I've been working on a time-history analysis for a forensic engineering project on a 8" pumped condensate line in a tunnel that failed at an anchor at a 1.5D elbow (ripped/busted anchor bolts from concrete).

I've been working with CII for 2 years and have taken the dynamics training. I started this analysis based off of a tutorial / training example from said class. I'm tapped out of resources and am writing to get an experienced reality check on my approach and results.

I've modeled this as B31.1, A106B and 140°F. Upstream of the anchor is 200' of straight pipe with a piston type EJ at the far end. Down stream of anchor, heading perpendicular, is 100' of straight pipe with a piston type EJ.

Anchor Input: Fx, Fy, Fz are 500,000 lbs/in. Mx, My and Mz are all as is. It's a stout anchor welded with thick plates.

Slug Assumptions: 10 gallons at 15 fps
Slug Force: ≈400 lbs
Rise Time: 630 ms
Sustained Time: 141 ms
Decay Time: 630 ms
Spectrum Definition (attached screen shot)
Force Set: 400 lbs in X; 400 lbs in Z
Time History Load Case: applied spectrum; factor 1.0 in Y direction.
Static/Dynamic Comb 1: D1 + S3

Results:

I'm getting stress levels in the 280,000 psi range and anchor forces in the 550,000 lb range all in this area I'm applying the slug flow.

Am I doing something wrong and/or inaccurate that I'm not catching? Are these loads reasonable based on my input?

Any help would add a few years to my life.

Louie

This sounds like you've applied the load^2. Specify your force magnitude in the "force set" only. Use a normalized "DLF/Ordinate" value with a maximum of 1.0.

Dear Louie,
Is this only slug that is responsible for distruction?I think there is something else.As what I understand is 400lbs is not very high force with 8" elbow..Please check your inputs before recomending any thing..

Excellent, thanks for the quick feedback.

As I was writing this all out, I wondered about that doubling up the load. So keeping the force set at the actual hand calculated load and the DLF at 1.0 makes sense. I'll give that a go.

I've got a ticket open too, with the same info. I'll update once I have a chance to run it.

Thanks!

As one would expect, using 400 lbs instead of 160,000 lbs makes a tremendous difference. (I had set the 400 lbs for DLF and 400 lbs for Force Set multiply)

John in tech support suggested using a .txt file to input the impulse data (from previous post). Does using the DLF generator creates a spectrum that is different from just the normal impulse wave form? If so, is it more or less conservative?

Anyway, I tried this text file approach and keep coming up with an error. The name of the file is "15fps10gal.txt"

*** ERROR *** Spectrum name or type incorrect

What is the correct format for importing a text file into a time history analysis?

Thanks!

Just read this in the help portion, for anyone interested...

"The spectrum name (or load profile) can be preceded by a (#) sign. The (#) sign instructs CAESAR II to read the spectrum table from a file having the same name as the spectrum *with no extension*. Several jobs in the current folder can then access this shock data. "

With no extension! That problem solved.