FEA Course Sample: Part of Example 5

Stability can catch you off guard from time to time. If you are used to stress analysis I would be extra cautious. In some models, I had calculated buckling happened with stresses as low as 25% of yield strength! This is something that you will never learn from linear static. This is also why we have made the linear buckling analysis.

Takeaways:

• If something is slender (aka “thin and long”) and under compression always check buckling! As you could see it did not cost a lot to perform LBA. It was only few settings and few minutes of computing.

• LBA won’t solve the problem for you in many cases. In fact it is only good with extremes. If you get a very low multiplier (that is below 1.0) you may be sure that your thing is going to fail. If the multiplier is very high you may be sure things will be ok. Stuff “in between” is more difficult. I would advice nonlinear analysis in those cases.

• Notice that I did not specify what a “very high” multiplier means. Be very careful about having a “magic number”. Sadly, what is acceptable depends on the thing you are designing. For a steel truss multiplier between 2-3 is usually ok. In a shell (like silos) the same multiplier would most likely mean failure. However, as you work in your industry you will get a sens of what is acceptable and what is not in many typical cases. This is the best way to go I think.

• Also, low eigenvalue (below 1.0) doesn’t always mean failure. In things that are bent, but want to go into membrane state” LBA will show very low multipliers (as this is a linear algorithm). In such cases it is best to do a nonlinear analysis. This however has nothing to do with our example here!

Want to learn more?

I really hope that you like my teaching style. If you do, you can learn A LOT MORE in the course itself. Click below to sign up!