I'm trying to understand the interpretation of loads that CII reports for jacketed lines. I have a core node 200 and a jacket node 1200 anchored together (CNode). The restraint definition is made on the input sheet for the 190-200 element as Node 200 CNode 1200 ANC. The resulting moments in the operating condition (W+T1+P1) are horrendous (six figures). Strangely, and I'm not sure this is even recommended, although no error messages get generated, if I define the CNode pair again on the jacket node, the moments halve, but are still bad. Apart from the fact that this system needs to be modified, I'm curious as to the use of CNodes (define once or twice?)

Read this thread, found using the [Search] facility.

The CNODE between 200 and 1200 defines an internal restraint. You should not define the same restraint (association) again, because now you have two restraints there. So this makes sense that the loads would drop by a factor of 2.

Jacketed systems are typically composed of two materials, so you can generate huge forces and moments in a jacketed system if you (erroneously) lock-up (or attempt to lock-up) the thermal growth. The problem isn't the CNODE, and assuming nothing else has been modeled at 200 and 1200, eliminate the restraint/CNODE here and change node 1200 to node 200. (So 1190-1200 becomes 1190-200, and 1200-1210 becomes 200-1210.) Your results shouldn't change. (Make sure you remove both restraint/CNODE associations at 200/1200.

Richard - Thanks. As a test, I simplified the file to a straight spool, anchored (to earth) at one end and supported at the other by a +Y support on the jacket. The other end has a flange - the core is anchored to the jacket here. I have one spider but I have a jacketed valve midway in the span. Everything looks OK except for huge X forces (X is the centerline of the spool) at either end of the valve which is modelled as a rigid (one 400 lb. rigid in the core and a separate 1 lb. rigid for the jacket). Each end of the rigid has a node for the core and node for the jacket anchored together with CNode. I'm trying to understand the reason for the large X forces because if I eliminate the valve all is well?

The two rigids, are they the same material at the same temperature - in other words is the coefficient of expansion the same? If not, that is the problem - you are trying to restrain two dissimilar expansion rates. That will give you huge forces every time.

That's it - the core is 316SS, the jacket A53B. I should eliminate the jacket CNodes and connect the jacket directly to the nodes defining the end points of the core rigid (valve), and eliminate the jacket rigid altogether? Thanks much for your help.

Many jacketed piping systems terminate jackets on the core pipe near valves and simply use jumpover pipes to connect to the jacketed volume on the other side of the vale. Make sure you have access to the details to your specific installation.

Using materials with such dissimilar expansion coefficients can be really tough on jacketed systems. You will need to make sure that spiders are far enough from elbows so that the differential expansion of the two pipes in the perpendicular leg can be taken up by flexure of the core pipe. And, asn you have already discovered, you need to be very careful about modeling what happens at valves and other components.

Thanks for the pointers Craig