Quantum

Terra Quantum and Empa Develop AI Model for Laser Welding Prediction

Read full article at source ↗

Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 29/09/2026, 12:04:24 UTC.

Briefings

Terra Quantum and Empa developed LP-FNO, a Fourier Neural Operator model predicting 3D temperature fields and melt-pool boundaries from laser welding parameters. Trained on Ti-6Al-4V simulations across 40-190W power and 0.1-1m/s speed, the model generates full 3D predictions in 8-88 milliseconds. It achieves up to 100,000x speedup over traditional multiphysics simulations, with approximately 2.5% average relative error in temperature and 0.9 IoU for melt-pool boundaries. The model handles both conduction and stable keyhole welding regimes via a quasi-steady reformulation, enabling real-time process control and digital twins.

Key points

• LP-FNO predicts full 3D temperature fields and melt-pool boundaries from laser welding parameters using a Fourier Neural Operator architecture.
• The model achieves up to 100,000x speedup over traditional multiphysics simulations while maintaining approximately 2.5% average relative error in temperature predictions.

Context and limitations

The model was trained on high-fidelity simulations of Ti-6Al-4V titanium alloy only. Coverage is limited to conduction and stable keyhole welding regimes; the model's performance in other regimes remains unverified.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 29/09/2026, 12:06:46 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeiafbpkxybajz4a4xd22efvv4g6wgrb3warycztgh3j7higkgnhlxu