Verifying Simulated Results

[Baisel008]

Hey all. I work for a company that produces aircraft avionics. The job requires thermal and mechanical analysis on circuit boards. We use Proe Creo Simulate for our simulated loads. I guess my question would be: What resources are out there to verify results? I'm usually given a Random Vibration profile. Id like to be able to know if the data I’m plugging into my simulation is accurate, and if the model is as true to real life as I can make.

 

[btrueblood]

Google "vibration test lab" and sift through the results to find one local to you. Call them and discuss what you are looking for. Contract them to do some testing on a couple of your boards/assemblies and pay extra if you need to to spend the day or so watching how they do the testing and gather results.

In the end, if your company is serious about what you do, you will end up purchasing a shaker table and control/data acquisition equipment.

 

[Baisel008]

Thanks btrueblood. My company does have a shaker table. I guess my next approach is to compare g loads on my model, to the accelerometer data from the actual unit.

 

[btrueblood]

You can also record strain gage readings, though it gets tricky (thermal compensation on a hot shaker head - take static readings before and after with temperature readings also). And strobes can help you define modes/nodes.

A crazy idea - why not source a small MEMS accelerometer and connector and have them mounted on a test board (if you have the real estate). Otherwise, you are always having to figure out the impact of the accelerometer and mounting means on your results.

 

[izax1]

Do not forget the stiffening effect and added masses of the components soldered to the PCB. They will give significant property changes from the original PCB.
Steinberg's book "Vibration analysis for electronic equipment" is a must if you are into vibration of electronics