I recommend performing a modal time history analysis with the iterative solver, if the software will allow it for your boundary conditions.
Recently, I was performing a seismic analysis for a customer, the model needed a small time step to capture the complex motion resulting in about 600,000 time steps. I started with the direct sparse solver as I typically do, since it will guarantee a solution as long as the setup is correct. After an hour, it hadn't even completed 1,000 of 600,000 time steps. I certainly couldn't wait that long. At that pace it would have been almost a month to complete!
Instead, I switched over to the FFEPlus iterative solver. It was magic. It finished the calculations in about an hour and a half!
During our training classes at Graphics Systems, customers are told about direct versus iterative solvers, but these are engineers just getting started with static problems. They might see 30 minutes instead of an hour, but the real benefits are in the more complex analyses, in this case hours over months.
Why is the iterative solver so much faster in the Modal Time History?
The iterative solver works by guessing at the displacements of the model, then iterating them higher and lower until all the forces in the model balance with the external forces. We find a huge benefit from this in small time-step linear dynamics because a good guess for the displacements of the next step is the displacements at the last step, leaving little calculation from step-to-step. This is totally opposite the direct solver that has to factor a huge matrix at each step.
There are two morals to this story. The first is that if something is taking a long time to simulate, there may be a much faster way to set it up. If you're not sure how to do that, contact someone who might. If Graphics Systems is your SOLIDWORKS Reseller give us a call!
The second moral is that Iterative solvers are usually much faster for a Modal Time History Analysis.
By: Brandon Donnelly, Simulation Applications Engineer