The aim of this study is to investigate the effects of surface roughness sizes on the discharge coefficient for a broad crested weirs. For this purpose, three models having different lengths of broad crested weirs were tested in a horizontal flume. In each model, the surface was roughed four times. Experimental results of all models showed that the logical negative effect of roughness increased on the discharge (Q) for different values of length. The performance of broad crested weir improved with decrease ratio of roughness to the weir height (Ks/P) and with the increase of the total Head to the Length (H/L). An empirical equation was obtained to estimate the variation of discharge coefficient C d in terms total head to length ratio, with total head to roughness ratio.

In this paper, constructive learning is used to train the neural networks. The results of neural networks are obtained but its result is not in comprehensible form or in a black box form. Our goal is to use an important and desirable model to identify sets of input variable which results in a desired output value. The nature of this model can help to find an optimal set of difficult input variables. Accuracy. Genetic algorithms are used as an interpretation of achieving neural network inversion. On the other hand the inversion of neural network enables to find one or more input patterns which satisfy a specific output. The input patterns obtained from the genetic algorithm can be used for building neural network system explanation facilities.

In an artificial environment, the most important key in the process equipment design is determining gas-liquid two-phase flow frictional pressure drop of pipes. To achieve this, an experimental investigation was carried out in this study to analyze the pressure drops of air-water two-phase flow in a 30mm internal diameter horizontal pipe with a length of 6m at different flow conditions. The study was carried out at 20Co using tap water and air. To cover the slug flow pattern, the volumetric flow rate of water varied from 30 to 80 LPM, and the volumetric flow rate of air from 40 to 200 LPM. Pressure transmitters were used to measure pressure at four different points along the test section, each 2m apart. The results of the experiments were compared to 8 models using 3 distinct methods: Mean Absolute Percentage Error (MAPE), Relative Performance Factor (RPF), and the percentage of data included in the range of the 30% error band. All methods produced similar results, with the Sun-Mishima model being the most accurate.

No abstract

No abstract