مجله هیدرولیک
سال هفتم شماره 1 (بهار 1391)

In this paper main factors associated with the accidents of Geisour-Gonabad water supply pipeline are investigated including continuous and transient pressures, water hammer control devices, air-valves, environmental conditions, and pipe material quality. The pipeline system is modeled for steady state condition, using WaterGems software, to calibrate Hazen-Williams roughness coefficient. A transient flow is initiated and the pressure heads are measured and compared with the computational pressures from Hammer software. More numerical tests indicate satisfactory performance of the system in the presence of surge control devices. Air-valves have no significant effect on transient flow. Laboratory tests demonstrate that water is neither corrosive nor scale forming. Finally, ensuring the satisfactory performance of protective devices, corrosion and low quality of pipe material are suggested as the predominant causes in the accidents of this pipeline. Estimating the maximum allowable length of corrosion and failure stresses based on ASME-B31-2009 code and comparing with the measured values confirm this assumption.

A study was carried out on the flow pattern around T-shaped submerged and deflecting spur dike located in 45 degree of 90 degree bend. Experiments were carried out in gentle bend with radius to width ratio of 4. To measure 3-D velocity and bed topography, Vectrino+ and point gage devices were used, respectively. First scour experiment was done and bed topography along overall bend was measured and after bed stabilization 3-D velocity was measured. Results indicated the presence of a secondary main clockwise flow and another secondary flow with opposite direction near the outer bank and surface water in first cross section (completely formed in cross section of 25o) and end bend (after end of spur dike influence on flow pattern). Presence of spur dike and change of bed topography leads to the change in secondary flow pattern in the cross section around spur dike. Observations indicated the presence of a separation interface in upstream spur dike and 0.3 depth flow level. In this separation interface two counterclockwise vortices were observed. These vortices are the main scour agent through creation of pressure fluctuations at the upstream wing edge of spur dike. Besides, a counterclockwise vortex with length as long as 2.5 times length of spur dike was observed downstream of spur dike. Analysis of results relevant to scour and flow pattern in cross and longitudinal sections, and different plans are among other points introduced in this paper.

Flood estimation using rainfall–runoff model is associated with uncertainty due to the uncertainty of input variables (i.e. rainfall), parameters of model (i.e. loss) and structure of the model itself. Estimation of design flood considering the uncertainty of all stochastic variables is an important issue during the design of hydraulic structures. This study aims to analyze uncertainty of estimated flood by using rain pattern generator (RPG) and HEC-HMS models using Monte Carlo simulation. Furthermore, error propagation of the input variables and parameters of HEC-HMS model into the estimated flood hydrograph were evaluated. The results showed the uncertainty of rainfall duration; rainfall pattern and losses are expected to incur the most noticeable effects, on the uncertainty of flood hydrograph.

Smoothed Particle Hydrodynamic (SPH), a new generation of mesh-free numerical methods to model free surface flow, is of great interest among many researchers nowadays. In this paper, ISPH was applied to simulate free-surface flow around hydraulic structures in several classical cases including dambreak, flow over a sharp-crested weir, and simultaneous operation of weir and gate. Due to enforcing incompressibility of flow, Cummins and Rudmans (1999) projection method is used in these cases. Error curves are used to evaluate the sensitivity of ISPH to approximate pressure terms. Two pressure formulations for approximating the left-hand-side of the Poisson equation result in a similar and acceptable trend of error. Comparison of several classical flow cases results using ISPH with experimental data and FVM-VOF shows good agreement between them.

Knowledge of technique of selective withdrawal is important in environmental hydrodynamic studies of dam reservoirs. The technique is used to obtain water of desired quality for drinking, industrial, and agricultural purposes, and is a water quality management tool of reservoirs.Considering the importance of the reservoir topography in selective withdrawal, in this paper the effects of side contraction on selective withdrawal of linearly stratified water from a point and a line sink in a reservoir is studied using Particle Image Velocimetry. The effects of contraction on propagation of shear waves and withdrawal layer thickness are discussed and the obtained critical Froude number is compared with the previous studies. The results show that the description of selective withdrawal in terms of motion of shear waves through contraction is a suitable method for predicting the critical Froude number.

In this research, Artificial Neural Network (ANN) was adapted to model the hydraulic jump length and conjugate depth of the stilling basin with convergent wall and inverse slope. More than 1682 experimental data on jump length and conjugate depth of hydraulic jumps for basins with convergent wall were used. The data was for rectangular section, with a -0.6%, -1.3%, -2% inverse slope and 2.7%, 4%, 5.3% convergent wall. In developing ANN, 15 configurations, each having different number of hidden layers and/or neurons, were investigated. The optimal models were capable of predicting jump length and conjugate depth of hydraulic jumps for a wide range of conditions. In each case, the configuration attained highest R2 value was selected as the optimal model. For stilling basin with convergent wall and inverse slope, the simplest ANN model for conjugate depth had a 6-3-1 configuration with 16 neurons in each of hidden layer and R2 =0.9974. The simplest ANN model for jump length had a 6-2-1 configuration, with 8 neuron in each of hidden layer and R2 =0.9865. The high values obtained for R2 in all cases, suggest close agreement between the ANN output and the experimental data.