An official website of the United States government
This paper presents a 3D model of a terahertz photoconductive antenna (PCA) using black phosphorus, an emerging 2D anisotropic material, as the semiconductor layer. This work aims at understanding the potential of black phosphorus (BP) to advance the signal generation and bandwidth of conventional terahertz (THz) PCAs. The COMSOL Multiphysics package, based on the finite element method, is utilized to model the 3D BP PCA emitter using four modules: the frequency domain RF module to solve Maxwell’s equations, the semiconductor module to calculate the photocurrent, the heat transfer in solids module to calculate the temperature variations, and the transient RF module to calculate the THz radiated electric field pulse. The proposed 3D model is computationally intensive where the PCA device includes thin layers of thicknesses ranging from nano- to microscale. The symmetry of the configuration was exploited by applying the perfect electric and magnetic boundary conditions to reduce the computational domain to only one quarter of the device in the RF module. The results showed that the temperature variation due to the conduction of current induced by the bias voltage increased by only 0.162 K. In addition, the electromagnetic power dissipation in the semiconductor due to the femtosecond laser source showed an increase in temperature by 0.441 K. The results show that the temperature variations caused the peak of the photocurrent to increase by and , respectively, under a maximum bias voltage of 1 V and average laser power of 1 mW. While simulating the active area of the antenna provided accurate results for the optical and semiconductor responses, simulating the thermal effect on the photocurrent requires a larger computational domain to avoid false rise in temperature. Finally, the simulated THz signal generation electric field pulse exhibits a trend in increasing the bandwidth of the proposed BP PCA compared with the measured pulse of a reference commercial LT-GaAs PCA. Enhancing signal generation and bandwidth will improve THz imaging and spectroscopy for biomedical and material characterization applications.
more »
« less
3D Model of Terahertz Photoconductive Antenna using COMSOL Multiphysics
This paper presents a 3D model of a photoconductive antenna (PCA) on semiconductor substrate. The simulations were conducted using the COMSOL Multiphysics package. The model considers the laser excitation and the carrier generation acceleration in the semiconductor layer. The computational work was achieved using the frequency-domain RF module and the semiconductor module. The results demonstrate that simulating the active area alone produces sufficient accuracy ~ 0.01% in the RF module solution (solution of the electric and magnetic fields) and ~ 0.23% in the semiconductor solution (photocurrent solution). The reduction in the simulated area helps minimizing the required CPU time and memory requirement in the 3D model at THz frequencies. The largest case in this study was simulated at the National XSEDE Supercomputing with ~ 0.3 billion unknowns and memory requirement of ~ 3.2TB in the RF module.
more »
« less
- Award ID(s):
- 1948255
- PAR ID:
- 10314401
- Date Published:
- Journal Name:
- Antennas and Propagation Society International Symposium
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
To accelerate the training efficiency of neural network-based machine learning, a memristor-based nonlinear computing module is designed and analyzed. Nonlinear computing operation is widely needed in neuromorphic computing and deep learning. The proposed nonlinear computing module can potentially realize a monotonic nonlinear function by successively placing memristors in a series combing with a simple amplifier. The proposed module is evaluated and optimized through the Long Short-term Memory with the digit number recognition application. The proposed nonlinear computing module can reduce the chip area from microscale to nanoscale, and potentially enhance the computing efficiency to O(1) while guaranteeing accuracy. Furthermore, the impact of the resistance variation of memristor switching on the training accuracy is simulated and analyzed using Long Short-term Memory as a benchmark.more » « less
-
Abstract In this study, a hybrid predictive model is used to predict 3D asymmetric lifting motion and assess potential musculoskeletal lower back injuries for asymmetric lifting tasks. The hybrid model has two modules: a skeletal module and an OpenSim musculoskeletal module. The skeletal module consists of a dynamic joint strength based 40 degrees of freedom spatial skeletal model. The skeletal module can predict the lifting motion, ground reaction forces (GRFs), and center of pressure (COP) trajectory using an inverse dynamics based optimization method. The equations of motion are built by recursive Lagrangian dynamics. The musculoskeletal module consists of a 324-muscle-actuated full-body lumbar spine model. Based on the generated kinematics, GRFs and COP data from the skeletal module, the musculoskeletal module estimates muscle activations using static optimization and joint reaction forces through the joint reaction analysis tool. Muscle activation results between simulated and experimental EMG are compared to validate the model. Finally, potential lower back injuries are evaluated for a specific-weight asymmetric lifting task. The shear and compression spine loads are compared to NIOSH recommended limits. At the beginning of the dynamic lifting process, the simulated compressive spine load beyond the NIOSH action limit but less than the permissible limit. This is due to the fatigue factors considered in NIOSH lifting equation.more » « less
-
Abstract Mitigating the adverse impacts caused by increasing flood risks in urban coastal communities requires effective flood prediction for prompt action. Typically, physics‐based 1‐D pipe/2‐D overland flow models are used to simulate urban pluvial flooding. Because these models require significant computational resources and have long run times, they are often unsuitable for real‐time flood prediction at a street scale. This study explores the potential of a machine learning method, Random Forest (RF), to serve as a surrogate model for urban flood predictions. The surrogate model was trained to relate topographic and environmental features to hourly water depths simulated by a high‐resolution 1‐D/2‐D physics‐based model at 16,914 road segments in the coastal city of Norfolk, Virginia, USA. Two training scenarios for the RF model were explored: (i) training on only the most flood‐prone street segments in the study area and (ii) training on all 16,914 street segments in the study area. The RF model yielded high predictive skill, especially for the scenario when the model was trained on only the most flood‐prone streets. The results also showed that the surrogate model reduced the computational run time of the physics‐based model by a factor of 3,000, making real‐time decision support more feasible compared to using the full physics‐based model. We concluded that machine learning surrogate models strategically trained on high‐resolution and high‐fidelity physics‐based models have the potential to significantly advance the ability to support decision making in real‐time flood management within urban communities.more » « less
-
We introduce Video Transformer (VidTr) with separableattention for video classification. Comparing with commonly used 3D networks, VidTr is able to aggregate spatiotemporal information via stacked attentions and provide better performance with higher efficiency. We first introduce the vanilla video transformer and show that transformer module is able to perform spatio-temporal modeling from raw pixels, but with heavy memory usage. We then present VidTr which reduces the memory cost by 3.3× while keeping the same performance. To further optimize the model, we propose the standard deviation based topK pooling for attention (pooltopK std), which reduces the computation by dropping non-informative features along temporal dimension. VidTr achieves state-of-the-art performance on five commonly used datasets with lower computational requirement, showing both the efficiency and effectiveness of our design. Finally, error analysis and visualization show that VidTr is especially good at predicting actions that require long-term temporal reasoning.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
