jacketed vessel design - differential temperature stress

[KevinNZ]

Hi, First time designing a pressure vessel with jacket.

It's only a small vessel made form DN100 pipe with a an external cooling jacket. (Figure 9-2 type 1)

Following ASME VIII.1 appendix 9, refers to "..particular attention to the effects of local internal and external loads and expansion differentials at design temperatures.." And refers to UG-22, this point the trail goes cold.

Looking at the formulars in Appendix 9 I do not see a temperature input.

What are the design methods for calculating the required strength of the jacket to vessel connections (closures) for differential temperature between the jacket and shell?

 

[XL83NL]

Interesting and good question.

Jumping into this thread to see what advice you’ll be given. I’ve seen similar situations where no provisions at all were supplied by the manufacturer (such as max delta T for jacket vs shell).

One simple I could think is to look for the weakest weld, determine the margin it still has (say the MAWP minus the design pressure) and relate that to an amount of extra stress in a weld (perhaps due to shear or longitudinal stress) that is a ratio to some sort of delta T.

Huub
- You never get what you expect, you only get what you inspect.

 

[r6155]

More information: material, design temperature & pressure in both sides.

Regards

 

[XL83NL]

Don’t see how that in particular would be necessary info. It seems like the OP is more looking into general thoughts for a potential method or approach to be applied.

Huub
- You never get what you expect, you only get what you inspect.

 

[LittleInch]

I would have thought you start from the basis of ambient temperature at zero differential thermal stress and the look at min inner temp vs max out shell temp (possibly when no fluid in the annulus?, so can the outside jacket see solar temp?) and ma inner shell vs max outer temp.

Then do them also with max design pressure.

The trick will be deciding what the most onerous temperatures are going to be, with and without fluid in the annulus.

100 DN pipe is pretty robust, but it will clearly be the fillet welds on the jacket seal to the main run pipe which are the crucial ones.

There is some guidance I think in ASME B 31.3 for jacketed pipes.

This is a pretty small Pressure Vessel....

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

 

[Snickster]

I believe you can approach the design the same as you would design jacketed piping. ASME PV Code don't give any method in Appendix 9 so it appears they leave it up to the engineer to determine the appropriate method considering all of the possible loads in UG-22. Basically you are looking for the local loads and resulting stresses at points of support of the jacket to the main pipe/vessel. For pressure vessel the local stresses are found by doing an analysis per WRC methods such as those found in WRC 107 once the loads are determined. Also you need to check the stresses in the connecting welds.

I belive in a piping sress analysis program such as Caesar II you can model the inner and outer piping as concentric pipes and attach them together with true anchors or flexible anchors. The program will then calculate forces and moments at each anchor point due to the differential expansion. You can then plug into a WRC type of simplified analysis or use a more accurate FEA analysis method.

I had to do a similar calculation once but it was for a 10 foot diameter refractory lined cat cracker overhead gas outlet pipe. In Caesar II it was possible to model the inner refractory lining and outer pipe as concentric pipes attached together with anchors along the length. This was the actual case since the refractory was attached with rods welded to the pipe at short distances between along the entire lentgth of the pipe.

Most stress analysis programs have methods and directions to analyze jacketed piping such as Caesar II or CAEPIPIE. Do a webserch for "stress analsis jacketed piping" and you will find a lot of links to references. You may even find a manual method.

 

[Snickster]

https://www.eng-tips.com/viewthread.cfm?qid=298572

 

[KevinNZ]

Following up, We have found

1. Piping modeling will tell you the high axial loads due to the difference in temperate of the pipe/vessel and the jacket. But does not tell you anything about the stress distribution in the closures.
2. modelled the vessel in Solidworks using thermal loading. This included the shells, clousure plates and the welds.
3. you need thick closure plates to take out the bending stress and transfer this to shear loads in the welds.
4. We also modelled a flat head in NozzlePro, but not as convincing as using Soildworks.

 

[XL83NL]

Interesting.

As per item 3), which thickness did you end up with before and after that modelling? How much (perceptual) increase did you find or require?

Huub
- You never get what you expect, you only get what you inspect.

 

[KevinNZ]

The required thickness of the closures is going to depend on more than just the temperature difference. Pressure and closure height all make a difference. In our case the jacket pressure is low and temperature difference high. Closure thickness ~ 5x required for pressure.

 

[r6155]

@ KevinNZ
Any sketch please.

Regards