Heat transfer and pressure drop measurements in prototypic heat exchanges for the supercritical carbon dioxide Brayton power cycles

An experimental facility was built to perform heat transfer and pressure drop measurements in supercritical carbon dioxide. Inlet temperatures ranged from 30--125 °C with mass velocities ranging from 118--1050 kg/m2s and system pressures of 7.5--10.2 MPa. Tests were performed in horizontal, upward, and downward flow conditions to test the influence of buoyancy forces on the heat transfer.;Horizontal tests showed that for system pressures of 8.1 MPa and up standard Nusselt correlations predicted the heat transfer behavior with good agreement. Tests performed at 7.5 MPa were not well predicted by existing correlations, due to large property variations. The data collected in this work can be used to better understand heat transfer near the critical point. The CFD package FLUENT was found to yield adequate prediction for the heat transfer behavior for low pressure cases, where standard correlations were inaccurate, however it was necessary to have fine mesh spacing (y+∼1) in order to capture the observed behavior.;Vertical tests found, under the test conditions considered, that flow orientation had little or no effect on the heat transfer behavior, even in flow regions where buoyancy forces should result in a difference between up and down flow heat transfer. CFD results found that for a given set of boundary conditions a large increase in the gravitational acceleration could cause noticeable heat transfer deterioration. Studies performed with CFD further led to the hypothesis that typical buoyancy induced heat transfer deterioration exhibited in supercritical flows were mitigated through a complex interaction with the inertial force, which is caused by bulk cooling of the flow. This hypothesis to explain the observed data requires further investigation.;Prototypic heat exchangers channels (i.e. zig-zag) proved that the heat transfer coefficient was consistently three to four times higher as compared to straight channel geometry. However, the form pressure loss due to the presence of the corners within the channels caused an increase in pressure drop by four to five times the pressure drop measured in the straight channel. Based on the results, more innovative geometries were recommended for future testing to reduce form losses found in the typical prototypic geometries.