We have an inline pressure balanced expansion joint in an expander inlet to achieve 0.8xNEMA SM23 load limit. Beyond the EXJ we have an anchor.

If I use, 90 to 100 percent cold pull, we can achieve low operating load with avaiable axial stiffness of the bellow. As it is an engineered item of close precision, we may not need 2/3 factor of ignorence in this calc.

Is it unwise to use 90% cold pull in this special application of very high temperature of the expander inlet ?

reg,
sam

You need to model the installed preloads properly and also the evaluate loads on the equipment in the cold/installed condition. The loads may be perfectly acceptable in the hot operating condition but if the casing is distorted in the installed case the impeller may hit the casing during heatup.

One must always consider the INSTALLED and OPERATING conditions for piping system evaluations.

I wonder how accurate the length will be. I've had ±3/16" tolerance on a 40" long inline EJ before. Per EJMA, ±1/4" tolerance is fine for such length. To lower the tolerance, we'd have to pay a lot more.

For applications on equipment with low allowable and high pressure (making the EJ stiffness higher), that much tolerance can make a difference. May not be such a big deal if the pipe fabricator in the field will do the final weld once the EJ arrives. But if project schedule doesn't allow it, you'll have another tolerance that can make it worse.

High precision cold pull is introduced in the shop itself in the expansion joint and shipping bars welded to keep it intact till fabrication of piping is complete. Then, shipping bars are cut.

Thus, for both welded and flanged expansion joints, cold cut can be easily applied with accuracy.

Getting lower stifness, if not with cold cut, to have sililarly low nozzle load at operating condition is costly!

reg,
sam

In line pressure balance expansion joints have higher stiffness than normal untied expansion joints due to number of bellows in series.
I have used 50% preset (extended) inline pressure balanced expansion joint but not on strain sensitive equipment system. Due to pre extension there will be tension load on nozzles in cold condition and compressive load in operating condition. Nozzle needs to be qualified for cyclic load accordingly.

We, too, have In line pressure balance expansion joints working more than 20years.

But, nowadays, people are acquiring false notions like cold pull is lke banned item!

When this is provided in code and as you mentioned - inline pressure balanced combined axial stiffness being high, it is prudent to make operating load less at the cost of increasing load in cold condition where the strain sensitive equipment will be hardly operating except at startup and shutdown for a small time.

Expansion joint spring rate issues are not so simple. When using the spring rates first find if the given values are working spring rates or initial elastic spring rates. Some companies publish spring rates that are based on their tests or on own developed formula. On small movement removing the force expansion joint returns to its original length. On larger movement it doesn't and you have to use force to get it back where it was. CAESAR II doesn't consider this.

In-line axial pressure balances unit is not easy to get to be in balance because of manufacturing limitations. Sometimes you get it sometimes not. On low pressure units, especially large diameter, there are friction issues...

Computer will tell you with 8 or more decimals what are the forces and moments. In reality results are thump sucks.

"Big boys" who manufacture expansion joints get the spring rate formula right. Remainder seem to have problems. Some get them right some not. Standards like EJMA have a formula to calculate axial spring rate but not bending or lateral. They do not give axial rate for pressure balanced expansion joint either. Expansion joint designer has to know how it is calculated. If software is used then the software developer has to know.

Engineers designing lines spend huge amount of time in trying to get nozzle loads within limits and worry about expansion joint spring rates. Often the requested spring rates are so low that the system becomes actually unsafe.

Try to get the design such that the spring rate is not an issue. Similar issue like guide gaps. If your results are OK on 3 mm gap but no good on 5 mm gap your design is wrong. You can have either zero gap using struts or unknown gap.

One training point: If your line design temperature is 400 C and you establish that the axial movement required is 30 mm compression and then you specify an expansion joint to be designed compression 30 mm and expansion 30 mm please let us know why that is specification is seriously faulty.