Daniel Cooper

Accurate, parametric cradle-to-gate Life Cycle Assessment (LCA) models of production systems are essential for environmentally motivated process comparison and optimization; however, commercially-available Life Cycle Inventory (LCI) databases are often too inaccurate for confident decision-making, while more accurate process and/or site-specific data-driven modeling can require prohibitively expensive experimental campaigns. In this work, a multi-fidelity transfer learning (MFTL) framework is developed for cradle-to-gate LCA that combines a low-fidelity reduced-order model, embedding basic physics or experience as a principled head start, with a high-fidelity adaptive boosting mechanism based on limited experimental data. The framework accelerates learning and helps prevent overfitting by augmenting the traditional residual-based loss with a correlation-aware term. The framework is demonstrated by modeling energy-related inventory data in two manufacturing processes: roller power consumption in aluminum cold rolling and energy use in polymer extrusion-based additive manufacturing (AM). The correlation-aware MFTL framework achieved R 2 validation scores of 0.99 and 0.98 using just 60% and 12% of the data collected from experiments on the rolling and AM case studies, respectively. This predictive accuracy outperformed Traditional MFTL ( R 2 validation scores of 0.88 and 0.83) and purely data-driven approaches such as a neural network and a support vector regressor ( R 2 validation scores of 0.85 and 0.84) when trained on the same amount of data. Overall, the framework enables accelerated, accurate, parametric modeling of LCI data, supporting the development of sustainable manufacturing systems.

Abstract Material flow analyses (MFAs) are powerful tools for highlighting resource efficiency opportunities in supply chains. MFAs are often represented as directed graphs, with nodes denoting processes and edges representing mass flows. However, network structure uncertainty—uncertainty in the presence or absence of flows between nodes—is common and can compromise flow predictions. While collection of more MFA data can reduce network structure uncertainty, an intelligent data acquisition strategy is crucial to optimize the resources (person‐hours and money spent on collecting and purchasing data) invested in constructing an MFA. In this study, we apply Bayesian optimal experimental design, based on the Kullback–Leibler divergence, to efficiently target high‐utility MFA data—data that minimizes network structure uncertainty. We introduce a new method with reduced bias for estimating expected utility, demonstrating its superior accuracy over traditional approaches. We illustrate these advances with a case study on the US steel sector MFA, where the expected utility of collecting specific single pieces of steel mass flow data aligns with the actual reduction in network structure uncertainty achieved by collecting said data from the United States Geological Survey and the World Steel Association. The results highlight that the optimal MFA data to collect depends on the total amount of data being gathered, making it sensitive to the scale of the data collection effort. Overall, our methods support intelligent data acquisition strategies, accelerating uncertainty reduction in MFAs and enhancing their utility for impact quantification and informed decision‐making.

Abstract Single-use consumer PLA products currently have two end of life avenues: industrial composting or landfill. Additive manufacturing holds the potential to create demand for recycled post-consumer PLA if the properties, cost, and performance of reclaimed PLA could match or rival its virgin counterpart. The paper investigates the viability of using post-consumer talc-enhanced polylactic acid (tPLA) for fused filament fabrication (FFF) 3D printing, which could both extend the life cycle of bioplastics and provide a sustainable alternative to virgin PLA. This research involves system modeling, material processing, and mechanical testing of printed specimens made from recycled tPLA derived from compostable cutlery. A comparison between recycled tPLA, uncontaminated tPLA, and virgin PLA was conducted, focusing on tensile and impact properties. Results indicate that while recycled tPLA can approach the mechanical properties of virgin PLA in specific orientations, print quality issues such as nozzle clogs and poor adherence remain prevalent challenges. Energy consumption analysis shows a 56% reduction compared to virgin PLA production, highlighting environmental benefits. Limitations identified include potential contaminants affecting print quality and mechanical performance. Future work should focus on optimizing decontamination processes, testing across various 3D printing platforms, and varying printing parameters to mitigate current drawbacks. This study provides foundational insights into making postconsumer tPLA a viable and sustainable material for additive manufacturing applications.

Abstract The supply chains of extruded aluminum are materially inefficient, with up to two-fifths of the billet being scrapped before the profile is incorporated into a final product. A significant source of process scrap arises from removing the tongue-shaped transverse weld—also known as the front-end defect or charge weld—that is formed between the consecutive billets being extruded, primarily because of concerns over weld integrity. Optimizing process settings and die geometry can reduce the transverse weld length—and thus the amount of scrapped material—but only by approximately 15%. We investigate a novel methodology for significant scrap reduction, where an initially profiled interface—rather than a flat one—between consecutively extruded billets compensates for the differential velocities of material across the billet cross-section as it moves through the die ports, resulting in shorter welds. This profiled interface is created using profiled billets that fit into a dummy block shaped with the inverse of the billet profile. We present a design process to define the shape of the profiled dummy block and billet. For a given part, we first determine the ideal shape by obtaining the velocity field from finite element simulations of the conventional extrusion process, assuming perfectly rigid tooling and no constraints on the creation of profiled tooling or billets. Next, we rationalize this shape by applying stress and deflection limits to the dummy block, ensuring it avoids plastic deformation and interference with the container wall. Additionally, we consider ductile damage limits for the billet to prevent cracking during a pre-extrusion hot forging stage, which is one method of generating profiled billets. The design process is applied to four profiles of increasing complexity: solid round and rectangular bars, a square-tube hollow, and a complex multi-hollow profile. Extrusion and forging trials using custom-built tooling are conducted to validate the design process. The experimental case studies demonstrate that profiled dummy blocks and billets can achieve weld length reductions of over 50% and that the same tooling can offer scrap savings across a range of similar extruded shapes. In the tests, a profiled dummy block with an air escape vent showed zero-to-negligible plastic deformation and neither air entrapment nor clogging of the vent during extrusion, while a conventional billet was hot-forged to produce profiled ends without cracking or deforming the forging tools. Overall, this study highlights that profiled billet extrusion is a promising technology for significantly reducing scrap from transverse weld removal in aluminum extrusions.

The U.S. manufacturing sector contends with an aging workforce, recruitment and retention challenges, and a strengthened national push for domestic production, leading to renewed calls for broadening workforce participation. Existing literature on broader engineering education suggests that both the focus of engineering curricula and student perceptions of a discipline’s career path likely impact workforce participation rates. To assess the current state of perceptions of manufacturing careers by undergraduate and graduate mechanical engineering students at an R1 public university, the authors of this paper collected and analyzed primary survey data in the 2023–2024 academic year. Perceptual differences existed between academic levels and between students with and without industry internship experience. Conversely, the findings revealed minimal differences in the perceived appeal and importance of manufacturing competencies and skills across racial and gender identities. Additionally, university courses and industry internships were identified as the primary factors influencing students’ perceptions of manufacturing careers. We propose that early exposure in mechanical engineering courses to real-world experiences that encompass a variety of manufacturing skillsets could foster more accurate career perceptions and potentially enhance participation rates.

Abstract Material flow analyses (MFAs) provide insight into supply chain‐level opportunities for resource efficiency. MFAs can be represented as networks with nodes that represent materials, processes, sectors, or locations. MFA network structure uncertainty (i.e., the existence or absence of flows between nodes) is pervasive and can undermine the reliability of the flow predictions. This article investigates MFA network structure uncertainty by proposing candidate node‐and‐flow structures and using Bayesian model selection to identify the most suitable structures and Bayesian model averaging to quantify the parametric mass flow uncertainty. The results of this holistic approach to MFA uncertainty are used in conjunction with the input‐output (I/O) method to make risk‐informed resource efficiency recommendations. These techniques are demonstrated using a case study on the US steel sector where 16 candidate structures are considered. The model selection highlights two networks as most probable based on data collected from the United States Geological Survey and the World Steel Association. Using the I/O method, we then show that the construction sector accounts for the greatest mean share of domestic US steel industry emissions while the automotive and steel products sectors have the highest mean emissions per unit of steel used in the end‐use sectors. The uncertainty in the results is used to analyze which end‐use sector should be the focus of demand reduction efforts under different appetites for risk. This article's methods generate holistic and transparent MFA uncertainty that accounts for structural uncertainty, enabling decisions whose outcomes are more robust to the uncertainty.

There is a growing interest in hydrogen for decarbonizing hard-to-abate industries. However, determining which industries to target, the scale of the opportunity, and how to meet the hydrogen demand out to 2050 is complex and highly uncertain. The most significant decarbonization opportunity identified in this review is in the refining and chemicals industries, where annual emissions could reduce by up to 24% by 2050 from 2021 levels if emissions-intensive grey hydrogen is replaced with clean hydrogen. New (clean) hydrogen applications include replacements for carbon-based reductants in steelmaking (≤18% steelmaking emissions reduction by 2050) and fuel for high-temperature heat in cement, aluminum , and glassmaking, with annual sectoral emissions reductions by 2050 of up to 23%, 3%, and 32% respectively. Hydrogen technologies have high readiness levels and face modest technical barriers in burner and furnace design. The primary challenge lies in reducing clean hydrogen production and delivery costs to $0.4–0.7/kgH 2 to compete with natural gas and scale its production from <1% of all U.S. hydrogen production today. The literature presents diverse U.S. industry clean hydrogen demand predictions (4–22 Mt/year by 2050) due to conflicting projections of industrial output, some incompatible with decarbonization goals; e.g., growth in gasoline production. After reconciling literature on hydrogen technology readiness, alternative decarbonization strategies, and U.S. climate targets, we estimate 2050 industrial clean hydrogen demand at 3.8–14.9 Mt/year, saving 28–133 MtCO 2eq (1.5-7.0% of current U.S. industry emissions). Green hydrogen production will require up to 682 TWh of low-carbon electricity, equivalent to 90% of current renewable generation .

Manufacturing industries are increasingly focused on achieving sustainability targets, which has driven the development of environmental impact models often based on life cycle assessment (LCA) methods and databases. However, these databases tend to be too generic to ensure accurate modelling (e.g., using global or regional average impact values per unit of mass processed). To improve accuracy, companies can generate customized data inventories through experiments or simulations, but these approaches are typically costly, time-consuming, and may disrupt daily operations. This article introduces a partial physics-based, multi-fidelity machine learning approach to generate low-cost, environmental impact models tailored to specific manufacturing systems. The framework uses reduced-order, low-fidelity, physics-based models to capture the process dynamics, followed by transfer learning with small volumes of high-fidelity (e.g., experimental) data. This allows for accurate gate-to-gate environmental impact predictions without the need for extensive experimental campaigns. The framework is demonstrated on a lab-scale metal rolling mill for predicting power consumption in gate-to-gate assessments. A simple slab analysis metal forming model trains the base learner, and adaptive boosting is used for transfer learning on experimental data. The framework achieved superior performance, requiring 13% less experimental data than a standalone machine learning model of the same accuracy trained solely on experimental data. This approach may offer a cost-effective solution for generating accurate predictive models in scenarios where data collection is challenging, either due to rigid use of standard process settings or data collection cost and time constraints.

High-quality recycling of aluminum from end-of-life planes into new aerospace-grade aluminum is starting to be researched by manufacturing and recycling companies. The intention of this work is to lay the groundwork for U.S. centric studies on aluminum end-of-life plane recycling by tracing the current lifespan of aerospace aluminum alloys from production to end-of-life to uncover and articulate opportunities and barriers to closed-loop recycling of end-of-life airplane aluminum in the United States. A dynamic material flow analysis (DMFA) of the aluminum from Boeing airplanes within the United States from 2024-2035 is performed to understand the material and financial scale of the opportunity. The current processes, challenges, barriers, and potential opportunities are informed by stakeholder and industry expert interviews. From the processes and analysis of barriers, a lifecycle map is created to visualize the current processes and problem junctions. Analysis of alloy makeup is used to identify why current scrap streams are unusable for high-value recycling. Recommendations on stakeholder actions and investments for increased high-quality recycling of aerospace alloys are included based on the barriers and challenges identified.

Increasing wind power is a key pillar of the pathway towards 100% clean electricity by 2035 in the US. To reach this goal, cumulative generation capacity will likely have to increase by 400-800% from 2023 to 2035. One constraint on this increase is the number of wind turbine blade molds in commission, which limits the turbine blade production rate. These molds are built using a resin-fiberglass composite with a lead time of 26-40 weeks. The molds are then typically discarded not due to physical degradation but due to changes in the blade geometry requirements for new wind sites. To minimize blade production lead time, a new modular mold is proposed that can be reconfigured to produce a range of different blade geometries. A cradle-to-gate manufacturing cost, energy and lead time model was constructed to compare a durable additively manufactured modular aluminum mold to the conventionally produced fiberglass composite mold. Modular molds are estimated to be quicker, but more expensive and energy-intensive to build than conventional molds. A US level wind turbine blade fleet analysis suggests modular molds could lead to around 25% fewer molds needing to be produced cumulatively by 2035 to meet the clean electricity target.

Material flow analyses (MFAs) are powerful tools for identifying and analyzing energy and material efficiency (resource efficiency) opportunities across a supply chain. MFAs are typically represented as directed graphs with key parameters including the mass of the material flows entering the system and the allocation of materials flowing through one node (typically representing a process or location) to other nodes in the system. Parametric uncertainty can hamper the credibility and usability of MFA results. Uncertainty may be reduced by collecting more data; however, an intelligent data acquisition strategy is needed given the limited resources available for completing a given MFA. In this article, we apply Bayesian optimal experimental design (BOED) derived from the Kullback-Leibler divergence to target the collection of high value data, which is then fed into a Bayesian framework to effectively reduce the MFA parametric uncertainty. The methodology is demonstrated using a case study on the 2012 U.S. steel sector. Bayesian inference is then used to validate the BOED results with data collected from the United States Geological Survey and the World Steel Association. This article’s methods allow efficient data collection to rapidly create MFAs with reduced and quantified parametric uncertainty, aiding decision makers in their efforts to pursue resource efficiency.

Additive manufacturing (AM) can reduce the environmental impacts of the materials in a manufactured component by lightweighting parts and reducing process scrap compared to subtractive manufacturing techniques such as machining. However, whether these savings materialize depends on the yield and energy and emissions intensity of powder manufacturing and utilization. Here we model the cumulative energy demand (CED) associated with the powder used to make a metal AM part. We use life cycle material flow analysis, combined with process energy intensities obtained from the literature. We perform Monte Carlo analysis to quantify the uncertainty in the results and sensitivity analysis to identify critical system parameters that drive CED for powder production and use. We then evaluate which critical system parameters have the most influence on the CED of AM powder that both producers and consumers can influence.

Improving industrial process performance while keeping costs low is a critical goal for organizations seeking to maintain competitiveness and profitability. This paper explores cost-effective strategies that can be implemented to improve efficiency, productivity, and quality in industrial production systems. We discuss various strategies such as lean manufacturing, process optimization, automation, and energy management, all of which can be adopted without significant capital investment. The paper highlights case studies, provides insights into the implementation of these strategies, and discusses the challenges and benefits associated with their application in industrial settings.

There is growing interest in hydrogen for decarbonizing hard-to-abate industries. However, determining which industries to target, the scale of the opportunity, and how to meet the hydrogen demand presents uncertainties. This article structures the work on these topics, focusing on U.S. industry through midcentury. The most significant decarbonization opportunity identified in the literature is in the refining and chemicals industries, where emissions-intensive grey hydrogen is already utilized. New opportunities arise in displacing carbon-based reductants (steelmaking) and/or using hydrogen for high-temperature heat in cement, aluminum, and glass manufacturing. Hydrogen technologies have high readiness levels and face modest technical barriers in burner and furnace design. The primary challenge lies in reducing the cost of clean hydrogen production and delivery to compete with natural gas, crucial for any hydrogen industry decarbonization strategy, and currently representing &lt;1% of U.S. hydrogen production. The literature presents diverse U.S. industry clean hydrogen demand predictions (5-22 Mt/year by 2050) due to conflicting projections of industrial output, some incompatible with decarbonization goals; e.g., growth in gasoline production. After reconciling literature on hydrogen technology readiness, alternative decarbonization strategies, and U.S. climate targets, we estimate 2050 industrial clean hydrogen demand at 3.8-14.9 Mt/year, saving 28-133 MtCO2eq (1-7% of current U.S. industry emissions). Green hydrogen production will require up to 886 TWh of low-carbon electricity, equivalent to 90% of current renewable generation.

Abstract Aluminum is the world's second most consumed metal, and its production contributes substantially to global greenhouse gas (GHG) emissions. When formulating decarbonization strategies, it is imperative to ensure their coherence and alignment with existing industrial practices and standards. A material flow analysis (MFA) is needed to gain a holistic and quantitative understanding of the flows and stocks of products/materials associated with all participants within the supply chain. To support risk-informed decision policymaking in decarbonizing aluminum manufacturing, this study develops a dynamic system model that maps global aluminum flows and computes their embedded GHG emissions. A baseline scenario is devised to reflect the current business and operation landscape, and three decarbonization strategies are proposed. Deterministic simulation is performed to generate dynamic material flows and performance metrics. Monte Carlo simulation is then implemented to evaluate the robustness of the system's performance under demand uncertainties. The results reveal the immense carbon implications of material efficiency, as well as the preponderant role of post-consumer scrap recycling in decarbonizing aluminum manufacturing. Informed by simulation outputs, macro decarbonization guidelines are formulated for various criteria. The object-oriented programming framework that underlies the dynamic MFA may be integrated with network analysis, agent-based simulation, and geospatial interfaces, which may lay the foundation for modeling more fine-grained material flows and supply chain structures.

The iron and steel sector is highly energy- and emissions intensive, accounting for 8% of global final energy use and 7% of global direct energy-related CO 2 emissions. In steel production, most emissions are generated when steel is produced from primary raw materials, but 60% to 80% of energy can be saved when steel is produced from scrap in electric arc furnaces (EAFs). This study provides a comprehensive update of the U.S. steel cycle for year 2017, demonstrating that the U.S. already uses much higher shares of scrap in steel production than the world average, with detailed information on primary and secondary steel production and the manufacture of steel into intermediate products and finished goods. Options to further reduce the embodied energy of steel in the U.S. will be discussed and include increasing secondary production through more efficient collection and separation, reducing the energy intensity of primary production through a shift from blast furnaces to direct reduced iron (DRI) technology, and a shift towards low-carbon electricity sources for DRI and EAF furnaces.

The embodied energy of vehicles is growing as energy-intensive materials such as aluminum auto body sheet (ABS) are used to deliver improved performance. This presents an opportunity for recyclers to shift towards high-value recycling into wrought alloys and for car makers to increase the end-of-life (EOL) recycled content of their sheet, reducing their material costs and energy burden. However, the current system cannot effectively recycle the aluminum and steel sheets. Shredded and contaminated EOL metal (e.g., mixed aluminum alloys with steel rivets and mixed steel alloys with embedded copper wiring) is often exported, downcycled to castings, or recycled as rebar. In this paper, we will discuss preliminary work on developing a design tool that couples the effects of vehicle design, recycling system practices and technologies, and alloy compositional tolerances, under different ecosystem scenarios from 2020-2050. Ultimately, design and R&D recommendations will be made using the tool to test the effect of cross-lifecycle design changes on the system metrics , which include cumulative energy demand (primary energy), greenhouse gas emissions, and primary metal demand associated with automotive sheet metal production.

This data is linked to the article published in The International Journal of Life Cycle Assessment with the same title. The data presents the results of three structural path analyses focusing on three different countries manufacturing chillers, i.e., China, Italy, and Czechia. A specific social impact category is investigated in each country, i.e., Association and barganing rights, Migration flows, and Gender wage gap, respectively. All three structural path analyses were performed using the pyspa package (https://github.com/hybridlca/pyspa).

This special issue presents a collaborative initiative between the ASME Manufacturing Engineering Division (MED) and the Design Engineering Division (DED) to promote research in sustainability within the design and manufacturing communities. As the need grows for methodologies and tools capable of supporting sustainable systems, this compilation presents recent research trends exploring the integration of sustainability principles into and progression toward sustainability goals by the design and implementation of engineered systems.We were pleased to receive submissions covering a wide range of design and manufacturing topics. The quality of the submissions and the enthusiastic participation of the community were notable. Following a rigorous peer-review process, 11 submissions were ultimately selected for publication. The papers are published within special sections of the Journal of Mechanical Design (JMD) and the Journal of Manufacturing Science and Engineering (JMSE).The chosen submissions for the special issue cover research within diverse sustainability-related topics categorized into six main groups: (1) design for remanufacturing, (2) advancements in sustainability assessment, (3) design interventions for user-sustainable behavior, (4) advancing disassembly practices, (5) techno-economic analysis of energy systems, and (6) innovative tools for teaching sustainability, as detailed below.Design for Remanufacturing: This category features two papers. The paper by Behtash and colleagues introduces a comprehensive framework, 'Reman Co-Design,' merging design and remanufacturing optimization for enhanced sustainability performance. The paper by Alves and co-workers addresses textile waste by proposing 12 design for sustainability and flexibility principles. The work showcases the development and testing of a mechanical textile recycling system for efficient material recovery.Advancements in Sustainability Assessment: This category explores innovative technologies for sustainability assessment, covering three papers. The paper by Mabey and co-workers introduces an approach for predicting social, environmental, and economic impacts of products through agent-based modeling and life cycle assessment. The work of Karkaria and colleagues proposes a machine learning-based framework for predicting tire life in the commercial freight industry by focusing on product usage data. The paper by Liao and co-workers investigates the capabilities of AI algorithms for the automated evaluation and rating of product repairability.Interventions for Sustainable Behavior: This category explores interventions designed to drive behavioral change toward improving sustainability performance. The paper by Halabieh and Shu explores the effect of limiting wastewater outflows in reducing water consumption and influencing water-conserving behavior. The paper by Rea and colleagues investigates the relationship between thermostat interaction and temperature selection in warm ambient conditions and shows how innovative intervention offers insights into designing products that foster energy-conscious behaviors.Advancing Disassembly Practices: This category investigates disassembly strategies critical for sustainable product life cycles. The paper by Rodríguez and Favi proposes an eco-design methodology for mechatronic products, focusing on repairability and circular economy principles. The paper by Lee and colleagues explores the prospects and challenges of introducing human-robot collaboration in product disassembly. Recognizing the inefficiency of manual disassembly, the study reviews recent progress in robotic disassembly, emphasizing the potential benefits of combining human skills with robotic precision.In addition to the aforementioned topics, the remaining two papers address techno-economic analysis and sustainability education. The paper by Li and Zhang examines the techno-economic dynamics of co-located wind and hydrogen energy systems within an integrated energy system. The paper by Raoufi and Haapala shifts focus to teaching sustainability concepts to non-experts using an analysis tool designed to facilitate sustainability performance analysis of manufacturing processes and systems.We hope these special sections of JMD and JMSE lay the foundations for ongoing and future research in sustainable design and manufacturing. Furthermore, we hope this compilation of papers has established a pathway for disseminating the attendant research findings to future visionary leaders across an array of disciplines to advance the application of sustainable systems engineering research within our society and industry.Special gratitude goes to Dr. Carolyn Seepersad, Editor-in-Chief of JMD; Dr. Albert Shih, Editor-in-Chief of JMSE; and Dr. Wei Chen, former Editor-in-Chief of JMD, for their invaluable support, leadership, and endorsement of the idea behind this joint special issue on Advances in Design and Manufacturing for Sustainability. We also extend our sincere appreciation to Amy E. Suski and Emily Bosco, editorial assistants of JMD and JMSE, for their significant support and efficient collaboration throughout the paper review and production process.We also express our gratitude to all contributors who responded to the call and enriched the joint special issue with their insightful submissions. A special acknowledgment goes to the reviewers for their generous time and insightful evaluations. In addition, our thanks extend to the technical committees of ASME DED and ASME MED for disseminating information about the special issue within their respective communities.

Abstract Laser powder bed fusion (L-PBF) can make injection molding tools with conformal cooling channels that can reduce plastic part warpage and production cycle times. In this study, we define the design space in which L-PBF tools are economically and environmentally beneficial compared to traditional toolmaking methods. We develop mechanistic production cost, lead time, and cradle-to-grave life cycle assessment models of the cumulative energy demand (CED) and greenhouse gas (GHG) emissions associated with injection molding (IM) tool inserts made from tool steel using a Renishaw AM500Q L-PBF machine versus conventional 3-axis machining of equivalent beryllium copper (BeCu) inserts. These models are applied to a set of seven tool inserts used to make glass fiber reinforced plastic engine intake filter housings. The models are informed using insert build time and electrical power measurements on the L-PBF machine. Across all the inserts, using L-PBF was found to be slower, costlier, and more energy and emissions intensive than conventional mold making. However, benefits from reduced IM cycle times and plastic part rejection rates when using the L-PBF inserts can lead to whole life cycle improvements. We calculate how the breakeven plastic part production volumes (above which L-PBF inserts are faster, cheaper, and more environmentally benign over the life cycle) change for varying cycle time and part reject rate improvements. Assuming the L-PBF inserts result in 15% and 2.5% reductions in cycle time and part rejects respectively, then the breakeven production volumes are 60,000 parts for lead time, 4,000 parts for CED and GHG emissions, and greater than the likely mold lifespan (100,000 parts) for cost.

Abstract Bayesian inference allows the transparent communication and systematic updating of model uncertainty as new data become available. When applied to material flow analysis (MFA), however, Bayesian inference is undermined by the difficulty of defining proper priors for the MFA parameters and quantifying the noise in the collected data. We start to address these issues by first deriving and implementing an expert elicitation procedure suitable for generating MFA parameter priors. Second, we propose to learn the data noise concurrent with the parametric uncertainty. These methods are demonstrated using a case study on the 2012 US steel flow. Eight experts are interviewed to elicit distributions on steel flow uncertainty from raw materials to intermediate goods. The experts' distributions are combined and weighted according to the expertise demonstrated in response to seeding questions. These aggregated distributions form our model parameters' informative priors. Sensible, weakly informative priors are adopted for learning the data noise. Bayesian inference is then performed to update the parametric and data noise uncertainty given MFA data collected from the United States Geological Survey and the World Steel Association. The results show a reduction in MFA parametric uncertainty when incorporating the collected data. Only a modest reduction in data noise uncertainty was observed using 2012 data; however, greater reductions were achieved when using data from multiple years in the inference. These methods generate transparent MFA and data noise uncertainties learned from data rather than pre‐assumed data noise levels, providing a more robust basis for decision‐making that affects the system.

Informed decision-making for sustainable manufacturing requires accurate manufacturing process environmental impact models with uncertainty quantification (UQ). For emerging manufacturing technologies, there is often insufficient process data available to derive accurate data-driven models. This paper explores an alternative mechanistic modeling approach using easy-to-access data from a given machine to perform Bayesian inference and reduce the uncertainty of model parameters. First, we derive mechanistic models of the cumulative energy demand (CED) for making aluminum (AlSi10) and nylon (PA12) parts using laser powder bed fusion (L-PBF). Initial parametric uncertainty is assigned to the model inputs informed by literature reviews and interviews with industry experts. Second, we identify the most critical sources of uncertainty using variance-based global sensitivity analyses; therefore, reducing the dimension of the problem. For metal and polymer L-PBF, critical uncertainty is related to the adiabatic efficiency of the process (a measure of the efficiency with which the laser energy is used to fuse the powder) and the recoating time per layer between laser scans. Data pertinent to both of these parameters include the part geometry (height and volume) and total build time. Between three and eight data points on part geometry and build time were collected on two different L-PBF machines and Bayesian inference was performed to reduce the uncertainty of the adiabatic efficiency and recoating time per layer on each machine. This approach was validated by subsequently taking direct parameter measurements on these machines during operation. The delivered electricity uncertainty is reduced by 40-70% after performing inference, highlighting the potential to construct accurate energy and environmental impact models of manufacturing processes using small easy-to-access datasets without interfering with the operations of the manufacturing facility.