Search for: All records

Abstract Shape memory alloys (SMAs) absorb and release large amounts of latent heat during martensitic transformation, making them ideal candidates for applications involving thermal energy storage and management. In this study, Cu–Zn–Al SMAs were investigated as lower-cost alternatives to NiTi-based SMAs for solid–solid phase change materials. The alloys were fabricated using an unconventional method of melting and solidification of the constituent elements sealed in quartz tubes under a pressurized Ar atmosphere. The alloys synthesized were found to exhibit superior figure of merit values for thermal energy storage, as compared to conventional solid–liquid phase change materials and NiTi-based SMAs, with thermal conductivity between 59 and 75 W/mK and latent heat values ranging from 3 to 6.5 J/g. Transformation temperature ranges (A_f–M_f) less than 20 °C were achieved within a wide operating temperature between − 145 °C and 100 °C. In addition, select CuZnAl compositions yielded excellent cyclic stability with only ± 2 °C shifts in transformation temperatures after 20 thermal cycles. The present study demonstrates the feasibility of CuZnAl SMAs for use in high heat flux thermal energy storage and management applications at a wider range of temperatures. 

Abstract The Global Navigation Satellite System (GNSS) airborne radio occultation (ARO) technique is used to retrieve profiles of the atmosphere during reconnaissance missions for atmospheric rivers (ARs) on the west coast of the United States. The measurements of refractive bending angle integrate the effects of variations in refractive index over long near‐horizontal ray‐paths from a spaceborne transmitter to a receiver onboard an aircraft. A forward operator is required to assimilate ARO observations, which are sensitive to pressure, temperature, and humidity, into numerical weather prediction models to support forecasting of ARs. A two‐dimensional (2D) bending angle operator is proposed to enable capturing key atmospheric features associated with strong ARs. Comparison to a one‐dimensional (1D) forward model supports the evidence of large bending angle departures within 3–7 km impact heights for observations collected in a region characterized by the integrated water vapor transport (IVT) magnitude above 500 kg . The assessment of the 2D forward model for ARO retrievals is based on a sequence of six flights leading up to a significant AR precipitation event in January 2021. Since the observations often sample regions outside the AR where moisture is low, the significance of horizontal variations is obscured in the average bending angle statistics. Examples from individual flights sampling the cross‐section of an AR support the need for the 2D forward model. Additional simulation experiments are performed to quantify forward modeling errors due to tangent point drift and horizontal gradients suggesting contributions on the order of 5% and 20%, respectively. 

As the demand for STEM jobs increases, central to the success of STEM education and careers is a strong foundation in mathematics. However, students’ interest in mathematics is often very low. Thus, it is imperative to cultivate interest in mathematics among high school students. To promote students’ interests and positive attitudes in mathematics, we implemented informal learning using design-based research (DBR). We show that DBR is a compelling and suitable methodology for our research aims. Then we report how DBR can extend from previous studies in using informal learning for mathematics and foster motivating learning ecology in a school setting. Our DBR project has completed four iterations. 

Neural network based computer vision systems are typically built on a backbone, a pretrained or randomly initialized feature extractor. Several years ago, the default option was an ImageNet-trained convolutional neural network. However, the recent past has seen the emergence of countless backbones pretrained using various algorithms and datasets. While this abundance of choice has led to performance increases for a range of systems, it is difficult for practitioners to make informed decisions about which backbone to choose. Battle of the Backbones (BoB) makes this choice easier by benchmarking a diverse suite of pretrained models, including vision-language models, those trained via self-supervised learning, and the Stable Diffusion backbone, across a diverse set of computer vision tasks ranging from classification to object detection to OOD generalization and more. Furthermore, BoB sheds light on promising directions for the research community to advance computer vision by illuminating strengths and weakness of existing approaches through a comprehensive analysis conducted on more than 1500 training runs. While vision transformers (ViTs) and self-supervised learning (SSL) are increasingly popular, we find that convolutional neural networks pretrained in a supervised fashion on large training sets still perform best on most tasks among the models we consider. Moreover, in apples-to-apples comparisons on the same architectures and similarly sized pretraining datasets, we find that SSL backbones are highly competitive, indicating that future works should perform SSL pretraining. 

Neural network based computer vision systems are typically built on a backbone, a pretrained or randomly initialized feature extractor. Several years ago, the default option was an ImageNet-trained convolutional neural network. However, the recent past has seen the emergence of countless backbones pretrained using various algorithms and datasets. While this abundance of choice has led to performance increases for a range of systems, it is difficult for practitioners to make informed decisions about which backbone to choose. Battle of the Backbones (BoB) makes this choice easier by benchmarking a diverse suite of pretrained models, including vision-language models, those trained via self-supervised learning, and the Stable Diffusion backbone, across a diverse set of computer vision tasks ranging from classification to object detection to OOD generalization and more. Furthermore, BoB sheds light on promising directions for the research community to advance computer vision by illuminating strengths and weakness of existing approaches through a comprehensive analysis conducted on more than 1500 training runs. While vision transformers (ViTs) and self-supervised learning (SSL) are increasingly popular, we find that convolutional neural networks pretrained in a supervised fashion on large training sets still perform best on most tasks among the models we consider. Moreover, in apples-to-apples comparisons on the same architectures and similarly sized pretraining datasets, we find that SSL backbones are highly competitive, indicating that future works should perform SSL pretraining. 

With a principled representation of uncertainty and closed form posterior updates, Gaussian processes (GPs) are a natural choice for online decision making. However, Gaussian processes typically require at least O(n2) computations for n training points, limiting their general applicability. Stochastic variational Gaussian processes (SVGPs) can provide scalable inference for a dataset of fixed size, but are difficult to efficiently condition on new data. We propose online variational conditioning (OVC), a procedure for efficiently conditioning SVGPs in an online setting that does not require re-training through the evidence lower bound with the addition of new data. OVC enables the pairing of SVGPs with advanced look-ahead acquisition functions for black-box optimization, even with non-Gaussian likelihoods. We show OVC provides compelling performance in a range of applications including active learning of malaria incidence, and reinforcement learning on MuJoCo simulated robotic control tasks. 

Bayesian optimization is a sample-efficient black-box optimization procedure that is typically applied to a small number of independent objectives. However, in practice we often wish to optimize objectives defined over many correlated outcomes (or “tasks”). For example, scientists may want to optimize the coverage of a cell tower network across a dense grid of locations. Similarly, engineers may seek to balance the performance of a robot across dozens of different environments via constrained or robust optimization. However, the Gaussian Process (GP) models typically used as probabilistic surrogates for multi-task Bayesian optimization scale poorly with the number of outcomes, greatly limiting applicability. We devise an efficient technique for exact multi-task GP sampling that combines exploiting Kronecker structure in the covariance matrices with Matheron’s identity, allowing us to perform Bayesian optimization using exact multi-task GP models with tens of thousands of correlated outputs. In doing so, we achieve substantial improvements in sample efficiency compared to existing approaches that model solely the outcome metrics. We demonstrate how this unlocks a new class of applications for Bayesian optimization across a range of tasks in science and engineering, including optimizing interference patterns of an optical interferometer with 65,000 outputs.