All experts;
I encountered a strange case while chekcing piping stress for furnace unit . If a rest on support is put at some specified node , code stress check fails to be qualified for sustained load case, and if the support be removed , all code stress is qualified.

Whether from mechanical theory or practice point of view, I can not understand the case written above.
Model file is too big (over 2M) to be attched.
Anyone has the similiar experience?
Any instruction and help is highly appreciated.

Regards

Gino

This is difficult to pin down without more information.

1) Was the failed load case a Sustained case or an Alt-Sustained case?

2) What else is different in the results? For example, are there spring hangers in the model to be designed. The presence (or not) of that support could change the hanger design resulting in load redistribution throughout the system.

I have encounter this situation. Sometimes if you put the support in one place and that support takes all the load that before was distributed to two or more supports, it is possible to get failure. Maybe you have thin wall or SIF or small size and this will lead to failure. If the file is big, you can try to attach only the ._a file which is smaller.

Mr. Rechard:
Thanks for your timely reply very much. You r right ,the concerned case is Alt-Sustain case. A few of spring hangers (including constant and variable types both) are employed in the system being analyzed.
Removal of the rest on support will surely cause load redistribution throughout the system to some extent, but I think it should not result in qualification checking differntly so much. Spring hangers data change is small between two support setting.
I have emalied model file to you, you may check it for more detals.

Regards

Gino

The "Alt-Sustained" load case is evaluated based on the converged restraint configuration of the (immediately) preceding OPE load case. Check that load case to see if the offending restrain is active. If it isn't, the solution removed it. There could be other differences caused by the non-linear solution.

I have passed your model on to the Tech Support group for further evaluation.