From scaling laws to the design space: How we design machines we cannot yet build

Value of the Information & Strength of the Argument

The talk provides valuable insights into the application of integrated modeling and optimization to fusion device design. The argumentation is solid, based on detailed simulations and comparisons with experimental data. The speaker clearly explains the limitations of scaling laws and demonstrates how a design-space approach can reveal trade-offs and sensitivities that point-design studies miss. The use of multi-objective optimization and large parameter sweeps is well-justified, and the discussion of uncertainties, such as the L-H transition threshold, adds depth. The presentation of the FUSE Explorer as a public resource is a significant contribution, enabling broader engagement with the data and models.

Scientific Rigor, Source Quality, Title Accuracy

The talk references specific publications (e.g., Slendebroek et al., Nucl. Fusion 66, 026032, 2026) and the FUSE framework, which is established in the literature. The methodology is described in sufficient detail to be reproducible, and the speaker is transparent about assumptions and limitations. The title accurately reflects the content, and the talk stays on topic. The presentation is rigorous, with clear explanations of the models and their validation. The speaker also acknowledges the need for further validation, particularly for negative triangularity configurations.

The title accurately reflects the content: the talk transitions from scaling laws to a design-space exploration approach for fusion devices.

8/10

The talk presents original research within an established integrated modeling framework (FUSE), with references to peer-reviewed publications and a publicly accessible database. The speaker is transparent about uncertainties and limitations, and the methodology is reproducible via the described tools.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 00:00Introduction: problem with scaling laws for fusion power plant design.
• 02:00Outline of the talk: database validation, design space, and predictions for ITER.
• 04:00Starting simulations from zero-dimensional parameters; FUSE workflow.
• 06:00Validation against ITPA confinement database; comparison with scaling laws.
• 08:00Example of scaling law failure: EU DEMO prediction vs. integrated modeling.
• 10:00Introduction to FUSE framework: components and design philosophy.
• 12:00Multi-objective optimization setup: objectives, constraints, and genetic algorithm.
• 14:00Comparison of positive and negative triangularity designs; trade-offs.
• 16:00Impact of heat exhaust proxy and L-H threshold uncertainty on design space.
• 18:00FUSE Explorer: public web tool for exploring design space and running simulations.
• 20:00Interactive example: running a different transport model on a design point.
• 22:00Large-scale parametric study of ITER-like plasmas: 1.5 million cases.
• 24:00Analysis of convergence and filtering: only a fraction of cases are viable.
• 26:00Implications for ITER design: mapping the design space and identifying risks.
• 28:00Future directions: agentic workflows and autonomous design exploration.

• IPAM Workshop: Fusion Device Design and Engineering — Workshop page where the talk was recorded.

• IPAM Workshop: Fusion Device Design and Engineering — Workshop page where the talk was recorded.