{"slug": "surrogate-model-for-temperature-fields-from-a-small-set-of-cfd-runs-which-to", "title": "Surrogate model for temperature fields from a small set of CFD runs: which architecture to start with?", "summary": "A practitioner building a surrogate model for steady-state temperature fields on heat sinks and electronics boards, trained on roughly 150–300 CFD simulation runs, was advised to stay with PCA plus a small model rather than adopt a Fourier neural operator or graph-based model at that data scale. The recommended fix for the model's missed hot spots near sharp features is to score on worst-case and 95th-percentile error and hotspot location, add a second small head predicting peak temperature and its position, or weight the loss toward high-gradient cells, and only then test a CNN on rasterized geometry against the same held-out geometries.", "body_md": "Hi everyone,\n\nI’m trying to build a surrogate model that predicts steady-state temperature fields on a [heat sink / electronics board] from its geometry parameters, so I can skip running a full simulation for every design change.\n\nMy setup so far:\n\n•        About [150–300] simulation runs, each with [N] geometry parameters (fin height, spacing, base thickness, etc.)\n\n•        Output is a temperature field on a [structured grid / unstructured mesh with ~X nodes]\n\n•        I’ve tried a simple MLP on the parameters, predicting a downsampled field. Mean error is acceptable, but it misses hot spots near sharp features.\n\nWith a dataset this small, is it more sensible to go with a Fourier neural operator, a graph-based model, or a CNN on a rasterized geometry? Or should I stay with a simpler model plus a PCA reduction of the field?\n\nIf there are open datasets or Hub models for physics surrogates that I could fine-tune or learn from, I’d appreciate pointers. I can share my training loop and a sample of the data if that helps.\n\nThanks!\n\n \n \nWith 150–300 runs, I would stay with PCA plus a small model for now — an FNO or graph model will likely spend that budget learning the field’s basics, while your current miss has a cheaper cause: a downsampled field trained on mean error can be right on average and still smooth away the peaks. Keep the simple model, but score it on worst-case and 95th-percentile error (and whether the hotspot lands in the right place), and add a second small head that predicts peak temperature and its location, or weight the loss toward high-gradient cells near sharp features. If that still misses, test a CNN on the rasterized geometry on the same held-out geometries — that comparison, not the architecture names, will tell you whether geometry representation is really the bottleneck.", "url": "https://wpnews.pro/news/surrogate-model-for-temperature-fields-from-a-small-set-of-cfd-runs-which-to", "canonical_source": "https://discuss.huggingface.co/t/surrogate-model-for-temperature-fields-from-a-small-set-of-cfd-runs-which-architecture-to-start-with/190855#post_2", "published_at": "2026-10-21 08:08:50+00:00", "updated_at": "2026-10-21 08:12:44.747301+00:00", "lang": "en", "topics": ["machine-learning", "ai-research", "neural-networks"], "entities": [], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/surrogate-model-for-temperature-fields-from-a-small-set-of-cfd-runs-which-to", "markdown": "https://wpnews.pro/news/surrogate-model-for-temperature-fields-from-a-small-set-of-cfd-runs-which-to.md", "text": "https://wpnews.pro/news/surrogate-model-for-temperature-fields-from-a-small-set-of-cfd-runs-which-to.txt", "jsonld": "https://wpnews.pro/news/surrogate-model-for-temperature-fields-from-a-small-set-of-cfd-runs-which-to.jsonld"}}