Half sine acceleration pulse FEA simuation.

[mechanicalFEA]

Hello all,

I wonder if anyone can help me with a simulation i am having problems with... here is the problem...

A flat disc with a central mass is subjected to a shock test. The disk is fixed to a rig which subjects it to a half sine acceleration pulse which lasts for 6ms (0.006s) with a peak acceleration of 500ms^2. I need to determine if it will fail.


(load case 1)
I have done a quick test using ansys mechanical by applying a global acceleration of 500ms^2 to the disk and fixing the outer bosses. i believe this is subjecting the disk to a far greater load than it would experience in the real test (neglecting brittle failure)

(load case 2)

I have also done a transient analysis by integrating the acceleration to get a velocity, and then integrating velocity to get the rate of change of displacement with respect to time. I split the 6ms into 30, 0.0002s load steps and its corresponding displacement as time increases. The displacement was applied upwards to the outer edge.


Questions..
1)
assuming the correct values of displacement with time were input into the fea model, and sufficiently split into small enough load steps to solve correctly, would load case 2 be an accurate way to model a half sine acceleration pulse?


2)
Is there an easier way to model this type of loading case, for example using frequency or modal analysis?

3)
What would be a good/practical way to validate some of the results i am getting. (other than f=ma, when i check my reaction forces they equal approx 170 (500*0.336) where 500=a, 0.336=m and f=170

Any help, advice or hints and tips with this would be very much appreciated, i am rattling my brains with this problem.

Thanks in advance,

 

[msquared48]

Just a thought, but wouldn't you get a better response by posting this in the Mechanical Other Topics forum?

If you choosde to, redflag this string to be removed, as posting the same post into more than one forum is not allowed.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[Ron]

Since this is a half-sine pulse, I would not treat it as a dynamic analysis. For a single short pulse, determine the instantaneous displacement from the impact and apply that as a deflection to the plate.

 

[GregLocock]

I believe the usual approach is to perform a dynamic (frequency response ) analysis of the plate, FFT the pulse, multiply the two together and then transform the resulting spectra back into the time domain. This assumes that the system is linear, which is reasonable, and that you know the damping of the plate, which is not necessarily reasonable. What is the frequency of the first resonance of the plate?





Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[mechanicalFEA]

Cheers everyone for you responces


greg,

i have just done a quick modal analysis and found the first frquency to be 83.98Hz which i beleive to be very similar to the frequency of the input half sine wave. (83.333 Hz) Im new to frequency and modal analysis but have tons of experience with statics. So where can i go from here?, I understand if the excitation frequency is similar to that of the natural resonant frequency then the part will resonate and is more likely to fail.

 

[GregLocock]

Wow. OK so now you know the mode shape and the peak deflection so can work out the strains at that frequency and see if they exceed your limits.

Oddly enough it'll be a very similar answer to Ron's approach because of Rayleigh Ritz. It wouldn't if the excitation frequency>> first mode.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[mechanicalFEA]

Greg,

i assume the first mode is the mode with the lowest frequency?

The displacement results i get from the modal analysis are not representative, nor the stress and strains. i thought modal analysis only tells the user how the part will behave, and at what frequencies are critical to excitement. To get realistic responses (i.e displacement, stress and strain) i have to excite the model with the important frequencies obtained from the initial modal analysis???

Can anyone briefly explain the typical process of doing this? im using catia and also have ansys.

any help is very much appreciated

thanks in advance