Transportation Energy Conversion

Lawrence Berkeley National Laboratory (LBNL) will develop a high power density, rapid-start, metal-supported solid oxide fuel cell (MS-SOFC), as part of a fuel cell hybrid vehicle system that would use liquid bio-ethanol fuel. In this concept, the SOFC would accept hydrogen fuel derived from on-board processing of the bio-ethanol and air, producing electricity to charge an on-board battery and operate the motor.

Princeton Optronics will develop a low-cost, high-temperature capable laser ignition system which can be mounted directly on the engine heads of stationary natural gas engines, just like regular spark plugs are today. This will be done using a newly developed high-temperature Vertical Cavity Surface Emitting Laser (VCSEL) pump combined with a solid-state laser gain material that can operate at temperatures typically experienced on a stationary natural gas engine.

Carnegie Mellon University will use deep reinforcement learning and atomistic machine learning potentials to predict catalyst surface stability under reaction conditions. Current methods for determining the metastability of bifunctional and complex surfaces undergoing reaction are difficult and expensive. Carnegie Mellon’s technology will enable stability analysis in both traditional catalysts and new classes of materials, including those used in tribology (friction), corrosion-resistant alloys, additive manufacturing, and battery materials.

The University of Texas at Austin proposes to create efficient, accurate, and scalable deep neural network (DNN) representations of design optimization problem solutions. The inputs to these DNN representations will be the vector of design requirement parameters, the outputs will be the optimal design variables, and the goal is to learn the map from inputs to outputs (i.e., inverse design). The team will focus on the problem of the optimal shape design of aerodynamic lifting surfaces—in particular aircraft wings—using Reynolds-Average Navier Stokes models for minimal drag and energy savings.

The University of Maryland (UMD) will create inverse design tools for the development of enhanced heat transfer surfaces at reduced computational cost. Heat transfer surfaces are used to increase the efficiency of many energy conversion systems, but they are currently designed in a slow, iterative fashion. UMD will use a direct inverse design method map from given environments and performance metrics to design variables or materials.

GE Research will develop design optimization tools for the laser powder bed fusion based additive manufacturing of turbomachinery components. The team will integrate the latest advances in multi-physics topology optimization with fast machine learning-based producibility evaluations extracted from large training datasets comprising high-fidelity physics-based simulations and experimental validation studies. The integrated methodology will be used to demonstrate simultaneous improvements in the producibility and thermodynamic efficiency of a multi-physics turbomachinery component.

The National Renewable Energy Laboratory (NREL) will develop a novel wind turbine design capability that enables designers to explore advanced technology concepts at a lower cost. This capability will harness the power of a deep neural network (DNN)-based inverse design methodology. To overcome challenges with the use of traditional DNNs in this application, NREL will develop innovative techniques to sparsify the neural network using active subspaces that will ensure that the model is invertible and can quickly zoom in on relevant designs at minimal cost.

The Massachusetts Institute of Technology (MIT) will develop machine learning (ML) enhanced tools to accelerate the development of catalysts that promote the oxygen evolution reaction (OER) or the oxygen reduction reaction (ORR). These reactions are critical to the cost-effective generation (OER) or oxidation (ORR) of hydrogen. Available catalysts for promoting these reactions include scarce and costly precious metals like platinum. Hence, their practical applications are limited by high cost and low abundance in addition to moderate stability.

GE Global Research will develop a probabilistic inverse design machine learning (ML) framework, Pro-ML IDeAS, to take performance and requirements as input and provide engineering designs as output. Pro-ML IDeAS will calculate the design explicitly without iteration and overcome the challenges of ill-posed inverse problems. Pro-ML IDeAS will use GE’s Bayesian hybrid modeling with multi-fidelity intelligent design and analysis of computer experiments and a novel probabilistic invertible neural network (INN). The proposed framework can be applied to general complex design problems.