negative eigenvalue for linear buckling 1

[jargon5]

Does a negative eigenvalue in linear buckling with a magnitude greater than one signify that the direction of the load produces such a safety factor? Or does it mean that the matrix solver could not converge? Help please.

 

[ThomasH]

Your first statement is correct. The sign is connected to the loads direction.

Try with a simple column and load it with both tension and compression. The tension case should giva negative eigenvalue and the compression should give positive value. Provided that you use the same load you should get the same absolute value.

Regards

Thomas

 

[jargon5]

Thanks alot Thomas.

 

[dmacx]

Just to emphasize the point, a negative eigenvalue means that if the load were reversed, it would buckle at the applied load times the magnitude of the eigenvalue. But, the load must be reversed so if that is not one of the loads being considered then look for the lowest positive eigenvalue as the critical mode. You can try to force the eigensolver to only output positive values by setting the lower frequency range to a small number as well as the number of modes desired in the ND field on the EIGRL.