pipingdesign: Impressed vibration

From: <Mandal>
Date: Fri Jan 02 2004 - 05:31:00 EST

Dear Sajit,

Could you find the mass participation for each mode? As a decent approximation (in case you do not have the mass participation which Caesar do not give for modal analysis), you may assume that the first mode (1 Hz, if I remember correctly) as your fundamental frequency, considering piping system as a uniformly distributed mass system with intermittent rigid supports (lock points or nodes). Normally, fundamental frequency should occur (for a uniformly distributed mass system) within the first six modes. Your approach is correct in comparing the natural frequency with the forcing function. But did you get data of the forcing function? However, you are still moving in the direction of arresting vibration using more rigid guides, vibration dampers and snubbers. Any luck in finding the source of vibration?

By the way, I am still unable to understand the source of support lift-off. Is the lift-off due to dynamics of piping system? My sincere feeling is that the lift-off is due to support settlement, more than anything else. And I am also not sure if support lift-off really addresses your problem of impressed vibration.

With best regards,
Pankaj Mandal.

-----Original Message-----
From: Steve McKenzie [mailto:mechproj@xtra.co.nz] Sent: 31 December 2003 17:11
To: <a href="/group/PipingDesign/post?postID=ogsOz4sPYXAv3ldma7v8VW9jdADBxg2D6dMHS_fyV5Ylp9_cCiiNpH_DgYzMkuoaNVQNryGyOTEQ3Kn0J1NZTSdvjw">PipingDesign@yahoogroups.com</a> Subject: RE: [Man.Distr]RE: [Man.Distr]RE: FW: [Man.Distr]RE:
[PipingDesign] Impressed vibration

Hi Sajit

Sorry, asleep at the wheel. Your understanding is correct. However did you take amplitude/fequency measurements when the pipe was vibrating? An on-site real world measurement will give some good calibration data for your model. The on-site vibration guys can probably help, and enable you to estimate the forcing function.
From your comment it would appear that the pipe lifts even in the absence of vibration. This means that either the original supports were only stops, or they did not function as intended. Your previous note said you had to cut off 20mm rods to fit your temporary supports. Were the original supports designed as stops or clamps? I think you will get the picture. The original plant designers should be willing to help.
5Hz is a number I have seen in petro specs, has been round for years, and one I use for informal calculation in the rare occasion that I need it. Your Caesar numbers indicate a primary frequency of 1Hz ignoring the ineffective supports. To me, the obvious answer is that the ineffective supports were not planned to be ineffective in the original design. Make them work and then redo the sums, including expansion effects. You will probably get a pleasant surprise, around 5Hz.

Your brief for an external agency with proven experience sounds like a prudent measure. Your superiors will probably be more comfortable with a brand name, and this is understandable, given current management structures.

In future I would prefer that you confine your correspondence to this forum and not direct to me. Unless you are really in the shit, when I will be more than happy to help in private.

Cheers

Steve McKenzie

-----Original Message-----
From: Sajit Viswan [mailto:sviswan@tebodin.co.om] Sent: Wednesday, December 31, 2003 4:51 PM To: <a href="/group/PipingDesign/post?postID=ogsOz4sPYXAv3ldma7v8VW9jdADBxg2D6dMHS_fyV5Ylp9_cCiiNpH_DgYzMkuoaNVQNryGyOTEQ3Kn0J1NZTSdvjw">PipingDesign@yahoogroups.com</a> Cc: Satyendra Bansal; Usiur Rahaman; Mike Mckenzie; Sudhansu Basu Subject: RE: [Man.Distr]RE: [Man.Distr]RE: FW: [Man.Distr]RE:
[PipingDesign] Impressed vibration

Please see the replies below.

-----Original Message-----
From: Steve McKenzie [mailto:mechproj@xtra.co.nz] Sent: Monday, December 29, 2003 3:15 PM
To: <a href="/group/PipingDesign/post?postID=ogsOz4sPYXAv3ldma7v8VW9jdADBxg2D6dMHS_fyV5Ylp9_cCiiNpH_DgYzMkuoaNVQNryGyOTEQ3Kn0J1NZTSdvjw">PipingDesign@yahoogroups.com</a> Subject: [Man.Distr]RE: [Man.Distr]RE: FW: [Man.Distr]RE: [PipingDesign] Impressed vibration

Hi Sajit

speaking out of school, several issues stand out:

from a visualisation perspective, natural frequencies do not mean much without the amplitude. Did Caesar give you the maximum amplitude and location for each frequency? What were they?
[Sajit Viswan]

My understanding on the natural frequency is that it is a characteristic of the system. Amplitude come in when there is a forcing frequency. I am trying to establish the natural frequency of the system for comparison to the forcing frequency. In this case the forcing frequency is suspected to be a flow induced one. I do not know any numerical methods for the computation of it. I can get the actual vibration readings (This is to be arranged by maintenance, equipment condition monitoring group) by allowing a controlled flow above the current levels to re-create the vibration that has happened.

B) Did Caesar predict static (non-vibration) liftoff of the supports in question?

[Sajit Viswan] Yes, it is. That is what I said in the earlier mail.

If not then it is clear that the model case is inaccurate. It would appear that the supports have been unfortunately located at the nodes instead of the midspan.

C) I saw only 3 jpgs that you sent Paul; 2 were of drawings (from the style I would estimate 1985 vintage - date?) and one showed your wedge supports. Have received no further visuals.

[Sajit Viswan] For the drawings, I am currently preparing a composite stress
isometric in A1 size which will be reduced and coverted to PDF. I shall then upload this for the group. I shall post the copies of the pids' as well, called by Shell as the PEFS.

D) Without doing sums, a 24" pipe restrained at 4m centres will have a natural frequency far in excess of the spectrum you indicate unless it

[Sajit Viswan] I understand from industry source that it is a general
practice in reducing flow induced resonance vibration by keeping the system natural frequency above 5 Hz. This is achieved by adding additional supports.

[Sajit Viswan] This can be resting and guide supports. But as the present
configuration of the 24" pipe is lifting off from the supports due to the static thermal expansion. I am inclined to think of snubbers, as a snubber can suppress vibration at the same time allow gradual expansion movement. Probably a clamp type support with a soft cushioning material inside may also do the work.
[Sajit Viswan] Have to see how much thermal expansion have to be
accomodated by the clamps.

........................
hold things......................Paul just sent me 2 JPGS of what looks like

a 4"
[Sajit Viswan] It is a 2" SS pipe.

pipe tapped into a big pipe with an around 3/4" valve (unsupported) joining the two. If this is the thing that broke then there should be quite a few people ashamed of themselves. A connection like that doesnt stand a chance. You must know that valves need to be supported. Bad, bad, bad. Look at the 3" line; no anchor, no nuthin that I can see. Do an analysis on the 4?" line and you may get a nasty surprise. Bad engineering.
[Sajit Viswan] It indeed is a case of bad supporting. This is being taken
care in the de-bottlenecking job. This vent line is also connected to the new 32" line.
Make the riser
smaller; build in some flexibility.

[Sajit Viswan] This facility was built between 98-99. The original design
was done on 3D model. Can make out from the drawings that are available. The drawings are 3D dumps to create the 2D deliverables. The plant was built by Snamprogetti (piping contractor) and the system that is a problem now which includes the launcher and the 24" bypass line is by Saipem (pipelines contractor). The demarcation between the pipelines and piping is were the 24" goes U/G. The 24" size was arrived at by Saipem based on the Shell DEP. Shell is the principal for PDO. The DEP gives the minimum bypass size allowed for the 48" to be 24". Therefore they chose the minimum. Whereas Snam sized their line to 32". The two join at the interface with a reducer. This 24" leg has been identified as the major pressure drop bottleneck by a study done by Tebodin for the debottleneck job which will bring another 32" line that will join the 48". Probably when this job is done this vibration problem will also disappear.

The picture has cut short my response in order to avoid hurting feelings.

No, numerical methods will not solve your problem(s) because they need a source of excitation. Perhaps some of the better programs may predict some self induced effects, but I think you will move faster by changing the natural frequency.

[Sajit Viswan] We can do the above work that I have said above. However
considering the criticality of the system, major safety risk that it involves and with the time constraint. I am not able to guarantee whether a resolution can be achieved at the end of the exercise. This is why I wanted to have an external expert agency to address the problem. The agency must have the necessary experience of having done similar work on flow induced vibrations and have to give evidence that their proposed solutions have in fact resolved the problem satisfactorily.

Scary!

Cheers

Steve McKenzie

[Non-text portions of this message have been removed]

Discussion List sponsor: The fluid flow calculations website - www.LMNOeng.com - LMNO Engineering, Research, and Software, Ltd.

To leave this list, send a blank email to: <a href="/group/PipingDesign/post?postID=HXMnnHSVtCoQ1ITggIsJnZMVBBoGg5Pahql4QCczCM1-6ljv_yiuUIiohuJqkx-yjguTi2I_ptS9dI_r5DMASs7K2tK6kopP4UPkrWvGTPo">PipingDesign-unsubscribe@yahoogroups.com</a>

List home: <a href="http://groups.yahoo.com/group/PipingDesign/">http://groups.yahoo.com/group/PipingDesign/</a> Main site: <a href="http://www.pipingdesign.com">http://www.pipingdesign.com</a>

Yahoo! Groups Links

To visit your group on the web, go to:
<a href="http://groups.yahoo.com/group/PipingDesign/">http://groups.yahoo.com/group/PipingDesign/</a>

To unsubscribe from this group, send an email to: <a href="/group/PipingDesign/post?postID=HXMnnHSVtCoQ1ITggIsJnZMVBBoGg5Pahql4QCczCM1-6ljv_yiuUIiohuJqkx-yjguTi2I_ptS9dI_r5DMASs7K2tK6kopP4UPkrWvGTPo">PipingDesign-unsubscribe@yahoogroups.com</a>

Your use of Yahoo! Groups is subject to: <a href="http://docs.yahoo.com/info/terms/">http://docs.yahoo.com/info/terms/</a>

Discussion List sponsor: The fluid flow calculations website - www.LMNOeng.com - LMNO Engineering, Research, and Software, Ltd.

List home: <a href="http://groups.yahoo.com/group/PipingDesign/">http://groups.yahoo.com/group/PipingDesign/</a> Main site: <a href="http://www.pipingdesign.com">http://www.pipingdesign.com</a>

Yahoo! Groups Links

To visit your group on the web, go to:
<a href="http://groups.yahoo.com/group/PipingDesign/">http://groups.yahoo.com/group/PipingDesign/</a>

Your use of Yahoo! Groups is subject to: <a href="http://docs.yahoo.com/info/terms/">http://docs.yahoo.com/info/terms/</a> Received on Fri Jan 02 05:31:00 2004

This archive was generated by hypermail 2.1.8 : Tue Mar 04 2008 - 11:40:31 EST