How to Deal With Nozzle Loads When Customer Did Not Specify Cartesian Direction 4

[m_ridzon]

I'm quoting a job for a pressure vessel. The head has 6 large nozzles on it. It will be analyzed in ANSYS FEA. The prospect gave me a table of shear loads (i.e., across the nozzle flange) and axial loads (i.e., into the nozzle flange). They specified the shear loads as SRSS (i.e., the resultant of the two shear components). In a similar way, they shared another table with bending and torsional loads. They never gave me a Cartesian coordinate system to align these with, and if I asked, my gut tells me they do not have that information. In the past when I've encountered this, I take one of these approaches:
[ul]
[li]Use deductive reasoning to determine which loads are additive with loads on the neighboring nozzle and analyze that scenario. This works for very simple vessel designs with only a couple nozzles.[/li]
[li]If the above option is not practical or possible, I will analyze all the permutations of the nozzle loading conditions; i.e., I will analyze various orientations on Nozzle A while also varying the orientations on Nozzle B.[/li]
[/ul]

Because of this convoluted design and nozzle pattern, only the second option seems possible. But because there are 6 nozzles, that seems like a ton of analysis as I vary the orientation of each nozzle load set and the neighbors. Therefore the quote for the job is going to skyrocket. I'm really trying to avoid overpricing this, but don't know how to address these nozzle loads. Has anyone run into this before and can share their experience?

 

[DriveMeNuts]

'Ask' with an explanation of "This is not possible without this information", seems to be the obvious next step.

 

[m_ridzon]

'Ask' with an explanation of "This is not possible without this information", seems to be the obvious next step.

Thanks for this feedback. Yes I agree. At this exact moment, management has me under pressure to generate the quote. So I'm slathering on the conservatism in the price to try and cover some of these unknowns.

 

[Trestala]

Nozzle loads are normally described as that (shear loads (Fc,Fl), and axial loads (Fr), same as for bending (Mc,Ml) and torsion (Mt)). Sounds like they provided loads per WRC107 convention. You should know how to translate from WRC107 to that of cartesian coordinate system.

 

[Geoff13]

When customer's have sent me nozzle loads they almost always turn out to be in their global piping coordinate system. Without asking for clarifications from whoever created the load table it's impossible to know what this is, and how they should be converted to the nozzle analysis coordinate system.

If the loads are in a consistent global coordinate system, then all the M1's are about the same axis, and the same for M2 and M3, but I don't think that's much help.

However at the bidding stage I understand it's almost impossible to ask for clarification. I would make a best guess and add a clarification "We have assumed ... for the nozzle loads. Nozzle loads must be finalized within X weeks of award. Any change from the assumptions may result in an extra." Of course your Sales group won't like including this statement but it's the reality.

 

[m_ridzon]

Nozzle loads are normally described as that (shear loads (Fc,Fl), and axial loads (Fr), same as for bending (Mc,Ml) and torsion (Mt)). Sounds like they provided loads per WRC107 convention. You should know how to translate from WRC107 to that of cartesian coordinate system.

Upon closer inspection, their request-for-quote makes reference to WRC107, but not explicitly in terms of load generation. I guess that doesn't really mean the loads did or did not come from there though. Nevertheless, I'm not familiar with WRC107 and how to translate load coordinate systems. Granted, I probably should be. But could you shine some more light on it?

However at the bidding stage I understand it's almost impossible to ask for clarification. I would make a best guess and add a clarification "We have assumed ... for the nozzle loads. Nozzle loads must be finalized within X weeks of award. Any change from the assumptions may result in an extra." Of course your Sales group won't like including this statement but it's the reality.

Time is short to generate the bid so we had to roll the dice and do something quick. I have told management how this is a significant issue and they understood. Therefore, we all agreed to slather on the conservatism in the quote to cover what we thought would be the worst case scenario. And we very clearly laid out our assumptions in the quote to cover ourselves.

 

[Geoff13]

Glad to hear your management is on-board. So often they only want to offer the cheapest price possible and not include appropriate contingency amounts.

With an FEA model it should be relatively easy to give the piping system designers nozzle flexibility coefficients. Since they know nothing about the vessel the piping engineers typically assume infinitely rigid end points for their piping.

If they are willing to re-run their analysis there can be significant reductions to the nozzle loads. This would still need to be within X weeks after award so you have time to adjust your final design to suit.

Good luck.

 

[SnTMan]

Backing up a bit:

I'm not familiar with WRC107 and how to translate load coordinate systems.

WRC 107, a publication of the Weld Research Council, is a method (not FEA) to anayze attachment loads (including nozzles) as to stresses induced into the shell component. Been around since the dinosaurs. Uses vessel coordinate system, Longitudinal, Circumferential, Radial, orthogonal axes. For WRC 107 analysis loads given in global coordinates must be translated into vessel coordinates. It contains no loads, they are specicied elsewhere. Loads are applied at the shell surface and nozzle CL intersection as three shear loads, Longitudinal, Circumferential and Radial, and three moment loads, Longitudinal, Circumferential and Torsional (about nozzle CL).
Loads specicifed as resultants are not really usable as such since they presumably cannot be resolved into the vessel coordinate ssystem.

The Weld Research Council offers newer publications that extend the analysis in various ways. I am for some reason not able to find a direct link to the WRC. Most vessel software incorporates one or more of these methods. Hope you find this useful.

Regards,

Mike


The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[CuMo]

Thanks for this feedback. Yes I agree. At this exact moment, management has me under pressure to generate the quote. So I'm slathering on the conservatism in the price to try and cover some of these unknowns.

Your management will point the finger at you if things go south and something failed because you designed it.
Get more information from your client or at least clearly highlight any assumptions made during the quote design stage and make sure everyone is onboard with these assumptions.