Saturday, March 26, 2011
Stress Analysis of a Natural Gas Engine Cylinder Head
1. Introduction
In the past, optimization of engine components such as cylinder heads was based on building a series of physical prototypes, and performing a series of different experiments and tests [1]. Unfortunately, the traditional process for designing and developing was time-consuming and difficult to build physical prototypes during the early stages of the design. The construction and testing of many prototypes is often required to meet a stringent design requirement [2]. This can turn into an expensive process and delay the entire design and development cycle. Although building and testing of the engine component prototypes can yield accurate design, detailed information is not available and the logic behind a specific design cannot be verified. As a result, engineers attain little and general information from each test [3]. Therefore, the finite element analysis (FEA) methodology is being used and becoming a systematic methodology in the early stages of engine design to save the time and cost of manufacturing process. Finite element analysis methodology (FEA) assist engineers to predict the best method for heat removal prior to the first prototype is built by calculating the temperature and stress distribution of each component. Therefore, finite element analysis (FEA) is considered as one of the most powerful computer-aided design tools for engineers [4]. In the process of an engineering analysis, a theoretical and numerical model is the starting point for researchers to develop or design an engineering system. This technique has been accepted for designing and developing complex geometry over a shorter period of time and at much lower cost.
The cylinder head is one of the most complicated and challenging parts of engine, where FEA plays an important role in its optimization [5]. A limited amount of information is available regarding thermal stresses in cylinder head. Komo and Bryzik investigated the development of thermal stresses in engine components with isolative ceramic coatings [6]. A twin-cam 16-valves cylinder head and cylinder block structure accompanied with several important subcomponents under firing load and assembly loads were investigated using FEM. The physical behavior of the gasket bead and liner, the stiffness distribution of cylinder head, the preload of the cylinder head bolts, the residual insertion loads of valve guides and valve seats, and firing pressure have been thoroughly discussed [7].
Other investigators carried out the sealing and structural response analyses under assembly and firing load cases for several areas of interest. Recommendations obtained from the project were forwarded to designers for successful incorporation into and adjustment of other areas for design evaluation. They provided information in regard to the nature and magnitude of thermal and mechanical stresses in the cylinder heads [8-10]. For calculation of boundary conditions from combustion chamber side, the model of the engine combustion was performed by commercial computer software [8-9, and 11-12]. The goals of the analysis were to provide: (1) Validation of the natural gas engine thermo-mechanical simulation results as are compared with the results of base diesel engine; (2) Achieving the thermal and mechanical stresses; (3) Comparison of the stress magnitude with the limits of elasticity [13-15].
2. Experimental Measurements
The engine specifications under the study are shown in Table 1. To apply accurate boundary conditions and input data to run the computational model, experiment was carried out to measure different parameters such as coolant flow rate, coolant inlet temperature, inlet air flow rate, inlet air pressure and temperature, exhaust gas flow rate, and outside cylinder head surface mean temperature (Table 2).
In this project, the engine performance data was measured by a dynamometer at 1850 RPM. The engine under investigation was equipped with six thermocouples. Three thermocouples were installed on the outside surface of cylinder and cylinder head to measure the average outside surface temperature. The inlet and outlet temperatures of the water jacket and the gas inlet port were measured using three thermocouples. The pressures at the inlet and outlet ports were measured using Piezo-Electric pressure transducers. The accuracy of temperature measurement were 0.1 C, pressure 0.001 KPa, and mass flow rate 0.1m3/s.
3. Computational Methodology
Details of the effort include model definition, meshing, model analysis, validation of the Finite Element Analysis (FEA) model, and applying thermal stress, and displacement boundary conditions. The results of the stress analysis, stress field in the firedeck, and the material evaluation are provided. The analysis procedure is shown in figure 1. The first stage in the process is to define the model geometry. This was accomplished using three-dimensional solid modeling using a computer-aided engineering tool, Solid Works [11]. The data is imported from the solid model to the mesh generation software. Mechanical boundary conditions and model constraints are defined and/or calculated to maximize the validity of the analysis given for the model [10]. Thermal and mechanical boundary conditions are applied to the finite element mode. The finite element analysis was carried out using a commercial finite element analysis software package, ANSYS [12]. The results are post processed into a form suitable for engineering assessment that accesses the analysis code's binary database and extracts appropriate results [13]. Local properties may be a function of surface finish, heat treatment, notch sensitivity, temperature, etc., and are an input by the user. Where the temperature dependency exists, user-input tables of temperature dependency calculate the local material property [14, 15]. The numerical analysis to calculate the temperature and stress distribution in cylinder head is achieved by multi-field technique. The essence of multi-field analysis is coupled-field analysis, which allows users to determine the combined effects of multiple physical phenomena (fields) of a design. If the input of one field analysis depends on the results of another analysis, the analyses are called multi-field. The applications of this technique include fluid-thermal and thermal-structure analysis. The computation processes for the analysis are shown in Figs 2 and 3.
3.1. Model Definition and Mesh Generation
The cylinder head model and its geometry are shown in Figure 4. The three dimensional solid modeling was performed by SOLIDWORKS [11]. The cylinder head mesh is constructed with ANSYS.
For thermo-hydraulic analysis, a model of the water jacket (Fig. 5) that receives local velocity and temperature of the cooling water was constructed. Flow characteristics of the water jacket, critical for keeping uniform firedeck cooling were analyzed using ANSYS. Also, brick-element mesh was constructed for water jacket with ANSYS. The combustion chamber is modeled by a computer software package, MATLAB, which calculate the gas temperature and pressure at each crank angle. The calculation is based on single zone model. For detail, see references [3, 16].
In any computational analysis, accurate mesh generation plays an important role. Therfore, sensitive areas are meshed with high resolution. The shape at the valve opening tapers outward near the firedeck, giving the valve bridge a smaller cross-section than any other location in the cylinder head. The valve bridge area is a region of concern and is finely meshed to determine accurately stress gradients as recommended by many investigators [4]. The completed three-dimensional model contains 507533 elements and 50965 nodes to model the cylinder head, and 733252 elements and 91244 nodes to model the water jacket. Element aspect ratios are chosen to be approximately 2.8 in the valve bridge area (Fig. 6). Away from the valve area, element aspect ratio less than 6.0 is used. However, this was not considered to be a significant problem because the stress gradients at these locations are very low. In the non-sensitive regions such as the top of the cylinder head, a coarse mesh is applied in order to reduce the number of elements and CPU time. The water jacket model of the cylinder head is meshed until the nodes of the interface of the cylinder head and water jacket models merge together. The results (along the x direction and across the y direction) indicate that the constructed finite element meshes with 7 x 3 elements in the bridge area would model the thermal process quite adequately. The thermal results of the finite element model that was constructed with this mesh criterion were compared with a highly refined mesh (two times refinement). The difference in results was considered acceptable (within 0.5 percent) for our study. In this analysis, it was found that the mesh types play an important role. Therefore, different types of meshes were examined until the mesh independency solution was achieved.
3.2. Thermal Boundary Conditions
In any thermal analysis, selection of proper boundary conditions is challenging, particularly for engine combustion chamber components, where boundary conditions may vary significantly both in space and time [16, 17]. The boundary conditions for stress analysis combine the results from the thermal analysis and displacement boundary conditions suitable for the cylinder head. In this analysis, only thermal and pressure loads from the combustion chamber were considered. This is not a serious limitation as thermal stress is the dominant form of stress in the cylinder head [18]. In this work, the regions of high stress were sought rather than a particular highly accurate stress value. To satisfy the thermal boundary conditions, the convective heat transfer coefficient should be calculated for all the following regions.
3.2.1 Outside Boundary Condition
To apply outside boundary conditions, the Rayleigh equation (1) is used for a free convection surface [17].
(1)
The value of is 0.25 and is 0.52
The Rayleigh and Grashoff numbers are calculated by the following equations,
(2)
(3)
3.2.2 Inlet and Outlet Ports Boundary Condition
To calculate the thermal heat transfer coefficient in the intake and exhaust ports, Christopher equation is used [19]:
(4)
Where,
(5)
The value of is acquired from CFD or experiment. Air temperature in this turbocharged engine at the inlet port is 337.1 K, and the gas temperature at the outlet port is assumed to be the same as the combustion chamber gas temperature when the exhaust valve is opened. It was calculated as follow [20]:
(6)
3.2.2 Water Jacket Side Boundary Condition
The prediction of the temperature distributions is achieved by solving the energy, momentum, and mass conservation equations simultaneously. Therefore, the water jacket inlet pressure and/or velocity are required to apply as boundary condition in the CFD model. For this purpose, the inlet water flow rate and temperature are measured. [21, 22].
3.2.3 Combustion Chamber Boundary Condition
To calculate the thermal heat transfer coefficient in the combustion chamber, the Woschni equation (7) is employed. To use this equation, pressure and temperature should be calculated at each crank angle.
(7)
In this equation b is the cylinder bore, is the average of gas velocity, which is almost equal to average of piston velocity. Figure 7 shows the gas pressure and temperature profiles in the combustion chamber
3.3. Stress Boundary Conditions
In order to reduce the complexity of the boundary conditions for stress analysis, the interaction between the cylinder head, cylinder head gasket, and cylinder block was not modeled. In fact, the cylinder head and the cylinder block were assumed to expand at the same rate for all points of contact between the cylinder block and cylinder head. This simplification does not allow the cylinder head bolts to constrain the thermal expansion of the cylinder head. Therefore, the thermal stresses are expected to be under-predicted. The other boundary condition is the pressure load that is applied to the combustion chamber. The maximum pressure of 10.5 (MPa) is used in this analysis. The boundary condition of the bolt is very significant since a pre-load is necessary to define tightening of the bolt. In this analysis, it is assumed that the surface interface of the cylinder head and bolt is moved inward. There is no outward movement for this surface. Therefore, for modeling, the contact element between the bolt and cylinder head is used [14].
4. Results and Discussion
As explained before, the first step in the analysis is to validate the computational model. The comparison of experimental and computational results is shown in Fig. 8. Nodes 1—8 in this figure are corresponding points to the numbers 1—8 in Fig.9. The calculation is performed for one temperature cycle at high coolant temperature. It is evident that a very good correlation exists between measured and calculated strains for lower temperatures where elastic material behavior predominates. For the higher wall temperatures exceeding 200 C, the calculation overestimates the influence of inelastic material deformation as compared with the measured values.
It is concluded (Fig. 8) that the Van Mises stress results are matched relativity very well. Therefore, the model can represent the baseline engine cylinder head and it can be further used for parametric studies.
The comparison of Diesel and Natural Gas Engines stress analysis results is shown in Figure 10. The results of Von Mises stress for Natural Gas Engine in cylinder 2 are 1.12 times more than corresponding points in the diesel engine. This is due to higher temperature of Natural Gas.Thermal and structural analysis results show the temperature and Von Mises distribution in cylinder 2 are higher than the others. So, our discussion is concentrated on cylinder 2. It is assumed that the temperature in the water jacket is constant. Figs.11 and 12 show that the temperature at the cylinder bridge is high in both engines; and consequently the temperature gradient at this region remains high. It is also concluded that the temperature distributions are different in all cylinders since the cooling water enters the left side of the bottom end of the block and exits from the left side of the top end of the cylinder head. Therefore, cylinder 1 is cooler than the others and the temperature gradient in cylinder 2 is higher than the other cylinders. Also, the highest temperature at the centre of any cylinder decreases away from the centre. This causes a high temperature gradient at the surface of the combustion chamber. This is due to the fact that far away from the centre, the cooling water flow rate is larger than at the centre of the firedeck. The results predict a large compressive strain and stress field at the valve bridge and seats on the firedeck of the cylinder head. These stresses are primarily due to relatively large temperature difference existing at the liner interface. The maximum temperature at firedeck is 616 K in Natural Gas and 549 K in Diesel engine. The results show that the temperature gradient at the gas side of the cylinder head of the natural gas engine is approximately 53 ºC/mm, and it is 5.6 ºC/mm for the coolant side of the cylinder head. Inside the liner, the temperatures on the firedeck change from 438 K to 616 K at high load operating conditions, while outside of the liner, the temperatures are relatively low at 293 K to 360 K.
The thermal expansion of the hot region is constrained by the stiffer cool region, which undergoes less thermal expansion. As a result, a compressive thermal stress field is created inside the liner. Figure 13 shows the temperature distribution contours of the cylinder head. The maximum stress on the cylinder head also showed similar results. In this work, first, the maximum cylinder gas pressure is applied to cylinder 2, then to the other cylinders. It is concluded that the average stress in the cylinder head reached its maximum value when the cylinder head 2 is under fire. When cylinders 1, 3, and 4 are under fire, the maximum stress is observed in cylinder 2 because the temperature gradient in cylinder 2 is more than in the other cylinders. Also, higher stress in the valve bridge and near the valve seat is seen in Figs.14 and 15 because the thickness of material in the valve bridge is higher and cooling flow rate is lower. Figures14, 15, 16, and 17 show larger compressive strain and stress at the valve bridge and seats on the firedeck of the cylinder head. An important observation from the result of the analysis is that the predicted Von Mises stresses exceed the limit of elasticity (dark areas in Fig. 18) for a typical cylinder head material [25]. This high stress would lead quickly to destruction of the cylinder head.
5. Conclusion
The maximum compressive stress is observed in the valve seats and valve bridge. Natural Gas engine stress is about 1.12 times higher than Diesel engine. High stresses at the valve bridge resulting from a constrained thermal expansion of the cylinder head are generally compressive. It is concluded that about 82%–87% of the total stress is thermal and the rest is due to pressure and mechanical stresses.
The temperature gradient at the surface of combustion chamber is not uniform. The maximum total stress was found in cylinder 2. The Von Mises stress value of the Natural Gas engine exceeded the elasticity limit of material used for the cylinder head. This high stress would lead to destruction of the cylinder head. It is recommended modifying the current cylinder head material from cast iron GG-26 to GG-30 in order to prevent failure of cylinder head. The engine cooling system should also be improved to reduce the area with maximum compressive stress.
REFERENCES
[1]. Spaniel, M., Macek, J., Divis, M., Tichanek, R., "Diesel engine head steady state analysis", Research Report of Technical University in Prague, 2003.
[2]. Assanis, N., "Multi-Dimensional modeling of Natural gas ignition under compression ignition conditions using detailed chemistry", SAE paper 980136, 1998.
[3]. Bryzik, W., Wood, M., Schwarz, E., and Glance, P., "High temperature engine component exploratory design development", SAE paper 890296, 1989.
[4]. Tichanek, R., Spaniel, M., Divis, M., "Steady state heat analysis of engine head", Research Report of Technical University in Prague, 2002.
[5]. Woods, M., Schwarz, E., and Bryzik, W. "Advances in high temperature components for the adiabatic engine", SAE paper 910460, 1991.
[6]. Komo, R., Bryzik, W., "Performance and durability of a ceramic coated adiabatic engine", ASME ETCC Symposium, New Orleans, LA, January, 1990.
[7]. Chyuan, S., "Finite element simulation of a twin-cam 16-valve cylinder structure", Finite Element in analysis and Design, Elsevier Science Publishers B. V., Amsterdam, Netherlands, 2000.
[8]. Roelle, M. J., Shaver, G. M., Gerdes, J. C., "A multi-mode combustion model of SI and HCCI for mode transition control", international mechanical engineering conference and exposition Anaheim, California, USA, November 13-19, 2004.
[9]. Cranfield, A., "Effects of diesel water emulsion combustion on diesel engine NOx emissions", MS Thesis of Florida university, 1999.
[10]. Trigui, N., Griaznov, V., Affes, H., Smith, D., "CFD based shape optimization of IC engine", J. of Oil & Gas science and Technology, Vol. 54, pp.297-307, 1999.
[11]. Reyes, A., "Beginner's guide to Solidworks", Schroff Development Corporation (SDC) Publications, India, 2005.
[12]. Lawrence, K., "Ansys workbench tutorial", Schroff Development Corporation (SDC) Publications, India, 2005.
[13]. Segerlind, L. J., "Applied finite element analysis", 2nd Edition, John Wiley, 1984.
[14]. Jorwekar, P., Birari, V., Nadgouda, M., "Cylinder head gasket contact pressure simulation for a hermetic compressor", International compressor engineering conference, Purdue, July 17-20, 2006.
[15]. Catania, A. E., Misul, D., Mittica, A., Spessa, E., "A refined two-zone heat release model for combustion analysis in SI engines", The 5th international symposium on modeling of Combustion in IC Engines, Comodia, 2001.
[16]. Ferguson, C. R., "Internal Combustion Engine", John Wiley, New York, 1986.
[17]. Wendl, M. C., "Fundamentals of heat transfer theory and applications", Class notes for ME 371 of Washington University, Version 2.1, 2005.
[18]. Tatschl, R., Basara, B., Schneider, J., Hanjalic, K., Popovac, M., Brohmer, A., Mehring, J., "Advanced turbulent heat transfer modeling for IC-Engine applications using AVL fire", Research report of international multidimensional engine modeling, Detroit MI, April 2, 2006.
[19]. Christopher D., Dennis A., "A universal heat transfer correlation for intake and exhaust flows in a spark-ignition internal combustion engine", SAE paper 2002-01-0372, 2002.
[20]. Pulkrabek, W., "Engineering Fundamentals of Internal Combustion Engine", Prentice Hall, 2nd Edition, 1997.
[21]. Chang, J., Guralp, O., Filipi, Z., Assanis, D., Kuo, T.W., Najt, P., Rask, R., "New heat transfer correlation for an HCCI engine derived from measurements of instantaneous surface heat flux", SAE paper 2004-01-2996, 2004.
[22]. Taylor, C. F., "The internal combustion engine in theory and practice", MIT Press, 2nd Edition, 1966.
[23]. Woschni, G. A., "A universally applicable equation for the instantaneous heat transfer coefficient in the internal combustion engine", SAE paper 670931, 1967.
[24]. Pischinger, S., "Internal combustion engine", VKA paper, institute for internal combustion engine, 2007.
[25]. Germany Industrial Norm, "Material standard DIN paper 1691", 1985.
Nomenclature
Area (m2)
Cylinder bore (m)
Average diameter of inlet and outlet ports (m)
Acceleration of gravity (m/s2)
Grashoff number
Convective heat transfer coefficient (w/m2K)
Thermal conductivity coefficient (w/mK)
Mass flow rate (kg/s)
Nusselt number
Gas pressure (pa)
Intake cylinder pressure (pa)
Exhaust gas pressure (pa)
Prandtl number
Rayleigh number
Reynolds number
Temperature (K)
Exhaust gas temperature (K)
Ambient temperature (K)
Gas velocity (m/s)
Displacement volume
Greece Symbols
Coefficient of volumetric thermal expansion (k-1)
Dynamic viscosity (kg/ms)
Kinematic viscosity (m2/s)
CAPTURES
Table 1. Engine specifications
Table 2. Natural gas engine experimental data to be used for computational modeling
Figure 1. Thermo-mechanical analysis procedure
Figure 2. Multi-field computations procedure for CFD-Thermal a: formulation b: loops sequence
Figure 3. Multi-field computations procedure for Thermal-Structural
Figure 4. Model of the engine cylinder head
Figure 5. Model of the water jacket for CFD analysis
Figure 6. Top view of nodes distribution in combustion chamber of the cylinder head.
Figure7. Pressures and temperature profiles in Natural gas and diesel engines.
Figure 8. Comparison of the experimental and computational stress results, node numbers
correspond to the points on the cylinder head in Fig. 9
Figure 9. Location of measured and computed stress in the cylinder head
Figure 10. Comparison of Von Mises stress distribution on fire-deck of combustion surface of Natural gas and Diesel engine cylinder head
Figure 11. Temperature contours on the firedeck of the combustion surface of Natural gas engine (K)
Figure 12. Temperature contours on the firedeck of the combustion surface of Diesel engine (K)
Figure 13. Temperature distribution contours on the firedeck combustion surface (K) of Natural gas engine
Figure 14. Von Mises stress contours on the firedeck combustion surface of Natural gas engine (MPa)
Figure 15. Von Mises stress contours on the firedeck combustion surface of Diesel engine ( MPa)
Figure 16. Von Mises strain contours on the firedeck combustion surface of Natural gas engine
Figure 17. Von Mises strain contours on the firedeck combustion surface of Diesel engine
Figure 18. Von Mises stress on the firedeck combustion surface more than yield stress
No. of Cylinder
4
Max Power
81 KW @ 2800
Max Torque
350 N.M @ 1600-2100 RPM
Compression Ratio
11:1
Cylinder Bore
97 mm
Stroke
128 mm
Volume
3780 cm3
Cooling Water Capacity
8 Liter
Fuel
Natural Gas
Air Induction System
Turbocharged and Intercooler
Fueling Strategy
Lean Burn
Table 1
Parameters
Measured
Coolant flow rate (m3/h)
Coolant inlet temperature ( K)
Coolant outlet temperature (K)
Inlet air flow rate per cylinder (kg/h)
Inlet air temperature ( K)
Exhaust flow rate per cylinder (kg/h)
Fuel flow rate per cylinder (kg/h)
Cylinder head mean temperature of the outside surfaces of ( K)
Inlet air pressure (KPa)
other posts you need to know . . .
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment