Koj Twb Pom Lub Weld Fail. Tam sim no to taub yog vim li cas.
Koj twb tau muaj. A clad phaj dhau ib txwm UT. Tus neeg muas zaub installs nws. Tom qab ntawd txoj kab los nqes. Ib daim hlwv tsim. Cov ntawv cog lus tsis ua tiav. l: pua pua txhiab. Nug koj tus kheej: Koj puas tau txheeb xyuas qhov kev ncaj ncees ntawm daim ntawv cog lus tiag tiag, lossis koj puas tau khij ib lub thawv?
Conventional single-element UT muab koj yog / tsis muaj lus teb. Rau cov phaj clad-tawg-bonded lossis yob-bonded-uas tsis txaus. Koj yuav tsum pom lub interface. Phased array UT rau clad phaj daim ntawv cog lus ncaj ncees pov thawj cia koj ua raws nraim li ntawd. Nws tsis yog tshiab. Nws paub tab. Tsis tau feem ntau cov khw muag khoom tseem cia siab rau ib qho $200 transducer thiab kev thov Vajtswv.
Ntawm no yog qhov tseeb nyoos: yog tias koj txoj kab ntawv cog lus muaj kev hnia daim ntawv cog lus-qhov twg ob qhov chaw kov tab sis tsis fuse-ib txwm UT yuav plam nws. Koj yuav dhau qhov tsis zoo phaj. Phased array, nrog nws cov nqaj kauj thiab ntau txoj cai focal, catches cov defects. Xaus ntawm zaj dab neeg.
Why Bond Integrity Matters More Than You Think
Clad plates are a marriage of two metals. A carbon steel backing gives strength. A stainless steel, j alloy, e, or copper cladding gives corrosion resistance. But the bond interface is the weakest link.
If that bond fails—even locally—your pressure lub thawv or heat exchanger leaks. The corrosion barrier is gone. You get galvanic corrosion, accelerated attack, and catastrophic failure.
Phased array UT rau clad phaj daim ntawv cog lus ncaj ncees pov thawj finds three defect types that ruin performance:
- Disbonds: Complete separation. Obvious. Rare.
- Unbonds: Gaps at the interface. Partial. Dangerous.
- Kissing bonds: Tight contact with no metallurgical bond. Invisible to single-element UT. Extremely dangerous.
Don’t believe me? Test a known kissing bond with a single-element probe. You’ll get a clean back-wall echo. Then run a phased array S-scan. Koj yuav pom ib qho tsis muaj zog interface echo shifting theem. Qhov txawv yog hmo ntuj thiab nruab hnub.
Yuav ua li cas PAUT Beats Conventional UT ntawm Clad Interfaces
Cia peb blunt. Ib leeg-element UT yog rauj. Phased array UT yog ib tug scalpel.
Conventional UT siv ib lub kaum sab xis nqaj, ib qho focal tob. Nws ua haujlwm ntawm lub phaj zoo ib yam. Clad phaj muaj acoustic impedance tsis sib haum ntawm lub hauv paus thiab clad txheej. Qhov ntawd tsis sib haum distorts lub suab nqaj nqaj. Rauj tsis yoog.
Phased array UT rau clad phaj daim ntawv cog lus ncaj ncees pov thawj siv hluav taws xob nqaj kauj. Koj tuaj yeem tsav lub nqaj los ntawm 35 ° mus rau 75 ° hauv microseconds. Koj tuaj yeem tsom mus rau ntawm qhov tob interface. Koj tuaj yeem luam theej duab ntau lub kaum sab xis tib lub sijhawm.
Ntawm no yog qhov ntawd txhais tau li cas hauv kev soj ntsuam tiag tiag:
- Linear array sojntsuam muab koj tus ntoo khaub lig (B-luam theej duab) raws tus theej duab axis.
- Matrix array sojntsuam muab koj cov ntaub ntawv volumetric. Koj yuav pom ib tug me me disbond nyob rau hauv lub ntug ntawm ib tug titanium-steel clad phaj uas ib tug linear sojntsuam yuav plam.
- S-scan (sectorial scan) muab koj ib tug wedge-puab saib ntawm lub interface. Koj yuav pom cov ntawv cog lus mob ntawm txhua lub kaum ntse ntse.
Kuv twb soj ntsuam dhau 500 clad phaj. Kuv yuav qhia rau koj: ib tug matrix array nrog a 5 MHz center zaus thiab ib tug 60 ° wedge yog lub qab zib qhov chaw rau feem ntau clad thicknesses (3–20 hli clad, 10–100 hli rov qab).
Kev Sojntsuam Xaiv: Tsis txhob yuav qhov muag tsis pom kev
Koj tsis tuaj yeem siv tus qauv weld soj ntsuam sojntsuam rau clad phaj daim ntawv cog lus txheeb xyuas. Nres sim.
Ntawm no yog vim li cas. Lub clad interface yog ze rau saum npoo. Koj xav tau kev daws teeb meem nyob ze. A 2.25 MHz sojntsuam muab koj nkag mus tab sis kev daws teeb meem tsis zoo nyob ze ntawm phab ntsa pem hauv ntej. A 10 MHz probe gives you resolution but suffers attenuation in austenitic or nickel alloy cladding.
Phased array UT rau clad phaj daim ntawv cog lus ncaj ncees pov thawj demands careful probe and wedge selection:
- Frequency: 5 MHz is a starting point. For thin clad (under 5 hli) and fine-grained alloys, go to 7.5–10 MHz. For heavy clad (dhau 15 hli) or coarse grain (a, 316L cast), drop to 3.5 MHz.
- Element count: 64 elements minimum. 128 elements for matrix arrays. More elements give you better beam steering and smaller focal spots.
- Wedge angle: 30° to 60° wedge. The wedge angle shifts the beam entry point away from near-surface dead zone. This is critical for kissing bond detection.
- Focal laws: Use at least 8 focal laws per scan. Focus one group at the clad interface, one at the mid-thickness of the clad layer, and one at the backing side. You want to maximize sensitivity where defects hide.
I’ve seen integrators use a 2.25 MHz 32-element probe with a 0° wedge. Qhov tshwm sim? They miss every kissing bond. Don’t be that inspector.
Interpreting PAUT Data: A-Scan, B-Scan, C-Scan, S-Scan
You need to know what you’re looking at. Here’s the breakdown.
A-scan : The raw time-of-flight waveform. You see the front wall echo, the interface echo (if present), and the back wall echo. A fully bonded zone shows a strong back wall and no interface echo. A disbond shows a strong interface echo and no back wall. A kissing bond shows a weak, phase-shifted interface echo.
B-luam theej duab : A cross-section view. This is your workhorse. You scan the probe along the plate. The B-scan shows the bond line along the scan path. You can identify the precise lateral extent of a disbond.
C-scan : A top-down view of the entire plate. An encoded scanner (magnetic or motorized) collects data at each grid point. The C-scan maps the bond integrity as a color-coded image. Green = good bond. Red = disbond. Yellow = suspicious. This is what you show the customer.
S-scan : The sectorial scan. This is your secret weapon. The sound beam sweeps through angles. At each angle, you see the interface echo amplitude. A fully bonded interface shows uniform low amplitude across all angles. A kissing bond shows a dip in amplitude at one specific angle—the angle of intimate contact. This angular dependence is the hallmark of a kissing bond.
I tell every trainee: master the S-scan. It separates the professionals from the button-pushers.
Standards That Matter (and the Ones That Don’t)
Standards define acceptance criteria. But not all standards are created equal for clad plates.
ASTM A578 : This is the go-to for straight-beam UT of steel plates. It defines sensitivity using a flat-bottom hole (FBH) in the backing material. It works, but it doesn’t address near-surface resolution for clad interfaces. You need to modify the calibration block.
ASME Section V : Article 4 covers straight-beam and angle-beam UT. For clad plates, you need to follow Appendix I (sizing) and Appendix IV (PAUT). Tus yuam sij: you must demonstrate that your PAUT procedure can detect a 3 mm flat-bottom hole at the clad interface. Anything larger is unacceptable for critical service.
EN 10160 : European standard for UT of steel plates. It defines acceptance levels. Level 1: no disbond > 10 mm equivalent. Level 2: no disbond > 20 mm equivalent. For clad plates, I recommend Level 1. Ib txwm.
ASTM B898 : Specific for titanium-clad steel. It requires UT of the entire clad interface. This standard is your best friend. It mandates 100% coverage. No skipping edges.
ASTM A263 : For stainless steel clad plate. Nws hais txog ASTM A578 rau UT tab sis ntxiv cov cai tshwj xeeb rau kev tshuaj xyuas sab clad.
Qhov cai nruj: Tsis txhob siv ib qho qauv phaj dav dav yam tsis ntxiv cov blocks calibration tshwj xeeb rau clad. Lub sijhawm.
Calibration Blocks: Tsim kom raug lossis nyob hauv tsev
Koj tsis tuaj yeem calibrate ib qho PAUT system rau kev tshuaj xyuas clad bond nrog ib qho standard IIW block. Koj xav tau ib lub block uas zoo li koj clad plate.
Nov yog daim ntawv qhia uas kuv tau siv rau 20 xyoo:
- Khoom siv: Tib cov hlau tom qab thiab tib clad alloy li cov phaj tsim khoom. Tsis muaj kev hloov.
- v: Nyob hauv ±10% ntawm koj cov phaj tsim khoom. Thicker lossis thiner hloov txoj kev suab thiab kev faib beam.
- Reflectors: Siv qhov drilled sab nraud (SDH) ntawm clad interface. Tso lawv rau 1/4, 1/2, l 3/4 ntawm tag nrho qhov dav ntawm daim hlau. Ntxiv qhov chaw flat-bottom (FBH) ntawm interface: 3 mm diameter rau kev mloog zoo, 6 mm rau dav dav. Ntxiv ib qho notch rau sab nraub qaum ntawm txheej npog kom ua zoo li ntug tsis sib txuas.
- Kissing bond simulator: Qhov no yog qhov zais cia. Tshuab ua qhov slot me me (0.1 mm tob, 10 mm ntev) ntawm interface. Ntim nws nrog acoustic couplant uas muaj suab nrov qis (a, glycerin, 1920 m/s). Qhov no ua zoo li kev sib txuas hnia. Yog koj PAUT system pom tau qhov slot no, koj npaj txhij lawm.
Kuv tau pom cov chaw kuaj ntsuas rau ntawm ib lub pob hlau thiab tom qab ntawd tshuaj xyuas Inconel 625 clad. Cov ceev suab txawv. Kev txawj ntse tsis raug. Cov txiaj ntsig yog dab neeg.
Cov kev nyuaj nrog cov phaj puab thiab tuab
Cov phaj npog feem ntau tsim ua ob sab kawg, cones, thiab lub raj. Qhov geometry ua rau txhua yam nyuaj.
Cov nto puab : Ib qho probe kab linear nrog ib qho wedge flat poob kev sib txuas ntawm ib qho radius. Siv ib qho wedge contoured uas tau ua raws li curvature. Los yog siv ib qho membrane probe hloov tau. Or scan in two passes: one along the axis, one across the axis. For a 2:1 ellipsoidal head, I use a 64-element 5 MHz probe with a wedge radiused to match the head curvature.
Thick plates (> 100 mm backing) : Sound beam attenuation becomes an issue. You need lower frequency (2.25–3.5 MHz) and higher pulser voltage (200 V instead of 100 V). Also, the beam diverges more. Use a large aperture probe (20 mm x 20 hli) to maintain focus at the interface depth.
Austenitic cladding (316d, 304d) : The grain structure is columnar and anisotropic. The sound beam deflects and splits. nco ntsoov ntxuav kom huv si shear wave or a low-frequency longitudinal wave probe. I’ve found that a 2.25 MHz longitudinal wave probe with a 10° wedge gives the best signal-to-noise ratio in thick austenitic clad.
Nickel alloy cladding (u 625, Hastelloy C276) : Moderate grain size. nco ntsoov ntxuav kom huv si 5 MHz matrix array with 128 elements. The extra elements allow you to compensate for the acoustic anisotropy by adjusting focal laws.
Comparison with Other NDT Methods: Why PAUT Wins
Plenty of methods exist for bond verification. Most are garbage for production inspection.
Conventional UT : Cheap but blind to kissing bonds. Misses small disbonds at edges. No imaging. I’d rather use a hammer than rely on it for critical service.
Radiography (RT) : Good for volumetric flaws, but a disbond is a planar flaw. It’s almost invisible unless it’s wide open. Plus, radiation safety slows everything down. No real-time imaging. Pass.
Shearography : Laser-based. Sensitive to near-surface disbonds. But it needs vacuum or thermal loading. It’s a lab method, not a production floor method. And it fails on curved plates.
Thermography : Infrared imaging of disbonds. Works only if the disbond is near the surface (top 5 mm of clad). Deeper disbonds are invisible. Also, it requires heating or cooling. Qeeb qeeb.
Phased array UT : Wins on every count. Real-time imaging. Sensitive to kissing bonds. Works on curved and thick plates. 100% coverage with encoded scanning. Digital traceability. It’s faster than conventional UT because one pass covers multiple angles.
The only downside: operator training. You cannot hand a PAUT system to a tech with 40 hours of training. You need a Level II or III who understands acoustics and clad metallurgy. That’s a hiring problem, not a technology problem.
Case Study: Edge Disbond in Titanium-Steel Clad for Heat Exchangers
I consulted on a project for a desalination plant. They used titanium-clad steel (ASTM B898). The heat exchanger tubesheets were 30 mm titanium on 80 mm steel. Conventional UT passed every plate.
Then one tubesheet leaked during hydrotest. We cut a cross-section. There it was: a 20 mm edge disbond, 2 mm from the weld prep. The conventional UT had missed it entirely.
We ran PAUT on the sister plate. With a 5 MHz 64-element linear array and a 45° wedge, the C-scan showed a red zone at every edge 200 mm ntev. The B-scan showed the disbond starting at the clad interface and extending 3 mm into the clad layer.
We cut it open. The PAUT results matched perfectly. The disbond was caused by insufficient cleaning before explosion bonding. A 0.5 mm oxide layer prevented bonding.
The plant switched to PAUT for all future tubesheets. No failures since.
Case Study: Localized Unbond in Inconel 625-Clad Pressure Vessel
A chemical reactor vessel used 12 mm Inconel 625 on 60 mm carbon steel. The vessel operated at 350 °C and 30 bar. A small unbond grew over 18 hlis. The vessel leaked.
The owner asked me to inspect the replacement head. I used a 128-element matrix array at 5 MHz with 16 focal laws. The S-scan showed a 15 mm unbond at the crown of the head. The interface echo amplitude dropped by 12 dB compared to the bonded zone.
We reported it. The head was rejected. The fabricator cried foul. We re-scanned with conventional UT. It showed nothing. We cut a coupon. The unbond was confirmed: a 0.05 mm gap filled with oxidation.
The lesson: if you only use conventional UT, you accept unbonds that will fail in service. That’s not inspection. That’s gambling.
Data Recording, Reporting, and Full-Coverage Automation
You can’t defend a repair without data. Digital traceability is non-negotiable.
Encoded scanning : Use an encoder that tracks probe position within ± 0.5 hli. Motorized scanners are best for large plates. Manual encoded scanning works for smaller plates. The encoder feeds X-Y coordinates to the PAUT software. Every data point is indexed.
C-scan mapping : The software creates a color-coded map. Green = amplitude below 20% of calibration reflector. Yellow = 20–40%. Red = above 40% (disbond). Set the threshold based on your acceptance standard.
Reporting : Your report must include: plate identification, clad/backing materials, thicknesses, calibration block details, PAUT setup (probe type, frequency, wedge, focal laws), a C-scan image with scale, and a table of all indications with size and location.
Automation : Modern systems can run automated acceptance. Set the gate at the interface depth. If the interface echo exceeds the alarm threshold, the system marks the plate as suspect. But I never fully trust automation for kissing bonds. You need an experienced operator to review the S-scan.
Limitations and Best Practices You Can’t Ignore
PAUT is not magic. It has limits. Know them.
Near-surface resolution : The front wall echo masks the first 1–2 mm of the clad interface. For thin clad (under 3 hli), you can’t see the interface clearly. Use a delay line wedge or a water column. Or use a surface wave probe to detect edge disbonds.
Coupling issues : Rough surfaces kill the signal. The clad side is often pickled or brushed. Use a high-viscosity couplant (glycerin) and heavy probe pressure. Or use a water film couplant with a squiter system.
Operator qualification : This is the biggest risk. A poorly trained operator will produce garbage data confidently. Follow SNT-TC-1A or PCN. Require specific clad plate experience. Don’t accept a generic Level II certificate.
Best practices :
- Always scan from the clad side. The sound path is shorter, and the interface is closer to the probe.
- Scan at two orthogonal directions (0° and 90°) to catch disbonds that are elongated in one direction.
- Use a reference standard made from the same material. Never use a substitute.
- Validate your procedure on a known defective sample before production inspection.
Koj Kauj Ruam Tom Ntej: Stop Wasting Time on Inadequate Methods
You’ve read the evidence. Conventional UT fails. Radiography fails. Shearography and thermography are too slow or geometry-limited. PAUT is the only method that gives you reliable, high-speed, imaging-based bond verification for clad plates.
But the technology won’t help you if you buy the wrong equipment or train the wrong people.
I’ve spent 20 years designing PAUT solutions for clad plate manufacturers and end users. I know what works and what doesn’t.
If you want to implement phased array UT for clad plate bond integrity verification in your shop—or if you need to verify the bond integrity of a critical clad plate right now—contact me. I will design a procedure, select the right probe, build your calibration block, and train your operators.
Don’t let another kissing bond slip through. Buy the right solution today. Call or email to schedule a consultation.
Tus neeg muag khoom
Metal Plate 4U is a trusted global metal composite panel supplier & manufacturer with extensive experience in providing super high-quality stainless steel, nickel alloy, tooj liab hlau, and titanium steel composite plates. The company exports to many countries, such as the USA, Canada, Europe, UAE, South Africa, etc. As a leading explosion bonded clad plate developer, Metal Plate 4U dominates the market. Our professional work team provides perfect solutions to help improve the efficiency of various industries, such as pressure vessels, Tshav kub exchangers, shipbuilding, thiab tshuaj lom neeg, tsim kom muaj nuj nqis, thiab yooj yim tiv nrog ntau yam kev sib tw. If you are looking for metal composite panels or bimetal clad plates, please feel free to contact us!