1 Introduction

European Welding Standards require test coupons that are made for
welding procedure qualification testing to be subjected to non-destructive
and then destructive testing.

The tests are called destructive tests because the welded joint is destroyed
when various types of test piece are taken from it.

Destructive tests can be divided into two groups:

 Those used to measure a mechanical property – quantitative tests

 Those used to assess the joint quality – qualitative tests

Mechanical tests are quantitative because a quantity is measured – a

mechanical property such as tensile strength, hardness or impact
toughness.

Qualitative tests are used to verify that the joint is free from defects – they
are of sound quality and examples of these are bend tests, macroscopic
examination and fracture tests (fillet fracture and nick-break).

2 Test Types, Test Pieces and Test Objectives

Various types of mechanical test are used by material manufacturers/
suppliers to verify that plates, pipes, forgings etc have the minimum property
values specified for particular grades.

Design engineers use the minimum property values listed for particular
grades of material as the basis for design and the most cost-effective
designs are based on an assumption that welded joints have properties that
are no worse than those of the base metal.

The quantitative (mechanical) tests carried out for welding procedure

qualification are intended to demonstrate that the joint properties satisfy
design requirements.

The emphasis in the following sub-sections is on the destructive tests and

test methods that are widely used for welded joints.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
2.1 Transverse tensile tests
Test objective
Welding procedure qualification tests always require transverse tensile tests
to show that the strength of the joint satisfies the design criterion.

Test specimens
A transverse tensile test piece typical of the type specified by European
Welding Standards is shown below.

Standards, such as EN 895, that specify dimensions for transverse tensile

test pieces require all excess weld metal to be removed and the surface to
be free from scratches.

Parallel
length

Test pieces may be machined to represent the full thickness of the joint but
for very thick joints it may be necessary to take several transverse tensile
test specimens to be able to test the full thickness.

Test method
Test specimens are accurately measured before testing. Specimens are
then fitted into the jaws of a tensile testing machine and subjected to a
continually increasing tensile force until the specimen fractures.

The tensile strength (Rm) is calculated by dividing the maximum load by the
cross-sectional area of the test specimen - measured before testing.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 The test is intended to measure the tensile strength of the joint and
thereby show that the basis for design, the base metal properties, remains
the valid criterion.

Acceptance criteria
If the test piece breaks in the weld metal, it is acceptable provided the
calculated strength is not less than the minimum tensile strength specified,
which is usually the minimum specified for the base metal material grade.

In the ASME IX code, if the test specimen breaks outside the weld or fusion
zone at a stress above 95% of the minimum base metal strength the test
result is acceptable.

2.2 All-weld tensile tests

Test objective
There may be occasions when it is necessary to measure the weld metal
strength as part of welding procedure qualification – particularly for elevated
temperature designs.

The test is carried out in order to measure tensile strength and also yield (or
proof strength) and tensile ductility.

All-weld tensile tests are also regularly carried out by welding consumable
manufacturers to verify that electrodes and filler wires satisfy the tensile
properties specified by the standard to which the consumables are certified.

Test specimens
As the name indicates, test specimens are machined from welds parallel
with their longitudinal axis and the specimen gauge length must be 100%
weld metal.

Round cross section

Round tensile specimen from a Round tensile specimen from

welding 2009
Rev 1 January procedure qualification an electrode classification test
test piece
Destructive Testing piece
Copyright  TWI Ltd 2009
 Test method
Specimens are subjected to a continually increasing force in the same way
that transverse tensile specimens are tested.

Yield (Re) or proof stress (Rp) are measured by means of an extensometer

that is attached to the parallel length of the specimen and is able to
accurately measure the extension of the gauge length as the load is
increased.

Typical load extension curves and their principal characteristics are shown
below.

Load extension curve for a steel that Load-extension curve for a steel (or other
shows a distinct yield point at the metal) that does not show a distinct yield
elastic limit point; proof stress is a measure of the
elastic limit

Tensile ductility is measured in two ways:

 Percent elongation of the gauge length

 Percent reduction of area at the point of fracture

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 The figure below illustrates these two ductility measurements.

2.3 Impact toughness tests

Test objective
Charpy V notch test pieces have become the internationally accepted
method for assessing resistance to brittle fracture by measuring the energy
to initiate, and propagate, a crack from a sharp notch in a standard sized
specimen subjected to an impact load.

Design engineers need to ensure that the toughness of the steel used for a
particular item will be high enough to avoid brittle fracture in service and so
impact specimens are tested at a temperature that is related to the design
temperature for the fabricated component.

C-Mn and low alloy steels undergo a sharp change in their resistance to
brittle fracture as their temperature is lowered so that a steel that may have
very good toughness at ambient temperature may show extreme brittleness
at sub-zero temperatures – as illustrated in following figure.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 Ductile fracture
Temperature range

47 Joules

Transition range Ductile/Brittle

transition
point

28 Joules
Energy absorbed
Brittle fracture
- 50 - 40 - 30 - 20 - 10 0
Testing temperature - Degrees Centigrade
Three specimens are normally tested at each temperature

The transition temperature is defined as the temperature that is mid-way

between the upper shelf (maximum toughness) and lower shelf (completely
brittle). In the above the transition temperature is -20°C.

Test specimens
The dimensions for test specimens have been standardised internationally
and are shown below for full sized specimens. There are also standard
dimensions for smaller sized specimens, for example 10x7.5mm and
10x5mm.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 Charpy V notch test piece dimensions for full sizes specimens

Specimens are machined from welded test plates with the notch position
located in different positions according to the testing requirements but
typically in the centre of the weld metal and at positions across the HAZ – as
shown below.

Typical notch positions for Charpy V notch test specimens from double V butt
welds

Test method
Test specimens are cooled to the specified test temperature by immersion in
an insulated bath containing a liquid that is held at the test temperature.

After allowing the specimen temperature to stabilise for a few minutes it is

quickly transferred to the anvil of the test machine and a pendulum hammer
quickly released so that the specimen experiences an impact load behind
the notch.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 The main features of an impact test machine are shown below.

Impact testing machine Impact specimen on the anvil

showing the hammer position
at point of impact.

Charpy V notch test pieces

before and after testing

The energy absorbed by the hammer when it strikes each test specimen is
shown by the position of the hammer pointer on the scale of the machine.
Energy values are given in Joules (or ft-lbs in US specifications).

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 Impact test specimens are taken in triplicate (three specimens for each
notch position) as there is always some degree of scatter in the results
particularly for weldments.

Acceptance criteria
Each test result is recorded and an average value calculated for each set of
three tests. These values are compared with those specified by the
application standard or client to establish whether specified requirements
have been met.

After impact testing, examination of the test specimens provides additional

information about their toughness characteristics and may be added to the
test report:

 Percent crystallinity – the % of the fracture face that has crystalline

appearance which indicates brittle fracture; 100% indicates completely
brittle fracture
 Lateral expansion – the increase in width of the back of the specimen
behind the notch – as indicated below; the larger the value the tougher
the specimen

A specimen that exhibits extreme brittleness will show a clean break, both
halves of the specimen having a completely flat fracture face with little or no
lateral expansion.

A specimen that exhibits very good toughness will show only a small degree
of crack extension, without fracture and a high value of lateral expansion.

2.4 Hardness testing

Test objective
The hardness of a metal is its’ resistance to plastic deformation. This is
determined by measuring the resistance to indentation by a particular type
of indenter.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 A steel weldment with hardness above a certain maximum may be
susceptible to cracking, either during fabrication or in service, and welding
procedure qualification testing for certain steels and applications requires
the test weld to be hardness surveyed to ensure there are no regions
exceed the maximum specified hardness.

Specimens prepared for macroscopic examination can also be used for

taking hardness measurements at various positions of the weldment -
referred to as a hardness survey.

Test methods
There are three widely used methods for hardness testing:

 Vickers hardness test - uses a square-base diamond pyramid indenter.

 Rockwell hardness test - uses a diamond cone indenter or steel ball.
 Brinell hardness test - uses a ball indenter.

The hardness value being given by the size of the indentation produced
under a standard load, the smaller the indentation, the harder the metal.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 The Vickers method of testing is illustrated below.

d1  d2
d
2

Both the Vickers and Brinell methods are suitable for carrying out hardness
surveys on specimens prepared for macroscopic examination of weldments.

A typical hardness survey requires the indenter to measure the hardness in

the base metal (on both sides of the weld), the weld metal and across the
HAZ (on both sides of the weld).

The Brinell method gives an indentation that is too large to accurately

measure the hardness in specific regions of the HAZ and is mainly used to
measure hardness of base metals.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 A typical hardness survey (using Vickers hardness indenter) is shown
below:

Hardness values are shown on test reports as a number followed by letters

indicating the test method, for example:

240HV10 = hardness 240, Vickers method, 10kg indenter load

22HRC = hardness 22, Rockwell method, diamond cone indenter (scale C)

238HBW = hardness 238, Brinell method, tungsten ball indenter

2.5 Crack tip opening displacement (CTOD) testing

Test objective
Charpy V notch testing enables engineers to make judgements about risks
of brittle fracture occurring in steels, but a CTOD test measures a material
property - fracture toughness.

Fracture toughness data enables engineers to carry out fracture mechanics

analyses such as:

 Calculating the size of a crack that would initiate a brittle fracture under
certain stress conditions at a particular temperature
 The stress that would cause a certain sized crack to give a brittle fracture
at a particular temperature

This data is essential for making an appropriate decision when a crack is

discovered during inspection of equipment that is in-service.

Test specimens
A CTOD specimen is prepared as a rectangular (or square) shaped bar cut
transverse to the axis of the butt weld. A V notch is machined at the centre
of the bar, which will be coincident with the test position - weld metal or
HAZ.

A shallow saw cut is made at the bottom of the notch and the specimen is
then put into a machine that induces a cyclic bending load until a shallow
fatigue crack initiates from the saw cut.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 The specimens are relatively large – typically having a cross section B x 2B
and length ~10B (B = full thickness of the weld). The test piece details are
shown below.

Test method
CTOD specimens are usually tested at a temperature below ambient and
the specimen temperature is controlled by immersion in a bath of liquid that
has been cooled to the required test temperature.

A load is applied to the specimen to cause bending and induce a

concentrated stress at the tip of the crack and a clip gauge, attached to the
specimen across the mouth of the machined notch, gives a reading of the
increase in width of the crack mouth as the load is gradually increased.

For each test condition (position of notch and test temperature) it is usual
practice to carry out three tests.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 The figures below illustrate the main features of the CTOD test.

Fracture toughness is expressed as the distance the crack tip opens without
initiation of a brittle crack.

The clip gauge enables a chart to be generated showing the increase in

width of the crack mouth against applied load from which a CTOD value is
calculated.

Acceptance criteria
An application standard or client may specify a minimum CTOD value that
indicates ductile tearing. Alternatively, the test may be for information so that
a value can be used for an engineering critical assessment.

A very tough steel weldment will allow the mouth of the crack to open widely
by ductile tearing at the tip of the crack whereas a very brittle weldment will
tend to fracture when the applied load is quite low and without any extension
at the tip of the crack.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 CTOD values are expressed in millimetres - typical values might be
<<~0.1mm = brittle behaviour; >~1mm = very tough behaviour.

2.6 Bend testing

Test objective
Bend tests are routinely taken from welding procedure qualification test
pieces and sometimes on welder qualification test pieces.

Subjecting specimens to bending is a simple method of verifying there are

no significant flaws in the joint. Some degree of ductility is also
demonstrated.

Ductility is not actually measured but it is demonstrated to be satisfactory if

test specimens can withstand being bent without fracture or fissures above
a certain length.

Test specimens
There are four types of bend specimen:

 Face: Specimen taken with axis transverse to butt welds up to ~12mm

thickness and bent so that the face of the weld is on the outside of the
bend (face in tension).
 Root: Test specimen taken with axis transverse to butt welds up to
~12mm thickness and bent so that the root of the weld is on the outside
of the bend (root in tension).
 Side: Test specimen taken as a transverse slice (~10mm) from the full
thickness of butt welds >~12mm and bent so that the full joint thickness
is tested (side in tension).
 Longitudinal bend: Test specimen taken with axis parallel to the
longitudinal axis of a butt weld; specimen thickness is ~12mm and the
face or root of weld may be tested in tension.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 Test method
Bend tests for welding procedure and welder qualification are usually
guided bend tests.

Guided means that the strain imposed on the specimen is uniformly

controlled by being bent around a former with a certain diameter.

The diameter of the former used for a particular test is specified in the code,
having been determined by the type of material being tested and the ductility
that can be expected from it after welding and any post weld heat treatment
(PWHT).

The diameter of the former is usually expressed as a multiple of the

specimen thickness (t) and for C-Mn steel it is typically 4t but for materials
that have lower tensile ductility the radius of the former may be greater than
10t.

The standard that specifies the test method will specify the minimum bend
angle that the specimen must experience and this is typically 120-180 0.

Acceptance criteria
Bend tests pieces should exhibit satisfactory soundness by not showing
cracks or any signs of significant fissures or cavities on the outside of the
bend.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 Small indications less than about 3mm in length may be allowed by some
standards.

2.7 Fracture tests

2.7.1 Fillet weld fractures
Test objective
The quality/soundness of a fillet weld can be assessed by fracturing test
pieces and examining the fracture surfaces.

This method for assessing the quality of fillet welds may be specified by
application standards as an alternative to macroscopic examination.

It is a test method that can be used for welder qualification testing according
to European Standards but is not used for welding procedure qualification.

Test specimens
A test weld is cut into short lengths (typically 50mm) and a longitudinal
notch is machined into the specimen as shown below. The notch profile may
be square, V or U shape.

Test method
Specimens are made to fracture through their throat by dynamic strokes
(hammering) or by pressing, as shown below. The welding standard or
application standard will specify the number of tests (typically four).

Hammer Moving press

stroke

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
 Acceptance criteria
The standard for welder qualification, or application standard, will specify the
acceptance criteria for imperfections such as lack of penetration into the root
of the joint and solid inclusions and porosity that are visible on the fracture
surfaces.

Test reports should also give a description of the appearance of the fracture
and location of any imperfection

2.7.2 Butt weld fractures (nick-break tests)

Test objective
The objective of these fracture tests is the same as for fillet fracture tests.

These tests are specified for welder qualification testing to European

Standards as an alternative to radiography. They are not used for welding
procedure qualification testing.

Test specimens
Test specimens are taken from a butt weld and notched so that the fracture
path will be in the central region of the weld. Typical test piece types are
shown below.

Test method
Test pieces are made to fracture by hammering or three-point bending.

Acceptance criteria
The standard for welder qualification, or application standard, will specify the
acceptance criteria for imperfections such as lack of fusion, solid inclusions
and porosity that are visible on the fracture surfaces.

Test reports should also give a description of the appearance of the fracture
and location of any imperfection.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009
3 Macroscopic Examination
Transverse sections from butt and fillet welds are required by the European
Standards for welding procedure qualification testing and may be required
for some welder qualification testing for assessing the quality of the welds.

This is considered in detail in a separate section of these course notes.

3.1 European Standards for destructive test methods

The following Standards are specified by the European Welding Standards
for destructive testing of welding procedure qualification test welds and for
some welder qualification test welds.

EN 875 Destructive tests on welds in metallic materials – Impact

tests – test specimen location, notch orientation and
examination.
EN 895 Destructive tests on welds in metallic materials –
transverse tensile test.
EN 910 Destructive tests on welds in metallic materials – bend
tests.
EN 1321 Destructive tests on welds in metallic materials –
macroscopic and microscopic examination of welds.
BS EN 10002 Metallic materials - Tensile testing. Part 1: Method of test
at ambient temperature.
BS EN 10002 Tensile testing of metallic materials. Part 5: Method of test
at elevated temperatures.

Rev 1 January 2009

Destructive Testing
Copyright  TWI Ltd 2009