3D printed heat sink elements have a much higher density than heat sink elements produced by conventional methods such as casting and/or brazing. As a result, the porosity in 3D printed heat sink elements has received less attention. And, the walls of 3D-printed heat sink elements can be designed to be thinner than those of heat sink elements manufactured using conventional methods, and the 3D printed heat sink elements are more compact and smaller. According to market research by 3D Science Valley, the cooling channel of the 3D printed heat sink element can be connected in series with the microchannel heat sink for cooling other components, saving the number of parts and avoiding the need for welding.
In this issue, through excerpts of recent domestic practice and research in the field of microchannels, 3D Science Valley and Gu You will appreciate the subtleties of microchannels.
© 3D Science Valley White Paper
Water cassette type based on micro-channel cooling
Structural optimization and experimental analysis of calorimeter
© 3D Science Valley White Paper
Wen Peng, Chen Lianzhong, Chen Ding, Chen Zhiming
China Academy of Aerospace Aerodynamics
Summary:
In the process of re-entry of high-speed aircraft, high-speed aircraft face a severe aerodynamic heating environment, and the maximum airflow temperature can reach tens of thousands of degrees. In order to solve the problem of continuous and accurate measurement of high heat flux on the surface of hypersonic vehicle model in ground test, a new water card calorimeter based on microfluidic channel cooling was developed, and its internal micro-runner structure was constructed by 3D printing technology. Through numerical simulation, the size and layout of the microfluidic channel of the new water card calorimeter were determined, and the temperature rise in the core area of the new water card calorimeter was reduced by nearly 50% compared with the traditional water card calorimeter with the same cold wall heat flow, which verified the testing ability and survivability of the microfluidic water card calorimeter in extreme thermal environments. The results of arc heating jet test show that the microfluidic water cassette calorimeter can measure the pressure, temperature and heat flow data synchronously, realize the integrated identification of high-temperature flow field parameters, and have good dynamic response characteristics, and the maximum measured heat flow value is more than 18MW/m2, and the absolute and average deviations of the measurement are controlled within 3.44% and ±1.72%, respectively.
Based on topology optimization methods
Study on the temperature uniformity of the heat dissipation structure of spider webs
© 3D Science Valley White Paper
Zhang Kan, Fu Ting, Wang Jiangbo
Wuhan University of Science and Technology
Summary:
The bionic spider web heat dissipation structure has a wide range of applications in the heat dissipation of high heat flux chips, but there is still the problem of uneven temperature distribution of the heat dissipation structure. In order to further improve the temperature uniformity of the heat dissipation structure of the spider web, the optimal design of the spider web structure was proposed by using variable density topology optimization. By analyzing the heat transfer performance of the topological runners obtained under different inlet and outlet layouts and different design domain shapes, it is found that the multi-inlet, multi-outlet and staggered arrangement can effectively improve the temperature uniformity of the topological runners, and the temperature uniformity of the topological runners is increasing when the number of design domain edges is less than ten. Finally, by finite element analysis, the traditional structure M1 and the three-dimensional topological reconstruction structures M2 and M3 with different design domain edges are simulated.
The results show that the temperature uniformity of M3 with 10 design domain edges is better than that of the traditional channel M1 and M2 with 6 design domain edges, and the thermal resistance of M3 is reduced by 18.48% compared with M1 at Re=1800, the temperature difference of the heat source surface is reduced by 25% compared with that of M1, and the PEC value of the comprehensive heat transfer performance evaluation factor reaches 1.22. In this study, the effectiveness of the topology optimization method in improving the temperature uniformity of the spider web heat dissipation structure is proposed and verified, which is helpful to promote the application of the topology optimization method in the heat dissipation structure.
Head structure of microchannel diffusion welded heat exchanger
Simulation analysis of the influence on flow characteristics
© 3D Science Valley White Paper
Fei Junjie1, Liu Minyun1,2, Xi Dapeng1, Tang Jia1, Liu Ruilong1, Zan Yuanfeng1, Huang Yanping1
1. Key Laboratory of Thermal and Hydraulic Technology of Nuclear Nuclear Reactors, China Nuclear Power Research and Design Institute2. Department of Engineering Physics, Tsinghua University
Summary:
In order to understand and grasp the influence mechanism of head geometry on the flow distribution capacity of microchannel diffusion welded heat exchanger (MCD) with supercritical carbon dioxide (SCO2) as working fluid, and to optimize the design of MCD head structure, improve the flow distribution uniformity of the heat exchanger, so as to improve the heat exchange efficiency and safety, the flow and flow distribution performance of MCD heads with different structures were studied by numerical simulation method. In order to solve the limitation of the number of meshes of complex heat exchanger models under hardware conditions, a UDF program that can be widely used in the simulation of MCD hydrodynamic properties was developed, which greatly reduced the repetitive mesh work and lowered the hardware threshold of simulation calculation. Fluent software was used to analyze the influence of local geometric parameters of the head (different head wall curve parameters, different porous baffle parameters, etc.) on the pressure drop, flow distribution performance and flow field.
The results show that the vortex generated in the inlet head cavity and the sudden shrinkage structure of the outlet head will cause pressure loss, and the low height of the quadratic curve wall head can effectively inhibit the vortex generation and reduce the pressure loss caused by the sudden shrinkage structure, so as to reduce the pressure drop of the head and improve the flow distribution performance.
High power based on new process technology
Heat dissipation technology for bare chip module microfluidic systems
© 3D Science Valley White Paper
Li Lidan, Qian Zifu, Zhang Qingjun, Liu Zhujun, Li Zhi, Li Peng
Sichuan Jiuzhou Electric Appliance Group Co., Ltd
Summary:
In order to solve the heat dissipation problem of high-power bare chips, a self-circulating integrated microfluidic channel heat dissipation system is designed on the power module cavity, and the heat dissipation characteristics of microfluidics, flat DC channels and cross-linked channels are compared. The results show that the heat dissipation characteristics of bare chips with microfluidic channels are better than those without microfluidics, and the heat dissipation characteristics of bare chips with cross-linked channels are better than those with flat direct channels; the bare die eutectic is welded to the diamond heat sink, and then the heat sink eutectic is welded to the power module cavity, and the conduction thermal resistance between the bare die and the power module cavity is reduced to 1/360~1/280 of the thermal resistance of the traditional process. The simulation and experiment can verify each other, and the maximum deviation is only 7.16%. The microfluidic system has strong heat dissipation capacity, which can solve the problem of bare die heat dissipation at an ambient temperature of 70°C and a heat flux density of 320 W/cm2.
Sweat based on biomimetic microfluidics
Structural design and application analysis of testing devices
Liu Hengxin
Qilu University of Technology
Summary:
Drawing on the knowledge of bionics, the structure of the open sweat collection microfluidic channel was optimized to control the flow of sweat and promote sweat transport. Based on theoretical analysis, structural design, and a research method combining numerical simulation and experimental verification, an open sweat collection microfluidic channel structure was designed and prepared. The main research contents of this project are as follows: (1) Drawing on the concept of biomimicry, a bionic capillary sweat collection structure based on vascular plants, a sweat collection microfluidic channel based on an open rectangular trough, and a wedge-shaped sweat collection microfluidic structure based on bionic cactus spines are designed. The results obtained by numerical simulation show that these types of water collection structures obtained by biomimetic design promote the collection of surface liquid to a certain extent. Its sweat collection is significantly increased. The disadvantage is that the liquid suction of the bionic capillary sweat collection structure is affected by the capillary arrangement, and the sweat collection is uneven and causes sweat retention. The open rectangular microfluidic channel and the bionic microfluidic channel based on the cactus wedge-shaped structure improve the shortcomings of uneven capillary absorption on the basis of optimizing the geometric structure, but a large amount of sweat trapped in the microfluidic channel causes the waste of sweat samples, which is not conducive to the acquisition of sweat specimen signals. In summary, starting from the concept of geometry and combining the basic elements of plane space, it is of great significance to expand the capillary effect of capillary into the open rectangular microchannel structure, which is of great significance for expanding the sweat collection surface and increasing the sweat collection volume. (2) In order to realize the effective transportation of sweat in the microfluidic channel, combined with the relevant theories of fluid mechanics, the structure design was carried out in the open microfluidic channel, and the characteristics of the leaf structure of Araucaria aucarcaria were cleverly used, and the open sweat collection microfluidic channel structure with the bionic structure of Araucaria araucaria leaves was designed. Taking the number of leaves in the transverse and longitudinal directions of Araucaria araucaria as a variable, the bionic structure arrangement of the blades with the transverse arrangement of "0.5+1+0.5", that is, a complete row of leaves in the middle, half a row of blades on both sides, and a number of longitudinal blades of 12 were obtained. Furthermore, it was concluded that the optimal liquid contact angle range for sweat collection was 30°~75°. The microfluidic channels in this arrangement meet the requirements of sweat collection and preferential accumulation, combined with the Laplace pressure of the capillary effect. (3) The light curing 3D printing prepared the structure of the sweat collection microfluidic channel and carried out the liquid adsorption experiment, the experimental results verified the reliability of the numerical simulation results, the internal blade layout of the microfluidic channel is transverse "0.5+1+0.5", and the liquid adsorption and transportation effect of the structure design with the number of longitudinal blades is 12. The design concept of sweat detection device based on open microfluidic channel was developed and the design ideas were given.
In summary, the bionic microfluidic channel structure designed and optimized in this study provides theoretical and experimental support for the realization of efficient collection of a large amount of sweat, and provides an idea for the application of open microfluidic channel in sweat detection devices.
Random variable width based on deep learning
Microfluidic concentration gradient chip design
© 3D Science Valley White Paper
Yu Junnan
Gangnam University
Summary:
Concentration gradient chips can generate drug solutions of different concentrations, and are widely used in personalized medicine, drug screening, and antibiotic susceptibility testing. At present, concentration gradient chips have problems such as narrow distribution range of outlet concentration and outlet flow rate, and lengthy design cycle. Therefore, it is necessary to find a concentration gradient chip structure with a wider distribution range of outlet concentration and outlet flow velocity, and realize the efficient and accurate design of concentration gradient chip. In this project, a design method of random variable width concentration gradient chip based on deep learning is innovatively proposed, and the research focuses on three aspects: mechanism analysis, design method research and experimental research, and the main contents and conclusions of this study are as follows: (1) In-depth analysis of the mass transfer mechanism of concentration gradient chip and convolutional neural network algorithm architecture. The basic characteristics and basic equations of microfluidic flow, as well as the working principle of the concentration gradient, were studied, and it was found that the outlet fluid behavior of the concentration gradient chip could be changed by a combination of microfluidic channels of different widths. The basic features, basic structure and solution process of the convolutional neural network algorithm are analyzed, and it is found that the geometric structure of the concentration gradient chip can be represented by the geometric feature matrix, which can effectively improve the training speed of the convolutional neural network model and reduce the computational complexity of the convolutional neural network model. (2) Through the combination of microfluidic channels with different widths, the design scheme of randomly variable width concentration gradient chip is proposed. Numerical simulation methods were used to study the outlet fluid behavior of a variety of randomly variable width concentration gradient chip structures. The results show that compared with the random equal-width concentration gradient chip, the three outlet concentration distribution ranges of the random variable width concentration gradient chip are extended by 9%, 16% and 11%, and the flow velocity distribution range is extended by 29%, 28% and 30%, respectively. When the number of microfluidic channel nodes is 5×5, that is, the total number of possible microfluidic channels is 40, and the occurrence probability of microfluidic channels is 70%, that is, the number of existing microfluidic channels is 28, the random variable width concentration gradient chip can obtain the best outlet concentration and outlet flow velocity distribution. (3) The convolutional neural network model is trained on the outlet flow velocity and outlet concentration dataset obtained by COMSOL simulation, and a design method for a random variable width concentration gradient chip is obtained. The results show that the prediction accuracy of the model is 92.68% and 91.51%, respectively, and the accurate prediction of the outlet fluid behavior is realized. Based on transfer learning, the c Transfernet model and the v Transfernet model were established to realize the rapid expansion of the concentration gradient chip database under different inlet conditions. The retrieval strategy and design process of random variable width concentration gradient chips were formulated to realize the minute-level design of concentration gradient chips. (4) On a 3D-printed random variable width concentration gradient chip, the accuracy of the outlet concentration prediction model is verified by experiments. The average absolute errors between the predicted and experimental values of outlet concentrations were 3.50%, 5.17% and 4.23%, respectively. The experimental results show that the method in this study can efficiently and accurately realize the automatic design of random variable width concentration gradient chips. In this project, a random variable width concentration gradient chip structure is proposed, and the automatic design of the concentration gradient chip is realized through the convolutional neural network model. Through mechanism analysis, the mass transfer mechanism of concentration gradient chip and the algorithm architecture of convolutional neural network were sorted out. Through the research of the design method, the key structural parameters of the random variable width concentration gradient chip are clarified, and a design method of the random variable width concentration gradient chip is obtained by fusing the convolutional neural network algorithm. Experimental studies verify the accuracy of the design method of the random variable width concentration gradient chip in this study.
The results show that the random variable width concentration gradient chip designed based on the convolutional neural network algorithm has the advantages of wide distribution range of outlet concentration and outlet flow velocity, high design efficiency and accuracy, and is expected to be applied to the automatic design of other microfluidic chips.
MEMS-based inkjet
Research on microfluidic channel design and control
© 3D Science Valley White Paper
Peng Xishun
Guizhou University
Summary:
In the field of additive manufacturing, inkjet printing chips have long been subject to foreign companies, and the production of inkjet printing chips with independent intellectual property rights in mainland China is imminent. Most of the designs of thermal bubble inkjet printing chips are limited to the reasoning of theoretical models, and there are still some problems in the whole set of chip processing and signal modulation. In this paper, a high-frequency thermal bubble inkjet printing chip based on MEMS is designed in the context of interdisciplinarity, and a drive signal control system is designed based on FPGA. The main contents of the paper are as follows: (1) In this paper, a high power density heating resistor based on TaN thin films is proposed. Considering the change of temperature distribution on the resistor surface, a layer of SiO2 film was added as a buffer layer between the TaN film and the Si3N4 film in order to avoid the occurrence of thermal stress over-causing the fracture of the resistor surface. Then, the finite element simulation is used to find the appropriate driving voltage and heating time, and the temperature distribution of the resistor surface is simulated, which further verifies the rationality of the model. (2) In this paper, an H-shaped flow-limiting structure is proposed, which shortens the cross-section between the chamber and the sprue and places a rectangular baffle to achieve the flow-limiting effect. This structure can effectively block the ink that is squeezed back into the sprue due to the expansion of air bubbles, so as to improve the utilization rate of ink in the chamber. Then, finite element simulation was used to find the appropriate size of the current-limiting structure, and the temperature change of the liquid in the chamber was studied. In addition, the entire inkjet process is simulated to estimate the inside of the chamber in the actual inkjet situation, and to further determine the appropriate driving voltage and duration. (3) Based on FPGA, a control system for inkjet drive signals is designed. Xilinx's Kintex-7-XC7K325T chip is used as the main control chip, the asynchronous FIFO is used as the cache device for small batches of data, and the DDR is used as the storage device for large-scale data, which provides conditions for offline printing. The system can control 10 printheads for inkjet printing at the same time, which meets the conditions of high-frequency printing. (4) Use MEMS technology to prepare hot bubble inkjet chips, and build a chip debugging environment. In this paper, the actual inkjet situation is tested by applying different voltages and compared with the previous simulation results. Then, by building a rotating disc as a verification method for inkjet frequency, an optical microscope was used to observe the distance between adjacent ink droplets on the film paper on the surface of the disc and observe the shape of the ink droplets, so as to observe the matching degree between the driving conditions and the simulation model in the actual situation. In addition, a single inkjet head is continuously inkjetted 200,000 times, and the volume of the inkjet is calculated to approximate the ink volume of a single inkjet to determine that the volume meets the desired level.
Based on the embedded cooling module
Research on microchannel heat dissipation technology
© 3D Science Valley White Paper
Li Xuebin
Guilin University of Electronic Technology
Summary:
The development of artificial intelligence and fifth-generation mobile communication technology is promoting the rapid development of electronic chips in the direction of miniaturization and high integration, and at the same time, it will also lead to the increasingly serious heating problem of electronic chips. Because in practical engineering applications, there are often cases where chip arrays work at the same time, in addition to considering the heat dissipation performance and temperature uniformity of each heat source, the temperature uniformity between heat sources is also one of the issues that need to be considered when dissipating heat. Therefore, it is necessary to propose a heat dissipation technology, which should have the following characteristics: it can effectively improve the heat dissipation capacity, temperature uniformity, and is suitable for heat dissipation of multiple heat sources. According to the requirements of the project, the author synthesized three enhanced heat transfer paths, proposed a microchannel heat dissipation technology based on embedded heat dissipation module, and carried out a series of studies around it, the specific research contents are as follows: 1. The flow and temperature fields of three microchannels with embedded heat dissipation modules and straight rectangular microchannels are compared and analyzed by numerical calculation method, and the performance improvement of three microchannels with embedded heat dissipation modules relative to MC-SR is analyzed. Through comprehensive comparison, the microchannel (MC-RPF) with pinfin-rib embedded heat dissipation module was selected as the core heat dissipation structure, and the four main structural parameters on MC-RPF were studied and analyzed. The results show that compared with other heat dissipation structures involved in this study, MC-RPF can increase the pressure drop by 26.71%, improve the heat transfer performance by 41.02%, and achieve the temperature uniformity by 59.6%. 2. On the basis of the first research content, the response surface analysis method and the Fast Non-Domination Sequencing Genetic Algorithm (NSGA-II.) were used to optimize the four parameters of MC-RPF, including rib height, needle fin height, number of needle fins and number of auxiliary flow channels, with the goal of voltage drop, thermal resistance and temperature standard deviation. The TOPSIS algorithm is used for multi-dimensional decision-making, and the structural parameters of MC-RPF with the best comprehensive performance are obtained. The results showed that MC-RPF had the best comprehensive performance when the rib height was 0.8 mm, the needle fin height was 0.8 mm, the number of needle fins was 19, and the number of auxiliary runners was 8. Subsequently, numerical calculations are used to compare and verify the optimized structure with the structure in response surface planning. The results show that the comprehensive performance of the optimized MC-RPF is increased by 124.31% and 3.04% compared with the unoptimized performance. 3. On the basis of the second research content, based on the idea of modular heat dissipation, a multi-heat source heat dissipation structure based on MC-RPF was proposed, and the structural design analysis and voltage drop optimization were carried out, and the structure used MC-RPF to dissipate heat for each heat source. The heat dissipation performance, temperature uniformity and fluidity of the three MC-RPF interconnected runners (series, parallel and hybrid MC-RPF interconnect runners) were comprehensively analyzed from three aspects: multi-field distribution, heat transfer performance between heat sources, and overall performance. The TOPSIS algorithm was used to comprehensively evaluate the three MC-RPF interconnected flow channels. The results show that the three MC-RPF interconnected runners are suitable for the heat dissipation of multiple heat sources in different flow ranges, and because the hybrid type has the widest flow range among the three MC-RPF interconnected runners, and also has good performance in other flow ranges, the hybrid type is selected as the MC-RPF interconnected runner with multiple heat source heat dissipation structure. Finally, according to the pressure loss correlation theory, the pressure drop of the hybrid MC-RPF multi-heat source heat dissipation structure was optimized, and the voltage drop of the optimized hybrid MC-RPF multi-heat source heat dissipation structure was reduced by 36.73% compared with that before optimization.
Process parameters for SLM forming CuCrZr alloys
Effect of microchannel roughness
HE Guohao11, LAO Zibin1, GAAN Honghai1, CAO Mingxuan1, FU Bin1, LIU Zhiping2, WANG Ying1, YUAN Minghui3
1. Faculty of Intelligent Manufacturing, Wuyi University2. School of Applied Physics and Materials, Wuyi University3. Department of Mechanical and Aeronautics and Astronautics, The Hong Kong University of Science and Technology
Summary:
In order to study the influence of process parameters on the surface roughness of the microfluidic channel of CuCrZr alloy formed by laser selective melting technology, the CuCrZr alloy specimen with microfluidic channel was prepared by orthogonal test method, and the roughness and micromorphology of the inner surface of the specimen were measured by three-dimensional morphometer and scanning electron microscope, and the influence law and internal mechanism of laser power, scanning rate and filling distance on the roughness of the overhanging surface and side surface of the microfluidic channel were analyzed. The results show that the scanning rate has the greatest influence on the roughness of the overhanging surface and the side surface, and the overhanging surface is affected by the mass of the melt pool itself, while the surface roughness of the side surface is greatly affected by the Marangoni effect, and the microfluidic sample with the minimum side surface roughness (16.91 μm) can be processed under the process parameters of laser power of 380 W, scanning rate of 520 mm/s and scanning spacing of 0.12 mm, and the microfluidic channel sample with the minimum side surface roughness (16.91 μm) can be machined under the process parameters of laser power of 380 W, scanning rate of 520 mm/s and scanning distance of 0.12 mm. With a filling distance of 0.14 mm, a microfluidic sample with a minimum roughness of the overhang surface (24.86 μm) can be processed. From the perspective of laser process window, this study provides a basis for the surface accuracy of laser selective melting technology.
l Valley Column l
Website Submissions l Send to [email protected]