A team led by Associate Professor Jinhui Yan and graduate student Adam Lawrence has won first place in the National Institute of Standards and Technology's Additive Manufacturing Benchmark Challenge (NIST AM-Bench), for Vat Photoplymerization. AM Bench is an initiative of NIST that seeks to provide industry benchmarks for additive manufacturing processes.
Every three years, AM-Bench presents a series of challenge problems related to additive manufacturing and invites submissions from engineers and modelers across the globe to predict benchmark measurement results. The competition provides researchers with an opportunity to test their simulations against rigorous, highly controlled experimental data in pursuit of industry-wide standards.
The challenge problem Yan and Lawrence took on focused on cure depth and required participants to model its dependence on resin composition and light exposure in VAT photopolymerization. Lawrence’s research with Yan in the Department of Civil and Environmental Engineering at The Grainger College of Engineering centers on exactly that: improving predictive modeling for light-based 3D printing by studying how resin cures under different light exposures and material systems.
Predicting how resin will behave under specific conditions is difficult because the ideal settings often vary from one printer to another. This can lead to failed prints, wasted material, and costly trial and error, all of which Yan and Lawrence’s methods are designed to reduce.
To improve predictions of resin behavior, they developed a novel multi-GPU-centric finite element model that captures three core aspects of the 3D printing process: light attenuation, curing chemistry, and heat transfer. Implemented on NVIDIA H200 GPUs, their framework can produce full-geometry and part-scale simulations with more than 150 million unstructured 3D elements in less than two minutes.
Such high-fidelity simulations will help manufacturers and researchers produce parts more consistently across labs, printers, and materials. Fewer errors and less waste can lower costs, shorten product development cycles, and improve confidence in 3D printing for real-world applications. The ultra-fast predictive capability of their, powered by state-of-the-art GPU hardware, also opens the door to real-time digital twin technologies—an area Yan’s group continues to explore.
The NIST AM-Bench Conference took place in November 2025 during ASME’s International Mechanical Engineering Congress & Exposition in Memphis, Tennessee, where the award was formally presented. Their winning team also included Shih-Teng Huang and Professor Dian-Ru Li of National Taiwan University.
Watch a time lapse of the cure process: