Learning Forum

[Gijoys4v]

Hello Ansys,

For the last few days I was trying to simulate a tensile experiment for the model given. The experimental setup and the obtained graph from the experimental work and from my simulation is given below.

when I tried to simulate the model I obtained answer that is too big than the obtained value . there is no problem in the material properties and I tried after mesh convergence study. I simulated the model by taking the symmetry. To one end symmetry boundary condition given and to the other end a displacement of 1% of total length of the model is given

the model is a four layered thin film. Also I had given some prestrain to the model which occurred during its cooling process. the length of the model is 300mm and thickness is only 0.1mm and the width is 5mm. 

I am stuck and cannot proceed. Any help will be really appreciated. Please help me to resolve my problem. please tell me where I went wrong.

The file is also attached

[peteroznewman]

I opened your model and changed the Displacement to 0 and solved to test the effect of your INISTATE command.

The result is a compressive reaction force and a negative strain. Is that what you intended to create as an initial state?

[Gijoys4v]

Hello Mr. Peter,

thanks for the reply. Yes I wanted to give a negative strain (which is actually the residual strain within in the thin film which occurred due to the cooling)

[peteroznewman]

Did you double the displacement in your half model to represent the full model when you plotted the force and compared it with the experiment?

[Gijoys4v]

To the half model, after giving symmetry to one end, I had given only 1 percent displacement of the length of the model taken. As I took only half of the model (which is 150mm), I gave displacement of 1.5mm only

[Gijoys4v]

And the force, I took from reaction prob.

[Gijoys4v]

Any suggestions or help will be really appreciated .

whether my method is correct?

[peteroznewman]

Rerun your model with double the number of elements through the thickness to see if that changes the result. If it does double it again. This is called a Mesh Refinement Study.

A screen snapshot of a spreadsheet would help with setup data such as Exp. Sample Length, Mod. Sample Length

and columns such as:

Exp. Disp., Exp. Strain, Exp. Force, Model Disp. Model Strain, Model Force

=========================================================

Next, assemble a list of all the parameters that have a range of uncertainty.

• Length


• Width


• Layer 1 Thickness


• Layer 2 Thickness


• Layer 3 Thickness


• Layer 4 Thickness


• Layer 1 Modulus


• Layer 2 Modulus


• Layer 3 Modulus


• Layer 4 Modulus


• Layer 1 CTE


• Layer 2 CTE


• Layer 3 CTE


• Layer 4 CTE


• Delta T for Layer 1


• Delta T for Layer 2


• Delta T for Layer 3


• Delta T for Layer 4

Assign a range for each uncertainty, then assign the worst case for each parameter that will push the Force higher or lower and label those values F+ and F-.  Run a model with all the parameters at F+ and run another model with all the parameters at F- and show the possible variation in the model caused by variations in the input to the model.  See if your experimental result falls between the F+ and the F- curves.