Bracket support movement

[NRP99]

I am checking vessel to vessel support bracket integrity using FEA with vessels having high operating temperature with thick insulation which is also covering bracket. I am getting very high stresses as I constrained the bolt holes of fixed bracket (horizontal both directions)which is not actual scenario. These stresses are due to thermal expansion getting restricted. Practically there will be relative movement in horizontal direction of the vessel with respect to support on which vessel rest, even if the bracket is fixed type. Sliding brackets anyway will allow the directional movement. Anyone here have an advice or practical aspects/input on this problem? Any reference which suggest that fixed brackets also move somewhat and How to calculate this somewhat?

This Problem may be similar to saddle where fixed saddle will also move somewhat. This somewhat movement will allow the vessel thermal expansion to ease. And high stress levels at constraint point as shown by FEA seems to be fictitious. Please suggest any practical and/or design advice for such scenarios. How general practice is followed?

 

[doct9960]

For direct-coupled vessels, the general practice is to have one vessel fixed and the other one sliding. The sliding vessel would have slotted holes in its support. We've done this on vessels that are supported on lugs or skirt. Like you said, the problem is similar to a vessel with saddles. One saddle is fixed and the other saddle (with slotted holes) is sliding. The foundation or structure on which the vessels are set on are considered fixed or rigid.

If you provide a sketch and more information, you will probably get a lot of response in this forum. Are the vessel supports on steel or concrete? Are they provided with slide plates?

 

[NRP99]

I have attached rough sketch of arrangement. The vessel is supported by four brackets as shown. The brackets are bolted to steel stool. The steel stool rests on and bolted to concrete pedastal. The main concern is bracket to shell/rf pad junction. By constraining bolt holes, there will be surely high stresses on bracket bolt hole as well as bracket to shell junction.

Just wanted to know whether there will be movement of fixed support to the extent of play provided between bolt and bolt holes in horizontal direction or equal to radial thermal expansion. What is general practice that is followed while designing? And also is there practical movement possible after fixing the brackets? Is this actual scenario after installation? If it is there, how to calculate it?

I thought if there will be slight movement of fixed support that will ease the stresses in brackets which are mainly due to thermal and without thermal load bracket stresses are within limits which is obvious.

 

[NRP99]

There are no slide plates/Teflon pads provided at this stage of design.

 

[TGS4]

You may have to simulate the flexibility/stiffness of those supports to get a reasonable answer.

 

[NRP99]

I forgot to thank doct9960. Thank you for your input.
TGS4 Thank you. I have checked by modelling steel stool as beam and connecting it directly to bottom surface. The stresses in bolt holes and brackets are reduced to some extent.

How to simulate the effect of steel stool and bolting connection of brackets to stool? Any advice? I am thinking of modelling stool and connecting stool holes and bracket holes by beam elements and fixing steel stool.Any other approach?

Even if we simulate all details, we still miss out practical aspect of some movement in bracket to support-radial and longitudinal direction. I wanted to know whether fixed supports will move to some extent after vessel is in operation.

Can anyone explain physics of supports? What are the practical aspects of support behaviour which should be kept in mind while designing as well as simulating?

 

[TGS4]

How are you modeling the vessel and supports? If it were me, I would model them using 3D-Solid elements. My preference would be to use the same approach for the steel stool and other support paraphernalia.

The physics of the support connection - well that depends on how much you tighten it, it at all. I would also keep in mind your assumptions about friction. My general approach is that if friction's presence helps, then I simulate a range from no friction (0.0) to a reasonable-assumed value (in the range of 0.3 to 0.5). If friction hurts, then I simulate a range from a reasonably-assumed value to some rather high value (like 2.0). If your vessel will be installed outside in a working process plant, then I would assume that all manner of crud will get between and gum up any surface that you expect to slide. Unless you actively design against that from happening.

 

[NRP99]

TGS4 Thank you.
I have modelled full vessel along with brackets except stool as solid model. To get flexibilty of stool and actual behaviour, I am thinking of modelling stool as solid and then connecting bolts as beam and providing frictional contact between bracket bottom surface and stool top surface. I will go with frictional coefficient as you suggested. Anything missing in this approach?
As far as physics is concerned, following is my logical thinking (even though I am somewhat doubtful about that). Please correct me people if I am wrong.

The fixed supports(saddles, brackets or any bolt connected supports used in vessels as well as piping) in high temperature application should move somewhat even we tight it fully. This is due to inherent flaw of bolted connection which only able to resist fractional movement in opposite to bolt axis direction as only frictional force is resisting it. Consider two plates bolted together in which one is fixed and other is free to move over fixed one. The movement of free plate is restricted by the bolt (clamp force+ weight)*frctional coefficient that is frictional force between the plates. Now consider for unit displacement of unit length bar of unit cross section requires force equal to modulus of material. This is very high force. Now consider 1000 unit length bar heated to delta T of 100 then expansion of bar is 1.15 unit. This means the thermal movement 1 unit due to delta T of 100 produces very high magnitude force. The free plate will overcome the negligible frictional force and will move if we heat it to delta T of more than 100.

I am doubtful because of magnitude of clamp force as well as friction coefficient which will form the resistance to movement.

 

[prex]

IMHO you don't need to assume movement at bolted flange: the flexibility of stool, pedestal and foundation is for sure much higher that that of vessel wall and won't require the flanges to slide.
I wouldn't model the whole system though: the supports may be easily modeled as beams and their flexibility estimated (though estimating the flexibility of foundation may be not easy, depending on how it is done, single plinths or slab). Then you can estimate the horizontal force acting on brackets under thermal growth and add it to your model of vessel.

prex
[URL unfurl="true"]http://www.xcalcs.com[/url] : Online engineering calculations
[URL unfurl="true"]http://www.megamag.it[/url] : Magnetic brakes and launchers for fun rides
[URL unfurl="true"]http://www.levitans.com[/url] : Air bearing pads

 

[lilliput1]

Trouble with modeling is the assumptions that components mau not behave perfectly per the ideal conditions made in establishing the simulation formulas. Hear about the sinking and tilting building in San Francisco? Why not provide Teflon slide plates to prevent damaging the concrete supports. Concrete is weak in tension the edge distance to the bolt may not be adequate to resist spalling. The concrete may not have steel reinforcement properly located to resist resulting tension from the vessel expansion. Repeated expansion and contraction may cause fatique stress that would eventually crack the concrete.

 

[NRP99]

prex and lilliput1. Thank you for your inputs.

@prex- You mean to say I can input opposing frictional force to thermal growth including flexibility of structural members. Yes. I can definitely do that but need some assumptions to do that.

@lilliput1-Will consider to suggest design team about providing Teflon/slide plates for less resistance to movement.

Anyone want to correct my understanding of physics of supports? Or better please explain someone their own understanding.

 

[r6155]

NRP99, please can you tell us more information?: diamenter, TL to TL, material, shell thickness......

Regards
r6155

 

[NRP99]

@r6155
Vessel dimension in mm.
2500ID*14thk, TL to TL-7500, Bracket to bracket-4700, Material- 387 gr11 class 2
Corrosive liquid storage vessel

 

[r6155]

NRP99 thank you. I prefer saddle support, why not?

Regards
r6155

 

[NRP99]

r6155

I guess brackets are preferred due to vessel is in pit, have high temperature operation and providing saddle will have cost implications such as total overhaul of pit for good foundation. This may be the reason to prefer inexpensive Brackets/lugs. Why do you prefer saddles? Any specific reason?

 

[r6155]

2 saddles supporte is the best design due to good distrubution of weight. With 4 support is difficult to guarantee good distribution of weight.

Regards
r6155

 

[NRP99]

r6155
You are somewhat right. But what about high temperature operation? The restraints provided by saddle is higher and ultimately stresses will be higher than bracketed vessel.

What do you think about the physics of support?

 

[r6155]

@NRP99
Design calculations shall tell you if stress is high and allowed or not.

Regards
r6155