Design of vessels for vibration

[EngAddict]

Hi all,

I am looking for some information on design of vessels for vibration. I have read the paper by Freese regarding Karman vortex oscillations but I was more after mechanical vibration from directly coupled compressors or agitators in horizontal vessels.

I have tried searching online for papers and I can't find any books on the subject either. I have books on vibration theory but I was after something regarding vessel design and evaluation. Such as best practices for attachments to the pressure envelope, effects of dampening by pipes and supports, design of supports for vibration, typical connections and welds, etc. Most pressure vessel design manuals don't cover this stuff.

I know the basics like concave weld face and well dressed toes. Plus design for resonance protection in oscillating systems and vibration isolation & absorption. But most of that does not help when you are faced with designing a vessel for vibration that requires an infinite life cycle.

What makes it worse is that a lot of specifications will require 70% of the nozzle loads to be applied for vibration analysis. This does not take into account the combined weights of the vibrating system, amplitude of oscillation or acceleration and frequency.. We could manually add dampening based on our model but to justify that w/o justifying the applied loads seems very unreliable. To get measurements taken now would take too long since when we got the job it was already overtime, particularly when the time to order custom forgings is taken into account.

Using the specified values means a modification of the design to what seems like a ridiculously over sized vessel. I could take one look at it and say it is more than capable of operating over an infinite life cycle based on the current design. But we have to design to a specification. To get the specification changed could take 6 months so it looks like we will have to make a stupidly sized and priced vessel. What makes it worse is the vessels that are being replaced have been fine but were designed before this current specification.

Anyway, back to what I was originally asking, does anyone have any good papers or sources of information for design for fatigue? Also does anyone have any good analysis methods? We are currently using FEA.

Cheers

 

[bernoullies123]

PD-5500 Annex C Assessment of vessels subject to fatigue - provides some good info.
They also have a worked out example in Annex W.B Calculation of cyclic stress due to external nozzle loads

 

[EngAddict]

Thanks, yeah one of the guys has been using that but we seemed to get some bogus values for stress when we used the extrapolation method. We compared it to FEA results and the stress seemed lower than it should have been. i.e. 45 deg fillet weld vs full radius fillet.

 

[LSThill]

ASM B31.3 APPENDIX W
HIGH CYCLE FATIGUE ASSESMENT OF PIPING SYSTEMS (Draft 3/25/09)


W300 Application

This appendix addresses the fatigue evaluation of B31.3 piping subjected to cyclic loadings when the total number of significant stress cycles due to all causes exceeds 100,000. A significant stress cycle is defined as a cycle with a computed stress range greater than 41.4 Mpa (6.0 ksi). The allowable displacement stress range requirements of 302, using the computed stress range in accordance with 319, provides an acceptable method of evaluating piping systems for fatigue when the number of significant stress cycles is less than 100,000. The piping cyclic loadings may be due to thermal expansion, anchor motion, vibration, inertial loads, wave motion or other sources

 

[unclesyd]

Here is a paper on the vibration in pressure vessels by Freese. Previously post by LSThill

http://www.codeware.com/support/papers/freese.pdf

 

[LSThill]

see attach

ASM B31.3 APPENDIX WHIGH CYCLE FATIGUE ASSESMENT OF PIPING SYSTEMS (Draft 3/25/09)W300 ApplicationThis appendix addresses the fatigue evaluation of B31.3 piping subjected to cyclic loadings when the total number of significant stress cycles due to all causes exceeds 100,000. A significant stress cycle is defined as a cycle with a computed stress range greater than 41.4 Mpa (6.0 ksi). The allowable displacement stress range requirements of 302, using the computed stress range in accordance with 319, provides an acceptable method of evaluating piping systems for fatigue when the number of significant stress cycles is less than 100,000. The piping cyclic loadings may be due to thermal expansion, anchor motion, vibration, inertial loads, wave motion or other sources

 

[DSB123]

LSThill,
The Appendix W you cite is for Piping Design whereas the question is about design of Vessels NOT piping.

 

[saplanti]

I believe you are correct with your toughts and the FEA analysis.

Are you the consultant for a (reciprocating) compressor manufacturer? Normally, the compressor manufacturers have all the basic information to give you a guidance. There is pulsation problem in addition to the mechanical vibration. The following papers might be helping you for more lights:

http://www.betamachinery.com/TechnicalArticles.aspx

http://www.onepetro.org/mslib/servlet/onepetropreview?id=OTC-2164-MS&soc=OTC

http://www.waset.org/journals/waset/v55/v55-56.pdf

http://www.turbomachinerymag.com/white papers/Screw compressors.pdf

Beta Machinery web site has a lot of papers on the subject. I am sure you will find one of them will give you the right direction.

Can you clarify why do they want you apply 70% of the nozzle load and how to apply? Do they think you need to isolate the vessel from the piping and consider 70% of the mass (instead of load) only? Probably they do not want to consider the damping effect of piping on the vessel. This might be the trick. What do you think?

I would expect the shell thickness is going to increase due to the minimum natural frequency requirement and use plate support to eliminate the transfer of vibration to the foundation in the force direction and make very rigid in the other. Tune up the height and other dimensions of the supports by the calculation (FEA).

I hope I did not confuse you more and it helps.

Ibrahim Demir