The Development of a Non-Equilibrium Dispersed Flow Film Boiling Heat Transfer Modeling Package
Open Access
- Author:
- Meholic, Michael James
- Graduate Program:
- Nuclear Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 02, 2011
- Committee Members:
- Fan Bill B Cheung, Dissertation Advisor/Co-Advisor
Fan Bill B Cheung, Committee Chair/Co-Chair
Arthur Thompson Motta, Committee Member
Kostadin Nikolov Ivanov, Committee Member
Cengiz Camci, Committee Member - Keywords:
- mechanistic modeling
DFFB
heat transfer
film boiling
dispersed flow
post-CHF
COBRA-TF - Abstract:
- The dispersed flow film boiling (DFFB) heat transfer regime is important to several applications including cryogenics, rocket engines, steam generators, and in the safety analysis of nuclear reactors. Most notably, DFFB is responsible for the heat transfer during the blowdown and reflood portions of the postulated loss-of-coolant accident (LOCA). Such analyses require the accurate predictions of the heat transfer resulting from the non-equilibrium conditions present in DFFB. A total of six, interrelated heat transfer paths need to be modeled accurately in order to quantify DFFB heat transfer. Within the nuclear industry, transient safety analysis codes, such as COBRA-TF, are used to ensure the safety of the reactor under various transient and accident scenarios. An extensive literature review of DFFB heat transfer highlighted a number of correlative, phenomenological, and mechanistic models. The Forslund-Rohsenow model is most commonly implemented throughout the nuclear industry. However, several of the models suggested by Forslund and Rohsenow to model DFFB phenomena are either inapplicable for nuclear reactors or do not provide an accurate physical representation of the true situation. Deficiencies among other DFFB heat transfer models in their applicability to nuclear reactors or in their computational expenses motivated the development of a mechanistically based DFFB model which accounted for each heat transfer mechanism explicitly. The heat transfer resulting from dispersed droplets contacting the heated wall in DFFB was often neglected in previous models. In this work, a first-principles approach was implemented to quantify the heat transfer attributed to direct contact. Lagrangian droplet trajectory calculations incorporating realistic radial vapor velocity and temperature profiles were performed to determine if droplets could contact the heated wall based upon the local conditions. These calculations were performed over a droplet size spectrum accounting for various droplet diameter effects. When contact was achieved, the heat transfer was quantified by coupling the mass flux of droplets contacting the wall with a mechanistic direct contact heat transfer coefficient. Unlike currently used methods, the proposed DFFB model shows the correct trends with respect to local vapor mass flux and wall superheats. The proposed DFFB model also extends previous models to predict the radiative and interfacial heat transfer. A nodal radiative heat flux model was adapted to account for the radiative heat transfer among the droplet size spectrum. The interfacial heat transfer provided by the dispersed droplets was predicted by adapting the Lee-Ryley model to account for the varying droplet velocity and interfacial area across the droplet size spectrum in addition to the vapor temperature distribution. A novel method was developed as part of the Lagrangian trajectory calculations to capture the convective enhancement due to the dispersed droplets intermittently altering the vapor temperature distribution. Following the development of the proposed DFFB heat transfer model, it was implemented within the existing framework of COBRA-TF. Effective heat transfer coefficients were utilized as a means of preserving the DFFB heat transfer contributions within COBRA-TF. The proposed DFFB model was assessed using data from 118 steady-state experiments in four separate facilities. Comparisons between the predictions of COBRA-TF utilizing the original and proposed DFFB models show significant improvements. Over all of the assessment cases, the proposed DFFB model reduced the root mean square error by 32.55 K. More significantly, the proposed model was more precise as it reduced the standard deviation in the wall temperature prediction error by 32.34 K. Throughout the assessment cases, the proposed DFFB model predicted both the magnitude and trend seen in the experimental data better than the original model based upon Forslund-Rohsenow. These results provided confidence that the proposed model reflects the underlying physics governing the DFFB heat transfer components and can be used in place of existing models in nuclear reactor safety analysis codes as a means to improve the accuracy in current large break LOCA simulations that include DFFB conditions. Overall, the proposed DFFB model marks a step-change in the use of mechanistically based DFFB models in reactor safety analysis codes. It also provides a functional modeling package which improves the predictive capabilities of nuclear reactor safety-analysis codes and serves as a starting point for future research and development in DFFB.