The other day, someone suggested a potential solution to an API-661 cooler nozzle loading issue I've been having and I'm trying to figure out if it will work the way that I hope it will. This person suggested installing some kind of flange support on the cooler nozzle itself and attaching that support to some kind of off-equipment structure. His thought was we could direct all of the nozzle load into this other structure and preserve the cooler nozzles without having to change the piping system. We're at our wits end here as this particular cooler nozzle has been giving us trouble for a while.

I want to like this solution, but I want to make sure it makes sense before I pursue it further. Does this seem like a viable solution?

Generally CAESAR results become suspicious when you have two supports put so close together - due to modelling technique more so than CAESAR itself.

Have you done any of the following:

1) Considered thermal expansion of the box header from the bottom of the box header to the nozzle elevation?

2) Considered the flexibility of the nozzle/box header connection? (NozzlePro offers this.)

3) Calculated stresses outside of API-661? (NozzlePro offers this.)

In addition, I'm not a fan of flange supports as typically stress analysts aren't prompted to do any additional work to qualify them and what they do to the bolts, and there's significant overhead to achieve this on a per-flange per-flange rating per-size per-flange support detail level.

I have followed the modeling procedure that you provided me almost exactly 3 years ago (http://forums.coade.com/ubbthreads/ubbthreads.php?ubb=showflat&Main=15864&Number=69632#Post69632), which has helped in a lot of situations. My two big problems right now are:

1) This is an interstage cooler for a compressor so vibration becomes a pretty big consideration. I want to remove supports & add loops to help thermal stresses, but the vibrations analyst doesn't like it.

2) The cooler vendor is completely unwilling to qualify nozzle loads above 1x API-661. With the edits I have made, I can get the problem loads down to 2x, but that's still not good enough.

I have not considered using NozzlePro mostly due to software issues on my computer. For reasons I don't understand, I'm not able to use the 2019 or 2018 versions of Caesar (the piping input window just doesn't open), but 2017 works ok for me. I haven't taken the time to reach out to support about it yet. All that to say, I can't get NozzlePro to work.

Show a snapshot(s) of your CII model. I have seen many ways to model Air Coolers and most are way off from reality. Better advice can be provided if you provide some more physical data.


As Michael eluded to, the flange support seems like a "paper" solution akin to grabbing at straws.

Sounds like it's time to get Support/IT to fix what's wrong with your machine.

I'll offer that CAESAR II also provides issues for me - I find that the farther the CAESAR file is from my computer (i.e. stored on a network across the world), the more likely it'll hang and not fully open the piping editor prepip.exe.

But if I kill prepip.exe it usually opens the 2nd or 3rd time.

I can't offer guidance on how to make NozzlePro work for you, but is the direction you should go, since you're hitting limitations of standard modeling practice.