Welding process testing and modelling

Objective

The objective of this pilot chain is to demonstrate how experimental welding-related processes can be linked with modelling and data analysis in a structured RI workflow. The chain connects physical testing, simulation, interpretation of results and preparation of reusable data outputs.

The pilot is intended to support research groups and industrial users that need evidence-based assessment of welding processes, heat-affected zones, deformation, residual stresses, defects or material performance after welding.

Workflow description

The workflow begins with the definition of a welding-related research or engineering task. This may include testing of a welded joint, analysis of heat distribution, assessment of deformation or modelling of thermomechanical behaviour.

Experimental data are collected through relevant testing or characterization procedures. These data can be used to define or validate numerical models, including finite element models, thermal simulations or simplified process models. Where appropriate, AI-assisted analysis can support defect detection, pattern recognition or interpretation of complex datasets.

The final stage includes comparison between experimental and modelled results, documentation of assumptions and parameters, preparation of output datasets, and description of the service pathway for future users.

Expected outputs

The expected outputs include a documented experimental-modelling workflow, structured testing data, simulation inputs and outputs, validation notes and a reusable service description.

The pilot chain should produce:

- a documented welding process scenario;

- testing or characterization data linked to the modelling task;

- simulation input files, model assumptions and output results;

- comparison between experimental and computational results;

- metadata describing samples, methods, parameters and outputs;

- a reusable workflow description for the RI service catalogue;

- recommendations for further validation or scaling.

The expected outputs include welding-related testing data, thermomechanical simulation results, modelling assumptions, validation notes, processed datasets, metadata records, README documentation, licence and citation information, and a repository-ready FAIR data package.

Inputs expected from partners

Partners are expected to provide a clearly formulated welding-related task, including material type, welding process, sample description and the specific problem to be addressed. The problem may concern weld quality, heat distribution, deformation, residual stress, defect formation or performance of welded joints.

The minimum input package should include:

- material and sample description;

- welding process or experimental setup description;

- available testing or characterization data;

- target parameters or properties to be assessed;

- expected modelling or simulation task;

- information on standards, constraints or industrial relevance;

- contact person responsible for interpretation of results.

For FAIR preparation, partners should also provide information on data ownership, publication restrictions, preferred licence and possible anonymisation requirements for industrial data.

Readiness notes and next actions

The current readiness level is suitable for a pilot demonstrator, but the chain requires careful harmonisation between experimental data, modelling assumptions and service documentation.

The main missing elements are standard templates for describing welding experiments, model parameters, validation criteria and reusable data packages. Additional work is also needed to define which data can be openly published and which data should remain restricted.

Next actions include selecting one representative welding scenario, aligning experimental and modelling metadata, documenting model assumptions and testing conditions, preparing reusable README and FAIR checklist templates, and testing repository deposit through DataverseUA.