Dislocation of Cold Clamped Shoe Support 1

[uaepiping]

Dear Piping Engineers,
Have a look at the video in below link:

https://youtu.be/oBdpx6xDvC0

There are many incidents of this type in our plants.
The reasons may be:
1) Loosening of bolts of clamps

a) Due to Installer mistake​

b) Due to compression of the cladding under the load of the piping​

c) Due to minor shrinkage of the PUF blocks​

d) Due to different coefficient of friction between clamp & cladding and shoe bottom & structure top​

2) Wrong size and number of bolts

What is the solution?
1) How to check (through calculations) that the number and size of bolts is adequate?
2) How to calculate proper bolt torque for sufficient group?
3) Will spring belleville spring washers be an answer for avoiding loosening of bolts?

 

[pennpiper]

Uaepiping.
Reasons #1 & #2 are not valid.
The reason for the failure shown in the video is the wrong type of support was selected.

The type of support required for this application is one that (a) does not conduct the Thermal properties of the Cryogenic Piping but is anchored in place to the pipe so it does not lose position lineally (on the pipe) during heat cycles.

Do some research here:

https://www.google.com/search?q=ima...XDJCYKHa9KCDQQsAQIJQ&biw=1024&bih=633#imgrc=_

Sometimes its possible to do all the right things and still get bad results

 

[mk3223]

Agreed with pennpiper for using worng pipe support in the first place.
IMO, it might be resulted from the cost concern of the installation material and labor. I thought it may help if the length of the clamp shoe is increased for sliding between the shoe and supporting steel. However, the engineer still needs to resolve the issue of the different friction forces between the shoe/steel and clamp/cold insulation.

 

[LittleInch]

The last supports I saw for PU foam insulated pipe had some quite vicious metal barbs which were pushed into the insulation by the clamping force - however that was for the "anchors". The normal supports used "disc spring" aka Bellville washers to provide the correct clamping force and allowance for shrinkage.

If the metal work is small enough in area then it won't act as a significant thermal path and only really penetrated the outer layer of three foam elements.

Try asking Carpenter and Paterson for a design....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[uaepiping]

@pennpiper,
"The type of support required for this application is one that (a) does not conduct the Thermal properties of the Cryogenic Piping but is anchored in place to the pipe so it does not lose position lineally (on the pipe) during heat cycles."
Anchoring a pipe in place is fine, but many standards are in practice which use only the support shown in video. And there are many incidents where the support gets displaced.
My point was that the loosening of bolts may have been due to three reasons, one - not enough torque applied at the time of installation, second - The PUF block may shrink little bit due to low temperature and this may collapse the cladding (seen in some incidents at site) and cause loosening of bolts, third - the compression of the PUF blocks may cause space at the top and hence cause loosening.
It is impossible to get the PUF block insulation with zero coefficient of linear contraction and with very high modulus of elasticity to have zero compression.
The answer to these is proper torque and use of the Belleville spring washers so that in case of shrinkage or compression enough torque should be there to avoid any slippage of clamp shoe.

 

[mk3223]

IMO, the support should be fixed to either the steel or the pipe. The support on the video shows that the problem is because both the pipe and the support can move freely.
The required torque for fighting up the PUF blocks may be based on the installation details which may be difficult to manage in the field. And, it may be a risk to damage it and create a new issue.

 

[LittleInch]

You might also look at low friction plates like Teflon or similar to reduce the friction force required but this might affect the stress analysis.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[uaepiping]

@mk3223,
Yes, the support is clamp type hence free to move (slide) on pipe and the base of the shoe is not fixed with the structure. This is the reason for dislocation.

@Littleinch,
The piping systems are existing, we are performing stress analysis. Now with PTFE, the friction coefficient between shoe and the structure will become less than the clamp and cladding.
But providing PTFE plates now is VERY difficult.

 

[lilliput1]

Google cryogen pipe insulated anchor to find prefabricated units. Locate them at points in the piping such that the in between sliding type supports will not travel because of expansion or contraction, beyond the support slide stops.

 

[uaepiping]

Dears,
My questions were:

1) How to check (through calculations) that the number and size of bolts is adequate?
2) How to calculate proper bolt torque for sufficient group?
3) Will spring belleville spring washers be an answer for avoiding loosening of bolts?

If I need to rephrase, then tell me I shall try to do just that!

 

[LittleInch]

Apart from getting a supplier to do it, then the answer is:

First work out your friction force from movement coming from the support. Then add safety factor ( 2?)

Then you can work out the compressive force and pressure required on your PUF jacket from the surface area of the support.

Make sure you're not going to crush the foam (compressive strength of the foam)
Then divide the force required by your number of bolts.

Bolt torque required for force is a well worn path and depends a lot on bolt lubrication - just do a search on bolt torque.

3) From what I've seen on other examples - yes Belleville washers, sometimes three or four sets, will compensate for the relaxation of the PUF under constant load.

Or put some barbs in the support and mechanically lock the support to the outer layer of PUF and glue the layers together.


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[uaepiping]

@Littleinch,
Thanks a lot for this great help.
Appreciated!