RE: dynamic loading Factor

Further discussion:

> When user's continue extracting eigenpairs beyond the range of the
> excitation time scale, and use a modal summation method, i.e. (SRSS,
> DSRSS,
> ABS, etc.) they will find that stresses can increase without bound,
> although
> displacements do not.

Chris wrote, "
This isn't likely. The kinetic energy of higher modes (reflected in the modal mass) drops off real fast in the higher modes. "

Unfortunately it happens. (That's the nature of the problem that Parag wrote about in his initial email on the subject.) Displacements are certainly lower as a function of the square of the frequency but stresses, because of the small radius of curvature associated with higher modes continue to add. This has been a problem with the approach used in CAESAR since 1987, (although this difficulty with spectral solutions I wouldn't expect is unique with CAESAR.) It is relatively easy to test. Take a typical pipe span including several straight runs (with intermediate supports), and at least five bends. Apply an impact load to it using the CAESAR DLF generator and spectrum analyzer, and use the default 33 Hz. cutoff. Then keep increasing the cutoff frequency so that more modes are included in the solution and watch the stress increase as the number of modes included increases. The phenomena occurs with even the slowest dynamic events, (and that is the nature of the problem.)

My experience with this problem is certainly dated by more than 10 years, but I have spent some time involved in addressing this condition which to my knowledge still exists. Received on Thu Apr 28 11:54:00 2005