Thermal fatigue in Inconel nozzles: a three-dimensional forensic analysis of a catastrophic failure

Last month, an experimental rocket engine failed violently during a static bench test. The nozzle, manufactured from Inconel 718 using powder bed fusion (PBF-LB), fractured into multiple fragments. The engineering team initiated a digital forensic analysis using GOM Inspect to scan the debris and reconstruct the original geometry. The objective was to determine whether the failure was caused by cyclic thermal fatigue or by a fusion defect in the internal layers of the component.

Volumetric reconstruction and thermal CFD simulation in ANSYS Fluent 🔥

The pipeline began with high-resolution optical scanning of the nozzle fragments. GOM Inspect allowed the pieces to be aligned and a surface mesh to be generated, which was exported to nTopology. There, a volumetric reconstruction was performed to fill in the missing areas and create a complete solid model. This model was imported into ANSYS Fluent for a coupled thermal CFD simulation. Boundary conditions representative of the test were applied: gas flow at 3200 K and chamber pressure of 70 bar. The results showed an extreme thermal gradient at the nozzle throat, with surface temperatures exceeding 1400 K. Stress analysis revealed that the highest stress zones coincided with the fracture lines observed in the debris.

Lessons on lack of fusion and fatigue in printed layers ⚙️

The microscopic study of the fracture surfaces, correlated with the stress maps from ANSYS, identified regions with excessive porosity. These areas exhibited a lack of fusion between adjacent layers, acting as stress concentrators. Cyclic thermal fatigue during the seconds prior to failure propagated cracks from these defects to catastrophic rupture. This case demonstrates that, in high-performance applications, material fatigue simulation must integrate real 3D scan data to capture manufacturing defects that ideal models omit.

What specific multiphysics simulation techniques allowed correlating the cracking patterns observed in the 3D forensic analysis of the Inconel nozzle with the thermal fatigue cycles during the catastrophic failure of the rocket engine?

(PS: Material fatigue is like yours after 10 hours of simulation.)