Hi, i am new member of this forum.

Is it a normal practice to take the advantage of Pump thermal expansion (low temperature service of 180 deg. Faren)in case of Stress Analysis of piping.

I request one of the memeber to give a suggestion regarding above issue.

Welcome to the forum!

Pumps are typically made of metal, as such, they expand at temperature just like the pipe they are connected to. Whether it is an advantage or not, depends on the situation.

That said, you need to be careful in how you model systems around pumps. For example, ANSI pumps are usually supported at the bottom of the case, so you would be inclined to model a rigid element from the centerline down to the base to account for it's expansion. But, we typically modeling piping as though it is supported at it's centerline, even though the support point is at the bottom of the pipe. Take those two together, and you can show a slightly greater expansion of your pump and have it "lift" the closest support, resulting in a overloaded pump nozzle. In such a case, Caesar is going to tell you you have a problem, when you really don't.



On the other hand, I've had 700°F bottoms pumps (API centerline supported) where the nozzle was 10" and the pipe coming at it was 20" with a eccentric reducer. In such a case, ignoring the growth up from the bottom of the pipe (and putting a hard support) could give a non-conservative result for the nozzle load.

There's no one cookie cutter approach. Each pump system has to be examined with a critical engineer's eye to determine how best to model so that the solution you recommend to the client is safe, effective, and reasonable.

Thankyou very much for your reply.

The actual detail for our case is as follows,

The pump type is as per API-610 - OH1,
End suction and Top discharge pump, with design temperature of 180 deg farenheit.
Distance from the Pump anchor point to the face of suction flange is 10" (horizontal distance) and vertical distance is zero inch.

Suction flange size is 8".

In such a case, I typically would model a 10" rigid element to get you from the face of flange to the center of the pump. The rigid element would be at the temperature of the line and I would put an anchor at the end of the element where the pump center is.

Also, you'll want to have a CNODE anchor between the end of the pipe and the rigid element of the pump so you can check the loads imposed on the face of the pump flange.

Thank you very much Mr. Edward for your quick response.

Dear Edward L. Klein,

As per your reply on 05.09.08, can you some how in details tell me how to model the pump flange be considering the CNODE.

Dear Paldex
Here concept of connecting node is just to get force & moment calculated data at the output restraint summery.
Say we have a straight pipe node 10 then 20(mid point) after that 30
We can model as
a) 10 to 20 then 20 to 30
b) 10 to 20 ,at 20 connecting node 21 with anchor , then 21 to 30

Above both represent the same thing. Only difference is that for option b at node 20 we will get force & moment data at restraint output summery. Option a also have same force & moment but it will not show in output result.

In your case since you are putting anchor at pump center line , Face of flange just connection so output restraint report will not so force & moment at flange interface.
If you make it with connecting node it will show the same & you can compare it with API pump allowable.

Regards
Habib