Aircraft engine component design and optimization

The main objective of present thesis is to design and optimize rotating disk of aircraft engines and calculate its burst speed and establish a metodology for disk design. In aviation industry, engine compressor disk are made of titanium or inconel alloys based on the metal temperatures of the part. Thus, for the front stages of compressor titanium alloys and for the aft stages of compressor inconel alloys are applicable due to material strength alteration with respect to temperature. In this study Ti 6AL 4V alloy have been chosed with respect to operating material temperatures of design with respect to benchmark engine. This thesis uses the CF6 engine as benchmark with respect to data shared by NASA articles. Initial disk geometry, operating temperatures and disk rotational speed are taken as reference from CF6 engine. In the disk geometry, 3D features like bolt holes, slots and blade dovetails have been excluded since this design approach does not include this kind of features in initial design. The disk design consists of a list of study like comparing the von Mises equivalent stress vs yielding of material, LCF and HCF analysis, creep analysis, fracture mechanics and modal analysis and burst speed calculations. However, it can be said that for an initial design comparing material yield strength vs engineering study results and checking burst speed is sufficient. This study includes 3-step analysis for disk optimization. First step presents the initial chosen geometry and verification of analitical calculations and finite element analysis without blade loading condition (external loading) but only applying centrifugal forces due to rotation and temperatures and termal expansion. In the second step, blade loading is included and verified by expected boundary condition on inner and outer diameter of the disk and comparing analitical and finite element method. The third step is the optimization of the disk geometry interms of finding the minimum weight and reaching the maximum allowable von Mises stress, which is almost equal to yield of material in this study. Discretization method have been applied on the structure by using analytic equations for the simulation of a real disk. By this way, disk behaviour can be simulated by smaller parts. Therefore, real disk geometry has been divided into 5 constant thickness disks and calculations have been conducted for each one with respect to equilibrium and compability equations, and boundary conditions. For the calculations without blade loading, boundary conditions for radial stresses are set to zero for the inner and outer diameter of actual disk since there is no element that will transmit the subject loading. For the calculations including the blade loadings the boundary condition for radial stress on the outer radius of disk is generated by blade loading and applied into disk discretization method. Geometry obtained by using discretization metod is analyzed by finite element method. Analysis have been conducted by using ANSYS program. 2D axisymmetric element SOLID272 has been used for disk modelling. This element has 3 degrees of freedom X, Y, and Z on translation direction with four nodes and sufficient for disk modelling. The blade is modelled by MASS21 element by assuming that blade is a point mass over outer radius of disk. MASS21 has 6 degrees of freedom on X, Y and Z axises on translation and rotation. Entire disk and blade was modelled as they will rotate about Y axis. Mass real constant set to infinitely small for MASS21 and the blade centrifugal load is applied on mass element as an external force. Temperature mapping was approximately uniform for each 5 parts as described in discrezation method and temperatures are increasing approximately from inner radius to outer radius. Finite element analysis results include axial stress component on especially inner radius of the disk where analytical calculation excludes. Axial stress component increases when there is discontinuty or sharp transition on the inner radius. The reason behind of that, it is expected to be non-uniform loading of inner radius due to centrifugal force. This condition imposes the inner radius to bend and therefore, axial stresses occur. However, 2D analytical formulations exclude the component in axial direction. 2D analytical calculations take account tangential stress which is in the hoop direction and radial stress in centrifugal force direction. Therefore, one reason of difference between stress results of finite element method and analytical calculations is the ratio of Poisson since it is expected from axial stresses to contribute stresses on hoop and radial direction. Comparsion results show that discretization is close to finite element method by %10. By using both approaches, stresses on the disks are shown and the critical locations where the designers should be careful, are determined. This study shows the first critical locations for the designed disk as inner diameter of disk which is named bore and the second one is the transition fillet radius to outer diameter which is named rim. Burst phenomenon is the break of disk due to disk stresses reaching the ultimate tensile strength of material. This condition is called as uncontained event. Because when a rotating disk fails and breakes into small fragments, engine case structure cannot contain disk pieces. It is expected from an engine not to operate close to burst speed. However, there could be many reasons that burst phenomenon can be seen like failure of engine shaft or break of engine control system that will result in overpseed. Because of this reason, burst speed which is limited and controlled by the aviation authorities like FAA and EASA, is calculated in this study. Burst calculations have been conducted by using Hallinan equation. Elastic and plastic behaviour of disk material have been considered during disk burst speed calcualtions. Elastic and plastic behaviour of the disk material have been accounted by S factor which is represented in Hallinan equation. The S factor is an empirical value which is usually generated by tensile test of a notched specimen. For highly plastic materials this factor is close to 1, and for highly elastic materials it is close to 0. The S factor represents the load sharing capability of the material during high loads. Hallinan equation consist of local maksimum tangential stress and average tangential stress for a specific selected section. The relation of local and average stress is arranged by S factor in the equation. Burst speed is calculated for bore location and the transition fillet radius to rim locations since these two locations have the highest tangential stresses. In this study, the S factor is chosen as 1, 0.5, and 0 to investigate its impact on burst speed on the critical locations of the disk. Results shows that for the case where the high loads shared by the cross section (S=1) of the critical locations have higher burst speed than the case which there is no load sharing along the cross section (S=0).