COMPARISON OF RESULTS OF DYNAMIC RECRYSTALLIZATION ... · design, therefore demand for it only increases in recent years [5]. ... rolling process was simulated using in the finite

Journal of Chemical Technology and Metallurgy, 53, 2, 2018

354

COMPARISON OF RESULTS OF DYNAMIC RECRYSTALLIZATION RESEARCH OF HC420LA STEEL BY TWO TYPES OF TESTS ON GLEEBLE 3800

Tatiana Zhuchkova1, Sergey Aksenov2, Valeriy Shkatov1, Igor Mazur1

ABSTRACT

The article shows an influence of kind of tests on material hot defamation behaviour during physical modelling on Gleeble 3800. As material, high-strength low-alloy automobile steel HC420LA was used. Stress-strain curves and material constants based on results of flow stress and plain strain tests were calculated and compared. Besides finite-element model-ing of rolling round bar on a smooth barrel was performed taking into account the calculated mechanical characteristics.

Keywords: hot deformation, plane strain test, uniaxial compression, Zener-Hollomon parameter, finite element modeling.

Received 14 September 2017Accepted 15 December 2017

Journal of Chemical Technology and Metallurgy, 53, 2, 2018, 354-359

1 Lipetsk State Technical University Moskowskaya St 30, 398600 Lipetsk, Russia2 National Research University - Higher School of Economics Myasnitskaya St 20, 101000 Moscow, Russia E-mail: [email protected]

INTRODUCTION

At the present time, physical modeling is an inher-ent part of the development of new production methods. Over time, it becomes more accessible for studying the mechanical properties during different steel forming processes. The results of physical modeling are used in further finite element simulation of production processes, such as rolling, forging and stamping. According to them energy-power parameters of processing are calculated, physical modeling also helps to predict the microstruc-ture of steels and alloys during and after machining

During the laboratory experiments constants and mechanical properties of steels and alloys are determined depending on various parameters: strain rate and tem-perature and chemical composition. Three main types of tests for hot deformation study can be distinguished: compression tests (plane-strain and uniaxial compres-sion), torsion and tensile tests. Tests results are “stress-strain” curves, their shape and values depend on test types as it is shown in [1 - 4]. Each test has limited ap-plication, advantages and disadvantages. Comparison of different kinds of tests are shown in [2]. The complexity

arises when various types of tests are used to simulate the same processes. For example, tests for plane-strain and uniaxial compression applied to the simulation of hot rolling. The aim of this work was to compare the results obtained using these two types of tests for high-strength low-alloy automobile steel HC420LA.

EXPERIMENTAL

The chemical composition of HC420LA used in the research is shown in Table 1. This type of steel (HSLA) is used in objects that require lightweight and robust design, therefore demand for it only increases in recent years [5].

The experimental procedures were carried out in the Czestochowa University of Technology (Poland) on physical simulation equipment Gleeble 3800 using Hydrawdge module described in [6]. The samples were heated with 10°C/s rate in a vacuum (10-5 Torr) till a heat-ing temperature 1100°C with three-minute soaking and cooled with 3 °C/s rate to a deformation temperature T (980°C and 1030°C), strain rates were 0.1, 1 and 10 s-1.

The tests on uniaxial compression were performed

Tatiana Zhuchkova, Sergey Aksenov, Valeriy Shkatov, Igor Mazur

355

on cylinders 15 mm height and 10 mm diameter. For the plane-strain test were used specimens in parallelepiped shape with sizes 10×15×20 mm, deformation zone width was 5 mm. To reduce friction between anvils and a sample tantalum and graphite plates and graphite paste were used.

Detailed description of plane-strain and uniaxial compression tests with given equations of values for true strain ε and flow stress σ for uniaxial and plane-stain

compression tests at Gleeble 3800 can be found in [6].

RESULTS AND DISCUSSION

Results of the experiment are “stress-strain” curves (Fig. 1). The obtained characteristics of plane-strain and uniaxial compression are fully consistent with the gen-erally accepted principles. In all observed curves stress decreases after the peak of deformation and subsequently

Table 1. Chemical composition of HC420LA. С Si Mn P Al Cr Ni Cu Ti V Nb 0,057 0,033 1,384 0,015 0,042 0,02 0,013 0,022 0,048 0,052 0,039

Fig. 1. Comparison of stress-strain curves for uniaxial and plane strain compression tests.

Journal of Chemical Technology and Metallurgy, 53, 2, 2018

356

becomes stable. This indicates the presence of dynamic recrystallization during hot forming as it isdescribed in [7]. Despite the full compliance with common depend-ences of stress from the temperature and strain rate for both types of experiments, there are differences in the values of such characteristics curves as peak stress and the steady stage stress. Fig. 1 shows that the greatest difference occurs at strain rates 0,1 and 1 s-1 up to 20 % while an error of the experiment was only about 5 % . Thus, a choice of type test to simulate hot forming can affect the calculation of material constants.

The combined influence of temperature and strain rate on the deformation behavior of steels and alloys is described by the parameter Zener-Hollomon called temperature-compensated strain rate [1]:

exp QZRT

ε =

(1)

where ε - strain rate, s-1, T - deformation temperature, K; Q - activation energy, J; R - universal gas constant equal to 8,31 J/(mol K).

The increase of Z-parameter is accompanied by an increase in the flow stress and the probability of occur-rence of dynamic recovery increases in comparison with dynamic recrystallization. Conversely, at low values of Z-parameter dynamic recrystallization processes are more likely [1].

Zener-Hollomon parameter Z can be expressed

mathematically as a function of flow stress as described in [1]:

''''exp( )

[sinh( )]

n

n

AZ A

A

σβσασ

=

)4()3()2(

where A, Aꞌ, Aꞌ’, n’, n, β, α (≈β/n’) - constants of de-formed material.

An equation of peak stresses can be derived from (1) and (4) [8]:

1 Zasinh A

n

pσα

=

(5)

In condition of constant deformation temperature coefficients n’, and β can be found using peak stress values [8]:

' [ ln / ln ]p Tn ε σ= ∂ ∂ (6)

' [ ln / ]p Tβ ε σ= ∂ ∂ (7)

The average values of the corresponding slopes were taken to find material constants. Figs. 2a and 2b show the results of n’ and β calculation for uniaxial and plane-strain compression, respectively.

The coefficient α according to these data is equal

Fig. 2. (a) calculation of n’ and β for uniaxial compression, (b) calculation of n’ and β for plane strain.

Tatiana Zhuchkova, Sergey Aksenov, Valeriy Shkatov, Igor Mazur

357

to 0,0067 for the first case and 0,0063 for the second. The coefficient n (in conditions of constant deforma-

tion temperature) and activation energy Q (in conditions of constant strain rate) can be calculated from equations (8) and (9) according to [8]

}{[ ln / ln sinh( ) ]Tn ε ασ= ∂ ∂ (8)

[ ln (sinh( )) / (1/ )]pQ Rn T εασ= ∂ ∂

(9)

Figs. 3(a) and 3(b) show the results of n and Q calculations. In Table 2 comparison of the coefficients found are given.

An average error of equations obtained using the co-efficients from Table 2 and eq. (5) is about 1 % and 2.7 % for uniaxial and plane-strain compression, respectively.

Different information about the behavior of the material during hot deformation obtained from different types of mechanical tests, can lead to misalignment in prediction of flow character during hot forming. In order

to estimate the effect of differences observed in previous section on simulation results, finite element simulation of rolling round bar on a smooth barrel was performed. The rolling process was simulated using in the finite element software SPLEN(Rolling) [9]. The simulated process was a rolling of a round bar of 30 mm initial diameter in flat rolls of 250 mm rotated at 10 min-1. The rolling gap was 20 mm. An isothermal quasi-static forming problem was solved for the 1/4 section of the bar. The three marker points were used to track changes in the effective strain during rolling. The mechanical proper-ties of the material were set in table form according the experimental results described in previous sections and corresponding to the temperature of 980°C.

The comparison of effective strain in a cross sec-tion of a rolled workpiece obtained by simulations using different material models is shown on Fig. 4a. It can be seen that the rolling shape and character of the effective strain distribution are very similar. In order to track the evolution of effective strain with time three

Fig. 3. (a) dependence lnσp from 10-4 1/T for hot plane stain test, activation energy 394 kJ/mole, (b) depend-ence lnσp from 10-4 1/T for hot uniaxial compression test, activation energy 402 kJ/mole.

Table 2. Comparison of calculated constants for uniaxial and plane strain com-pression.

n' β α n Q A uniaxial compression

8,4 0,056 0,0067 7,7 433 14,3·1016

plane strain 10,1 0,064 0,0063 6,5 407 1,3·1016

Journal of Chemical Technology and Metallurgy, 53, 2, 2018

358

marker points were placed in the specimen as shown in Fig. 4(a). The evolution of the effective strain in these points is presented in Fig. 4b.

Much larger deviations were observed in rolling forces predicted by FE simulations. The force calculated using the results of uniaxial compression tests was found at 145 kN, calculations performed using the results of plane strain compression tests give 156 kN.

CONCLUSIONS

Uniaxial and plane strain compression tests were performed on Gleeble 3800 for HC420LA steel. The stress-strain curves obtained by different experimental techniques differ from each other. The conclusion is that the experimental technique used for the investigation of a material behaviour affects the character of stress-strain curve as well as the values of peak strain and peak stress. These differences can be explained by two possible factors. First of all, the procedure of calculation of the effective strain and the effective stress based on the measured values of tools displacement and forces acting on the tools do not take into account nonuniform distribution of effective strain and strain rate in the specimen volume. The second possible reason is that the stress state can affect the dynamic microstructure development of the material and consequently its flow behaviour. Subsequent use of the flow data obtained by different experimental methods in finite element simula-tion may result in differences of force characteristics of

simulated process, at the same time the differences in prediction of deformation are neglectable.

AcknowledgementsResearch has been carried out in the framework of

the state order of the Ministry of Education and Science of the Russian Federation according to the project № 11.1446.2017/4.6 .

REFERENCES

1. S.S. Gorelik, S.V. Dobatkin, L.M. Kaputkina, Re-crystallization of metals and alloys, Moskow, 2005, p. 432, (in Russian).

2. E. Hadasik, M. Tkocz, R. Kawalla, S.T. Mandziej, Comparison of the results of hot torsion and hot compression tests, 9th International Metallurgical Conference, 2000.

3. G.E. Dieter, H.A. Kuhn, S.L. Semiatin, G.E. Dieter, Handbook of Workability and Process Design, ASM International, USA, 2003, p. 390.

4. T.I. Cherkashina, I.P. Mazur, S.A. Aksenov, Soft reduc-tion of a cast ingot on the incomplete crystallization stage, Material Science Forum, 2013, pp. 261-265.

5. T. Zhuchkova, V. Shkatov, I. Mazur et al., Physical and mathematical modeling of flow stress curves of low-alloyed structural steel during hot deformation, XV International Scientific Conference: New tech-nologies and achievements in metallurgy, material en-gineering and production engineering, Czestochowa,

Fig. 4. (a) finite element mesh, marker points (1,2,3) position and effective strain in a cross section of a rolled rod, (b) evolution of effective strain in marker points in the simulation using the experimental data of an uniaxial (I) and plane strain (II) compression.

Tatiana Zhuchkova, Sergey Aksenov, Valeriy Shkatov, Igor Mazur

359

2014, pp. 161-164.6. H. Dyja, A. Galkin, M. Knapinski, Reologia metali dk-

sztalcanych plastycznie, Czestochowa, 2010, p. 371.7. D. Kuc, J. Gawad, Modelling of microstructure

changes during hot deformation using cellular au-tomata, Archives of metallurgy and materials, 56, 2011, 523-532.

8. Seyed Ali Asghar Akbari Mousavi, Majid Meisami,

Investigation on Zener – Hollomon parameter of a medium carbon low alloy of 1Cr – 1Mn – 1,5Ni – 0,1V under hot compression test, International Conference on Metallurgy and Materials METAL, Roznov pod Radhostem, 2010.

9. S.A. Aksenov, E.N. Chumachenko, I.V. Logashina, T. Kubina, Numerical simulation in roll pass design for bar rolling, Metalurgija, 54, 2015, 75-78.