Direct simulation of hot jet in cold channel crossflow with adiabatic thermal boundary conditions
Description
This is the data of a direct simulation of a hot jet in cold channel crossflow. The adiabatic thermal boundary conditions are applied to the top and bottom wall. Fo details about the computational configurations, methods and results, please refer to "Direct Numerical Simulation of a Low Momentum Round Jet in Channel Crossflow" which has been accepted to Nuclear Engineering and Design, or "DNS of a Jet in Cross Flow with Passive Scalar Mixing" in Turbulence, Heat and Mass Transfert 8, 2015 (see links below) We also have the data of the same configuration but with iso-thermal boundary condition or conjugate heat transfer. A paper discussing the difference of the thermal fields under different thermal b.c. has been submitted to the International Journal of Heat and Mass Transfer. A paper on validating selected RANS models using the current database is under preparation. A paper discussing the coherent structures of current flow using proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD) is under preparation. We have the 3D time-averaged statistics of mean velocity, Reynolds stress, heat fluxes, temperature fluctuations and their budgets. We also have more than 4000 full resolution instantaneous snapshots for each case (only 200 for the adiabatic case). The complete dataset is about 20TB!!! More data will be released in the future. Please feel free to contact me if you need more information or more data. The authors acknowledge the use of ARCHER HPC allocated via UK EPSRC Turbulence Consortium (EPSRC grant EP/L000261/1) and the use of BlueGene/Q supercomputer sponsored by EDF R&D centre Chatou. The authors are grateful to Cedric Flageul for help in implementing heat transfer in the DNS code and Sofiane Benhamadouche for valuable discussions and advice. E-mail: zzwz02@163.com or zhao.wu@manchester.ac.uk
Files
Steps to reproduce
The data is the time-averaged statistics and stored in MATLAB format (*.mat). The data are stored every 4 original simulation nodes and contains 52*49*64 nodes. The corresponding x, y and z coordinate are also stored in the .mat file. If the data point is the exact point stored in the data, just use the matrix to get the value. The data at the point (x,y,z)=( x(i), y(j), z(k) ) is stored at (j,i,k) in the time-averaged statistic matrix (x and y are swapped). For example, if you want the time-averaged u velocity (um) at (x, y, z)=(1.5, 0.3203, 0). First look at the vectors 'x' and we found x=1.5 corresponds to the 25th element in the vector 'x'. And y is the 17th, z is the 32nd, i.e. (x, y, z)=(1.5, 0.3203, 0)=( x(25), y(17), z(32) ). Then um(17,25,32) contains the data at this point. If the data point is not the exact point stored in the data, you should interpolate the data.
Institutions
- The University of Manchester School of Mechanical Aerospace and Civil Engineering