Implementing TI Billet Phased Array on the Shop Floor By Mark Pompe, West Penn Testing Group Jeff Umbach, Pratt & Whitney Experience at West Penn Testing
Implementing TI Billet Phased Array on the Shop Floor
By
Mark Pompe, West Penn Testing Group
Jeff Umbach, Pratt & Whitney
Experience at West Penn Testing
Overview• History of WPTG involvement with Ti Billet
Inspection• Comparison to Existing Multizone Inspection• Implementation Of Phased Array at WP• Lessons Learned• Q & A
MultiZone Ti Billet at WPSioux City Accident
MultiZone Ti Billet at West Penn• Operated GE MultiZone system 1995
through 2004• First Production System installed after Initial
RMI test bed• Over 22 million lbs of rotor grade Ti billet
were inspected• Testing has transitioned to billet producers
MultiZone Ti Billet at WP• 3 Separate
Systems– Motion– Acquisition– Analysis
MultiZone Ti Billet at West Penn
• Billet followers held 4 to 7 MultiZone transducers
WPTG Phased Array Involvement
• Operating Phased array systems since 2000 on wrought product and Composites
• Volunteered to be a test bed for Ti Billet Phased Array Project in 2008
Modifications for PA trial at WP
• Modified Follower to hold 5” diameter Phased Array Transducer
• Re-configured MZ prototype gimbal, commercial gimbals too fragile, test can be violent with bent billets.
• Added auxiliary encoders for X axis and rotational position – Passive System
Modifications for PA trial at WP• Follower/transducer configuration shown for start of billet scan• The lead rollers are on the billet and the back rollers, not visible in
the photographs, are not yet on the surface.
Phased Array transducer
Why Consider Phased Array?
Phased Array
• 5 min setup– One Transducer 6” to
10”• Evaluation 2 C-scan
images• No over scan for rack
of transducers
Multi Zone
• 2 to 4 Hour Setup– 3 sets of 4 to 6
transducers• Evaluate 4 to 6 C-
scan images• 12” to 24” overscan
Alignment For Multizone Probes is on Billet Back Surface Reflection
or FBHs• Internal Ti Billet structure steers the beam
off axis, normal to back wall is not normal to FBHs
• Can add hours to setup for 4 to 6 transducers
Alignment for Phased Array is performed on the billet front surface• 9 elements of the transducer used to align the
transducer on the surface of the sample• Alignment completed in ~5 min
time
Alignment is performed on the billet surface
Mis-alignment in the circumferential direction is noted because on pair of elements is closer to the surface than the other pair.
Multizone Calibration
• Each Transducer is calibrated on a FBH at the start and end of the zone• The two FBHs are equalized by Water Path Adjustment to within 3 dB.• Complicated by difficulty in normalizing
Phased Array Calibration• A DAC (TCG) is used in each zone with typical FBH standards• The setup for 6” diameter inspection is shown
Channel gain = 1 dB
Sum gain = 42 dB
Filter = 5 MHz band pass
Channel gain = 2 dB
Sum gain = 42 dB
Filter = 5 MHz band pass
• An acquisition file is made utilizing 7 sweeps to help guarantee that the backwall is sampled with the beam as normal as possible
Gate on the backwall signal
7 sweep angles with the beams focused at the backwall
Attenuation correction
49.2% is used for the attenuation comparison to the billet
More Accurate Attenuation Comparison
PA Data Analysis• Software automatically evaluates the C-scans for Amplitude and Signal
to Noise Ratio Defects• Similar to MZ
• The automated software applies the criteria for both zones and displays the rejectable and reportable signals
• The table is also written to an Excel file with scan information to be used for a report
• Excel file contains results from the automated analysis
PA Billet Inspection Report
Data File: D:\PA data\Ti Billet Inspection\West Penn\6-in dia\B3W3A rerun 4-13-10 Report.xls 40281Time: 16:11:23 4/13/2010 at West Penn
attn cor.Operator: J Umbach NZ DZ 0.4256Billet ID: B3W3A 4-13-10 channel channel Billet Length (inches): 34.8 gain=2 gain=1Axis Inspection Range(inches): 34.9857Billet start axis position for evaluation: 0.5 20 20.08512Billet end axis position for evaluation: 34.3 20 20.2128Calibration Amplitude (%FSH of #2 FBH calibration):80 22.8 15 23.14048 15.34048Calibration Gain(dB): 0.43 25.8 7.9 26.332 8.432Transducer SN: 6511 4.3 5.02352Amplitude reject level (%FSH of #2 FBH calibration):70 4 4.91504Signal to Noise Ratio reject value: 2.5 7.7 8.80656Reportable amplitude value(%FSH): 40 9.3 10.59808OR Reportable Amplitude/Averge Noise Rate: 3 10.8 12.2896Search Angle Width: 20 Angle Incr 15Seacrh Axial Width: 2 Axial Incr 1.75
Indications: 4 Inspection Results: Reject
# Channel Gate Axial positionAngle PositionTOF Amp(%) SNR Noise Status1 2 A 22.5857 161.6 5.55 84.31 9.73 11.11 Reject 84.50%2 2 A 22.6226 162.8 5.54 63.53 6.72 10.81 Reject3 2 A 23.0655 161.6 5.38 39.61 3.84 10.98 Reject4 2 A 26.2024 105.6 9.78 32.55 2.51 11.26 Reject 32.50%
peaked values recorded by the operator
Lessons Learned• P/A set up at WP on several occasions in 2010 on two systems, 6” and 10”
Billet• Encoder vs position• Follower Stability
– 1/16th to ±1/4 typical spec for Water Path– Best Possible for PA < 1/16”, you can see it in the data.
• Operator Training – prior c-scan experience – 2 operators were quick studies.
• Tomoview Software designed for engineers• Procedures can be worked out for operators• PW generating focal laws initially, need a source for this.• Single Probe, Alignment and DAC dramatically reduces setup• New Standards needed, can use simpler design
To Accomplish Implementation
• FAA approval (Beginning 4/2011)
• P&W specification
• RR, Honeywell Specifications
• AMS Specification for Generic Use
• Equivalency to AMS-2631
Cost Comparison MZ vs PA• Instrumentation – MZ built and maintained by one source with 1990s
components. P/A should have multiple sources (Olympus, Peak, others possible)
• Transducers – Phased Array is 1/3rd of cost for 6” to 10”, – One Phased Array Probe vs 3 sets of 4 to 6 MZ probes. – Spares are needed for either. – Both single source (Same source), more expense to develop another source.
• Integration – Possible through inspectionware with peak instrument, Separate motion control and data collection/Analysis with Olympus.
• Followers – readily modify off the shelf units (Tac Tic for ours) vs complex custom builds
• Phased Array requires new calibration standards with additional FBHs• Evaluation PC integrated with data collection pc, simplifies maintaining PCs,
PCs are off the shelf running modern operating systems.• Phased Array purchase price is lower, maintenance is significantly lower.
Q & A