case study bringing a solution to market inspection of ... · presented by bruce stetler ... ect...
TRANSCRIPT
Case Study
Bringing a solution to market
Inspection of Twisted tubes in
Heat Exchangers
Presented by Bruce Stetler
NDTMA February 12, 2014
Topics discussed
• Background on Heat Exchangers, both Standard and Twisted tube design
• Engineering and modeling a prototype • Carrying out tests on the prototype • Results on calibra?on standards • Field trials • Conclusions
Ø However, there are limitations associated with these type of exchangers, which include inefficient usage of shell side pressure drop, dead or low flow zones around the baffles where fouling and corrosion can occur, and flow induced tube vibration, which can ultimately result in equipment failure.
STANDARD EXCHANGERS
Ø Over 85% of all heat exchangers in oil and gas, chemical, petro-chemical, and power generation are accommodated through the use of conventional shell and tube type heat exchangers.
TWISTED TUBE HEAT EXCHANGER AVANTAGES TO THE INDUSTRY
• Over the last 20 years, TWISTED TUBE heat exchangers have made their way into the Refinery and Chemical industries.
• TWISTED TUBE tubes geometrical design creates a swirl-flow or forced fluid motion along the curvature of the tube by inducing greater heat transfer.
• The tubes are interlocking and free of baffles or supports, which allow them to have a higher heat transfer coefficient than other tubular heat exchangers (up to 40% higher).
• Tubes come in a variety of sizes from .625” up to 1” inch in diameter with material types such as stainless, titanium, brass, monel, carbon steel, duplex, and nickel (to name a few).
Ini?al assessment
As we know Eddy Current is a electromagnetic application used in the industry for the inspection of non-ferrous material in such alloys as: Stainless Steel, Titanium, Brass, Hastelloy, Monel…
EDDY CURRENT TESTING
Eddy currents have the sensitivity to detect a wide range of corrosion and erosion (process or mechanical) such as:
• baffle fretting (vibration), impingement erosion, tool gouges, corrosion, pitting, stress corrosion cracking, MIC, etc.
EQUIPMENT USED FOR R&D
• ECTANE – ERNMI • MAGNIFI 3.2B2
www.eddyfi.com
Booth #24
Conventional ECT bobbin probe
For reference we used a standard bobbin probe 11mm (.433”) diameter (Eddyfi probe). Tube material was Titanium .750” x .049” nominal wall.
Bottom (Dip)
Top (Crest)
Tube Dimensions Tube OD Tube ID Tube WT Straight sec6on 19.05mm (.750") 16.28mm (.604") 1.38mm (.054")
Twisted Top (Crest) 22.23mm (.875") 19.46mm (.766") 1.38mm (.054") Twisted Bo>om (Dip) 15.43mm (.607") 12.66mm (.498") 1.38mm (.054")
Constant ID 12.71mm (0.500")
Bottom (Dip)
Top (Crest)
Two Titanium calibration standards
No.# 2 calibration standard / Top (crest) pit Standard and one thru-wall bottom (dip) and 2 in straight section.
No.# 1 calibration standard / bottom (dip) pit Standard, but also has several top (crest) defects.
# Machined defects Diameter Depth Loca6on 1 Thru wall .052" 100% Straight 2 OD Flat boYom hole .187" 40% Straight 3 Thru wall .052" 100% BoYom 4 2x Thru wall (at 180?) .052" 100% Top 5 OD Flat boYom hole .078" 80% Top 6 OD Flat boYom hole .109" 60% Top 7 OD Flat boYom hole .187" 40% Top 8 4x OD Flat boYom hole (at 90) .187" 20% Top 9 2x Thru wall (at 180) .078" 100% Top
# Machined defects Diameter Depth Loca6on 1 Axial EDM notch .40" 100% Top 2 Oblique EDM notch .40" 100% Top 3 Circ EDM notch .40" 100% Mid 4 Thru wall .052" 100% BoYom 5 OD Flat boYom hole .078" 80% BoYom 6 OD Flat boYom hole .109" 60% BoYom 7 OD Flat boYom hole .187" 40% BoYom 8 OD Flat boYom hole .187" 20% BoYom 9 4x OD Flat boYom hole (at 90?) .187" 20% Each 90 ? 10 Thru wall .078" 100% Top
Noise level
11mm (.433”) Standard ECT bobbin probe results bottom dip hole (on calibration standard #1)
Calibration done at 1V, 40˚
Calibration done at 1V, 22˚
Noise / lift-off
Calibration done at 1V, 22˚
Calibration done at 1V, 40˚
No Cir. detection
20% absorb in noise
100% bottom thru-wall hole
Bottom
Top
11mm (.433”) ECT bobbin probe results top crest hole (on calibra+on standard #1)
Noise / lift-off
Top (crest) 100% hole @ 74 deg. (rotates clockwise approx. 52 deg.)
Calibration done at 1V, 74˚
Bottom (dip) 1V, 22˚
Bottom
Top
Bottom (Dip)
Top (Crest)
Eddyfi ECT- twisted probe concept
Twisted bobbin coils (Diff. & Abs.): The differential signal on the twisted bobbin will be used to create sizing curves for analysis (Top/crest and bottom/dip defect responses).
2 absolute coils (180 deg. Apart): The absolute coils will be used to discriminate top (crest) defects from the bottom (dip) defects.
Slip ring assembly: Allows the probe tip to rotate through the twist.
Standard bobbin coils (Diff. & Abs.) To be added for inspection of straight sections
Diff. & Abs)
ABS -1
ABS-2
Eddyfi ECT-‐ twisted probe design movement
ECT Twisted bobbin bottom defects calibration #1
• To reduce lift-off and noise on both twisted bobbin and top coils we rotated each signal to get the noise level horizontally.
• #1 calibrate standard we used the bottom 100% thru-wall hole at 1volt, 19˚,
then we created sizing curves for bottom (dip) defects. • Next two slides will also show the top coils detection, to confirm top defects
location (see blue circles) • By using the <Slew> (aligning the strip chart data) process from Magnifi we
can now align the top coils signal with the twisted bobbin signal. This will allow us to distinguish between the location of the indication in the bottom (dip) or top (crest).
ECT Twisted results (on calibration standard #1 / bottom /dip)
Calibration done at 1V, 19˚
TOP 1 Derivate signal TOP 2 Derivate signal TWISTED DIF HP
Filter signal
TWISTED DIF
TOP 1 ABS TOP 2 ABS
TOP 1 Derivate signal TOP 2 Derivate signal TWISTED DIF HP
Filter signal
TWISTED DIF.
TOP 1 ABS TOP 2 ABS
ECT Twisted results (on calibration standard #1 / disguising top/crest defects)
100% TWH TOP/CREST @ 59 deg. 61%
ECT Twisted bobbin using top coils calibration standard #2
• Again to reduce lift-off and noise signal on both twisted bobbin and top coils we rotated each signal to get the noise level horizontally (no calibration required).
• Then for the top defects we will create new curve with calibration standard #2. • By using the <Slew> (aligning the strip chart data) process from Magnifi we can
now align the top coils signal with the twisted bobbin signal. This will allow us to distinguish between the location of the indication in the bottom (dip) or top (crest).
ECT Twisted results (on calibration standard #2 Top/Crest defects)
TOP 1 Derivate signal TOP 2 Derivate signal TWISTED DIF HP Filter signal
Hole located on top crest
TOP 1 TOP 2
Bundle simulation
• At this ?me only calibra?on tubes were scanned to determine the probe detec?on limita?on.
• Both calibra?on tubes were surrounded by similar ?tanium twisted tube samples to simulate field condi?ons.
• Top (crest) defects are s?ll highlighted by blue circle.
Bundle simulation on calibration standard #1 (bottom/dip), surrounded by similar twisted tube
TOP 1 Derivate signal TOP 2 Derivate signal TWISTED DIF HP
Filter signal
Calibration hole located on bottom appears at 97%
TWISTED DIF
TOP 1 TOP 2
Bottom (dip) #1 calibration curve of approximately 104 degree separation.
BOTTOM CURVES for ECT
Size Measure
100o % 19 deg
90o % 25 deg
80o % 37 deg
70o % 50 deg
60o % 62 deg
50o % 74 deg
40o % 86 deg
30o % 98 deg
20o % 111 deg
10o % 123 deg
0o % 135 deg
Bundle simulation on calibration standard #2 (top/crest), surrounded by similar twisted tube
TOP 1 Derivate signal TOP 2 Derivate signal TWISTED DIF HP
Filter signal
Hole located on top appears at 93%
TOP 1 TOP 2
Top (crest) #2 calibration curve of approximately 60 degree phase separation.
TOP CURVES -‐ Bundle for ECT
Size Measure
100o % 46 deg
90o % 52 deg
80o % 59 deg
70o % 66 deg
60o % 73 deg
50o % 79 deg
40o % 86 deg
30o % 93 deg
20o % 100 deg
10o % 106 deg
0o % 113 deg
ECT Titanium comparative results
• By comparing the standard bobbin probe we found the Eddyfi ECT – twisted probe concept had very good detection for bottom (dip) defects and top (crest) defects, even if surrounded by other twisted tubes.
• By using the twisted bobbin coils for analysis we can use the two top (crest)
coils to locate the defect position and then select the proper sizing curve for analysis.
• Note: these results can be obtained only by using Magnifi software with its specific process such as median filter, slew, and automatic landmark recognition.
• We expect data collection speeds to be at 12 inches per/sec.
CONFIDENTIAL
Probe head time life After used the twisted probe on site we were able to inspect between 1000 – 1500 tubes, originally we were expected to inspect at least 500. We pushed the probe in at 2ft/sec then with a pulling speed of 1ft/s (see picture below)
© 2010 Eddyfi. All rights reserved. 25
CONFIDENTIAL
Next probe design We believe that a closed design like below will be eventually better, as this design will push internal deposit outside of the tube where the first design cannot, moreover this design will be more robust and easier to assemble from a manufacturing perspective The first design has epoxy over the crest coils, and not by positioning both crest coils into both shoes, so instead of the epoxy in contact, it will be machined plastic shoes, we think that the improved design will increase longevity. (target 2000 tubes)
© 2010 Eddyfi. All rights reserved. 26
Engineering ?me
• Ini?al Assessment 4 hours • Applica?on Specialist & Modeling 20 hours • Prototype development 40 Hours • POD of Prototype 20 Hours • Manufacturing 5 hours • Tes?ng of finished product 30 hours • Total 119 Hours
Eddyfi RFT twisted tube design
Single driver coil (exciter coil)
Twisted bobbin coils (receiver)
2 Top (receiver absolute coils)
Slip ring assembly
*Preliminary tests have been performed on a carbon steel twisted tube (1.0” OD / .109” WT) with a prototype probe. Results are pending…
Ques?ons?