Julio Cesar Pacio

Modelling Two-Phase Heat Exchanger Performance in the Annular-Mist Flow Regime Considering Entrainment and Deposition Phenomena

Two-phase flow heat exchangers are main components of large cryogenics, power generation, refrigeration and liquefaction of natural gas plants, both in terms of capital cost and technical challenges. A major challenge in their design is the prediction of local heat transfer coefficients and pressure gradients for the evaporating or condensing fluids. Traditional heat exchanger models are based on one single correlation for predicting the heat transfer in the entire saturated boiling regime, disregarding the flow structure. However, the structure of the flow dictates how the different physical processes (nucleate boiling, convective heat transfer to the liquid and vapour phase, thin film evaporation) interact and contribute to the total heat transfer. In particular, a relevant flow-regime transition for the sizing of heat exchangers is the occurrence of dryout during the evaporation process in the annular-mist flow regime. The objective of this work is to present a three-field model for describing the annular-mist flow considering a liquid film, liquid droplets and a vapor phase, and predicting the occurrence of dryout. The flow structure is affected by the entrainment, deposition and evaporation. These processes are studied on the base of semi-empirical models. The final mathematical model is implemented into an in-house solver. The model is validated with uniform heat flux data available in the open literature. While the model performs well in the case of water flows (within 10% error), the uncertainties are larger for other fluids, probably due to the applicability range of the empirical models. Finally, two numerical examples considering the sensitivity of the input parameters and axial power distribution are studied.Copyright

Sensitivity Analysis of Heat Exchangers Using Perturbative Methods

The solution of heat exchanger models is usually influenced by different parameters related to fluid properties, geometry and flow conditions. This implies that several simulations are usually required for design and optimization purposes. In this work, a sensitivity analysis using perturbative methods is presented as an alternative for reducing the computational cost of studying the sensitivity of the solution to different parameters. The method is based on the computation of the adjoint problem, and thus only one extra simulation is required for studying the sensitivity of an integral response, regardless of the number of parameters. Both the direct and the adjoint problem were solved using a least squares formulation.