Heat pipes are capillary-pumped two-phase devices that transport heat from localized sources in electronics to a heat sink through continuous evaporation and condensation of an internal working fluid. Given the reliance of a heat pipe on capillary transport of liquid through internal wicking structures to the evaporator section where heat is applied, operating it at a power exceeding the so-called capillary limit can lead to dryout at the evaporator and subsequent device failure. However, in response to highly dynamic workloads in the electronics being cooled, heat pipes may more typically be exposed to heat loads above the capillary limit only intermittently over brief time intervals. Understanding the heat pipe response to these transient workloads is critical, as designing heat pipes for steady-state operation at the peak transient heat load would represent an expensive overdesign. Our previous work has experimentally characterized the transient heat pipe response to power pulses exceeding the capillary limit. It was demonstrated that a pulse must be sustained for a minimum duration termed the timeto-dryout before dryout is initiated. Once a pulse-induced transient dryout does occur in a heat pipe, its thermal resistance does not necessarily recover back to the pre-dryout performance even after the power input drops below the capillary limit. This behavior, termed thermal hysteresis, can be circumvented if the power is lowered (or throttled) to a sufficient threshold below the capillary limit for an extended time interval. In the current work, a first-of-its-kind transient heat pipe model is developed to predict the salient features of heat pipe response to pulse-load-induced dryout as well as recovery from dryout. The model uniquely considers spatiotemporal variations in local liquid saturation in the wick (i.e., the fraction of pore volume occupied by the liquid). Experiments are performed using commercial heat pipe samples that span a range of sizes and wick types to validate the model predictions. It is shown that the model can predict the transient thermal response, including thermal hysteresis, of the heat pipe during pre-dryout, dryout (pulse load), and post-dryout (recovery) stages with good accuracy. The model results are also validated against experiments for heat pipes spanning a range of wick types, heat pipe lengths, and heat pipe thicknesses. The capability to accurately predict crucial temporal events during dryout and recovery is key to establishing power expenditure strategies in electronics and designing heat pipes with improved dryout and recovery performance.
原文链接
分享
引用
