Impeller Design for a Microjet Engine
Transcription
Impeller Design for a Microjet Engine
Impeller Design for a Microjet Engine Cuneyt KENGER TUSAŞ Motor Sanayii A.Ş., Muttalip Mevkii 26003, Eskişehir, Turkey Bulent ACAR TUSAŞ Motor Sanayii A.Ş., Muttalip Mevkii 26003, Eskişehir, Turkey Ferhat SAHİN TUSAŞ Motor Sanayii A.Ş., Muttalip Mevkii 26003, Eskişehir, Turkey Abstract The use of microjets has found application areas in the propulsion of target drones and missiles. These propulsion systems have their own challenging topics due to their small sizes and very high rotational speeds (in the order of 100000 rpm). The complete engine consists of the following main components: Inlet, Impeller, Diffuser, Combustion Chamber, Turbine and Nozzle. In this paper, the focus was given to present the methodologies used in the design of the impeller (radial compressor) only, which is one of the most critical components of an engine. The main function of an impeller is to efficiently compress the air coming from the inlet and to deliver it to the combustion chamber via a diffuser. The compressor pressure ratio is highly dependent on the impeller tip speed. Hence, high tip speeds are essential to obtain high pressure ratios. Therefore, in microjets, high rotational speeds are inevitable. This results in many challenges such as vibratory behavior, transonic flow, shock waves in diffusers and high stresses. Furthermore, tip clearance is one of the most important concerns affecting impeller performance. Special care must be taken to maintain tip clearance close to an optimum value under operating conditions. Introduction TEI was established in 1986 and has been assembling, manufacturing, testing and designing aircraft engine parts in Turkey. The initial area of activity during TEI’s development was engine assembly which then expanded into gas turbine engine parts manufacturing. TEI started its research and development activities in 1996 and has been participating in international development programs of military aircraft engines. In addition to participating in collaborative projects, TEI is also undertaking internal projects in order to continuously enhance its design capabilities. In this scope, TEI started working on developing and manufacturing a turboprop engine (see Figure 1) to be used in unmanned air vehicles (UAV) in target drones (see Figure 2). This project was started in September 2004 and planned to be completed in two phases. As part of the first development phase, TEI-TJ-1X (see Figure 3) was developed and first engine test and performance tests (see Figure 4) were completed in May 2005. This turbojet engine mainly consists of inlet, compressor, combustion chamber, turbine and nozzle. In the second phase of the project, a turboprop engine (TEI-TP-1X) will be developed using the core of the turbojet engine. Once this turboprop engine has been developed various performance tests will be conducted. Figure 1. TEI-TP-1X Turboprop engine Figure 2. The target drone TAI S-38 Turna Figure 3. TEI-TJ-1X Turbojet’s sectional view Figure 4. TEI-TJ-1X is seen during open-air tests In this paper, the following analyses for the impeller of TEI-TJ-1X have been presented with details in the following order; • • • Structural Analyses of the Impeller Modal Analysis of the Impeller Full Modal Analysis of the Rotor In addition to these analyses, detailed rotor dynamic and fatigue analysis were performed, which will not be discussed in this paper. Structural Analysis of the Impeller TEI-TJ-1X impeller (radial compressor) shown in Figure 5 was machined from 7XXX series aluminum. During the design process of this impeller, investigations were carried out in order to determine stresses and deformations, which led to the identification of possible optimization regions. Various geometries were analyzed under mechanical loads, among which the best ones were selected for further analysis including application of aerodynamic (fluid) loads as well. The impact of accurate estimation of fluid loads on deformations was studied by applying the loads successively (i.e. mechanical, mechanical+fluid pressure, mechanical+fluid pressure+fluid temperature) and comparing the impact of each load type on stresses and deformations. Figure 5. Impeller & Inlet Definition of the Problem Since the impeller geometry is cyclic symmetric, 1/6 of the geometry was modeled in UG. The sector was constructed using the fluid volume boundaries in order to enable data transfer from CFD analysis. For each geometry, finite element model was created with identical mesh density, solid 92 elements were used (see Figure 6) and each geometry was meshed with nearly 60000 elements. In order to take the effect of aerodynamic loads into account, fluid pressure and temperature values were transferred from CFD model into ANSYS (Reference 1) by an in house developed code (FAT-P/T). Mesh densities of CFD and FEM models do not need to be the same for transfer operation. FAT-P/T is capable of interpolating for different mesh patterns automatically (Reference 2). Figure 6. Impeller Mesh full and cyclic model Mechanical and aerodynamic loads (fluid temperature and pressure) were applied on the geometries and due to lack of sufficient knowledge on some boundaries; assumptions simulating the worst cases were applied. As mechanical loads, max. 120000 rpm rotational speed and 2200 N bolt pre-load were applied. As for fluid load; pressure and temperature distribution for air passages, ambient pressure and temperature, backface pressure (0.2MPa-0.3MPa), backface temperature (120-150ºC) were applied. Results Compressor tip clearance has an important effect on the overall performance and operability of a compressor and has wide-spread consequences involving all components such as rotor, blades and casing design. The outer contour of the impeller was designed in such a way that 0.2 mm constant tip clearance will be maintained at all locations, while the engine runs at max speed. According to the performed analyses of inlet and impeller, the contours are determined respectively. Due to the geometrical form of impeller (radial compressor), control of tip clearance is not as easy as in axial compressors or turbines. Therefore, one of the most challenging aspects of the impeller design was to handle the tip clearances. Thermal expansion of the impeller, inlet and rotor, radial expansion due to centrifugal forces, rotor orbits due to unbalance, bearing clearances and production tolerances are some of the most important factors in determining tip clearances. If rubbing occurs due to any reason between rotating impeller and stationary inlet (casing), this may result in a catastrophic engine failure. For this reason, most of the effort during impeller design was dedicated to maintaining a safe and efficient tip clearance value. During open air tests, engine was tested at various speeds including max. speed of 125000 rpm to simulate transient effects, no sign of rubbing either on impeller blades or inside of the inlet was observed. As stated previously, the biggest impact on tip clearance comes from both mechanical (centrifugal ) and aerodynamic (fluid pressure + temperature) loads. This effect was investigated for various loads. As a result of finite element calculations, it can be concluded that fluid pressure has the least effect among all loading types. Mechanical loads are especially dominant around the inlet section of the impeller and thermal loads are as dominant as mechanical loads for exit section of impeller, where fluid temperature reaches max. 150º C. Figure-7 depicts radial deformations for each different type of loading. Basically, tip deformations were investigated under 3 different load cases, additionally sensitivity analyses were performed to understand the impeller response against different magnitudes of aerodynamic loads taking place at backface. - Mechanical (centrifugal) loads Mechanical loads + fluid pressure Mechanical loads + fluid pressure + Temperature Figure 7. Effects of loads on radial deformations Similar studies were performed for axial displacements, since axial deformations also contribute to tip deformations considerably. Tangential deformations, on the other hand, have no critical effect on calculation of tip clearances. Figure 8 shows axial and radial deformations of the impeller of TEI-TJ-1X turbojet engine. MIN MAX Figure 8. Axial Deformations (left) and Radial Deformations (right) The tip clearance of an impeller must meet the following requirements: • avoidance of rubbing for normal flight maneuvers and surges • acceptable clearance during cold starts • acceptable clearance during idle to maintain adequate surge margin • minimum clearance for conditions of cruise and max. climb to achieve the best efficiency • no blade rub during hot reburst (transient performance) The Figure 9 shows the tip clearance when the impeller and inlet are deformed at max rotational speed. Figure 9. The tip clearance when the engine is running at max rotational speed. The deformations of the inlet inner surface were seen to be very small in structural analyses (about 3 microns). So the deformations of the inlet inner side are not taken into consideration. In the figure 10, static tip clearances are shown: Figure 10. Static and deformed contours of impeller and cover As the stresses are considered, there are three hot spots for the impeller depicted in Figure 11 : Figure 11: Hot Spots for the impeller geometry Results of the optimization study for the bore of base and other impeller models are given in the below Figure12. Through optimization studies, various bore geometries were generated and consequently finite element analyses were performed. Due to both production difficulties and unsatisfactory results, no change in the bore region was made. Base Line Modified 1 Modified 2 Modified 3 MIN MAX Figure 12: Von Misses Stresses under mechanical and aerodynamic loads Results of the optimization study for the backface of Base Impeller and the corresponding equivalent stress distributions are presented in Figure 13. As expected, stress values around the transition radius were decreased substantially, after introducing a generous radius in this region. On the other hand, this change had no considerable impact on the stresses around splitter radius and bore. The only drawback of this optimization is a weight increase of 5% (5 grams). Weight: 94 g MIN Weight: 92 g Weight: 99 g MAX Figure 13. Von Misses Stresses under mechanical and fluid loads for various cases Modal Analysis of the Impeller Definition of the Problem In this study, to determine dynamic characteristics of the impeller, several modal analyses were performed. Vibration characteristics of the impeller must be investigated in detail to prevent high cycle fatigue (HCF) and rubbing due to interaction of impeller blade tips with stationary inlet as a result of excessive vibration. Subsequently, campbell diagram was constructed. According to results, none of the impeller frequencies intersect with any engine order for the operational range of impeller speed between 100000-120000 rpm. Results Figure 14 is a Campbell diagram showing eigenfrequencies for the designed impeller against the rotational speed. The frequencies change slightly with the engine speed due to both temperature dependent Young’s Modulus and stress stiffening characteristic. Engine orders are also plotted as multiples of engine speed as sloping lines originating from 0 rpm. Impeller resonance takes place, provided that the lines of each impeller frequency crosses through the relevant engine orders. But this itself is not the only criteria. Details will be discussed later. The design target is to keep the speed range between idle and redline speed free from resonances (i.e. crossings with the basic engine orders) and engine orders equal to the airfoil numbers of the upstream and downstream blade rows. Figure 14. Campbell Diagram for the Impeller Pre-stressed modal analysis were performed according to the given procedure ( Figure 15) as explained in ANSYS documents. Figure 15. Procedure for pre-stressed modal analysis for cyclic geometries Since cyclic sector angle is 60º. One can get only 3 harmonic index. To illustrate the mode shapes at each harmonic index (i.e. 0,1,2,3), 4 pictures were presented below in Figure 16. An interference diagram presented in Figure 17 similar to campbell diagram can be plotted to see whether operating speed crosses any nodal diameter as well. The idea behind using Interference diagram is to check engine orders with nodal diameters. Campbell diagram itself is not self explanatory. To have a resonance, frequency crossing with engine order is not a unique condition as stated previously. Nodal diameter and engine order must be the same and to visualize it, interference diagram plays an important role, whose x axis shows nodal diameter and whose y axis shows frequencies for the case presented in Figure-17, operation limits of 100.000 rpm and 120.000 rpm are used and 2 lines representing these rpms were drawn and checked against any crossings with nodal diameters. Intersections show potential resonance frequencies. To avoid resonance problem, various models were created and analyzed and then stress and displacement values were recalculated. As expected, there is a contradiction between lower stress values and higher natural frequencies. At this stage of the design, an optimization is inevitable. There is no obstacles like guide vanes in upstream for the impeller but there are some vanes located on the diffuser in downstream. Number of them was taken into account in campbell diagram and in the final diagram we realized that there were some frequency-crossings in operational range + safety margin area at higher frequencies and modes, which are in general not dangerous. To make sure, we run the engine at various speeds and observed nothing wrong in terms vibration and rubbing. Figure 16. Mode Shapes of Impeller for various harmonic indexes Figure 17. Interference Diagram (frequecy v.s. nodal diameters) Full Modal Analysis of the Rotor Definition of the Problem In this study, dynamic analysis of rotor (see Figure 18) including impeller, shaft and turbine is performed to determine resonant frequencies of each component separately. The most difficult part of the full modal analyses of the rotor was to define realistic boundary conditions of bearings and bearing housings, whose stiffnesses affect modal response substantially. In order to calculate the bearing housing stiffness values correctly and precisely, whole engine model (WEM) was created. Contact option was used to tie the different mesh patterns belonging to the rotor parts shown in figure 19 for dynamic analysis. Figure 18. Machined Rotor and its 3-D CAD Model Figure 19. Rotor FE Model Results According to results, critical resonance of the impeller is around 60.000 rpm, for turbine this value is around 90.000 rpm and first critical shaft speed is 157.000 rpm , which is well above the max operating speed of 125000 rpm with a safety margin of 25 %. During engine run, between 30.000 and 120.000 rpm, no resonance was observed for any of the part. The reason is maybe due to enough damping at bearings (i.e. one o-ring is placed outer periphery of the front bearing & lubrication). Figure 20. Mode shape of the Impeller Figure 21. Mode shape of the Turbine Figure 22. Mode shape of the Shaft Conclusion Completion of the design and manufacturing processes has resulted in the development of the first gas turbine engine . The capabilities gained through this project will be used in subsequent national and international projects. Eventually, as FE group at TEI, we have demonstrated that ANSYS itself was a satisfactory tool for structural, thermal and dynamic analysis in the design of a microjet engine. Reference 1) ANSYS Inc., 2004, Users Guide for Release 9.0. 2) ANSYS User Meeting Turkey, 2005, Bulent Acar