High heat transfer plant-inspired neural network structure controlled by variable magnetic field

Abstract

Efficient heat dissipation and thermal protection present urgent challenges in high-power integrated circuits (ICs). Although applying a coating of highly thermally conductive materials on the surface of ICs is anticipated to mitigate heat concentration issues, ensuring thermal protection for adjacent devices continues to pose a challenge. Inspired by the microstructure of unidirectional nutrient transport in plant roots, this study utilizes magnetic liquid metal droplets to develop a high thermal conductivity network capable of adaptively manipulating the heat transfer path. This approach aims to tackle the challenges of heat concentration, disordered thermal dissipation, and thermal protection for high-power ICs, thereby enhancing thermal management efficiency. By controlling the distribution of the magnetic field, this study orchestrates the structure of the thermal conduction network to ensure rapid and orderly heat dissipation of ICs while simultaneously validating the network's thermal protection performance. The temperature in the IC thermal concentration zone reaches thermal equilibrium at 399.1 K when the ambient temperature is at 295 K. As the ambient temperature rises to 333 K, the temperature in the IC heat concentration zone stabilizes at approximately 400 K. Simultaneously, the temperature at a specific point in the thermal path of the network registers at 341 K, with the temperatures of the devices flanking this point at 314 K. The magnetron thermal conduction network, inspired by the root structure of bionic plants, not only boosts the thermal management efficiency of ICs but also shows promising application prospects in aerospace, electronics, and other related industries.