Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Diffusion Model Inverse Design”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Diffusion Model-Guided Inverse Design of Bimetallic Catalysts for Ammonia Decomposition

In the past decade, artificial intelligence and deep learning have played increasingly prominent roles in materials design and discovery. Among these, generative AI models, known for their ability to create unique and complex structures, have emerged as state-of-the-art tools for materials screening due to their high efficiency and low computational cost. In catalysis, one of the major challenges is identifying promising material candidates within an immense chemical space. This challenge can be addressed using generative approaches, such as diffusion-based inverse design models. In this study, we present a machine learning-guided workflow that employed a diffusion model for the inverse design of bimetallic alloy catalysts for low-carbon ammonia decomposition, a key reaction for ammonia emission control and sustainable hydrogen production. Catalyst candidates were evaluated using nitrogen adsorption energy as the key descriptor, inspired by multiscale modeling. The proposed workflow identified low-cost, environmentally friendly catalysts with excellent catalytic performance, which have been validated theoretically and experimentally. Our framework decoupled the generative and property-prediction components, enhancing both flexibility and accuracy in the catalytic material design process.

Adsorption↗
DOE OSTI2026

Interpretable Deep Learning for Advancing Field-Enhanced Catalysis

This DOE Early Career project developed a physics-informed, interpretable AI-and-modeling framework to understand and exploit electric-field effects in heterogeneous catalysis, with ammonia cracking and synthesis as a representative pathway. The team built and validated methods to map local electric fields on metal surfaces and nanoparticles, showing that low-coordination features (tips/edges/corners) can concentrate fields by several-fold relative to flat facets. Using DFT-generated datasets, the project created physics-guided machine learning models that rapidly predict local electric fields and field-dependent adsorption energetics with near-DFT accuracy while reducing computational cost by orders of magnitude. These predictions were integrated with microkinetic modeling to quantify how field-dipole interactions reshape reaction energetics and mechanisms, enabling large increases in predicted catalytic rates and substantial reductions in operating temperature under favorable field conditions. To accelerate discovery of earth-abundant catalysts, the project combined interpretable ML screening (with electronic-structure descriptors identified as key drivers) with a generative inverse-design workflow based on diffusion models and physics constraints. The resulting closed-loop approach, linking simulation, mechanistic modeling, and AI, provides reusable tools and datasets for designing catalysts and operating conditions in field-enhanced catalysis, with broad relevance to electrostatic catalysis, plasma catalysis, electrocatalysis, and other energy-related chemical transformations.

30 DIRECT ENERGY CONVERSION↗
DOE OSTI2024

Denoising diffusion probabilistic models for generative alloy design

Inverse material design is an extremely challenging optimization task made difficult by, in part, the highly nonlinear relationship linking performance with composition. Quantitative approaches have improved significantly owing to advances in high throughput experimentation and computational thermodynamics. However, existing physics-based tools are mostly forward models; input a chemistry and obtain a prediction. More recently the materials community has leveraged advances in the machine learning community to establish novel inverse design frameworks. Very recently denoising diffusion probabilistic models have been shown to be extremely powerful generators producing synthetic data of various modalities e.g. images, text, audio, tables, etc.. In this work a novel framework for alloy design and optimization is proposed leveraging these class of models. Five key generative tasks are demonstrated (1) unconditional generation (2) composition conditioned generation (3) property conditioned generation (4) multi-feedstock conditioned generation and (5) generative optimization. These methods were tested on three case studies: high entropy alloy design, superalloy binder jet additive manufacturing, and in-situ dual-feedstock wire-arc additive manufacturing. Results indicate that the established models are extremely flexible, expressive, and robust. The architecture’s flexibility and training procedure empower the model to learn complex intra-compositional and composition-property relationships. Furthermore, the probabilistic nature of these models makes them well suited for addressing solution non-uniqueness and tackling uncertainty quantification tasks. While the fidelity and quantity of the underlying training data is paramount, we envision that future alloy design frameworks will make extensive use of these kinds of machine learning models as “search” tools bolstering the utility of experimental and computational approaches.

36 MATERIALS SCIENCE↗
DOE OSTI2026

EvoDiffMol: evolutionary diffusion framework for 3D molecular design with optimized properties

Designing molecules with specific target properties remains a fundamental challenge in computational chemistry. While existing approaches show promise, most rely on simplified representations like SMILES strings or 2D graphs that lack essential three-dimensional geometric information. We present EvoDiffMol, a computational framework that integrates evolutionary algorithms with three-dimensional diffusion models for property-driven molecular generation. The method operates through adaptive evolutionary optimization, where population-based selection guides the generation process toward desired property landscapes. EvoDiffMol supports both unconstrained molecular design and scaffold-constrained generation that preserves fixed substructures while optimizing complementary regions. Comprehensive evaluation demonstrates exceptional performance, achieving the highest drug-likeness score (0.94) among all compared state-of-the-art methods while maintaining excellent validity, uniqueness, and novelty. Beyond single property optimization, the framework demonstrates flexible multi-property optimization capabilities, simultaneously controlling multiple molecular descriptors including synthetic accessibility, lipophilicity, topological polar surface area, and clinically relevant ADMET properties such as cardiotoxicity (hERG) and intestinal permeability (Caco-2). This adaptability spans from simple descriptors to practical pharmaceutical endpoints without requiring complete model retraining. The framework achieves precise control over target property values, generating molecules with properties closely matching specified targets for both single and multiple descriptors. Scaffold-constrained experiments preserve fixed molecular cores while maintaining effective property optimization. The three-dimensional representation offers advantages in maintaining structural validity during iterative optimization, with potential for geometry-aware applications in materials science and drug discovery.

3D molecular generation↗
DOE OSTI2025

Morphogenic Growth 3D Printing

Inspired by nature's morphogenesis, a new 3D printing process –growth printing (GP)– takes advantage of a self‐propagating curing front to produce 3D polymeric parts following a growth‐like development plan. The propagation of the curing front is driven by the exothermic polymerization of dicyclopentadiene (DCPD), which transforms the liquid resin into a stiff polymer as it propagates at 1 mm s −1 . GP is triggered when a heated initiator contacts the uncured liquid resin in an open container. The initiator nucleates the frontal polymerization reaction and the isotropic radial propagation of the growth front. Simultaneously, the initiator is moved up across the free surface of the resin, pulling the cured object out of the uncured resin. The motion trajectory of the initiator with respect to the free resin surface controls the growth morphology of the 3D part. An inverse design algorithm is developed to produce 3D parts by modeling the reaction‐diffusion‐driven solidification process. This process has substantial energy savings and high printing speeds.

3D printing↗
DOE OSTI2025

Physically constrained 3D diffusion for inverse design of fiber-reinforced polymer composite materials

Designing fiber-reinforced polymer composites (FRPCs) with a tailored nonlinear stress-strain response is crucial for applications such as energy absorption in crash structures, flexible robotics, and impact-resistant protective gear. However, the inherent complexities of composite materials and the multitude of parameters involved, render traditional design and optimization methods inadequate for achieving effective inverse design of composites. In this paper, we present an AI-based inverse design framework that effectively and efficiently generates FRPCs with targeted nonlinear stress-strain responses. We introduce a physically constrained diffusion model (PC3D_Diffusion) capable of managing the complexities of composite materials and producing detailed, high-quality designs. We propose a loss-guided, learning-free approach to generate physically feasible microstructure designs by explicitly enforcing physical constraints during the generation process. For training purposes, 1.35 million FRPC samples were created, and their corresponding stress-strain curves were computed using established physics-based computational models. The results show that PC3D_Diffusion consistently generates high-quality designs with tailored mechanical behaviors, while guaranteeing compliance with the physical constraints. PC3D_Diffusion advances FRPC inverse design and may facilitate the inverse design of other 3D materials, offering potential applications in industries reliant on materials with custom mechanical properties.

Xu, Pei [Clemson Univ., SC (United States)]↗
DOE OSTI2023

Towards inverse microstructure-centered materials design using generative phase-field modeling and deep variational autoencoders

The field of Integrated Computational Materials Engineering (ICME) combines a broad range of methods to study materials’ responses over a spectrum of length scales. A relatively unexplored aspect of microstructure-sensitive materials design is uncertainty propagation and quantification (UP/UQ) of materials’ microstructure, as well as establishing process-structure–property (PSP) relationships for inverse material design. In this study, an efficient UP technique built on the idea of changing probability measures and a deep generative unsupervised representative machine learning method for microstructure-based design of thermal conductivity of materials is proposed. Probability measures are used to represent microstructure space, and Wasserstein metrics are used to test the efficiency of the UP method. By using deep Variational AutoEncoder (VAE), we identify the correlations between the material/process parameters and the thermal conductivity of heterogeneous dual-phase microstructures. Through high-throughput screening, UP, and the deep-generative VAE method, PSP relationships that are too complex can be revealed by exploiting the materials’ design space with an emphasis on microstructures. As a last point, we demonstrate generative machine learning serves as a useful tool for inverse microstructure-centered materials design, and we demonstrate this by examining the inverse design of thermal conductivity in nano-structured materials. Here, the results reveal the effects of morphology, volume fraction, characteristic length scale, and the individual thermal diffusivity of phases on the thermal conductivity of dual-phase alloys. Our findings emphasize the advantages of high-throughput phase-field modeling and generative deep learning for linking PSP and inverse microstructure-centered materials design.

36 MATERIALS SCIENCE↗
DOE OSTI2025

High‐Asymmetry Metasurface: A New Solution for Terahertz Resonance via Active Learning‐Augmented Diffusion Model

Terahertz (THz) metamaterials with high‐figure‐of‐merit (high‐FoM) performance resonance are essential for advancing sensors, detectors, and imagers. Conventional designs focus on symmetric or low‐asymmetry geometric structures, leaving high‐asymmetry designs largely unexplored due to the inefficiency of trial‐and‐error‐based rational design. Recent deep learning techniques offer automation and acceleration but are constrained by the need for large datasets inherent to their data‐driven nature. Here, a novel prior knowledge‐guided generative model augmented by a physics‐constrained active learning mechanism to design high‐asymmetry metamaterials. An advanced diffusion model learns features from a small set of classical structures with high‐FoM THz resonance and generates new high‐asymmetry structures. To mitigate the limited number of classical structures, the generated high‐asymmetry structures are actively selected and integrated into the initial training dataset based on their physical characteristics. Experimental results demonstrate the superior resonance performance of the generated high‐asymmetry metamaterials over classical designs, exhibiting improvements exceeding 30% in key resonance metrics. Remarkably, this performance is attained using only 68 classical structures as the initial training dataset, significantly reducing the data requirements for deep learning‐based metamaterial design. The proposed scheme for generating high‐asymmetry structures provides a new effective and efficient solution for high‐FoM resonance, expanding applications in high‐sensitivity THz metadevices.

diffusion model↗
DOE OSTI2023

Denoising diffusion algorithm for inverse design of microstructures with fine-tuned nonlinear material properties

Here we introduce a denoising diffusion algorithm to discover microstructures with nonlinear fine-tuned properties. Denoising diffusion probabilistic models are generative models that use diffusion-based dynamics to gradually denoise images and generate realistic synthetic samples. By learning the reverse of a Markov diffusion process, we design an artificial intelligence to efficiently manipulate the topology of microstructures to generate a massive number of prototypes that exhibit constitutive responses sufficiently close to designated nonlinear constitutive behaviors. To identify the subset of microcstructures with sufficiently precise fine-tuned properties, a convolutional neural network surrogate is trained to replace high-fidelity finite element simulations to filter out prototypes outside the admissible range. Results of this study indicate that the denoising diffusion process is capable of creating microstructures of fine-tuned nonlinear material properties within the latent space of the training data. More importantly, this denoising diffusion algorithm can be easily extended to incorporate additional topological and geometric modifications by introducing high-dimensional structures embedded in the latent space. Numerical experiments are conducted on the open-source mechanical MNIST data set (Lejeune, 2020). Consequently, this algorithm is not only capable of performing inverse design of nonlinear effective media, but also learns the nonlinear structure–property map to quantitatively understand the multiscale interplay among the geometry, topology, and their effective macroscopic properties.

42 ENGINEERING↗
DOE OSTI2022

Optimal Experimental Design for Inverse Problems in the Presence of Observation Correlations

Optimal experimental design (OED) is the general formalism of sensor placement and decisions on the data collection strategy for engineered or natural experiments. This problem is prevalent in many critical fields such as battery design, numerical weather prediction, geosciences, environmental and urban studies. State-of-the-art computational methods for experimental design do not accommodate correlation structure in observational errors produced by many expensive-to-operate devices such as X-ray machines, radars, and satellites. Discarding evident data correlations leads to biased results, higher expenses, and waste of valuable resources. We present a general formulation of the OED formalism for model-constrained large-scale Bayesian linear inverse problems, where measurement errors are generally correlated. The proposed approach utilizes the Hadamard product of matrices to formulate the weighted-likelihood, and is valid for both finite as well as infinite-dimensional Bayesian inverse problems. Extensive numerical experiments are carried out for empirical verification of the proposed approach using an advection-diffusion model, where the objective is to optimally place a small set of sensors, under a limited budget, to predict the concentration of a contaminant in a closed and bounded domain.

Correlated observations↗
NASA NTRS2004

A Variable Turbulent Schmidt Number Formulation for Scramjet Application

In high speed engines, thorough turbulent mixing of fuel and air is required to obtain high performance and high efficiency. Thus, the ability to predict turbulent mixing is crucial in obtaining accurate numerical simulation of an engine and its performance. Current state of the art in CFD simulation is to assume both turbulent Prandtl number and Schmidt numbers to be constants. However, since the mixing of fuel and air is inversely proportional to the Schmidt number, a value of 0.45 for the Schmidt number will produce twice as much diffusion as that with a value of 0.9. Because of this, current CFD tools and models have not been able to provide the needed guidance required for the efficient design of a scramjet engine. The goal of this investigation is to develop the framework needed to calculate turbulent Prandtl and Schmidt numbers as part of the solution. This requires four additional equations: two for the temperature variance and its dissipation rate and two for the concentration variance and its dissipation rate. In the current investigation emphasis will be placed on studying mixing without reactions. For such flows, variable Prandtl number does not play a major role in determining the flow. This, however, will have to be addressed when combustion is present. The approach to be used is similar to that used to develop the k-zeta model. In this approach, relevant equations are derived from the exact Navier-Stokes equations and each individual correlation is modeled. This ensures that relevant physics is incorporated into the model equations. This task has been accomplished. The final set of equations have no wall or damping functions. Moreover, they are tensorially consistent and Galilean invariant. The derivation of the model equations is rather lengthy and thus will not be incorporated into this abstract, but will be included in the final paper. As a preliminary to formulating the proposed model, the original k-zeta model with constant turbulent Prandtl and Schmidt numbers is used to model the supersonic coaxial jet mixing experiments involving He, O2 and air.

Xiao, X.↗
NASA NTRS1993

A numerical treatment of geodynamic viscous flow problems involving the advection of material interfaces

Effective numerical treatment of multicomponent viscous flow problems involving the advection of sharp interfaces between materials of differing physical properties requires correction techniques to prevent spurious diffusion and dispersion. We develop a particular algorithm, based on modern shock-capture techniques, employing a two-step nonlinear method. The first step involves the global application of a high-order upwind scheme to a hyperbolic advection equation used to model the distribution of distinct material components in a flow field. The second step is corrective and involves the application of a global filter designed to remove dispersion errors that result from the advection of discontinuities (e.g., material interfaces) by high-order, minimally dissipative schemes. The filter introduces no additional diffusion error. Nonuniform viscosity across a material interface is allowed for by the implementation of a compositionally weighted-inverse interface viscosity scheme. The combined method approaches the optimal accuracy of modern shock-capture techniques with a minimal increase in computational time and memory. A key advantage of this method is its simplicity to incorporate into preexisting codes be they finite difference, element, or volume of two or three dimensions.

Lenardic, A.↗
NASA NTRS2018

MODIS Solar Diffuser Degradation Determination and Its Spectral Dependency

This study presents a modeling approach to improve solar diffuser (SD) degradation determination from SD stability monitor (SDSM) measurements. The MODIS instrument uses a SDto calibrate its reflective solar bands (RSBs) on-orbit. Due to the imperfectly designed SDSM sun view screen, the SD reflectance tracked by SDSM has large noise. The SDSM measurements noise is spectrally coherent and can be minimized by normalizing measurements to the least degradeddetector 9 (936 nm). In this study, a SD degradation model is used to determine the SDdegradation's wavelength dependency and the detector 9 degradation is estimated by the model solution.The results show the SD degradations measured at 6 SDSM detectors (554 _ 936 nm) have stable relationships, where the degradation is inversely proportion to 1/wavelength^4. The model estimated SD degradation at SDSM detector 9 wavelength (936 nm) is ~0.9% from 2002 to 2018.Based on the SD degradation model solution, the SD degradation at short/mid wave bands are estimated to improve short/mid wave bands calibration. The model can also be used to improve interpolating SD degradation at SDSM detectors to RSB wavelengths. Compared to linear interpolation, bands 9 and 10 show the largest differences of up to 0.3 and 0.4% respectively. These differences directly impact the calibration coefficients of these bands.

radiometric calibration↗
NASA NTRS2014

SWIM: A Semi-Analytical Ocean Color Inversion Algorithm for Optically Shallow Waters

In clear shallow waters, light that is transmitted downward through the water column can reflect off the sea floor and thereby influence the water-leaving radiance signal. This effect can confound contemporary ocean color algorithms designed for deep waters where the seafloor has little or no effect on the water-leaving radiance. Thus, inappropriate use of deep water ocean color algorithms in optically shallow regions can lead to inaccurate retrievals of inherent optical properties (IOPs) and therefore have a detrimental impact on IOP-based estimates of marine parameters, including chlorophyll-a and the diffuse attenuation coefficient. In order to improve IOP retrievals in optically shallow regions, a semi-analytical inversion algorithm, the Shallow Water Inversion Model (SWIM), has been developed. Unlike established ocean color algorithms, SWIM considers both the water column depth and the benthic albedo. A radiative transfer study was conducted that demonstrated how SWIM and two contemporary ocean color algorithms, the Generalized Inherent Optical Properties algorithm (GIOP) and Quasi-Analytical Algorithm (QAA), performed in optically deep and shallow scenarios. The results showed that SWIM performed well, whilst both GIOP and QAA showed distinct positive bias in IOP retrievals in optically shallow waters. The SWIM algorithm was also applied to a test region: the Great Barrier Reef, Australia. Using a single test scene and time series data collected by NASA's MODIS-Aqua sensor (2002-2013), a comparison of IOPs retrieved by SWIM, GIOP and QAA was conducted.

shallow waters↗
DOE OSTI2023

Evaluation of dual-weighted residual and machine learning error estimation for projection-based reduced-order models of steady partial differential equations

Projection-based reduced-order models (pROMs) show great promise as a means to accelerate many-query applications such as forward error propagation, solving inverse problems, and design optimization. In order to deploy pROMs in the context of high-consequence decision making, accurate error estimates are required to determine the region(s) of applicability in the parameter space. The following paper considers the dual-weighted residual (DWR) error estimate for pROMs and compares it to another promising pROM error estimate, machine learned error models (MLEM). Here, we show how DWR can be applied to ROMs and then evaluate DWR on two partial differential equations (PDEs): a two-dimensional linear convection–reaction–diffusion equation, and a three-dimensional static hyper-elastic beam. It is shown that DWR is able to estimate errors for pROMs extrapolating outside of their training set while MLEM is best suited for pROMs used to interpolate within the pROM training set.

42 ENGINEERING↗
NASA NTRS1998

Helioseismic Constraints on New Solar Models from the MoSEC Code

Evolutionary solar models are computed using a new stellar evolution code, MOSEC (Modular Stellar Evolution Code). This code has been designed with carefully controlled truncation errors in order to achieve a precision which reflects the increasingly accurate determination of solar interior structure by helioseismology. A series of models is constructed to investigate the effects of the choice of equation of state (OPAL or MHD-E, the latter being a version of the MHD equation of state recalculated by the author), the inclusion of helium and heavy-element settling and diffusion, and the inclusion of a simple model of mixing associated with the solar tachocline. The neutrino flux predictions are discussed, while the sound speed of the computed models is compared to that of the sun via the latest inversion of SOI-NMI p-mode frequency data. The comparison between models calculated with the OPAL and MHD-E equations of state is particularly interesting because the MHD-E equation of state includes relativistic effects for the electrons, whereas neither MHD nor OPAL do. This has a significant effect on the sound speed of the computed model, worsening the agreement with the solar sound speed. Using the OPAL equation of state and including the settling and diffusion of helium and heavy elements produces agreement in sound speed with the helioseismic results to within about +.-0.2%; the inclusion of mixing slightly improves the agreement.

Elliott, J. R.↗
DOE OSTI2023

Bayesian sequential optimal experimental design for nonlinear models using policy gradient reinforcement learning

We present a mathematical framework and computational methods for optimally designing a finite sequence of experiments. This sequential optimal experimental design (sOED) problem is formulated as a finite-horizon partially observable Markov decision process (POMDP) under a Bayesian setting and with information-theoretic utilities. The formulation is general and may accommodate continuous random variables, non-Gaussian posteriors, and nonlinear forward models. The sOED design policy incorporates elements of feedback and lookahead simultaneously, and we show it to generalize the commonly-used batch and greedy design strategies. We solve for the sOED policy using the policy gradient (PG) method from reinforcement learning, and provide a derivation for the PG expression in the sOED context. Adopting an actor-critic approach, the policy and value functions are parameterized using deep neural networks and improved via PG estimates produced from simulated episodes of designs and observations. The new PG-sOED algorithm is first validated on a linear-Gaussian benchmark, and then compared against other design baselines on a sensor movement problem for contaminant source inversion in a convection-diffusion field. As a result, we provide explanation for the policy behaviors using knowledge of the underlying physical process.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

High Performance PEFC Electrode Structures

Raytheon Technologies Research Center (RTRC) in collaboration with Ion Power Inc. and the University of Arkansas at Little Rock and the FC-PAD consortium executed a project that: (1) developed comprehensive models of oxygen and proton transport in cathode catalyst layers that provided new insights about losses associated with reducing platinum loading to 0.1 mg/cm 2 , (2) designed, fabricated, characterized, and tested cells featuring carbon-supported catalyst layers with low platinum loadings that met efficiency and power density targets set forth by the Hydrogen and Fuel Cells Technology Office, and (3) designed, fabricated, characterized, and tested novel catalyst structures. The modeling efforts produced eight scientific publications that shared new descriptions of oxygen and proton transport in catalyst layers with the broader technical community. The key physical insights from the models are expressed in straightforward algebraic expressions, facilitating easy adoption by other researchers. The models developed during the project explain why resistance to oxygen transport is inversely proportional to interfacial area of platinum – an observation that eluded quantitative explanation for a decade. The models quantitatively predict polarization when measured morphological parameters and transport properties are input. Membrane-electrode assemblies made during the project met stated goals for current at high voltage and power density at rated voltage for transportation fuel cells at the end of budget period 2. High performance was achieved by developing a diffusion layer with low resistance to oxygen transport and a PtCo/C cathode catalyst layer capable of supporting rapid proton and oxygen transport. The project was subsequently directed to curtail experimental work and focus on modeling during budget period 3. Ion Power Inc. developed new expertise in fabricating catalyst layers containing platinum-cobalt catalysts, and processing very thin membranes. The University of Arkansas at Little Rock developed novel nanocolumnar self-supported thin film electrocatalysts that can be grown on carbon supports using a simple high pressure sputter deposition method. The nanocolumnar microstructure provides adequate surface-to-volume ratio for efficient platinum utilization, and the conformal platinum shell with larger crystal grain sizes covering the carbon support surface may eliminate durability issues associated with catalyst dissolution, agglomeration, and carbon corrosion.

08 HYDROGEN↗