CWI Destructive Testing Study Guide: Bend, Tension, Charpy

Why destructive tests matter to an inspector

Destructive tests cut, bend, pull or break a sample so its properties can be measured directly. They are used to qualify welding procedures and welders, to classify filler metals and to investigate failures. An inspector often witnesses specimen removal and testing. You need to recognize a valid test and tell a passing result from a failing one. The standard reference is AWS B4.0, Standard Methods for Mechanical Testing of Welds.

Many exam questions ask you to match a test to the property it measures:

Which tests a procedure or welder qualification calls for is covered in the welding performance guide.

Tension tests

Transverse reduced-section tension test. The specimen crosses the weld, so it contains weld metal, HAZ and base metal. It measures the tensile strength of the joint as a whole and can break in any zone. Most codes accept the result if the strength meets the specified minimum tensile strength of the base metal, wherever the break occurs.

All-weld-metal tension test. The specimen is machined entirely from weld metal. It gives the yield strength, tensile strength, elongation and reduction of area of the weld metal itself. This is the test used to classify filler metals.

Strength is load divided by original area (worked examples are in the welding calculations guide). Ductility is reported two ways. Percent elongation is the change in gauge length over the original length. Reduction of area is the change in cross-sectional area over the original area.

Guided bend tests

A guided bend test bends a strip cut from the weld around a mandrel to a set radius. The outside of the bend stretches, so discontinuities there open up. The test shows both soundness (lack of fusion, slag, porosity) and ductility.

• Face bend: the weld face is on the outside, in tension.
• Root bend: the root surface is on the outside, in tension. It exposes incomplete penetration and root fusion problems.
• Side bend: the specimen is cut across the weld so the full thickness is on the bent surface.

Side bends replace face and root bends on thicker material. A thick face or root specimen would need huge force and would only test one surface region. A side bend tests the full weld thickness in a strip thin enough to bend. Codes set acceptance limits on the size and total length of open discontinuities, and on cracks at the specimen corners.

Charpy V-notch impact test

A small bar with a machined V-notch is cooled or heated to a set temperature and broken by a swinging pendulum. The energy absorbed, in ft-lbf or joules, measures notch toughness. Lateral expansion and percent shear fracture may also be reported.

Temperature matters. Many steels have a transition range: tough above it, brittle below it. That is why toughness requirements always state a test temperature. The metallurgy and heat control guide explains what heat input does to toughness. Specimens may be notched in the weld metal or in the HAZ, depending on what the code wants to check.

Hardness tests

• Brinell: a large ball makes a big impression; good for base metal, too coarse for narrow weld zones.
• Rockwell: B scale uses a ball, C scale uses a diamond cone; quick readings from a dial or display.
• Vickers: a square-based diamond pyramid with light loads makes small indents. This lets hardness be mapped across weld metal, HAZ and base metal.

Hardness surveys are used to check for hard HAZ structures that may be prone to cracking, and to verify heat treatment.

Macroetch test

A cross section is cut, ground, polished and etched with a mild acid. The weld metal, fusion boundary and HAZ become visible to the eye or at low magnification. A macroetch shows:

• Weld profile and fillet leg sizes.
• Root penetration and fusion.
• The extent of the HAZ.
• Gross discontinuities such as large porosity or cracks.

It is quick and cheap, which is why it is used in fillet weld qualification and in production checks.

Fracture tests

Nick-break test. Used in pipeline standards such as API 1104. Notches are cut at the weld edges so the specimen breaks through the weld metal. The fracture faces are then examined for gas pockets, slag, lack of fusion and other discontinuities and judged against the standard's limits.

Fillet weld break test. A fillet weld is loaded so it breaks through the weld. The fracture surface is then examined for complete fusion to the root and for inclusions or porosity. It is one way codes such as AWS D1.1 check welder skill on fillet welds.

Reading fracture surfaces

• Ductile fracture: dull, fibrous surface with visible deformation, necking or shear lips at the edges. It absorbs a lot of energy.
• Brittle fracture: bright, crystalline or granular surface with little deformation. It absorbs little energy.
• Fatigue fracture: smooth zones with beach marks spreading from an origin, ending in a rougher final overload area.

Specimen preparation and witnessing

A test can fail because of a poor specimen rather than a poor weld. Saw cuts, grinding marks across the specimen, sharp edges and overheating during preparation can all cause false results. Codes specify where specimens are removed, so results cannot be chosen selectively. The inspector's role is to confirm the specimen locations, preparation and test method, and to record results accurately.

Common exam traps

• Confusing the transverse tension test (whole joint) with the all-weld-metal test (weld metal only).
• Thinking bend tests measure strength; they measure soundness and ductility.
• Forgetting that Charpy results mean little without the test temperature.
• Choosing Brinell for a narrow HAZ, where Vickers is the better fit.
• Mixing up which surface is in tension in face and root bends.

Mark J. Curry's tip: if an exam question names a property, name the test first. Then check which specimen orientation or location the question describes.

Work through the free destructive testing practice questions. To see which Part A topic to study first, take the free CWI diagnostic test.

Sources

The Destructive Tests questions cite these documents, by document and, where we can confirm it, chapter or section. We never reproduce them.

1. AWS A5.1 (1 question)
2. AWS D1.1 (1 question)

Found a mistake? Use "Report a problem" under any question, and Mark J. Curry checks it.

Questions? support@cwiready.com