CMMI-EPSRC: Predictive Models for Manufacturing Stronger, Lighter Polymer Composites for Bone-Repair Implants

Surgeons who repair complex fractures or rebuild bone lost to trauma or cancer rely on implants such as plates and screws to stabilize the skeleton while it heals. Most of these devices are made from metal, which is strong but much stiffer than bone, can interfere with medical imaging, and can be costly to produce in patient-specific shapes. This joint National Science Foundation-United Kingdom Research and Innovation (NSF-UKRI) Engineering and Physical Sciences Research Council (EPSRC) project advances a new generation of lightweight, bone-like implants made from high-strength polymers reinforced with carbon fiber. The central obstacle is that these materials are difficult to manufacture reliably and affordably, which has slowed their clinical adoption. By pairing advanced manufacturing with predictive computer modeling, the research seeks to make these composites strong, consistent, and economical enough for routine clinical use, with the potential to improve patient recovery and quality of life. The work also strengthens advanced-manufacturing capability shared between the United States and the United Kingdom, trains post-doctoral and graduate engineers, and produces educational materials for students from high school through university. This grant is awarded under the NSF-UKRI Lead Agency Opportunity. The research develops and validates a hybrid additive manufacturing process for carbon fiber reinforced polyaryletherketones, a family high-performance polymers with an established record in medical implants. The process combines layer-by-layer printing with a subsequent consolidation step that applies heat and pressure to remove internal voids and increase strength. Fabrication parameters for both stages are optimized using a structured design-of-experiments approach. Printed and consolidated specimens, including realistic trauma-plate surrogates, undergo characterization of chemistry, crystal structure, porosity, microstructure, and mechanical performance using advanced characterization techniques. In parallel, a multi-scale finite element framework spanning fiber-matrix interactions, layer deposition, and full-part consolidation predicts thermal history, crystallinity, and porosity. These physics-based simulations train machine-learning surrogate models, including Gaussian Process Regression and Artificial Neural Networks, that map manufacturing parameters to material properties and through inverse design and identify conditions yielding target performance. Accelerated aging under hydrothermal, repeated steam-sterilization, and lipid-exposure conditions, combined with diffusion-based modeling, assesses durability under clinically relevant conditions. The project delivers an integrated process-structure-property framework with methods transferable to other composites and to sectors such as aerospace and energy. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.