Wiliam Regone et al. Metalurgia & Materiais - scielo.br · Wiliam Regone et al. Metalurgia & Materiais Numerical and experimental analysis of the microstructural evolution during

495REM: R. Esc. Minas, Ouro Preto, 62(4): 495-502, out. dez. 2009

Wiliam Regone et al.

Metalurgia & Materiais

Numerical and experimentalanalysis of the

microstructural evolutionduring cross wedge rollingof V-Ti microalloyed steel

(Análise numérica e experimental da evolucãomicroestrutural durante a laminação

transversal com cunha de um açomicroligado ao V-Ti)

ResumoO desenvolvimento de processos de conformação para a melhoria da qualidade

dos produtos, bem como a diminuição das perdas de matéria-prima e, também, daenergia são os principais enfoques de várias pesquisas no Brasil e no exterior.Entre essas pesquisas está o processo Cross Wedge Rolling (CWR), que vemsendo estudado nos últimos anos para a substituição do recalque horizontal aquente na fabricação em uma única operação de eixos escalonados, pinos, eixosexcêntricos e outros componentes. Esse processo é caracterizado pela deformaçãoplástica de um tarugo cilíndrico pela ação de ferramentas em formato de cunhamontadas sobre placas planas que se movem tangencialmente uma em relação àoutra. O objetivo desse trabalho foi analisar o comportamento termomecânico deum aço laminado no CWR por simulação numérica e a evolução microestruturaldada pela simulação física. Os resultados mostram um comportamento do materialde recristalização dinâmica e suas microestruturas do campo austenítico comaspectos de alto grau de refinamento.

Palavras-chave: Conformação de metais, método dos elementos finitos, evoluçãomicroestrutural.

AbstractThe improvement of manufacturing processes to assure product quality and

reduce the amount of raw material and energy is the main objective of many

recent researches. Some of them study cross wedge rolling (CWR) as a substitute

to hot upsetting in the manufacture of stepped shafts, pins, eccentric shafts and

many other mechanical parts. In this process a cylinder is deformed by two

wedge tools assembled on plane plates that move tangentially one against the

Wiliam Regone

Doutor, Pesquisador Colaborador,Laboratório de ConformaçãoMecânica - FEM- UNICAMP

E-mail: [email protected]

Mário Luiz Nunes da Silva

Doutor, Pesquisador Colaborador,Laboratório de ConformaçãoMecânica - FEM- UNICAMP


Sérgio Tonini Button

Professor AssociadoLaboratório de ConformaçãoMecânica - FEM- UNICAMP


REM: R. Esc. Minas, Ouro Preto, 62(4): 495-502, out. dez. 2009496

Numerical and experimental analysis of the microstructural evolution during cross wedge rolling of ...

other. The main objective of this work is

to study the thermomechanical

behaviour of medium carbon steel

during hot CWR by means of a numerical

analysis. The numerical results will be

compared to the microstructure of

microalloyed steel samples which were

submitted to CWR experimental tests.

The results suggested that dynamic

recrystallization was present during

CWR and that microstructures in the

austenitic region were very refined.

Keywords: Metal forming, finite element

method, microstructural evolution.

1. IntroductionDuring the last few years, many

industries have been substituting hotupsetting by CWR to manufacturestepped shafts, pins and eccentric shafts,because CWR presents manyadvantages when compared to hotupsetting like better dimensional quality,better mechanical properties, highproductivity, reduction of discards, easyautomation, and lower levels of noise andvibration during operation.

In this process a cylindrical billet isdeformed by two wedge tools assembledon plane plates or cylinders with arelative tangential movement to eachother (Figure 1).

Stress and deformation statespresent in CWR are very complexcomprehending radial compression, axialelongation and transverse shear stresses[Weronski and Pater, 1992; Qiang Li etal, 2002 and Silva, Regone and Button,2005].

Despite these many advantages,CWR is stable only within specificconditions related to part and toolgeometry, process temperature, andstrain rate. These parameters arecorrelated and the incorrect choice ofprocess parameters can cause defectsclassified into three groups: internal(cavities and pores), superficial, andcross section deviations. The mostcommon defect in CWR is the formation

of a central cavity, also known as theMannesmann effect [Dean and Fu, 1993and Danno and Tanaka, 1984].

During the hot forming of steelparts, the density of dislocations in theaustenitic region is continuously reducedby thermal softening caused by dynamicrecovery and recrystallization.

If recovery is the only softeningmechanism, stress will raise until astationary condition is achieved when ahardening-softening balance isestablished. Dynamic recovery is lesslikely to happen in metallic alloys withlow stacking fault energy [McQueen andJonas, 1976].

Therefore, the dislocation densityreaches a value high enough to initiatedynamic recrystallization when a criticalstrain is achieved. Then a stress peak isobserved with a subsequent andcontinuous reduction of strainhardening. The softening rate thenreaches a maximum value and flow stressis reduced to a stationary state and arefined microstructure is formed[Regone, Jorge and Balancin, 2000].

The main objective of this work isto study the thermomechanicalbehaviour of medium carbon steel whendeformed by hot CWR. Numericalanalysis with commercial finite element

software was done to simulate theprocess and to evaluate the variation oftemperatures, strains and stresses.Experimental tests with CWR werecarried out to analyze the microstructuralevolution of Ti-V microalloyed steeldeformed in the austenitic region, and tostudy the products of the austenitetransformation in the CWR products.

2. Materials and methods2.1 Experimental analysis

Cylindrical billets of microalloyedsteel with the chemical compositionshown in Table 1 and with a size of17 mm in diameter and 90 mm in lengthwere machined. The billets were heatedto 1150°C, using a soaking time of20 minutes. Thereafter the billets wererolled at 200 mm/s in CWR equipment(Figure 2) [Gentile, 2004] and then air- orwater-cooled. Samples were cut in themedium cross-section region formicrostructural analysis.

The water-cooled samples werepolished and etched with an aqueoussolution, containing picric acid anddetergent, at 80°C during 60 to 120seconds, to reveal the prior austeniticgrains. On the other hand, air-cooledsamples were polished and etched with

Figure 1 - Five tooling assemblies for CWR [Dean, T.A e Fu, X. P., 1993].



Nital 2% to reveal the microstructuresformed by the austenite decomposition.The samples were analyzed by opticalmicroscopy and the volumetric fractionof the present phases was determined inaccordance with the ASTM E-562standard.

2.2 Numerical analysis

Commercial software was used tosimulate CWR with the finite elementmethod. Medium carbon steel wasdefined in the numerical analysis insteadof the microalloyed steel, since resultsfrom previous studies showed that bothsteels present similar thermomechanicalproperties.

The billet and tools were modelledto a three-dimensional analysis,considering the billet material as anelasto-plastic material and the tools as arigid material. The process conditionswere simulated to be similar to those ofthe experimental tests (temperature,rolling speed, friction coefficient, thermalconstants, and material properties).

To analyze some process variables(like effective strain, effective stress,temperature and effective strain rate) fivepoints were selected along the workpieceas shown in Figure 3. These points,located in the medium longitudinal planeof the billet, present the evolution of thebillet on the CWR tool during theprocess. P1 is the point at the centre ofthe workpiece, points P2 and P3 arelocated along the radius with P3 beingthe more distant to the center and thenearest to the surface. Points P4 and P5are in the workpiece axis and near to theend of the billet.

3. Results anddiscussion3.1 Numerical results

Figures 4a and 4b show theevolution of billet temperature andeffective strain during rolling. Figure 5shows the effective strain in points P1to P5 as a function of rolling time.

Table 1 - Ti-V microalloyed steel chemical composition (weight %).

Figure 2 - Laboratory CWR equipment and tooling.

Figure 3 - Modification of the billet during CWR and points defined on the medium

longitudinal section.

Deformation begins with the contactbetween billet and tools, and is graduallyincreased in all points. If results in pointsP2 and P3 are compared to P1, a greaterdeformation in P3 is observed, since thispoint is the nearest to the billet surfacealong the radial direction and is closer to

the tool surfaces. Points P4 and P5present smaller strains because they arefar from the billet centre and justexperience lateral deformation (pointsbeyond P5 must show no deformationkeeping the billet with a constantdiameter in that region).



Equivalent strains present during CWR are usually greaterthan in other metal forming processes because of multi-axialdeformations present in that process. It can be assumed thatthese greater deformations are beneficial to reaching the criticaldeformation necessary to activate the mechanisms that initiatedynamic recrystallization.

During CWR a complex stress state with compressiveand tensile stresses is established after some revolutions ofthe billet. Figure 6 shows the effective stress calculated forpoints P1 to P5 as a function of process time.

For points P1, P2 and P3, equivalent stress presents asimilar cyclic behaviour with significant variation of stress foreach moment the billet touch the dies. Also observed is aninitial period of stress increase when points P1 to P3 are directlyunder (or above) the tool wedge, followed by a period of stressdecrease when the tool edge moves away from these points.Point 3 shows the most significant variation in the maximumand minimum stresses because the greater deformationobserved in this point associated to cyclic behaviour.

Points P4 and P5 present a different behaviour. There isan initial increase of stress that remains constant till the end ofthe process, since these points mainly experience axialdeformation that causes billet elongation. Therefore, for pointsP4 and P5, equivalent stresses are directly related to axial tensilestress.

Significant thermal processes occur during CWR: heat istransferred between billet and tools, also generated by thedeformation, and finally transferred by convection to theatmosphere. The variation of temperature caused by theseconditions is important information for understanding thechanges of microstructure and mechanical properties thathappen along the process.

Figure 4 - a) Variation of temperature at the workpiece surface during CWR. b) Variation of effective strain at the workpiece

surface during CWR.

Figure 5 - Effective strain versus process time.

Figure 6 - Effective stress versus process time.

a) b)



The temperature variation withinthe billet during CWR is shown in Figure7a, again for the points P1 to P5.

Points P1 and P2 show an initialincrease of temperature caused by theheat generated by billet deformation andthen a small decrease caused by heatconduction along the billet.

The same behaviour is observed forpoint P4 with a small delay because thispoint is further from the centre of thebillet than P1 and P2. The variation oftemperature in P5 is almost negligible;the small temperature increase isbalanced by the decrease caused byconvection.

Like in the analysis of theequivalent stress, point P3 presents a

significant variation of temperature in theprocess beginning, again caused by thecyclic behaviour of the process and bythe higher deformation observed for thispoint. The intense heat transfer betweentools and billet causes the significanttemperature decrease observed in P3,followed by a temperature increase dueto the heat generated by metaldeformation.

These observations can beconfirmed in Figure 7.b that shows thevariation of temperature as a function ofthe equivalent strain.

During CWR, the microstructure ofeach region of the billet is affected bythese changes in strain, stress andtemperature, and especially in strain rate

that defines the process time availablefor thermal mechanisms like phasetransformation, dynamic recovery andrecrystallization, and grain growth.

Figure 8 shows the variation ofeffective strain rate as a function of theprocess time. At the process beginning,high strains are developed in short timesand therefore high strain rates areobserved in point P3 with each billet-tool contact. Points P1 and P2 also showthe same behaviour at smaller rates. Atpoints P4 and P5, there are low strainsduring the process and consequentlysmaller and almost negligible strain rates.

Flow stress during hot deformationis strongly influenced by these process

Figure 7 - a) Temperature versus process time. b) Temperature versus effective strain.

Figure 8 - Strain rate versus process time. Figure 9 - Flow stress curves.

a) b)



variables as can be observed in Figure 9for points P1, P4 and P5. Initially flowstress increases due to strain hardening,then a critical deformation is reached anda stress peak is observed correspondingto dynamic recrystallization. Thereaftera steady state is established by thebalance between strain hardening andthermal softening [McQueen and Jonas,1976 and Jonas, 1994].

3.2 Experimental results

a) Microstructural analysis ofwater-cooled samples

Figures 10a to 10d present themicrostructures of representativesamples from the surface to the centre ofthe workpiece, respectively. All themicrographs were observed in theautenitic region after CWR and water-cooling.

These micrographs were randomlyselected among many samplespresenting similar microstructures. Inall these micrographs, it was verydifficult to reveal the austenite grainboundaries and to define theirmorphology or orientation becausethey were very refined due to the highstrains and strain rates shown inFigures 5 and 8.

Figure 10 - a) Microstructure of samples from the region near point P3 (Optical microscope – picric acid). b) Microstructure of

samples from the region between points P3 and P2 (Optical microscope – picric acid). c) Microstructure of samples from the region

between points P2 and P1 (Optical microscope – picric acid). d) Microstructure of samples from the region near point P1 (Optical

microscope – picric acid).

a) b)

c) d)



In Figures 10c and 10d, the dark regions represent thecentral cavities generated by the Mannesmann effect [Silva,Regone and Button, 2005]. One important observation in thesefigures is that the cavities are isolated and not connected bysmall cracks.

b) Microstructural analysis of air-cooled samples

Figures 11a to 11d present the microstructures ofrepresentative samples from the surface to the centre of the

workpiece, respectively. All the micrographs were observed insamples CWR rolled and air-cooled. They present amicrostructure formed by pearlite and ferrite grains.

Figure 11a shows the microstructure near the billet surfacewith refined ferrite grains (mean size near to 5 µm) also shownin Figure 11b. In Figure 11c, near the billet centre, the coarseningof the pearlite colonies is observed, which is also observed inthe centre of the billet near the large cavities. Despite thecomplex stress state present in CWR, no other phases wereformed by the austenite decomposition.

Figure 11 - a) Microstructure of samples from the region near point P3 (Optical microscope - Nital 2%). b) Microstructure of samples

from the region between points P3 and P2 (Optical microscope - Nital 2%). c) Microstructure of samples from the region between

points P2 and P1 (Optical microscope - Nital 2%). d) Microstructure of samples from the region near point P1 (Optical microscope -

Nital 2%).



Table 2 shows the measuredproportion (with M - mean of fivemeasures and s - standard deviation) offerrite and pearlite present in the air-cooled samples related to themicrostructures shown in Figures 10a to10d.

It is observed that pearlite increasestoward the billet centre, which can beexplained by the larger strains observedin the surface (Point 3) that possiblygenerate a large number of sites for thenucleation of ferritic grains as confirmedby the numerical results shown in Figure8 for the flow stress in that point.

4. ConclusionsNumerical analysis of the CWR

process provided good results for thegradients of equivalent strain, stress,temperature and strain rate and allowedprediction of the microstructuralevolution in the microalloyed steel bythe evaluation of the stress flow curve.

Microstructural analysis showedthat CWR promoted an intenserefinement of the austenite grains duringdeformation, and that central cavities areformed from isolated nuclei present nearthe center line of the workpiece.

Experimental results also showedmicrostructures with a larger proportionof pearlitic phase in samples near thecentre of the billet. This can be explainedby the high amount of strain and largestrain rates at the billet surface thatpromoted the nucleation andrecrystallization of ferritic grains in thatregion.

5. AknowledgmentsAuthors wish to thank FAPESP - Fundação de Amparo à Pesquisa do Estado

de São Paulo and CNPq - Conselho Nacional de Desenvolvimento Científico eTecnológico, for their financial support to this work.

6. ReferencesDANNO, A., TANAKA, T. Hot forming of stepped shafts by wedge rolling with three rolls.

Journal of Mechanical Working Technology, v.9, p.21-35, 1984.

DEAN, T. A., FU, X. P. Past developments, current applications and trends in the crosswedge rolling process. International Journal of Machine Tools and Manufacture, v.33,n.3, p.367-400, 1993.

GENTILE, F.C. Estudo do processo de laminação cruzada com cunha (cross wedge rolling)

para fabricação de eixos escalonados. UNICAMP, agosto de 2004. (Tese de Doutorado).

JONAS, J. J. Dynamic recrystallization-scientific curiosity or industrial tool? Mat. Scie. and

Eng., v. A184, p. 155-165, 1994.

McQUEEN, H. J., JONAS, J. J. Recovery and recrystallization during high temperature

deformation. In: ARSENAUT, R. J. (ed.). Treatise on Materials Science and Technology.New York: Academic Press, 1976. v. 6, p. 393-493.

QIANG LI, LOVELL, M. R., SLAUGHTER, W., TAGAVI, K. Journal of Materials

Processing Technology, p. 125-126, 248-257, 2002.

REGONE, W., JORGE JÚNIOR, A.M., BALANCIN, O. Metodologia para determinar os

tipos de amaciamentos que atuam em processos termomecânicos. CBECIMAT, 14.SãoPedro - SP, 3 a 6 de dezembro de 2000.

SILVA, M.L.N. DA, REGONE, W., BUTTON, S.T. Microstructure and mechanical propertiesof microalloyed steel forgings manufactured from cross-wedge-rolled preforms. Scripta

Materialia, v. 54, p. 213-217, 2005.

WERONSKI, W., PATER, Zb. Selection of geometric parametrs of transverse wedge rollingtools. Journal of Materials Processing Technology, v. 34, p. 273-280, 1992.

Artigo recebido em 09/10/2008 e aprovado em 21/08/2009.

Table 2 - Fraction proportion of ferrite and pearlite (%).

A REM tem novo endereço:

FUNDAÇÃO GORCEIX - REMRua Carlos Walter Marinho Campos, 57 - Vila Itacolomy

35400-000 - Ouro Preto - MG

www.rem.com.br

Documents

Wiliam Regone et al. Metalurgia & Materiais - scielo.br · Wiliam Regone et al. Metalurgia & Materiais Numerical and experimental analysis of the microstructural evolution during