Integration of material design and product design a CFD based approach.

Prakash, K. Soorya ; Nazirudeen, S.S. Mohamed ; Malvinraj, M. Joseph 等

Introduction

Valves are components which controls the fluid flow and pressure of a system. Types of applications for valves differ on their own and are normally used on safety and flow control grounds. While these valves are used for flow control, it is obvious that the dynamics of the valve and its control loop has to match strictly with dynamics of the control system. In the course of study on these facets it is over and done with establishment of relationship between valve positions, pressure drop and flow which is commendable as highly non-linear. The able-bodied documentation of valves is not admirable for the reason; these facts features technical hitch to predict the properties down or up-sized.

In any flow system the familiar flow restrictors are the valves and hence its design and performance analysis will be a significant task. The selection of the valve types, design and material plays a vital role in its performance and reliability. A number of researches have been experimented in valves for its shape, size, fluid types, operating parameters, discharge coefficient, and eroding characteristics for the improvement of valve technology [5]. It is quite disturbing that detailed investigation was not done over the integration of material design and valve design to suit a specific functionality. This research work is focussed on establishment of relationships between the design variable of both material and product design domains. Computational fluid dynamics a powerful analysis tool is utilized to compute flow restrictions in the form of resistance co-efficient and flow volume co-efficient. In CFD analysis the pressure distribution across the required area is analyzed by gratifying the boundary condition applied.

In order to perform the design, optimization and the analysis of the valve performance, for a particular application dynamic fluid analysis is performed on two types of valve viz. stem valve and ball valve. Valve body is chosen for generating relation between type of fluid flow, geometry and pressure loses using CFD, since they have a simple mechanical construction, utmost exposure during fluid flow, attains critical deformation, and more importantly, give a low head loss. CFD analysis is capable to reveal the complex flow structure and the sonic characteristics around the valve, which the experiments hardly ever provide. Even otherwise, experimentation needs to be supplemented with CFD analysis because of intricate geometry as well as complexities like turbulence during the sonic flow through a valve. The experimental results were validated by CFD or vice versa.

Methodology

Valves, when used in high pressure systems, often get damaged by the frequent and severe movement of the fluid molecules, back and forth, against each other over an extended period of time and by shock waves during the sudden opening and closing of valves. Leakage is another problem in the valves used in high-pressure systems. The repair or replacement requires demounting of the valve components from an existing piping structure. The inconvenience and attendant damage to the surrounding piping structure poses major impediment to the valve replacement. Another shortcoming in the valves employed in high-pressure applications is the requirement of considerable pressure for actuating the valve [1]. This is because the high pressure and related high friction forces are necessary to insure sealing. Further, the contaminants affect the seal life so severely that the life of the valves reduces to nearly half of its prescribed lifetime. In addition to all these, the geometry of the valve and the drop in velocity around the downstream region involves the risk of particle accumulation.

In short, for the effective functionality and reliability of any type of valve it is of utmost importance to consider

* The nature and adhering characteristics like pressure, temperature, viscosity of fluid flow,

* Material factors inclusive of composition, physical and mechanical properties and

* Design parameters comprehending thickness, inlet diameter and shape of valve body.

All these significant aspects are to be predetermined at the concept level of product design as proposed by the consumer. The present arena of research is hence focussed on material and design integration at the early stage of product development and when implemented a customized product of choice pleasing all these criterions can be revealed.

In the present study, analysis of a specific stem and ball valve design has been carried out for pressurized flow of water, lubricant and diesel to recognize the performance and reliability characteristics. The flow pattern analysis has specifically done on different ports and opening modes of valves to eventually determine the optimization of material, design and operation. The valve is subjected to a large out of balance moment imposed by huge weights that act close to the valve and is positioned between long horizontal sections of pipe fed by a constant head source. The methodology adopted in these systems focus on the design, analyses and numerical validation of the results [5]. Hence measurement of pressures and head losses at various sections of valves for different modes of opening is done.

Material prospects

The valves selected for the research work is well applicable for high pressure services and hence the comparable material in particular for these types of valve is ASTM A487 Gr 4C. In broad-spectrum, real time experiments were conducted in a foundry by varying the chemistry within the specified composition standard limits. In present work, the elemental content in the proposed steel was spotted within the following limits as per ASTM A487 Gr 4C: 0.2 to 0.3 % Carbon; 0.4 to 0.8 % Silicon; 0.8 to 1.0 % Manganese; 0 to 0.03 % Sulphur; 0 to 0.03 % Phosphorus; 0.4 to 0.8 % Chromium; 0.4 to 0.8 % Nickel; 0.15 to 0.3 % Molybdenum; maximum 0.5% Copper; 0 to 0.03 % Vanadium; 0 to 0.1 % Tungsten and a maximum of 0.60% of unspecified alloying elements. The melting range of the specified steel is about 2740-2800[degrees]F. The test castings were poured in C [O.sup.2]--sodium silicide moulds, knocked out, risers were removed and then subjected to normalising, tempering sort of heat treatments and various material properties such hardness, passion ratio and so on were also studied and appropriate selection has been done.

Researchers have extensively experimented with all categories of valves in order to identify their specific applications [3]. Numerous mathematical and numerical (CFD and Fluid-Structure-Interaction (FSI)) models have been developed to model the performance of valves. With reference to the explicit purpose and industrial relevance this study has well thought-out for an assortment of valves. As a consequence ball valve ([v.sub.1]) with 30 and 45 degree opening modes and stem valve ([v.sub.2]) with10 mm and 70mm modes are observed for the same material.

Numerical scheme of analysis

The CFD code of fluent 6.0 for finite volume method has been utilised to solve the discretization of continuity equation and Navier stokes equation. The CFD code is commonly used to solve since it has high capability of solving the transient, compressible, turbulent and reacting flows in the finite volume grid with boundary condition and meshes [2]. As per the requirements of study, hexahedral meshes are incorporated in handling the complex geometry and in enabling the compensational domain. The methodology of computation is based on the pressure concentrations and temperature distribution. The temporal discretization and spatial discretization is used for the momentum, energy and turbulence equation. In addition the various valve openings were considered to test out numerical stability of CFD calculation.

Modelling approach

By the process of integration the models of studied valves has been imported from modelling software to CFD software and thereby transient pressure loads acting on the structure is applied [6]. Governing equations of fluid flow such as continuity equation, heat transfer equation, momentum equation and Reynolds equation for a turbulent fluid flow of a compressible fluid has been adopted to solve the CFD analysis [7]. Equation 1 illustrates the continuity equation of turbulence

[partial derivative] [rho] / [partial derivative] t + div ([rho] U) = 0 (1)

Where p is the fluid density and U is vector of the mean velocity. Equation 2, 3 and 4 exemplify the Reynolds equation of turbulent flow used in specific to this application.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)

where P is pressure, |i is the fluid viscosity, u', v' and w' are turbulent fluctuating components of velocity. In adherence to the same application, Equation 5 demonstrates heat transfer of turbulence flow.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)

Where in CP is the specific heat, T is temperature, and [GAMMA]T is heat conduction coefficient and T' is turbulent fluctuant of temperature. In context to the above said relations, equations 6& 7are the standard k-[epsilon] turbulent model and equation 8 illustrates momentum equation of the turbulence flow model.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (6)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (7)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (8)

The velocity distribution is non-uniform due to viscous effects which in turn set up shear stresses. If when considered for the fluid element, equation 9 illustrate the kinetic energy K generated by the turbulent fluid during the course of flow and equation 10 & 11exemplifies for the calculation of net shear force and the net pressure force on the element [8].

[partial derivative] ([rho] k) / [partial derivative] t + [DELTA] ([rho] U k) = [DELTA]. [([mu] + [[mu].sub.t] / [sigma] k) [[DELTA].sub.k]] + [P.sub.k] - [rho] [epsilon] (9)

Net shear force on the element =

([[tau].sup.+] ([partial derivative] [tau] / [partial derivative] y) dy) dx.dz- [tau] dx.dz (10)

Net pressure force on the element =

p.dy.dz-([p.sup.+] ([partial derivative]p/[partial derivative]x))dy.dz (11)

It is observable that the flow-solver fluent 6.0 uses the differential transport equation for the turbulence kinetic energy and turbulence dissipation for supplementary analysis.

Geometry and Boundary Conditions

The valves are assumed to be isothermal and independent with temperature thereby considering the property of the fluid flowing through the valve in synchronization with geometric flow resistances and density of the viscous Newtonian fluid as stated in table1. For performing the CFD analysis the imported geometry is fixed along the x-y plane of the coordinate axes.

CFD process

A. Pre-processing

This is the initial step in building and analyzing a flow model. The valve geometry is modeled and all dimensions are interlinked to ensure stability of the valve geometry during parametric updates. Figure 1(a) & 1(b) demonstrates the geometry fashioned using modeling software. CFD geometry model is created with assistance of solid modeling software wherein considerations of real time design of valves is appreciated and imported to FLUENT using .STP and .IGES data formats as exhibited in figure 2 (a) & 2(b). The imported geometry is refined and meshing is carried out subsequently with the help of ANSA software. The required boundary conditions and fluid properties for solving are furnished as stated in tabulation 1. The meshed model is then cleaned up with help of the conditioning tool and the model is made ready for solving. Parametric variations are also made inherent to the process [10].

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

B. Solving

The flow calculations for ball and stem valves are carried out in this step by the CFD software. It is obvious that the code generated by the software has high feasibility which enables modification at any stage of the process thereby refinement of designs is done more professionally. Initially the no: of iterations is set to 100 and successive increase in the no: of iterations up to 5000 produced results in a refined manner.

The solution obtained for different valves subjected to this study in this step are shown in clear pictures. These solved CFD model geometries are incorporated to the post processing and final output generation is processed.

C. Post processing

This is the ultimate step in CFD analysis, and it involves the organization and interpretation of the flow data and the production of CFD images and animations. The post processing tools enables several levels of reporting [9]. High-resolution images and animations obtained through post processing in a quick, efficient and sufficient manner are publicized in figures 4 to 27. The regions of maximum static pressure, dynamic pressure and the areas of recirculation are identified and are presented in the graphical form.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

[FIGURE 13 OMITTED]

[FIGURE 14 OMITTED]

[FIGURE 15 OMITTED]

Computational results & discussions

The study focuses to predict the capability of the valve body material towards application of pressure through fluid flow. Consequently to make the analysis valid the exertion of fluid pathway by different viscous fluid for different valve openings is observed. Since valves are commonly used as the flow control regulator for high-pressure system and due to seepage and higher attrition resistance characteristics it is made necessary to conduct analysis in different opening modes, and hence analysis is carried out for 10 mm and 70 mm release of stem valve. Similarly various types of analysis inclusive of 300 and 450 openings are performed over ball valve. To notice the fluid flow conduit, province of maximum pressure concentration and velocity vectors over definite areas of the valve body, the same analysis is conducted for different types of fluids. Among the three different fluids considered, the pressure exerted by water flow was higher than any other fluid considered in this study. This variation predicted is due to its high density and low viscosity characteristics expedited at a maximum standard pressure and controlled temperature. It is obvious that considerations on water flow through the valve are enough to study the behaviour of the material of the valve.

Observations over the outputs as seen in figures 4, 8, 12, 16, 20, 24 the areas of recirculation are found to be maximum in 30 degree opening mode at the outlet side of the ball valve for considering water as fluid flow. Also, in case of stem valve the areas of recirculation are found to be maximum in 10 mm opening mode at the outlet side of the valve considering water as fluid flow. In depth analysis makes clear that the above said regulations are due to maximum pressure exerted by fluid flows viz. Water in these cases. The regions of maximum pressure concentration in case of stem valve are identified as the regions on and above the stem and the same are clarified from figures 5, 6, 9, 10, 13, 14. On the other hand the regions of maximum pressure concentration for ball valve are identified as the areas around the ball and corresponding valves of pressure concentration is noticed by referring to the figures 17, 18, 21,22,25,26. It is further noticed that the pressure concentration is more when the valve is maintained at 45 degree open mode for ball valve and 10 mm open mode for stem valve.

The valve delivering high velocities and the respective regions of delivering high velocity is identified as the region near the position of an obstacle. The obstacles of fluid flow through these valves are the ball and stem. It is clearly understood from the velocity profiles showcased in figure 7, 11,15,18,22 and 27 the velocity vectors are critical in 450 opening mode of ball valve. The same prediction when extended to stem valve reveals that the velocity vectors are in peak for 10 mm opening mode. In addition to all these facets, the areas of maximum recirculation and the regions of critical pressure excreted at particular section is perceived and the same is notified as the inlet boundary condition for the further study using FEA.

[FIGURE 16(a) OMITTED]

[FIGURE 16(b) OMITTED]

[FIGURE 17(a) OMITTED]

[FIGURE 17(b) OMITTED]

[FIGURE 18(a) OMITTED]

[FIGURE 18(b) OMITTED]

[FIGURE 19(a) OMITTED]

[FIGURE 19(b) OMITTED]

[FIGURE 20(a) OMITTED]

[FIGURE 20(b) OMITTED]

[FIGURE 21(a) OMITTED]

[FIGURE 21(b) OMITTED]

[FIGURE 22(a) OMITTED]

[FIGURE 22(b) OMITTED]

[FIGURE 23(a) OMITTED]

[FIGURE 23(b) OMITTED]

[FIGURE 24(a) OMITTED]

[FIGURE 24(b) OMITTED]

[FIGURE 25(a) OMITTED]

[FIGURE 25(b) OMITTED]

[FIGURE 26(a) OMITTED]

[FIGURE 26(b) OMITTED]

[FIGURE 27(a) OMITTED]

[FIGURE 27(b) OMITTED]

Conclusions

CFD module for the ball and stem valve is successfully analyzed and the regions of maximum pressure concentration is determined. It is found that the maximum pressure is attained in the fluid flow viz. Water for both systems of valves. The same is attained since the viscosity of water is on the lower side when compared to any other fluid considered in this study. The maximum pressure in the regions around the inner walls are predicted for both types of valves and so the valve body manufactured by this particular ASTM Grade A487 material has the ability to withstand a maximum dynamic pressure of 330 bar approximately. The analyzed results for 17 mm thickness of valve body in case of ball valve and 75 mm in case of stem valve have the ability to withstand the pressure notified. As future prospects, the ability of the proposed valve material in relation to fatigue characteristics can be identified using FEA by applying the observed results as input parameters in adoption to variation in thickness of ball valve and OD of stem valve. Hence the study ascertains a robust affiliation between the design variables of material design domain and product design domain.

References

[1] Borghi, M., Milani, M., and Poluzzi, R. Stationary Axial Flow Force Analysis on Compensated Spool Valves, International Journal of Fluid Power, 1 (1): 17-25, (2000).

[2] Cody McKinley and John Lumkes. Using Computational Fluid Dynamics (CFD) to Simulate a Cylinder Head Flow Test, Fluid Power Journal, Systems Integrator Directory, 40-43, (2009).

[3] Del Vescovo, G., and Lippolis, A.Three-dimensional Analysis of Flow Forces on Directional Control Valves, International Journal of Fluid Power, 4(2): 15-24, (2003).

[4] Exponent Failure Analysis Associates, Inc. Analysis of Residential Excess Flow Valves for Fuel-Gas Piping Systems, Prepared for GAMA--An Association of Equipment and Appliance Manufacturers, Doc. No.0700158.000 A0T0 0907 AO01, (2007).

[5] Jan Forsberg, Tommy Persson. CFD-tools in Valve Design Validation of Simulated 3-D Flow through a Butterfly Valve, Report, (2005).

[6] Mookherjee, S., Acharyya, S., Majumdar, K., and Sanyal, D. Static-Performance Based Computer-aided Design of a DDV and its Sensitivity Analysis, International Journal of. Fluid Power, 2(2): 47-63, (2001).

[7] Qinghui Yuan, Perry Y. Li. Using steady flow force for unstable valve design: modeling and experiments, Journal of Dynamic Systems, Measurement, and Control, 127: 451-462, (2005).

[8] Wendy Hardyono Kurniawan, Shahrir Abdullah and Azhari Shamsudeen. A computational fluid dynamics study of cold-flow analysis for a mixture preparation in a motored four-stroke direct injection engine, Journal Of Applied Sciences, 7(19): 2710-2724, (2007).

[9] Yang, R. Hydraulic Spool Valve Metering Notch Characterization Using CFD, International Mechanical engineering Congress, 10: 11-17, (2003).

[10] Yansheng Jiang, Antonio Carlos Valdiero, Pedro Luis Andrighetto, Wang Chong, Luis Antonio Bortolaia. Analysis of pneumatic directional proportional valve with CFX mesh motion technique, ABCM Symposium Series in Mechatronics, 3: 510-518, (2008).

Biographical notes

K.Soorya prakash is a faculty in the Department of Mechanical Engineering at ANNA University, Coimbatore. He has completed his post graduation in Production engineering from PSG College of technology. His research interest is mainly focused on Identification and analysis of materials for newer and emerging processes.

Dr.S.S. Mohamed Nazirudeen, is Currently working as Dean (Student affairs) at PSG College of technology. He has obtained his Ph.d from ANNA University,

Chennai. His research area includes Failure analysis design, Metallurgical property analysis. He has to his credit many national and international Publications.

M.Joseph Malvinraj is currently pursuing his Post graduation in Mechanical Engineering in ANNA University, Coimbatore. His research area includes Alloy development, Processing, Testing, Characterisation of materials and FEA.

T.Manohar is currently pursuing his Post graduation in Engineering Design in ANNA University, Coimbatore. He has about 6 years of Industrial experience. His research interest is mainly on Material design, Processing Testing of Materials and Non Destructive testing.

K. Soorya Prakash ($), S.S. Mohamed Nazirudeen (#), M. Joseph Malvinraj * and T. Manohar (@)

($) Faculty of Mechanical Engineering, ANNA University Coimbatore

(#) Faculty of Metallurgical Engineering, PSG College of Technology, Coimbatore

* PG Student of Mechanical Engineering, ANNA University Coimbatore

(@) PG Student of Mechanical Engineering, ANNA University Coimbatore

Corresponding author E - mail: sooryaprakash.k@gmail.com

Table 1: Inlet boundary conditions.

Parameters/Fluid                      Water   Lubricant   Diesel

Density (kg/[m.sup.3])                 951       875       834
Viscosity (cp)                          1       22.2        4
Inlet pressure (bar)                   350       280       180
Temperature ([degrees]C)               110       100       15.6
[Dia.sub.in] (mm)    [Valve.sub.1]     87        87         87
                     [Valve.sub.2]     51        51         51
[Dia.sub.out] (mm)   [Valve.sub.1]     54        54         54
                     [Valve.sub.2]     75        75         75