Why do I get duplicate modes warning after running a buckling analysis?

Duplicate modes are possible in a buckling analysis due to the way this analysis is performed.

 

Since you are really only interested in the first (simplest and lowest) modes you are not really concerned with the higher ones up near the boundary. 
Therefore, you do not really need to worry about the warning, however here is an explanation of what goes on internally (applicable to Subspace only):

1. You request N modes
2. The Eigen solver runs. It searches for 2N modes, and iterates until the bottom N have converged (lower modes converge quicker)
3. Finally a check is run – how many modes SHOULD have been found below the Nth mode?
   1. This can be determined accurately and exactly
   2. Remember the Eigen solution is approximate, and is only as accurate as the convergence tolerance
   3. If the number of modes expected = N there is no problem.
   4. If the number of modes expected < N Eigenvalues were missed
   5. If the number of modes expected > N it's likely duplicate values exist near the boundary (e.g. a cluster around the Nth mode).
      
      If you see this, try re-running the buckling analysis, but this time ask for the number of predicted modes (in the case of the screenshot above = 20) and you will likely get a clean solve. You will see a cluster with a similar value between the number you asked for originally and 20.

 

Where do I change the number of modes requested?

This can be changed in the Analysis ribbon > Options

 

Image