Maximum step size and print interval in Transient Analysis

Davide_21943

Hello,

If I set hmax=1e-3 and print interval = 0.1, does the print interval affects the performance of the simulation or just the visualization of the results?

I'm talking about a generic transient analysis.

Praful

Hello Davide

Print interval represents the frequency of recording the results and affects only visualization. The maximum step size controls the steps that the solver takes during the simulation and affects the performance. Except, in the case where the print interval is smaller than the hmax, the solver would force the hmax to be equal to print interval.

Hope this helps

Regards

Praful

 

 

Praful

Hello Davide

Print interval represents the frequency of recording the results and affects only visualization. The maximum step size controls the steps that the solver takes during the simulation and affects the performance. Except, in the case where the print interval is smaller than the hmax, the solver would force the hmax to be equal to print interval.

Hope this helps

Regards

Praful

 

 

Samir Zemam

Hello;

I am training now on MotionSolve software and I had questions on Print interval and Maximum step size and their settings, then I found your excellent answer to my question. Now I am taking this opportunity to ask you another question in the same topic. How do we choose the Print interval & Maximum step size values to get good visualization and performance (for good compromise between them !) ?

 

thanks !

Praful

Samir Zemam said:

Hello;

I am training now on MotionSolve software and I had questions on Print interval and Maximum step size and their settings, then I found your excellent answer to my question. Now I am taking this opportunity to ask you another question in the same topic. How do we choose the Print interval & Maximum step size values to get good visualization and performance (for good compromise between them !) ?

 

thanks !

Hello Samir. Welcome to the MotionSolve community.

Generally, print interval and hmax (max. step size) are not in any conflict that you have to do some kind of a compromise between the two. Of course print interval can force a hmax value.

As I said, print interval is the interval that the solver saves the output. A default of 0.01 sec is a good value, but the value should be chosen such that it captures well the force/displacement/velocity/acceleration peaks and provides enough resolution in the animation that aides the understanding of how the model is behaving. E.g.If there are high transients such as impulse events, you might have to chose a lower print interval of 0.001 or even lower.If the simulation has to be compared with test results, for correlation, the print interval might have to be matched with sampling interval of the test. This is also true if high frequency events have to be captured. 1/10th or 1/20th of the sampling interval is typical.

A lower print interval could lead to huge result files in terms of size. If that is a concern, the simulation can be controlled around the time of interest by having a lower print interval for a certain time range of the simulation and a coarser value at other times.

Regarding hmax, if the model behavior is continuous, a higher hmax works. The solver has mechanisms to auto-cut the step size if it has trouble converging. If the model has severe transients or discontinuities then a lower hmax is demanded for a robust simulation. A general good practice is to begin with 0.01 and get lower value by some trial and errors to obtain a robust simulation as well as time taken. If the solver repeatedly wavers between a higher and lower step size, it is better to set a value lower so that the solver marches smoothly.

It is all the more important to create good models that the solver can handle. I would recommend to go through these modeling tips.

 

Good luck !

Regards

Praful