Problem finding yield point

[abc123ali]

Respected ABAQUS users, I actually performed shear fracture test of the joint formed by curing a binder between two plates with varying composition of binder containing metal nanoparticles. In doing a test I get some binder composition in which the stress-strain curve is almost linear but very slightly curved. It seems that may be there is lack of yielding or my be there is a yielding but its very unclear and difficult to find. I need to know how to find yield point inn such a case to develop input data of plastic yielding?

Secondly if the material is more britlle with very little yielding but it have fatigue life experimentally during the cyclic loading experiment I performed. So how to simulate it using abaqus becasue it has no yield strain and yield stress properly defined or may not have these value instead it have value of fracture strength. How to get hysteresis loop of such a problem or which input data will be supported by ABAQUS to perform hyteresis of brittle material during cyclic loading as just for ductile materials we use plasticity data? The out put data is plotted here for stress and strain curve in SI units.

Thanks for cooperation

 

[FEA way]

Are you sure there's any plastic behavior (permanent deformation) before fracture ? Maybe it's just a nonlinear elastic response.

Abaqus has procedures for cyclic loading and low-cycle fatigue calculations but you need quite a lot of properly determined material properties to obtain useful results. You should look for a way to do more experiments on your adhesive.

 

[abc123ali]

I have performed many experiments but got just like these type of results except for highly flexible ductile binder? There is no any yileding or plastic deformation

 

[FEA way]

Then maybe you should use a different material model like hyperelasticity for this adhesive. There are many models to choose from, some need only a few parameters and there are tools in Abaqus to calibrate them.

 

[abc123ali]

Why should I consider it as hyperelastic model and not consider it as brittle and then perform cyclic hysteresis on it using ABAQUS?

 

[SWComposites]

The plotted data is missing. Post a picture of the stress strain data directly, not as an attachment.

If the adhesive material is brittle, then you need to use a different approach for fatigue analysis. This is not a FEA issue, it is s fundamental strength/fatigue prediction issue.

For bonded joints in general you cannot use static test data to predict fatigue strength. You need run fatigue tests and either determine S-N curves (if there is no crack growth before fatigue failure) or determine da/dN curves for crack growth as a function of load level. The da/dN data can be used in a fracture analysis implemented in FEA.

 

[abc123ali]

What I want to do actually are not fatigue analysis but to perform single cycle or a few cycle plasticity to observe the hysteretic energy stored by the joint along with stiffness degradation and not too many cycles to predict fatigue life. If S-N curve considering low cycle fatigue is not an issue of ABAQUS FEM then why we need or what kind of cyclic loading models we solve here or what kind of results the cyclic loops gives us here as it lacks to provide fatigue life?

here is the data
0.0127886 0
0.050383195 0.0092351
0.075997413 0.0184751
0.09925967 0.02771415
0.125645625 0.03695025
0.151994826 0.0461883
0.177204794 0.055438
0.198556105 0.0646985
0.221414112 0.073956
0.246256606 0.0831385
0.273010035 0.092479
0.299395978 0.1016565
0.326149429 0.110928
0.352902858 0.120243
0.381971517 0.1295045
0.416111561 0.138766
0.448708133 0.1479795
0.484759112 0.157272
0.5219861 0.1665355
0.570862541 0.175766
0.618563065 0.1849825
0.669754737 0.19431
0.718998733 0.203592
0.770190405 0.2127815
0.827592661 0.222075
0.879960388 0.231317
0.935782384 0.2405865
0.987745861 0.249879
1.048271802 0.2591325
1.10993701 0.26836
1.161900488 0.2775765
1.227828557 0.286914
1.289861274 0.2961395
1.349211298 0.305404
1.403122359 0.314638
1.459752947 0.3239375
1.518698721 0.333204
1.572205531 0.3424175
1.610608482 0.351638
1.666834821 0.3609735
1.721921983 0.370247
1.781235173 0.379472
1.839409187 0.3887055
1.896811627 0.398
1.946827336 0.4072555
1.999966638 0.41652
2.05582556 0.425791
2.108560704 0.435049
2.164787042 0.444315
2.212487474 0.4535335
2.26445086 0.462807
2.309468782 0.472088
2.356360714 0.4813355
2.3974831 0.490596
2.438201328 0.4998665
2.478551954 0.5091
2.52974358 0.518355
2.584022242 0.52762
2.637161544 0.536885
2.698826754 0.546155
2.75659661 0.555415
2.811683772 0.56467
2.87140112 0.573935
2.925312366 0.583125
2.986573232 0.592405
4.270260526 0.951685
4.306715986 0.96095
4.34126014 0.970215
4.371100628 0.979425
4.402521192 0.988735
4.430046402 0.99798
4.456432438 1.00728
4.481238028 1.0166
4.505308786 1.025845
4.522728016 1.03509
4.533201412 1.044305
4.544079154 1.053645
4.554920336 1.06288
4.565026318 1.07214
4.572008706 1.08141
4.580534614 1.090615
4.587516632 1.09993
4.591375428 1.109195
4.596042538 1.118375
4.611183416 1.127705
4.640656118 1.13697
4.664285978 1.146225
4.690672012 1.15546
4.715881946 1.16474
4.741863638 1.174185
4.76240646 1.18344
4.784896774 1.192705
4.80702004 1.201945
4.829106382 1.21122
4.84968576 1.220435
4.872176444 1.229705
4.895842858 1.23896
4.91987669 1.248265
4.942771716 1.257505
4.96603379 1.26677
4.986172638 1.27603
5.007119434 1.285255
5.028470942 1.2947
5.046698486 1.30397
5.065293448 1.313255
5.080801742 1.32251
5.10020539 1.331765
5.117256832 1.34104
5.13044985 1.350265
5.147906006 1.35953
5.16962456 1.368805
5.188623864 1.37806
5.20879927 1.387295
5.227431158 1.39651
5.24679788 1.40584
5.266973284 1.4151
5.283657312 1.424335
5.302656986 1.433685
5.321251948 1.44291
5.339883834 1.45223
5.355024344 1.46143
5.372443572 1.47073
5.38835621 1.479965
5.404635894 1.48926
5.421724264 1.49846
5.437599976 1.50778
5.451564752 1.51704
5.467881362 1.52629
5.482213556 1.53556
5.49577399 1.54481
5.511282284 1.55408
5.524475302 1.56333
5.538440078 1.572555
5.550860968 1.581855

 

[SWComposites]

f

What are the units and how exactly on the test specimen is the data measured? What test method standard was used?

At what load level (percent of static strength) do you want to predict fatigue life?
And do you have any fatigue test data to compare to?

 

[abc123ali]

The units set were MPa for stress while N-mm for force-displacement plots. ASTM F606 AND ISO 898, ASTMD1002 is the test standard. I dont have any fatigue data and I want to predict hysteretic loops and not the fatigue life because it may need S-N curve depending on scalar stiffness degradation for a few cycles at the load level of yielding point which may not be clear here becasue the joint is more brittle but less ductile. If I get yield point or may be where this loop give big hysteresis energy but much before failure but in plastic region to sure energy dissipation which may not be achived in elastic region. What point you better suggest or the load or stress value to perform the task?

Secondly is there any requirement in which is needed in the plasticity data to execute scalar stiffness degration or only plasticity is pretty enough?

Thirdly if we get the fatigue life S-N curve by experiment what is an advantage of performing fatigue life analysis in ABAQUS over the experiment is enough?

Fourthly I want to ask very basic question when I look at the contour plot of this simulation there are some element for which the contour plots show the Von mises stress much beyond the fracture stress of the stress strain plot. Why is it so? HoW is it possible that software keeps those elements whch must be eliminated automatically and not be the part of the process after reaching FRACTURE value. How they contribute as supporting ELEMENTS TO applied load in that situation?

Fifth; Do the above data enough for isotropic hardening without any parameter to the expansion of yield surface. Is it fine to show stifness degradation after a few cycles during and beyond the hardening process? Is it pretty enough to show fretting due to numerous crack propagation but not just same to analysis of crack intensity for the single crack analysis or finite number of cracks?

Thanks

 

[FEA way]

Secondly is there any requirement in which is needed in the plasticity data to execute scalar stiffness degration or only plasticity is pretty enough?

Abaqus has damage models if that's what you mean. Check the documentation chapter "About Progressive Damage and Failure".

Thirdly if we get the fatigue life S-N curve by experiment what is an advantage of performing fatigue life analysis in ABAQUS over the experiment is enough?

Fatigue analyses (e.g. in fe-safe which is software from the same vendor as Abaqus but meant for fatigue) require the S-N curve data as input for stress-based fatigue.

Fourthly I want to ask very basic question when I look at the contour plot of this simulation there are some element for which the contour plots show the Von mises stress much beyond the fracture stress of the stress strain plot. Why is it so? HoW is it possible that software keeps those elements whch must be eliminated automatically and not be the part of the process after reaching FRACTURE value. How they contribute as supporting ELEMENTS TO applied load in that situation?

See the first answer - you need a damage model to account for that. There are also special techniques for crack propagation (also fatigue crack propagation) modeling. Plasticity itself can't account for all those effects. It's just a model describing metal yielding with a given type of hardening (also kinematic for cyclic loading). You can read more about it in the chapter "Classical Metal Plasticity". There are also additional models for rate dependence, creep and so on.