Experimental and numerical investigation of flow distribution pattern at a T-shape roadway crossing under extreme storms

Figures & data

Figure 1. Schematic of the experimental system.

Figure 2. Site photo of the experimental setups.

Figure 3. Boundary conditions of CFD model.

Figure 4. Computational mesh in the horizontal and vertical directions.

Table 1. Factors and levels selected for the 2^k factorial design experiments.

Factors	Low level (−1)	High level (+1)
Qu(m^3/h)	10	40
Sy(%)	0.3	5
Sx(%)	1	2
R(m)	1.3	3.2

Table 2. Setups of the 2^k factorial design experiments.

ID	Qu(m^3/h)	Sy(%)	Sx(%)	R(m)
1	10	0.3	1	1.3
2	10	0.3	1	3.2
3	10	0.3	2	1.3
4	10	0.3	2	3.2
5	10	5	1	1.3
6	10	5	1	3.2
7	10	5	2	1.3
8	10	5	2	3.2
9	40	0.3	1	1.3
10	40	0.3	1	3.2
11	40	0.3	2	1.3
12	40	0.3	2	3.2
13	40	5	1	1.3
14	40	5	1	3.2
15	40	5	2	1.3
16	40	5	2	3.2

Table 3. Parameter combinations used in the CFD model validation tests.

Longitude Slopes (Sy)	Upstream inflow flowrate Qu (m^3/h)
1%	24.85	46.55	4.99
3%	24.22	16.47	4.81
5%	24.30	14.88	4.48

Figure 5. Comparison of experimental results and numerical predictions: (a) flow distribution ratio; (b) water depth, where A1 is 1% longitude slope, A2 is3% longitude slope, and A3 is 5% longitude slope for all roadways.

Figure 6. Comparison of water depth measured experimentally and predicted numerically along the channel.

Figure 7. Pareto chart results of 2^k factorial design experiments, where the dash line represents the threshold that a variable is considered to be important to the flowrate distribution ratio.

Table 4. Factors and levels used in the multi-level full sequence simulations.

Factors	Three levels of factor values	Description
R(m)	1.3, 2, 2.6, 3.2	Roadway turning radius
Sy(%)	0.3, 1, 3, 5	Roadway longitude slope
Qu(m^3/h)	10, 20, 30, 40	Upstream inflow flowrate

Figure 8. Relationship of distribution ratio R_q versus inflow flowrate Q_u at different turning radius R when longitude slope S_y is fixed.

Figure 9. Numerically simulated streamlines at the road crossing under different combinations of turning radius R and longitudeslope S_y.

Figure 10. Comparison of Flowrate distribution ratio from this study and Ghostine’s study on flat-bottom diversion channels (Ghostine et al., 2013).

Figure 11. Comparison of streamlines at S_u= S_b= S_d= 1%, (a), (c), (e) are road crossings with side slope and a turning radius of 0.6 m; (b), (d), (f) is the flat bottom diversion channels.

Figure 12. Relationship of distribution ratio R_q versus inflow flowrate Q_u at different longitude slope S_y when turning radius R is fixed.

Figure 13. Relationship of distribution ratio R_q versus inflow flowrate Q_u at different downstream standing water depth in the main road.

Figure 14. Relationship of distribution ratio R_q versus inflow flowrate Q_u at different downstream standing water depth in the main road and branch road.

Ghostine, R., Vazquez, J., Terfous, A., Riviere, N., Ghenaim, A., & Mose, R. (2013). A comparative study of 1D and two-dimensional approaches for simulating flows at right angled dividing junctions. Applied Mathematics and Computation, 219(10), 5070–5082. https://doi.org/10.1016/j.amc.2012.11.048

 Web of Science ®Google Scholar