When is it appropriate to do a response spectrum analysis?


Usually only when required to do so by the specification your customer has written for the analysis. It's most common in the nuclear industry. You may also want to do one as a quick way of obtaining the peak response to a time-history base excitation. In that case the RESP command in POST26 can be used to create a response spectrum from the time-history information. A new feature in 14.0 is the capability to use an acceleration time-history with the RESP command. A sample of converting a time-history data to a response spectrum is below. The data could also be read from a text file with *VREAD.

! creation of a response spectrum from a time history stored in
! array parameters var1 (time) and var2 (data) the array parameter
! freq contains the frequencies to be computed

*dim,var1,,20 ! store input in 2 array parameters
*dim,var2,,20
*dim,freq,,30
! time points of time-history (could *VREAD)
*vfil,var1(1),ramp,.001,.001
*stat,var1
*vfil,var2(1),ramp,0,0 ! data of time-history (could *VREAD)
var2(10)=1 ! create impulse time history - should produce
! flat unit spectrum
*stat,var2
*vfil,freq(1),ramp,5,1 ! frequencies 5, 6, ... , 34

*stat,freq

/prep7 ! make a dummy model
n,1
fini

/post26
store,alloc,30 ! allocate the number of time points
vput,var1(1),1 ! vput the array parameters into post26 variables
vput,var2(1),3
vput,freq(1),2
prvar,2,3
resp,4,2,3,1,.05,.001
prvar,2,4 ! variable 4 vs 2 is the response spectrum
xvar,2
plvar,4
fini
/exit,nosa





Show Form
No comments yet. Be the first to add a comment!