I am trying to model a copper busbar as a flexible body for simulating it in the Simscape Multibody environment.
I have built the model, but I am failing to generate the flexibility. I will describe the entire process. Please comment if it is the right method. If wrong, I kindly request you to suggest the actual method to be followed.
I started off by performing modal analysis on the imported STL busbar component on Matlab. The PDE toolbox was utilized. Multipoint point constraints were applied to the critical points on the body and was CB (Craig Bampton) reduced to obtain the mass and stiffness matrices. Analytically the damping matrix was also developed.
One of the points has to be considered as the reference frame, with respect to which deflections will be measured, and hence its DOFs have to be eliminated. So, as a result of rigid body modes, 6 rows and columns were manually deleted from all the matrices. Finally state space matrices were obtained.
First question, should we use an FE software like ANSYS or is it right to reduce the model and obtain the mass and stiffness matrices in Matlab using the PDE toolbox.
Once the matrices were obtained, it was superimposed with the rigid body model in simulink using joints at appropriate locations. For the rigid body model, I imported the STL file again. Because the frames created in simulink have no inertia of their own, I split the busbar into 3 solid parts and connected each frame to the split parts. I dont see the deflection upon simulation and it stops simulating showing me an error
"Nonlinear iteration is not converging with step size reduced to hmin (1.9776e-22) at time 5.56646e-08. Try reducing the minimum step size and/or relax the relative error tolerance."
Second question, Is this approach correct?
Looking forward for valuable suggestions.
Regards, Goutham Sajja