TURBINE BLADE Tulsidas. D*,Dr.Shantharaja. M,a Dr. Kumar. Kb
Address for Correspondence
* Mechanical Engineering, School of Engineering & Technology, Jain University, India. a Mechanical Engineering, UVCE, Bangalore, India b Mechanical Engineering, PESCE, Mandya, India ABSTRACT The large variety of turbo-machinery blade root geometries used in industry prompted the question if an optimum geometry could be found. An optimum blade root was defined, as a root with practical geometry which, when loaded returns the minimum fillet stress concentration factor. The present paper outlines the design modification for fillet stresses and a special attention is made on SCF of the blade root (T-root) which fails and to guarantee for safe and reliable operation under all possible service conditions. Finite Element Analysis is used to determine the fillet stresses and Peterson’s Stress Concentration Factor chart is effectively utilized to modify the blade root. The root is modified due to the difficulty in manufacturing the butting surface of the tang which grips the blade to the disk crowns having small contact area. KEYWORDS Steam Turbine Blade Root, Stress Concentration Factor, Fillet Stresses, Disk stresses. 1.0. INTRODUCTION have become available to optimize the shape and In most axial flow turbo machinery used today the minimize the stress values so as to improve the blades which direct the flow of steam are structural integrity of the bladed disk [3]. manufactured separately to the turbine disks on The tang in steam turbine blades play a vital role in which they are then mounted. The blade rotor preventing the disk opening during operating connection is highly and complexly stressed area condition as well off design condition but the which if not properly designed can fail with serious criticality lies in establishing the contact between the consequences. In critical areas the fillet stresses are blade tang and disk crown across the circumference, modified using the Petersons stress concentration Since the contact area is approximately less than factor chart. 1mm.The manufacturing feasibility would be a real Stress Concentration Factors are significant in challenge. Estimating the butting forces at the tang machine elements as it gives rise to localized stresses, region is an another challenge. In the present work any change in the design of surface or abrupt change one such effort is made in looking at the blade root in the cross section. Almost all machine components without the tang with minor modifications at the and structural members act as ‘Stress risers’. These blade root butting area in the blade and the disk root discontinuities are very dangerous which lead to fillet to reduce the peak stresses as an initial work at failure due to the Peak Stresses introducing cracks. blade root fillet and to make a sensitivity study using These cracks propagate to catastrophic failure before Peterson’s chart and to modify the blade root to the desired design life. Hence, it is very much reduce the stress concentration factor at blade root essential to analyze the stress concentration factors fillet as well the disk fillet stresses. for critical applications like Turbine Rotor blades. In 2.0. TURBINE BLADE OVERVIEW the light of the above Finite Element Analysis (FEA) Different stages of the turbine parts of turbo with fine mesh in the vicinity of the blade root of machinery use different blades, the properties of Steam Turbine Rotor is applied to determine stress which are designed to meet the specific operating concentration factors. criteria of those stages in which they are used. The Geometric discontinuities cause a large variation of turbine stages often use blades with internal cooling stress locally, and often produce a significant while the very long blades in the final stages of large increase in stress called as ‘Stress Concentration’. steam turbines are pretwisted so that when the This can also appear when loads are applied over a centrifugal force acting on them untwists the blades small area or at a point. Geometric discontinuities are to the correct angular alignment with the steam flow often called as ‘Stress Risers’. Examples of stress is achieved [4]. risers include holes, notches, fillets and treads in a Often the high pressure turbine stage and the last low structural member. Geometrically the ratio of pressure stage use multiple tangs T-root type blade maximum stress to the average or nominal stress is roots while the short and medium impulse and the called as Stress Concentration Factor (SCF) and is low pressure stages use single tang T-root type blade denoted by Kt [1]. roots. The design differs from one machine to the The most comprehensive source of stress other and a lot depends on the design policy followed concentration factors for commonly encountered by the specific manufacturer. For instance some geometries has been compiled by Peterson. The companies use a large number of very small tangs on relation between two and three dimensional stress their roots while others prefer to use only a few tangs. concentration factors is made by assuming an There are various ways of fitting the roots into the analogy exists between a circumferential fillet and a rotor. Some roots are fitted circumferentially into the circumferential groove. This is the limitation of the rotor while others are fitted axially. Peterson charts for estimation of the stress 3.0. BLADE TERMINOLOGY concentration factors [2]. The optimization methods The blade terminology is as shown in Fig.1.
International Journal of Advanced Engineering Technology E-ISSN 0976-3945
manufacture the tang will be more which increases
the product cycle time. Hence the manufacturing cost will be high. So to reduce the manufacturing cost, time and the difficulty in manufacturing the tang, the tang is removed in the modified blade and attained the stresses within the limits or nearer to it. 7.0. DESIGN CONSIDERATIONS AND APPROACH The design rule margins are necessary for uncertainties present in material, manufacturing, assembly and on site operating conditions. In the present work as per API standards 121% margin is considered for 11% over-speed due to full fuel throw–off condition. Additional margin of 15% is Fig.1Blade Nomenclature considered for manufacturing uncertainties [5]. 4.0. OBJECTIVES OF THE PRESENT WORK 7.1. Factor of Safety limits: The present work is divided into two parts: • To identify the modifications of blade root required using Peterson’s chart within the linear zone to reduce the peak fillet stresses in the blade and disk. • Verify the same using Finite Element Yield strength at room temperature = 550 MPa Analysis for two cases with and without the Yield strength at operating temperature (1000) = 540 tang in the blade. Firstly, to study the fillet MPa stresses with tang and then Petersons chart is Factor of safety at operating speed = 1.68 used to reduce the peak stresses with the Factor of safety at overspeed = 1.15 modification to the butting area and Allowable stress at operating speed: reducing the fillet radius. To conduct the σ minimum sensitivity analysis for the fillet stresses in σ allowable= blade and disk using FEA and then to (overspeed) 2 * (addiionalFOS ) compare the performance of blade and disk 540 fillet stresses for the above two cases. σ allowable = 5.0. BLADE ROOT MODEL 1.68 The original model of blade root and disk with tang is σ allowable = 321 MPa as shown in the Fig.2. In this model the butting Allowable stress at over speed: surface of the tang in root and disk is the difficult part σ min imum to manufacture. Based on the optimization and with σ allowable = the use of Petersons SCF chart the blade root is overspeed modified as shown in the Fig.3. The tang which was 540 difficult to maintain the contact butting surface σ allowable = between the disk and the root in the original model is 1.15 modified by removing the tang and attaining the same satisfactory results. The geometry of the base σ allowable = 469 MPa line model and modified model are same in all the The axial deformation allowable is 2mm between dimensions except the modifications in r, d, D and M. shroud and the casing for preventing the tip rub. 7.2. SCF using Peterson’s Chart [6] Base line model with tang For r/d and D/d ratio Kt is 2.1. Modified model without tang For r/d and D/d ratio Kt is 1.95. Using Peterson’s Stress concentration factor chart and the knowledge based design criteria the stress concentration factor is reduced by modifying the blade root dimensions of base line model and Figs.2 & 3 Base line & Modified T-root model completely removing the tang in the modified model 6.0. BLADE WITH TANG are shown in Fig 4. The blade and disk assembly with tang provides the blade safety in taking off the blade from the disk. The tang holds the blade coming or pulling out of the disk when the centrifugal force is applied. The Vonmises stresses are high at the tang which leads to crack growth and leads to fracture near the tang. The main difficulty in using the tang for the model is manufacturing the tang with a small contact area Fig. 4 Blade root dimensions to be modified & which is a very difficult as well as the time taken to modified dimensions IJAET/Vol.III/ Issue I/January-March, 2012/XXXX International Journal of Advanced Engineering Technology E-ISSN 0976-3945
σnom = 65.12 N/mm2
8.0. BLADE CENTRIFUGAL FORCE σmax =126.96 N/mm2 CALCULATION 9.0. FINITE ELEMENT MODEL The governing equation for centrifugal force is: F = mω2r (1) Where m is the mass of the moving object, r is the distance of the object from the centre of rotation (radius of curvature) and ω is the angular velocity. Consider a small segment of mass ‘δm’ of length having width ‘δr’ at a distance r from the centre. Then the equation for the force ‘δF’ on this small segment is given by: Fig.5 FE Model of blade root and disk(Base line δF = δmω2r (2) model) Assuming the blade to have a constant cross sectional The geometry of one blade root with disk is modeled and the finite element model of bladed disk is area A and material density ρ , we can write: generated using ANSYS 12.0 as shown in Fig.5. δm = ρAδr PLANE 42 Elements is considered for finite element Then equation (2) is written has model and meshed with defining the element size. δF = ρA ω2rδr Surface to Surface contact pairs were created Or formally it is written has between the blade and disk mating surfaces. The dF = ρA ω2rdr (3) plane boundary conditions and different constraints Let r1 be the radius of the disk and r2 is the distance are applied assuming the blade and disk are made up between the centre of the disk and tip of the blade. of Chrome steel (X28CrMoNiV49) material having Then, integrating equation (3) along the total length properties like Young’s modulus, density, Poisson’s of the blade, the total force acting on the blade is ratio and yield strength are 2.1x105 MPa, 7.85x10-9 given by: kg/mm3 , 0.3 and 550 MPa are the input to ANSYS. r2 The following discussions for the comparison and F = ρA ω2 ∫ rdr design decision are based on the stresses for the base r1 line model and the modified model. The geometry So, F = ρA ω2 (r22-r12) /2 (4) shows the Vonmises stress of 125MPa and 1st Principal stress of 133MPa at the tang in a blade root The angular velocity is converted from revolutions and disk as shown in Fig.6. with the design speed of per minute (rpm) to radians per second using the 3000rpm. relationship: 2π n ω= (5) 60 Knowing the values for the cross-sectional area, density, angular velocity and radii, the force on one blade is calculated. Once the force is calculated we can estimate the nominal stress σ on the blade root using the relation: F Fig.6 Vonmises stress and First Principal stress σ= (6) A root with design speed of 3000rpm. Where Aroot is the cross-sectional area of the blade For 121% over speed, the Vonmises stress is of root [7] [8]. 184MPa and 1st Principal stress of 194MPa of the 8.1. Analytical Calculations Base line model are as shown in Fig.7. The given are the following data. Material Density, ρ= 7.85x10-9 kg/mm3 Speed, n = 300rpm Angular velocity, ω = 314.2 rad/sec Blade tip radius, r2 = 273mm Disk radius, r1 = 250 mm Blade length, (r2-r1) = 23 mm Substituting the above values in equation (4), we get force and stresses for Base line model as F= 10906.8 N Fig. 7 Vonmises stress and First Principal stress σnom=F/Aroot with over speed of 3630rpm. σnom = 57.09 N/mm2 The Aspect ratio is checked for both baseline model σmax =Κt σnom and the modified model. Aspect ratio is defined as σmax =119.9 N/mm2 the ratios of the longest edge to the shortest edge.The For Modified model Aspect ratio of the base line model were well within F= 12640.7 N the limit of 5 as shown in the Fig.8 and Fig.9.
IJAET/Vol.III/ Issue I/January-March, 2012/XXXX
International Journal of Advanced Engineering Technology E-ISSN 0976-3945
For 121% over speed the Vonmises stress of 200MPa
and 1st Principal stress of 195MPa in the modified model are as shown in Fig.12. The Aspect ratio is also checked for the modified blade root model without tang and it was well within the limit of 5 for Quad elements as shown in Fig.13.
Fig. 8 Aspect ratio at Fillets of the Disk
Fig. 13 Aspect ratio at Fillets of the Disk and
Blade Root (Modified). CONCLUSION Fig.9 Aspect ratio at Fillets of the Blade Root. The design considerations made to address the structural integrity of the blade and the sensitivity analysis for geometric parameters with aid of finite element analysis has resulted in the following observations during the course of study: • Since the blade root geometry was straight root, the design principles of Peterson’s Stress concentration factor chart was effectively utilized to achieve the blade root dimensions for reduced stress concentration factor and fillet Fig.10 FE Model of Modified blade root and disk. stresses comparatively for two cases. • The base line model and the modified root model was subjected to centrifugal loads and the resulting centrifugal stress was compared with hand calculations by which it can be concluded that effective usage of blade roots without tang can reduce the uncertainties in manufacturing and malfunctioning of tang root blades which are well within the design considerations. REFERENCES 1. R. Nagendra Babu & Dr. J. A. Tamboli, “Determination Fig.11Vonmises stress and First Principal stress of Stress Concentration Factors for Filleted Shafts in with design speed of 3000rpm. Tension using FEA”, Journal of Shivaji University, Volume No.38, 2003. In the modified model the tang is eliminated and the 2. R.Nagendra Babu, K.V.Ramana and K.Mallikarjuna fillet is modified in which the blade root area will be Rao, “Determination of Stress Concentration Factors of more and the stresses generated will be more in the A Steam Turbine Rotor by FEA”, World Academy of modified model. The FE model is as shown in Fig.10. Science,Engineering and Technology, Volume No.39, 2008. The modified Finite element model is analyzed by 3. J.S.Rao and S.Suresh, “Blade Root Shape applying the same boundary conditions and solved, Optimization”, Altair CAE Users conference, Aug 3-5, the results of the Vonmises stress is of 138MPa and 2006, Bangalore. 1st Principal stress is of 133MPa as shown in Fig.11. 4. George Hettasch, “Optimization of Fir-Tree Type Turbine Blade Roots Using Photoelasticity”, University The of Stellenbosh, Dec 1992. Vonmises and 1st Principal stresses are maximum at 5. Mahesh Shankar, K.Kumar and S.L.Ajit Prasad, “T- the blade root fillet because the area in the blade root Root Blades in A Steam Turbine Rotor: A Case Study,” will be more when the tang is removed in the Engineering Failure Analysis 2010. 6. Peterson.R.E, “Stress Concentration Factors”, John modified model. Wiley & Sons, ISBN 0-471-683329-9, Page No.57. 7. Gary,“Forces on Large Steam Turbine Blades”, RWE npower Industry, The Royal Academy of Engineering. 8. G.D.Robinson,“Review of Fatigue Failure in Spindle Blades of A 120MW Steam Turbine”, IPENZ Transactions,2000,vol.27No.1/Gen25.
Fig.12 Vonmises stress and First Principal stress
with over speed of 3360rpm. IJAET/Vol.III/ Issue I/January-March, 2012/XXXX