Material Modelling of PVC for Change in Tensile Properties with Variation in Strain Rate

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 04 Issue: 07 | July -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 2352
Material Modelling of PVC for change in tensile properties with
variation in Strain Rate
Nilesh Chavan1, M.V.Walame2, Mihir Ponkshe3
1P.G Student, Vishwakarma Institute of Technology, Mechanical Department, Maharashtra, Pune
2Professor, Vishwakarma Institute of Technology, Mechanical Department, Maharashtra, Pune
3 Home Cleaning Department, Gtec-Whirlpool of india Ltd., Pune
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Design of plastic part is totally different in
comparison to metal parts. PVC belongs to polymer whose
modulus does not change with change in strain rate. Material
modelling of such material can be done in LS-DYNAwithMAT-
24 model. During impact or crash simulation of polymer
material modelling was first step to perform as material
properties of polymer varies with change in strain rate of
loading. Tensile testing with change in strain rate is carried
out on PVC specimen and validation performed using LS-DYNA
software. The linear fit of yield strength against log of strain
rate was developed for PVC and refined material model was
developed using modelling parameters.
Key Words: MAT24, LS-Dyna, Eyring equation, Hypermesh
1.INTRODUCTION
Polymer shows complex behaviour in comparison of metal
parts under loading conditions. To account thisbehaviour in
CAE package with available material model, material
modelling of polymer is important [3]. Parts made of
polymers were common in many systems of the home
appliances and automobiles duetoadvantagespolymersuch
as light in weight, easy to form complex shape, cost effective
etc. The material selected for study is used fordisplaylens in
home appliances which have electronic components
enclosed in it. To account non-linear behaviour of the
material simple material models were defined in LS-DYNA
[8]. During impact simulation selection of exact material
model as per material behaviour is important [4]. Parts
supporting electronic, electrical components and part of
aesthetics assembly were subjected to impact loadtestingin
home appliances [10]. Hence material modelling of selected
material with change strain rate is required. Objectives of
study as follows:-
 Verify effect of element form, element size and
meshing type on simulation results.
 To have equation for predicting change in yield
strength as per change in strain rate of the material.
 Material modelling ofthegivenmaterial inLS-DYNA
to account change in yield strength value with
variation in strain rate of loading.
2. POLYMER (PVC) BEHAVIOUR
The polymer behavior cannot be categorized as totally
nonlinear and hyper-elastic. The mechanical properties of
plastic show variation with change in strain rate, change in
temperature and other parameters. Most of plastic shows
nonlinear elastic behaviour prior to yielding region. Trendof
polymer response depends upon the type of polymer [5].The
behaviour of the PVC may be consideredasanaloguestosteel
in yielding region [2]. PVC shows reaching maximum stress
beyond which it mayundergoneckingformationwhichcause
strain hardening of material and sudden increase in stress
value after showing yielding behaviour.Theyieldvalueofthe
PVC depends on strain rateand temperature.Duetoincrease
in strain rate and decrease in temperature its value got more
influenced.
In this paper, PVC is used for material modelling. The
behaviour of the selected PVC is slightly different than
normal PVC behaviour. It does not have abrupt stress
increase in post yield region. Hence material modellling of it
is quite difficult than other PVC. In general during simulation
negatives slopes in stress-strain curve were mostly avoided
[3].
3. MATERIAL MODELLING AND SPECIMEN
PREPARATION
The material modelling starts withspecimen preparation for
tensile testing. For impact simulations it was essential to
have material data with change in strain rate of testing. For
fidelity of results obtained from experimentation ASTM
standards needs to follow as per material. Figure.1 shows
general flow chart to followinmaterial modelling procedure.

Figure.1: Flowchart for material Modelling
The main objective of material modellingistoreplicateclose
behaviour of the any material under givenloadingcondition.
For good modelling purpose experimentation data with less
noise data was important, Hence experimentation as per
standards was important. The process starts with
preparation of the specimen as per ASTM standards and
carrying out the experimentation [7]. The data obtained
from the test has been processed and Force-Deflectioncurve
was generated which will be usedforvalidationpurpose. For
creating material model which can account variation of
strain rate in modelling, tensiletestdata withdifferentstrain
rate was required. By using experimentation data material
model parameters were calculated and processed for
simulation validation. In Ls-Dyna for polymer materials
generally linear elastic plastic model was used [4]. The
simulations with different combinations will be repeatedtill
close validation with Force-deflection curve was not
established.
The specimen geometry for experimentation purpose was
prepared as per ASTM D638 standard. The type of specimen
used depends on thickness of material and its availability.
The specimen has been cut from Injection moulded
component. The variation in thickness of the specimen is
within the tolerance limit as per standards. The shape of the
specimen was dumbbell shape as shown in fig. 2.
Figure 2: Tensile testing specimen
Table 1: Dimensions of the Specimen
Abbrevation Description units Dimension
s
G Gage length mm 25
L Length of Narrow
section
mm 33
D Distance between
Grips
mm 65
Lo Overall length of
the specimen
mm 115
W Width of the
narrow section
mm 6
Wo Width overall mm 19+6.4
T Thickness mm 3
Ro Outer Radius mm 25
R Radius of fillet mm 14
4. EXPERIMENTATION
The tensile testing was performed on Universal testing
machine. The UTM was Advanced Computerized
Electromechanical System machine. This machine consist of
PC controlling system which works in close loop control for
measuring parameters such as cross head stroke, loading
force etc. The load cell used inside of machinehascapacityof
10000N with least count of 0.1N force. Contact type high
precision EX1 extensiometer was used for displacement
measurement.
Figure 3: Tensile testing Specimen
The testing was carried out at different strain rates. Initially
data of PVC for 50mm/min was collected by repeating same
test number of times at same strain rate as per ASTM
Standard. The curves obtained from these test helps to have
smoothed material data available for modelling purpose. To
account strain rate variable in model dumbbell shape
specimen testing was carried out at 500mm/min,
50mm/min, and 10mm/min.
Specimen preparation as ASTM standards
Experimental Tensile Testing
Data preparation for CAE Package
Prepare Model in LS-Dyna, Simulation trial
Simulation Result interpretation and comparison
Use for plastic part simulation
Yes No

a. UTM Machine b. Extensometer Position
Figure 4: UTM machine and Extensometer Position
It was tough to measure data with accuracyatlowstrainrate
less than 10mm/min. To ensure less noise in test results
contact extensometer was used for strain measurement
purpose. After that graphs obtained from these testing has
been compared with graph obtained from 50mm/min
results to check proper data collection and use average
engineering stress-strain curve for modelling. The graphs
obtained from testing were stress-strain graphs. By using
numerical formulae’s force Vs deflection curve and True
Stress Vs True Strain curve was generated.
True Stress= σ/〖(1-µ*ε)2
True strain = ln (1+ ε )
Where, σ = Engineering Stress
ε = Engineering Strain
µ = Poisson’s Ratio
σT = True Stress
The strain hardening curve has been derived from available
stress strain curve results. This curvewill beuseful foruseof
LCSS option in Material modelling. If LCSS option was used
then use of constants C,P which were related to cowper-
symond’s equation will not be required. This option allows
user to give no. of stress-strain curves as input to model.
These curves were entered in the form of table which allows
material model to interpolate stress-strain curve with
change in strain rate.
Figure 5: Engineering Stress-Strain curve with different
strain rates
From testing following things were observed, withchangein
strain rate the mode of failure of the specimen changes.
From stress-strain graphs for all tests it has been seen that
the curve overlaps each other approximatelyininitial elastic
region and it shows shift in yield point with progressive
change in strain rate of testing. This behaviour of plastic
shows that PVC belongs topolymercategorywhosemodulus
does not changes with change in strain rate, Hence Material
model MAT24 of LS-Dyna can be used for its modelling
purpose [4].The experimental results were influenced by
noise factors involved in specimen preparation,
experimentation facility and data collection system. The
results obtained from the tensile testing can be plotted on
log scale as shown in fig. 6
Figure 6: Plot of yield Stress Vs log on strain rate
The figure 6 shows change in yield value with respect to log
of strain rate. This graph is helpful to predict the Eyring
equation which helps to interpolate yield value as per
change in strain rate. In CAE package to make use of LCSR
option Eyring equation is helpful. Eyring equation predicts
yield value shift more accuratecomparetoCowper-symonds
equation [3].From figure.6, equation giving close fit was
found out, as follows: Y = 57.36 +5.3456 x

5. SIMULATION
The validation process starts with checking the effect of
variable options available within the software for modelling
of the specimen. The variables were type of meshing,
element form, and no. of elements along thickness of
specimen etc. Making use of different type of mesh mainly
depends on shape of the component. The accuracy of the
results obtained with different meshingdependsonelement
form used for particular meshing in LS-Dyna [6]. To check
effect of element form, number of elements across thickness
two simple simulations of plates with load at free end in
cantilever condition and center load in fixed support
conditions has been performed. These simulations show
effect of element form and type of meshing on simulation
results. From it element form with respect to type of
meshing was decided and used in tensile testing simulation.
From simulation and theoretical calculations it has been
concluded that for HEX mesh 4elements across thickness
and element form 1 gives good co-relation with analytical
result. R-TRIA Mesh shows more stiffness than TRIA mesh
elements. Hence use of R-TRIA will give under-predicted
results. TRIA mesh with element form 13 shows good co-
relation. For this simulation trial PVC material was used.
For simulation with polymer, linearelastic plastic model was
used. Within these typesof modelsMAT24wasmostpopular
model used for simulation. MAT24 has ability to treat
material in bi-linear manner which differentiate its
behaviour in pre yield and post yield region. Using stress-
strain curve obtained from testing true stress strain curve
and its modulus was found out. The pre-processing was
completed in HYPERMESH software and model solvedin LS-
Dyna.
By making use of results obtained from element analysis
element form 13 was used as meshingwasdonewith TETRA
mesh. Total numbers of elements were 41638 andthenodes
were 10336.Due to more number of elements it requires
more time for solution. Boundary conditions hastoreplicate
exactly same as Experiment. The left side of specimen was
rigidly fixed while on right gripping portion cross-head
speed was applied.
First simulation was carried out to decide modulus valuefor
the material model. From results it has been seen with
E=3000MPa shows over prediction of results whereas
E=2500MPa will give under prediction of results. Hence
E=2800MPa has been selected which gives approximately
same result as experimental curve.
To model post yield region LCSS option was available. By
defining stress-plastic strain curve post yield region can be
formulate. The limitation with PVC was LCSS option will not
account negative slope curve [3]. To predict the post yield
behaviour of the material tangent modulus option wasused.
The other variable which affect the modelling results were
vp setting. Hence simulation using different vp setting was
carried out to select parameter predicting close
experimental results. The effect of vp setting with 0 and 1
shows some variation but with vp setting 1 and -1 it gives
same result.
Figure 7: Force-displacement curve with vp in LS-Dyna
The reason for less predominant effect of vp setting was use
of tangent modulus option. During these simulation trials
point need to consider is correct application of boundary
conditions. The nodes of specimen at moving ends were
constrained in Y and Z directions else it shows wrong
behaviour of specimen. The change in portion of fixing and
moving elements changes the position of failure in
simulation results. To check effect of poisson’s ratio on
results same simulations were performed with change in its
value. It shows increase in maximum force value.Toaccount
for shift in yield point of the plastic withchangeinstrain rate
LCSR option was used. This option helps to provide ratios of
yield stress vs strain rate inside the material card. After
selecting vp setting as 1 simulation trials with change in
strain rate was carried out. The graphs obtained from
simulation were compared with experimental results.
Figure 8: Stress-Strain curve
The simulation result shows close resemblance with
experimental results till yielding region. Simulation with
MAT24 shows necking phenomenon which was not seen in
testing at all strain rates, as shown in figure 9 (b)

(a) CAE specimen failure
(b) Experimental specimen failure
Figure 9: Specimen failure
6. CONCLUSIONS
The material modelling helps to understand effect of
different variables on reliability of results obtained from
simulation. We can conclude that,
1. Type of element form used has more influence on
simulation results for LS-DYNA. In lower order elements
TYPE (13) for TRIA mesh and TYPE (1) for HEX mesh will be
recommended.
2. The equation obtained from fig.6 allows interpolating
yield stress analytically with change in strain rate within
range of strain rate. Noise factors involved in
experimentation, specimen preparation,partmanufacturing
etc. gives variation in results obtained.
3. The refined MAT24 model predicts approximately good
results up to yield point. At highstrainratesimulationshows
non-linear shift prior to yield point if compared with actual
result due to neckingphenomenuminsimulationandchange
in tangent modulus at same time.
4. The MAT24 can’t account for change in shape of curve in
post-yield region. Failure strain variation with change in
strain rate can’t be accounted by MAT24.
ACKNOWLEDGMENT
I want to extend my sincere gratitude towards Mr.Mihir
Ponkshe and Mr. Sandeep patil from Gtec-Whirlpool ofindia
Ltd.,Pune for allowing the experimentation required for the
modelling work and guidance in simulation work.
REFERENCES
[1] Mohan K Neelam,Sriram Kalga. Elastic properties of
PVS pipes. Journal ofstructural engineering(2002)vol
29. pp 91-96.
[2] Yuzo Nakagawa, Satoshi Okuda, On the yield
Behaviour of Un-plasticized PVC(Analogy with the
Yielding of Mild Steel).Bulletin of JSME(1959)vol 2.pp
97-102
[3] Hubert lobo,Brian Croop.A Robust Methodology to
calibrate crash material models for polymers(2009).
[4] Lobo H,Advances in the Measurement and Modelling
of plastics for impact simulations,9th International LS-
Dyna User’s Conference (2006).,Detroit
[5] Juan A Hurtado, Hubert Lobo. Characterization and
modelling nonlinear behaviour of Plastics, ABAQUS
User conference(2006),Boston
[6] Suri Bala,Jim Day.General guidelinesforcrashanalysis
in LS-Dyna (2006).
[7] Standard test method for tensilepropertiesofplastics.
ASTM International (2014).
[8] LS-DYNA Keywords User’s Manual, Livermore
Software Technology Corporation (LSTC). Volume II
Material models (2014).Revision 5442.pp 154-160.
[9] Designing with Plastic-The fundamentals (2006),
Ticona Engineering Polymers.
[10] UL746C: Polymeric Material--Use in Electrical
Equipment Evaluations (1997). Underwriters
Laboratories. pp 75-76.

Material Modelling of PVC for Change in Tensile Properties with Variation in Strain Rate

More Related Content

What's hot (18)

Similar to Material Modelling of PVC for Change in Tensile Properties with Variation in Strain Rate (20)

More from IRJET Journal (20)

Recently uploaded (20)

Material Modelling of PVC for Change in Tensile Properties with Variation in Strain Rate