Approximating the Steady-State Temperature of 3D Electronic Systems with Convolutional Neural Networks

Thermal simulations are an important part of the design process in many engineering disciplines. In simulation-based design approaches, a considerable amount of time is spent by repeated simulations. An alternative, fast simulation tool would be a welcome addition to any automatized and simulation-based optimisation workflow. In this work, we present a proof-of-concept study of the application of convolutional neural networks to accelerate thermal simulations. We focus on the thermal aspect of electronic systems. The goal of such a tool is to provide accurate approximations of a full solution, in order to quickly select promising designs for more detailed investigations. Based on a training set of randomly generated circuits with corresponding finite element solutions, the full 3D steady-state temperature field is estimated using a fully convolutional neural network. A custom network architecture is proposed which captures the long-range correlations present in heat conduction problems. We test the network on a separate dataset and find that the mean relative error is around 2% and the typical evaluation time is 35 ms per sample ( 2 ms for evaluation, 33 ms for data transfer). The benefit of this neural-network-based approach is that, once training is completed, the network can be applied to any system within the design space spanned by the randomized training dataset (which includes different components, material properties, different positioning of components on a PCB, etc.).

On analytical methods for calculating the steady-state temperature field in a circular cylindrical shell

The present paper presents two analytical methods for calculating the steady-state temperature field in a circular cylindrical shell. The effectiveness of the methods in terms of accuracy in comparison with the classical approach, based on Bessel functions, is analyzed. The proposed analytical algorithms contain relatively simple computational operations. Since they do not use special functions, the algorithms can be used to solve a wide range of problems.

Approximating the Steady-state Temperature of 3D Electronic Systems with Convolutional Neural Networks

Thermal simulations are an important part in the design of electronic systems, especially as systems with high power density become common. In simulation-based design approaches, a considerable amount of time is spent by repeated simulations. In this work, we present a proof-of-concept study of the application of convolutional neural networks to accelerate those thermal simulations. The goal is not to replace standard simulation tools but to provide a method to quickly select promising samples for more detailed investigations. Based on a training set of randomly generated circuits with corresponding Finite Element solutions, the full 3D steady-state temperature field is estimated using a fully convolutional neural network. A custom network architecture is proposed which captures the long-range correlations present in heat conduction problems. We test the network on a separate dataset and find that the mean relative error is around 2 % and the typical evaluation time is 35 ms per sample ( 2 ms for evaluation, 33 ms for data transfer). The benefit of this neural-network-based approach is that, once training is completed, the network can be applied to any system within the design space spanned by the randomised training dataset (which includes different components, material properties, different positioning of components on a PCB, etc.).

Maximum Allowable Pressure Calculation of Water Tube Boiler during Operation

In the steam boiler industrial sector, pressure and temperature of the water tube are the two main factors that affect the safety and efficiency of a steam boiler. Explosions may be occurring because of a sudden drop in pressure without a corresponding drop in temperature. Therefore, understanding the temperature distribution of the water tube boiler is essential to control the failure and explosion of the boiler. Once the temperature distribution is known than the limiting factors that affect the water tube life such as the maximum allowable pressure can be determined. ANSYS software will be used to determine the temperature distribution in the water tube of a utility boiler during operation at elevated inlet water and furnace temperature. The theory of axisymmetric has been utilized since the water- tube is cylindrical in shape. In axisymmetric theory, a three-dimensional cylindrical problem like a water tube can be reduced to two-dimensional by ignoring the circumferential Ө, while the r-axis and z-axis became x-axis and y-axis or Cartesian coordinate. Then two-dimensional rectangular elements meshing for the profile cross-section along the water tube in r and z axes were implemented in a computerized simulation using ANSYS 10 to find out the steady-state temperature distribution of the water tube.

Thermal model of an asynchronous traction motor of a diesel locomotive

Objective: Evaluation of the temperature regime of the windings of an asynchronous traction electric motor with a squirrel-cage rotor (ATED) of the DAT-350 type of a 2TE25A diesel locomotive using the de- Современные технологии – транспорту 467 ISSN 1815-588Х. Известия ПГУПС 2021/4 veloped thermal model. Methods: The research was carried out by the fi nite element method in the application for calculating steady-state temperature fi elds SolidWorks. Resuls: To ensure the specifi ed service life of the ATED, the temperature rise of the windings must be limited to safe values, based on the insulation class. Therefore, it is crucial that designers and operators have access to accurate and computati onally effi cient, physics-based ATED thermal behavior modeling tools. Practical results: Using the proposed methods, it is possible to accurately and effectively evaluate the thermal characteristics of ATED of promising diesel locomotives. In addition, designers can easily and quickly adjust the parameters and performance characteristics of the ATED in such a way that they are conducive to improving the overall performance of the locomotive.

Resistive Heating of a Shape Memory Composite: Analytical, Numerical and Experimental Study

This work investigates in detail the Joule resistive heating phenomenon of electroactive Shape Memory Composites (SMC) when an electric current is injected at constant power. The SMC is a covalent poly(ε-caprolactone) network filled with 3 wt% of multiwall carbon nanotubes. The resistive heating of the SMC is studied by means of surface temperature measurements, analytical formulas and a coupled 3D thermo-electric numerical model. Analytical expressions are derived for the 2D temperature distribution within a parallelepipedic SMC, either with constant or linearly-dependent electrical resistivity. These analytical expressions can be used to investigate the influence of geometrical and material parameters in the steady-state temperature and its distribution across the sample. The results also allow one to identify the parameters that are crucial for predicting the temperature rise due to resistive heating: the temperature dependence of the resistivity has little effect on the steady-state temperature, whereas the thermal conductivity plays a significant role. The time-dependent temperature is shown to be related to the particular temperature dependence of heat capacity. Furthermore, the presence of external objects (clamps or grips) used during the shape memory cycle must be taken into consideration for a certain temperature to be reached since they result in a lower steady-state temperature and a slower resistive heating phenomenon. With the findings presented in this work, accurate resistive heating can be predicted for a SMC upon the injection of an electric current at constant power.

A low-cost, portable, and practical LAMP device for point-of-diagnosis in the field

Transition of rapid, ready-to-use, and low-cost nucleic acid-based detection technologies from laboratories to points of sample collection has drastically accelerated. However, most of these approaches are still incapable of diagnosis starting from sampling, through nucleic acid isolation and detection in the field. Here, we developed a simple, portable, low-cost, colorimetric, and remotely controllable platform for reliable, high-throughput, and rapid diagnosis using loop mediated isothermal amplification (LAMP) assays. It consists of a thermally isolated cup, low-cost electronic components, a polydimethylsiloxane sample well, and a fast prototyped case that covers electronic components. The steady-state temperature error of the system is less than 1%. We performed LAMP, Colony-LAMP, and Colony PCR reactions using the yaiO2 primer set for Escherichia coli and Pseudomonas aeruginosa samples at 65˚C and 30 min. We detected the end-point colorimetric readouts by the naked eye under day light. We confirmed the specificity and sensitivity of our approach using pure genomic DNA and crude bacterial colonies. We benchmarked our Colony-LAMP detection against Colony PCR. The number of samples tested can easily be modified for higher throughput in our system. We strongly believe that our platform can greatly contribute rapid and reliable diagnosis in versatile operational environments.

An experimental study of the temperature evolution of a single fin submerged in a phase changing material during melting at different heat fluxes

An experimental setup has been designed to study a single cylindrical fin placed in a cylindrical enclosure filled with phase changing material (PCM). The heat flux to the fin is measured at the top of the fin. The temperature evolution at different fin heights is measured by thermocouples placed internally in the fin. The evolution of these temperatures has been studied for different heat fluxes. This provides insight in the contribution of the different fin heights to the total heat transfer to the PCM during the different stages of the melting process. As such they can be used to assess the effectiveness of the fin over its length. After approximately 6h, the fin temperature stabilizes during melting. Due to the temperature drop over the fin, the bottom temperature reached is significantly lower than the temperature at the top and the contribution of this lower part to the total heat transfer is lower as well. For heat fluxes higher than 3805±75 W/m2, the steady-state temperatures at fin locations in contact with the melting PCM are similar. For low heat fluxes, this steady-state temperature is not reached during a 12h experiment. Longer experiments are thus needed to study the steady-state behaviour at these lower heat fluxes.