Heat flux measurements were performed in an air-cooled utility engine using a fast-response coaxial-type surface thermocouple. The surface heat flux was calculated using both analytical and numerical models. The heat flux was found to be a strong function of engine load. The peak heat flux and initial heat flux rise rate increase with engine load. The measured heat flux data were used to estimate a global heat transfer rate, and this was compared with the heat transfer rate calculated by a single-zone heat release analysis. The measured values of heat transfer were higher than the calculated values largely because of the lack of spatial averaging. The high load data showed an unexplainable negative heat flux during the expansion stroke while the gas temperature was still high. A 1D and 2D finite difference numerical model utilizing an adaptive timestep Crank-Nicholson (CN) integration routine was developed to investigate the surface temperature measurement. Applying the measured surface temperature profile to the 1D model, the resultant surface heat flux showed excellent agreement with the analytical inversion solution and captured the reversal of the energy flow back into the cylinder during the expansion stroke. The 2D numerical model was developed to observe transient lateral conduction effects within the probe and incorporated the various materials used in the construction and assembly of the heat flux sensor. The resulting average heat flux profile for the test case is shown to be slightly higher in peak and longer in duration when compared with the results from the 1D analytical inversion, and this is attributed to contributions from the high thermal diffusivity constituents in the sensor. Furthermore, the negative heat flux at high load was not eliminated suggesting that factors other than lateral conduction may be affecting the measurement accuracy.
R&D on divertor plasma facing components at the Institute for Plasma Research
