
Re: Solidworks simulation premium, how to simulate a "sine waveform vibration" test?
Jared Conway Oct 21, 2014 1:48 AM (in response to Jing Liu)1, if you have checked, do the freq at the dynamic study level and report the issue to your var.
2. Take the time aspect out and run 2 harmonic analyses. Or make a curve, run a time history and get some popcorn for a long simulation.

Re: Solidworks simulation premium, how to simulate a "sine waveform vibration" test?
Jing Liu Oct 22, 2014 1:15 PM (in response to Jared Conway)Thank you, Jared, for your suggestion.
For 1. : Our VAR is Hawkridge. I have sent an email regarding this problem, haven't got any reply yet.
For 2. : could you please be more specific with the "run 2 harmonic analysis", do you mean run one first, and then copy the result to the next one and run? Also, you recommend "run a time history", my question is: how to set up the parameters to restore the real test as I described earlier?
Thank you!


Re: Solidworks simulation premium, how to simulate a "sine waveform vibration" test?
Mark Ankrom Oct 21, 2014 1:30 PM (in response to Jing Liu)One of my major complaints about simulation is the model is dependent on the study(statics, freq response, transient). The copying model to each "study" is very inefficient from a modeling and updating perspective.
As for the sine sweep simulation. At that sweep rate (about 2 octaves/min), a lightly damped structure will hit steady state for each of the frequencies. Just perform the harmonic analysis and use resonant frequencies plus some uniformly distributed points in between. This will allow you to determine peak loading and make xy plots.
Fatigue cycles can be determined by integrating the sweep function (usually logarithmic). If I remember correctly there are 22 cycles per hertz for 2 octaves/min. You can use this with the response plot to estimate the fatigue damage.
Mark

Re: Solidworks simulation premium, how to simulate a "sine waveform vibration" test?
Jing Liu Oct 22, 2014 1:20 PM (in response to Mark Ankrom)I agree with you, the copying to the new study IS inefficient sometime. I got 23 modes from the frequency analysis between 102000 Hz. Do you suggest me to run the harmonic analysis for each of the mode? How to "plus some uniformly distributed points in between"?
Just began to do the simulation, feel like I have so much to learn

Re: Solidworks simulation premium, how to simulate a "sine waveform vibration" test?
Mark Ankrom Oct 23, 2014 12:21 AM (in response to Jing Liu)Since you already have the test data you should have a collection of xy plots of response versus frequency. your simulation objective should be to generate the same plots as the test. The harmonic analysis requires user input specifying the frequencies for calculating the steady state response.
if you only choose the natural frequencies you will only get some discontinues peaks. To make a meaningful plot you need to add points before , between, and after the natural frequencies. Depending on the structure you may only need to analyze a much smaller freq range. Analysis below the first mode will be rigid body and only requires a few points. Analysis above your mode 23 will roll off in about an octave. There is not much reason to analyze more than an octave above the last mode.



Re: Solidworks simulation premium, how to simulate a "sine waveform vibration" test?
Sean Screws Oct 23, 2014 10:56 AM (in response to Jing Liu)Looks like you have an idea of what your outputs should be. If you need RMS, then you'll need Random Vibe. If you need max value (i.e. max displacement), then you'll need Harmonic. If you need results based on time, Model Time History.
A Harmonic analysis will look at X points around each frequency within a given bandwidth. If you assume that the transient effects are negligible (taking time out of the equation), then two harmonic analyses should give you two response plots that you can export to .csv and superimpose in excel,matlab,etc.