The Verification Problem in Practice
Marking and paperwork are the first evidence of a pipe's product form, but they are not proof. Paint stencils fade, bundles are re-bundled, and mixed stock in a yard is common. When the specification excludes a seam, or when a buyer suspects that a welded length was supplied against a seamless order, the claim has to be settled by physical examination rather than by argument. This article sets out the methods that actually work, in order of evidential strength, together with the sample preparation, the applicable standards and the interpretation rules.
The methods fall into three groups: metallographic and macro-etch examination, which detect the bond line directly; non-destructive and physical techniques, which detect it indirectly; and documentary and dimensional checks, which raise or lower suspicion but do not decide the question.
Method 1: Macro-Etch and Metallographic Examination
This is the definitive method. It works because a weld, even a forged high-frequency weld with no filler metal, leaves a narrow bond line whose microstructure differs from the parent metal unless the whole length has been fully normalised. High-frequency welds are particularly narrow, so a rough grind followed by etching may fail to reveal the seam; proper preparation is essential.
Sampling: cut a ring or coupon from the end of the pipe, at least 40 mm long, taken from a location that includes any visible seam indication. Where the seam is not visible, cut a full ring so the entire circumference can be examined. Record the pipe identification, heat number and orientation before cutting.
Preparation follows the sequence in ASTM E3 for metallographic specimen preparation: sectioning with coolant to avoid overheating, mounting if the section is small, coarse grinding, fine grinding through progressively finer abrasives, then polishing with diamond or alumina suspension to a scratch-free finish. For macro examination only, grinding to a fine finish and etching is sufficient.
Etching: a 2 to 5 percent nitric acid in ethanol solution, the classical nital etchant, reveals ferrite grain boundaries, the bond line and the heat-affected zone in carbon and low-alloy steel within seconds to a minute. ASTM E407 lists accepted etchants and their application, and more selective reagents such as picral are used where carbide or grain boundary detail matters. Stainless grades need different reagents entirely, typically an electrolytic etch or aqua regia based mixture, because nital does not attack austenite effectively.
Interpretation: in the as-welded high-frequency condition the bond line appears as a narrow, often light-etching band with a decarburised or ferrite-rich edge, flanked by a heat-affected zone of finer or acicular structure. In a seamless section the structure is continuous and follows the deformation pattern from piercing and rolling, with no linear discontinuity. Widmanstatten ferrite, banding and non-metallic inclusion stringers may be present in either form and must not be mistaken for a weld; the discriminating feature is a continuous line running through the wall thickness at constant angular position around the circumference.
Etching methods also expose seam defects that a casual visual check misses: lack of fusion, oxide notches at the bond line, and penetrator or cold weld in high-frequency joints. A purchaser who suspects a substitution should ask for the etched macro section to be photographed at magnification alongside a scale, and the record retained with the test report.
Method 2: Hardness Traverse Across the Suspected Seam
Hardness is a rapid, semi-quantitative discriminator when a weld is present. A traverse is run across the suspected seam location at mid-wall, using a portable or bench hardness tester with a defined load, and the readings are plotted against distance.
In an un-normalised high-frequency weld the bond line frequently shows a hardness peak, sometimes 30 to 60 HV above the parent metal, with a dip or a plateau in the heat-affected zone on either side. In a normalised or fully annealed product the traverse is flat within normal scatter. Bench testing to ASTM E92 or ASTM E384 gives the most reliable resolution; portable Leeb or UCI instruments are acceptable for screening but must be correlated before use in a dispute.
Hardness also has a compliance role independent of identification: under ISO 15156 / NACE MR0175 a hardness ceiling applies to the weld and heat-affected zone in sour service, so the same traverse serves both purposes.
Method 3: Non-Destructive Seam Detection
Where cutting is not acceptable, the seam can often be located without damaging the pipe. Ultrasonic testing with a angle beam probe scanned circumferentially will return a distinct geometric reflection from a fusion line that is absent in a seamless wall; the technique follows the principles of ASTM E213 for tubular ultrasonic examination and of ISO 10893-11 for automated weld seam testing. Eddy current testing with a differential probe is even more sensitive to the conductivity and permeability change at a bond line on the outside surface, and it is the standard in-line method used by mills to verify seam tracking.
Magnetic particle testing under ASTM E709 is useful on magnetic steels where the seam is at or near the surface and where local permeability changes are strong enough to produce an indication; it is less reliable on a fully normalised fine-grained weld. Radiography under ASTM E94 is not a good identification tool, because a forged high-frequency weld has no density difference to image; it works only where a filler-metal weld with a distinct root profile exists, as in submerged arc or electric fusion welded pipe.
The practical limitation of all non-destructive methods for this purpose is that they detect an anomaly, not a product form. A positive indication plus a metallographic confirmation is a defensible conclusion; an ultrasonic indication alone is a reason to cut a sample, not a finding.
Method 4: Dimensional and Geometric Evidence
Several geometric features correlate strongly with product form and are worth checking first because they cost nothing.
A longitudinal weld is usually accompanied by a slight wall thickness step, a weld bead crown on the outside in as-welded condition, and a faint straight line along the length. Internal flash, unless scarfed, produces a raised land on the bore that can be felt with a finger or seen with a borescope. A helical weld produces a continuous spiral line at a consistent helix angle, which no seamless process can imitate.
Wall thickness distribution is also informative. Seamless pipe made by piercing tends to show a sinusoidal eccentricity pattern around the circumference, with the thick and thin sides roughly 180 degrees apart, reflecting the piercer and mill axis. Welded pipe generally shows a more uniform wall with a local anomaly at the seam, and its ovality is a forming artefact rather than an eccentricity artefact. A four-point or eight-point ultrasonic wall survey makes this pattern visible.
Straightness, end squareness and the presence of a heat tint or oxide band along a line are supporting evidence only. None of these observations proves product form on its own.
Comparison of Identification Methods
| Method | Evidence strength | Destructive | Typical turnaround | Best applied to |
|---|---|---|---|---|
| Metallographic examination after nital etch | Definitive | Yes, coupon required | Hours to one day | Disputed deliveries, specification exclusions |
| Macro etch of full-thickness section | High | Yes | Under one hour | Rapid screening at mill or third-party lab |
| Hardness traverse across seam | Supporting to high | Semi, small indents | Minutes | Field screening plus sour service compliance |
| Ultrasonic angle beam scan | Supporting | No | Minutes per length | Locating the seam before sampling |
| Eddy current differential probe | Supporting | No | Continuous | High-speed surface seam detection |
| Magnetic particle testing | Low to supporting | No | Minutes | Surface-breaking bond line indications |
| Visual and borescope inspection | Low | No | Seconds | Flash, bead and spiral seam evidence |
| Wall thickness mapping | Low to supporting | No | Minutes | Eccentricity pattern versus seam anomaly |
| Certificate and marking review | Documentary only | No | Immediate | Raising suspicion, never concluding |
Sample Selection Rules That Avoid Wrong Conclusions
Most failed identifications are sampling failures rather than technique failures. Three rules prevent them.
First, sample the pipe, not the end of the bundle. A single coupon from an accessible pipe end may come from a different length than the one under suspicion. Second, when the seam is not visible externally, take a full ring so that the whole circumference is available; a coupon cut at an arbitrary angular position has a high chance of missing a narrow high-frequency bond line. Third, take duplicate sections from two different locations along the length, because a defect or an inclusion stringer that appears once may be misinterpreted as a seam.
Where the dispute concerns a whole consignment, statistical sampling is appropriate: define the lot, take samples at random from across the lot, and base the conclusion on the lot rather than on a single length. The result should be recorded in terms of the number of lengths examined, the number showing a bond line, and the method used.
Documentary Evidence and What It Can Establish
The inspection certificate is the primary documentary control. Under EN 10204 type 3.1 the certificate is issued by the manufacturer's independent inspection representative and states the product form, the standard, the grade, the heat number, the dimensions and the test results. Under API SPEC 5L the pipe marking includes the product specification level, the grade and the manufacturing process designation, so a length marked as seamless that shows a bond line is a marking non-conformance in addition to a product non-conformance.
Two documentary checks are worth routine attention. Confirm that the standard cited on the certificate actually covers the product form; ASTM A53 Type S is seamless while Type E and Type F are welded, and ASTM A106 is seamless only, so a certificate citing ASTM A106 for an obviously welded pipe is internally inconsistent. Then confirm that the non-destructive examination method reported matches what the product form can support: a full-body ultrasonic report with no seam examination is normal for seamless and suspicious for submerged arc welded pipe.
FAQ
Q: Why does a rough grind and etch sometimes not reveal the seam?
High-frequency welds are extremely narrow and may have been normalised in line, so the bond line can be nearly invisible after a coarse preparation. Fine grinding, polishing and a controlled nital etch, or examination at higher magnification, is needed to reveal it.
Q: Is a hardness peak along a line enough to prove the pipe is welded?
No. A hardness peak is strong supporting evidence of an un-normalised weld, but identification should be confirmed by metallographic examination of a section. Hardness is best used for screening and for sour service compliance.
Q: Can ultrasonic testing identify the product form without cutting?
It can locate a suspected seam reliably enough to justify sampling. A distinct circumferential reflection at a constant angular position indicates a fusion line, but the conclusion should be confirmed metallographically.
Q: What magnification is appropriate for seam identification?
A macro examination at 5 to 20 times is usually sufficient to reveal the bond line and heat-affected zone in carbon steel. Higher magnification is used to characterise weld defects, decarburisation or grain structure once the seam has been located.
Q: Do all welded pipes show an internal flash line?
No. Internal flash is commonly scarfed in line, so the bore may be smooth. Its absence proves nothing, and identification then relies on external evidence, hardness or metallography.
Q: What sample size is reasonable for a consignment dispute?
There is no universal number. A defensible approach is to define the lot, take random samples across it, and report the number examined and the number found with a bond line, so the conclusion is stated as a lot-level finding rather than as an anecdote.





