arXiv subjects

Douglas Allaire

Publications and source records attributed to Douglas Allaire.

16 recordsLinked to original sources

arXiv2026

Portfolio-Based Constrained Multi-Objective Bayesian Optimization for Materials Design

Materials discovery and design campaigns can be formulated as constrained multi-objective Bayesian optimization (CMOBO) problems, within which each experimental decision negotiates between two coupled but competing goals: discovering feasible candidates and refining the underlying Pareto front. Here we recast acquisition-function choice as an adaptive policy-selection problem over a portfolio of conventional and feasibility-focused acquisition functions. This was done using two controllers: UCB-Bandit, a modified UCB multi-armed bandit, and Agentic-Switch, a multi-agent decision system driven by a large language model (LLM). Both were evaluated against fixed-policy baselines in silico across five synthetic benchmark functions and two materials design case studies. The adaptive policies performed competitively in terms of both cumulative feasibility count and feasible hypervolume improvement, while each individual acquisition function performed well for only one metric, suggesting that adaptive policies are better suited for constrained materials science problems.

cond-mat.mtrl-sci↗

arXiv2026

Surrogate-Gated Generation and Foundation-Model Embeddings for Bayesian Materials Design

Closed-loop materials discovery iterates between proposing candidate structures and evaluating their properties, and property evaluation dominates the cost. In the generative variant, a learned prior proposes candidate crystals and a property oracle scores them; we ask whether a cheap probabilistic surrogate can triage the generator's output, and what such a surrogate must do well. Across three architecturally distinct pretrained diffusion priors (MatterGen, CrystalFlow, ADiT) and two targets (room-temperature heat capacity and bulk modulus), we insert a Gaussian process acquisition gate between structure generation and the oracle in an RL-steered generative workflow. The gate matches or exceeds ungated fine-tuning of the generative model while capping oracle calls at a fixed per-cycle budget. Budget-matched ablations isolate the mechanism. At an identical four-call budget, ranking-based selection outperforms arbitrary selection, confirming that the gain comes from the surrogate's choice; the gate comes within $\sim$9\% of exhaustive oracle spending at roughly one-fifth of the calls. A density-functional-theory check of the bulk-modulus discoveries confirms the learned oracle to within 2.5\% on average and the surrogate's ranking of the generated structures at Spearman $ρ= 0.94$. A cross-factorial benchmark of surrogate performance spanning mechanical, electronic, and vibrational properties identifies pretrained ORB embeddings with a Gaussian process as the most reliable combination, which we adopt as the building blocks of the proposed workflow. The complete pipeline is released as open-source software.

cond-mat.mtrl-sci↗

arXiv2026

A BRAVE Alloy Design Campaign (Bayesian Risk-aware Alloy discoVery and Exploration)

In constrained alloy optimization, the compositions with the highest performance potential often reside at the boundary of phase stability -- where the risk of experimental failure is also highest. This work demonstrates this principle through a risk-aware Bayesian optimization campaign on single-phase FCC high-entropy alloys in the Al-V-Cr-Mn-Fe-Co-Ni-Cu system. A learned feasibility classifier, integrated directly into the multi-objective acquisition function, probabilistically penalizes candidates likely to produce failed experiments while preserving access to high-performing boundary compositions. From approximately 27,000 CALPHAD-screened candidates, 48 alloys were synthesized over three closed-loop iterations targeting five objectives (yield strength, UTS/YS ratio, strain at UTS, dynamic-to-quasi-static hardness ratio, and simulated depth of penetration), exploring 0.12\% of the feasible space. Two compositional regimes emerged: a V-rich, Ni-rich high-strength regime (UTS up to ${\sim}1480$~MPa at 50% elongation) and a Mn-containing high-ductility regime (UTS/YS up to 4.20 at $>$50% elongation). Among feasible alloys, vanadium simultaneously drives yield strength ($r = 0.84$) and sigma-phase formation ($r = 0.54$ with infeasibility); at V = 24~at.%, the three strongest alloys and three sigma failures share the same compositional point. Additionally, the strongest performing alloys cluster in a narrow region of compositional space (V $\geq$ 20 at.%, Ni $\geq$ 36 at.%), representing ${\sim}100$ of $27,074$ feasible candidates -- a probability of $P \approx 6.5 \times 10^{-6}$ under random sampling. This dual role -- consistent with the KKT prediction that constrained optima lie on active constraint boundaries -- required feasibility-aware acquisition to access; hard filtering would have excluded this region entirely.

cond-mat.mtrl-sci↗

arXiv2026

The Kernel Manifold: A Geometric Approach to Gaussian Process Model Selection

Gaussian Process (GP) regression is a powerful nonparametric Bayesian framework, but its performance depends critically on the choice of covariance kernel. Selecting an appropriate kernel is therefore central to model quality, yet remains one of the most challenging and computationally expensive steps in probabilistic modeling. We present a Bayesian optimization framework built on kernel-of-kernels geometry, using expected divergence-based distances between GP priors to explore kernel space efficiently. A multidimensional scaling (MDS) embedding of this distance matrix maps a discrete kernel library into a continuous Euclidean manifold, enabling smooth BO. In this formulation, the input space comprises kernel compositions, the objective is the log marginal likelihood, and featurization is given by the MDS coordinates. When the divergence yields a valid metric, the embedding preserves geometry and produces a stable BO landscape. We demonstrate the approach on synthetic benchmarks, real-world time-series datasets, and an additive manufacturing case study predicting melt-pool geometry, achieving superior predictive accuracy and uncertainty calibration relative to baselines including Large Language Model (LLM)-guided search. This framework establishes a reusable probabilistic geometry for kernel search, with direct relevance to GP modeling and deep kernel learning.

cs.LG↗

arXiv2026

Simulated Annealing-based Candidate Optimization for Batch Acquisition Functions

Bayesian Optimization with multi-objective acquisition functions such as q-Expected Hypervolume Improvement (qEHVI) requires efficient candidate optimization to maximize acquisition function values. Traditional approaches rely on continuous optimization methods like Sequential Least Squares Programming (SLSQP) for candidate selection. However, these gradient-based methods can become trapped in local optima, particularly in complex or high-dimensional objective landscapes. This paper presents a simulated annealing-based approach for candidate optimization in batch acquisition functions as an alternative to conventional continuous optimization methods. We evaluate our simulated annealing approach against SLSQP across four benchmark multi-objective optimization problems: ZDT1 (30D, 2 objectives), DTLZ2 (7D, 3 objectives), Kursawe (3D, 2 objectives), and Latent-Aware (4D, 2 objectives). Our results demonstrate that simulated annealing consistently achieves superior hypervolume performance compared to SLSQP in most test functions. The improvement is particularly pronounced for DTLZ2 and Latent-Aware problems, where simulated annealing reaches significantly higher hypervolume values and maintains better convergence characteristics. The histogram analysis of objective space coverage further reveals that simulated annealing explores more diverse and optimal regions of the Pareto front. These findings suggest that metaheuristic optimization approaches like simulated annealing can provide more robust and effective candidate optimization for multi-objective Bayesian optimization, offering a promising alternative to traditional gradient-based methods for batch acquisition function optimization.

cs.LG↗

arXiv2025

Deep Gaussian Process-based Cost-Aware Batch Bayesian Optimization for Complex Materials Design Campaigns

The accelerating pace and expanding scope of materials discovery demand optimization frameworks that efficiently navigate vast, nonlinear design spaces while judiciously allocating limited evaluation resources. We present a cost-aware, batch Bayesian optimization scheme powered by deep Gaussian process (DGP) surrogates and a heterotopic querying strategy. Our DGP surrogate, formed by stacking GP layers, models complex hierarchical relationships among high-dimensional compositional features and captures correlations across multiple target properties, propagating uncertainty through successive layers. We integrate evaluation cost into an upper-confidence-bound acquisition extension, which, together with heterotopic querying, proposes small batches of candidates in parallel, balancing exploration of under-characterized regions with exploitation of high-mean, low-variance predictions across correlated properties. Applied to refractory high-entropy alloys for high-temperature applications, our framework converges to optimal formulations in fewer iterations with cost-aware queries than conventional GP-based BO, highlighting the value of deep, uncertainty-aware, cost-sensitive strategies in materials campaigns.

cond-mat.mtrl-sci↗

arXiv2025

Data Driven Insights into Composition Property Relationships in FCC High Entropy Alloys

Structural High Entropy Alloys (HEAs) are crucial in advancing technology across various sectors, including aerospace, automotive, and defense industries. However, the scarcity of integrated chemistry, process, structure, and property data presents significant challenges for predictive property modeling. Given the vast design space of these alloys, uncovering the underlying patterns is essential yet difficult, requiring advanced methods capable of learning from limited and heterogeneous datasets. This work presents several sensitivity analyses, highlighting key elemental contributions to mechanical behavior, including insights into the compositional factors associated with brittle and fractured responses observed during nanoindentation testing in the BIRDSHOT center NiCoFeCrVMnCuAl system dataset. Several encoder decoder based chemistry property models, carefully tuned through Bayesian multi objective hyperparameter optimization, are evaluated for mapping alloy composition to six mechanical properties. The models achieve competitive or superior performance to conventional regressors across all properties, particularly for yield strength and the UTS/YS ratio, demonstrating their effectiveness in capturing complex composition property relationships.

cond-mat.mtrl-sci↗

arXiv2025

Mapping of Microstructure Transitions during Rapid Alloy Solidification Using Bayesian-Guided Phase-Field Simulations

This study addresses microstructure selection mechanisms in rapid solidification, specifically targeting the transition from cellular/dendritic to planar interface morphologies under conditions relevant to additive manufacturing. We use a phase-field model that quantitatively captures solute trapping, kinetic undercooling, and morphological instabilities across a broad range of growth velocities ($V$) and thermal gradients ($G$), and apply it to a binary Fe-Cr alloy, as a surrogate for 316L stainless steel. By combining high-fidelity phase-field simulations with a Gaussian Process-based Bayesian active learning approach, we efficiently map the microstructure transitions in the multi-dimensional space of composition, growth velocity, and temperature gradient. We compare our PF results to classical theories for rapid solidification. The classical KGT model yields an accurate prediction of the value of $G$ above which the interface is planar for any growth velocity. Microstructures transition from dendrites to cells as the temperature gradient increases close to this value of $G$. We also identify the occurrence of unstable "intermediate" microstructures at the border between dendritic and planar at low $G$, in the absence of banding instability in this Fe-Cr alloy. Our results highlight the capabilities of Bayesian-guided PF approaches in exploring complex microstructural transitions in multidimensional parameters spaces, thereby providing a robust computational tool for designing process parameters to achieve targeted microstructures and properties in rapidly solidified metallic alloys.

cond-mat.mtrl-sci↗

arXiv2025

Accurate and Uncertainty-Aware Multi-Task Prediction of HEA Properties Using Prior-Guided Deep Gaussian Processes

Surrogate modeling techniques have become indispensable in accelerating the discovery and optimization of high-entropy alloys(HEAs), especially when integrating computational predictions with sparse experimental observations. This study systematically evaluates the fitting performance of four prominent surrogate models conventional Gaussian Processes(cGP), Deep Gaussian Processes(DGP), encoder-decoder neural networks for multi-output regression and XGBoost applied to a hybrid dataset of experimental and computational properties in the AlCoCrCuFeMnNiV HEA system. We specifically assess their capabilities in predicting correlated material properties, including yield strength, hardness, modulus, ultimate tensile strength, elongation, and average hardness under dynamic and quasi-static conditions, alongside auxiliary computational properties. The comparison highlights the strengths of hierarchical and deep modeling approaches in handling heteroscedastic, heterotopic, and incomplete data commonly encountered in materials informatics. Our findings illustrate that DGP infused with machine learning-based prior outperform other surrogates by effectively capturing inter-property correlations and input-dependent uncertainty. This enhanced predictive accuracy positions advanced surrogate models as powerful tools for robust and data-efficient materials design.

cs.LG↗

arXiv2025

Accelerated Multi-Objective Alloy Discovery through Efficient Bayesian Methods: Application to the FCC Alloy Space

This study introduces BIRDSHOT, an integrated Bayesian materials discovery framework designed to efficiently explore complex compositional spaces while optimizing multiple material properties. We applied this framework to the CoCrFeNiVAl FCC high entropy alloy (HEA) system, targeting three key performance objectives: ultimate tensile strength/yield strength ratio, hardness, and strain rate sensitivity. The experimental campaign employed an integrated cyber-physical approach that combined vacuum arc melting (VAM) for alloy synthesis with advanced mechanical testing, including tensile and high-strain-rate nanoindentation testing. By incorporating batch Bayesian optimization schemes that allowed the parallel exploration of the alloy space, we completed five iterative design-make-test-learn loops, identifying a non-trivial three-objective Pareto set in a high-dimensional alloy space. Notably, this was achieved by exploring only 0.15% of the feasible design space, representing a significant acceleration in discovery rate relative to traditional methods. This work demonstrates the capability of BIRDSHOT to navigate complex, multi-objective optimization challenges and highlights its potential for broader application in accelerating materials discovery.

cond-mat.mtrl-sci↗

arXiv2024

Hierarchical Gaussian Process-Based Bayesian Optimization for Materials Discovery in High Entropy Alloy Spaces

Bayesian optimization (BO) is a powerful and data-efficient method for iterative materials discovery and design, particularly valuable when prior knowledge is limited, underlying functional relationships are complex or unknown, and the cost of querying the materials space is significant. Traditional BO methodologies typically utilize conventional Gaussian Processes (cGPs) to model the relationships between material inputs and properties, as well as correlations within the input space. However, cGP-BO approaches often fall short in multi-objective optimization scenarios, where they are unable to fully exploit correlations between distinct material properties. Leveraging these correlations can significantly enhance the discovery process, as information about one property can inform and improve predictions about others. This study addresses this limitation by employing advanced kernel structures to capture and model multi-dimensional property correlations through multi-task (MTGPs) or deep Gaussian Processes (DGPs), thus accelerating the discovery process. We demonstrate the effectiveness of MTGP-BO and DGP-BO in rapidly and robustly solving complex materials design challenges that occur within the context of complex multi-objective optimization -- carried out by leveraging the pyiron workflow manager over FCC FeCrNiCoCu high entropy alloy (HEA) spaces, where traditional cGP-BO approaches fail. Furthermore, we highlight how the differential costs associated with querying various material properties can be strategically leveraged to make the materials discovery process more cost-efficient.

cond-mat.mtrl-sci↗

arXiv2024

Illustrating an Effective Workflow for Accelerated Materials Discovery

Algorithmic materials discovery is a multi-disciplinary domain that integrates insights from specialists in alloy design, synthesis, characterization, experimental methodologies, computational modeling, and optimization. Central to this effort is a robust data management system paired with an interactive work platform. This platform should empower users to not only access others data but also integrate their analyses, paving the way for sophisticated data pipelines. To realize this vision, there is a need for an integrative collaboration platform, streamlined data sharing and analysis tools, and efficient communication channels. Such a collaborative mechanism should transcend geographical barriers, facilitating remote interaction and fostering a challenge-response dynamic. In this paper, we present our ongoing efforts in addressing the critical challenges related to an accelerated Materials Discovery Framework as a part of the High-Throughput Materials Discovery for Extreme Conditions Initiative. Our BIRDSHOT Center has successfully harnessed various tools and strategies, including the utilization of cloud-based storage, a standardized sample naming convention, a structured file system, the implementation of sample travelers, a robust sample tracking method, and the incorporation of knowledge graphs for efficient data management. Additionally, we present the development of a data collection platform, reinforcing seamless collaboration among our team members. In summary, this paper provides an illustration and insight into the various elements of an efficient and effective workflow within an accelerated materials discovery framework while highlighting the dynamic and adaptable nature of the data management tools and sharing platforms.

cond-mat.mtrl-sci↗

arXiv2019

Uncertainty Propagation in a Multiscale CALPHAD-Reinforced Elastochemical Phase-field Model

ICME approaches provide decision support for materials design by establishing quantitative process-structure-property relations. Confidence in the decision support, however, must be achieved by establishing uncertainty bounds in ICME model chains. The quantification and propagation of uncertainty in computational materials science, however, remains a rather unexplored aspect of computational materials science approaches. Moreover, traditional uncertainty propagation frameworks tend to be limited in cases with computationally expensive simulations. A rather common and important model chain is that of CALPHAD-based thermodynamic models of phase stability coupled to phase field models for microstructure evolution. Propagation of uncertainty in these cases is challenging not only due to the sheer computational cost of the simulations but also because of the high dimensionality of the input space. In this work, we present a framework for the quantification and propagation of uncertainty in a CALPHAD-based elasto-chemical phase field model. We motivate our work by investigating the microstructure evolution in Mg$_2$(Si$_x$Sn$_{1-x}$) thermoelectric materials. We first carry out a Markov Chain Monte Carlo-based inference of the CALPHAD model parameters for this pseudobinary system and then use advanced sampling schemes to propagate uncertainties across a high-dimensional simulation input space. Through high-throughput phase field simulations we generate 200,000 time series of synthetic microstructures and use machine learning approaches to understand the effects of propagated uncertainties on the microstructure landscape of the system under study. The microstructure dataset has been curated in the Open Phase-field Microstructure Database (OPMD), available at \href{http://microstructures.net}{http://microstructures.net}.

cond-mat.mtrl-sci↗

arXiv2019

Uncertainty Analysis of Microsegregation during Laser Powder Bed Fusion

Quality control in additive manufacturing can be achieved through variation control of the quantity of interest (QoI). We choose in this work the microstructural microsegregation to be our QoI. Microsegregation results from the spatial redistribution of a solute element across the solid-liquid interface that forms during solidification of an alloy melt pool during the laser powder bed fusion process. Since the process as well as the alloy parameters contribute to the statistical variation in microstructural features, uncertainty analysis of the QoI is essential. High-throughput phase-field simulations estimate the solid-liquid interfaces that grow for the melt pool solidification conditions that were estimated from finite element simulations. Microsegregation was determined from the simulated interfaces for different process and alloy parameters. Correlation, regression, and surrogate model analyses were used to quantify the contribution of different sources of uncertainty to the QoI variability. We found negligible contributions of thermal gradient and Gibbs-Thomson coefficient and considerable contributions of solidification velocity, liquid diffusivity, and segregation coefficient on the QoI. Cumulative distribution functions and probability density functions were used to analyze the distribution of the QoI during solidification. Our approach, for the first time, identifies the uncertainty sources and frequency densities of the QoI in the solidification regime relevant to additive manufacturing.

physics.app-ph↗

arXiv2018

Bayesian Uncertainty Quantification and Information Fusion in CALPHAD-based Thermodynamic Modeling

Calculation of phase diagrams is one of the fundamental tools in alloy design---more specifically under the framework of Integrated Computational Materials Engineering. Uncertainty quantification of phase diagrams is the first step required to provide confidence for decision making in property- or performance-based design. As a manner of illustration, a thorough probabilistic assessment of the CALPHAD model parameters is performed against the available data for a Hf-Si binary case study using a Markov Chain Monte Carlo sampling approach. The plausible optimum values and uncertainties of the parameters are thus obtained, which can be propagated to the resulting phase diagram. Using the parameter values obtained from deterministic optimization in a computational thermodynamic assessment tool (in this case Thermo-Calc) as the prior information for the parameter values and ranges in the sampling process is often necessary to achieve a reasonable cost for uncertainty quantification. This brings up the problem of finding an appropriate CALPHAD model with high-level of confidence which is a very hard and costly task that requires considerable expert skill. A Bayesian hypothesis testing based on Bayes' factors is proposed to fulfill the need of model selection in this case, which is applied to compare four recommended models for the Hf-Si system. However, it is demonstrated that information fusion approaches, i.e., Bayesian model averaging and an error correlation-based model fusion, can be used to combine the useful information existing in all the given models rather than just using the best selected model, which may lack some information about the system being modelled.

cond-mat.mtrl-sci↗

arXiv2018

Multi-Information Source Fusion and Optimization to Realize ICME: Application to Dual Phase Materials

Integrated Computational Materials Engineering (ICME) calls for the integration of computational tools into the materials and parts development cycle, while the Materials Genome Initiative (MGI) calls for the acceleration of the materials development cycle through the combination of experiments, simulation, and data. As they stand, both ICME and MGI do not prescribe how to achieve the necessary tool integration or how to efficiently exploit the computational tools, in combination with experiments, to accelerate the development of new materials and materials systems. This paper addresses the first issue by putting forward a framework for the fusion of information that exploits correlations among sources/models and between the sources and `ground truth'. The second issue is addressed through a multi-information source optimization framework that identifies, given current knowledge, the next best information source to query and where in the input space to query it via a novel value-gradient policy. The querying decision takes into account the ability to learn correlations between information sources, the resource cost of querying an information source, and what a query is expected to provide in terms of improvement over the current state. The framework is demonstrated on the optimization of a dual-phase steel to maximize its strength-normalized strain hardening rate. The ground truth is represented by a microstructure-based finite element model while three low fidelity information sources---i.e. reduced order models---based on different homogenization assumptions---isostrain, isostress and isowork---are used to efficiently and optimally query the materials design space.

physics.app-ph↗