Engineering guide

WRC 537 vs FEA: Choosing a Method for Nozzle Load Assessment

Choosing between WRC 537 vs FEA for nozzle loads comes down to whether the nozzle geometry fits the assumptions behind the WRC curves, and how much margin the result leaves. This guide explains what each method does, where WRC 537 stops being valid, how results are judged against ASME Section VIII and the common mistakes that lead to wrong answers in either method.

Last updated:

Why nozzle loads need checking

Pressure design under ASME Section VIII, PD 5500 or EN 13445 sizes the shell and checks nozzle reinforcement for internal pressure. It does not, by itself, account for the forces and moments that connected piping applies to the nozzle. Those loads create local membrane and bending stresses in the shell around the nozzle, which add to the pressure stresses.

Piping loads come from thermal expansion, weight, pressure thrust and occasional loads such as wind, seismic or relief-valve reaction. They are usually supplied by the pipe stress engineer, or taken from an allowable nozzle load table in the project specification and checked against the actual pipe stress results later.

What WRC 537 is and how it works

WRC Bulletin 537 is a re-issue of the long-established WRC 107 method, presenting the same dimensionless curves for local stresses in cylindrical and spherical shells as equations that are easier to program. The curves were derived from shell theory for a rigid attachment on a shell, so the method calculates local membrane and bending stresses from a radial load, two moments and torsion at a limited number of points around the attachment.

Because it is quick and widely available in vessel design software, WRC 537 is the default for routine nozzle checks. WRC 297 is an alternative for nozzles in cylindrical shells. It treats the nozzle as a cylinder intersecting a cylinder rather than a rigid attachment, and gives stresses in the nozzle neck as well as the shell.

Where WRC 537 stops being valid

The WRC curves are only valid within the geometric parameter ranges they were derived for. Outside those ranges the software may still return a number, but the number is an extrapolation and should not be relied on.

  • Nozzle diameter large relative to the shell diameter, beyond the curve limits
  • Very thin or very thick shells relative to their diameter
  • Non-radial nozzles: hillside, tangential or angled entries
  • Nozzles on heads other than spherical sections, such as torispherical knuckles or conical sections
  • Nozzles close to other nozzles, supports, skirts, cone junctions or other discontinuities
  • Reinforcing pads, which can only be represented approximately by checking at the pad edge and at the nozzle with combined thickness
  • Cases where stresses inside the nozzle neck, not the shell, govern

Inputs for either method

Both methods need the same core data, and most disagreements between a WRC result and an FEA result trace back to different inputs rather than different methods. Agree the inputs in writing before running either.

Load cases should be kept separate. Sustained loads (weight and pressure thrust) are primary and are judged against primary stress limits, while thermal expansion loads are secondary and are judged against the wider primary-plus-secondary range. Occasional loads such as wind and seismic are combined with sustained loads and may use increased allowables where the code permits.

  • Shell and nozzle outside diameters, thicknesses and corrosion allowances
  • Reinforcing pad size and thickness, if fitted
  • Distance to the nearest discontinuities: other nozzles, heads, cones, supports
  • Design pressure and temperature, and material allowable stresses at temperature
  • Piping forces and moments for each load case, with the axis convention stated
  • Any project-specific acceptance criteria or allowable load tables

What FEA adds

Finite-element analysis models the actual geometry: the shell, nozzle neck, pad, welds and nearby features, with the pressure and piping loads applied together. It removes the geometric restrictions of WRC 537 and shows where peak stresses actually occur, which is often not at one of the four standard WRC points.

For ASME Section VIII, the evaluation normally follows the design-by-analysis rules of Division 2, Part 5, which is commonly applied to Division 1 vessels for loads that the Division 1 rules do not address. In an elastic analysis, stresses are linearised through the wall and classified as primary membrane, local membrane, primary plus secondary and peak, then compared with the corresponding allowable limits. Elastic-plastic or limit-load methods are alternatives where elastic results are too conservative. Our guide on when FEA is required for an ASME VIII vessel explains this route in more detail.

FEA also gives the nozzle's local flexibility, which can be passed back to the pipe stress engineer. Modelling the nozzle as flexible rather than rigid sometimes reduces the calculated piping loads enough to resolve an overstress without hardware changes.

Side-by-side comparison

Neither method is better in every case. The table summarises how they compare on the factors that usually decide the choice.

Comparison of nozzle load assessment methods
FactorWRC 537FEA
Geometry rangeRadial nozzles on cylinders and spheres within curve limitsAny geometry that can be modelled
Nearby discontinuitiesNot accounted forIncluded directly
Reinforcing padsApproximate, two separate checksModelled with the pad and welds
Stress locationsFixed points around the attachmentFull stress field, peaks located
Effort and review timeLowHigher; needs mesh and model checks
Typical useRoutine nozzle checks with clear marginLarge, non-radial or highly loaded nozzles and marginal results

Common mistakes

Load direction errors are the most frequent. A short table that maps each piping load component to the WRC or FEA axis system, checked by a second engineer, removes most of them.

  • Sign conventions: mixing the piping software's global axes with the vessel nozzle's local axes, so moments are applied in the wrong direction
  • Running WRC 537 outside its parameter limits without noticing the warning
  • Combining sustained and thermal loads into one case and comparing with a single allowable
  • Using nominal instead of corroded thickness
  • In FEA, judging peak stresses at a sharp re-entrant corner against membrane limits instead of linearising
  • In FEA, using a mesh too coarse through the thickness at the junction, or restraints that stiffen the shell artificially

Choosing a method and when to involve an engineer

A practical approach is to run WRC 537 first where the geometry is within its range. If utilisation is comfortably low, the result stands. If the geometry is outside the range, the result is close to the limit, or the nozzle is near another discontinuity, move to FEA. Where loads exceed allowables, the options are to reduce loads through piping changes, add a pad or thicker insert, or justify the actual stress with a more detailed analysis.

SLETEC carries out both WRC-based checks and nozzle FEA, including a heat-exchanger nozzle-load assessment to WRC 537 in our project references. Our pressure vessel local stress case study shows a typical FEA route, and our FEA analysis service covers the wider scope.

Frequently asked questions

Is WRC 537 the same as WRC 107?

WRC 537 presents the same method and curves as WRC 107, rewritten as equations that are easier to implement in software. Results should be essentially the same for the same inputs. Many specifications now refer to WRC 537 rather than WRC 107.

When is FEA required instead of WRC 537?

FEA is needed when the nozzle geometry is outside the WRC parameter limits, the nozzle is non-radial or close to another discontinuity, or the WRC result is marginal. Some project specifications also require FEA above a certain nozzle size or load level.

Which allowable stresses are used for nozzle load checks?

Results are usually classified and compared with the design-by-analysis limits of ASME Section VIII Division 2, Part 5, for local membrane, primary plus secondary and, where needed, peak stress. The project specification may set additional criteria.

What is WRC 297 used for?

WRC 297 treats the nozzle as a flexible cylinder intersecting a cylindrical shell, rather than a rigid attachment. It gives stresses in the nozzle neck as well as the shell. Like WRC 537, it has its own parameter limits that must be checked.

Discuss your project with an engineer

Send drawings, datasheets or a short scope – an engineer replies with scope, schedule and price, usually within 24 hours.