Files in this item

FilesDescriptionFormat

application/pdf

application/pdfXIAO-DISSERTATION-2019.pdf (6MB)Restricted to U of Illinois
(no description provided)PDF

Description

Title:Non-equilibrium effects on in-tube condensation from superheated vapor
Author(s):Xiao, Jiange
Director of Research:Hrnjak, Pega
Doctoral Committee Chair(s):Hrnjak, Pega
Doctoral Committee Member(s):Jacobi, Anthony; Elbel, Stefan; Dutton, Craig; Radermacher, Reinhard
Department / Program:Mechanical Sci & Engineering
Discipline:Mechanical Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:Ph.D.
Genre:Dissertation
Subject(s):Condensation
Superheated vapor
Non-equilibrium
Abstract:In a vapor-compression system, refrigerant leaves compressor as superheated vapor. After entering the condenser, energy is transferred out of the refrigerant flow. As a result, the thermodynamically averaged temperature, namely the bulk temperature, of the refrigerant decreases. Conventionally when a 3-zone approach is adopted, as the specific enthalpy of the refrigerant flow becomes that of bulk quality one, the first droplet of condensate forms, indicating the onset of condensation. This approach is valid as long as the refrigerant flow is in thermal equilibrium. However, thermal equilibrium almost never happens during a heat transfer process due to the temperature gradient required by the Newton’s cooling law. During a condensation process, the vapor refrigerant has the highest temperature at the center of the tube and the lowest temperature at the tube wall. Hence, as the bulk temperature of the refrigerant decreases, the temperature of the inner tube wall drops to the saturation temperature before the refrigerant. Assuming that the tube is made out of hydrophilic material, which is typically the case in a vapor-compression system, liquid film forms as a result, regardless of the bulk temperature and specific enthalpy. Therefore, even though the specific enthalpy suggests that the refrigerant is still superheated vapor, the refrigerant flow is already two-phase flow. Similarly, when the bulk quality of refrigerant flow becomes zero, the liquid is already subcooled due to the temperature gradient. There has to be vapor left uncondensed to make the bulk temperature of the flow the saturation temperature. Therefore, even though the specific enthalpy suggests that the refrigerant is already subcooled, the refrigerant flow can still be two-phase flow. Two extra two-phase regions inside the traditionally defined single-phase regions occur because the two-phase flow is not in equilibrium during the condensation process. This phenomenon has a significant impact on important parameters such as void fraction, flow regime, pressure drop, and heat transfer. The previous understandings of the non-equilibrium effects on the condensation from superheated vapor can be improved, and methods for quantifying those effects are almost nonexistent. This work illustrates the effects of non-equilibrium on the condensation process, and modeling it using a 5-zone approach. The thesis first presents the flow visualizations from high-speed camera and film thickness measurements using a non-intrusive method to measure the film thickness in an operating facility. The flow visualizations and film thickness measurement confirm the existence of the two-phase regions in the conventionally defined single-phase regions. The film thickness measurements demonstrate the development of liquid film from the onset of condensation to lower specific enthalpies as condensation proceeds. The effects of mass flux and heat flux on the film thickness are also presented with R134a and R32 condensing in a 6 mm tube. The film thickness measurements are then converted into the void fractions, which are compared to some of the void fraction correlations. The large discrepancies between the predictions and experimental data are observed due to the fact that the void fraction starts to drop not at bulk quality one, but at the real onset of condensation. “Superficial quality” is proposed to better represent the mass fraction of vapor flow in the two-phase flow, and serves as a correction to the current void fraction correlations. The second part of the thesis is on the flow regimes determined from the flow visualizations. Effects of the mass flux, heat flux, refrigerant properties, and tube size on the flow regimes during condensation from superheated vapor are discussed. When comparing the experimentally determined flow regimes to a conventional flow regime map for condensation, three major defects of the flow regime map are identified. Firstly, there is no information on flow regimes beyond bulk quality one and zero, where there is two-phase flow. Secondly, the film generating mechanisms at the beginning of the condensation is not properly taken into account. This is evident because there are impossible flow regimes at the wrongly perceived “onset of condensation”. Thirdly, the flow transition criteria between different flow regimes are not clearly described; thus the correlations for the transition lines are highly empirical and cannot be used for a microchannel tube. A new flow regime map is introduced to remedy those defects. The third part of the thesis continues to explore the pressure drop during condensation from superheated vapor. After the onset of condensation, the interactions between the liquid and vapor create waves on the liquid-vapor interface, which dissipate more energy than a smooth tube wall. Consequently, the pressure drop rises according to the experimental data from different refrigerants condensing in different sized tube under varied mass fluxes. If thermal equilibrium is assumed, and a 3-zone approach is used to model the pressure drop, pressure drop continues to decrease right after the onset of condensation because the density of the flow keeps increasing. This mechanism is non-existent in a 3-zone approach. “Wave-enhancement factor” is proposed to fill the vacancy. A new mechanistic model for pressure drop is developed following development of the flow. It provides seamless transitions through all five regions, as well as more accurate results. This is because the two-phase flow, instead of single-phase flow, is taken into account, after and before the onset and end of condensation. The model is validated by experimental data from R32, R134a, and R1233zd(E) with heat fluxes of 5 to15 kW m-2 and mass flux of 100 to 400 kg m-2 s-1, showing that more than 90% of experimental data are within 10% of model predictions. The fourth part of the thesis demonstrates the effects of non-equilibrium on the heat transfer coefficient. The experimental data on heat transfer coefficient in the two extra regions demonstrates that the single-phase heat transfer correlations cannot involve the presence of latent heat, thus fail to provide accurate predictions. Besides, the discussions regarding the heat transfer behavior around bulk quality one are rare in the previous studies. “Film heat transfer coefficient” is proposed to represent the thermal resistance through the liquid film. Because it directly links the heat transfer coefficient to the film thickness, which can be calculated knowing the void fraction and flow regime, the film heat transfer coefficient is modeled and then converted back to the bulk heat transfer coefficient for consistency. A new mechanistic model for heat transfer is developed. It creates a seamless line across all five zones, unlike models using a 3-zone approach, because it takes into account the non-equilibrium effects. Additionally, the model is validated by experimental data from R134a, R32, R1234ze(E), R410A, as well as R744 condensing at mass fluxes from 100 to 200 kg m-2 s-1 and heat flux from 5 to 25 kW m-2, showing that most experimental data are within 15% of the model predictions. As for the effects of non-equilibrium on condenser design, there is an obvious benefit by using 5-zone approach and taking the non-equilibrium into account. Having higher heat transfer coefficient than single-phase transfer means it is possible to reduce the size of the condenser without losing capacity. The scenarios where the size reduction is more apparent, and the opposite, are demonstrated through sizing the condenser under simplified but realistic conditions. The results show that as the air-side heat transfer improves, non-equilibrium effects will play a bigger part in the sizing of a condenser. On top of the single component refrigerants, heat transfer coefficients of R32/R1234ze(E) mixture are also explored in the final part of the thesis.
Issue Date:2019-04-08
Type:Text
URI:http://hdl.handle.net/2142/104988
Rights Information:Copyright 2019 Jiange Xiao
Date Available in IDEALS:2019-08-23
Date Deposited:2019-05


This item appears in the following Collection(s)

Item Statistics