Welding Simulation – How Much Validation is Good Enough?

Welding Simulation – How Much Validation is Good Enough?

Finite Element (FE) model analyses are employed to simulate welding processes by industry and academia for a wide range of applications. In many cases, the focus is primarily on the prediction of Residual Stress (RS) distributions in high integrity welds, for use in ‘fitness for service’ defect tolerance assessments. The use of validated best-estimate RS profiles can significantly reduce conservatism, compared with upper-bound RS profiles derived for common weld geometries in industrial codes. Such assessments are used to support safety justifications for plant life operation, or to plan maintenance/repair, thereby reducing costs associated with outages and inspections.

As confidence in FE weld modelling tools has increased, they are also being utilised in support of design and manufacturing applications. Examples include the optimisation of welding process parameters, joint geometry, fit-up and restraint. Subsequently, the effects of post weld machining operations and Post Weld Heat Treatment (PWHT) can be investigated. FE model predictions of temperatures, distortions and RS can inform key decisions, thereby reducing costs and timescales associated with fabricating and testing specimens and prototypic components.

However, welding processes and material behaviour are complex. Accurate FE models require physical and mechanical properties from ambient temperature to near melting point. The development of realistic material models requires extensive testing and calibration. Examples include the characterisation of cyclic strain hardening and microstructural phase transformations.

Software developers have responded to industry demand by incorporating user-subroutines into FE software packages to facilitate the application of sophisticated material models. Software enhancements are also being made to improve computational efficiency for multi-pass welds, mainly in support of Metal Additive Manufacturing (MAM).

Current ‘state-of-art’ academic research is pushing back the boundaries further by developing multi-physics tools to represent fundamental behaviour, such as the molten pool, weld metal solidification and grain growth.

These issues beg the question of ‘how much validation is good enough’, noting the driver to reduce costs where feasible in competitive markets? The answer probably depends on the physical phenomena and length-scales of most interest, as well as the significance of analysis predictions.  

  • Is a lower level of validation acceptable when investigating the viability of a low Technology Readiness Level (TRL) process, for example by performing model parametric studies, compared to higher TRL supporting product release for manufacture?
  • Where can simplified modelling methods be used with acceptable accuracy, such 2D analyses with bead lumping, advantage of geometric symmetry?
  • Is it acceptable to perform measurements on simplified test-pieces, such as plate welds (e.g. to calibrate heat input and material models), before scaling up to real structure, noting it might not be feasible to apply the best measurement techniques on actual components?
  • Alternatively, can measurements on one representative geometry, provide sufficient model validation for wider use, where structural boundary conditions are modified (e.g. pipe weld)?
  • What levels of uncertainty should be accounted for in measurements and FE predictions?
  • What is the demand for NAFEMS best practice guidance documents, benchmarks, and tutorials to aid FE analysts?

Who should attend?

This seminar is recommended for structural integrity engineers/analysts, designers, manufacturing engineers, researchers and software developers to share their experiences and ideas.


An agenda will be available shortly.


Event Type Seminar
Member Price £180.00 | $223.46 | €199.25
Non-member Price £260.00 | $322.77 | €287.81
Credit Price Free when using 2 Member Credits


Start Date End Date Location
13 May 202013 May 2020Coventry, UK


To be confirmed in Coventry


We would like to extend an invitation to your company to be part of this exciting event. There are several outstanding opportunities available for your company to sponsor or exhibit at the seminar, giving you maximum exposure to a highly targeted audience of delegates, who are all directly involved in simulation, analysis, and design.

Please contact Jo Davenport - jo.davenport@nafems.org to discuss the options available.