FLOW-PIPE-SEEPAGE COUPLING ANALYSIS OF SPANNING INITIATION OF A PARTIALLY-EMBEDDED PIPELINE

The initiation of pipeline spanning involves the coupling between the flow over the pipeline and the seepage-flow in the soil underneath the pipeline. The pipeline spanning initiation is experimentally observed and discussed in this article. It is qualitatively indicated that the pressure-drop induced soil seepage failure is the predominant cause for pipeline spanning initiation. A flow-pipe-seepage sequential coupling Finite Element Method (FEM) model is proposed to simulate the coupling between the water flow-field and the soil seepage-field. A critical hydraulic gradient is obtained for oblique seepage failure of the sand in the direction tangent to the pipe. Parametric study is performed to investigate the effects of inflow velocity, pipe embedment on the pressure-drop, and the effects of soil internal friction angle and pipe embedment-to-diameter ratio on the critical flow velocity for pipeline spanning initiation. It is indicated that the dimensionless critical flow velocity changes approximately linearly with the soil internal friction angle for the submarine pipeline partially-embedded in a sandy seabed.

CFD analysis of pneumatic conveying in a double-tube-socket (DTS (R)) pipe

A pipeline with a bypass is widely used for the pneumatic conveying of material. The double-tube-socket (DTS (R)) technology, which uses a special inner bypass, represents the current state of the art. Here, a new methodology is proposed based on the use of computational fluid dynamics (CFD) to predict the energy consumption of DTS conveying. The predicted results are in good agreement with those from pilot-scale experiments. (C) 2010 Elsevier Inc. All rights reserved.

A 2-kW COIL with a square pipe-array JSOG and nitrogen buffer gas

A 2-kW-class chemical oxygen-iodine laser (COIL) using nitrogen buffer gas has been developed and tested since industrial applications of COIL devices will require the use of nitrogen as the buffer gas. The laser, with a gain length of 11.7 cm, is energized by a square pipe-array jet-type singlet oxygen generator (SPJSOG) and employs a nozzle bank with a designed Mach number of 2.5. The SPJSOG has advantages over the traditional plate-type JSOG in that it has less requirements on basic hydrogen peroxide (BHP) pump, and more important, it has much better operational stability. The SPJSOG without a cold trap and a gas-liquid separator could provide reliable operations for a total gas flow rate up to 450 mmol/s and with a low liquid driving pressure of around 0.7 atm or even lower. The nozzle bank was specially designed for a COIL using nitrogen as the buffer gas. The cavity was designed for a Mach number of 2.5, in order to provide a gas speed and static temperature in the cavity similar to that for a traditional COIL with helium buffer gas and a Mach 2 nozzle. An output power of 2.6 kW was obtained for a chlorine flow rate of 140 mmol/s, corresponding to a chemical efficiency of 20.4%. When the chlorine flow rate was reduced to 115 mmol/s, a higher chemical efficiency of 22.7% was attained. Measurements showed that the SPJSOG during normal operation could provide a singlet oxygen yield Y greater than or equal to 55%, a chlorine utilization U greater than or equal to 85%, and a relative water vapor concentration w = [H2O]/([O-2] + [Cl-2]) less than or equal to 0.1.

Modelling of Unsteady Gas-Dynamic Flow in a Pipe and at its Exit Using CFD

Traditionally the simulation of the thermodynamic aspects of the internal combustion engine has been undertaken using one-dimensional gas-dynamic models to represent the intake and exhaust systems. CFD analysis of engines has been restricted to modelling of in-cylinder flow structures. With the increasing accessibility of CFD software it is now worth considering its use for complete gas-dynamic engine simulation. This paper appraises the accuracy of various CFD models in comparison to a 1D gas-dynamic simulation. All of the models are compared to experimental data acquired on an apparatus that generates a single gas-dynamic pressure wave. The progress of the wave along a constant area pipe and its subsequent reflection from the open pipe end are recorded with a number of high speed pressure transducers. It was found that there was little to choose between the accuracy of the 1D model and the best CFD model. The CFD model did not require experimentally derived loss coefficients to accurately represent the open pipe end; however, it took several hundred times longer to complete its analysis. The best congruency between the CFD models and the experimental data was achieved using the RNG k-e turbulence model. The open end of the pipe was most effectively represented by surrounding it with a relatively small volume of cells connected to the rest of the environment using a pressure boundary.