Earthquake simulation step size question (Linear Dynamic Modal Time History)
For the model I have, I am applying uniform base excitation in the x-axis, y-axis, and z-axis directions with real earthquake data. I’m using 30 seconds of acceleration vs time data for each direction and the step size of the data is 0.01 seconds. Thus I have 3001 time values and 3001 acceleration values for each of the three directions (starting from 0 to 30 seconds).
I changed the simulation time to start at 0 and go to 30 seconds. And in the same place I tried setting step size to 0.01. As you can expect I was greeted with a very long process. Basically it seemed it was going to take more than 5 hours for this to be solved. After about ten minutes almost 100 out of 3000 steps were solved according to dialog box.
If I changed the step size to 1 instead of 0.01 would this be faster and better in my case? Or is it better to match the input accelerations I am using and have the same step size as those (0.01)?
My goal is to obtain the response accelerations in the three axis directions. Would my results be too inaccurate? Is it better to do the 5+ hour simulation? What would you suggest?