MODELING TWO-PHASE FLOWS USING EFFICIENT MULTI-SCALE PHASE CHANGE CLOSURE METHODS: CONTACT LINE EFFECT

Framework and objectives

The high integration and densification in power electronics and new carbon-free engine (due higher adiabatic flame temperature) in aeronautics and transport present increasing challenges for more efficient heat control and cooling means for which phase change systems are a good candidate. One of the possible solutions to be used are passive two-phase loops, in which heat transfer is associated with the change of state of the fluid. It consists of a closed duct partially filled with a liquid in a close-to-thermodynamic-equilibrium state with its vapor phase.

Once charged, i.e., heated from a side called the evaporator and cooled from another called the condenser, the fluid onsets a motion due to phase change and buoyancy and/or capillarity, depending on the technology and application field targeted. The motion insures the super-heat transfer capabilities of such a device. The main issue to address is usually how to design the device in a way that it reached its lowest thermal resistance, shortest response time, lowest noise, highest resilience to failure and largest working stability. Flexibility may also be a variable to consider. To assess such a device, modeling effort has to be made and worth noticing that robust and reliable models at the device scale are quite scarce. Nevertheless, Pprime Institute has put considerable efforts to develop a two-phase flow model with either mixed or separated phases. These models were able to simulate a two-phase gravity loop. This has been possible in a first step using a one-dimensional mixture model and lastly using a three-dimensional segregated model.

Work program and means

Yet, many issues remain unanswered to date. First, the present model is extremely computer resources consuming. Although the model being dimension-free, performing a full three-dimensional simulation of the whole loop is at the present time out of reach, indeed.

Therefore, optimized algorithms for efficient interface advection are sought for. Then, liquid film and/or vapor film dynamics close the contact line (or micro-region) is lacking in the present model. Bearing in mind that the heat transfer is enhanced considerably close to this line, an appropriate representation of this phenomenon is of paramount importance and is definitely a step forward to a more realistic loop model. In addition, a more accurate scheme has to be developed in order to enhance the mass conservativeness in compressible flow context. Finally, some minor improvements have to be done in order to take into account Kelvin effect for example, i.e., the departure from saturation conditions whenever the interface is no longer planar, the recoil pressure, etc. Test cases for validation are available from both Pprime Institute experimental data and literature. Using the developed tool, original loop can be designed and assessed afterward.

This subject has two essentially components. Its first step is to optimize the detailed model in order to better simulate the behavior of the two-phase real in reasonable computation time. In a second step, an analytical model will be developed describing the dynamics of phase change close to the contact line. This model will be implemented as a more-accurate boundary condition that feeds the numerical model with data unreachable otherwise.

Pprime Institute (P’) is a renowned research laboratory specialized in the fields of Physics and Engineering Sciences. It is a CNRS institution linked with the University of Poitiers (The Faculty of Fundamental and Applied Sciences, ENSIP (National Higher School of Engineers of Poitiers)) and ISAE-ENSMA (National Higher School of Mechanics and Aerospace Engineering).

Applicant profile, prerequisites

The attendee will have:

1. Good knowledge of PDE for fluid mechanics/heat and mass transfer

2. Good knowledge of numerical methods/computation

3. Good practice of symbolic calculus

4. Good knowledge of Python scripting language

5. Good written communication skills in English

Appreciated (but not mandatory) skills are:

1. Knowledge of asymptotic developments

2. Basic knowledge of openFOAM library

3. Knowledge of C or C++ language