accelerated Life Data Analysis (ALTA - from 4500$), System Analysis Software using RBDs or Fault Trees (BlockSim – from 3150$), Software for Reliability Growth Analysis and Repairable System Analysis (RGA – from 3150$), Reliability Centered Maintenance Software (RCM – from 4500$), etc., more than 200 software packages for FMEA, maintenance, DOE, etc. Supplementary are organized courses, training and consulting services; 2.Reliability Workbench 11, offered by Isograph Ltd [10], is a complex software package, with applications to maintainability prediction, FMECA and FMEA, FaultTree+ Fault Tree Analysis, Reliability Block Diagram analysis, Reliability Allocation and Growth, Event Tree and Markov Analysis, Weibull Analysis of historical failure data, Parts Libraries, etc.


79

3.Reliass Inc. [11], offers ASENT toolkit- reliability and maintainability computer-aided engineering solutions, EAGLE - an enhanced integrated logistics support software system, the Advanced Integrated Maintenance Support System (AIMSS), LOGAN- Fault and Event Tree module, LOGAN Monte Carlo analysis module, Raptor -Reliability Simulation Tool, Plant Availability Modelling with RAMP, etc. 4.DfRSoft (Design for Reliability) [9] provides low priced design (only 249$, based in Excel) for reliability software with applications for Reliability Plotting, System Reliability Analysis, Field Return Analysis, Accelerated Reliability Growth Multi Test Modules and Traditional Growth Analysis, Design-FMEA With Failure Mode Look-Up Table, Distributions and Confidences, Cpk Assessment, etc.

Fig.3. Cumulative density function plot [16] Fig.4. Vapor pressure failure function: [19] 5. Regression and curve fitting for reliability data Regression is a conceptually simple technique for investigating functional relationship between output and input decision variables of a manufacturing process and may be useful for process data description, parameter estimation, and control [17]. A large body of techniques for carrying out regression analysis has been developed. Familiar methods such as linear regression and ordinary least squares regression are parametric, in that the regression function is defined in terms of a finite number of unknown parameters that are estimated from the data. Nonparametric regression refers to techniques that allow the regression function to lie in a specified set of functions, which may be infinite-dimensional [6]. Curve fitting is the process of constructing a curve, or mathematical function, that has the best fit to a series of data points, possibly subject to constraints [5]. All the important statistical packages offer a regression module [19], but there are also low cost solutions. Of course, the use of regression and curve fitting to determine the reliability from field data suppose a well known practical situation, to choose the adequate model. A first accessible opportunity to calculate the regression curve for the reliability field data (fig.4) is the software CurveExpert, a comprehensive curve fitting system (freeware, Curveexpert professional for Win 70$)[3]. A similar opportunity is LABFit [7], a curve fitting software, with nonlinear regression - least squares method, Levenberg-Marquardt algorithm -, almost 500 functions at the library with one and two independent variables, functions finder, etc. (shareware, professional 85$).


80

6. Conclusions A few elements of field reliability data processing are presented in the paper. A concise overview of statistical software with life data analysis is developed with extension to specialized reliability software. The usual costs of these packages are also detailed. A final choice for adequate software for field reliability data processing should consider the customer conditions for accuracy, data volume and necessary investment. As an example, for a few data samples and less details it is cheap to use low cost or freeware/shareware software packages. References [1]. Abernethy, R., B. - The New Weibull Handbook, 4th ed. Gulf Publishing Co., AT Houston, Texas 2005. [2]. Collins, M. - Error Prediction Models and Field Data, Carnegie Mellon University Spring 1998 (http://www.ece.cmu.edu/~koopman/des_s99/traditional_reliability/index.html) [3]. http://curveexpert.software.informer.com/1.3/ [4]. http://en.wikipedia.org/wiki/Comparison_of_statistical_packages#General_information [5]. http://en.wikipedia.org/wiki/Curve_fitting [6]. http://en.wikipedia.org/wiki/Regression_analysis [7]. http://webcache.googleusercontent.com/search?q=cache:http://www.labfit.net/ [8]. http://weibull.reliasoft.com/?gclid=CJ3h98LRy6gCFRWazAod5yvjqw [9]. http://www.dfrsoft.com/ [10]. http://www.isograph-software.com/index.htm [11]. http://www.reliability-safety-software.com/products/product_asent.htm [12]. http://www.theriac.org/ (EPRD) [13]. Lawless, J.,F. - Statistical Models and Methods for Lifetime Data, 2nd edition. John Wiley and Sons, Hoboken (2003). [14]. Lawless, J., F., Kalbfleisch, J., D., - Some Issues in the collection and analysis of field reliability data. In: Survival analysis: state of the art, vol.211, NATO ASI Series, Applied Sciences, NATO Advanced Study Institute, NATO Science Series E, Editors, Klein, J., P., Goel, P., K., Springer Verlag, 1992, pp141-152 [15]. Liu, C., C. - A Comparison between the Weibull and Lognormal models used to analyse the reliability data, Ph.D. Thesis (University of Nottingham 1997) [16]. Paris, A., S. - Models in mechanical reliability data 3rd Symposyum „DURABILITY AND RELIABILITY OF MECHANICAL SISTEMS” Univ C. Brancusi Tg. Jiu, mai 2010, In: Fiability and Durability, no. 1(5)/2010, Editura “Academica Brâncuşi” , Târgu Jiu, ISSN 1844 – 640X p.81-86 [17]. Paris, A,. S., Târcolea, C., - Regression models applied to manufacturing systems, (2010), Proceedings in Manufacturing Systems, Editura Academiei Române, vol.5, nr.4, pp. 249-253. [18]. Smith, D., J. - Reliability, maintainability and risk: practical methods for engineers, Ed. Butterworth-Heinemann, Oxford, UK, 2005 [19]. Târcolea, C., Paris, A., S., , Discriminant Analysis and Applied Regression UPB, BSG Proceedings 18, DGDS-2010, Geometry Balkan Press, 2011, pp. 93-98


81

STUDY THE CHARACTERISTICS OF SMALL AND VERY SMALL

SPAN WINGS, USED ON SHIPS

Professor Engineer PhD. Beazit ALI Professor Engineer PhD. Gheorghe SAMOILESCU “Mircea cel Bătrân” Naval Academy

Abstract: This scientific work presents the way in which the small, and very small span wings can be obtained starting from the great span wings and using the two scales of the similarity theory. Basing on two scales model it can transcribe from model at nature the coefficients xc , yc and lengthening λ of Gottingen - 612 profile. Keywords: wing, similarity theory, two scales model, elongation, span.

1. Introduction

In the following we will set out the coordinates (polars) of wings of small and very small elongation, which can be obtained starting from the coordinates of the wings of big elongation, if the theory of similitude is used at two scales. In order to obtain this it is necessary to transcribe from model to nature the coefficients cx, cy and elongation λ, according to the model at two scales of the wing. 2. Transcription of the coefficients cx, cy and cm according to the model at two scales

In the case of the rectangular wing in a plane, having string c constant all along the span (in this case the wing is called aerodynamically twisted with the angle of attack variable along the span) the surface of the wing is determined with the relation:

lcS ⋅= (1) The relative elongation of the wing is:

cl

cll

Sl

=⋅

==22

λ (2)

Since Kl = Kz şi Kc = Kx the elongation scale is: mn K λλ ⋅= 1

Since Kl = Kz şi Kc = Kx the scale of the elongation is:

1KKK

KKK

m

n

x

z

c

l ====λλ

λ (3)

from which results the relation of elongation transcription:

mn K λλ ⋅= 1 (4) K1-ratio of distort If we write the wing’s bearing force like:

2

2vScR yy⋅

⋅⋅=ρ

(5)

We have: 22vxzcvScRR KKKKKKKKKKK

yyy⋅⋅⋅⋅=⋅⋅⋅== ρρ (6)

The scale of the forces can also be written like: 22222

1 vzxvRR KKKKKKKKKy

⋅⋅=⋅⋅⋅== ρρ (7)


82

And by the equalization of the relations (6) şi (7) we get: 222vzvxzc KKKKKKKK

y⋅⋅=⋅⋅⋅⋅ ρρ (8)

hence resulting the scale of the unitary coefficient of the bearing force:

1KKK

cc

Kx

z

y

yc

m

n

y=== (9)

having thus:

mn yy cKc ⋅= 1 (10) The advance resistence being:

2

2vScR xx⋅

⋅⋅=ρ

(11)and taking into account the

relation (8), because the scale of forces is dependent on their nature we can write: 222vzvxzcRR KKKKKKKKKK

xx⋅⋅=⋅⋅⋅⋅== ρρ (12)

from which results the relation:

1KKK

cc

Kx

z

x

xc

m

n

x=== (13)

having:

mn xx cKc ⋅= 1 (14) As it is known for a given profile the coefficients cx, cy and cM are functions of the

incidence angle α, and the criteria of similitude Re, Fr, Sh şi Eu; also, the covement conditions of the wing in the unlimited fluid or in the vicinity of a solid or fluid surface have a great influence. Let’s examine now the scale of cM in conditions of similitude at two scales, with small angles of attack. In this case we can write the relation:

eRM y ⋅= (15) In which e represents the distance from the pressure centre of the profile to its board of attack. So, we can write:

xzvcRM KKKKKKK ⋅⋅⋅=⋅= ρ22 (16)

and

xvxzcM KKKKKKKM

⋅⋅⋅⋅⋅= 2ρ (17)

Equalizing (16) cu (17) we get: 2222vxzcxzv KKKKKKKKK

M⋅⋅⋅=⋅⋅⋅ ρρ (18)

From which results:

1)()( K

KK

ccK

x

z

mM

nMcM

=== (19)

Or:

mMnM cKc )()( 1= (20)


83

By the help of the relations (3), (10), (14), and (20) it is possible to transcribe the non

dimensional λ, cy, cx and cM from the model to nature, which as seen, have in nature the values from the model multiplied by the distortion ratio K1. Being non dimensional, these coefficients vary to the same extent when they shift from model to nature.

We should also say that in order to obtain the nature wing’s hydrodynamic coefficients we can also use the following formula: If we write the speed on the model like:

m

mmm c

vv ⋅=

Re (21)

And having known that between the speed of the nature wing and model wing is the following relation of similitude:

z

xmn K

Kvv ⋅= (22)

we get:

mm

n

m

n

m

mmn

cl

cc

cvv

⋅

⋅⋅

=

λ

Re

(23)

from which:

nm

mmnmn lc

ccvv

2

Re ⋅⋅⋅=

λ (24)

or:

nmm

nn

m

m

cvlv

cc

⋅⋅⋅

=Re (25)

nn

mnmmmm lv

cvcc

⋅

⋅⋅⋅=

λRe (26)

obtaining in this way the relation of determination of the model’s string’s length cm:

3

2Re

⎟⎟⎠

⎞⎜⎜⎝

⎛

⋅⋅⋅⋅

=nn

mmmmm lv

cvc

λ (27)

Using the definition relation of the relative elongation we can determine the span of the model wing:

mmm cl ⋅= λ (28) We calculate the scale of the string cK and the scale of the span lK :

m

nxc c

cKK == (29)


84

m

nzl l

lKK ==

(30)

We determine the distortion ratio 1K :

x

z

c

l

KK

KKK ==1 (31)

We state the scales of density, speed, and force.

m

nKρρ

ρ = (32)

m

n

z

xv v

vK

KK == (33)

m

n

m

n

x

x

y

yzxR R

RRR

KKKK ==⋅⋅= 2ρ (34)

Both the model and the real wing are rectangular in plan, and we can determine, with the known data the areas of the surfaces:

mmm clS ⋅= (35)

nnn clS ⋅= (36) According to the law of the model we calculate the speed of the nature (real) wing:

z

xmvmn K

KvKvv ⋅=⋅= (37)

With the known data we can further determine the bearing force of the model wing:

2

2mm

myyvScR

mm

⋅⋅⋅=

ρ (38)

Using the law of model or the relation (7), we will calculate the bearing force of the real wing:

mmn yzxyRy RKKKRKR ⋅⋅⋅=⋅= ρ2 (39)

From which the coefficient of the real wing bearing force:

2

2nn

n

yy vS

Rc n

n ⋅⋅

=ρ (40)

We calculate the advance resistance of the model wing:

2

2mm

mxxvScR

mm

⋅⋅⋅=ρ

(41)

and on the basis of the law of model we get the advance resistance of the real wing:

mmn xzxxRx RKKKRKR ⋅⋅⋅=⋅= ρ2 (42)

From which the coefficient of advance resistance of the real wing is deduced:


85

2

2nn

n

xx vS

Rc n

n ⋅⋅

=ρ (43)

In conclusion, taking into account what we have mentioned before, we can say that the values of the coefficients

nyc and nxc of the real (nature) wing, do not depend on

dimensions of the model wing; they depend only on the relative elongation of the wing, and for every single elongation of the wing only one polar is established.

It is true that if we extend the relations (10) and (14) we get:

m

ny

m

n

m

n

yx

zyy mmmn

c

ccll

cKKcc

λλ

⋅=⋅=⋅= (44)

Thus, for the mcn 3,0= ,smvn 25=

Thus, for the mcn 3,0= ,

smvn 25=

(45)

This is to confirm once more that within the relations between coefficients the dimensions of model wing do not interfere.

3. Tracing the Gottingen – 612 profile’s polar with relative elongation λ = 3, knowing the corresponding profile’s polar corresponding to the relative elongation λ = 5

The string’s length cn = 0,3 m and the ship’s speed vn = 25m/s is considered for the nature wing. We also stress that the initial polar was drawn in the aerodynamic tunnel, on the small span wing under observation will function in water.

Cinematic viscosity values of the two fluids are: s

maer

2

0000143,0=υ ;s

mapa

2610191,1 −⋅=υ

Thus, for the mcn 3,0= , smvn 25=

and

sm

apa

2610191,1 −⋅=ν there results:

6

6 1027,610191,1

3,025Re ⋅=⋅⋅

=⋅

= −apa

nn cvν

(46)

Ren = 6.300.000. The Gottingen-612 profile is characterized by: λm = 5 and Ren = 420.000. The following data are to be found in the specialty literature.

4. Conclusions The distortion ratios for the three elongations

1nλ = 3; 2nλ = 2 and

3nλ = 1, will be:

6,0531'

1 ===m

nKλλ

(47)

4,0522"

1 ===m

nKλλ

(48)


86

2,0513'''

1 ===m

nKλλ

(49)

Using the equations obtained through the theory of similitude (10) and (14) we can draw the nature wing with elongation

1nλ =3. Going on in the same manner, that is starting from the polars of big span wings and

using the theory of similitude at two scales, the polars of other profiles (of small and very small span), which were analysed, can be built; for example: Gottingen- 439, Gottingen- 480, NACA- 4409, Clark Y, RAF- 32, Gottingen-565, Gottingen-670, Gottingen-682, Gottingen- 507, and NACA- 6412, (for λ = 3, λ = 2 and λ =1) . References [1] Niestoj W., Profile pentru aeromodele, Varşovia, 1976 , p.29-36. [2] Vasilescu, AL. A., Analiză dimensională şi teoria similitudinii, Editura Academiei, Bucureşti, 1969, p.45-56. [3] Al.A. Vasilescu, Similitudinea sistemelor elastice, Editura Academiei, Bucureşti,1969. [4] E.Carafoli, V.N. Constantinescu, Dinamica fluidelor incompresibile, Editura Academiei, Bucureşti, 1981, p. 114-135 [5] Beazit Ali, “Stabilirea punţii de legătură între teoria aripii de mică anvergură şi teoria aripii de mare anvergură pe fondul teoriei similitudinii la două scări” Referat de doctorat, Universitatea“ Dunărea de Jos” Galaţi, 1995, p. 67-71. [6] Beazit Ali, Obţinerea polarelor aripilor de mică anvergură plecând de la polarele aripilor de mare anvergură, folosind teoria similitudinii la două scări, Buletinul „Tehmar”, Constanţa, 1996, p.3. [7] Beazit Ali, Traian Florea, Study on the upward small span profile based on the two scale similarity theory, The XII the National Conference on Thermotechnics with International Participation, “Mircea cel Bătrân” Naval Academy, Constanta, 2002, p 3-4.


87

LINEAR QUADRATIC OPTIMAL CONTROL UNDER STOCHASTIC UNIFORM OBSERVABILITY AND DETECTABILITY CONDITIONS

V.M. Ungureanu, Constantin Brâncuşi University, Tg-Jiu, ROMANIA

Abstract. In this paper we apply the results in [1] - [3] to solve some linear quadratic control problems under either stochastic uniform observability conditions or detectability conditions.

Keywords: linear quadratic control, stochastic observability, detectability, stochastic differential equations

1. Statement of the problem Let V,U,H be real separable Hilbert spaces and let ),0[J ∞=⊂ +R be an interval. If E is a Banach space endowed with the norm . , then )E,J(C is the space of all mappings

EJ:G → that are continuous. By ))H(L,J(Cs we denote the space of all strongly continuous mappings )H(LJ:G → and by ))H(L,J(Cb the subspace of ))H(L,J(Cs which consist of all mappings such that ( ) .tGsup Jt ∞<∈

Assume the following hypothesis:

G1: ( )s,tU is a strong evolution operator generated by the family ),0[t),t(A ∞∈ ; there exists a sequence ))H(L),,0([C}A{ snn ∞⊂∈N such that for every N∈n the family

( ) 0tn }tA{ ≥ generates the strong evolution operator ( )s,tUn and

( ) ( ) ,Hx,xs,tUxs,tUnn ∈→

∞→

uniformly on bounded subsets of }ts0),s,t{( ≤≤=Δ .

If (G1) holds we often say that G1 )A( holds. We will say that G1 )A( ∗ holds if the families ( ) ( ) ),0[t,n,tA,tA n ∞∈∈∗∗ N satisfy the hypothesis (G1). Let us consider the following stochastic system with control }H:G,B:A{ ii

( ),Hx)s(y

,st),t(dw)t(u)t(H)t(y)t(G

dt)t(u)t(Bdt)t(y)t(A)t(dy

iii

m

1i∈=

≥++

++=

∑=

where )),H,U(L,(CH,B bi +∈ R C)),U,H(L,(CH,B bi +∗∗ ∈ R ))V,H(L,(Cb +∈ R ∈∗C,

)),V,H(L,(Cb +R ,(CG,G bii +∗ ∈ R ,...,1i)),H(L = ,m,m ∗∈ N ))U(L,(CK b +∈ R and

there exists 0>δ such that for all

I)t(K,t δ≥∈ +R .

The control u is assumed to belong to a set of admissible controls ,Uad which consists in all )U,(Lu 2 Ω×∈ +R such that u is −tF adapted and


88

∞<≥

2

st)t(yEsup ,

where ( )ty is the mild solution of }H:G,B:A{ ii .

The system without control }H:G,B:A{ ii (i.e B = 0 and }m,...,1{i,0Hi ∈= ) will be denoted by }G;A{ i . We deal with the following

Quadratic control problem.

Let us consider the set adqad Uu{ ∈=U such that 0)t(yE 2 → as ∞→t , where ( )ty is

the solution of }H:G,B:A{ ii } . We look for an optimal control qadu U∈ which

minimizes the quadratic cost

.dt)t(u),t(u)t(K)t(y)t(CE)u(I 2

sx,s ><+= ∫

∞

(1)

2. Linear quadratic control problem Let us recall the following result from [4]. THEOREM 1. Assume that }H:G,B:A{ ii is stabilizable and either }C;G,A{ i is stochastically uniformly observable or }C;G,A{ i is detectable and let us consider the control problem }H:G,B:A{ ii , (1). Then the optimal control is given by the feedback law

( ) ( )( )[ ] ( )( ) ( )tytPtPtu~ 1BK −−= , (2)

where P is the unique bounded and uniformly positive solution of

[ ] [ ] 0)P()P()P(CCPGGPAPAP 1ii

m

1i

=−++++′ −∗∗∗

=

∗ ∑ BKB , (3)

where )s(G)s(P)s(H)s(P)s(B)s)(P( ii

m

1i

∗

=

∗ ∑+=B and ).s(H)s(P)s(H)s(K)s)(P( ii

m

1i

∗

=∑+=K and

)t(y is the solution of }H:G,B:A{ ii corresponding to the control u~ .

The optimal cost is

( ) ( ) x,xsPu~I x,s = . (4)

Let w be a real Wiener process and let )1,0(H)1,0(H)(D: 2102

2∩=Λ−=Λ

ξ∂∂

)1,0(L2→ be the infinitesimal generator of an analytic semigroup on the Hilbert space )1,0(L2 [3].

It is well known that ( ) ),1,0(LDH 22/1 ⊕Λ= equipped with the inner product ,u,xu,xu,x

)1,0(L22)1,0(L12/1

12/1

H 22 +ΛΛ= ( ) ( ) ,Hu,uu,x,xx T21

T21 ∈== is a real

separable Hilbert space. Also the linear operator


89

( ) ( ) ( ) .xx

0I0

xA,HDD:A2

12/1⎟⎟⎠

⎞⎜⎜⎝

⎛⎟⎟⎠

⎞⎜⎜⎝

⎛Λ−

=→Λ⊕Λ (5)

is the infinitesimal generator (see [2],[1]) of a strongly continuous semigroup ( )tS in .H

Furthermore, ( )tS is a contraction semigroup and

( )( )[ ]

[ ] ,tncos,xtnsin,xn2

tnsin,xtncos,x2

xtS

n)1,0(Ln2)1,0(Ln11n

n)1,0(Ln2n1

)1,0(Ln11n

22

22

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛

φπφ+πφπ−∑

φπφ+πφ∑

=

∞

=

π

∞

=

where πξ=φ nsinn . We consider the stochastic differential system

( ) ( ) ( ) ( ) ( )( ) ( )

( ) ( ) ( ) ( )( ) ( )⎪

⎪⎩

⎪⎪⎨

⎧

Ω∈ω≥∈ξ∈ωξ=∈ξξ=ξξ=ξ

≥==ξ+ξ+=ξ

ξ∂ξ∂

,0t],1,0[,H,t,u,u,uu].1,0[,y0,y,y0,y

0t,0t,1y,0t,0y,tdwt,ydtt,udtt,dy

T21

1t0

t2t,y

t 2

2

(6)

with the observation relation

( ) ( )( ) .t,y,0tz Tt ξ= (7)

We study the problems of finding the optimal control which minimizes the cost functional

( ) ( ) ( ) .y,yx,u,uu,dt)t(ut,yE)u(I T100

T21

2

H

2

)1,0(Lt0

x,0 20==+ξ= ∫

∞

(8)

Before starting to solve the above problems, we see that the above stochastic differential system can be equivalently rewritten as

( ) ( ) ( ) ( ) ( ) ( ) ( ) ( )( ) ( ) ( ),txtCtz

x0x,tdwtxtGdttutBdttAxtdx 0=

=++=

where ( ) ( ) ( )( ),y,yx,)x,x(x 100T

21 ξξ=ξ= ( ) ( ) ( ) ,I000

tCtGtB ⎟⎟⎠

⎞⎜⎜⎝

⎛=== t ≥ 0.

Comparing it with }H:G,B:A{ ii and (1) we see that ,1m = ( ) ( ) 0tHtH 1 == and ( ) HVU,0t,ItK ==≥= .

Obviously, the system is in the time invariant case. The hypothesis (G5) holds and A is the infinitesimal generator of a strongly continuous semigroup of contractions ( )tS . We shall prove that (6)-(t) is stochastically uniformly observable and }H:G,B:A{ is stabilizable.

Applying the results in [5], [6] it follows that }C;G,A{ i is stochastically uniformly observable. The algebraic Riccati equation associated with the discussed control problems is


90

0PPBBCCPGGPAPA =−+++ ∗∗∗∗ (9)

We recall that AA −=∗ [1]. It is immediate that ( )I5P 21

21 += is a nonnegative

solution of (9) and we conclude that it is also a nonnegative and bounded solution of (3). Theorem 18 from [4] shows that P is stabilizing for the system }0:G,B:A{ i .

By Theorem 1, it follows that the optimal cost of the quadratic control problem is

( ) 2

H021

21

x,s x5)u(I0

+=

and the optimal control is

( ) ( ) ( )( ) ,t,y~2/12/5,0t,u~T

t ξ+−=ξ

where ( )t,y~t ξ is the solution of the equation

( ) ( ) ( ) ( ) ( ) ( )( ) ( ) ( ) ( ).y0,y~,y0,y~

,tdwt,y~dtt,y~2/12/5dtt,y~t,y~d

1t0

tt2

2

t

ξ=ξξ=ξ

ξ+ξ+−ξ∂ξ∂

=ξ

For another application of Theorem 1, the reader can consult [4].

In what follows we solve a linear quadratic control problem for a stochastic system which is detectable but it is not stochastically uniformly observable.

A parabolic equation. Consider the stochastic system

( ) ( ) ( ) ( )( ) [ ]

( ) ( ).t,yt,z,1,0,sin20,y,0t,0)t,1(y)t,0(y

)t(dwt,ydtt,udtt,yt,dy 2

2

ξ=ξ∈ξπξ=ξ≥==

ξ+ξ+ξ∂ξ∂

=ξ

where )t(w is a real Wiener process. Our purpose is to find an admissible control, which minimizes the cost functional

( ) .dt)t(ut,yE)u(I 2

)1,0(L

2

)1,0(L0

y,0 220+ξ= ∫

∞

For this problem ),1,0(LVUH 2=== ,GG,0H,1mnot

11 === HIKCGB ==== and Λ=A , .H)1,0(H)1,0(H)(D: 21

02

2 →∩=Λ=Λξ∂

∂

Then the above system can be equivalently rewritten as

( ) ( ) ( ) ( ) ( ) ( ) ( )( ) ( ) ( ).tytCtz

;y0y,tdwtytGdttBudttAytdy 0=

=++=

Since A generates a compact semigroup ( )tS on H , }C;G,A{ is not stochastically uniformly observable(see [6]). Now we associate the following algebraic Riccati equation


91

,0PBPBKCCPGGPAAP 1 =−+++ ∗−∗∗

or equivalently, .0PIPPAAP 2 =−+++ A nonnegative solution of this equation is

( ) IAAIP 221

21 ++++= i.e.

( )

.nsin2e

,ee,y121nn

21yP

n

nn

22222

1n

πξ=

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛+⎟

⎠⎞

⎜⎝⎛ +π−+π−= ∑

∞

=

Now we show that the hypotheses of Theorem 1 are verified. Since P is also a nonnegative and bounded solution of (3) we only have to prove that }C;G,A{ is detectable.

We recall that }G,CA{ − is uniformly exponentially stable if and only if the Lyapunov equation 0IPPAAP =+++ has a unique positive solution. It is easy to see that this is equivalent to the assertion: }IA{ 2

1+ is uniformly exponentially stable, which is obviously true. Therefore, }G,CA{ − is uniformly exponentially stable and, consequently, }C;G,A{ is detectable. In view of Theorem 1, the optimal control is

( ) ( ),t,Pyt,u ξ−=ξ

where ( ),y tξ is the solution of the closed loop system

( ) ( ) ( ) ( )

( ) .ey0y

,tdwtydtty121A

21tdy

10

2

==

+⎟⎟

⎠

⎞

⎜⎜

⎝

⎛+⎟

⎠⎞

⎜⎝⎛ +−−=

The minimum value of the cost functional is

.121

21y,Py)u(I

222

H000 +⎟⎠⎞

⎜⎝⎛ +π−+π−==

References [1] R. Curtain, J.Pritchard, Infinite dimensional linear systems theory, Springer Verlag, Berlin, 1978. [2] G.Da.Prato, J. Zabczyk, Stochastic Equations in Infinite Dimensions, University Press Cambridge, 1992. [3] A. Pazy , Semigroups of linear operators and applications to partial differential equations, Applied Mathematical Sciences 44,Springer- Verlag, Berlin, New -York, 1983. [4] V. M. Ungureanu, Optimal control of linear stochastic evolution equations in Hilbert spaces and uniform observability, Czechoslovak Mathematical Journal, Vol. 59, Issue: 2, 2009, 317-342. [5] V. M. Ungureanu, Stochastic uniform observability of linear differential equations with multiplicative noise,J. Math. Anal. Appl. 343 (2008), no. 1, 446--463. [6] V. M. Ungureanu, Stochastic uniform observability of general linear differential equations, Dynamical Systems, Vol. 23, No. 3, September 2008, 333-350.


92

SCHEME ANALYSIS TREE DIMENSIONS AND TOLERANCES PROCESSING

S.L.dr.ing. Constanta RADULESCU1,

Prof.univ. dr. ing. Liviu Marius CÎRŢÎNĂ2, Prof.univ.dr.ing. Constantin MILITARU3

1,2Faculty of Engineering „Constantin Brâncusi” University of Tg- Jiu 3University Politehnica of Bucharest Romania

Abstract: This paper presents one of the steps that help us to determine the optimal tolerances depending on the technological capability of processing equipment. To determine the tolerances in this way is necessary to take the study and to represent schematically the operations are used in technological process of making a piece. Also in this phase will make the tree diagram of the dimensions and machining tolerances, dimensions and tolerances shown that the design execution. Determination processes, and operations of the dimensions and tolerances tree scheme will make for a machined piece is both indoor and outdoor. Keyworks: tolerance, cost, process, capability.

1. Introduction This paper (1) was presented a model for optimizing technological tolerances depending

on the capability of processing equipment. To determine the optimal tolerances is thus necessary to take the study to represent schematically the operations are used in technological process of making a piece, which is why before proceeding to calculate standard deviations and the calculation function optimization, which in this case is to minimize the cost of running the play, it should be presented in tabular form, the technological process of track operations studied. Operations are presented in Table technological process have a minimum of information, but very important, concerning: the name of the operation, the reference surface, the surface is processed and the size of the process. Also in this stage, we help determine optimal tolerances, it is necessary to make the tree diagram of the dimensions and machining tolerances, dimensions and tolerances shown that the design execution.

2. Case Stud To determine what is said in his practice (1) and above, use an example for the song in

figure 1, which is considered to be achieved for large series and mass production.

Fig.1. Chain elements tolerances of the workpiece dimensions.


93

The piece is worked both outside and inside and will include operations such as turning, milling, grinding and heat treatment (hardening + recovery). In figure 1 presents the workpiece, and operations are presented in table 1 indicates the reference surfaces and surfaces prepared for each operation. It will be noted in large areas of the workpiece. In addition, BC and GC are used to indicate heat-treated surfaces of B and G, G1 and B1 and B are generated surfaces respectively, G. To see chains processing tolerances can use the information given in table 1, and with it will draw a tree representation as shown in figure 2. The arrows represent the relevant processing operations. Areas that are connected ends of the arrows are flat reference surfaces. Areas that are connected arrowheads are manufactured surfaces (surfaces in bold). The songs were performed on numerically controlled machine tools. Table 1. Workpiece processing sequence.

Number Operations

Operation Reference area Processed area Working size (mm)

1 Turning

A G 19,70

2 Turning

G B 19,40

3 Turning

G C 17,80

4 Turning

B F 17,30


94

5 Milling

B D 2,60

6 Grinding

B G1 19,30

7 Grinding

G1 B1 19,20

8 T.T hardening +recovery) B Bc 0,25 9 T.T (hardening + recovery) G Gc 0,25

10 Turning

F E 1,00

NOTE: The table is not complied with the order process technology operations. Using figure 2, where are the tree sizes and processing tolerances and dimensions and tolerances shown on the drawing performance can be obtained conveniently tolerances chains of processing operations.

Fig.2. Tree representation of the dimensions and tolerances in processing(L,t)

and dimensions and tolerances shown on the drawings (Tpd).


95

The tolerance chain dimensions have been obtained on the basis that tolerances results must be less than or equal to those of design specifications for implementation. Based on figure 2, were established bilateral ties maximum tolerance of the machining processes and machine tool gears. The standard deviations of processes depend on the processed surface, the size of machines and cutting tools to machine tools. Based on the above information can move on determination standard deviations in writing optimization function at presentation restrictions and finally to determine optimal tolerances.

3. Conclusions. Operations are presented in table technological process have a minimum of information, but very important, concerning: the name of the operation, the reference surface, the surface is processed and the size of the process. Using figure 2, where are the tree sizes and processing tolerances and dimensions and tolerances shown on the drawing performance can be obtained conveniently tolerances chains of processing operations. From the foregoing it is apparent that in order to apply an optimal method for allocating tolerances of components, design engineers and engineering technology should consider the process of implementation throughout the piece. References. [1].C., Rădulescu, L.M.Cîrţînă - Theoretical studies regarding optimization of technological tolerances depending on the capacity of the processing- Analele Universităţii ,,Constantin Brâncuşi,, din Tg-Jiu, nr.4/2010, 217-224, ISSN1842-4856. [2].Dragu, D., ş.a., - Tolerances and technical measures, Pedagogic and Didactic Publishing House, Bucharest, 1980. [3].Rădulescu, C. Contributions to the study of technical and economic aspects in solving chain dimensions in the composition of machinery and industrial equipment. Doctoral Thesis, Bucharest Polytechnic 2009. [4]. Wei, C.C and Y.C.Lee - Determining the process tolerances, based on the manufacturing process capability, Intenational Jurnal of Advanced Manufacturing Tehnology, 1995


96

EXPERIMENTAL RESEARCH FOCUSED ON YIELDS OF LINEARE HYDRAULIC MOTORS USED TO DRIVE THE BOTTOM PUMPS

Dr.ing. Petre SĂVULESCU, Universitatea Petrol-Gaze din Ploiesti,

Abstracts:This paper presents the authors’s concerns for determining the functional parameters and its yields for a linear hydraulic engine with double effect. Functional parameters are determined both at unloaded running and loaded running of the linear hydraulic motor. The yields was determined on loaded running.

Key words: yields, linear hydraulic motor, bottom pump

1. Introduction Based both on the presenting data on [1; 2] and on the experience [3; 4], in some conditions of controlled drilling or as a result of some sensitive deviations during wells drilling or on the situation of exploration of oil at high deep, clasical pumping units use it became irational because of the huge energetic consum or of the imposibility to use them at deeps more than 3000 m.

On these situations are used oil extraction units with hydraulic action. These units include the extraction pump and a hyrdaulic motor on which the reversing movement is autocontrolled. By all these reasons the author of this work has designed and realised a lineare hydraulic engine to do the experiments for registing both the functional parameters and the its yeld.

The laboratory test allow to establish some mesures which can not be registed in the field because of the construction of the deep hydraulic unit and because of the high deep of laying. The author also has realised two desplacement transducers boath to regists the desplacement of dispenser distributor and of piston rod displacement of the linear hydraulic motor. In such cases are used for oil extraction units with hydraulic action.

These are consisting of the extraction pump and a hydraulic motor which is an autocontrolled reversal movement one. For these reasons, the author of this work, conceived, designed and built a linear hydraulic motor for conducting experimental research to determine both functional parameters and its return. Experimental research in laboratory conditions, allowing the establishment of the sizes that can not be determined on the field conditions, because of the depth hydraulic unit buunloaded running and of its location at hight deep. Also the author has made and second displacement transducers, both for recording the movement of the distributor drawer and of displacement of motor piston rod of linear hydraulic motor.

2. Experimental test results The general bench view on which are doing the tests is presented by figure 1.

The simultaneous registered of seven parameters (two displacements, three tenssions and two flows as function of time) at diferent pump functional regimes alows to put on evidence diferent quqntitative and calitative aspects. The reading of oscilograph was done by direct mesuring and then the results were analized by the help of EXCEL program.


97

Fig. 1. The general stand view To record the displacement of the draw of lineare hydraulic engine it was realised a

inductive transducer, transformer type, which was introduced in the balanced hydraulic tube which is presented in figure 2. Figure 3 presents the functional diagrame of lineare hydraulic engine unloaded running, at an angular speed of pump by ωp=157rad/s. The run on duty it was realised by using a vane with a nail on the exit pipe. The functional diagrame is presented in figure 4.

Fig. 2. Displacement transducer and balanced hydraulic tube system .


98

Fig. 3. Functional diagrame of lineare hydraulic engine unloaded running ωp=157rad/s

Fig. 4. Operation diagram of linerae hydraulic motor loaded running , ωp=157rad/s

Stroke capacity of lineare hydraulic motor

xs[dm] xm[dm]

pa[MN/m2] pe[MN/m2] pcj[MN/m2] Qa104[m3/s] Qe104[m3/s]


99

33 m103180 −⋅== ,SAV mm , (1)

where:

Am is aria of piston transversal section; S – stroke size . Theoretical frequency of MHL function is:

,,m

ptm V

Qf = (2)

where Qp is the debit realised by pump.

The real operating frequency of the linear hydraulic motor can be determined experimentally by the help of operating charts, resulted on an operating cycle time (Tc).

Given this observation it can be written:

c

em Tf 120

, = , (3)

where Tc is:

Tc=ta+td+2tcd , (4)

where:

ta is time for making the upward stroke; td – time for making the downward stroke; tcd–swiching time of distributor drewer . Theoretical values and experimental values of lineare hydraulic engine are presented in table 1.

Table 1. Frequency values of lineare hydraulic engine unloaded running

Crt. no ωp [rad/s] Qp[104m3/s] fm,t [strokes/min] fm,e [strokes/min]

1. 115 2,54 47,9 42,8

2. 125 2,77 52,2 46,5

3. 136 3,00 56,6 52,2

4. 146 3,22 60,7 53,5

5. 157 3,46 65,3 58,8

6. 167 3,69 69,6 61,8


100

The dependence between the run frequency of the linear hydraulic motor and pump flow is shown in Figure 5. Is apparent that there is a constant difference (10-11%) between actual working frequency and theoretical one for linear hydraulic motor. This is explained by the fact that a part of the flow is used for switching flow distributor drawer at the head race and because of dead space existence.

Fig 5.The dependence between the run frequency of the linear hydraulic motor and pump flow: fm,t –theoretical frecvency on unloaded running run; fm,e– experimental frecvency unloaded running .

When operating under loading run (the pressure on exhaust of hydraulic linear engine

((pe 2MN/m5,1≅ ), the frequency of operation loaded running (fm,s) of hydraulic linear motor shows a decrease, compared with the unload run. This is shown in Figure 6.

Fig.6. Operating frequency of the linear hydraulic motor loaded running (pe=1,5 MN/m2).

By examining the diagrams of linear hydraulic motor operating at different angular

speeds of the pump it can made some observations. It will further examine the operating diagrams of the linear hydraulic motor at ωp=157 rad/s.


101

2.1 Operation of linear hydraulic motor unloaded running It can be observed that the switching time of drewer distributor is tcd=0,10 s. The

upward stroke is done in 0,72 s and the downward one in 0,70 s, the two races runing time being about the same. Standing time of piston between upward stroke and downward one is 0.08 s. The feed power (pa) between the two races has an increasing from 1,167 MN/m2 la 2,33 MN/m2, which can be explained by the the suddenly change of direction of movement of the piston of linear hydraulic engine.

2.2 Operation of linear hydraulic motor loaded running From the begining it is observed that the realization time of races is higher. The

distributor's switching time is tcd=0,08s. Time to achieve the upward race is ta=1,02 s, while the downward race td=0,98s. It results a difference for the two races of 0,04 s. Parking between the ascending and descending flight is 0,10s.At the changing of direction of movement, the feed pressure increased from 2,39 to 2,89 MN/m2. It follows an increase of 20% above the working pressure, an increase that can be taken into account in setting the safety valve.The acceleration of the piston for upward race is ama=0,38 m/s2 and for the downward race is amd=0,41m/s2 From the analyse of charts at upload run and load run of linear hydraulic motor is can be noted that the duration of an operating cycle decreases in the same time with the increasing of angular velocity from the pump. The operating time cycle loaded running run is higher than the unload operation at the same angular speed of the pump. This is shown in Figure 7.

Fig. 7. Operating cycle according to the angular speed of the pump (ωp):

Tc,g – time of operation cycle unloaded running ; Tc,s – time of operation cycle loaded running .

3.The yield of linear hydraulic motor The yield of linerae hydraulic motor (ηm) can be established by the following relation:

,maa

mm

h

m

QpvF

PP

⋅⋅

==η (5)

where:


102

mP is mechanical power; hP – hydraulic power; mF – the power realised by the lineare hydraulic motor;

mv – the medium displacement speed of the piston ; ap – feed presure; aQ – feed flow.

The power realised by lineare hydraulic engine as a function of feed pressure is presented in figure 8.

.

Fig. 8. The power realised by lineare hydraulic engine as a function of feed pressure

To evaluate the mechanic power it is calculed a medium speed of piston displacement which is presented in figure 9.

Fig. 9. Medium speed of piston displacement as a function of feed debit (Qa)

The engine yield is calculated by relation (3.1) an it is presented in figure 10.

Fig. 10. Yield of linerare hydraulic engine variation as a function of feed pressure (pa)


103

It is noted that the overall efficiency of linear hydraulic motor is placed at around 0.8,

a value that can be considered good. 4. Conclusions

The experimental research focused on linear engine which are using to drive hydraulic pumps for deep extraction of oil, led to a number of conclusions, as follows:

• linear hydraulic motor hydraulic balanced led to a leveling of supply pressure on the two strokes (up and down);

• watching the drawer movement and the movement of piston distributor engine requires to carry out two displacement transducers (inductive type);

• inductive displacement transducers made by the author have proved to be reliable in operation of linear hydraulic motor (under pressure, the working environment being an oil H30);

• simultaneous register boath of distributor drawer movement, of the piston displacement and the flow and pressure values showed a proper operation of the linear hydraulic motor;

• up and down trips operation time is insignificant different (difference 0.02 to 0.04 s) • switching time of supply purpose (traveling time of the drawer distributor), is in the

range 0.08 to 0.10 s; • operating frequency of linear hydraulic engine, experimentally determined, is with 10-

11% less than the operating frequency theoretically determined; • an operating cycle period (Tc), is decreasing with the ungular velocity growth on

pump and is smaller on unload operation of linear hydraulic engine; • yield value of linear hydraulic engine t is approximately constant and lies around 0.8.

References [1]. S ă v u l e s c u , P . P a n a , I . Experimental research focused on functional parameters of linear hydraulic engines used to drive deep pumps, in Proceedings of the VII International symposium" University's Day ", University" Constantin Brancusi 'of Targu Jiu, May 24 to 26, Section 3 , paper 36, "Technology machinery and equipment", volume published on CDROM, ISBN 973-8436-06-0, Targu Jiu, 24 to 26 May 2002. [2]. S ă v u l e s c u , P . P a n a , I . Contributions concerning linear motors use to drive hydraulic deep pumps,. Jubilee Session-drilling equipment in oil extraction, May 28 to 30, Ploiesti, 1998. [3]. S ă v u l e s c u , P . P a n a , I . Experimental research on linear engines use to drive hydraulic deep pumps for oil extraction, Scientific Research contract no. 196/1996 of C.N.C.S.U., Ploiesti, 1996. [4]. S ă v u l e s c u , P . Research focused on oilfield equipment to operate wells in unconventional ways, Doctorall Thesis, Petroleum-Gas University Ploiesti, 2004.


104

REGRESSIONS MODELING OF SURFACE ROUGHNESS IN FINISH TURNING OF HARDENED 205Cr115 STEEL USING FACTORIAL

DESIGN METHODOLOGY

Alexandru STANIMIR, Marius ZAMFIRACHE, Nicolae Cătălin EFTENIE University of Craiova

Abstract In this study, in order to find out a mathematical relation of polynomial second degree type which describe, in finish turning of hardened 205 Cr115 steel, the roughness parameter Ra dependence on cutting edge wear, depth of cut, feed rate and cutting speed, a factorial design methodology was used. The experimental tests were done according to a composed, central, four-factor five-level factorial program.

Key words: roughness, hardened steel, finish turning, regression, factorial design

1. Introduction A large range of components encountered in the aerospace defense and auto vehicles

production are made from ferrous materials thermal treated, where their hardness usual vary between 45-70 HRC and which can be dry processed using polycrystalline cubic boron nitride cutting tools. For many of these parts is imposed a high process precision and a high level smoothness of functional surfaces.

Some researchers have investigated wear characteristics of PCBN cutting tools [1, 2, 5] and the influences of work material properties, tool geometry and cutting conditions on surface integrity of the finish-machined parts [3, 6, 9]. They reported challenges in the identifying various processes, equipment and tooling related factors affecting surface quality, tool life and productivity. Hard turning with PCBN cutting tools demand prudent design of experiment by reason of the lower toughness than other common tools materials, thus chipping is more likely. Also PCBN cutting tools designed for hard turning feature negative rake angle and a chamfer and/or a hone of the edge. The effect of edge geometry on surface quality is significant [3, 6].

Due to the changes in properties of hardened workpiece material, chip formation and shearing process differ in hard turning [11], and the workpiece hardness has a profound effect on the performance of the PCBN tools [9], and also on the integrity of the finish machined surfaces [8].

The cutting conditions i.e., depth of cut, feed rate and cutting speed influence the performance of PCBN cutting tools and the roughness of the finish machined surfaces. Increasing cutting speed and depth of cut result in increased temperatures at the cutting zone and at elevated temperatures chemical wear of PCBN becomes a leading wear mechanism and often accelerates weakening of cutting edge, resulting in premature tool failure, namely edge breakage of the cutting tool [9]. In the machining of hard materials, the profile of cutting edge is transferred to the machined surface with reasonable accuracy. Then, as long as the flank wear is uniform, it does not increase the surface roughness [8]. Another parameter that degrades the surface roughness and tool life is the vibrations. In order to reduce tool vibration it is necessary to provide sufficiently rigid tooling and work piece fixtures. Extremely rigid, high power and high precision machine tools are required for hard turning because PCBN tools are brittle and prone to chipping. It is also suggested that having higher rigidity in machine tool-clamping-tooling system achieves better surface quality on the part [9].


105

2. Experimental equipment Workpiece material The work piece material used in this study was the 205Cr115 steel, with the chemical

composition presented in the table 1. Table 1.

Chemical composition of 205Cr115 steel C% Mn% Si% Cr% Ni% 1,8...2,2 0,15...0,450,15...0,3511,0...13,0 0,35

The cylindrical rings 205Cr115 specimen (fig. 1) that are used in this experiments had a outside diameter of 62 mm, a inside diameter of 53 mm and a length of 16 mm. The specimens were heat treated at 61-63 HRC.

Fig. 1. Workpiece

Technological system and apparatus The experimental machining tests was performed on a SP630 CNC lathe and the

roughness parameter Ra was measured with a Ranck Tailor Form Talysurf Series 2 profilometer.

Also solid top PCBN inserts BZ 435 TPGN 22 04 16 were used with a Sandvik

Coromant CTENR 2525M22 (ISO 5608) tool holder, that give the following tool geometry: K=60º, α=17º, γ=-6º and λ=-6º. In order to machining the outside cylindrical surface of the specimens a special clamping system (fig 2) was used.


106

Fig. 2. Clamping system for workpiece

3. Experimental design In order to establish the second degree polynomial relationship coefficients for the Ra

roughness parameter calculus (1) a four-factor five-level factorial design was used [4], [7], [10].

214

213

212

2111098

76543210

VBbVbfbabVBVbVBfbVfb

VBabVabfabVBbVbfbabbR

cpcc

pcppcpa

⋅+⋅+⋅+⋅+⋅⋅+⋅⋅+⋅⋅+

⋅⋅+⋅⋅+⋅⋅+⋅+⋅+⋅+⋅+= (1)

In the table 2 is presented the planning of the experiments, where y represents the

response to the studied phenomena, in this case Ra, and x1, x2, x3 and x4 are the independent variables. The five levels of the independent variables codified (-2), (-1), (0), (1) and (2) were: - x1 – the cutting speed, Vc=80m/min; 120m/min; 160m/min; 200m/min; 240m/min; - x2 – the feed, f=0,025mm/rev; 0,050mm/rev; 0,075mm/rev; 0,100mm/rev; 0,125mm/rev; - x3 – the depth of cut, ap=0,05mm; 0,10mm; 0,15mm; 0,20mm; 0,25mm; - x4 – the flank wear, VB= 0mm; 0,05mm; 0,10mm; 0,15mm; 0,20mm; 0,25mm.

Table 2. The planning of experiments

Current no.

Independent variables levels Y Ra[μm] X1

Vc[m/min] X2

f[mm/rev] X3

ap[mm] X4

VB[mm] 1 200 (1) 0,05 (-1) 0,1 (-1) 0,05 (-1) 0,2705 2 120 (-1) 0,05 (-1) 0,1 (-1) 0,05 (-1) 0,2695 3 200 (1) 0,1 (1) 0,1 (-1) 0,05 (-1) 0,3425 4 120 (-1) 0,1 (1) 0,1 (-1) 0,05 (-1) 0,3690 5 200 (1) 0,05 (-1) 0,2 (1) 0,05 (-1) 0,2762 6 120 (-1) 0,05 (-1) 0,2 (1) 0,05 (-1) 0,2648 7 200 (1) 0,1 (1) 0,2 (1) 0,05 (-1) 0,3590


107

8 120 (-1) 0,1 (1) 0,2 (1) 0,05 (-1) 0,3339 9 200 (1) 0,05 (-1) 0,1 (-1) 0,15 (1) 0,2711 10 120 (-1) 0,05 (-1) 0,1 (-1) 0,15 (1) 0,2872 11 200 (1) 0,1 (1) 0,1 (-1) 0,15 (1) 0,4815 12 120 (-1) 0,1 (1) 0,1 (-1) 0,15 (1) 0,4331 13 200 (1) 0,05 (-1) 0,2 (1) 0,15 (1) 0,2759 14 120 (-1) 0,05 (-1) 0,2 (1) 0,15 (1) 0,2211 15 200 (1) 0,1 (1) 0,2 (1) 0,15 (1) 0,4621 16 120 (-1) 0,1 (1) 0,2 (1) 0,15 (1) 0,4463 17 240 (2) 0,075 (0) 0,15 (0) 0,1 (0) 0,3546 18 80 (-2) 0,075 (0) 0,15 (0) 0,1 (0) 0,2551 19 160 (0) 0,125 (2) 0,15 (0) 0,1 (0) 0,3557 20 160 (0) 0,025 (-2) 0,15 (0) 0,1 (0) 0,2413 21 160 (0) 0,075 (0) 0,25 (2) 0,1 (0) 0,2519 22 160 (0) 0,075 (0) 0,05 (-2) 0,1 (0) 0,2704 23 160 (0) 0,075 (0) 0,15 (0) 0,2 (2) 0,5111 24 160 (0) 0,075 (0) 0,15 (0) 0 (-2) 0,3189 25 160 (0) 0,075 (0) 0,15 (0) 0,1 (0) 0,2329

4. Results and discussion

In order to find the values of the coefficients from the relationship (1), the experimental test results were processed with the Mathcad 2.5 program. The relationship established for the calculus of Ra roughness parameter is:

[ ]mVBV

faVBVVBfVfVBa

VafaVBVfaR

c

pccp

cppcpa

μ225

22

234.1910

345.30849.3003.0215.220007.0248.1

003.0775.1138.50046.0955.4766.1042.1

⋅++

⋅+⋅+⋅⋅+⋅⋅+⋅⋅+⋅⋅−

⋅⋅+⋅⋅+⋅−⋅+⋅−⋅−=

−

(2)

Fig. 3. Roughness surface variation with depth of cut and feed rate


108

Fig. 4. Roughness surface variation with cutting speed and flank wear

In the figures 3 and 4 are presented the graphic diagrams of the relationship (2), made

with the Mathematica 5.2 software, for the doublets ap-f, with command (3), and Vc-VB, with command (4). The parameters that don’t appear on axis were considered at the central value so: Vc=160[m/min], VB=0,1[mm] , ap=0,15[mm] and f=0,075[mm/rev].

}] , ][ , ][{ →+++"""/"""AxesLabel{y,0,0.2},,{x,80,240}

y^2,*19.234x^2*0.00001y*x*0.003y*3.6591-x*0.0041-595Plot3D[1.4Rarevmmfmma

(3)

}] , ][ , ][ { →+++"""""min/"AxesLabel{y,0,0.2},,{x,80,240}

y^2,*19.234x^2*0.00001y*x*0.003y*3.6591-x*0.0041-595Plot3D[1.4RammVBmVc

(4)

These diagrams give an eloquent image about the influence of each factor studied

upon the machined roughness surfaces. So, from figure 3 we can see that the roughness increase with the feed and presents a minimum with depth of cut variation. And also figure 4 shows that roughness of machined surface present points of minimum with the cutting speed and the wear cutting tool flank variation.

The maximum roughness, Ra=0,511μm, corresponds to the following experimental conditions: ap=0,15mm, f=0,075mm/rev, Vc=160m/min and VB=0,2mm and the minimum roughness value in this experimental study is Ra=0,2211μm, obtained in the following experimental conditions: ap=0,2mm, f=0,05mm/rev, Vc=120m/min and VB=0,15mm. However the most small values of roughness ≈2,5μm were obtained when the flank wear was VB≈0,1mm.

The calculated minimum of surface roughness {0.0808553, {u→0.0963638, x→0.183896, y→0.0496018, z→186.225}} resulted from the evaluation of the relation (2) in Mathematica 5.2 with the relation:


109

}],,{,∗19.234+∗0.00001+∗30.345+∗3.849+

∗∗0.003+∗∗22.215+∗∗0.0007+∗∗1.248−∗∗0.003+∗∗1.775−∗5.138−∗0.0046−∗4.955−∗1.766−[1.042

uzyxuzyxuzuyzxuxzx

yxuzyx

,

Minimize

2222

(5),

in witch are used the following notations : x= ap , y=f , z=Vc and u= VB. This result show that the minimum Ra=0,0808553μm, can be archived in the following cutting conditions ap=0,183896mm, f=0,0496018mm/rev, Vc=186,225m/min and VB=0.0963638mm. The last value shows that, the best roughness don’t correspond to the new tool If the flank wear is uniform, it does not increase the surface roughness [11].

5. Conclusions

The established second degree polynomial relationship for the Ra calculus as a function of the cutting conditions and the flank wear approximate in a satisfactory way the studied phenomena.

The main influence on the surface roughness is exerting by the feed rate and flank wear. However, we have to notice that the cutting speed and the depth of cut influence indirectly the roughness surface because the increase of this parameters result in increased chemical wear of PCBN cutting tool. Also, the interactions of some parameters as feed rate and flank wear have a great influence on the roughness values of the machined surface.

References

[1] Barry J. Byrne G. – Cutting tool wear in the machining of hardened steels. Part II: CBN cutting tool wear. Wear 247/2001, pp 152-160. [2] Benga G. Stanimir Al. – Wear patterns of polycristalline cubic boron nitride cutting tools when machining hardened bearing steel. Annals of DAAAM for 2007 & Proceedings of the 18th International DAAAM Symposium 24-27th October, 2007, Zadar, Croatia, pp. 73-74. [3] Chou YK. Evans CJ. Barash MM. – Experimental investigation on CBN turning of AISI 52100 steel, J.Mater. Process Technol. 134/2003, pp.1-9. [4] Cochran. W.G. , Cox.C.M., Experimental Design, Willey, New York , 1957. [5] Kishawi, H.A., Elbestawi, M.A. – Tool wear and surface integrity during high speed turning of hardened steel, Proc.inst.Mech. Eng. B, 215/2001, pp. 755-767. [6] Matsumoto Y. Hashimoto F. Lahoti G. – Surface integrity generated by precision hard turning, Annals of the CIRP, 48/1/1999, pp. 59-62. [7] Mihai.R., Introducere in strategia experimentarii , cu aplicaţii in tehnologia chimica , Editura Ştiinţifica şi Enciclopedica, Bucuresti 1976 [8] Nakayama K. Arai M. Kanda T., - Machining caracteristics of hard materials, Annals of the CIRP, 37/1/1988, pp. 89-92. [9] Ozel T. Hsu T.K. Zeren E. – Effects of cutting edge geometry, workpiece hardness, feed rate and cutting speed on surface roughness and forces in finish turning of hardened AISI H13 steel, Int. J. Adv. Manuf. Technol. 25/2005, pp. 262-269. [10] Stanimir. Al., Pascu. I., Ivanus. Radu., Experimental program for the determination of second degree polynomial functions with four variables used in the calculus of the surface roughness – International Conference UNIVERSITY S DAYS, 24-26 mai, Tg-Jiu, 2002. [11] Stanimir. Al., Pascu. I., Buzatu St. Geonea I., Chip formation and cutting forces in the machining of hardened Rul1V steel, Annals of the Oradea University, 2008.

Documents

EFFECT OF NANOPOWDERED ADDITIONS OF TUNGSTEN … · 2011-11-09 · Hard alloy BK8 ГОСТ 3882-74 was used as an initial material, tungsten monocarbide nanopowders of 5 mass.%, 10