Can we perform sine on random vibration analysis in solidworks simulation . according to military standards MIL-STD-810 G
Yes you can do that. You need the "Linear Dynamic" analsysis package. I believe it comes with what is called "Simulation Professional", but check with your VAR.
Yes the "sine on random" is a commonly deployed test technique. You can calculate the energy input from the sine function (or a summation of several known excitation frequencies), then combine the energy levels to the random source (linearity allows principle of superposition to hold for this case). Then you can use the "Random vibration" analysis option in SW Simulation.
Good to know. My experience is that those curves are just referred to as random vibration curves. When a sine is involved, either harmonic or time history is usually called for.
But definitely sim prem.
can you share me the procedure to do so??....any docu such kind..??
Which part? The sine to random or setting it up in solidworks simulation?
The sine to random ...
do you have a copy of the standard?
sine and random vibe don't really go together from my experience.
but anthony is right, if you have a dynamic analysis, you need a higher level of solidworks, solidworsk premium. if you dont' have that license, drop me a line at email@example.com, we can help you with that analysis. if you have the license but need a push in the right direction, we also offer training and mentoring on those analyses.
I am from VAR too..:). Sine on RVA can be done in solidworks simulation premium.
In Sine (harmonic) analysis we give acceleration vs real time as the input...
In RVA we give acceleration2/hz vs frequency as input.....
How to combine these two to do Sine on RVA.
I hope Anthony has got answers..
from what I recall you have to convert the sine input to have the same units as the random vibration input, then just add them. I believe you will need to calculate a GRMS power value for each sine frequency and add it to the appropriate frequency band in the GRMS of the random vibration signal. Then, just input the whole band on the loading curve of Simulation. I know that Steinberg's "Vibration Analysis for Electronic Equipment", 1988 (Wiley-Interscience), has some details about this. Hope that helps.
as you wrote for sine: you have the acceleration vs time record so you can generate a set of frequency 'bands' where the acceleration is nearly constant, square it, and divide by the bandwidth. You will need some software to sort-out (identify) the frequencies into 'bands' which have level (or nearly level) acceleration values.
Example: the band from 1000 to 5000 Hz has fairly const acceleration within some tolerance, say 10 m/s^2, the bandwidth is 4000 Hz and the center frequency point is at 3000 Hz [1000 +(5000-1000)/2]. So the first point will be: 100/4000 = 0.025 (accel2/Hz) at 3000 Hz. Go up the spectrum from 5001 Hz to say, 10,000 Hz or for as many bands as you can find with nearly const acceleration. I believe you could get some open source software if you look hard enough, or just write your own alorithm to do that.
True. I have done it the other way - convert random into sine, but from some internet reading it seems most experts say to do it like you are describing, convert the sine peaks into narrow band psd rectangles. Some say to double their value to get closer to a 2 sigma value.
Retrieving data ...