Search NASA⌕ Search

SEARCH · Search NASA

Results for “LIGHTWEIGHT”

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 343 records · Page 19
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Parameter extraction approaches for compact modeling of thermoelectric modules

Thermoelectric (TE) cooling has experienced rapid advancements with the foundational understanding of TE materials. TE modules, compact and lightweight devices, have become the prevalent approach for implementing TE technologies. Accurately quantifying TE physical parameters (Seebeck coefficient α, thermal conductivity κ, and thermal resistance ρ) is challenging due to the dynamic temperature changes in operation. Furthermore, extracting lumped property parameters is crucial for designing energy systems using TE modules. Existing research has several limitations, such as lack of comparative analysis between prevalent formulae, reliance on potentially inaccurate vendor-supplied data, disregard for fundamental assumptions, and absence of empirical measurements. Further, this study addresses these gaps by conducting TE material characterization, comparing three existing formulae using vendor datasheets, designing a laboratory test facility for model validation and refinement, and outlining a structured data extraction procedure. The study's novelty lies in multiple key contributions: (1) a detailed comparative analysis of existing formulae for extracting TE property parameter; (2) executing experimental work in a laboratory setting to validate the model and elucidate its limitations; (3) highlighting potential risks; (4) clarifying possible assumptions from both material and engineering perspectives; and (5) considering temperature differential impacts. This comprehensive approach addresses the current research gaps and provides valuable insights into the design and application of TE modules in various energy systems.

36 MATERIALS SCIENCE↗
DOE OSTI2026

A generative machine learning model for designing metal hydrides applied to hydrogen storage

Developing new metal hydrides is a critical step toward efficient hydrogen storage in carbon-neutral energy systems. However, existing materials databases, such as the Materials Project, contain a limited number of well-characterized hydrides, which constrains the discovery of optimal candidates. This work presents a framework that integrates causal discovery with a lightweight generative machine learning model to generate novel metal hydride candidates that may not exist in current databases. Using a dataset of 450 samples (270 training, 90 validation, and 90 testing), the model generates 1000 candidates. After ranking and filtering, six previously unreported chemical formulas and crystal structures are identified, four of which are validated by density functional theory simulations and show strong potential for future experimental investigation. Overall, the proposed framework provides a scalable and time-efficient approach for expanding hydrogen storage datasets and accelerating materials discovery.

generative model↗
DOE OSTI2021

Shear assisted processing and extrusion of enhanced strength aluminum alloy tubing

Hollow extrusions are used in many industries to make strong, lightweight, and affordable structures. Shear Assisted Processing and Extrusion (ShAPE) enables the extrusion of many alloys, often with noteworthy properties. In this investigation, ShAPE was used to extrude 1 and 2 mm wall tubes of aluminum alloy 6063 measuring 12 mm in diameter at extrusion speeds up to 3.8 m/min. This speed is 10 and 15 times faster than has been reported for any alloy using ShAPE and friction stir back extrusion, respectively. Increasing the extrusion speed from 0.7 to 3.8 m/s resulted in using 68% less process energy at steady state. Microstructural analysis and tensile testing were performed on as-extruded tubes and tubes with a direct T5 heat treatment. Grain refinement from 65 µm to below 20 µm was observed. Growth of macroscale Mg 2 Si strengthening precipitates was not observed. ShAPE processing resulted in recrystallization, and extrusion speed was shown to influence final extrudate texture. Here, transmission electron microscopy analysis revealed that the as-extruded microstructure was free of nanoscale ß" strengthening precipitates. Nanoscale ß" was prominent in direct-aged (T5) ShAPE tubes, similar to a slightly over-aged conventional T6 microstructure. As-extruded tubes had ultimate tensile strengths on par with conventional T5 extrusions and approximately double the total elongation. Tubes that underwent a T5 heat treatment had yield and ultimate strengths averaging 198 and 234 MPa, respectively, with an average total elongation of 11%. The tensile test results are comparable to a conventional T6 heat treatment, but without the solutionization heat treatment step. Mechanical properties either remained flat or improved with increased extrusion speed, which suggests that ShAPE can be scaled up to an industrial process to create energy- and economically efficient high-strength extrusions.

36 MATERIALS SCIENCE↗
DOE OSTI2023

KEBLM: Knowledge-Enhanced Biomedical Language Models

Pretrained language models (PLMs) have demonstrated strong performance on many natural language processing (NLP) tasks. Despite their great success, these PLMs are typically pretrained only on unstructured free texts without leveraging existing structured knowledge bases that are readily available for many domains, especially scientific domains. As a result, these PLMs may not achieve satisfactory performance on knowledge-intensive tasks such as biomedical NLP. Comprehending a complex biomedical document without domain-specific knowledge is challenging, even for humans. Inspired by this observation, we propose a general framework for incorporating various types of domain knowledge from multiple sources into biomedical PLMs. We encode domain knowledge using lightweight adapter modules, bottleneck feed-forward networks that are inserted into different locations of a backbone PLM. For each knowledge source of interest, we pretrain an adapter module to capture the knowledge in a self-supervised way. We design a wide range of self-supervised objectives to accommodate diverse types of knowledge, ranging from entity relations to description sentences. Once a set of pretrained adapters is available, we employ fusion layers to combine the knowledge encoded within these adapters for downstream tasks. Each fusion layer is a parameterized mixer of the available trained adapters that can identify and activate the most useful adapters for a given input. Our method diverges from prior work by including a knowledge consolidation phase, during which we teach the fusion layers to effectively combine knowledge from both the original PLM and newly-acquired external knowledge using a large collection of unannotated texts. After the consolidation phase, the complete knowledge-enhanced model can be fine-tuned for any downstream task of interest to achieve optimal performance. Extensive experiments on many biomedical NLP datasets show that our proposed framework consistently improves the performance of the underlying PLMs on various downstream tasks such as natural language inference, question answering, and entity linking. These results demonstrate the benefits of using multiple sources of external knowledge to enhance PLMs and the effectiveness of the framework for incorporating knowledge into PLMs. Finally, while primarily focused on the biomedical domain in this work, our framework is highly adaptable and can be easily applied to other domains, such as the bioenergy sector.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
DOE OSTI2022

Building energy efficiency and load flexibility optimization using phase change materials under futuristic grid scenario

This work investigates the energy savings and load flexibility capacity of phase change material (PCM) integrated into building envelope under a future energy generation scenario, where 80% of the energy load comes from renewable sources. Using community-scale modeling, we first determine the net thermal storage required to fully manage the demand variability at a utility grid, and then using whole-building simulations we optimize operating parameters - such as PCM thickness, latent heat, interior setpoint profile, PCM distribution, and heat transfer coefficient - to determine the optimal conditions required for maximum energy reduction and load flexibility without compromising occupants' thermal comfort. The optimal PCM-integrated envelope proposed in this study can provide annual load flexibility up to 33.6% and annual energy savings up to 10.8% in a lightweight residential building located in Baltimore, MD. This study is relevant given the increasing contributions of renewable energy in the total energy generation mix, leading to a significant time-imbalance between peak energy demand and peak energy production. While thermal energy storage using PCM is a recognized technique, there is no known prior study dedicated to examining the thermal performance of PCM-integrated envelope under future energy generation scenarios. This study bridges the research gap by investigating a practical way to implement PCMs in the buildings and maximize the energy efficiency as well as load flexibility related benefits.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2024

Life cycle assessment of coir fiber-reinforced composites for automotive applications

Past decades have seen an increasing prevalence of natural fiber-reinforced composites (NFRCs) due to growing conscientiousness around sustainability and a push towards vehicle lightweighting. The environmentally friendly and sustainable claims of NFRCs need to be validated due to their large variability and variety, particularly where material substitutions are concerned, such as in substituting glass fiber with natural fiber. Additionally, the objective of this work is to determine the cumulative energy demand (CED) and greenhouse gas emissions (GHG) associated with an automotive part (of volume 0.001 m3) made from 40 wt% coir fiber-reinforced polypropylene (PP) and compared with a similar part made from 40 wt% glass fiber reinforced PP. SimaPro v. 9.0.0.49 was used for the analysis, whereas inventory data were collected from databases, such as Ecoinvent 3, Transportation Energy Databook, Greet model 2022, and published papers. The results showed that CED and GHG associated with the coir fiber-reinforced composite part were lower than the glass fiber-reinforced composite part for both cradle-to-gate (~34–40%) and cradle-to-grave (excluding end-of-life) (~24%) analysis.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Active operator learning with predictive uncertainty quantification for partial differential equations

With the increased prevalence of neural operators being used to provide rapid solutions to partial differential equations (PDEs), understanding the accuracy of model predictions and the associated error levels is necessary for deploying reliable surrogate models in scientific applications. Existing uncertainty quantification (UQ) frameworks employ ensembles or Bayesian methods, which can incur substantial computational costs during both training and inference. Here, we propose a lightweight predictive UQ method tailored for Deep operator networks (DeepONets) that also generalizes to other operator networks. Numerical experiments on linear and nonlinear PDEs demonstrate that the framework’s uncertainty estimates are unbiased and provide accurate out-of-distribution uncertainty predictions with a sufficiently large training dataset. Our framework provides fast inference and uncertainty estimates that can efficiently drive outer-loop analyses that would be prohibitively expensive with conventional solvers. We demonstrate how predictive uncertainties can be used in the context of Bayesian optimization and active learning problems to yield improvements in accuracy and data-efficiency for outer-loop optimization procedures. In the active learning setup, we extend the framework to Fourier Neural Operators (FNO) and describe a generalized method for other operator networks. To enable real-time deployment, we introduce an inference strategy based on precomputed trunk outputs and a sparse placement matrix, reducing evaluation time by more than a factor of five. Our method provides a practical route to uncertainty-aware operator learning in time-sensitive settings.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Synergistic cellulase–xylanase formulations for enhanced dewatering and fiber bonding toward energy-efficient and sustainable paper and packaging production

A mechanistic understanding of the synergistic effects of enzymes on cellulosic fibers dewatering and fiber bonding is essential for advancing energy-efficiency and lightweight production of paper and packaging materials. This study investigates the impact of varying cellulase and xylanase formulations on equilibrium moisture content (EMC) after pressing and tensile strength of cellulosic fiber webs using a factorial experimental design. Nine custom enzyme formulations were evaluated at controlled dosages ranging from ∼20 to 76 FPU/mL cellulase and ∼500–1130 IU/mL xylanase. Response surface modeling revealed a significant synergy, particularly at 20–40 FPU/mL cellulase combined with ≥ 1000 IU/mL xylanase. Under these conditions, EMC decreased by up to 3.6% compared with the untreated refined control, while tensile index gains of up to 16% were statistically significant for optimized blends (p < 0.05). The interaction between cellulase and xylanase was also significant for the tensile response (p = 0.010). Protein efficiency analysis showed that optimized formulations containing 30–40% less protein outperformed the commercial benchmark. The nonlinear synergy between cellulase and xylanase is attributed to their complementary substrate specificities. Endoglucanase- and β‑glucosidase‑rich cellulases hydrolyze internal β‑1,4‑glycosidic bonds in amorphous cellulose, loosening fiber walls and increasing flexibility, while xylanases target hemicellulose, primarily xylan-rich domains, enhancing porosity and improve cellulase accessibility. Tailoring enzyme formulations at low loadings overcomes traditional trade-offs between strength and dewatering, enabling cost-effective, energy-efficient, low-carbon solutions for sustainable packaging and hygiene products.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Mitigation of distortion of Al/steel part under simulated paint baking condition: Experiment and numerical model studies

Multi-material joining of lightweight structures is essential to reduce vehicle weight for more energy savings and less greenhouse gas emission. However, mismatch of thermal expansion coefficient for dissimilar materials during the paint baking process can induce part distortion and joint failure for adhesive bonding. Here, in the present work, a thermomechanical model based on contact mechanics and large deformation theory was developed for dissimilar high-strength Al alloy and steel components to study the distortion mechanism and influential factors of the residual gap. The established model was used to optimize joint conditions, such as pitch distance and part geometry. When a weld pitch is shorter than 100 mm, the maximum gap between Al and steel part can be greatly reduced to 0.1 mm, and the local stress and plastic strain around the joint during the oven heating and cooling cycle are also substantially reduced compared with the long pitch case (900 mm). The numerical modeling results revealed that a comparable bending stiffness ratio between the steel and Al cross sections is critical to the minimization of gap and distortion under paint baking condition. Digital image correlation technique was used to measure the overall part distortion and local strain distribution that were used to validate the model prediction. Weld bonding (adhesive bonding with friction bit joining) process was successfully employed to join Al to steel component without gap opening in adhesive after the paint baking and cooling.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Simulation-driven design optimization of reaction injection molding (RIM) process for polydicyclopentadiene (pDCPD): Minimizing cycle time, defects, and warpage

Replacing metal components in trucks, trailers, and buses with lightweight polymer composites is challenging due to high temperatures and complex manufacturing. The Reaction Injection Molding (RIM) process using Dicyclopentadiene (DCPD) resin offers a solution by producing robust parts with excellent stiffness, impact strength, and resistance properties. Simulations are essential for optimizing this process, predicting defects, and improving quality. However, most commercial software is tailored for thermoplastics, requiring thermoset users to generate their own datasets. In this study, a material data card for DCPD was developed to perform RIM simulations. Design of Experiments (DOE) was used to identify key factors affecting filling, curing, and warpage, aiming to minimize cycle time and defects. The simulations explored varying injection gate parameters (size, location, number) and process conditions (mold/resin temperature, injection/curing pressure). Results showed that gate design significantly impacts filling behavior and defects. A single central gate provided balanced flow with fewer defects, while two corner gates led to more defects. Additionally, lower injection pressure increased filling time, while higher mold temperature accelerated curing but led to more warpage. In conclusion, this optimization framework aims to enhance DCPD part performance and promote sustainable manufacturing by reducing waste and energy consumption.

42 ENGINEERING↗
DOE OSTI2026

Multi-material additive manufacturing of aluminum 6061-T6 alloy with stainless steel 304: Suppression of intermetallic compounds and interface growth mechanism

Fabricating integrated aluminum-steel structures is difficult because of metallurgical incompatibilities that promote intermetallic compound (IMC) formation at their interfaces. This work introduces a practical route for additively manufacturing 6061-T6 aluminum alloy (AA6061-T6) directly onto 304 stainless steel (SS304) using a high shear strain rate-assisted interfacial bonding mechanism. Through the friction extrusion deposition-based additive manufacturing method, both single- and multi-layer deposits of AA6061-T6 were successfully fabricated on SS304, yielding a robust hybrid multi-material structure. Comprehensive analyses of deposition quality, interface porosity, and bonding performance showed that the interface achieved a tensile strength exceeding 154 MPa under quasi-static loading. Microstructural observations revealed that the deposited aluminum layers experienced severe plastic deformation, resulting in pronounced grain refinement. Importantly, the interfacial zone was found to be free of brittle IMCs, instead containing a thin amorphous layer that evolved through a non-linear reaction-zone growth law. This solid-state additive strategy establishes a promising pathway for lightweight structural systems, nuclear energy component cladding, and multifunctional engineering components, redefining how dissimilar metals can be integrated for advanced applications.

Aluminum alloy↗
DOE OSTI2024

Machine learning for the redox potential prediction of molecules in organic redox flow battery

Here, organic redox flow batteries (ORFB) are recognized as an innovative technology for the large-scale storage of renewable energy. The redox potential of organic redox-active molecules plays a vital role in their performance. Advanced screening techniques like high-throughput experiment and machine learning (ML) have significantly enhanced organic material performance and transformed the field of ORFB. However, the scarcity of experimental data poses a considerable challenge for ML model development in this domain. In our study, we developed lightweight graph-based Gaussian process regression (GPR) models with GPU-accelerated marginalized graph kernel and hybrid kernel to predict the redox potentials of organic redox-active molecules for ORFBs, specifically focusing on small datasets. To evaluate model accuracy, we created a new experimental database of organic redox-active molecules by the data from hundreds of published papers and assembled previous computational datasets. We also considered some key parameters, such as pH conditions and solvent type, to assess their impact on redox potential prediction. Our GPR model predicted redox potentials with high accuracy across all datasets using minimal training data. The study provides powerful tools for molecule screening and design and delivers valuable guidance on designing training datasets for costly experiments.

25 ENERGY STORAGE↗
DOE OSTI2022

On the high-temperature stability of the Al 8 Cu 3 Ce intermetallic in an additively manufactured Al-Cu-Ce-Zr alloy

High-temperature resistant eutectic Al alloys are crucial materials for lightweight and energy efficient design in the automotive and aviation industries. Additive manufacturing offers a pathway to refine eutectic microstructures and develop novel alloys with superior high-temperature strength. High-volume fraction intermetallic Al-Cu-Ce alloys have been developed to deliver high-temperature strength in combination with reduced hot-tearing susceptibility. Zr is added to provide additional strengthening via nanoscale Al 3 Zr precipitation, and to stabilize and avoid coarsening of the Al 8 Cu 3 Ce phase. However, the detailed interaction between Zr and Al 8 Cu 3 Ce remains unexplored. In this work, we show with synchrotron X-ray diffraction that laser powder bed fusion fabricated Al-Cu-Ce and Al-Cu-Ce-Zr alloys contain predominantly the Al 8 Cu 3 Ce intermetallic in the as-fabricated condition. Heat treatment of the Al-Cu-Ce alloy results in the Al 8 Cu 3 Ce → Al 8 Cu 4 Ce phase transformation. In the Al-Cu-Ce-Zr alloy, minor fractions of (Al,Cu,Si) 4 Ce and Al 2 Cu-θ are found in the as-fabricated condition, while Al 8 Cu 3 Ce remains stable during heat treatment. Atom probe microscopy quantifies intermetallic stoichiometries and reveals how Zr is enriched at the Al-matrix/Al 8 Cu 3 Ce interface acting as a diffusion barrier against solute exchange. Calibrated thermodynamic modeling underpins this as a kinetic effect. Here, a qualitative microstructural model summarizes, how Zr stabilizes Al 8 Cu 3 Ce against phase transformations and coarsening.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Achieving superior mechanical properties: Tailoring multicomponent microstructure in AISI 9254 spring steel through a two-stage Q&P process and nanoscale carbide integration

In the pursuit of lightweight, durable steel, we have successfully developed a multicomponent structure in AISI 9254 spring steel using a two-stage quenching and partitioning (Q&P) process. The primary objective of this process was to engineer an optimized microstructure consisting of nanobainite, martensite, and nano-carbides. Utilizing the insights gained from the results of advanced techniques such as X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and atom probe tomography (APT) performed on the as-received AISI 9254 spring steel, we refined the quenching and partitioning (Q&P) path, leading to the successful establishment of a bainitic transformation for superior mechanical properties. Our tensile tests revealed a high yield strength (≈ 1600 ± 25 MPa) and ultimate tensile strength (≈ 1850 ± 50 MPa), along with considerable elongation (≈ 11.15 ± 0.25%). We also identified that pre-formed martensite lath defects and high silicon content play crucial roles during the Q&P process, preventing carbide coalescence and increasing strain-hardening capacity. Finally, this study demonstrates the potential of a Q&P process to generate high-strength, ductile steel for automotive and aerospace applications.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Elucidating microstructural evolution and hardness variation across friction self-piercing riveted Al-7055 using synchrotron X-ray scattering and advanced microscopy techniques

Friction self-piercing riveting (FSPR) is a unique hybrid joining technique that combines the advantages of mechanical interlocking, frictional heat, and solid-state joining (if metallurgically compatible) to produce crack free joints in high strength and/or low-ductility alloys at room temperature. Here, in the current study, Al-7055 sheets were joined using FSPR for lightweight automotive applications and significant microhardness variations were observed across the joint cross-section. A detailed microstructural characterization at multiple length scales was carried out using advanced electron microscopy and X-ray scattering techniques to provide a fundamental understanding of the process-structure-property relationships. The relative contributions of microstructural characteristics at various length scales (i.e., grain size, dislocation density, solute concentration, precipitate nature) to strengthening were estimated using existent formulations (i.e., Hall-Petch, Taylor, precipitate bypass/shear equations) and correlated to the observed microhardness values across different regions. Small-angle X-ray scattering and scanning transmission electron microscopy revealed significant changes in the size and volume fraction of precipitate species, i.e., GP-I Zones, η′, and Mg/Zn solute co-clusters, depending on the process region. It was observed that the dissolution of the small η′/GP-I zones (T ∼ 150–200 °C) in the heat-affected zone were the key reason for the hardness drop. Further, it was shown that solid-solution, dislocation, grain size and solute co-cluster strengthening played a key role in the thermo-mechanically affected zone and grain-refined zone (GRZ). Finally, these observations were leveraged along with the Zener-Holloman relationship and grain size in the GRZ to estimate the peak joining temperature of the GRZ (∼ 350 °C) near the steel rivet.

aluminum 7xxx alloy↗
DOE OSTI2025

Property optimized energy absorber for automotive bumpers utilizing multi-material and structural design strategies

This study proposes a novel design for automotive bumper using optimized lattice structures and multi-materials to balance low-speed collision and high-speed pedestrian impact performance. Different blends of 20 % carbon fiber-reinforced acrylonitrile butadiene styrene with thermoplastic polyurethane were used to tailor material properties. The energy absorber features lattice structures with customized mechanical responses, created by varying the incline angle θ from 0 to 180°. We conducted 576 finite element simulations on a half-scale model to optimize energy absorption and stiffness, leading to 66 optimized designs that met both low-speed and high-speed impact criteria. Two sub-scale optimized energy absorbers with different peak forces—both meeting low-speed impact requirements—were 3D printed and validated through drop-weight testing. The one with lower peak stress demonstrated a more compliant response, exhibiting approximately 90 % lower initial peak force and an increase in energy absorption of around 33 % (from 24 J to 32 J). Compared to the baseline triangular lattice, the optimized absorber increased energy absorption by 68 % from (19 J to 32 J) and reduced peak stress by 70 %. It also showed near-complete recovery with minimal fractures, making it suitable for repeated use. This design improves safety while offering a lightweight, durable, and cost-effective bumper system.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Improving adhesive bonding of short carbon fiber thermoplastic composites to aluminum alloys with a hybrid laser-plasma surface modification strategy

This study investigates hybrid laser–plasma surface modification strategies for metal–CFRTP (carbon-fiber-reinforced thermoplastic polymer) dissimilar joints to improve their bonding performance, in contrast to existing literature that mostly focuses on either plasma or laser treatment alone. By conducting double cantilever beam (DCB) tests on adhesively-bonded AA5052 and CFRPA66 (carbon-fiber-reinforced polyamide 66) joints, as an example of metal–CFRTP joints, it was found that laser engraving on the metal surface combined with plasma treatment on the CFRTP surface significantly improved the specific fracture energy of the joint by 187% and 31% compared to as-received and plasma-treated-only joints, respectively. However, the hybrid treatment of laser engraving and plasma on the investigated CFRTP surface did not improve the bonding performance of the joints. The underlying mechanisms related to hybrid laser-plasma surface modification strategies were further investigated by examining the surface and cross-sectional morphologies after DCB testing using microscopy. Computational modeling was performed to elucidate the interaction between grooves on the metal substrate and the CFRTP–adhesive interfacial bonding in metal–CFRTP joints. This study provides new insights into developing surface modification methods for achieving strong metal–CFRTP adhesive joints, aimed at lightweighting structural components in automotive, aerospace, and other applications.

Adhesive bonding↗
DOE OSTI2025

Multi-objective automatic discovery of optimized metamaterials for varying velocity impact protection

Mechanical metamaterials have demonstrated exceptional impact performance while remaining lightweight. Impact resistance has traditionally been investigated using quasi-static simulations, often with the assumption that performance will translate to high-velocity impact scenarios. However, critical crash protection parameters—such as peak stress and absorbed energy—are highly sensitive to impact velocity, leading to inconsistent performance under dynamic loading. To address this, we introduce a strain-rate-aware, active deep learning framework that enables multi-objective optimization of impact protection metrics across a wide range of impact velocities. Our framework captures the strain-rate sensitivity of architected lattices by learning to control spatial gradation in cellular metamaterials, resulting in over 200 % enhancement in impact protection relative to state-of-the-art designs such as Voronoi and re-entrant lattices. We demonstrate its practical utility by designing next-generation lattice structures for automotive bumper systems that satisfy multiple, velocity-specific safety criteria—capabilities beyond those of conventional designs. More than just a predictive tool, this framework marks the first step towards enabling adaptable impact-resistant structures across dynamic regimes.

Deep learning↗