Custom EN 13749 methodology development for a bespoke railway bogie

While large bogie manufacturers compete on volume and rely on fully standardized designs, Ganz targets the niche custom market, such as operators in the US, India and Russia who manufacture bespoke vehicles outside European tooling norms.

For one such program, Ganz had to deliver a custom-engineered motor bogie, the Ganz 225.1-11-B4, to a US railway operator. The bogie is a welded S355NL steel structure that needed to be qualified under EN 13749, the European standard for bogie structural requirements.

The customer was not satisfied with a standards-only compliance package: they wanted a full analytical narrative behind every load case, every assumption, and every safety margin.

With the US contract signed and specifications defined, Ganz identified the analytical scope the operator demanded as beyond the in-house team's capacity. They engaged eCon to develop a fatigue evaluation method specific to the 225.1-11-B4's welded connections and to run it as the final EN 13749 verification step.

A bespoke evaluation tool was implemented in ANSYS and Python, running as the final post-processing step of the FE analysis. The tool extracts the linearised stress tensor at every shell element along every weld seam, computes the in-plane stress component normal to the weld line, evaluates the bending ratio, applies mean stress correction, shifts the SN curve to the required probability of survival, applies a weld-length correction, and returns per-element fatigue life. The code was validated at three levels: against hand-calculation of a defined reference stress state, against physical fatigue experiments on welded specimens, and against an enlarged specimen geometry to verify the weld-length correction.

Standard rail weld fatigue methods such as BS7608, Eurocode 3, EN13445 are calibrated for thick plate structures and were considered too coarse for the analytical defensibility this program required. The method takes the Volvo-Chalmers automotive weld fatigue approach as its baseline as automotive methodology is more refined because vehicle bodies must be aggressively weight-optimised. It adapts it to the shell-element weld stress evaluation eCon had established with Ganz. The σp stress component governing mode-A crack propagation is captured directly from the shell element results; membrane and bending contributions are separated; the SN curve is interpolated between pure-tension and pure-bending as a function of bending ratio. A weakest-link weld-length correction was added on top. A concept familiar from aerospace fatigue work but generally absent from rail-standard practice: a longer weld under the same load exhibits lower expected fatigue life than a short specimen, because the larger stressed volume is statistically more likely to contain a microstructural defect.

Two representative weld types - V-shaped butt welds and one-sided fillet welds in S355NL steel - were fatigue-tested on dedicated specimens in eCon's materials laboratory. The lab holds a high-frequency fatigue testing machine, capable of cycling a specimen continuously for weeks, alongside the composite specimen fixtures required for carbon-fibre testing.

The tests produced the SN-curve inputs (slope, intercept, life scatter, R-ratio and endurance limit) that feed the evaluation method. Plate-thickness corrections were then applied analytically across the eight steel thicknesses used in the bogie. A separate on-site deformation measurement was performed on the fabricated bogie under defined loads, to validate the shell FE model's stiffness behaviour against physical reality.

The result is a verification chain in which every analytical step is traceable to either an eCon-developed method or an eCon-measured input.