An Intelligent Model of Process Selection for Sustainable Composites

Abstract

In pursuit of addressing the critical challenge of plastic waste management and for ensuring sustainable consumption and production, there has been notable initiatives in manufacturing industries. These include a shift from conventional materials to sustainable lightweight composites. Natural fiber reinforced polymer composites (NFRPCs) offer a solution, being eco-friendly, degradable, and sourced from renewable materials as well as providing comparable mechanical properties to traditional materials. Their adoption aligns with UN SDG 12, fostering more sustainable production practices and offering socioeconomic benefits. With the exponential growth in demand for NFRPCs, it is evident that integrating decision making tools with manufacturing is imperative. Decision support systems have emerged as a disruptive intervention in sustainable manufacturing. During the product design and development phase, designers often confront complex decisions regarding selection of best material and manufacturing process, which are intricately linked. In order to optimize material design, all process variables must be concurrently considered. Early consideration of manufacturing aspects through design for manufacturing is crucial, as decisions made in the initial design stages heavily influence downstream activities and costs. An integrated approach to material and process selection ensures compatibility and optimizes manufacturing efficiency, resulting in higher quality and cost-effective products. Existing literature indicates the availability of numerous commercial tools and methodologies for optimal material selection, particularly for conventional materials, and a few for sustainable composites. However, the domain of process selection appears largely underexplored, with only a handful of tools available, predominantly tailored for traditional materials. Hence, the current research endeavor aims to develop a holistic framework for the judicious selection of constituents of composite materials tailored for specific applications. Thereafter, it proposes a novel methodology of process selection which is integrated with the graphic user interface for an optimal process selection. The developed selection tool is named CompoCraft. The present research investigation is divided into three phases. The first phase underscores the multidimensional nature of selecting manufacturing processes, stressing on the factors including material properties, part characteristics (shape/complexity, size, mass, section thickness, tolerance, and roughness), production requirements (production volume and rate), and cost implications. It explores the complex relationship between composite constituents, their properties, and manufacturing processes in defining final product characteristics and performance. It emphasizes the importance of balancing fiber weight fraction (FWf) with fiber aspect ratio (FAR) in NFRPCs to optimize the composite performance. A case study on short coir fiber-based polymer composites is illustrated to analyze the interplay of material and process selection. The results reveal that injection molded coir-based composites yielded better mechanical properties as compared to compression molded composites. During the second phase of research work, a holistic framework has been discussed and developed for the judicious selection of composite constituents (matrix and reinforcement) tailored for specific application. Other than the optimal material selection, it can aid decision makers by providing information on the most critical criterion influencing the decision, accordingly the designers can allocate resources effectively, prioritize design objectives, and optimize product performance. In the case of selection of a natural fiber for sustainable composites, specifically for semi-structural applications, the developed hybrid MCDM framework comprised of objective weightage methods as Entropy and CRITIC method for calculating the criteria weights and the TOPSIS, VIKOR, and PROMETHEE II as ranking methods. After aggregating the various ranking lists using degree of membership method, the framework suggested Basalt as the optimal fiber. The FAR followed by the moisture content were found to be critical criteria for selection of Basalt fiber. Similarly, the analysis for selection of a biodegradable polymer that utilized WSM, WPM, WASPAS, and TOPSIS, concluded that PLA is the most appropriate biodegradable polymer for the food packaging applications. Water vapor permeation that prolongs the shelf life and cost which ensures commercial viability were the critical criteria for deciding the optimal biodegradable polymer for food packaging applications. The sensitivity analysis was conducted by varying the criteria weights. In the third phase, the knowledge based expert system: CompoCraft was conceptualized and developed for cost-effective manufacturing process selection with a focus on sustainable composites. The knowledge base for the expert system was developed from the questionnaire survey responses of the domain experts including academicians and industrialists. The process-criteria compatibility mapping was accomplished in fuzzy ratings. The process selection methodology involved a two phased approach: screening based on the composite constituents (material) and shape/complexity parameters; and ranking based on the part design specifications and production requirements. In the ranking phase, process compatibility score (PCS) was calculated for the screened processes. Effective process selection necessitated a balanced consideration of PCS and cost rating, enabling informed decision making to achieve the desired balance between efficiency, performance, and economic viability. Sensitivity analysis was performed to assess the impact of variations in the criteria weights on the PCS score and the ranks of the screened processes, providing insights into the robustness and reliability of the developed expert system. The selection methodology/inference system was integrated with the graphic user interface to develop the CompoCraft tool using App Designer application on MATLAB R2023a. The validation of the expert system was performed by showcasing three case studies for the NFRPC products. The CompoCraft system selected compression molding as the optimal manufacturing process for the case study extrapolated from the scientific literature wherein an automotive bumper beam was fabricated with flax/PLA composite. In a similar in-house case study, the decision support system (CompoCraft) selected compression molding as the optimal manufacturing process for chair boards fabricated with jute/epoxy composite. Injection molding was selected as the optimal manufacturing process for the third case of an industrial product, i.e., screw mounted fluted panel knob. Finally, the robust expert system; CompoCraft can be effectively used by the domain practitioners, designers, or researchers during the conceptual design stage of product development, specifically based on sustainable materials.