【摘要】 芯片的冷却问题已经成为限制电子设备发展的技术瓶颈之一。本文通过数值模拟，对某人脸识别电子芯片在有限空间内的散热问题展开优化研究。研究芯片温升与轴流风扇流量、进风面积、芯片工作功率的关系，并通过多目标优化得到更加符合当前目标需求的芯片及风扇的工作参数。研究结果表明，轴流风扇流量从0增加到9CFM范围内，芯片温升下降约56%;风扇进风面积从0.031 m~2增大到0.196 m~2,芯片温升上升13%;随着芯片功率增加，芯片稳定工作的温升呈线性提升。通过多目标优化设计，得到的芯片及风扇的工作参数为P=5.87 W,Q=1.056 CFM,A=0.041 m~2。更多还原

【Abstract】 The cooling problem of chips has become one of the technical bottlenecks limiting the development of electronic devices. In this paper, through numerical simulation, the heat dissipation problem of a face recognition electronic chip in a limited space is studied. The relationship between the chip temperature rise and the axial fan flow rate, air inlet area, and chip operating power is studied, and the operating parameters of the chip and the fan are obtained through multi-objective optimisation to better meet the current target requirements. The results show that the temperature rise of the chip decreases by about 56% when the axial fan flow increases from 0 to 9 CFM; the temperature rise of the chip increases by 13% when the fan inlet area increases from 0.031 m~2 to 0.196 m~2; and the temperature rise of the stable operation of the chip increases linearly with the increase of the chip power. Through the multi-objective optimisation design, the operating parameters of the chip and fan obtained are P=5.87 W, Q=1.056 CFM, A=0.041 m~2.更多还原