Production aluminum catamaran · weld quality
A practical briefing for an owner’s team that wants to inspect aluminum catamaran welds independently of the yard. Costs are 2025-order-of-magnitude USD ranges (new equipment, North America / Western Europe). Training assumes ISO 9712 or ASNT SNT-TC-1A Level II for people who will actually accept or reject welds. Always confirm against the class society, flag, and contract specification.
| Method | Equipment cost | Training (days / dollars) | What it finds | Top sources |
|---|---|---|---|---|
| 1. VT | $150–$8,000 basic; $3,000–$25,000 with video borescope / measurement |
2–5 days course $400–$2,000 + documented experience (often 1–3 months) |
Surface only: profile, undercut, porosity, cracks open to surface, misalignment, distortion, incomplete fusion at the face, weld size | AWS, ASNT, class rules; G.A.L. Gage; Evident/Olympus IPLEX; Fluke; Waygate Mentor Visual iQ |
| 2. Leak testing | Soap / air gauges: $200–$2,000 Vacuum boxes: $800–$4,000 Pressure decay / helium: $8,000–$40,000+ |
2–5 days $600–$2,500 experience relatively short |
Through-wall paths only: pinholes, missed welds, cracked fillets that leak. Not a map of internal weld quality | Inficon, Pfeiffer, Agilent; vacuum-box makers; IACS / class tightness tests |
| 3. UT (conventional) | Flaw detector kit: $4,000–$18,000 Calibration blocks, probes, encoder extras $1,000–$6,000 |
10–20 days classroom typical for Level II $1,800–$5,500 + 3–9 months experience |
Internal planar and volumetric: lack of fusion, incomplete penetration, cracks, some porosity; also thickness. Weak on very thin plate and unskilled setup | Evident Epoch, Sonatest, Waygate/Krautkramer, Zetec; IIW blocks; ISO 17640 / AWS D1.2 |
| 4. PAUT | Instrument + probes: $25,000–$90,000 Encoded scanners / software $5,000–$40,000 |
UT first, then 5–15 extra days $2,500–$8,000 experience on top of UT |
Same family as UT, with images, better coverage, better recording. Often combined with TOFD. Strong RT replacement candidate | Evident OmniScan, Eddyfi/M2M, Zetec, Sonatest Veo; ISO 13588; ASME V / class PAUT procedures |
| 5. PT (dye penetrant) | Visible kits: $80–$600 Fluorescent + UV-A lamp: $400–$3,000 Production spray inventory ongoing |
2–4 days $400–$1,800 + 1–3 months experience |
Surface-breaking only: fine cracks, open porosity, toe cracks, crater cracks. Excellent sensitivity if cleaning is honest | Magnaflux, Sherwin, Met-L-Chek, BabbCo; ISO 3452; ASTM E1417 |
| 6. ET (eddy current) | $4,000–$25,000 instrument Weld / array probes $500–$8,000 |
8–15 days Level II $1,800–$5,000 steeper theory than PT |
Surface and near-surface cracks in conductive metal; coating/paint remaining in some setups; conductivity / heat-damage clues. Not deep volumetric | Eddyfi, Evident, Zetec, Rohmann, Fischer; ISO 17643; weld-scan probes |
| 7. RT (X-ray / gamma) | X-ray tube: $25,000–$150,000+ Gamma projector + source: $15,000–$60,000 plus source swaps Digital/CR: $15,000–$120,000 Licensing, barriers, dosimetry extra |
10–20 days NDT + radiation safety $2,500–$8,000 plus regulator licensing — this is the expensive part |
Volumetric: porosity, hollow root, many incomplete penetrations, some inclusions. Planar cracks can be missed if unfavorably oriented. Permanent image | Waygate, Yxlon, Varex, Carestream, Fujifilm, QSA Global; ISO 17636; IAEA / national radiation rules |
Ranges assume buying capable industrial kit, not the cheapest Amazon gadget and not a full automated gantry. Used equipment can cut capital 30–60%. Certification fees, travel, recertification (typically 5 years), and a Level III to write procedures are not in the course price.
Surface Always first. Most rejected aluminum welds fail VT: undercut, suck-back, wrong throat, porosity on the face, stop/start craters, distortion of thin plate.
Through-wall Proves watertightness of hull, tanks, and voids. Complements weld NDT; it does not replace it.
Internal Workhorse for butts and accessible T-joints if the team is skilled and the procedure is written for aluminum.
Internal + record Best owner-side volumetric tool for a modern aluminum yard if you will not run radiography.
Surface Cheap, sensitive, messy. Excellent on aluminum if surfaces are clean and not too rough.
Near-surface Strong on aluminum because the metal is conductive. Useful in service more than during dirty production welding.
Radiation Classic volumetric picture. Safety, access, and licensing dominate the decision — not image quality.
What “well enough” means:
Practical path to drop RT:
You still cannot inspect from the office. PAUT interpretation on fillets and nodes is a skill. Without a competent Level II and a procedure, skipping RT is just skipping inspection.
A serious first-year budget including training, blocks, and a Level III procedure package is often $60k–$150k, not including salaries.
Define the job carefully. A humanoid that walks the yard, preps the surface, deploys couplant, scans a changing 3D weld, interprets against a procedure, and signs a class-accepted report is not a near-term product.
Encoded PAUT crawlers, cobot arms with UT/ET probes, tank crawlers, drones for VT, and AI assistance on radiographs or PAUT plots. These work best on repeatable geometry: pipes, tanks, plate butts, wind towers.
Changing access, overhead and vertical fillets, confined voids, grinding dust, temporary staging, tack welds, heat distortion, and the fact that codes still want a certified person responsible. Aluminum T-joints produce ugly geometry echoes.
VT assist / AI flagging: now–3 years, human still accepts.
Specialized robots on selected long butts: 2–6 years in advanced yards.
General humanoid doing certified weld NDT unsupervised: unlikely before the 2035–2045 window, and liability/class acceptance may lag the hardware. Keep humans in the loop for production-critical acceptance.
Both, but they are different inspections.
This is process control. You are looking for welding defects created in the yard: lack of fusion, porosity, undercut, missed welds, wrong size, distortion, leaks. Once the boat is faired, painted, and in the water, many of those defects are expensive or impossible to re-inspect fully. Do VT/PT/PAUT/leak tests during assembly, not at the champagne launch.