I'm new to dynamic modeling in Caesar II, so please bear with me.

I'm analyzing a gas hydrotreater on which the heater feed piping is vibrating. (A little background: Hydrogen gas is dumped into the gas oil prior to a series of interchangers. After the interchangers, the 2-phase mixture is fed into the heater.)

Our client wanted not only a static analysis to resolve issues with the pipe supports, but also a 2-phase flow analysis and a dynamic analysis.

Our process guys indicate that the predicted 2-phase flow regime is not the cause of the vibration. So, I am having to do a dynamic analysis in Caesar II to try to find out what's causing it.

My question is, if all I have are flow velocities, pressure drop, and pipe and fluid properties, is this enough information to complete a dynamic analysis in Caesar II?

I should note that we do not have slug flow (it's annular-dispersed), and no occasional shock loads (i.e., no quick-closing valves, no wind loads, no vibrating equipment, etc). I'm also curious to know just what Caesar's capabilities are, and will my static analysis (placement/type of supports) be affected by the outcome of the dynamic analysis?

Thanks in advance for your help.

No this isn't enough. CAESAR II won't do anything with this information - it is not a hydraulics package. You have to define the dynamic load to CAESAR II. Once you can describe the load, you tell CAESAR II whether the analysis needs to be harmonic, response spectrum, or time history. Then based on this selection, you define the appropriate data to define the loading.

Yes the statics analysis could be affected, if you have to change the support configuration as a result of the dynamica analysis.

Though I am not expert in dynamic analysis, I will try to reply as per my understanding.

Though sometimes, piping does not fall into the slug regime, sometimes, two-phase flow can be a cause of vibration.

The dynamic analysis can highly affect your static support configuration as dynamic analysis is always associated with the minimum natural frequency requirement.

If you think slug forces are considerable, you can the generate the of Time Vs Force spectra by process parameters like Fluid desity, velocity, Liquid and vapor volume fraction,and other piping data.

You can feed the spectra into the caesar. Caesar II will perform the modal analysis first. At that time you can check the frequency of system. If you find that frequecy is very low, that can be probabale reason of vibration even though slug forces are not considerable. Remember slug force spectra will just generate a DLF pattern which will add some shock loads to the static system, but does not change the frequency of the system.

OR

when you calculate the forces and find that values are very low, simply go for modal analysis and try to elevate the system natural frequency (atleast 4-5 Hz) by adding or modifying the supports. (This is what is done in case of Mechanical Study of Reciprocating Compressor to avoid any chances of resonance/vibrations- but in that case you a have reference of minimum frequecy requirement with respect to the particular harmonics of reciprocating compressor)

I think there are many more knowledgable experts on the forum who can guide you better on this topic and can correct me if I am wrong somewhere.

Best Regards,

Come on... lets think about this..... So if your Chem E's are right the line will not vibrate with no flow in the line???!!!!! :rolleyes:

You may have flow induced vibration but the source is undeniably and undoubtedly the mass flow!

CAESAR II is a structural analysis program for the structural analysis of the piping it does not provide any fluid mechanic solutions.

I suggest to get the modal analysis done by simulating the piping in Caesar-II with all structural boundary conditions like heater end terminal, anchor/restraints in line etc for the vibrating line; get all the first few natural frequencies, corresponding mode shapes & try to identify the vibration frequency & mode shape from signature study, if possible. By observation also, the vibrating modeshape can be identified. Then, modify the restraints to raise/alter the resonating natural frequency.

Similar cases occur in power plant feed water heater drain piping downstream control valve. There, the pipings are designed with very low mixture velocity i.e. large size & made very stiff - mostly control valves sit very near to the condenser-end nozzle, anchored.

Luf Sir is right. No doubt, this vibration is flow induced. But, may be - a very flexible piping of small size is the reason.

regards,

sam

John,

I only meant to say that we were aware that the vibration was not caused by *slug* flow. With this being South Louisiana, we don't have too many other means of inducing vibration in piping other than mass flow or rotating equipment--it's not exactly a region for earthquakes or constant high winds....

Anyway, I neglected to mention before that our vibration is intermittent and this piping is 10" S/100. Not small, not flexible. One theory is that the dispersed gas bubbles are finding a wide spot in the line and forming a vapor pocket, resulting in slug flow when the fluid just ahead of that vapor pocket slams into a 90° bend just downstream of the HEXs. I don't know how much credit I give this approach--it seems like a stretch.

On the other hand, we may be facing a system that possesses a very low natural frequency--I am hoping to verify this with the modal analysis. I have calculated the forces and now I just have to give myself a crash course in building the dynamic load cases. Therein lies the problem. I'm still learning this stuff!

You guys surely must remember what it was like being a young engineer....right?? Fumble through things with minimal guidance all while keeping one eye on the clock...!

Thanks guys. Your advice is truly appreciated.

Well you can whack away at this from a couple of view points....

1)Based upon observed behaviour.. displacements and frequency you can "guess" at a forcing function... and then confirm the CAESAR II output versus the observed behaviour... being new to the field I would not recomend you trying this approach.

2)Run a natural frequency check try to get the modes dispersed and the lowest frquency above ~ 8-10 hz. This will require additional restraints but will not give you a design load for said restraints see 1 above.

The bottom line... go find an old guy to help you with this drippy messy bag of oozing !#@!@#!~#

Hi Kelly,
I had very similar experience with a horizontal re-boiler and its 36" return line. Initially,
process guys were saying that the flow was annular, but when I proved to them that slugs can be the only cause of the behavior as observed, they found out that during startup the flow goes through "transition regime" in which slugs are created. They also found out that 90 deg. bends could be causing local flow disturbances.
Most importantly, you have to go to site, measure movements, and their frequencies at different locations (elbows). If vibrations are high enough and frequency low enough, you will not need fancy instruments for this exercise.
Based on site observations and the worst case velocities of gas / liquid phases from process, I worked out a possible dynamic force. Important note is that apart from force, the "rise time" has big influence on your resulting movements. So you will have to make quite a few assumptions. Anyway I used Time History analysis and did a quite a few runs and tries until I got a simulation very similar to observed vibrations on site.
For Time History analysis, check step by step procedure as explained in one of the news letters (do a search) and I think in one of the manuals as well.
Good luck!