Description of the archive files
The cold leg mixing benchmark exercise has been proposed by OECD/NEA within the framework of Verification, Validation and Uncertainty Quantification (VVUQ) of computational fluid dynamics (CFD) codes. The corresponding experimental facility has been established by Texas A&M University. It consists of two vessels connected by a leg, including a valve separating at the initial state two isothermal mixtures of a light and a heavy fluid. When the valve is opened, the density gradient causes the heavy fluid to flow into the cold leg and to discharge into the light vessel. Simultaneously, an opposite flow of the light mixture appears in the cold leg. The facility mimics the critical scenario of a pressurized thermal shock in a reactor vessel, where the accurate prediction of temperature stresses is essential to avoid a rupture of the vessel.
The archive files available on this page are extracted from the large eddy simulations of the cold leg mixing benchmark conducted with the open-source code TrioCFD. Several simulations were performed using on different grid refinements (meshes M1 to M4) in order to check the mesh convergence. The value of the Atwood number under consideration corresponds to a density change of 10% between the heavy and the light fluids. As the fluids are miscible, the problem was treated as a single-phase flow. Two physical modelings were investigated: the first one uses a uniform density and accounts for buoyancy effects by means of the Boussinesq approximation. The second one is a low-Mach (quasi-compressible) model where density and viscosity are supposed to vary linearly in function of concentration. The physical quantities of interest are the components of time-average and RMS velocity, and the tangential component of the Reynolds-stress tensor.
Publication
These results are presented in the following publication:
Angeli, PE. Wall-Resolved Large Eddy Simulations of the Transient Turbulent Fluid Mixing in a Closed System Replicating a Pressurized Thermal Shock. Flow Turbulence Combust (2021). https://doi.org/10.1007/s10494-021-00272-z
Content
1) The files are stored in three directories:
Straight_probes_cold_leg : vertical profiles in the cold leg
Straight_probes_downcomer : horizontal profiles in the downcomer
Circular_probes_downcomer : circular profiles in the downcomer
2) The prefix in the file names are:
VD4 : solutions on mesh M4 of the Variable-Density model defined by equations (17)-(19) in the article
BA4 : solutions on mesh M4 of the Boussinesq model defined by equations (21)-(23) in the article
3) The locations of the profiles are:
x1, x2, x3, x4, x5, x6, x7, x8 : position of vertical lines in the cold leg (Fig. 14 of the article corresponds to x4) with
x1 <-> x = 70.1156 cm
x2 <-> x = 72.5600 cm
x3 <-> x = 75.0044 cm
x4 <-> x = 77.4489 cm
x5 <-> x = 79.8933 cm
x6 <-> x = 82.3378 cm
x7 <-> x = 84.7822 cm
x8 <-> x = 87.0044 cm
y1, y2, y3, y4, y5, y6, y7, y8 : position of vertical lines in the downcomer (Fig. 16 and Fig. 19 of the article correspond to y4) with
y1 <-> y = -14.5502 cm
y2 <-> y = -15.5536 cm
y3 <-> y = -16.5571 cm
y4 <-> y = -17.5605 cm
y5 <-> y = -18.5640 cm
y6 <-> y = -19.5674 cm
y7 <-> y = -20.5709 cm
y8 <-> y = -21.5743 cm
4) The files have two columns:
First column = coordinate (y in [m] for files in 'Straight_probes_cold_leg', x in[m] for files in 'Straight_probes_cold_leg', theta [in deg] for files in 'Circular_probes_downcomer')
Second column = value of the field (WITHOUT nondimensionalization)
5) The name of the field is included in the file name:
* umean, vmean, wmean : components of the average velocity in x, y, and z directions respectively (averaging over the time interval [6.38 s ; 16.38 s] for the cold leg, and [10.84 s ; 20.84 s] for the downcomer)
* usd, vsd, wsd : components of the standard deviation of velocity in x, y, and z directions respectively (same time averaging as previoulsy)
* cmean, csd : average and standard deviation of the mass fraction ('phi' in the article)
Contact
For any question or request, please contact: pierre-emmanuel.angeli@cea.fr
Last update of the archive: 2021/02/17
Video of flow evolution
Evolution in real time of the mass fraction phi for 0 < t < 20.84 seconds (quasi-compressible model, mesh M4):