In one of my analysis, i m trying to use Inlined Pressure balanced bellow of size 48" (1200Dia), Pressure - 5 Kscg, Temp-270 deg.C near to the blower nozzle. When i used axial stiff, Lateral stiff as 50 kg/mm & Torsion as 1.0E^5, the loads on the blower nozzle is well within the manufacturer's limit but when i used Torsion stiff as 1.0e^12 the loads(MY) on the blower nozzle are exceeding the manufaturer's limit by 6times. Is it correct to use torsion value as 1.0E^5 ? & whether Inlined pressure balanced bellows can be manufactured for this pressure & size of pipe?

I would check with the manufacturer on both of those questions.

The use of the 1.0E12 stiffness value is an analysis technique to call attention to a potential problem with the bellows. The actual torsional stiffness is nowhere near this high. However, if you use the actual value (say 1.0E5) you may not know what is an unacceptable torsional rotation.

By using the large stiffness, if there is any tendancy to torque the joint, you'll see very large loads - which you did. When you reduced the torsional stiffness the loads dropped. BUT, you likely have torsional deflections at the joint, and expansion joints don't work well in torsion.

So as Loren states above, you should contact the Expansion Joint vendor for assistance.

Richard is correct. Bellows do not like torsion. Elements get destroyed. EJMA has calculation, which can be used to calculate maximum allowable moment. However normally only the manufacturer can use this calculation method as the calculation requires details that are not normally published by the suppliers. If I remember correctly the allowable stress for torsion is 20% of the published allowable stress.

Please do not get an idea to ask bellows manufacture to design torsion restrainer. Doesn't work. Hinge and gimbal compensators can carry relatively high torsion moments because of the hinge and gimbal ring.

Standard inline pressure balanced compensator is not designed for lateral movement. If you have any then talk to the supplier.

That stiffness you have seems low. I do not have my own computer near by to check. Please note that the axial stiffness is sum of 3 spring rates and for standard inline the lateral stiffness is high to very high.

48" axial pressure balanced unit can easily have a mass of few metric tons. Make sure you get correct mass for it. Also one critical issue and often totally overlooked is the fact that rod plates bend and rods extend under pressure load. I have seen cases where the combination of these is more than the specified axial movement of the compensator! Last set of bellows I redesigned had at least 3 times the specified movement in the original design. Only experienced bellows designer with correct software can figure out the strain and deflections. For rod plates you need FEA.

In the future would you plese use SI or imperial units. These units used in your part of the world requires software or books to figure out.

Richard is correct. Bellows do not like torsion. Elements get destroyed. EJMA has calculation, which can be used to calculate maximum allowable moment. However normally only the manufacturer can use this calculation method as the calculation requires details that are not normally published by the suppliers. If I remember correctly the allowable stress for torsion is 20% of the published allowable stress.

Please do not get an idea to ask bellows manufacture to design torsion restrainer. Doesn't work. Hinge and gimbal compensators can carry relatively high torsion moments because of the hinge and gimbal ring.

Standard inline pressure balanced compensator is not designed for lateral movement. If you have any then talk to the supplier.

That stiffness you have seems low. I do not have my own computer near by to check. Please note that the axial stiffness is sum of 3 spring rates and for standard inline the lateral stiffness is high to very high.

48" axial pressure balanced unit can easily have a mass of few metric tons. Make sure you get correct mass for it. Also one critical issue and often totally overlooked is the fact that rod plates bend and rods extend under pressure load. I have seen cases where the combination of these is more than the specified axial movement of the compensator! Last set of bellows I redesigned had at least 3 times the specified movement in the original design. Only experienced bellows designer with correct software can figure out the strain and deflections. For rod plates you need FEA.

In the future would you plese use SI or imperial units. These units used in your part of the world requires software or books to figure out.

Apologies for double posting. System problem.

Thanks to everyone.