Numerical study of distributed jet microchannels with Micro-pin-fins: Structural comparison and temperature uniformity optimization

With the rapid development of high-power and large-area chips, temperature non-uniformity and hotspots induced by non-uniform heat flux have become major bottlenecks for performance and reliability. This study targets large-area high-power chips with strongly non-uniform heat flux and investigates a distributed jet microchannel heat sink integrated with micro-pin-fins. Under uniform-heat-source conditions, experiments were performed to compare a confined manifold and a fractal distributed-jet supply structure. The fractal design achieved significantly improved temperature uniformity, reducing the experimental temperature standard deviation from 6.61 K to 3.88 K. Numerical analysis further revealed the underlying mechanisms: the confined structure suffers from crossflow-induced maldistribution, whereas the fractal configuration maintains uniform jet impingement and suppresses heat-transfer dead zones. For non-uniform heat sources, a CFD-driven adaptive micro-pin-fin height optimization algorithm was developed. Under a 2000 W load, this method reduced the maximum temperature difference from 22.0 °C to 5.9 °C and decreased the temperature standard deviation by 85.6 %, demonstrating highly effective hotspot suppression. Overall, combining distributed jet arrays, micro-pin-fins, and adaptive height optimization provides a high-performance cooling solution capable of achieving both efficient heat dissipation and outstanding temperature uniformity, making it well suited for next-generation large-area high-power chips.