Sorry John, I thought I stated my position earlier in this thread.

One of the qualities of secondary stresses is that a small amount of localized yielding will reduce the stress. When a pipe lifts up off of a support, the dead weight load that was on this support is re-distributed elsewhere in the system. Unlike thermal forces that reduce with some yeilding, forces due to the weight of the piping system in the hot position will remain constant after any yielding so these stresses, which in my opinion are primary stresses, will remain constant as well.

This is why I feel that removing the +Y support or clamping it down is the more correct way to evaluate sustained stresses. Still, I know this is not perfect. It doesn't account for those situations where a support reaction is 3 kips in the sustained case but only 100 lbs in the operating case. Though the pipe still touches the support, for all practical purposes this support is gone and the dead weight load has been re-distributed elsewhere. For these reasons, I not only evaluate the +Y supports that lift off but other supports that have significant differences betweeen the sustained and operating loads. A complicated system may require more than one CAESAR run with different support definitions for a satisfactory evaluation.

I'm not completely convinced that the load cases advised by Richard Havard's auditor correctly evaluates this situation either. The "T" alone is not an accurate behavior. This load case set-up may be OK if all of the Y supports were coded as double acting and only the ones that want to lift are coded as +Y.

About SUPERPIPER's comments, I don't know how you make all of your supports sit all the time, even generally. If you have hot piping with vertical offsets, you're going to get lift-off quite often. Many times that support is not required, so there is no reason to add flexibilty just so it doesn't lift off of the steel.

Also, your choice of the words 'dangling off supports' lets me think that your experience with lift-off has resulted in unstable piping installations. For flexible systems, I like to add a guide at the supports that lifts off for stability.

That's my 2¢,