Report on Simulations with Thermal Fluid-Structure Interaction

Kurzfassung

Scope of the deliverable The numerical results generated in WP 4.4 are presented and discussed in comparison to the respective experiments. The two focal points of the investigation are a sensitivity analysis and the coupled simulations of the different passive concepts in comparison to the experimental data. The sensitivity analysis of the heat loads introduced to the model by the flow field includes geometric aspect, wind tunnel conditions as well as physical models. It aims at the extraction of the major influence parameters to improve the understanding of the experiments at the wind tunnel conditions. The already published results of coupled analysis at the nominal flow conditions [1] are not repeated here in detail but key results are shown in the Annex A: Numerical prove of concept 1b. The deviations of those results in comparison to the experimental values motivated the adjustments to the catalysis model applied for the coupled simulations discussed here. The coupled simulations concentrate on the generic leading edge model with a radius of 20mm at an angle of attack of -10° which is defined as the THOR reference configuration. The evaluation of the passive concepts for this model includes the plain model without a passive concept as a reference, each passive concept 1a and 1b as well as the combination of both concepts 1a+1b. The same four cases are repeated for a smaller radius of the leading edge providing even higher temperatures in the stagnation region. Results The major fluid mechanical influence on the chosen model geometry in the L3K wind tunnel is well described via a two dimensional analysis. The remaining 3D effects are small on the rear part of the model and can be neglected at the leading edge. Concerning the surface heat flux prediction, thermal non equilibrium and Argon as air component can be neglected. The positioning of the model inside the test section is negligible taking the measuring accuracy of the data into account. The reservoir pressure influences the peak heating but, with respect to the accuracy of its measurement, is of minor influence. The major parameter for the heat flux prediction under the investigated flow condition is the assumption about the catalysis at the surface. An open issue remains the rarefaction influence on the heat fluxes predicted by the CFD. The measured temperature distribution on the surface of the reference model (20mm radius of the leading edge) is reproduced in a coupled simulation applying a tailored catalysis model which predicts a non catalytic surface for the high temperatures in the stagnation region, a highly catalytic surface at lower temperatures on the longer plate and a smooth transition between both extreme states. Keeping the CFD model fixed, coupled simulations are conducted with the different structural models for highly conductive fibres, internal radiative heat transfer in a cavity of the model and the combination of both concepts. The influence on the peak heating is in all cases well predicted by the numerical simulations compared to the experiments. The rise in temperature at the beginning of the longer plate calculated for the cavity model is in good agreement with the experiment. In the cases with the highly conductive fibres, the experimental temperature level at that position cannot be reproduced. A possible source for the deviation is the unknown conductivity of the ceramic material with the highly conductive fibres at temperatures above 1200K. A first sensitivity analysis shows the influence of different extrapolations into the higher temperature regime. In case of the smaller leading edge radius (10mm), the agreement with the experiments is less convincing. The predicted tendencies are similar to the results achieved for the larger radius. But, the experimental values do not follow the same tendencies for both radii. The predicted range of peak temperature reductions is twice as large as measured in the experiments. Another observation is the excessive peak temperature of the clean configuration (without highly conductive fibres and without cavity). Assuming already a non catalytic wall at the high temperatures of the stagnation region, there is no significant sensitivity left in the parameter space of the CFD model to reduce the peak heat load sufficiently. This hints at the presence of a rarefied flow regime for which the heat load on a surface is known to be smaller than predicted under the assumption of continuum.