Jean Michel Puybouffat 1, Dionisio Agostinho 2 ABSTRACT

Transcription

1 The 12 th International Conference of the Slovenian Society of Non-Destructive Testing»Application of Contemporary Non-Destructive Testing in Engineering«September 4-6, 2013, Portorož, Slovenia More info about this article: ULTRASONIC EXAMINATION OF LP STEAM TURBINE SHRUNK-ON DISCS METHOD (DESIGNED THROUGH MODELING), PROCESS AND TOOLS, DATA CAPTURE, ANALYSIS & POST PROCESSING) Jean Michel Puybouffat 1, Dionisio Agostinho 2 1 NDT Engineer, Rotek Engineering, Johannesburg, South Africa, jm@vsonomatic.co.za 2 NDT Engineer, Rotek Engineering, Johannesburg, South Africa, dion.agostinho@eskom.co.za ABSTRACT The different manufacturing technologies of Low Pressure steam turbines for the power generation industry include solid turbines (mono-block), welded turbines and shrunk-on disc turbines. For shrunk-on disc LP turbines, the turbine rotor is constructed by thermally expanding the discs and mounting each disc onto the spindle allowing it to shrink fit. This widely used technology allows for ease of manufacture and maintenance. The discs are aligned and locked in their circumferential positions by way of a pin and drive pin hole arrangement. However, this type of assembly is a source of risk of cracking due to the inherent stress. Additional factors that contribute to the risk of cracking are stress corrosion, fretting action and fatigue, due to operating condition.the areas of concern are the shrink-fit area between shaft and disc and the drive pin holes on the lateral face of the discs. During the regular maintenance program the client requires that a non-intrusive examination of these areas be carried out, without dismantling the turbine. The only means of inspecting the suspect area is to use an ultrasonic test method. This paper presents the process followed from the stage of receiving the specification through the development of the procedure, to the performance of the inspection; The process of examination design, through the use of modeling (ray tracing and CIVA)- The probe arrangements, optimal positioning through modeling, probe positioning assemblies The ultrasonic package, typical data captures, analysis methods, and reporting. The displayed results come from existing UT software package applied on LP in service inspections worldwide.. Key words: Steam Turbines, Shrunk-on discs, cracks, modeling, ultrasonics). 1. Introduction The scope of examination of this design of LP shrunk-on disc turbine covers mainly the discs and sometimes the keyways of the center spacer (when present) for dual flow machines. For the discs, that is topic of this paper, their most stressed area that is necessary to monitor, are the drive pin holes and the bore of the disc, as indicated on the drawing below. 97

2 BORE DPH Stressed areas Figure 1: Shrunk-on Disc LP turbine The monitoring is applicable for the drive pin hole (DPH) both above, below and on the bottom, as well as for bore on the highly stressed part the shrink fit. The direction of propagation of possible defects is as described below. Radially Axially (DPH) Axially (Disc) Figure 2: Direction of propagation of indications The indication reference sizes defined are; for above DP (ADP) 3mmx6mm,below DP( BDP) 6mmx6 mm, Bore 6x6mm or 6mx9 mm ( HxL).The requirement for the inspection imposed by the utility, is for the detection and sizing to be performed either with the rotor in-situ (in casing) or on rollers and to be carried out in the shortest possible time. 2. Assumptions From the requirements and conditions of examination some assumptions were necesary: Ultrasonic design : The ultrasonic design would define the package to carry out the examination. The package to be used for the inspection would be dependent on: the applied method of inspection, type of probes, wave mode, design of wedges, required refraction angle, coverage of each probe and number of probes. The key points of design are distributed as below: Principles of ultrasonic design 98

3 PE, TOFD 1) Basic inputs 2) Wave Refraction angle per method 4) UT package 3) Coverage assessment, probe location Probe (Ø, F number, type of wedges, PCS, location) Figure 3: Ultrasonic design process 3. Ultrasonic design 3. 1 Background Disc geometric data and basic beam aperture calculation for each area to be scanned was used to compile the UT package. The effect of size of crystal and frequency on resultant beam aperture was evaluated. The specific disc geometry was introduced in this computation for each position of the probe : seated on hub, seated on the wing and the specific axial location. Finally the incident angle on the target area defined the refracted angle A computation spread sheet was compiled from inputs data (method, mechanical, probe choice) provided the main outputs required to set the basis for the examination: probes, refracted angle ( mono-angle and bi angle wedges), optimized probe locations, coverage. Computations were then performed for all the variables enabling the drawing up of sketches to visualize the examination process and conditions. All evaluations were made with commonly used software: Excel, Solid works, Maple. Figure 5: Examination set up 3 2 Modeling principles The validated background data are the inputs for the modeling. CIVA (release 10) was used for this task. CIVA models are based on the pencil method, in which matrices take into account the propagation in each medium and at each interface from a source point. In load sample CAD file, all physical characteristics are introduced following a mesh (elastic constant, and limit conditions). UT sensors are defined and activated: Beam propagation pattern is calculated and 99

4 drawn (2D, 3D). With defect response module for flaws inserted in CAD, responses are calculated and displayed through views A, B, D, S-Scan, with tools to assess analysis. At each step of mesh the wave front is computed (with the physical characteristics and the sum of each contribution set the beam status. This method provides the physical image of the beam 4. Modeling works and results Figure 6: Pencil Method principle (by courtesy of Extende) Each configuration of probe arrangement and scan area was modelled in order to evaluate responses and performance for the arrangement as well as create a sample display to be used for training Drive Pin Hole Examination TOFD Modeling The DP scanning involves three areas; Above, Below and Bottom (DP rear face). The process of examination requires detection and sizing. The modelling is applied to disc 3 TS IT (turbine 250 MWe AEG-Siemens) ADP (above drive pin hole) TOFD The results from the background studies define a probe set up: 6 mm- 3.7 MHz, refracted angle 56 LW PCS 8 Hub radius 399, 7 mm. Expected indications are mainly generated from the top surface growing vertically and after circumferentially due to effect of the revolution. The cracking is due to fatigue effect, stress due to the pin and corrosion according the disc and steam conditions Ultrasonic setting, scan, beam behaviour Computations: LW, SW, mode conversion, Lateral wave, Backwall, shadow effect The display show: LW, Ind Top DP, response from side DP and SW response from Ind Top DP (conversion) Figure 7: Ray and Pencil tracing ADP with mode conversion and Modelling results Bscan Ascan for AD 100

5 4.1.2 BDP (below drive pin hole) TOFD The results from the background studies define a probe set up: 6mm Ø, 3.7 MHz, refracted angle 50, LW, PCS 20-25, Hub radius 399,7mm. Expected indications are mainly generated from the bottom surface initially growing vertically and then circumferentially due to effect of the rotation RDP (rear face drive pin hole) TOFD modelling The characteristic of crack indications in this area is that they initiate from rear face and grow axially. To pick up the tip of diffraction, the beam must be orientated square as possible to the wider face of the indication. 4.2 Bore Sides TOFD modeling The ultrasonic design for this examination requires locating probes on the hub and probes on the chamfer due to the geometry and coverage required Bore centre part TOFD modeling The principle due to the geometry is to pick up diffraction tip with skewed shots crossing the centre part called TOFD with oblique PCS. 4.3 BDP (below drive pin hole) PE modelling 2D computation, probe 3 mm 3.7 MHz, interaction defect beam Kirchhoff 1 Ind H3 mm x L 6mm above DP, 1 Ind H6mm x L6mm below DP. Bscan display: Response DP 2 above Ind Response DP2 below Ind Both are highlighted by corner effect and diffraction tip 4.5 Bore examination PE Modeling Figure 8 101

6 4 5 1 Bore examination-hub side for Detection Beam computation To insure the coverage from the ultrasonic design results two sets of probes are necessary (P29L/R refraction 29, P52L/R refraction 30 ). One set on hub and one set on chamfer. +21dB/max pressure Figure 4: disc 3 TS IT pencil tracing beam computation Results : Pencil tracing (aperture 6 ) to assess the coverage Beam pattern on area to scan : Coverage required: 90mm, probe on 36 mm, probe on 80 mm, overlap 20mm Gain enhancement with reference to point of max pressure 21 db Beam behaviour - existing side lobe for P29 probe Defect response Conditions of defect response: 3D model, probe 3 mm 3.7 MHz Design of wedges: mono angle and bi angle for chamfer adaptation. 3D computation, direct beam SW, Kirchhoff interaction, calibration Ø3mm SDH axial location: 30 mm /80 mm Indication 3mmx9 mm separate 2 and 4 5. LP turbine examination results In order to assist with the analysis of the data obtained from the inspection complex scans were modelled to provide readily available comparison of what should be expected in the inspection. A reference library was compiled to assist in the analysis of the inspection data. 5.1 TOFD ADP and BDP (comparison of modeling and actual) Modelling ADP Computations: LW,SW, mode conversion, Lateral wave, Backwall, shadow effect Results: Lateral wave: 16.5µs Ind: h3mmxl6mm17.48µs Top DP (385 mm) 18.14µs 102

7 Actual Scan ADP Probe delay 5.1µs, LW 16.7µs Ind: 17.4µsT,op DP 18µs Figure 10 Comparison of modelling and actual results Remark: The comparisons modelling/ actual scans show similar patterns and particularly describe the effect of geometry and interaction between defect and beam correctly. The deviation in time comes from a difference of dimensions between model and actual specimen. 6. Conclusions Non-destructive examinations are key items in the maintenance, management and design maintenance. Within the targets of maintenance the non-destructive examinations must provide results and information that is reliable but that also makes sense for engineering maintenance. The process described above is motivated by the following principles: Clear understanding of requirements dapted and explained system responses The tools of CAD software and physical modelling are efficient solutions to analyse and set a process, including the hardware. Particularly for turbine examinations, the size of component, the complex geometry, the software are a powerful way to design an examination contributing to: Process design, Expectations, explanation and validation of results. process and set up of examination package At the data analysis stage, the comparison of practical examination data and modelled results, allow assessment of assumptions and guide any corrective actions, with validated information. For specific practical results the software offers post-processing, enabling the correct analysis. For this application of examination of shrunk-on disc LP turbines in service, the tools of CAD software and physical modelling have fulfilled the objectives of the non-destructive examinations. They provided the results that could be adapted to the requirements of engineering maintenance, i.e. clear results (exam report), tools that enabled clear explanations of results from a complex geometry and enabled post-processing for any forecasting. 103

8 7. References CIVA: An expertise platform for simulation and processing NDT data: P Carmon, S Mahout, S Carillon, R Railing Pencil method for ultrasonic beam computation: N Gingerer CEA DRT LIST A semi- analytic FEM hybrid model for simulating configurations involving complicated interactions of waves with defects: N Gengembre, Alain Lhemery, Ryuji Omote, Thierry Fouquet Andreas Schuman CEA DRT LIST, EDF /R&D/Sinetics Method for the ultrasonic testing on disk shrink fitted onto shafts in vicinity of shrink fitted seats and apparatus carrying out the method: Hans Achim Jestrich (Ratingen), Ewald Werntgen (Mulheim) KWU Aktien gesellshaft Advanced Ultrasonic 2D Phased Array Probes: Frederic Reverdy(CEA list), G Ithurralde (EADS), Nicolas Dominguez (EADS) Ultrasonic Phased Array Inspection Turbine Components: Waheed Abbasi, Michael Fair, Siemens Power Generation Pittsburg USA Inspection of shrunk-on steam turbine disks using advanced ultrasonic: Paul A Zayeick (Gastonia NC) Greg P Selby (Charlotte NC), Mark Dennis (Charlotte NC) 2002 Flehr Hohback Albritton & Herbert LPP four embarcadero center suite 3400 San Francisco CA 104