Acta Metallurgica Slovaca, 10, 2004, 4 (271 - 278) 271

COMPARISON OF THE FLOW STRESS FUNCTIONS DETERMINED

IN THE HOT COMPRESSION AND TORSION TESTS

Hadasik E.1, Jędrusik D.1,2, Kawalla R.2

1
Silesian Technical University, Dept. of Process Modelling and Medical Engineering
Krasińskiego 8, 40-019 Katowice, [email protected]
2
Technische Universität Bergakademie Freiberg, Institut für Metallformung, Bernhard-von-
Cotta-Straße 4, 09599 Freiberg, Germany, [email protected]

POROVNANIE NAPÄTOVO – DEFORMAČNÝCH KRIVIEK ZO SKÚŠOK

ZA TEPLA V TLAKU A KRUTE

Hadasik E.1, Jędrusik D.1,2, Kawalla R.2

1
Sliezska Technická univerzita, Katedra modelovania procesov a medicínskeho inžinierstva,
Krasińskiego 8, 40-019 Katowice, Poľsko
2
Technická univerzita Bergakadémia Freiberg, Ústav tvárnenia kovov, Bernhard-von-Cotta-
Straβe 4, 09599 Freiberg, Nemecko

Abstrakt
V článku sú prezentované vzťahy dosiahnuté zo skúšok za tepla v krute a tlaku medzi
priebehom napätia, deformačnej rýchlosti a teploty pre austenitickú oceľ triedy 18-8. Krutová
skúška bola vykonaná na torznom plastometri na Ústave tvárnenia kovov TU Bergakadémie
Freiberg (Nemecko) a Katedre modelovania procesov a medicínskeho inžinierstva Sliezskej TU
( Poľsko). Skúška v tlaku bola realizovaná na servohydraulickom zariadení TU Bergakadémie
Freiberg. Sledovaná oblasť teplôt bola: 900, 1000, 1050, 1100, 11500C, pričom deformačné
rýchlosti boli 0,036, 0,36 a 3,6 s-1. Výsledky dosiahnuté z plastometrických testov boli použité
za základ pre odvodenie analytickej funkcie priebehu napätia, známej v literatúre ako Hensel –
Spittelova funkcia. Konštanty v rovnici popisujúcej napäťovo – deformačnú krivku boli určené
pomocou metódy mnohonásobnej lineárnej regresie. Sú demonštrované základné rozdiely medzi
krivkami napätie- deformácia a kvantitatívnym definovaním plasticity materiálu pomocou
maximálneho napätia, deformácie zodpovedajúcej maximálnemu napätiu a deformácie do lomu.
Diversifikácia napäťovo – deformačných kriviek určených experimentálne viedla k určeniu
základných diversifikovaných konštant v Hensel – Spittelovej rovnici. Obdržané rozdiely sú
spôsobené rozdielnosťou skúšobných metód: tlak a krut, technických postupov na výskumných
zariadeniach, zahrňujúcich geometriu vzoriek a metódy pre prenos zaznamenaných signálov do
vzťahov medzi napätím a deformáciou, určením termického efektu a diferenciácie
deformačných rýchlostí. Uvedené rozdielnosti vedú k rozdielnosti parametrov v Hensel –
Spittelovej rovnici.

Abstract
In the article is presented the comparison of relationships obtained in the hot torsion
and compression tests between the flow stress, strain rate and temperature for an exemplary
austenitic steel 18-8 grade. The torsion test was performed with use of torsion plastometers
being the research equipment in Institut fűr Metallformung TU Bergakademie Freiberg
Acta Metallurgica Slovaca, 10, 2004, 4 (271 - 278) 272

(Germany), the Process Modelling and Medical Engineering Dept. at Silesian Technical
University (Poland), while the compression test was made on servo-hydraulic machine in TU
Bergakademie Freiberg. Applicable range of temperatures was 900, 1000, 1050, 1100 and
1150°C, while the range of strain rates was 0.036, 0.36 and 3.6 s-1. The results obtained from the
torsion and compression tests were taken as basis for determination of analytic function of the
flow stress known in the literature as Hensel-Spittel function. The constants in flow stress
relation were determined using the method of multiple linear regression. Demonstrated are
essential differences both, in the curves of flow stress relations and in the quantities defining the
plasticity of material, such as: peak flow stress, deformation corresponding to peak stress and the
strain to fracture. Diversification of the flow curves as defined in experiments leads to
obtainment of essentially diversified constants in the Hensel-Spittel equation. Obtained
differences are due to application of different research methods: compression and torsion; from
the technical advance of research equipment, applicable research procedures, including the
geometry of samples and the method for conversion of recorded measurement signals into
relation between the flow stress and strain, accounting for the thermal effect and differentiation
of strain rates. The appearing incompatibilities are leading to diversification of parameters of
Hensel-Spittel function.

Key words: Flow stress, strain to fracture, hot torsion test, hot compression test, material
plasticity, austenitic steel, Hensel-Spittel function

1. Introduction
The torsion and compression tests are generally approved tests for determination of
plasticity characteristics of material [1 – 3]. They have both, the advantages and disadvantages.
In compression test there appear a friction between the tool and sample, what has unfavourable
effect at determination of plasticity characteristics and must be accounted for in calculations of
the flow stress [4]. On the other hand, the torsion test is characterized by non-uniform
distribution of deformation over the length and cross-section of sample, what obstructs a
calculation of relation between a torque moment and number of twists onto relation of flow
stress and strain [5]. It can be stated [5-8] that diversification in modernity of testing equipment,
free choice in execution of test, difficulties in accounting for thermal effect and heterogeneity of
deformation can lead to essential differentiation of results obtained in different research centres.
In present work is made a comparison of the results from the hot torsion and
compression tests of austenitic steel 18-8 grade. The tests were executed at Silesian Technical
University and TU Bergakademie Freiberg.

2. Test procedure
2.1Tested material
For plastometric examinations was taken an austenitic steel 0H18N9 (X2CrNi18.9)
grade in form of rolled rods of 15 mm dia. Chemical composition of tested steel is presented in
Table 1.
Table 1. Chemical composition of tested steel 18-9.
C Cr Mn Ni Si P S B
Steel
%
0H18N9
0,028 18,40 0,187 8,90 - 0,036 0,012 -
X2CrNi18 9
Acta Metallurgica Slovaca, 10, 2004, 4 (271 - 278) 273

Prior to execution of plastometric samples the rods were solution treated from
temperature 1150°C with holding time 60 min.

2.2Determination of plasticity characteristics

Examinations adopted for determination of the flow characteristics of material were
executed in two research centres. The torsion test No.1 was performed at torsional plastometer in
Department of Process Modelling and Medical Engineering of Silesian Technical University in
Katowice (Poland), while test No.2 in the Institute of Plastic Working TU Bergakademie
Freiberg (Germany). In Bergakademie Freiberg was additionally executed a compression test. It
can be assumed that all tests made in German Institute were executed in the same deformation
conditions as those in Polish University. For tests were taken five temperatures within the range
900°C – 1150°C and three strain rates 0.036 [s-1], 0.36 [s-1] and 3.6 [s-1]. Tested material was
heated up to temperature 1200°C and held in this temperature through 5 minutes, then subjected
to deformation. All recorded data were used for determination of the flow stress and strain. To
define these parameters were taken the following relations:
- flow stress

σ P = 3 ⋅τ (1)

3M
τ= (2)
2πR 3

- strain

γ
*

ϕ= N
(3)
3

while

γ γ* γ* γ*
*
= 2 ln( + + 1 ) = ar sinh (4)
N 2 4 2

In Nadai-Mises' formula there is additionally considered the representative radius

being in relation to external radius R*/R = 0.75. A formula for non-dilatational strain has then
the following form:
* 2πRN
γ = 0,75 (5)
L
where: R – external radius of sample,
L – measured length of sample,
N – number of twists

Such approach to this problem allows to avoid additional divergencies, which can be
provoked by application of different relations for calculation of deformation, what can be
noticed in Fig.1. The calculations made for test No.1 are showing that for identical deformation
conditions, the obtained flow curves are achieving different values of the strain limit, and
simultaneously various courses of the flow curves. Application of the
Acta Metallurgica Slovaca, 10, 2004, 4 (271 - 278) 274

130
120
110

Flow stress σ p [MPa]

100
90
80
70
60 Mises 0,036[1/s]
50 Nadai-Mises 0,036[1/s]
40 Mises 0,36[1/s]
Nadai-Mises 0,36[1/s]
30
Mises 3,6[1/s]
20
Nadai-Mises 3,6[1/s]
10
0
0 0,4 0,8 1,2 1,6 2 2,4 2,8 3,2 3,6 4 4,4 4,8 5,2 5,6

Strain
Fig.1 Relationship between flow stress and strain with application of different relations for calculation of strain
(T=1000°C)

Misses' relation allows for calculation of the equivalent strain:

γ
ϕ= (6)
3

while:

2πRN
γ= (7)
L

allows to obtain larger deformations. Therefore, it is very important that at

comparison of curves obtained from two research centres, thus generated at different torsional
elastometers, to choose one method for calculation of the flow stress and strain.
In the compression test of cylindrical samples the flow stress was calculated
according to relationship, which accounts for effect of friction between a tool and sample:

σś
σP = (8)
µ dx
1+ ⋅
3 hx

In the formula µ represents the friction coefficient, which in hot working processes is
adopted as 0.175, while σś being a forming resistance (average unit pressure) is expressed by
means of formula:

F ⋅ hx F
σś = = (9)
h0 ⋅ A0 A

The strain is calculated from relation:

⎛ h0 ⎞
ϕ = ln⎜⎜ ⎟
⎟ (10)
⎝ hx ⎠
Acta Metallurgica Slovaca, 10, 2004, 4 (271 - 278) 275

All calculations were executed in AUK programme (Auswertung Umformtechnischer

Kenwerten) in the Institute of Plastic Working at TU Bergakademie Freiberg. The obtained flow
curves were additionally corrected accounting for differentiation of temperature.

3. Test results and discussion

3.1 Comparison between torsion tests
In Fig.2 are presented the flow curves obtained from torsion tests executed at Silesian
Technical University (sample No.1) and at TU Bergakademie (sample No.2) at the strain rate
0.36 [s-1] for three selected temperatures.
A characteristic feature of all flow curve courses is their mutual resemblance, in spite
that the flow curves for austenitic steel 0H18N9 in torsion test No.1 are achieving lower values
of flow stress in all performed tests.

200
180
160
Flow stress σP [MPa]

140
120
100
80
60
40 sample No. 1 T=1000°C sample No. 2 T=1000°C
sample No. 1 T=1050°C sample No. 2 T=1050°C
20 sample No. 1 T=1100°C sample No. 2 T=1100°C
0
0 0,5 1 1,5 2 2,5 3 3,5 4

Strain

Fig.2 Flow curve courses in torsion tests determined in two different research centres.

In Fig.3 are presented diversified values of flow stress obtained in both torsion tests in
function of temperature, for four different strain values. As demonstrated, a diversification of the
flow stress is appearing both, for the low and high strain values. Diversification is higher, the
higher is a strain rate and lower the deformation temperature.
The presented diversification does not result from adopted methodology of
calculations, because calculations were conducted using the same relationships. It can be
assumed that this diversification results from different technical mode of test execution, heating
condition (resistance or induction type) and various geometry of sample. In tests executed at
Silesian Technical University the sample had a measured length 50 mm, while that at TU
Bergakademie Freiberg had 15 mm.

3.2 Comparison between the torsion and compression tests

In Fig.4 are presented the flow curves for austenitic steel 0H18N9 determined in the
hot torsion and compression tests. The curve from test No.1, independently from adopted
conditions of deformation is situated below a curve defined in the compression and torsion test
No.2. Diversification of flow curves depends on the temperature and strain rate, however, one
Acta Metallurgica Slovaca, 10, 2004, 4 (271 - 278) 276

cannot ascertain such evident dependency on the parameters of deformation. This is due to a
large scatter of results. It should be noted that in torsion tests was used one sample only to one
strain condition, while in compression tests there were two samples. It is probable, that
execution of larger number of repetitions would bring a better explanation to the diversification
existing between tests executed at Silesian Technical University and TU Bergakademie Freiberg.

66
60
Strain = 0,2
54
Strain = 0,4
48
Strain = 0,6
∆σp [MPa] 42
36 Strain = 0,8
30
24
18
12
6
0
850 900 950 1000 1050 1100 1150 1200

Temperature [0C]

66 Strain = 0,2
60 Strain = 0,4
54 Strain = 0,6
48 Strain = 0,8
[MPa]

42
36
30
P

24
∆σ

18
12
6
0
850 900 950 1000 1050 1100 1150 1200
Temperature [ 0C]

Fig.3 Differences between flow stresses in function of temperature and strain for torsion test No.1 and No.2 at various
strain rates.

180

160

140
Flow stress σP [MPa]

120

100
torsion No. 1 T=1000°C
80
torsion No. 2 T=1000°C
60 compression T=1000°C
torsion No. 1 T=1050°C
40
torsion No. 2 T=1050°C
20 compression T=1050°C
0
0 0,4 0,8 1,2 1,6 2 2,4 2,8
Strain

Fig.4 Flow curves from the compression and torsion tests for T=10000C and 10500C at strain rate 0.36 [s-1].

3.3 Function of flow stress

The flow curves determined in different research centres and in different tests were
used as initial data for determination of the function of flow stress. For the calculations was
selected a function based on thermodynamic parameters, which was elaborated in the German
institute [9]:
Acta Metallurgica Slovaca, 10, 2004, 4 (271 - 278) 277

⎛ m4 ⎞
⎜ ⎟
m2 ⎜⎝ ϕ ⎟⎠
σ F = Ae m1ϑ
ϕ e (1 + ϕ )(m5ϑ ) e (m7ϕ )ϕ& m8*ϑ (11)

The determined constants (A, ml, ....) in equation (11) by the linear regression method
are collected in Table 2. The constants refer to the data from torsion test No.1, torsion test No.2,
the compression test and the common data from torsion test No.1 and No.2.
Table 2 Constants defined in equation (11)
Constants Torsion test Torsion test Torsion test Compression test
No.1 No.2 No.1 and No.2
A 19077,8 20774 20518,1 5786,27
m1 -0,00408 -0,00449 -0,0045 -0,00262
m2 0,337209 0,215451 0,212194 0,403876
m4 0,001025 -0,00047 -4,4E-05 -0,00026
m5 -0,00133 6,18E-05 -0,00035 -0,00263
m7 0,311025 -0,14797 0,101659 0,952045
m8 0,00012 0,000104 0,000118 0,000125
R2 0,9782 0,9812 0,9639 0,9881

The differences defined between the constants in equation (11) are inevitable, since as
presented in Fig.2 and 3 there exists essential diversification of flow curves for steel 0H18N9.
However, it can be stated that the parameter ml characterizing the influence of temperature is
showing similar values in torsion test No.1 and No.2, while another value in compression tests.
Parameter m6 characterizing the effect of strain rate is showing similar values for the torsion and
compression tests.

120

100
[MPa]

80
P
Flow stress

60
torsion test No. 1
40 torsion test No. 2

20
torsion test No. 1and No. 2
compression test
0
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2
Strain
Fig.5 Comparison of flow curves defined by equation (11) for T=11000C and strain rate 0.36 [s-1].

It is difficult to evaluate the influence of deformation upon flow stress, as it is

characterized by four parameters m2, m4, m5, m7 having much diversified values for the torsion
and compression tests. In Fig.5 are showed exemplary courses of curves obtained by means of
equation (11). A similar character is showing the curve courses from compression test made at
TU Bergakademie Freiberg and the torsion test executed at Silesian Technical University.
Though the differences in values of the flow stress are achieving up to 20 [Mpa], yet still both
Acta Metallurgica Slovaca, 10, 2004, 4 (271 - 278) 278

courses are are showing a distinct maximum the exceeding of which results in drop of stress
untill coming to a steady state.
4. Summary
Execution of plastometric tests in different research centres, aimed at determination of
material plasticity characteristics in hot forming processes, leads to significant diversification of
the courses of flow curves and defined values of flow stress and strain to fracture. This
diversification results from adopted test (compression, torsion), applied testing equipment,
possibilities to maintain required conditions of deformation, as well as adopted methodology for
execution of tests and computing of recorded values. Any incompatibilities can lead to
diversification of defined parameters in function of the flow stress, although it is proven that the
influence of temperature and strain rate on the flow stress is defined in similar way in the torsion
and compression tests.

Literature
[1] Schotten K.: Mathematische Beschreibung der Fließkurve bei der Warmumformung
verschiedener Stähle. Shaker Verlag, D82 (Diss. RWTH Aachen) Band 14/2000
[2] Hensger K.E.: Dynamische Realstrukturprozesse bei Warmumformung.
Freibe rger Forschungshefte B323 (1982)
[3] Gronostajski Z.: Modele konstytutywne opisujące zachowanie się wybranych
stopów miedzi w zakresie dużych odkształceń plastycznych. Wrocław 20 00
[4] Doege E., Nolkemper, H.M., Saeed, J. Fließkurvenatlas metallischer Werkstoffe
und eine grundlegenden Einführung. M ünchen 1986.
[5] Hadasik E.: Metodyka wyznaczania charakterystyk plastyczności w próbie
skręcania na gorąco. Politechnika Śląska, Zeszyty naukowe Nr 1546, Gl iwice
2002
[6] Hadasik E., Płachta A., Gronostajski Z.: Anali za sposobów wyznaczania
naprężenia uplastyczniającego w próbie skręcania na gorąco. Rudy i metale
nieżelazne R-46 2001 nr 11, s. 504-507
[7] Hense l A., Spittel T.: Kraft- und Arbeitbedarf bildsamer Formgebungsverfahren.
VED Deutscher Verlag für Grundstoffindustrie Leipzig 1978
[8] Spittel M., Spittel T.: Ver- und Entfestigung bei der Warmumformung. Freiberger
Forschungshefte B235 (1982)
[9] Spittel T., Spittel M., Tachert H.: Umformeigenschaften von Aluminium und
Aluminiumlegierungen. Aluminium 70 (1994), s. 69-75