Experimental data Koen Blanckaert – Explanations
______
The explanations concern two laboratory experiments in the same laboratory flume, one with flat bed (named Q89, or also F16_90_00) and one with mobile bed (named M89, or M16_90_00). Both experiments have been simulated numerically by teams in Iowa (Zeng, Constantinescu) and Delft (van Balen).
Flat Bed experiment
There are two matlab workspaces containing the data:
-FlatBed_flowdepth.mat: this workspace contains data on the flow depth. Although the bottom is horizontal, the flow depth is not constant, because there is a transversal inclination of the water surface in the bend.
-FlatBed_expdata.mat: this workspace contains all the flow data:
- Measurements have been done in 12 cross-sections around the flume:
- Cross-section in the straight inflow reach at 2.5m and 0.5 m upstream of the bend. These are indicated by the appendixes _m25 and _m05 in the names of the variables
- Cross-sections in the bend at 15°, 30°, 60°, 90°, 120°, 150° and 180°, also indicated by the appendix in the name of the variables (for example xxxxx_090)
- Cross-sections in the straight outflow at 0.5 m, 1.5 m and 2.5 m downstream of the bend, also indicated by the appendix in the name of the variables.
- In each of these cross-sections, vertical profiles were measured at transverse locations indicated by the variable n_Q89_crosssection. The variable n_Q89_120, for example, gives the transverse locations in the cross-sections at 120° in the bend. By definition, n = 0 at the centreline, n = -0.65 at the inner bank and n = 0.65 at the outer bank.
- Each vertical profile consists of grid point at elevations given by the variable zn_Q89_crosssection. These vertical elevations are given in normalized form: zn = 0 at the bed and zn = 1 at the water surface. The dimensional elevations in m are obtained by multiplying zn with the local flow depth.
- The mean velocity has three components, which I call in my papers (vs, vn, vz). They are given in variables that are called V_corr_Q89_crosssection. V_corr_Q89_090, for example, contains the velocity data in the cross-section at 90° in the bend. This variable is an array of size 51 x 50 x 3. Here, 51 stands for the vertical dimension, and corresponds to the size of zn_Q89_090. In the same sense, 50 stands for the size of n_Q89_090 and indicates the transverse grid. And the last dimension, 3, stands for the three velocity components (vs, vn, vz).
- In the same sense, the variable vii_Gcorr_Q89_crosssection contains the turbulent normal stresses. They are ordered as
- In the same sense, the variable vij_Gcorr_Q89_crosssection contains the turbulent shear stresses. They are ordered as
- Near the water surface, the measurements are of bad quality. I have removed these measurements, or replaced them by extrapolated values. The variables section_V_Q89_crosssection indicate which data points I have removed in the mean velocities, and section_v_G_Q89_crosssection indicate the removed points in the turbulent stresses. But there seem to be an error in these variables. Anyway, these variables are not essential for you.
Mobile Bed experiment
There are two matlab workspaces containing the data:
-surface_bottom_results_mob_interp_2008.mat: this workspace contains the bed level and the water surface level. They are given in spline format, which allows you to extract them on the grid of your choice. You can extract the data and make plots with the file “M16_bed_surface_figures.m”. This file uses functions that are included in the old matlab toolbox “splines_Matlab5”. You will have to add this toolbox to the path of matlab. Attention, I have notices that there is an error in the name of the workspace to load, line 6 of the program.
-MobielBed_expdata.mat: this workspace contains the experimental data. The format is essentially the same as for the flat bed experiment. All data have been resampled on the same vertical grid called zn_sigma.