Tube expansion

[mielke]

we have a box with a tube bundle in it. Hot gas (~1000F) is flowing on the box side (shell side) and cold water is flowing through the tubes (~200F).

The hot gas is being cooled to almost 200F and there is very little temp rise in the water. We want to design for clearance between end of tubes and the box wall to compensate for tube growth. If we use the very very conservative approach and assume the hot gas heats the first couple of tubes to about 1000F and base linear expansion of that.

But what about the box. The mean metal temperature for shell and tubes says the shell would on average be about 500F and for tubes around 200F. So there would be more expansion in the shell than the tubes. So the clearance will increase with increasing temperature. Is this correct reasoning? fyi there is just ambient atmosphere around box.

Thank You

 

[mielke]

These are U-tubes. and the clearance question is between end of utube and box wall. FYI

 

[chicopee]

Only if the box and tubes coefficient of thermal expansion are the same.

 

[mielke]

chicopee please clearify...
Are you affirming that the thermal expansion design about the above situation should be based on the mean metal temperature for shell and tubes (per TEMA). And that clearance between utube ends and box wall will increase with increasing temperature.

 

[StrykerTECH]

That is a complicated problem. If you have and internal temp of the box at 1000F and an ambient air temp outside of that box, then depending on the heat exchange off that box you may have some serious thermal stress on the box surface causing warpage. It would be great if you could model up a reference drawing or model to get a rough idea of wall thickness and material. You would have to see what the heat transfer coefficient is from the box to the outer air. Then find your steady state resultant temperature. That will give you your thermal stress which you can then figure out how much it will warp. Then you can figure out its expansion. The tubes will be pretty constant but the stress in the tubes will be much greater. I am assuming turbulent flow through the tubes.
Depending on wall thickness of the tubes, those things will turn into bananas with an 800F delta between the outer and inner surfaces. A rough model would be very helpful.

StrykerTECH Engineering Staff

http://www.stryker-tech.com/

 

[racookpe1978]

You've got intermediate (startup/shutdown) factores there where the box and U-tubes are going to face extremes: High initial heat flow with 100% heat flux into the heat exchanger sides and almost no cooling effect from the cold water inside the tubes) for example, and shutdown (when there is no hot gas flow but 100% cooling water is still flowing.)

So design for the two extremes and the normal operating condition falls in the "safe" (moderate) middle.