Stress Analysis

Is that enough?
Self evidently not.
If there was "excessive vibration" you have to cure that first.

When is performing stress analysis deemed necessary for a piping arrangement?
When you have had a failure, or ideally before.
Which is more expensive, repairing the failure and preventing a recurrance, or doing a stress analysis?
It would be interesting to see if the stress analysis would have predicted the failure, if so, then it answers your question.
Del

GA Del.

Anytime any critical machine prematurely breaks down, somebody else should examine the machine to identify why the failure happened. Remember this great definition of insanity: doing the exact same thing but expecting a different result is a symptom of insanity.

At first, a stress analysis is required for that piping system and to check the cause of vibration and how to overcome. And stress analysis is required for sensitive machinery equipment like Pumps, Compressors and Turbines. And stress analysis is required for piping systems with relatively higher pressure and temperatures.

When you need to carry out a stress analysis for a piping system?

See CR4 Threads: How to determine pipe need to stress & Theory of failure in Piping Stress Analysis.

1. Find the source of vibration and cure that. You didn't mention whe the vibration ocurred. If if is during shutdown, consider installing an air-dam.

2. just welding the material is not sufficient to relieve the inherent stress in the weldment, unless it is supposed to be "prestressed." It is highly unlikely that you could properly prestress the weldment. The weldment must be "Normalized" to relieve welding stresses. Consider the weldment to be a perferation in an envelope. The weld is stronger than the parent material, so the parent material will begin to fail. probably in the HAZ (Heat Affectd Zone). The metal will fatigue and fail, quite catastophically.

After normalization, and depending on the thickness and surface of the weldment, consider the appropriate NDE. MT, usually Magniflux, may not be sufficient.

A stress analysis will not necessarily give you the answer you need as most FEA programmes look at static stress. From your description of "excessive vibration" you need to do a number of things:

1: Identify the source of the vibration

Worn bearings

Rubbish mounting arrangement

Insufficient supports / tie backs

Insufficient / wrong damping in the system

Fluid pulses

Out of balance motor

- I can't be more specific as you haven't told us what it is you've been welding...other than a pipe.

2: Modal Analysis of current system

This will tell you which are the dangerous frequencies for the systems, and the ones you MUST design out to prevent damage.

You design these out by adding damping - changing the design to change the natural frequency. Adding weight, adding stiffeners, clamping are all possible solutions.

3: Fatigue Analysis

Make sure whoever is doing your stress analysis work includes calculating the fatigue life - for which they'll need to be inputting oscillating stress fields. Not all analysis includes this.

Materials can fail at 10x less than their yield strength under fatigue conditions - this is the effect you used when you repeatedly bend your credit card to break it in half.

And Finally....

It sounds to me as if at the very least you need some new pipe hangers.

Are you going to blame Squirrel Kris?
Power, Work, Dynamics and Time
Remember he is a past master at this
A cat down a well, not a mine

It's the wheelie bins not the well that worry me now menow meooow.
Del

According to her mother the lady likes cats!

Wonder what she might have done if she didn't like them?