New IAEA Research Project Uses Machine Learning to Better Predict Polymer Changes under Radiation
| Source: International Atomic Energy Agency | Original article
The International Atomic Energy Agency (IAEA) has launched a coordinated research project that will bring machine‑learning techniques to bear on the long‑standing challenge of predicting how ionising radiation alters polymer structures. The agency’s call for proposals, issued this week, invites universities, national labs and industry partners to develop data‑driven models that can forecast chain scission, cross‑linking and embrittlement in the wide range of polymers used in nuclear reactors, medical devices, space hardware and radioactive waste containers.
Radiation‑induced degradation is a critical reliability issue: polymer seals, cable jackets and shielding foils can fail unexpectedly, prompting costly shutdowns or safety incidents. Traditional approaches rely on time‑consuming experiments and physics‑based simulations that struggle to capture the complex chemistry of high‑energy particle interactions. By training algorithms on existing degradation datasets and on new measurements generated under the project, researchers aim to produce predictive tools that run in minutes rather than weeks, enabling designers to screen materials early and to plan replacement schedules with greater confidence.
The IAEA’s initiative dovetails with a broader push to embed artificial intelligence in nuclear science, echoing recent work on reinforcement‑learning‑enhanced simulations and neuro‑symbolic models for process monitoring. Success could accelerate the rollout of radiation‑resistant polymers, lower maintenance costs for power plants, and improve the durability of medical implants that operate in radiotherapy environments.
Watch for the deadline for project proposals, slated for late May, and for the first consortium announcement expected in the autumn. Subsequent milestones will include the release of an open‑access polymer‑radiation database, benchmark ML models, and pilot validation studies in operating reactors and hospital settings. The outcomes will likely inform future IAEA safety guidelines and could set a new standard for AI‑driven materials engineering in the nuclear sector.
Sources
Back to AIPULSEN