cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

General questions regarding static analysis

SOLVED
Highlighted
Newbie

General questions regarding static analysis

I have carried out a few design analyses with Pro Mechanica and understand that the following steps need to be executed to analyze any part. Please correct me if I am wrong below : -

 

1. Design the part.

2. Open Mechanica application from the menu.

3. Define loads and constraints.

4.Define material properties (SS, graphite, etc)

5. Define mesh (generally AutoGEM creates a mesh automatically, but we can define the element size)

6. Open analysis window and define type of analysis(single pass, multi-pass etc)

7. Start analysis

8. View reports

 

Based on the video Lec 21 | MIT Finite Element Procedures for Solids and Structures, Nonlinear Analysis - YouTube , I came across a few queries that need clarification : -

 

1. Is it possible to analyse a part with only loads and no constraints(tensile loads in the opposite directions at opposite ends of a part or compressive loads squeezing a part)?

 

2. Where to I check the plasticity and permanent deformation effects on a part? The reports in point 8 above show different levels of von mises stress across the part so I assume that if the stress levels cross the yield strength of the material then the deformation is permanent. Can anyone confirm this?

 

Edit: After my How to calculate uniform load across a surface earlier on this forum, I went ahead and took up FEA to develop better understanding of the topic. My understanding stems from reading texts/watching videos of Klaus Jurgen Bathe and might be outdated in the current context so feel free to add to my knowledge wherever you find fit.

 

Thanks in advance,

Lucky

1 ACCEPTED SOLUTION

Accepted Solutions

Re: General questions regarding static analysis

1. I believe it is possible, using "inertia relief", but i highly recommend against it. The problem is that the software compensates for any static movements, but it does not show you it is doing this. If there is something wrong in your model it will likely be compensated for and you will never notice. It is much better to think about good constraints, even though it may need some more thoughts. In your case, maybe you can use a symmetry constraint, or otherwise use the 3-point constraint system where 1 point is constrainted in all directions, 1 point in 2 directions, and 1 point in 1 direction, so that the part is free to deform in all directions while still being fully constrained. Edit: look here for an explanation on the 3-point constraint system.

2. If the von Mises stress is higher than the yield stress, then yes you will have plastic deformation... but this may not be the best way to check. You can also output "plastic strain", and if that is larger than 0 you always have plastic deformation (as the elastic strain is excluded from this variable). Another way to do it of course is to unload the model and check deformations, but that will add calculation time to your analysis.

Here is a good PDF on the subject, posted by gfraulini:

Elasto plastic material.pdf

1 REPLY 1

Re: General questions regarding static analysis

1. I believe it is possible, using "inertia relief", but i highly recommend against it. The problem is that the software compensates for any static movements, but it does not show you it is doing this. If there is something wrong in your model it will likely be compensated for and you will never notice. It is much better to think about good constraints, even though it may need some more thoughts. In your case, maybe you can use a symmetry constraint, or otherwise use the 3-point constraint system where 1 point is constrainted in all directions, 1 point in 2 directions, and 1 point in 1 direction, so that the part is free to deform in all directions while still being fully constrained. Edit: look here for an explanation on the 3-point constraint system.

2. If the von Mises stress is higher than the yield stress, then yes you will have plastic deformation... but this may not be the best way to check. You can also output "plastic strain", and if that is larger than 0 you always have plastic deformation (as the elastic strain is excluded from this variable). Another way to do it of course is to unload the model and check deformations, but that will add calculation time to your analysis.

Here is a good PDF on the subject, posted by gfraulini:

Elasto plastic material.pdf