Bi-Linear material model & resulting stresses?

[cnuk]

I have a part setup with a bilinear material model. Yield and ultimate are 85 ksi and tangent modulus is zero. I run my analysis on a simple part and I was surprised to see stresses of 95 ksi. How can there be stresses of 95 ksi when I have a material model that should limit the stress to 85 ksi? Both principal and equivalent stress were greater than 85 ksi.

 

[Drej]

It's hard to give advice with such little detail. Expand on your problem.

 

[cnuk]

I have a shaft that is subjected to torsion. The torque is resisted by something like a hex socket. I have compression only supports on the surfaces of the hex to resist the torque. When the torque is applied there is localized yielding near a radius at the bottom of the hex. I setup a bilinear material model to investigate how much yielding takes place in the hex under load. Problem is, when I review the stress plot, the stresses at the radius of the hex drive are 95+ksi. With a material model set to 85 ksi yield strength and zero tangent modulus I cannot understand why the stress is over 85 ksi? The plastic strain plot shows some local yielding as expected, just the stress does not make sense.

 

[Drej]

1. Which component of stress are you looking at?
2. Have you done a mesh sensitivity study?
3. Have you plotted the stress-strain history at the location to see what's happening i.e. to check that your input stress-strain data matches your response?
4. Are all of the components the same material i.e. do they share the same mechanical properties? Is the point where you stresses are higher at a material boundary?
5. Which version of ANSYS are you using? Workbench? ANSYS? Version number?
6. What type of elements are you using? What is the ANSYS designation name of these elements?