No abstract

Abstract:In this paper, a numerical model is developed by combining thermodynamics with heat transfer theory. Taking inner and external multi-irreversibility into account, it is with a complementary equation for heat circulation in air gaps of a steady cooling system with commercial thermoelectric modules operating in refrigeration mode. With two modes concerned, the equation presents the heat flowing through air gaps which forms heat circulations between both sides of thermoelectric coolers (TECs). In numerical modelling, a TEC is separated as two temperature controlled constant heat flux reservoirs in a thermal resistance network. In order to obtain the parameter values, an experimental apparatus with a commercial thermoelectric cooler was built to characterize the performance of a TEC with heat source and sink assembly. At constant power dissipation, steady temperatures of heat source and both sides of the thermoelectric cooler were compared with those in a standard numerical model. The method displayed that the relationship between f and the ratio ′ c / c was linear as expected. Then, for verifying the accuracy of proposed numerical model, the data in another system were recorded. It is evident that the experimental results are in good agreement with simulation(proposed model) data at different heat transfer rates. The error is small and mainly results from the instabilities of thermal resistances with temperature change and heat flux, heat loss of the device vertical surfaces and measurements.

No abstract

No abstract

In this paper, a universal loss calculation model in simulation is established in MATLAB / Simulink which can be used in not only ideal but also non-ideal conditions (non-sinusoidal load current and grid voltage, unfixed system parameters such as stator resistance for induction motors, unstable DC voltage with fluctuation, aperiodic modulation strategies with unfixed switching frequency, and so on). Considering the influence of dead time, steady operating states of each device can be obtained by the measurement of the load current and gate triggering pulses of every IGBT in a phase arm which can be easily measured in experiment rather than the currents and voltages of devices. Analyses in two-level topology and Neutral-Point-Clamped (NPC) three-level topology were done as examples. In periodic pulse width modulation with sinusoidal modulation wave, consistency among results from the proposed simulation model, the formula commonly used and commercially available loss software provided by power device manufacturer was shown. Through the design of a two-level inverter in hysteretic control, experimental verification of proposed method was applied. Calculating power loss in this method can be adapted to different converter topologies with various control strategies which is not even for ideal conditions. Nomenclature A Quadratic coefficient for quadratic polynomial fitting of switching loss and current according to the least-square method [mJ/A 2 ] B Monomial coefficient for quadratic polynomial fitting of switching loss and current according to the least-square method [mJ/A] C Constant term for quadratic polynomial fitting of switching loss and current according to the least-square method [mJ] E Power loss [W] f Frequency [Hz] () it Real-time current through the device [A] I Current amplitude [A] K Correction coefficient for junction temperature relative to the base junction temperature M Modulation ratio r Conduction resistance [  ] R Conduction resistance at the base junction temperature t Time[s] T Temperature [℃] () ut Real-time voltage of device [V] U Voltage amplitude [V]