Expansion Joint Design 1

[MQM90]

I want to understand that in such component designing what are general recommendations i.e.

a) Cycles to be used in analysis - Any general IDEA
b) Flange and Flued joint design thickness limitations ?
c) Thin walled joint design thickness limitations ?
d) Which type of expansion joints are relatively better for BEM type exchanger ?

Best regards,
Osama

Osama Nusrat Ali

 

[SnTMan]

MQM90, I'm a little rusty on these things, but:

a) Cycles used must be based on a realistic evaluation of the operating conditions. If a vessel is started and stopped once a day, 1000 cycles won't get you much life.

b) Thickness of a F&F joint will be dependent on the applied stresses versus allowables. However, as the joint gets thicker, the spring rate also increases such that the joint may not prevent overloading of the tubes or tube-tubesheet joint. A good general rule of thumb is that if you don't really need an expansion joint, a F&F will work just fine.

c) See above for thickness versus applied stress versus allowables. Often client specs will not allow multi-ply joints.

d) The best joint is NONE. However, this is not always possible and so the selection will really depend on the differential expansion and to some extent design pressure. Very small amounts can often be accomodated by a F&F. Anything over about 1/8" will most likely require a bellows joint.

Some other pros and cons are that you generally have to design a F&F yourself, and then find a vendor to make the design, for which they will propose something a little different dimensionally which you then get to re-analyze. Repeat as required.

For a bellows joint, you can furnish the manufacturers with the design conditions and other desired features of the joint and they will design it. You then "only" have to review the documents, insure delivery is as promised, etc, etc, as when using any vendor.

A F&F joint is very rugged, however a bellows is very much less so and needs to be treated delicately. Bubba can take one out with his big rodeo belt buckle, no offense to Bubba intended. It can blow up at hydro with no warning, possibly requiring the services of the manufacturer to repair it on an emergency basis.

Note that my remarks are not related to a particular code of construction as you did not mention one, although my experiences are with ASME Section VIII, Div. 1. You would want to carefully read your actual code used, if any.

This is all kind of general, perhaps some of the other participants can give better information for you.

Regards,

Mike

 

[MQM90]

Dear SnTMan,

You wrote "A good general rule of thumb is that if you don't really need an expansion joint, a F&F will work just fine"

But here you meant that a good general rule of thumb is "where someone really need an expansion joint, a F&F will work just fine".

I think don't is a mistake. Yes or No ?

My applied or actual stresses calculated thur software and repeat analyis that my F&F type clears roughly at 9.52 mm with pressure upto 3 kg/cm2g. But according to your above suggestion "increasing thickness also increases spring rate which may not hold the bending stresses"

That means i have to go for Bellow design and by the Vendors.

Please let me know if i am going in right direction ?

Osama Nusrat Ali

 

[SnTMan]

MQM90, what I meant by my rule or thumb is that typically, F&F joints will only accomodate very small amounts of movement. Many times a F&F joint can be eliminated by other means such as adding tube supports, increasing the tube-ts joint strength, etc.

If the calculations show the F&F joint stresses to be acceptable, and the fixed tubesheet calculations are acceptable using the actual joint dimensions and spring rate then you are good to go, provinding your vendor can supply the joint to those dimensions.

If, on the other hand the calculations cannot be made acceptable with a F&F joint, a bellows is your next choice. Is is rare that you can increase the F&F thickness without increasing the joint diameter, the increased thickness increases the spring rate, the increased diameter increases the hydrostatic end force. These effects can quickly feed on themselves such that a F&F joint cannot be made to work, either due stresses in the joint itself, or stresses in tubes or tube-ts joint of the exchanger.

A F&F joint can also accomodate more movement by adding convolutions. I have once used a joint with two convolutions but is was a nasty thing to make, given that it had five round seams and four couplings for vent and drain.

I hope this is a little more clear to you, sorry for any misunderstandings.

Regards,

Mike

 

[MQM90]

Dear SnTMan

Ok i am getting the concept now.

Conclusion of YOUR emails are:

1. F&F JOINT is good to accomodate very small amounts of movement and if stresses of tubes and tube-ts JOINT are acceptable provided dimensions and spring rate are inputted correctly.

"This is also clear that when i only increase the thickness of annular plate, the spring rate value and diameter of Expanion joint changes which simply DOESNOT let the F&F JOINT work accurately Hah "

2. And if the F&F JOINT doesnt clear with one or more convulsions due to the applied stresses then ultimately second choice is a Bellow type Expansion joint i.e. a thin walled with convolsions, as much as required. In this case, my software automatically calculates the spring rate. So no worries when i will go for Bellow type JOINT.

Question:

What value of spring rate i have adopted in F&F JOINT calculations at present?
I have used the value of spring rate from Bellow JOINT calculations as software calculates it for me and therefore I have inputted the same spring rate value for F&F Joint calculations.

Mike, I want to confirm how to validate the spring rate values for F&F expansion JOINT i.e. is there any reference available or any table which is referred for taking values for spring rate.

Appreciate your efforts to answer my questions.

Thanks

Osama Nusrat Ali

 

[SnTMan]

MQM90, to your statement 1), I would add that the expansion joint stresses must also be acceptable, other wise it and your statement 2) appear correct.

However, the spring rate of a F&F joint will be very much higher than a bellows.

The only means I am aware of to calculate the F&F spring rate is per the Standards of the Tubular Exchanger Manufacturers (TEMA), however commercial software should incorporate this or some other method to do the calulation. If yours does not, you will have to get the TEMA handbook and calcuate it separately, and I assume, enter it into your software.

Regards,

Mike

 

[MQM90]

Thankyou SnTMan.




Osama Nusrat Ali

 

[SnTMan]

Well, no good deed goes unpunished, as it looks like I will have to do one of these soon. An old design. With flanged only heads. Per Part UHX. Yuk.

MQM90, you're welcome, I may be asking YOU questions soon

Regards,

Mike