000168489 001__ 168489 000168489 005__ 20260209162330.0 000168489 0247_ $$2doi$$a10.1016/j.rineng.2025.108418 000168489 0248_ $$2sideral$$a146722 000168489 037__ $$aART-2026-146722 000168489 041__ $$aeng 000168489 100__ $$0(orcid)0009-0007-8488-6089$$aTierz, Alicia$$uUniversidad de Zaragoza 000168489 245__ $$aVariational rank reduction autoencoders for generative thermal design 000168489 260__ $$c2026 000168489 5060_ $$aAccess copy available to the general public$$fUnrestricted 000168489 5203_ $$aGenerative thermal design for complex geometries is fundamental in many areas of engineering, yet it faces two main challenges: the high computational cost of high-fidelity simulations and the limitations of conventional generative models. Approaches such as autoencoders (AEs) and variational autoencoders (VAEs) often produce unstructured latent spaces with discontinuities, which restricts their capacity to explore designs and generate physically consistent solutions. To address these limitations, we propose a hybrid framework that combines Variational Rank-Reduction Autoencoders (VRRAEs) with Deep Operator Networks (DeepONets). The VRRAE introduces a truncated SVD within the latent space, leading to continuous, interpretable, and well-structured representations that mitigate posterior collapse and improve geometric reconstruction. The DeepONet then exploits this compact latent encoding in its branch network, together with spatial coordinates in the trunk network, to predict temperature gradients efficiently and accurately. This hybrid approach not only enhances the quality of generated geometries and the accuracy of gradient prediction, but also provides a substantial advantage in inference efficiency compared to traditional numerical solvers. Overall, the study underscores the importance of structured latent representations for operator learning and highlights the potential of combining generative models and operator networks in thermal design and broader engineering applications. 000168489 536__ $$9info:eu-repo/grantAgreement/ES/MICINN/PID2023-147373OB-I00$$9info:eu-repo/grantAgreement/ES/MTFP/TSI-100930-2023-1 000168489 540__ $$9info:eu-repo/semantics/openAccess$$aby$$uhttps://creativecommons.org/licenses/by/4.0/deed.es 000168489 655_4 $$ainfo:eu-repo/semantics/article$$vinfo:eu-repo/semantics/publishedVersion 000168489 700__ $$aMounayer, Jad 000168489 700__ $$0(orcid)0000-0001-5483-6012$$aMoya, Beatriz 000168489 700__ $$aChinesta, Francisco 000168489 7102_ $$15007$$2570$$aUniversidad de Zaragoza$$bDpto. Informát.Ingenie.Sistms.$$cÁrea Lenguajes y Sistemas Inf. 000168489 773__ $$g29 (2026), 108418 [13 pp.]$$tResults in Engineering$$x2590-1230 000168489 8564_ $$s35808464$$uhttps://zaguan.unizar.es/record/168489/files/texto_completo.pdf$$yVersión publicada 000168489 8564_ $$s2729755$$uhttps://zaguan.unizar.es/record/168489/files/texto_completo.jpg?subformat=icon$$xicon$$yVersión publicada 000168489 909CO $$ooai:zaguan.unizar.es:168489$$particulos$$pdriver 000168489 951__ $$a2026-02-09-14:42:01 000168489 980__ $$aARTICLE