Hello Stressers,

Does anyone have encountered this criteria of (how many percentage of liquid to vapor) as a guide to which is most like to create vibration?

If anybody encounters this criteria, what's that figure (% liquid to vapor) that most likely to cause vibration.

I think I've heard this from one of my senior before but forgot what's the figure to watch out for.

That's why the spreadsheet developed in my experience before shows this value taken from process input.

Many Thanks

Dear Borzki,
When there is a two-phase flow, as a stress engineer, we should design our piping stiff only. because, as per my knowledge, from % of vapour and liquid, one can not judge whether vibrations will occur or not.

Also, it depends on size of pipe too. So better go for stiff design to avoid any practical problem at site.

I hope it clarifies. Other members can throw light on same for more clarity.

Pipe orientation matters, too.

In my opinion, it's not a clear cut science because there will always be an amount of randomness associated with two phase flow. In addition, the research that is available will likely never be 100% comparable to real world conditions due to precisely controlled laboratory conditions.

I've PDF'd a website that does a decent job at describing types of flows. You'd need to classify your flow before you could determine your risk for vibrations.

Bubbly and misty flow are the least "two phase" of the two phase flows, and thus can be treated as liquid and gas respectively.

Stratified, if in horizontal pipe, is also not a risk for vibrations.

Slug, plug flow are most at risk for outright failure due to slug forces.

Annular flow is possibly at risk for fatigue failure, but is generally the preferred flow regime in that it results in smallest pipe and generally works regardless of orientation.

What could be one flow type in horizontal piping can be a different flow type in vertical piping.

Thanks Michael and Ankit for your opinion. You're both right. Vibration as we all know is hard to take into account in the design stage as it's hard to predict what will be the forcing frequency that a two phase system will produce. Industry practice is to use 4 Hz as a guideline to make the piping system stiffer by adding guides and stopper in the system. And as Michael says it's not clear cut science since we don't know beforehand what will be the forcing frequency, just hoping that it will not match with the natural frequency of the piping system to avoid resonance. Although there are now guidelines being developed like the EIG so that vibration problem can be solved proactively (in the design stage) rather than reactively (solving vibration as it happens in the field) but I think it's still not 100% guarantee.

Any other opinion is greatly appreciated.

Cheers!!!

Very good answer by Borzki. EI guidelines are most useful to assess the vibration severity, but those guidelines shall not provide the mitigation measure. The actual mitigation measure shall be decided based on stress analysis.

Can anyone clarify, if the piping system has supported with spring hangers, in that case 4Hz is applicable or not? Do we need to consider the natural frequency?

Thanks in advance,

If you potentially have turbulent or slug flow, natural frequency is always important.

If you can guarantee stratified flow, it doesn't matter at all.

Thank you.

Dear Michael,

The natural frequency can be increased by making the system rigid enough. But if we have spring supports, the systems have enough flexibility to movie up and down..How can we increase the natural frequency to 4 hz.

Kindly clarify.

Thanks in advance.

Hi Char,

If there is really a potential for vibration due to two phase flow in the slug regime then maybe you need a snubber (not sure if it's ok for two phase flow application). This will allow for thermal movement while restraining in the dynamic event. Actually, the 4 Hz requirement is somewhat subjective unless you have really a well defined forcing frequency (say the frequency at which the slug hits the elbow). Usually a +/-10% frequency sweep is assumed to take into account some uncertainties like the support stiffness, etc. It may happen that the forcing frequency may still coincide with the natural frequency of pipe during operation and experience resonance. But if the 4 Hz is really a requirement from client spec (I've read some client spec requires even 5Hz) then you have to find a way to solve both static and dynamic problem (in which the solution is opposite). But first, as per Michael's suggestion confirm first with the process engineer what type of two phase flow regime you have.

Any other opinion is greatly appreciated.

Cheers!!!

As a generality, natural frequency is a function of moment of inertia (i.e. diameter and thickness), modulus of elasticity, and length(s) and direction(s) of line segments able to be affected by vibration.

You may wish to verify your supports. Standard CAESAR II supports tend to offer more play than real life. E.G. a pipe shoe generally doesn't allow the pipe to pitch easily. Also, if your static case shows your guide or anchor isn't activated, CAESAR II assumes it just goes away for dynamic analysis.

If your line is insulated, check the frequency without it. CAESAR II insulation adds mass without stiffness, thereby making natural frequency worse. Mind you, natural frequency IS worse due to insulation, but it's not quite as bad as CAESAR II might indicate. Perhaps lighter insulation might help, right?

Are valves causing you problems? Move them closer to supports. Are you using the right mass?

Add more supports if possible. Add clamps. Sure, your static stresses go up, but it's better than the alternative.

Michael is correct. CAESAR II will linearized the dynamic model based on a reference (solved) static load case, which you specify in the dynamic input. You can review the "Active BC (boundary condition) report to see which restraints are active.

The CAESAR II default restraint stiffness is 1E12 lb/in, which you can change in the Configuration. If you decide to do this you must be able to justify your number.

Thank you Michael and Richard.
In my case, 36" piping from ground level to 28meters vertically going and connecting to the vessel. Vertical part have slug flow and vibration. The spring hanger supports from vessel cleats were used to manage vertical thermal expansion. But due to slug mixture, the springs are actively working and pipe is shaking. Analysis is going on. Thinking to stop the vibrations with all round guide supports (0mm gap) with keeping the hangers same. Kindly advise if any.

Hi Bab,

You can also request to take some vibration measurements and perform a modal & harmonic analysis to have a better understanding at what frequency the system is being excited (in case there is resonance)or it's undergoing a forced vibration (no resonance). This is a bit tedious task of matching the actual mode shape experienced by the system comparing it to the mode shapes in Caesar II. But once you have a good understanding of what mode shapes to kill then you will know where to put guides or stops to avoid coupling of fluid/structure interaction.

I've read some discussion forum from my previous company, saying even though you can shift the mode shape to what is observed in the field, but after modification of the piping system there maybe an instance where the system can still be excited at other frequencies. So solving vibration is really a challenging job. You can also suggest if process engineer can do something on the flow regime.

So far I've only encountered one vibration problem in my 20 years as stress engineer and actually the measured vibration is low compared to some guidelines so further stress analysis was not undertaken. We just recommended to closely monitor the vibration amplitudes and see if it's getting worse.

At the end of the day, fatigue is the main concern which is a dangerous failure mechanism especially if there are stress raisers.


Any other opinion is greatly appreciated.

Cheers!!

Hi Borzki,

Thank you for detailed advise. We are performing the vibration readings at various operating conditions and following some guideline. I have been solving many vibration issues since past 15 years, but this is really typical problem which I shared in this form. Keeping system modes away from resonance, keeping the natural frequency more with less static stresses is big challenge here.

Any other suggestions please..