Hi,
When performing normal modes analysis on a shell with a laminate attached to it (e.g a aluminum/aluminum sandwich panel), Nastran seems to be unable to find the correct normal modes when picking a low, non-zero value for the Young's modulus of the core. A value of zero or a value of about 1000 MPa or higher seems to give correct answers but lower ranges show incorrect normal modes.
I was wondering if somebody else has experienced this or knows what causes it.
Any insight would be appreciated.
Hi @W_A_R_W_
Numerical solutions can lose accuracy when the ratio of (maximum stiffness)/(minimum stiffness) becomes "too large". The stiffness is related to the material properties (mainly the modulus of elasticity) and the mesh (the size and distortion of the elements). It is not surprising that the results change when the material properties change.
The question is how the various stiffness contributes to the real answer. For example,
So if 1000 MPa is the lowest core stiffness that gives reasonable answers, try running the model with higher and higher a core stiffness. The graph of the results will show whether a stiffness lower than 1000 MPa will really make any difference or not, and if so, you may be able to extrapolate out to the desired stiffness.
I'm curious about your statement "unable to find the correct normal modes". Does this mean the analysis fails? Or does it calculate the modes but they are wrong? (And if wrong, how obvious is it that they are wrong?)
John
Can't find what you're looking for? Ask the community or share your knowledge.