Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
"Failed to compute elastoplastic strain variables" Error With Shrink Fit
Posted Oct 13, 2016, 9:34 p.m. EDT 2 Replies
Please login with a confirmed email address before reporting spam
Hello,
I am attempting to simulate the shrink fit of two concentric cylinders with the effects of plasticity considered.
To simulate the shrink/press fit:
1) Pair with the inner diameter of the outer cylinder set as the source boundary and the outer diameter of the inner cylinder set as the destination boundary and search method is fast
2) Contact condition using the Penalty method and Friction applied as the Static Coulomb Friction model using a coefficient of friction of 0.2
This method works quite well for the Linear Elastic case.
Once I turn on the Plasticity model, things become quite unstable. I have performed plastic analysis on other simulations in COMSOL and am quite sure I have the right settings. As an overview:
1) Material with defined Density, Young's modulus, Poisson's ratio, and Initial Yield Stress
2) Hardening function is a user-defined interpolation using Piecewise cubic interpolation and constant extrapolation. Argument units are set to "1" and the Function is set to "Pa". This defines my stress-strain curve for my material (a type of steel).
3) Plasticity model is created under "Linear Elastic Material" node and is set for "Small plastic strains" and von Mises stress is the yield function. Isotropic hardening is used and is directed to use the hardening function data. The hardening function in the material is set as a function of epe (effective plastic strain).
Upon attempting to solve, I get the error "Failed to compute elastoplastic strain variables". I've seen a Knowledge Base article and a few threads on this and I believe it is happening because my initial state is outside the yield surface. To combat this, I attempted to perform a parametric sweep of the inner cylinder's outer diameter, slowly allowing it to grow outward into the full desired shrink fit diameter so that the shrink fit is slowly introduced to circumvent this error. This unfortunately has not worked either.
Is there another way that I should be approaching this to resolve this error?
Any help is appreciated. Thank you.
I am attempting to simulate the shrink fit of two concentric cylinders with the effects of plasticity considered.
To simulate the shrink/press fit:
1) Pair with the inner diameter of the outer cylinder set as the source boundary and the outer diameter of the inner cylinder set as the destination boundary and search method is fast
2) Contact condition using the Penalty method and Friction applied as the Static Coulomb Friction model using a coefficient of friction of 0.2
This method works quite well for the Linear Elastic case.
Once I turn on the Plasticity model, things become quite unstable. I have performed plastic analysis on other simulations in COMSOL and am quite sure I have the right settings. As an overview:
1) Material with defined Density, Young's modulus, Poisson's ratio, and Initial Yield Stress
2) Hardening function is a user-defined interpolation using Piecewise cubic interpolation and constant extrapolation. Argument units are set to "1" and the Function is set to "Pa". This defines my stress-strain curve for my material (a type of steel).
3) Plasticity model is created under "Linear Elastic Material" node and is set for "Small plastic strains" and von Mises stress is the yield function. Isotropic hardening is used and is directed to use the hardening function data. The hardening function in the material is set as a function of epe (effective plastic strain).
Upon attempting to solve, I get the error "Failed to compute elastoplastic strain variables". I've seen a Knowledge Base article and a few threads on this and I believe it is happening because my initial state is outside the yield surface. To combat this, I attempted to perform a parametric sweep of the inner cylinder's outer diameter, slowly allowing it to grow outward into the full desired shrink fit diameter so that the shrink fit is slowly introduced to circumvent this error. This unfortunately has not worked either.
Is there another way that I should be approaching this to resolve this error?
Any help is appreciated. Thank you.
2 Replies Last Post Oct 17, 2016, 9:59 a.m. EDT
Please login with a confirmed email address before reporting spam
Posted:
8 years ago
Hi,
The idea is right, but if you run a parametric sweep changing the geometry, each solution starts from scratch, so this will not help you.
You need an Auxiliary Sweep, which cannot change the geometry. Instead you should use the Offset property in the Contact node, and increase it using the sweep parameter.
By the way: Your clear problem description made it easy to pinpoint the problem. I think that there would be many more questions answered on this forum if the questions were as clearly formulated as this one.
Regards,
Henrik
The idea is right, but if you run a parametric sweep changing the geometry, each solution starts from scratch, so this will not help you.
You need an Auxiliary Sweep, which cannot change the geometry. Instead you should use the Offset property in the Contact node, and increase it using the sweep parameter.
By the way: Your clear problem description made it easy to pinpoint the problem. I think that there would be many more questions answered on this forum if the questions were as clearly formulated as this one.
Regards,
Henrik
Please login with a confirmed email address before reporting spam
Posted:
8 years ago
Hi Henrik,
Thank you for the reply and guidance.
Yes I had gathered I needed a sweep of some kind but had also read that Auxiliary sweep was incapable of geometrical changes. I will give this method of increasing the offset parameter and see how my results fare. And thank you for the problem statement praise.
Thank you,
Chris
Thank you for the reply and guidance.
Yes I had gathered I needed a sweep of some kind but had also read that Auxiliary sweep was incapable of geometrical changes. I will give this method of increasing the offset parameter and see how my results fare. And thank you for the problem statement praise.
Thank you,
Chris