Basics
Basics
(OP)
Lost! I'm wrestling with an aging brain and an education that happened too long ago and involved too much beer and too little class. I have three vibration requirements and am trying to compare the two for their severity.
1) 13Hz, 3G, 1.5M cycles.
2) 30Hz, 3G, 0.5M cycles.
3) Hz g^2/Hz
5 0.01374
12.5 0.04121
15 0.02747
20 0.00385
25 0.00302
30 0.00385
45 0.00769
100 0.0011
Grms = 0.79
a) Comparing 1) and 2) how do I calucate which is a harsher test? Is there a way to calucate the total energy acting on the test subject over the life of the test? I understand that one of those frequencies may be resonant tot he test subject so it could implode regardless if the energy level is less severe?
b) For 3) I have a random vibration profile and have got as far as understanding that the SQRT of the area under the curve gets me Grms. This is the average G applied to the test subject throughout the test...correct? So then can I multiply it by 1.414 to get a peak G to get 1.11G and thus assume that this test is about 1/3 the severity of 1) and 2)?
c) Now what if I have someone telling me to run a vibration test from 5 Hz to 20 Hz and back to 5Hz over a two minute period while maintaining 0.5G. Is this the same as
Hz g^2/Hz
5 0.25
20 0.25
1) 13Hz, 3G, 1.5M cycles.
2) 30Hz, 3G, 0.5M cycles.
3) Hz g^2/Hz
5 0.01374
12.5 0.04121
15 0.02747
20 0.00385
25 0.00302
30 0.00385
45 0.00769
100 0.0011
Grms = 0.79
a) Comparing 1) and 2) how do I calucate which is a harsher test? Is there a way to calucate the total energy acting on the test subject over the life of the test? I understand that one of those frequencies may be resonant tot he test subject so it could implode regardless if the energy level is less severe?
b) For 3) I have a random vibration profile and have got as far as understanding that the SQRT of the area under the curve gets me Grms. This is the average G applied to the test subject throughout the test...correct? So then can I multiply it by 1.414 to get a peak G to get 1.11G and thus assume that this test is about 1/3 the severity of 1) and 2)?
c) Now what if I have someone telling me to run a vibration test from 5 Hz to 20 Hz and back to 5Hz over a two minute period while maintaining 0.5G. Is this the same as
Hz g^2/Hz
5 0.25
20 0.25





RE: Basics
For 1) I assume that 1 and 2 are sine vibrations and the G value is peak to peak. Make sure you know how your acceleroometers are calibrated. Othwise you could be off by 1.414. The 30 hz vibration has a slightly smaller displacement than the 13 hertz but the 13 hz is for a longer period. The 13 Hz is more severe.
For 2) You are comparing apples and oranges to some extent. What you say about the G level is correct. The problem is that in random vibration all the frequencies are present. In case 1) it is only at 13 Hz. In case 2) it is all the frequencies between 5 Hz and 100 Hz. That may make case 2) more severe than case 1) but how long are you running case 2)? Fatigue and resonant issues can come into play too.
For 3) No, assuming that you are using swept sine in the running the fequencies back and forth. The second test looks like random. See 2) above about all the frequencies being present.
Aslo in going down to 5 hz. I hope you are not using piezoelectric accelerometers. We never used them below 20 Hz. They roll off at the lower frequencies.
Also, I am curious, what are you trying to accomplish? Is the a qual test, Mil-Std 810, stress screenig, or something else?
Regards
Dave
RE: Basics
Thanks for the reply. These are various auto suppliers requirements for validating chassis components.
RE: Basics
Some little remarks:
Question A:
if you are considering rigid body displacement:
accel=(2.pi.freq)^2.disp
so
3 G-peak @ 13 Hz -> 4,4 mm displacement
3 G-peak @ 30 Hz -> 0,82 mm displacement
Not considering resonance..
So 13 Hz is much more severe.
Question B:
g-rms is used to define the size of the vibratory table necessary to perform the test. You can't use this g-rms to make direct comparisons. Nevertheless, there are some approximation used to convert APSD into G-levels, only for sizing objective. I'll try to find the reference and equations.
Question C:
It seems a sine sweep. Do-RTCA-160 has good explanations of how those tests are performed. A summary can be found at Harris and Piersol Handbook of Vibration.
RE: Basics
That depends partly on the mounting location and direction of your accelerometer, I'm assuming that you are quoting the table vibration.
The same problem applies with 3.
Cheers
Greg Locock
SIG:Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of Eng-Tips.
RE: Basics
If I understand correctly uglyfatjoe already performed a swept sine test and from the data it is clear that around 13Hz the density of energy is the highest meaning there may be a resonance freq. in that area. Doesn't that mean that test 1 (@ 13Hz) is the harsher test?
RE: Basics
Cheers
Greg Locock
SIG:Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of Eng-Tips.
RE: Basics
You should first perform (3) and based on the data you may decide how to preoceed. Maybe (1) and (2) are not required at all, or maybe both required. Maybe you'll find another resonance frequency and decide to test there... lots of (?)
Good luck