Many heat transfer processes do not allow temperature measurements with advanced instruments like IR cameras, and thus require the use of thermocouples – often inserted sheathed thermocouples – in order to measure the temperature in some key-regions. Nevertheless, thermocouples have an intrinsic response time that can dampen substantially temperature measurements and, when applicable, affect heat flux estimations by inverse methods. In this study, a thermocouple measurement correction method is proposed, especially for fast thermal transients like quenching, to avoid underestimation of the boundary heat fluxes due to delayed temperature responses. A simplified energy balance at the hot junction allowed modeling the temperature measurement by a thermocouple, in which heat losses and the response time were taken into account so they can be estimated using the least squares method on experimental data. A series of tests was performed to validate the present method with a heated jet impinging on a copper body instrumented with two thermocouples: one ungrounded and sheathed, then with high response time (”slow”); and another with exposed welded wires, hence with low response time (”fast”). After having estimated the slow thermocouple response time and heat loss parameters, the model allowed a good reconstruction of the fast thermocouple signal using the slow thermocouple measurements. Also, the heat flux estimation by the inverse method using the reconstructed signal resulted in practically the same results obtained using the fast thermocouple data. A parametric sensitivity analysis showed the heat loss parameters can be neglected for heat flux estimations, while simulations showed that temperature noises degrade the response time estimation but data filtering can mitigate this noise effect. Finally, the present method was applied to a large scale jet-cooling of a hot plate – near industrial conditions – and showed a substantially higher dissipated heat flux than the initially estimated. In fact, the results using reconstructed signals demonstrated that the heat flux dissipated by the jet has a higher dependence on the jet Reynolds number than the observed with the original thermocouple measurements, showing how important the temperature measurement correction is to evaluate fast thermal processes.
Read full abstract