I have been given the delta P vs. time spectrums to do a time history analysis in order to explore the impact of surge on a piping system. my problem is the spectrums are "too" long; the pressure fluctuation spectrum is given within 60 seconds and the time steps are 10ms! (every spectrum contains 6000 rows of information!) I know that it is impossible to enter more than 500 rows in CII spectrum input. is it correct to only pick a "fragment" of the total spectrum in which the highest delta P occurs?

following the above situation, I have some more questions;
1- Since the part which I am analyzing is in the downstream of shut-off valve, (the peak point of the delta Ps are all negative figures) shall I assume the force vectors in the opposite direction of the flow? (unlike the upstream part where the force vectors would be along the direction of the flow)

These questions cannot be answered here in this Forum. Please open a Support Request on Smart Support and send in you job files.

Follow the pressures. Imagine you are looking up and down the pipe from the inside. Every surface that you see has a pressure. The loads you are modeling are the changing pressures on these areas. So, downstream of the closing valve, immediately after starting the transient, I would guess higher pressures remain on the downstream surfaces. So I would state that, initially, the forces act along the direction of flow. But they bounce.