2.1 CFD
CFD is the design and analysis method commonly used in the valve industry. In this way, three-dimensional flows inside the valve are simulated using various equations that describe the flow conditions. Its importance for designers is being possible to carry out all the analyses in the virtual environment before the prototype is produced, and to transfer the product to the serial environment conditions, based on the right foundations at the first time and with less labour and material loss. The three-dimensional designs of the products were made with the Solidworks drawing program. ANSYS computer aided simulation program was used for CFD analysis and FEA. A butterfly valve with a diameter of 200 mm was selected according to the capacity of the test setup in order to carry out the studies with experimental verifications. In the industry, as defined in the standard, the nominal diameters of the products are classified in mm by adding DN before them, and the pressures at which the products are used are expressed with PN. The designs of the products that can operate under 16 bar pressure, which are the most widely used in applications, have been chosen with DN 200 and nominal pressures were expressed as PN 16. In CFD analysis, it was solved with the K-epsilon turbulence model, and improvements were made in the boundary layers and mesh by keeping the y + value at 3 and below.
2.2 CFD Formulations
Flow coefficient and pressure loss coefficient formulas are defined according to EN 1267:2012 (E) standard. Accordingly, the flow coefficient (Kv) is calculated by Eq. 1.
$$Kv=Q\sqrt{\frac{\rho }{\varDelta p\times \rho 0}}$$
1
Kv (m3/h) in the equation is the flow coefficient, Q is the flow rate in m3/h, ρ is the density of the water in kg/m3, ρ0 is the density of the water at 15°C in kg/m3, ∆p (The difference between the inlet pressure and the outlet pressure) It represents the pressure loss in the valve in bar.
The fluid resistance coefficient ζ (zeta) is considered according to Eq. 2. In Eq. 2, ΔP is the pressure loss in the valve in pascal, u is the flow rate in m/s, and ρ is the density of water in kg/m3.
$${\zeta }=\frac{2\times \varDelta P}{\rho \times {u}^{2}}$$
2
According to the EN 1267 flow resistance test standard, the valve was modelled in the Solidworks drawing program with the valve in the fully open position. The modelling performed is shown in Fig. 2.
2.3 FEA
The strength values of the valve designed in FEA were determined according to the boundary conditions defined in the EN 1704 standard. In FEA, the parts were handled with two different numerical methods, singular and assembled, and interpolation solution was realized with the Rayleigh-Ritz method. The boundary conditions specified in the standard were applied exactly and were defined as one and a half times the nominal pressure value for the body part and ten percent more than the nominal pressure value for the disc part.
2.4 Materials Selection
While modelling the body part in contact with the fluid, as a result of the analyses made according to EN 12266 and EN 1074 standards, the body material that can withstand one and a half times the pressure value of the body pressure was selected. The throttle material that will withstand ten percent of the pressure value was selected by the same method, and the material was selected according to the stresses and deformation results in the parts according to the body and throttle strength analyses. Referring to the finite element mechanical analysis, EN GJS 400 − 15 material was selected.
The stresses in the system were handled with the product and process design were carried out in a virtual-simulated environment. In the finite element analysis, pressure was defined on the body fluid contact surfaces and the body was given a curved structure, thus reducing weight. Figure 3 shows the current product’s body (a) and designed curved body structure (b)
Disc material was selected the same methodology with the body structure and the new design comparison with the previous version was shown in Fig. 4.
2.4 Flow Simulation
After all the flow and pressure analyses were performed, the analyses of the hydrodynamic forces on the disc were performed. Therefore, Fig. 5 shows the CFD mesh analysis image of the hydrodynamic forces on the disc. The purpose of the mesh analysis was to break a complex volume into small segments to be simulated. By definition, a mesh was a network of cells and points. It could have almost any shape in any size.
Disc subjected to hydrodynamic load and the pressure position change on the surface is given in Fig. 6. This figure indicates the force distribution on the operating disc assembled to the valve body.
2.5 Design Verification, Validation and Prototype Production
The casting model of the valve, which was designed by analysing its hydrodynamic properties, and the production of the core box were carried out on CNC machines with CAM software (Fig. 7).
After the GGG50 - DDK50 - ASTM 70-50-05 cast iron alloy was prepared by the company, the liquid metal sample taken from the ladle was solidified before casting and analysed in the Foundry-Master Xpert spectroscopy device. Analysis results are shared in Table 1.
Table 1
The chemical analysis of GGG50 alloy.
| Chemical Analysis | C | Si | Mn | S | Mg | P | Balance |
| (%) | 3,55 | 2,75 | 0,25 | 0,01 | 0,06 | 0,02 | Rest |
Following the trial castings, hot tearing defects were observed in certain parts of the part. The images of the error are shared in Fig. 8.
Hot tearing is one of the most serious defects encountered in castings. When Fig. 8. was examined carefully, it was concluded that the cause of the error was the sudden cross-sectional change in that region. The formation and propagation of hot tear is directly related to parameters such as cooling rate, solidification time, temperature gradient, chemical composition of the alloy and casting geometry. The schematic view of the tensile stress occurring in this region is given in Fig. 9.
At this stage of the study, considering that the cooling rates are different due to the difference in cross-sectional thicknesses in the defected region, a change was made in the casting design and the simulation study was started by designing the full core (Fig. 10), which was made in the wrong product design, in the form of shell core.
In order to eliminate the error, it was decided to change the casting design and to return to the shell type core design, considering that the solid core mass causes rapid cooling of the thin areas in the casting. The cooling rate analysis image of the shell type core design casting simulation is given in Fig. 11.
The solid model for verification of the design was carried out by Solidworks as given in Fig. 12. This stage had a final and critical importance in which the manufacturability decision of the product would be made virtually.
With the help of computer-aided design and simulation, it was determined that the product could be produced without any problems in the virtual environment, and the production stages in the real environment were started. The prototype developed product was carried out by considering the design parameters and its image is given in Fig. 13.
The microstructure photographs of the cast GGG-50 spherical grafted cast iron sample are given in Fig. 14. The sample microstructure consists of spherical graphite, ferrite (white areas) and perlite (dark areas). No carbide or casting defects were found in the microstructure. The hardness value in the poured state was measured as approximately 175 HB.