How many test specimens does a PQR groove weld test plate require under AWS D1.1?

For a standard groove weld PQR test plate in AWS D1.1:2025, you need at minimum two reduced-section tensile specimens and four bend specimens (typically two face bends and two root bends, or four side bends for material 3/8 inch and thicker). If CVN toughness testing is required under Table 6.8, additional Charpy specimens are cut from a separate location in the test plate.

Can the CWI performing the weld cut the test specimens themselves?

Specimen cutting is normally performed by the test laboratory or a mechanical shop using a saw or flame-cut with grinding cleanup. The CWI or welding engineer is responsible for identifying the specimen locations on the test plate, witnessing or verifying the cut, and confirming that the specimens come from the required locations — not from the discarded runoff tabs or outside the qualified groove area.

What happens to the discarded sections of the PQR test plate?

AWS D1.1 requires that 1-inch minimum discard sections be removed from each end of the test plate before cutting specimens. These end sections contain the start and stop craters and arc strike areas, which are not representative of production welding. The remaining plate is the testable area; specimens come from this zone only.

Does AWS D1.1 require the PQR test to be witnessed?

AWS D1.1:2025 does not mandate that a CWI or third-party inspector witness the PQR weld, but many contractors and owners require it by contract. The PQR must be a true record of what was welded and how — witnessed or not, any falsification of parameters invalidates the qualification. Witnessing is strongly recommended for new procedures because it provides an independent record that the parameters are correct.

PQR Test Coupon Cutting and Specimen Identification Under AWS D1.1

Where you cut tensile and bend specimens from a PQR test plate matters under AWS D1.1:2025. This guide covers coupon location, specimen marking, and documentation before the test lab gets involved.

A procedure qualification record (PQR) is only as reliable as the test specimens extracted from it. AWS D1.1:2025 prescribes not just how many specimens are required and what tests they undergo, but where in the test plate they must come from. Cutting specimens from the wrong location — or failing to document the extraction properly — can invalidate a qualification test even when all the mechanical results pass.

This article walks through the PQR test plate requirements from the end of welding to delivery at the test laboratory: discard zones, specimen location, coupon dimensions, identification marking, and the chain of documentation that connects a test report to a WPS.

The PQR test plate: dimensions and discard requirements

AWS D1.1:2025 specifies the minimum length and width of the PQR test plate based on the number and type of specimens required. For a standard groove weld qualification, the test plate must be large enough to yield the required specimens after the mandatory discard sections are removed.

End discards. At least 1 inch (25 mm) of material must be discarded from each end of the groove weld plate before any test specimens are cut. These end sections are where the welder starts and stops the weld. Arc strikes at the start may produce a cold lap or lack of fusion; stop craters at the end contain a depression and potential porosity or shrinkage cavities. Neither location is representative of continuous production welding, and AWS D1.1 excludes them from the qualified test area.

Runoff tab removal. Many PQR groove welds are made with extension bars (runoff tabs) clamped to the ends of the test plate to allow the welder to start and stop the arc outside the testable zone. When runoff tabs are used, they are removed after welding (by sawing and grinding to the plate edge), and the discard requirement is measured from the edge of the base plate, not the edge of the runoff tab. Attempting to use the runoff tab zone as part of the testable area is a common misunderstanding.

Test plate thickness. The test plate must be welded to the thickness for which qualification is sought. AWS D1.1:2025 Table 6.6 (essential variables) lists base metal thickness as a variable with defined qualification ranges — a PQR welded at 3/4-inch thickness qualifies a WPS for a specific range of production thicknesses. If the PQR test plate is thinner or thicker than intended, the qualification range shifts.

Specimen types and their locations

For a complete joint penetration groove weld PQR, AWS D1.1:2025 requires the following specimen types (the exact quantity and type depend on plate thickness):

Reduced-section tensile specimens. These are machined to a parallel-sided gauge section that includes the full weld and heat-affected zone. The tensile specimen tests the weld and HAZ in tension; the minimum tensile strength must meet or exceed the specified minimum tensile strength of the base metal or the matching filler metal, whichever is less.

Tensile specimens are cut transverse to the weld axis, spanning the full weld cross-section from one plate face to the other. The specimen location is centered in the testable zone of the plate, spaced to avoid the edge discards and the bend specimen locations.

Bend specimens. Bend tests evaluate ductility and detect planar discontinuities (LOF, cracks) that may not show in the tensile test. The specimen is bent over a mandrel of specified diameter; the test side is placed in tension during bending so that discontinuities open up under the bend.

For base metal thickness less than 3/8 inch (10 mm), face bends and root bends are typically required — the weld face side and root side are each placed in tension on separate specimens. For base metal 3/8 inch and thicker, side bends are used instead: the specimen is oriented with the weld cross-section facing the mandrel, so both the face and root are tested simultaneously in the same specimen. Side bends give a more complete picture of weld quality through the full joint thickness.

Bend specimen locations are staggered relative to the tensile specimens to prevent overlap. AWS D1.1:2025 provides dimension tables and coupon cutting plans showing the approximate locations — these plans are a starting point, and the actual layout on each specific test plate must account for the testable zone dimensions after discards.

CVN Charpy specimens (when required). If the WPS requires compliance with supplementary essential variables under AWS D1.1:2025 Table 6.8 — meaning the procedure will be used for welds with required CVN toughness — the PQR must include Charpy V-notch impact testing. Three Charpy specimens per test location are standard: one with the notch in the weld metal, one with the notch in the heat-affected zone, and sometimes one at a specified distance from the fusion line.

CVN specimens are cut from a separate zone of the test plate and are oriented longitudinally in the weld — parallel to the weld axis — so the notch crosses the weld, HAZ, or fusion line as intended. The test temperature is specified in the WPS or by the governing code or contract documents.

For more on when CVN testing is required, see CVN supplementary essential variables under AWS D1.1:2025 Table 6.8.

Specimen marking and identification

Specimen identification is the link between the test result and the physical weld. If a specimen's origin cannot be traced back to a specific PQR test plate, welded by a specific procedure, the test result cannot support the PQR.

Mark each specimen before it is cut from the plate. Use a low-stress stamping method — vibro-engraving or a felt-tip marker is acceptable for initial identification; hard stamp marking directly on the gauge section of a tensile specimen is not acceptable because the notch effect can initiate fracture at the stamp. Stamp the specimen blank (before machining) in a non-gauge area.

The identification marking system should capture at minimum:

PQR number or procedure number
Specimen type (T = tensile, FB = face bend, RB = root bend, SB = side bend, CVN)
Specimen position in the plate (e.g., 1, 2, 3 from left to right in the testable zone)
Test plate heat number or material certification number (for traceability of the base metal used in qualification)

Maintain a cutting diagram — a sketch or drawing that shows the test plate dimensions, the location of each specimen cut, and the identification numbers of each specimen. The cutting diagram becomes part of the PQR documentation package and allows anyone reviewing the PQR to confirm that specimens came from the required zones.

Before sending to the lab: CWI documentation role

A CWI or responsible welding engineer should review and document the following before releasing the test plate for cutting:

Confirm welding parameters were recorded. The PQR must record the actual parameters used during welding, not the intended parameters. This includes amperage, voltage, travel speed, preheat, interpass temperature, electrode classification and diameter, and any process-specific variables such as gas flow rate or wire feed speed. If parameters were not recorded during welding, the PQR cannot be reconstructed after the fact. For more on what the PQR record must contain, see ASME IX QW-482/QW-483 WPS and PQR forms explained (the ASME format parallels the AWS D1.1 documentation intent).

Photograph the as-welded test plate. A photograph before cutting establishes the visual condition of the weld surface and shows that no post-weld grinding or re-melting occurred that would misrepresent the production weld. Photograph both the face and root sides.

Visually inspect the test plate. AWS D1.1 requires the test plate to pass visual inspection before it is submitted for mechanical testing. Surface cracks, excessive undercut, or porosity on the weld face that exceeds visual acceptance criteria means the procedure failed before any specimen is cut. A procedure that fails visual inspection must be re-welded; sending a failing test plate to the lab wastes test fees and delays qualification.

Verify the weld is representative of the intended WPS. If a parameter significantly exceeded the intended WPS range during the PQR weld — due to equipment variance, technique, or process instability — the PQR can only support a WPS that encompasses that parameter range. If the intent was to qualify a WPS with an upper heat input of 60 kJ/in and the actual PQR heat input was 80 kJ/in, the WPS qualified by that PQR must acknowledge the 80 kJ/in upper bound.

At the test laboratory

Select a laboratory accredited for mechanical testing of welds, preferably with AASHTO or A2LA accreditation if the work involves bridge or infrastructure projects. The lab receives the test plate (or pre-cut blanks) and machines the final specimen geometry per the specified standard.

Provide the lab with the following:

Cutting diagram showing specimen layout
Specimen identification list
Material test reports for the base metal plate (so the lab can verify tensile strength against the specimens)
The required test standard (AWS D1.1:2025 geometry and acceptance criteria)
CVN test temperature (if applicable)
A contact for any questions during specimen preparation

The lab's report will reference the specimen identification numbers; your cutting diagram ties those numbers back to locations in the weld. That connection is the chain of traceability from "weld test" to "WPS support document."

For a complete view of PQR laboratory selection considerations, see PQR test laboratory selection under AWS D1.1. For software that tracks WPS and PQR status across multiple qualification runs, see our pricing page.

Rule library based on AWS D1.1:2025; verify against your governing edition. The authority having jurisdiction (AHJ) or contract may specify AWS D1.1:2020 or an earlier edition.