Conference Agenda

A new experimental setup has been constructed for a cold swirling turbulent jet issued through a pressure-swirl atomizer generating a spray at Reynolds number up to Re=10⁶.

The cold spray injection is used for steam condensation and pressure regulation in the pressurizer of a pressurized water reactor (PWR).

For large spray flowrates the droplets also reach the pressurizer tank wall, which acts as an undesired thermal load.

Current simplified prediction tools for transient load calculations lead to conservative estimations of the loads and under predicted lifetime.

More advanced tools, e.g. computational fluid dynamics (CFD) require better models for two-phase flow heat transfer in order to get more reliable lifetime predictions in a long term operation (LTO) context.

Measurements have been conducted in a 1:1.84 lab scale model of the spray two-phase flow characterizing the liquid fraction, droplet size and velocity distributions dependence on the spray flow rate and surface tension.

The spray cooling heat transfer has also been measured using a unique heat transfer sensor developed at CEA in France.

The experimental data base will be used for validation of more advance CFD models, being developed in conjunction to the present experimental campaign.

With the development of surface engineering, porous surfaces have emerged as a significant research subject in boiling heat transfer. The latter, in turn, plays a crucial role in various industries such as power plants, distillation plants, and microelectronic technology. In this paper, the Molecular Dynamics method is adopted to investigate the wicking dynamics and boiling dynamics of two porous surfaces: foam, which exhibits randomly distributed pores, and mesh, composed of ordered square wires with relatively uniform pore sizes. Three wettability, namely hydrophilic, neutral, and hydrophobic wetting states, are assigned to the two porous surfaces de-coupling the effect of wettability from surface structure. Results reveal that, during the wicking process, the foam surface shows better wetting ability as it absorbs liquid under both hydrophilic and neutral wettability. Comparatively, the mesh surface has the fastest wicking speed under hydrophilic wettability yet it becomes non-wetting under neutral wettability. During the boiling process, the boiling dynamics differ greatly under three wettability. More importantly, the difference in surface structure makes the foam surface possess a better heat transfer whereas the mesh surface causes gentle pressure variation. Our findings provide insights into the design of artificial porous surfaces for certain purpose and their potential application.

Water entrainment at T-junctions is of upmost importance in some nuclear safety analyses. Such a phenomenon directly impacts the core liquid inventory, hence its coolability. Numerical modelling using system-scale codes is essential for characterizing the two-phase flow interactions at the T-junction and in the branch line upstream. In light of this, code validation must be carried out through comparison to experimental data (Separate Effect Tests).

The test section represents, at a lower scale compared to reactor scale, an upper core plenum, a hot leg, and a pressuriser surge line. The test loop is operated at atmospheric conditions. Thus, the transposition issue (geometry and thermal hydraulics conditions) also has to be tackled.

The aim of this paper is to present the set of calculations, the comparison to experimental data and the scaling approach through the confrontation of CFD and system-scale code predictions.

The system-scale computations are performed with the thermal hydraulics code CATHARE and the CFD calculations with NEPTUNE_CFD, an in-house code.

CATHARE and NEPTUNE_CFD results are first compared to the experimental data, both, qualitatively (video recording) and quantitatively (water height) for two configurations (vertical upward and inclined T-junctions). This allows an assessment of the codes’ accuracy regarding the phenomenon of water entrainment at a T-junction and raises reflections on the physics, and the modelling.

Then, CATHARE and NEPTUNE_CFD calculations are performed at reactor scale and thermal hydraulics conditions. It is assumed that CFD better copes with scaling and is taken as a reference for the code-to-code comparison.

Finally, future work is proposed.

The efficient heat dissipation of the permanent magnet propulsion motor rotor is crucial to the development of advanced propulsion systems. As an advanced thermal management technology, the rotating heat pipe enables effective cooling of rotating components through internal phase-change heat transfer and natural circulation. Based on this research background, our team has built a rotating heat pipe experimental system and conducted experiments with a 70% filling ratio, a length of 500mm, and a diameter of 30mm using a stepped rotating heat pipe. The results show that its heat transfer capability gradually increases with the rotational speed. Under the same conditions, the heat transfer performance of the parallel-axis rotating heat pipe reached twice that of the coaxial rotating heat pipe, with an equivalent thermal conductivity of up to 1438.08 W/(m·K). This study provides experimental data support for the application of rotating heat pipes in the cooling of permanent magnet propulsion motor rotors.

The pool boiling experiment for the observation of corrosion products deposition is carried out to better understand the fouling mechanism under nucleate boiling conditions. The experimental apparatus comprises the quartz glass cavity, test piece (aluminum), high-speed camera and heater set-up. The deposition tests are performed in dilute colloidal solution (Fe₃O₄) with different wall temperature and bulk temperature at atmospheric pressure. The experimental observations indicate that the deposits exhibit a circular distribution and a thickness of approximately a few micrometers under nucleate boiling. The fouling ring is distinguished by a lower central thickness and a higher edge thickness. To gain further insight into the flow field distribution during the bubble growth process, the numerical simulation of the bubble growth and detachment process is conducted using the CFD method. It has been demonstrated that corrosion products are transported to the contact line of the bubble as a consequence of turbulence vortex. Besides, the micro-layer situated at the base of the bubble will undergo a process from thinning to drying out, resulting in the deposition of corrosion products on the heated surface. Through a combination of experimental and numerical techniques, the transport mechanism of corrosion products under the influence of nucleate bubbles has been elucidated, and the model between the evaporation flux and deposition rate of corrosion products has been developed.

The Best Estimate Plus Uncertainty (BEPU) methodology, developed over several decades, has seen numerous innovations aimed at enhancing the efficiency and quality of the BEPU procedure. Wilks’ formula characterized by nonparametric statistics is widely used for uncertainty evaluation, while it is time consuming. Deterministic sampling (DS) methodology assesses the uncertainty of outputs through the first two orders of moments of the input uncertain parameters. The reduction in computational effort achieved by using fewer sampling times, compared to the Wilks method, presents a promising alternative for enhancing the BEPU methodology. 16 input parameters and 3 safety-related output parameters as the Figure of Merits (FoMs) are chosen in ESBWR initiated by main steamline break for BEPU evaluation using RELAP5. First order Wilks’ method and three DS (DS-Standard, DS-Simplex, and DS-Hadamard) methods are applied. Subsequently, a preliminary sensitivity analysis of the Wilks’ results is performed to identify the input parameters with a significant impact on the FoMs. The downscaled parameters were then used as inputs for BEPU calculations using three DS methods. The degree of envelopment and conservatism of the three results (Wilks’ results with 16 input parameters, three DS results with 16 input parameters, and three DS results with downscaled parameters) relative to the experimental data were compared to determine whether the downscaled input results could be considered valid for the final BEPU analysis under the given conditions.