Proper design of high performance industrial heat transfer equipment relies on accurate knowledge and prediction of the thermal boundary conditions. In order to enhance the overall gas turbine efficiency, advancements in cooling technology for gas turbines and related applications are continuously investigated to increase the turbine inlet temperature without compromising the durability of the materials used. For detailed design, local distributions are needed in addition to bulk quantities. Detailed local distributions require advanced experimental techniques whereas they are readily available using numerical tools. Numerical predictions using a computational fluid dynamics approach with popular turbulence models are benchmarked against a semi-empirical correlation for the friction in a circular channel with repeated-rib roughness to demonstrate some shortcomings of the models used. Numerical predictions varied widely depending on the turbulence modelling approach used. The need for a compatible experimental dataset to accompany numerical simulations was discussed. An exact, closed-form analytical solution to the enhanced lumped capacitance model is derived. The temperature evolution in a representative 2D turbulated surface is simulated using Fluent to validate the model and its exact solution. A case including an interface contact resistance was included as well as various rib sizes to test the validity of the model over a range of conditions. The analysis was extended to the inter-rib region to investigate the extent and magnitude of the influence of the metallic rib features on the apparent heat transfer coefficients in the inter-rib region. It was found that the thermal contamination is limited only to the regions closest to the base of the rib feature. An experimental setup was developed, capable of measuring the local heat transfer distributions on all four channel walls of a rectangular channel (with aspect ratios between 1 and 5) at Reynolds numbers up to 150,000. The setup utilizes a transient thermochromic liquid crystals technique using narrow band crystals and a four camera setup. The setup is used to test a square channel with ribs applied to one wall. Using the transient thermochromic liquid crystals technique and applying it underneath high conductivity, metallic surface features, it is possible to calculate the heat transfer coefficient using a lumped heat capacitance approach. The enhanced lumped capacitance model is used to account for heat conduction into the substrate material. Rohacell and aluminum ribs adhered to the surface were used to tandem to validate the hybrid technique against the standard technique. Local data was also used to investigate the effect of thermal contamination. Thermal contamination observed empirically was more optimistic than numerical predictions. Traditional transient thermochromic liquid crystals technique utilizes the time-to-arrival of the peak intensity of the green color signal. The technique has been extended to utilize both the red and green color signals, increasing the throughput by recovering unused data while also allowing for a reduction in the experimental uncertainty of the calculated heat transfer coefficient. The over-determined system was solved using an un-weighted least squares approach. Uncertainty analysis of the multi-color technique demonstrated its superior performance over the single-color technique. The multi-color technique has the advantage of improved experimental uncertainty while being easy to implement.
| Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:etd-5662 |
| Date | 01 January 2014 |
| Creators | Tran, Lucky |
| Publisher | STARS |
| Source Sets | University of Central Florida |
| Language | English |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Electronic Theses and Dissertations |
Page generated in 0.003 seconds