Using a pulsed electrical current enables a temporary state in which a Peltier element achieves temperatures below that obtained with a steady current. This is referred to as supercooling. Supercooling is followed by a period of superheating during which Peltier heat transfer is diminished and the surface temperatures increases. Most studies have found that the duration of superheating is longer than the duration of supercooling. As a result, the current pulse generates a net heating instead of enhanced cooling. There are limited studies that have shown the possibility of net cooling during a current pulse. The objective of this paper is to discuss the operating conditions for which net cooling is possible and maximized. The interaction between pulse duration and pulse height using isosceles shaped current pulses on net cooling was investigated using response surfaces generated using electrical-thermal analogies in SPICE. Pulse duration ranged from 0.1 to 10.0 seconds and pulse height ranging from 1.01 to 6.0 times steady current. Response surfaces were used to map a variety of performance factors; including the transient time to achieve a minimum temperature, the pulse cooling enhancement, transient penalty and transient advantage. The optimal combination of pulse duration and pulse height was identified.

No abstract

Described herein is a detailed and comprehensive multiphysics model of a thermoelectric generator (TEG). The one-dimensional model uses electricalthermal analogies solved for transient response using SPICE. There are many advantages and applications of thermoelectric generators. Wider use and application advancements are generally limited by the tools available for engineering and scientific studies. Currently, available modeling tools are limited by some combination of speed, platform capabilities, or missing physics that are not used or assumed to be negligible. The TEG module model herein is made up of two sub-models, the thermoelement model and the non-thermoelement model. Rather than a lumped thermoelement model, the model herein makes use of distributed physics that include the following: Thomson effect, temperature dependence, mass, Joule heat, thermal resistance, Seebeck effect, and electrical resistance. The non-thermoelement model takes into account temperature dependence and simulates Joule heat generation, thermal resistances, thermal and electrical interface resistances, and mass for and between the ceramic, copper, and solder. The comprehensive model herein was correlated to experimental data that simultaneously varied electrical current and hot and cold side temperatures with time. Very minimal adjustments to reported thermoelectric properties were required to almost perfectly match the experimental transient power output. The effects of the non-thermoelement model, distributed Thomson effect model and distributed temperature dependent property model were quantified. The model ran very quickly, taking 2.5 real-time seconds to run a 4000 s transient simulation.

Applying a current pulse enables a short-term transitory state where the cold junction of a Peltier couple reaches temperatures below that obtainable via maximum temperature delta steady-state current. Short-term cooling applications like on-chip hot spot and pulsed laser sensor cooling have been studied using pulsed cooling. Some studies have proposed applications that utilize consecutive repeating pulses for longer term cooling applications. These studies have found or theorized increased cooling and coefficient of performance (C O P). Considering these studies, it is desirable to have a more detailed analysis of how the additional cooling and C O P are achieved. The objective herein is to provide a detailed analysis of cooling rate and C O P during pulses using a realistically modeled system simulated in SPICE. It was found that cooling rate for long term consecutive pulse cooling applications can be increased over steady-state but C O P in most cases is reduced during current pulses. The reasons why this happens are studied in depth.

Applying an electrical current pulse enables a transitory state in which the cold junction of a Peltier couple reaches temperatures below that obtained with an optimized steady-state current. This is known as supercooling. This supercooling is followed by a period of superheating. It has been shown for optimized isosceles triangle shaped current pulses, the sum of supercooling and superheating can provide a net cooling advantage. In most cases, supercooling has been studied as a standalone couple and not as a system and only from the perspective of cold side temperature. The objective of this paper is to gain insight about the sensitivity of the system performance metrics of COP, power consumption, cold junction cooling rate ( Qc ) and average cooled heat generating object temperatures during pulsed operation. A comprehensive system variable parametric study was performed. The study used a system model that utilized electrical-thermal analogies in SPICE. It was demonstrated that Qc over an entire pulse event can be improved over I max steady operation but not steady Iopt operation. Qc can be improved over I opt operation only during the early part of the pulse event. COP is reduced during a pulse due to the fast time constant of power consumption relative to Qc . Time delayed Joule heat and Seebeck voltage contribute to further lowering of COP. During part of the pulse transient, lower performance interface materials improved Qc and COP which may mean interface resistance is an optimizable parameter for transient operation.