Abstract

The results of a feasibility study for ultrasonic inspection of L-0 blade root inspection are presented. Two linear phased-array probes were used to detect the EDM notches in specific reference blocks and an actual spare blade root. The ray-tracing model was used to develop wedges, to optimize the detection, and to explain detection by mode-converted waves. Data analysis for three techniques was performed using TomoViewTM and 2-D / 3-D layout of blade/block. EDM notches as small as 2 x 0.5 mm could be detected up to 65 mm from the inlet / outlet faces. Larger EDM notches (6 x 2 mm, 9 x 3 mm and 12 x 4 mm) could be detected and sized up to a depth of 120 mm. The sizing accuracy for length is about 1 mm, and 0.5 mm for the height. Representative detection and sizing examples are also included.

Introduction

The supplier of Ontario Hydro's Darlington station recommended de-blading the lower pressure (LP) rotors after 50,000 hours of operation, to inspect L-0 blade root / steeple. This operation could lead to an increase of outage time by 8 - 10 days. The de-blading operation will not provide utility with a reliable maintenance tool. Darlington NGS (DNGS) is running 4 CANDU units of 960 MW, with 3 LP rotors per unit and 126 L-0 blades / rotor. During the regular outage, 2 LP rotors are opened for NDT / maintenance inspection. Due to the complexity of blade geometry, Specialized Inspection and Maintenance Department ( SIMD ) - Nuclear Technology Services Division and RDTech-Quebec City proposed to DNGS to perform a feasibility study using phased-array technology. Similar techniques are used by other ISI companies - utilities [ 1 -7 ] for complex parts inspection, when the probe is limited in movement and many focused angle beams are required to cover the part. Should this study prove to be successful, an industrial system will be developed and commissioned.

This paper presents the results of part 1 of feasibility study using linear-array probes and advanced data analysis software. The following topics will be discussed:

• UT techniques to cover the first 75 mm of blade volume (inlet/outlet)
• Reference blocks and targets
• UT ray tracing in blade root geometry
• Ultrasonic results
• Data analysis using TomoViewTM
• Summary of UT data ( detection / sizing capability, accuracy, S/N ratio )

RDTech-Quebec City and Ontario Hydro - SIMD-Pickering performed the first part of feasibility study. Companies from France and Canada contributed with specific tasks to this project. Some input regarding ray tracing, probe development and engineering assessment was received from companies / institutions from USA, Scotland, Switzerland, Sweden, Japan, Germany and Spain.

The first 75 mm from inlet-outlet face were inspected, using two linear probes and three techniques. The synthesis of these UT results will be presented to this workshop.

Experimental program

The first goal of this feasibility study was to evaluate the capability of detection and sizing of EDM notches located on specific reference blocks, using 3 (three) techniques (see Figure 1) .

Figure 1: Ultrasonic techniques used in feasibility study with linear array probe.

A variety of reference blocks, from simple to complex shapes, were manufactured and EDM notches were placed in the platform / hook areas. Table 1 listed the blocks used for probe characterization and detection / sizing assessment.

TABLE 1. Reference blocks used for probe characterization and detection / sizing.
Block ID	Thickness / reflectors	Scope
IOW modified	75 mm / SDH	Probe characterization
RDTech-S1	120 mm / SDH	"
NS-2	27 mm / FBH	"
NS-4	1-11 mm / FBH	"
NS-5	16 - 27 mm / FBH	"
Oblique block	120 - 250 mm / EDM notches	Detection / sizing
Block #1	130 mm / EDM notches	"
Block #2	130 mm + 80 mm wing / EDM notches	"
Block # 3	130 mm + 80 mm half wing / EDM notches	"
" Silver" blade L-0	130 mm / EDM notches	"
Spare blade L-0	130 mm	Comparison study / detection

An example of L-0 blade mock-up is presented in Figure 2.

Figure 1: Ultrasonic techniques used in feasibility study with linear array probe.	Figure 3: Ray tracing model simulating a detection of a linear defect 6 mm x 2 mm, located at 50 mm on hook # 1, from the inlet face, using technique # 1 and focused probe.

TABLE 2: The main features of phased-array probes.
Feature	Probe # 1	Probe #2
Center frequency	7 MHz	11 MHz
Range	10 - 110 mm	1 - 25 mm
Skew angle	± 60°	± 30°
Number of elements	32	20
S / N for technique 1	> 12 dB	> 16 dB
S / N for technique 2 / 3	> 30 dB	> 36 dB

The reverse engineering results of blade root and 6" of the wing were used to create the necessary 3-D drawing, .dxf mesh file, 2-D cross section for specific positions. The 3-D drawing was incorporated into a ray-tracing program - Imagine 3D, developed by Ontario Hydro Technologies[ 8 ]. The data from this model were used to optimize the wedge design, detection angles and to confirm TOF values from specific targets. Blade root 3-D drawing was used to develop the 3-D UTCAD software. An example of ray tracing is presented in Figure 3.

Two prototype linear array probes were designed and manufactured. The main features of these probes are presented in Table 2.

The following hardware was used: Tomoscan 3.5R8, Focus 32 / 64 piezo-composite phased array driver, MDU, Traker and Rover manipulators. RDTech-Quebec City manufactured the all above-mentioned hardware.

Ultrasonic results

The ultrasonic data were analyzed using TomoViewTM software. The 2-D layout of each scan / UT presentation (side-top-front, and A-scans associated with the cursors positions were displayed and measured). An example of detection, sizing and plotting is presented in Figure 4. An example of sectorial scan used for technique #2 to detect and size two EDM notches of 6 x 1 mm and 6 x 4 mm is presented in Figure 5.

Figure 4: Data analysis and 2-D plotting into reference block dimensions. A 6 mm x 2 mm EDM notch on hook #3 was sized as 6.2 mm x 1.8 mm, and located within 1.4 mm in all three coordinates of the specimen.	Figure 5: Detection and UT display in multi-scan of two EDM notches in block #1.

The first probe could size EDM notches within an accuracy of about ± 1 mm for the length and ± 0.5 mm for the height. The second probe could size EDM notches within an accuracy of about ± 0.6 mm for the length and ± 0.2 mm for the height. The summary of detection and sizing capability for technique #1 and #2 are presented in Table 3 and 4.

TABLE 3: Technique #1 summary of UT detection and sizing capability.
UT results	Probe # 1	Probe #2
Detection capability	2 x 0.5 mm up to 30 mm 3 x 1 mm up to 80 mm 6 x 2 mm up to 90 mm	3 x 1 mm from 5 mm up to 20 mm
Sizing capability	3 x 1 mm with (L = + 0.5 mm (h = - 0.5 mm	3 x 1 mm with (L = + 0.6 mm (h = + 0.2 mm
S / N ratio	> 14 dB, with tip signals > 6 dB	> 20 mm, with tip signals > 10 dB

TABLE 4: Technique #2 summary of UT detection and sizing capability.
UT results	Probe # 1	Probe #2
Detection capability	2 x 0.5 mm up to 65 mm - H2 3 x 1 mm up to 85 mm - H3 6 x 2 mm up to 110 mm - H4	2 x 1 mm from 5 mm up to 35 mm - H1 3 x 1 mm up to 35 mm - H1 4 x 2 mm up to 65 mm - H2
Sizing capability	2 x 1 mm with (L = +1 mm (h = + 0.5 mm 6 x 2 mm with (L = ( 0.3 mm (h = + 0.5 mm	3 x 1 mm with (L = + 0.8 mm (h = - 0.3 mm
S / N ratio	> 30 dB, with tip signals > 12 dB	> 40 dB, with tip signals > 16 dB

The UT data for height measurement using probe # 1 (technique #1 and #2) are presented in Figure 6. The length accuracy for the same set-up is presented in Figure 7 and Figure 8.

The UT results with probe #2 were focused on 3x1 mm EDM notch. They are presented in Figure 9, for the height, and in Figure 10 for the length.

Figure 6 : Height accuracy for probe #1 - technique 1 and 2 Figure 7 : Length accuracy for probe #1 - technique 1 and 2 - short EDM notches. Figure 8 : Length accuracy for probe #1 - technique 1 and 2 - long EDM notches. Figure 9 : Accuracy in height sizing for probe #2 - technique 2. Figure 10 : Length accuracy for probe #2 - 3 x 1 mm EDM notch - technique #2. Figure 11 : Sectorial scan used for sizing an EDM notch on platform.

The results for technique #3 using the first probe concluded that detection is feasible up to hook #3. Sizing was performed in sectorial scan and B-scan. The accuracy for length evaluation is based on encoder positioning (± 1 mm ).The height was measured based on tip diffracted signals, and the actual angle for tip and corner. Its accuracy is about -0.5 mm. An example is illustrated in Figure 11.

Conclusions

The first part of feasibility study to inspect the first 75 mm from inlet / outlet face, using linear phased- arrayed probes proved to be successful. EDM notches as small as 3 mm x 1 mm could be detected and sized within less than 1 mm accuracy, and with a good signal-to-noise ratio. The most efficient technique (#2) could be used only on limited zones. The high accuracy results were obtained due to the following factors:
• Ray tracing model, which provided the probe-defect interaction, wedge design and time of flight values
• High precision manipulator / encoder system
• Novel software - PASS - to design the phased array probes
• In-deep field characterization of linear array probes
• Novel software developed to plot the UT data into 2-D layout of the part.

Part 2 of feasibility study will investigate the detection and sizing for deeper range (75 - 200 mm), as well as optimization of technique #1 for the first 100 mm from the inlet / outlet face.

Acknowledgements: : the authors wish to thank Darlington Nuclear Power Plant - Power Train Engineering (Dave Bruce and Dale Craig), and Nuclear Technology Services Division for their support and granting the publication of this paper.

References:

1. Goto,M., et.al.: " Automated phased array UT system for turbine blade inspection " , in 10th Int. Conf. On NDE in the Nuclear and P.V. Industries , Glasgow, 1990, ASM Int., pp.583-590.
2. Brekow, G., et.al.:" Ultrasonic inspection at complicated geometries of pressure vessels with phased array probes ", ibid, pp.423-429
3. Wustenberg,H., et.al.: " Ultrasonic phased arrays for non-destructive inspection of forgings ", in Mat. Eval., June 1993, pp.669-672.
4. Nottingham,L.D., Solomon, K.R., Presson, H.J.:" Phased array ultrasonic approach to turbine blade attachment inspection", in 12th Int. Conf. On NDE in the Nuclear and P.V. Industries , Philadelphia , 1992, ASM Int., pp.493-497.
5. Schenk,G., Erhardt,A., Hauser,T.:" Development of two advanced ultrasonic inspection systems with phased arrays", in ASNT Fall Conf., Seatlle 1996, pp.218-220.
6. Desruelles,D., Burat,O., Pierre, G.: " Inspection of tube using ultrasonic phased array", in 14th WCNDT,New Delhi, India, 1996, pp.2031-2034 (vol.4).
7. Goto,M., Matsunaga,Y.:" Application of portable phased array UT system ", ibid, pp.2095-2100.
8. Mair, H.D., Ciorau,P.:" Ray-tracing 3-D model: a useful tool for feasibility study and ISI solving problems" To be published in Mat. Eval. 1997.

Authors:

A. Lamarre - alamarre@rd-tech.com
RDTech - Quebec City - Canada
N. Dubé - RDTech - Quebec City - Canada
P. Ciorau - Ontario Hydro - Pickering - Canada
B. Bevins - Ontario Hydro - Pickering - Canada

For more information see backup issue: NDTnet 12/97: NDT in Power Generation.