UHX 13.5.9 and tube buckling 1

[Peregrino7]

In Fixed tubesheets calculations ASME - UHX.13.5.9 and TEMA RCB-7.24 I am having problems with tube buckling.
How can I avoid this problem? I have tried to decrease the unsupported span length but with no success, and expansion joint is not allowed for this equipment.
Best regards,


G

 

[unclesyd]

One thing we have done on long Hx's with SS tubes/CS shell is to put a short section of SS in the shell to compensate. We have this arrangement on several tubular Hx's and numerous long tubular reactors.


Is a Flued Expansion Joint allowed?

 

[ijzer]

Change the design to a floating head type.

 

[Peregrino7]

Thank you both for your answer,

The design has to be one pass in tube side, this makes difficult for a floating head type, and we have tried other possibilities with the thermal department changing TEMA types... but there are other limitations.

My problem is because I have SS tubes/CS shell, I wanted to ask to Unclesyd how did you put this short section of SS in the shell to compensate??? How does it work in order to compensate ??? Did you put it in the middle ??? I have to show with calculations that this will solve the problem... Can I do that ???
Thanks for your answer and help...!

GH

 

[unclesyd]

We put the stainless section on the end and tied it into the tubesheet. Normally the SS is on the hottest end of the HX.
I don't recall the Hx length where this becomes effective. As I recall the Hx's where it is used are fairly long though not as long as the reactors. The tubular reactors I mentioned are 24' to 26' long and have a 5' or 6' section of SS in the shell. They operate at 250°C to 280°C.

The SS shell sections are just long enough to get the tubes just out of the buckling range due to thermal expansion while operating. The caveat is that this approach doesn't normally allow for steaming if the steaming is above operating temperature.

I have never personally carried out the calculations but they are apparently not very complex as this scheme was used BC (before computers), then as a Fortran program, then Lotus, and the last I saw was with Mathcad and/or Excel.

We have numerous single pass floating head Hx’s all with SS tubes with some SS and CS shells.
We have single pass Ti tube floating head Hx’s with both SS and CS shells.

If space is a problem there is a flued expansion joint design where the ring, for lack of a better word, is concave. This is used to take the height out of the flanges by allowing inward expansion of the ring instead of bending the flanges.

 

[johnnymist2003]

Hi,

These problems arise due to the different thermal expansions of the S/S and C/S, and the thermal expansion is dependant on the mean metal temperatures of the shell and tube sides of the exchanger. Have you confirmed that the mean metal temperatures are accurate (your process guys should be able to give an accurate value). It is not sufficiently accurate to estimate these temperatures in situations like this. You may manage to reduce the compressive stresses with a combination of increasing the tube wall thickness and also reducing the unsupported tube span.Sometimes also changing the tubesheet to stainless steel also helps.Your process guys would need to check the process for the increased wall thickness. I have a comment about inserting the stainless steel shell section - at the circumferential weld there may be high stresses due to the different thermal expansions of the S/S & C/S and I'd advise that you look at this also. I know of at least one end user specification that prohibits welding dis-similar materials in this manner (ie. S/S to C/S).I sympathise with you, I've had to sort out this sort of thing many times, sometimes the only option is to go fully S/S. Good luck.
John