I want to install an EJ that was designed for 6.3" compression in an application that requires 4" extension. If I precompress the EJ by 4" for installation so it can expand back to its free length in service, how do I specify the spring rate? At free length the spring rate is 235 lb/in and with 4" of precompression the force it applies to the pipe is 940 lb but it goes to zero after 4" of extension in service back to its free length. The pressure force would still be present. I tried to enter a negative number for the spring rate but CAESAR II won't accept that. Thanks for any help.

You can't have a "negative stiffness".

What you want to do is specify the stiffness, AND apply equal+opposite axial forces to compress the bellows the desired amount.

Thanks for the quick reply.
I can see that the equal and opposite forces would give me the initial compression but when the joint is extended in service and allowed to approach its free length the forces would decrease to near zero. Is that inconsequential since they balance each other?

Actually the force(s) will be overcome by whatever causes the joint to extend.

Here's another way of doing it. It may not be worth the effort but it's a good mental exercise...
Model the XJ without those extra (compression) forces but add a CUT LONG to your model.
Select a pipe on one side of the XJ or the other, break that pipe into two elements where one element (the one closer to the XJ) is now the length of the XJ pre-compression. Reset the material of this element to CUT LONG (material index #19). Be sure the trailing element is set back to your proper material. Now add CS to every load case name. Check the results.
That "cold spring" will double the length of that cut element. You could even run CS as a separate analysis to see the compression separately.
Make sure your XJ vendor is aware of this pre-compression. It affects the overall fatigue life of the joint.

Very good instrction!Thanks.