Dear All,

We have analysed bare(non-insulated) line to keep natural frequency above required level by using anchor u-bolt support. Every suport location having U-blot we have modelled guide with zero gap and hold down with zero gap.
Line is not having temperature hence we could use standard u-bolt support(refer attached) with zero gap.

however when client did actual testing of natural frequency of line values are much lower than C-II results.

Please clarify below doubt.
1) Will use of arrester type support(with belting material) will improve natrual frequency.

2) By the use of arrester type support will we be able to match measured natrual frequecy after installation with that of C-II results.

2) Is there other way to match actual natural frequencies(measure after installation) with C-II natrual frequency.

Avinash,
Most of the common U bolt supports are provided with some gap in lateral direction. Check your company support standard detail for Gap between pipe OD and U bolt ID. Also in attached U bolt drawing , only one nut is shown. Single nut without lock nut gets loos in vibration prone piping. It is better to use lock nut in vibration prone piping.
U bolt with vibration suppressing material such as rubber or belting material can dampen vibration.
In stress analysis we consider support structure as rigid while in actual condition support structure can be less rigid (L post , T post etc.) and because of this frequency calculated by Caesar may not match actual frequency at site.

Hi Khalid,
U-bolt used is common U-bolt with lock nut, however as you said there is some gap in lateral direction.

Because of non right support structure frequency will not match with actual at site. In our results are totally out of order(frequency is almost half of C-II calculation).

However I would like to know what is best method of supporting lines to get results closure to C-II calculations. Will strap arrester will improve frequencies by considerable margin.

I think that you have to check all supports configuration whether or not suit for vibration on your system. For example, to be checked support attachment type (u-bolt or stap u-bolt) and applying proper anchors and guides on your system. Friction between strap and pipe (what you are about to use) will help but I think it is not enough if there are not enough anchors and guides on pipe direction changes on your system.

Many of the stiffness and mass terms used in initial design are estimates and even though one might say stiffer and heavier might be conservative when evaluating static response, these assumptions provide little help in determining system natural frequencies. Things like rigid restraint stiffness, rigid element stiffness, branch flexibility usually require some fine adjustments to better-reflect the actual dynamic model.

Without more specific information, it's difficult to say.

• CAESAR's spring values for friction might be off if you have locations without u-bolts.
• Support points are substantially more flexible in reality than what CAESAR assumes and are allowing the pipe to flex more readily.
• Do we know the real world natural frequency measurement methodology to be accurate? If they're counting half-cycles instead of full-cycles, that would explain why it's off by a factor of 2.

While any clamping will perform better than a u-bolt to some degree, it will still be subject to inaccuracies presented above.