In this study, equivalent fan characteristic curve that is a new approach for jet fan-ventilated tunnels is presented for the first time. Computational fluid dynamics (CFD) method is used to obtain the equivalent fan characteristic curve for different conditions through simulating the flow field in tunnels with different cross sections and jet fan locations FLUENT commercial software. The results indicate that the flow rate exiting ventilation tunnels with different cross sections and equal cross-sectional areas is approximately the same. Then, we tried to find the optimum location of jet fans in various transverse cross sections during the tunnel length. The results show that if we install the jet fan between tunnel entrance (3 m from tunnel entrance) and 90 m to tunnel exit, the flow rate through the tunnel will not change significantly. At the end, a correlation in terms of tunnel geometric parameters and the distance between jet fan center and the tunnel roof in circular tunnels was presented to obtain characteristic curves of various tunnels with different distances between jet fan and tunnel roof. This correlation can be used as an alternative approach to replace 2D solution and axial symmetry with time-consuming and costly 3D solution. In order to verify our correlation, three diameter ratios (circular tunnel diameter to jet fan diameter) of 6, 9 and 15 and two distances of 0.9 and 1.5 m between jet fan center and tunnel roof is selected. The results show the desirable accuracy of correlation for common distances between jet fan and tunnel roof in tunnels with various cross sections. The effect of jet fan angles of 4 and 6 degree which were suggested in literature on flow rate through the tunnel and thus the equivalent fan curve characteristic was investigated.

 In the present thesis, the simulation of natural gas flow in urban gas pipeline networks has been studied. In this research, the flow inside the pipe has been assumed as steady-state, isotherm and compressible and the working fluid is natural gas. The result of the code development is available to simulate urban gas supply networks. Also, the developed program occupies the minimum space in the system. The solution of pipe networks includes determining the amount of flow in pipes, determining the velocity at the beginning and at the end of all network pipes one by one and also the pressure in nodes and connections providing that the network structure, pressure at the entrance and at the beginning of the network, the amount of gas consumption in nodes, physical dimensions of the pipes and physical properties of the fluid flowing in the network are given. First, the equations governing a pipeline placed between two nodes and the relationship between the pressure (head) at both ends of the pipe line (nodes) and the flow rate of that pipeline were obtained. Since the equation governing the pipeline is non-linear and a network with multiple interconnected pipelines should be analyzed, hence the non-linear equations of the pipes should be solved using Couple Method. In this study, all the non-linear equations have been concurrently solved according to the two methods of linear theory and Newton–Raphson method in an algorithm. In the following, the pressure amounts at nodes, the flow rate in the pipelines and the velocity values at the inlets and outlets of all the pipelines were obtained for several sample networks using the combined algorithm. The combined method has provided the possibility to simultaneously benefit from the advantages of both the linear theory and Newton-Raphson method and each of the methods compensates the drawbacks of the other. Finally, the results of one of these networks have been validated by the software results.

In present study, simulation of three dimensional heat conduction with nonlinear contact resistance in nonhomogeneous regions is carried out using boundary element method. First, heat transfer equation is solved homogeneously. In the next step, boundary equation integral form of governing equations is obtained. Continue to the results, upon numerical solution of boundary integral with the aid of traditional boundary element method, a pipe with nonlinear boundary condition is studied using iteration method and the results are validated with proper previous results. In three dimensional case, it is assumed that two pipelines are far from each other and therefore the potential distribution around pipe has been neglected. In this case, upon using pipe elements potential distribution across the pipe axis is obtained which is regarded as three dimensional in homogeneous state. In the nonhomogeneous case, the equations are obtained using boundary element method but due to defining a new term of equation on all the domain, dual reciprocity method for integral transfer of this term from the volume to the domain boundaries is presented. In the case of two dimensional nonhomogeneous, results of this method are compared with point to point Jacobi finite difference method. For nonhomogeneous three dimensional domain, heat transfer analysis is done using pipe elements and dual reciprocity method. Moreover, some important factors in potential analysis such as tube sizes and tube distances from each other are investigated in both homogeneous and nonhomogeneous cases. Comparison of the results also show that the proposed dual reciprocity method with proper accuracy can help through heat transfer analysis of studied models in infinite regions with any function of medium conduction coefficient.

 In this thesis, gas flow is simulated in a turbine flow meter of urban gas pressure reduction station. Standard installation of turbine meters requires straight pipes of 20 diameter length before and after the meter. However, for packed pressure reduction stations, there are not enogh spaces to install straight pipes. Usually there are elbows at the input and output pipes with much smaller length of 20D. Therefore, the turbine meter may have acceptable measurement error. In this study, the effect of these cases on the measurement accuracy of the turbine flowmeter are examined. In this regard, the performance of turbine flowmeter in the gas flow measurement process is investigated in 4 different meter settings and at 3 different flow rates. Also, flow fields within input and output pipes along with the elbow are studied. Computational network is made with structural grid. By using the above grid, flow field inside the meter is simulated. Then the meter revolution, pressure and velocity fields are obtained. Results show that the accuracy of the turbine meter is strongly dependent on the development of inlet velocity profile. Numerical results show that using the elbow at the entrance has the greatest effect on measurement accuracy of the turbine meter. As the elbow location approaches near the inlet of meter, measurement error increases. In the case where elbow location is at 10 diameters from the inlet of the meter, the measurement results errors are less than 2%, which correspond to the standard IGS-IN-102. So accuracy of the solution results confirms. Also, the results show that using elbow at outlet pipe has smaller effect on measurement accuracy of the turbine meter

 In this thesis, the performance of Hoffman brick-buming kilns at various pressures is to be considered. In order to accomplish this, the effects of reducing the inlet gas pressure on kilns temprature distribution and bricks resistance were verified both experimentally and analytically by installing accurate exact data logging instrument. The results were presented for 30 psig pressure as the system present condition and for 15, 20 and 25 psig pressures. Studies showed that despite the pressure reduction, the temprature distribution in the kiln was not changed significantly and also the bricks resistance was not considerably affected. The final result extracted from performing the project showed almost 20 percent of reduction in the fuel consumption just by reducing the gas pressure from 30 to 15 psig. In the thermal analysis after evaluating the kilns stability and extracting the relations, the kilns were assessed regarding to entered, absorbed and exited heat values. In this report, the description for all of the stages such as study phase, procurement and installation of logging instrument, thermometry data, thermal analysis and other relevant cases is presented.

 In this thesis, two types of gas flow filtration systems of city gas pressure reducing stations are simulated. The gas flow for the first system is from the inside to the outside of the filter element, but for the second one, it is from the outside to the inside. Each of these systems has been investigated for four gas flow rates, from minimum to nominal flow rate passing through the station. For each case, the filter element is considered with four different porosities, from completely new to no longer in service. Since the considered filtration system has a complex geometry, a lot of efforts have been made to achieve suitable computational grids, so that in the generated computational grids, the percentage of inappropriate cells with a skew of more than 0.6, is less than 2.5%. The results show that the filter element increases the gas flow turbulence and the larger vorticities appear on the surface of element. Also the numerical results indicate that the pressure drop of filter abruptly increases by reducing the porosity of that to less than 0.4. Therefore, due to excessive pressure drop in porosity of 0.2, filter elements need to be replaced before they reach this level of porosity. Otherwise, it causes excessive pressure drop in the gas flow path. This pressure drop is variable from about 6 psi up to 28 psi from minimum to maximum flow rate passing through the station. Also the second type system, has more uniform distribution of gas flow through the entire surface of the element in comparison with the first one.

 Lack of appropriate ventilation and consequently increase in temperature, humidity and pollutant emissions and similarly lack of oxygen in some parts of Karoon III hydropower lead to define a project toward solving the expressed problems. Upon evaluations, simulating ventilation system as a way to express solution of Karoon III hydropower’s problems was deemed necessary. Placing the hydroelectric Power Collection below ground level, inclusion of all pathes, existence of different source of moisture and HVAC instruments, Short and undeveloped flow pathways in Power plant, and also the environmental impacts set the reasons to limit using software to simulate existing ventilation system. First, the underlying problems identified in plants and original project was divided into several sub-project. Geometric data collection requirements and clarifications, the measurement data for the simulation prepared by the apparatus and methods of the study were extracted. Using the pressure equations and Newton-Raphson method lead to close initial guess of discharges in each direction in the compiled code. After that the Hardy-Cross method were used to involving the rings and fans information into the solution and the final discharges quantity in each path is obtained. Pressure of each node is calculated by resistance and discharghes value of each path. Thermodynamic data of each route is calculated using the existing relationships and rate of discharche quantity of each path respectivly. By comparing the information obtained from compiled code and messured date, the code was verified. Redirecting and also increase or decrease the resistance in different directions was used to improve ventilation system.

 One of the pollution that enters into the gas networks is the pollution caused by rich and after-pig gases. In this project, the destructive effects of these two gases were evaluated for pressure reducing station of Abadan power plant. Rich gas phase diagram was plotted by pipesime software and it was observed that the increase of steam in the gas causes formation of water droplets. By checking station during the pigging period, it was evident that the solid particles increase in the gas after pigging. To solve the problem of existence of solid and liquid particles in gas, using a multi-cyclone was proposed in the entrance of Abadan power plant station. A cyclone of this equipment was analyzed by Fluent software and the pressure and velocity contours were achieved. Then separation efficiency of cyclone was evaluated in facing with solid and liquid particles and it was observed that the cyclone with over 99% efficiency is able to separate solid particles with diameters more than 4 micrometers and water drops with diameters more than 8 micrometers. Then, according to the station's flow rate, the number of cyclones for a good efficiency in the multi-cyclone between 32 and 39 was determined. Finally, it was concluded that the use of multi-cyclone in Abadan power plant station is affordable according to the high cost of repairs and gas prices. By using multi-cyclone, the cost of repairs will be offset after three years.

 One of the challenges that Gas industry faces it in natural gas lines is to forming dew and appear liquid phase in these lines. Appearing liquid phase can be lead to unfavorable factors as forming Hydrate, corrosion of metal surfaces, reducing the line capacity in pipe lines and gas network facilities. So it will be necessary to study and recognize the effective factors on dew phenomena in natural gas in order to prevent from appearing above problems and also control gas quality. Gas Networks Research Center of University Shahid Chamran of Ahvaz Following the request Gas Company of Khuzestan province, this thesis has defined. The main aim of this study is to survey the dew phenomena in gas Pipelines and effective factors on it. In this regard, in addition to define exact dew phenomena as actually, the kinds of dew diagrams in natural gas were investigated and recognized the actual effect amount of each gas component on the dew point curve. Pipesim software, SRK equation of state and Moody flow equation has used to simulate in this thesis. The most important purpose of this study is to evaluate reasons of dew formation on the Abadan 36-30 inch Pipeline. Results show that high amount of water in gas is the reason of leading to this problem and if the amount of water is less than 200 ppm, never this does not happen. Lack of complete drying pipelines after Hydrostatic test is one of the reasons that is caused Staying moisture in the gas pipeline. So reasons of this problem, Lack of complete drying pipelines after Hydrostatic test, are studied in this thesis. After studying the various standards, visit the drying process pipeline and holding expert meetings, it was found that the root of this problem is in executive guidelines that has developed by National Iranian Gas Company.

 In present research, Potential flow about an under water vehicle for calculation of hydrodynamic coefficients of lift and pitching moment, has been investigated. For steady flow It is possible to calculating the velocity and pressure distribution of under water vehicle with Boundary element method and distribution of cosntant source and doublet combination on the elements. With calculation of pressure coefficient, other coefficients can be calculated. The potential flow is modeled with Laplace function of potential. Then, Drichlet boundary condition and Kutta condition for creating Circulation, are applied. With transforming the Laplace equation by Integral equation and applying the boundary conditions, the equation convert to a Matrix of algebraic equation. The lift coefficient of separate wing and body, calculated by Boundary element method. Next, the lift and pitching moment coefficient of wing- body combination, calculated by Component Build-up Method. Resuts show that Panel method and Component build-up method perform calculations with high accuracy and high speed, so that it is 78 percent time saving in calculations.

 In this thesis, pipe networks of a special gas lab have been modeled and some of producible networks in that with assumption of one dimensional, adiabatic flow have been numerically simulated. So, first necessary prerequisites and criterions for design of this lab related to its purposes have been presented. Next, required piping components for design and way to select them based on standards and literature in piping field have been mentioned. Also in the final of design step, three dimensional modeling of this lab according selected piping components in PDMS software have been done. For result verification of mentioned modeled lab after start up, one pipeline and two producible different networks in that with assumption of one dimensional, adiabatic flow have been numerically simulated. simulations for both air and methane gases because of using of air in this lab networks for results comparison and differences them with together have been done. Comparison of obtained results for both gases and scrutiny variation of flow parameters showed that total pressure profiles almost has been matched. In addition,similar result in static pressure profile for both gases has been observed. Also there is some little difference in Mach number and static pressure profiles. But considerable difference showed in obtained flow rates for air and methane in different simulated networks. The maximum difference in obtained flow rates for air and methane in pipeline and two simulated networks has been calculated about 33.35, 35.819 and 36.452% respectively.

In this thesis, high velocity compressible flows along with heat transfer effects have been simulated in order to calculate the amount of released gas. At first, high velocity pipe flow is simulated in both adiabatic and non-isothermal conditions using the Euler equations. different algorithms have been developed to simulate pipe flows in each thermal conditions. In order to ensure the accuracy of simulations, the results of numerical solutions have been compared to other available exprimental and numerical datas which indicates the reliability of present simulation. Next, the effects of heat transfer on the flow parameters have been studied different thermal conditions. Furthermore, the present simulation has been extended from a pipe to a network of pipes by considering the minor losses through the junctions along with the existing compressibility effects. Then, Proper algorithms are proposed due to simulate the purge flow of gas in several examples of branched and looped networks. Finaly, using these algorithms the purge flow of gas has been simulated in mentioned networks. Results of simulations indicate that the heat transfer in cases like buried pipes, have a small effect on the amount of released gas and the maximum difference between this case and adiabatic thermal condition in a pipe and network has been calculated about 3.6 and 4.5% respectively. these results for unburied pipes has been calculated about 4.3 and 14.7% for a pipe and network respectively.

 In the present thesis, reduced-order modeling (ROM) of unsteady conduction heat transfer is performed based on the system eigenmodes. Eigenvectors of the corresponding eigensystem are chosen to construct the related reduced order model. Dual Reciprocity Boundary element method is used for the present numerical analysis. In continue, the eigenvalues and the right and left eigenvectors are computed and used to construct the desired reduced order model. In the present analysis, static correction was necessary to construct the reduced order model. Static correction application will reduce the efficiency of ROM. In order to reduce the static correction effects, another system of equations is constructed based on the non-zero eigenvalues. It is seen that the number of zero eigenvalues are equal to the number of the nodes with dirichlet boundary condition. The obtained results show that the present reduced order model can give satisfactory results using only a few dominant eigenmodes. For this reason, seven problems with different boundary conditions in two and three dimension are considered. Results show that with a bigger system of equations and more number of dirichlet conditions, better results will be obtained and in some cases, ROM can save more than 90% in computational time. Moreover, one can find out about the stability of the numerical method using the behavior of the eigenvalues.

In this research, steady state heat conduction in two-dimensional and three-dimensional geometry has been studied in numerical method. The method is boundary element method. For more efficiency domain decomposition technique is also used in this paper. For this reason; first boundary integral form of governing equation was obtained. Some two-dimensional and three-dimensional examples were analyzed by boundary element method and constant elements. And the results were validated by reliable sources. Then these examples were analyzed by domain decomposition technique. First, main domain in regard with the geometry was divided into two or more sub-domain; the initial guess for virtual boundaries was elected and analyzed by current boundary element method. Afterwards the results for sub-domains were corrected with a specific algorithm and were selected for the candid answer again. This procedure is continuing till the final answer. Finally; the results in this chapter were compared with the previous ones. So that while declaring accuracy and efficiency of analyzing heat conduction problems by boundary element method and domain decomposition technique, necessary arrangements for parallel processing for this method was unlimbered.

The fast multipole method (FMM) as been regarded as one of the top 10 algorithms in scientific computing that were developed in the 20th century. Combined with the FMM, the boundary element method (BEM) can now solve large-scale problems on a PC that this opened up a wide range of applications for the boundary element method. In this research the fast multipole boundary element method (FMBEM) was introduced and this method applies to solution laplace's equation outcome of steady state heat conduction without heat source.The basic concept and main procedures in the fast multipole method for solving boundary integral equations are described in 2D and 3D. Finally numerical examples are presented to further demonstrate the efficendy, accuracy and potential of fast multipole boundary element method for solving large-scale problems. For both FMBEM and conventional BEM converge quickly to the exact solution for few elements. In general, the fast multipole boundary element method is found to be equally accurate as the conventional BEM. The CPU times used for both approaches in calculations shows significant advantage of the fast multipole boundary element method in the saving compared with the BEM.The FMBEM has less accurate than the conventional BEM for larger models because of the truncation error introduced in the multipole moment expansion.