AIJSTPME (2009) 2(4): 85-92

Finite Element Analysis of Reverse Engineered Internal Combustion Engine Piston

Gudimetal P. School of Engineering Systems, Queensland University of Technology, Gardens Point Campus, P. O. Box 2434, 2 George Street, Brisbane Q4001 Australia E-mail: p.gudimetla@qut.edu.au Gopinath C.V. Principal, Chaitanya Engineering College, Kommadi, Visakhapatnam-41, INDIA E-mail: gopinathcv@yahoo.com Abstract Damaged or broken parts are generally too expensive to replace, or are no longer available and this is particularly relevant to the automobile industry owing to the ever-increasing accidents. Reverse Engineering (RE) has been successfully employed to for possible recovery of a damaged or broken part. In this paper, we present a framework which successfully uses RE to generate a CAD model of a damaged internal combustion (IC) engine piston and then use the state-of-the-art ANSYS finite element analysis package to perform a linear static and a coupled thermal-structural analysis of the component. Further, a parameteric evaluation of the material properties vis--vis operating conditions is carried out to generate a relational database for the piston to arrive at optimal design solutions under different operating conditions. Keywords: Reverse engineering, Laser scanning, Surface reconstruction, Surface modeling, Thermalstructural analysis, ANSYS 1 Introduction Damaged or broken parts are generally too expensive to replace, or are no longer available. RE can be defined as: Systematic evaluation of a product with the purpose of replication. This involves design of a new part, copy of an existing part, recovery of a damaged or broken part, improvement of model precision and inspection of a numerical model. Advantages of the technique include immediate feedback, data reduction, direct generation of geometry and higher precision of the final product. Reverse engineering has become a very important branch of design and manufacture, and the technique has been widely recognized as being an important step in the product development cycle. The use of RE has resulted in a decrease the manufacturing time and costs, among other aspects. RE is the process of producing design details in the form of CAD model from the physical part in the process of the product design. In contrast to the traditional production sequence, reverse engineering typically starts with measuring an existing object, so that a solid model can be deduced in order to make use of the advantages of CAD/CAM/CAE technologies [1]. Afterwards, CAD models are used for manufacturing or rapid prototyping applications. RE technology can be used to aid in manufacturing of spare parts when original parts inventories are exhausted. For mechanical parts the process involves sensing the geometry and then passing the sensed data to an appropriate CAD/CAM system for manufacturing. Reverse engineering of mechanical parts require extraction of information about an instance of the particular part sufficient to replicate the part using appropriate fabrication techniques. The resulting models can be directly imported into feature based CAD system without loss of semantics and topological information inherent in the feature based representation. In recent times, with the advent of state-of-the-art finite element analysis packages such as ANSYS, the RE generated models can be used to perform a comprehensive performance analysis for optimized solutions.

King Mongkuts University of Technology North Bangkok Press, Bangkok, Thailand

85

Gudimetal P. and Gopinath C.V. 2 Geometry acquisition and digitization Several authors have researched regarding the reverse engineering, especially by focusing on scanning methods (advantages and weaknesses of different scanning systems), reverse engineering applications based on image processing and vision aided, multi-probing approaches, integration with rapid prototyping and other processes, scanning path planning, data-point preprocessing and reduction methods, surface fitting algorithms and solving approaches. Y. M. Chiang, F. L. Chen [1] propose a new architecture based on look-up table that keeps the estimated normal vectors of the measured data to refine the data points digitized by CMM. The digitized data are first fitted into several NURBS curves by interpolation. ChangXue Feng and Shang Xiao [2] presents a Computer Aided Reverse Engineering (CARE) approach. In this approach, a CMM is used to digitize an existing mechanical object, and then the IGES files of the point cloud data from CMM digitization are generated using a software called ScanPak. Pro/Engineer is used to create the solid model of the object and finally the laminated object manufacturing process is used to duplicate the object. T. Shen, J. Huang, and C-H Menq[3] have described a multiplesensor coordinate measuring system and its applications to automated part localization and rapid surface digitization. C.X. Feng [4] presents the methodology of Internet-based reverse engineering with a case study illustrating its applications in integrating CAD and CAM. Yin Zhongwei and Jiang Shouwei [5] report on the automatic segmentation and approximation of three-dimensional digitized points for reverse engineering. Based on an innovation that uses the properties of a Non-Uniform Rational B-Spline (NURBS) or B-spline and makes ordered digitized points, which takes less computation time than traditional algorithms in calculating surface normals and curvatures at digitized points, an algorithm was developed for automatic segmentation and NURBS surfaces fitting for digitized points. Kettil, P and Wiberg, N. E [8] presented the work on the integrated use of computational methods (e.g. geometric modeling, simulation, visualization and optimization) for structural analysis and design. The focus in this paper is on 3D solid modeling and dynamic simulation. The paper describes a program system comprising the in-house program FEM90 integrated with commercial CAD and optimization programs. The program system is tested on a major 86 bridge structure example. The example shows that 3D solid modeling and simulation are versatile tools for design of structures. Yasar Deger [9] performed a study focused on possible local re-design of the device in order to ensure a resonance-free operation. In a first step the significant eigen frequencies with corresponding mode shapes were obtained by means of an experimental modal analysis (EMA). Subsequently, the dynamic behavior of the platform was simulated using an ABAQUS FE model. The comparison of the eigen frequencies based on FE calculations with their experimental counterparts proved in general quite satisfactory correlation. 3 Reverse engineering process If only one original part is available, it has to be handled with utmost care during the digitization process as the original part is crucial for validation. The component is studied thoroughly and the prominence of every feature effecting the working of that component must be estimated also the feature which can be measured manually must be identified. Such features encompass prismatic, geometric shapes. All other features like free-formed surfaces and complex contours and 3D surfaces and to be measured through other techniques like scanning, acoustics and optical methods. All the dimensions which can be measured manually are taken with the help of available measuring devices like vernier calipers, height gauge, etc. The features, which cannot be measured manually, can be obtained through any available digitization techniques. The typical Reverse Engineering process can be summarized in sequence as under 1. Physical model which needs to be redesigned or to be used as the base for new product. 2. Scanning the physical model to get the point cloud. The scanning can be done using various scanners available in the market. 3. Processing the points cloud includes merging of points cloud if the part is scanned in several settings. The outlines and noise is eliminated. If too many points are collected then sampling of the points should be possible. 4. To create the polygon model and prepare .stl files for rapid prototyping. 5. To prepare the surface model to be sent to CAD/CAM packages for analysis. 6. Tool path generation with CAM package for suitable CNC machine manufacturing of final part on the CNC machine.

King Mongkuts University of Technology North Bangkok Press, Bangkok, Thailand

Finite Element Analysis of Reverse Engineered Internal Combustion Engine Piston 3.1 Digitization Digitization is the process of capturing the data of the physical model and converting it into digital form. However, with the introduction of new technologies the term digitizing is now used as the generic description for the process of acquiring data from undefined surfaces out of which a complete CAD model of the part is made, to facilitate analysis and to incorporate changes easily. The part is studied in detail, an idea of the probable manufacturing process helps in identification of the proper strategy for the digitizing. Higher the accuracy required higher will be the cost; hence one must strike the balance between the desired accuracy and the cost incurred. Use of Coordinate Measuring Machine (CMM) which can be attributed as semi automatic digitization technique is justified in cases, where 3D surfaces and holes are to be considered. The features to be scanned are identified and appropriate sections are made on the part. In the vicinity where there is a sudde n change in profile maximum number of sections are to be made. Then scanning is done along the predetermined sections and the point cloud data obtained from the parts. The data is normally in the form of a point cloud containing the X, Y, and Z coordinates. This digital data in general, can be used for generation of surface models or .stl files. This data can also be used for inspection purposes where the point cloud is superimposed on the CAD model and calculate the deviation at the respective points on the component with respect to CAD model. There should be a powerful package which can handle the collected point cloud to form the files for Rapid prototyping or for any CAD application. 3.2 Surface reconstruction Two main categories in surface reconstruction are free form fittings and quadric fitting. The surface meshes generated in these approaches are inadequate for current CAD/CAM software which typically require surface to be represented in the Non Uniform Rational BSplines (NURBS) form. CAD model has the flexibilit y of incorporating changes and interfacing with CAM. Thus through a CAD, one can achieve faster production rates. Surface Modeling and Solid Modeling of the part is done from the cloud point data using modeling packages like Pro-E and CATIA. 4 The present work RE of IC engine piston The piston is modeled from cloud points obtained by laser beam scanning. The points are joined by lines and then creating bounding areas and filling the volume. Later the geometric solid model is imported to CATIA package. 4.1 Laser beam scanning The following procedure followed to obtain the cloud point data. 1. Clamp the piston with three jaw chuck to restrict the degrees of freedom. 2. All surfaces are scanned separately and patches are created for each surface. 3. The surfaces and patches created in step ii, in IGES format, directly imported to CATIA to create the solid model. 4. Using the patches full extended surfaces are obtained for each face of the piston by extrapolating, extending, blending, intersecting and trimming operations of wire frame and surface design module of CATIA to create a closed surface model (Fig.3). 5. The 3D Solid model of the piston is then created from closed surface model, and is imported to ANSYS.

Figure 1: Final Solid Model of the Piston from RE Point Cloud 5 Design and analysis 5.1 Piston assembly For any meaningful analysis of a piston, it is imperative that an appropriate piston pin and connecting rod assembly is incorporated into the study since the dynamic performance of the piston is a function of the contact pressure between the pin and the bearing system, and the flexing and oval distortion of the pin which are all important criteria

King Mongkuts University of Technology North Bangkok Press, Bangkok, Thailand

87

Gudimetal P. and Gopinath C.V. for assessment of whether the forces occurring can be transferred from the piston to the pin safely. Due to the complex geometries and the interactions involved, the finite element analysis approach is an attractive and easy means to understand such interactions. Consequently, in this study, an appropriate assembly was designed to suit the reversed engineered piston. This assembly is as shown in the Figure 2 below. The model was meshed with Solid186 elements using standard mechanical shape checking in ANSYS workbench using 44603 nodes and 23234 elements. A preliminary mesh convergence analysis showed that this level of mesh refinement was adequate to perform both the linear static and thermal-structural analysis of the model. aluminum alloy which has a thermal expansion coefficient, 80% higher than the cylinder bore material made of cast iron. This leads to some differences between running and the design clearances. Therefore, analysis of the piston thermal behaviour is extremely crucial in designing more efficient engine. The thermal analysis of piston is important from different perspectives. First, the highest temperature of any point in piston must not exceed more than 66% of the melting point temperature of the alloy [10]. This limit temperature for the current engine piston alloy is about 640 K. Temperature distribution leads to thermal deformations and thermal stresses. It is generally assumed for piston temperature to remain constant throughout a working cycle and not dependent on the operating states. This assumption is justified for zones within the piston. Thin surface layers on the piston head are also subjected to cyclic temperature exposure within the working cycle, resulting in thermally induced stresses that constitute a high cycle load on the material. The piston thermal deformation has an important role in piston skirt design which has a potential to reduce friction and piston slap. In this design, both of the thermal and mechanical stresses must be considered indicating the importance of piston thermal analysis. The heat transfer coefficients for the different parts of the piston were calculated using the procedure outlined in [10,12]. Assuming a steady state thermal case, Figure 3 below shows the heat transfer coefficients for the piston at different positions.

Figure 2: Meshing in ANSYS of the Piston with a connecting rod 5.2 Boundary conditions for thermal analysis It is important to calculate the piston temperature distribution in order to control the thermal stresses and deformations within acceptable levels. The temperature distribution enables us to optimize the thermal aspects of the piston design at lower cost, before the first prototype is constructed. As much as 60% of the total engine mechanical power lost is generated by piston ring assembly [10]. The piston skirt surface slides on the cylinder bore. A lubricant film fills the clearance between the surfaces. The small values of the clearance increase the frictional losses and the high values increase the secondary motion of the piston. Most of the Internal Combustion (IC) engine pistons are made of an 88

Figure 3: Thermal boundary conditions used in the study

King Mongkuts University of Technology North Bangkok Press, Bangkok, Thailand

Finite Element Analysis of Reverse Engineered Internal Combustion Engine Piston Figure 4 shows the structural boundary conditions that were used to simulate the piston behaviour accurately. Apart from the fixed support at the bottom end of the connecting rod, a frictionless support was declared on the circumference of the piston to simulate the movement within the cylinder. Further, the connection between the piston and piston pin and the piston pin and connecting rod were changed from the default bonded to no separation to permit relative translational and rotary motions respectively. Shaft Power, IP P Transmission efficiency Hence, Break Power, BP Stroke length L Speed, N Diameter of bore D Cross sectional area of, piston head A Mean effective pressure, pm Hence, Pressure acting on the piston head, pm = 12 kW = 0.8 =12X0.8 = 7.6 kW = 75 mm = 1400 rpm = 47.76 mm = 1792.23 mm2 = 3.02 N/mm2 = 3.02 N/mm2

These boundary conditions were incorporated into the finite element model after invoking the map of analysis tree within ANSYS Workbench 6 Results and postprocessing Figure 5a and Figure 5b show the temperature distribution and the von Mises equivalent stress distribution within the piston below. These results have been scoped and sectioned for clarity. It can be seen from Figure 5a that for the prescribed operating temperature, peak temperature from the combustion of gas in the chamber penetrates the piston crown through nearly 90% of its thickness before the piston ring dissipates some of the heat, while the average piston temperature beneath the piston ring is around 280 C through a depth of 45mm. Figure 5b shows that the compressive stresses on the piston are negligible and the forces acting on the piston head are transmitted to the piston pin with a factor of safety in the vicinity of 0.9 for the peak operating pressure, as can be seen in Figure 6. This essentially shows that the material of the piston will work well under the prescribed operating conditions. It should be noted that an analysis of the piston pin for shear stresses is beyond the scope of this work, which can be however achieved by an appropriate material model for the piston pin and examining the shear stresses in the cross section of the pin.

Figure 4: Structural Boundary Conditions used in the study 5.3 Mechanical load calculations During operation, the piston is subjected to loads exerted by the gas, FG and masses FmK given by

FK = FG + FmK

The gas forces result via multiplication of the cylinder pressure pG acting on the piston head across the head surface area AK. The gas force can thus be given as

FG = AK pG
On account of the considerable variation in the maximum gas pressure subjected to the piston head, the mean effective gas pressure is calculated from design first principles. The design data for the considered piston is presented below.

King Mongkuts University of Technology North Bangkok Press, Bangkok, Thailand

89

Gudimetal P. and Gopinath C.V. 7 Material Modelling Further to the first generation FEA shown in Figure 6 above, the piston performance was evaluated for different classes of Aluminium alloys (MAHLE 124, MAHLE 138 and MAHLE 142 series). Table 1 presents the material properties of these different alloys. Table 2 below shows the results of the parametric analyses of the three different alloys that were considered in this study. The results indicated that MAHLE 142 material model is best suited for the range of operating conditions for the piston under consideration. Further to this study, a load variation analysis was performed for the MAHLE 142 material model. The load was varied from 50% to 225% of the mean calculated load with a target value of the induced stress being equal to the compressive yield strength of the material. It can be seen from Table 3, that the MAHLE 142 is safe for up to a peak operating load of 200% the mean load (or about 6 MPa) acting on the piston head. The deformation for this loading was found to be 0.119 mm while the induced compressive stresses were very close to the compressive yield strength of the material (244 MPa). Figure 7 shows a graphical representation of the relationship between the induced stresses versus the pressure variation. 7 Conclusions The results of this study show that reverse engineering can be used in conjunction with finite element analysis to model obsolete product geometries and material models to redesign and generate new models to suit existing cylinders and other assemblies. We have also shown that the DesignXplorer within the ANSYS Workbench environment can be used very effectively to realize any parametric optimization. Acknowledgement The authors are thankful to M/s SAMKRG Pistons Ltd, Pydibhimavaram, Srikakulam (Dist) AP, INDIA for permitting and providing the necessary technical data related to pistons to execute this project and to Messers D.Pavan Kumarm, T.Sonesh and A. Srinivas (Final year B.E (Mech) Students, Department of Mechanical Engineering GITAM College of Engineering VISAKHAPATNAM 530 045 AP INDIA for carrying out the measurements and data processing.

(a)

(b) Figure 5: Temperature Profile and Equivalent Stress Distribution of the Piston

Figure 6: Safety Factor for the Piston

90

King Mongkuts University of Technology North Bangkok Press, Bangkok, Thailand

Finite Element Analysis of Reverse Engineered Internal Combustion Engine Piston Table 1: Physical & Mechanical Properties of MAHLE Al Piston Alloys [13] Properties Youngs Modulus E [N/mm2] Temperature (C) 20 150 250 350 20 150 250 350 20 100 20 200 20 300 20 400 20 20 150 250 350 MAHLE 124 80 000 77 000 72 000 65 000 155 156 159 164 20 21 21.9 22.8 2.7 200 250 180 230 100 150 40 65 MAHLE 138 84 000 80 000 75 000 71 000 143 147 150 156 18.6 19.5 20.2 20.8 2.68 180 220 170 210 100 140 60 80 MAHLE 142 84 000 79 000 75 000 70 000 130 136 142 146 19.2 20.5 21.1 21.8 2.77 200 280 180 240 100 160 50 70

Thermal Conduction coefficient, [W/mK] Mean, linear thermal expansion [1/k x 10-6] Density [g/cm3] Tensile Strength [N/mm2]

Table 2: Comparative performance of the four different aluminium alloys ANSYS Results Maximum Maximum Deformation Induced Stress mm N/mm2 0.062 168 0.055 151 0.052 149 0.079 146 Material Properties Permissible Yield stress stress(F.S=2) N/mm2 2 N/mm 190 380 210 420 230 460 260 520

S. No 1 2 2 3

Material Cast Aluminum MAHLE 124 MAHLE 138 MAHLE 142

Table 3: Performance of MAHLE 142 for different loading conditions Type of Loading 50% 75% 100% 150% 175% 200% 225% Mean Effective Pressure, Pm N/mm2 1.51 2.265 3.02 4.53 5.285 6.04 6.795 Deformation mm 0.051 0.062 0.079 0.091 0.101 0.119 0.128 Induced stress N/mm2 78 102 146 201 223 244 286

King Mongkuts University of Technology North Bangkok Press, Bangkok, Thailand

91

Gudimetal P. and Gopinath C.V.

350

Stress Analysis in ANSYS

300

250

Induced Stress, Mpa

200

150

Induced stress N/mm2 Permissible stress, N/mm2

100

50

0 0% 50% 100% 150% 200% 250% Pressure Load Variation in %

Figure 7: Stress Analysis of Piston References [1] Chiang Y. M. and Chen F. L., 1999. Sculptured surface reconstruction from CMM measurement data by a software iterative approach, Int. J. of Production Research, 37(8). [2] Chang-Xue Feng and Shang Xiao., 2000. Computer-Aided Reverse Engineering with CMM for digitization and LOM for duplication, in Proceedings of the 4th Intl Conference on Frontiers of Design and Manufacturing, Intl Academic Press, Beijing, China, 256-262. [3] Tzung-Sz Shen., J. Huang and C-H Menq., 2000, Multiple-Sensor Integration for Rapid and High-Precision Coordinate Metrology 110 IEEE/ASME Trans. on Mechatronics, 5(2). [4] Feng C.-X., 2003. Internet-Based Reverse Engineering, Int. J. of Advanced Manufacturing Technology, 21(2): 138 144. [5] Zhongwei Y. and Shouwei J., 2003. Automatic segmentation and approximation of digitized data for reverse engineering, Int. J. of Production Research, 41(13): 3045 3058. [6] Swanson J., Schok D., Kelley A. and Callow D., 2003. Investigation of Reverse Engineering and Redesign of Milkshake Maker, Internal report submitted for Mechanical & Aerospace Engineering Department, Arizona State University. [7] Markine V.L., Deman A.P. and Esveld C. A procedure for design and optimization of a railway track structure, 1-10. [8] Kettil P. and Wiberg N. E., 2002. Application of 3D Solid Modeling and Simulation Programs to a Bridge Structure, Engineering with Computers, 18(2): 160-169. [9] Yasar Deger., Optimisation of the Dynamic Behaviour of a Machine Tool Mounting Device FENET/NAFEMS Seminar: FEM in Structural Dynamics, Wiesbaden, Germany [10] C.H. Li., Piston thermal deformation and friction considerations, SAE Paper 820086, 1982. [11] Properties and Selection: Irons, Steels and High Performance Alloy, ASM Handbook, vol. 1, ASM International, 1990. [12] Y. Liu. and R.D. Reitz, Multidimensional modeling of combustion chamber surface temperatures, SAE Paper 971539, 1997. [13] Handbook of Internal Combustion Engines, SAE International.

92

King Mongkuts University of Technology North Bangkok Press, Bangkok, Thailand