10.
Testing of Welded Joints
10. Testing of Welded Joints
Ls
126
The basic test for determination of material
a
S S
S S
in test area
in test area
S
S
b
b1
S
S
Generally, it is carried out using a round
L0
Lc
r
behaviour is the tensile test.
specimen. When determining the strength of
a welded joint, also standardised flat speci-
Lt
total length
head width
Lt
b1
width of parallel length
plates
b
tubes
b
1 2
Lc
r
parallel length ) )
radius of throat
depends on test unit
b + 12
12 with a £ 2
25 with a > 2
6 with D £ 50
12 with 50 < D £
168,3
³ L S + 60
³ 25
mens are used. Figure 10.1 shows both standard specimen shapes for that test. A
specimen is ruptured by a test machine while
the actual force and the elongation of the
1
d1
d
S
S
) for pressure welding and beam welding, L S = 0.
2
) for some other metallic materials (e.g.aluminium, copper and their alloys)
__ L c ³ L S +100 may be required
r
is typical for this test, Figure 10.2.
L0 = measurement length
(L0 = k ÖS0 with k = 5,65)
Lt = total length
S0 = initial cross-section within
test length
br-er10-01.cdr
ment values, tension σ and strain ε are calculated. If σ is plotted over ε, the drawn diagram
LO
LC
Lt
d = specimen diameter
d1 = head diameter depending
on clamping device
LC = test length = L0 + d/2
r = 2 mm
specimen is measured. With these measure-
Normally, if a steel with a bcc lattice structure
© ISF 2002
Flat and Round Tensile Test Specimen
to EN 895, EN 876, and EN 10 002
is tested, a curve with a clear yield point is
obtained (upper picture). Steels with a fcc
lattice structure show a curve without yield
Figure 10.1
point.
The most important characteristic values
s
which are determined by this test are: yield
stress ReL, tensile strength Rm, and elongation
Rm
ReH
Rel
sf
A.
To determine the deformability of a weld, a
e
ALud
Ag
A
bending test to DIN EN 910 is used, Figure
10.3. In this test, the specimen is put onto two
s
supporting rollers and a former is pressed
Rm
through between the rollers. The distance of
RP0,2
RP0,01
sf
the supporting rollers is Lf = d + 3a (former
diameter + three times specimen thickness).
e
0,2 %
0,01 %
Ag
is observed. If a surface crack develops, the
A
br-er10-02.cdr
© ISF 2002
Stress-Strain Diagram With
and Without Distinct Yield Point
Figure 10.2
The backside of the specimen (tension side)
test will be stopped and the angle to which
the specimen could be bent is measured. The
10. Testing of Welded Joints
127
test result is the bending angle and the diameter of the used former. A bending angle of 180°
is reached, if the specimen is pressed through the supporting rollers without development of
a crack. In Figure 10.3 specimen shapes of this test are shown. Depending on the direction
the weld is bent, one distinguishes (from top to bottom) transverse, side, and longitudinal
bending specimen. The tension side of all three specimen types is machined to eliminate any
influences
through
Specimen
on
the
notch
test
effects.
thickness
of
transverse and longitudinal
specimens
is
thickness.
Side
the
plate
bending
specimens are normally only
used with very thick plates,
here the specimen thickness
is fixed at 10 mm.
A
determination
of
the
toughness of a material or
Figure 10.3
welded joint is carried out
with the notched bar impact test. A cuboid specimen with a V-notch is placed on a support
and then hit by a pendulum ram of the impact testing machine (with very tough materials, the
specimen will be bent and
drawn through the supports). The used energy is
measured.
Figure
10.4
represents sample shape,
notch
shape
(Iso-V-
specimen), and a schematic presentation of test
results.
Figure 10.4
10. Testing of Welded Joints
128
Three specimens are tested at each test tem-
b
Designation
VWS a/b
Dicke
a
RL
VWS a/b
(fusion
weld)
Fusion line/bonding zone
perature, and the average values as well as
b
Weld centre
Designation
RL
the range of scatter are entered on the impact
a
Dicke
b
b
energy-temperature diagram (AV-T curve).
VWT 0/b
VHT 0/b
This graph is divided into an area of high im-
a
b
b
pact energy values, a transition range, and an
VHT a/b
a
area of low values. A transition temperature is
VWT 0/b
b
a
b
VWT a/b
VHT a/b
b
b
VWT a/b
drop of toughness values. When the tempera-
a
RL
RL
VHT a/b
a
RL
takes place.
V = Charpy-V notch
W = notch in weld metal; reference line is centre line of weld
H = notch in heat affected zone; reference line is fusion line or bonding zone
(notch should be in heat affected zone)
S = notched area parallel to surface
T = notch through thickness
a = distance of notch centre from reference line (if a is on centre line of weld, a = 0 and
should be marked)
b = distance between top side of welded joint and nearest surface of the specimen
(if b is on the weld surface, then b = 0 and should be marked)
br-er10-05.cdr
As this steep drop mostly extends across a
certain area, a precise assignment of transi© ISF 2002
Position of Charpy-V Impact Test
Specimen in Welded Joints to EN 875
Figure 10.5
ture falls below this transition temperature, a
transition of tough to brittle fracture behaviour
a
RL
assigned to the transition range, i.e. the rapid
tion temperature cannot be carried out. Following DIN 50 115, three definitions of the
transition temperature are useful, i.e. to fix TÜ
to:
1.) a temperature where the level of impact values is half of the level of the high range,
2.) a temperature, where the fracture area of the specimen shows still 50% of tough fracture
behaviour
3.) a temperature with an impact energy value of 27 J.
Figure 10.5 illustrates a specimen position and notch position related to the weld according to
DIN EN 875. By modifying the notch position, the impact energy of the individual areas like
HAZ, fusion line, weld metal, and base metal can be determined in a relatively accurate way.
Figure 10.6 presents the influence of various alloy elements on the AV-T - curve. Three basically different influences can be seen. Increasing manganese contents increase the impact
values in the area of the high level and move the transition temperature to lower values. The
values of the low levels remain unchanged, thus the steepness of the drop becomes clearer
with increasing Mn-content. Carbon acts exactly in the opposite way. An increasing carbon
content increases the transition temperature and lowers the values of the high level, the steel
becomes more brittle. Nickel decreases slightly the values of the high level, but increases the
10. Testing of Welded Joints
129
values of the low level with increasing con-
specimen position:
core longitudinal
J
tent. Starting with a certain Nickel content
specimen shape:
ISO V
(depends also from other alloy elements), a
300
2% Mn
steep drop does not happen, even at lowest
1% Mn
200
0,5% Mn
temperature the steel shows a tough fracture
behaviour.
Charpy impact energy AV
100
0% Mn
27
200
In Figure 10.7, the AV-T – curves of some
J
100
27
13% Ni
8,5%
5% 3,5%
2% Ni
commonly used steels are collected. These
0% Ni
curves are marked with points for impact en-
200
ergy values of AV = 27 J as well as with points
0,1% C
J
where the level of impact energy has fallen to
100
0,4% C
half of the high level. It can clearly be seen
0,8% C
27
-150
-100
-50
0
Temperature
50
°C 100
© ISF 2002
br-er10-06.cdr
Influence of Mn, Ni, and C
on the Av-T-Curve
that mild steels have the lowest impact energy values together with the highest transition temperature. The development of finegrain structural steels resulted in a clear im-
Figure 10.6
provement of impact energy values and in
addition, the application of such steels could be extended to a considerably lower temperature range.
With the example of the
steels St E 355 and St E
690 it is clearly visible that
an increase of strength goes
mostly hand in hand with a
decrease of the impact energy
level.
provement
Another
showed
imthe
application of a thermomechanical
treatment
(con-
trolled rolling during heat
treatment). The application
of this treatment resulted in
an increase of strength and
Figure 10.7
10. Testing of Welded Joints
130
impact energy values together with a parallel saving of alloy elements. To make a comparison, the AV-T - curve of the cryogenic and high alloyed steel X8Ni9 was plotted onto the diagram. The material is tested under very high
P
test speed in the impact energy test, thus
C
growth and fracture mechanisms.
1,2h ± 0,25
there are no reliable findings about crack
0,55h ± 0,25
C
P
a
b
CT - specimen
L
h
1,25h ± 0,13
Figure 10.8 shows two commonly used
specimen height h = 2b ± 0,25
specimen width b
total crack length a = (0,50 ± 0,05)h
test load P
specimen shapes for a fracture mechanics
a
h
test to determine crack initiation and crack
growth. The lower figure to the right shows a
2,1h
2,1h
possibility how to observe a crack propaga-
b
S
tion in a compact tensile specimen. During
SENB -specimen
3PB
specimen width b
bearing distance S = 4h
sample height h = 2b ± 0,05
total crack length a = (0,50 ± 0,05)h
F,U
crack initiation
U
F
the test, a current I flows through the speci-
UE,aE
U
men, and the tension drop above the notch is
UO
measured.
V
V
br-er10-08.cdr
© ISF 2002
Fracture Mechanics Test
Sample Shape and Evaluation
As soon as a crack propagates through the
material, the current conveying cross section
Figure 10.8
decreases, resulting in an increased voltage
drop. Below to the left a measurement graph of such a test is shown. If the force F is plotted
across the widening V, the drawn curve does not indicate precisely the crack initiation.
F
Analogous to the stress-
F
strain diagram, a decrease
of force is caused by a reduction of
the stressed
h
cross-section. If the voltage
drop is plotted over the
d
force, then the start of
d
d1
2
crack initiation can be determined with suitable accuracy,
and
the
crack
br-er-10-09.cdr
Hardness Testing to Brinell and Vickers
Figure 10.9
propagation can be observed.
10. Testing of Welded Joints
131
Another typical characteristic of material behaviour is the hardness of the workpiece. Figure
10.9 shows hardness test methods to Brinell (standardised to DIN 50 351) and Vickers (DIN
50 133). When testing to Brinell, a steel ball is pressed with a known load to the surface of
the tested workpiece. The diameter of the resulting impression is measured and is a magnitude of hardness. The hardness value is calculated from test load, ball diameter, and diameter of rim of the impression (you find the formulas in the standards). The hardness
information contains in addition to the hardness magnitude the ball diameter in mm, applied
load in kp and time of influence of the test load in s. This information is not required for a ball
diameter of 10 mm, a test load of 3000 kp (29420 N), and a time of influence of 10 to 15 s.
This hardness test method may be used only
0,200 mm
6
2
7
10
3
6
7
7
0
8,9
reference
level for
measurement
10
3
10
specimen surface
6
130
30
0
hardness
scale
hardness
scale
100
6
4
5
3
8
130
30
0
specimen surface
0,200 mm
Instead of a ball, a diamond pyramid is
1
3
100
0
Hardness testing to Vickers is analogous.
This method is standardised to DIN 50133.
4 5
3
8
0,200 mm
Hardness Number).
3
1
0,200 mm
on soft materials up to 450 BHN (Brinell
8,9
reference
level for
measurement
7
10
pressed into the workpiece. The lengths of
the two diagonals of the impression are
Terms
Abbreviation
ball diameter = 1,5875 mm ( 1/16 inch)
-
cone angle = 120°
2
-
radius of curvature of cone tip = 0,200 mm
3
F0
test preload
4
F1
test load
lated from their average and the test load.
5
F
total test load = F0 + F1
6
t0
penetration depth in mm under test preload F0. This defines the reference level
for measurement of tb.
The impressions of the test body are always
7
t1
total penetrationn depth in mm under test load F1
8
tb
resulting penetration depth in mm, measured after release of F1 to F0
geometrically similar, so that the hardness
9
e
resulting penetration depth, expressed in units of 0,002 mm:
tb / 0,002
10
HRC
HRA
measured and the hardness value is calcu-
1
value is normally independent from the size
of the test load. In practice, there is a hard-
HRB
HRF
Rockwell hardness = 130 - e
br-er10-10.cdr
ness increase under a lower test load be-
© ISF 2002
Hardness Test to
Rockwell
cause of an increase of the elastic part of the
deformation.
Rockwell hardness = 100 - e
e =
Figure 10.10
Hardness testing to Vickers is almost universally applicable. It covers the entire range of materials (from 3 VHN for lead up to 1500 VHN for hard metal). In addition, a hardness test can
be carried out in the micro-range or with thin layers.
Figure 10.10 illustrates a hardness test to Rockwell. In DIN 50103 are various methods standardised which are based on the same principle.
10. Testing of Welded Joints
132
With this method, the penetration depth of a penetrator is measured.
At first, the penetrator is put on the workpiece by application of a pre-test load. The purpose
is to get a firm contact between workpiece and penetrator and to compensate for possible
play of the device.
Then the test load is applied in a shock-free way (at least four times the pre-force) and held
for a certain time. Afterwards it is released to reach minor load. The remaining penetration
depth is characteristic for the hardness. If the display instrument is suitably scaled, the hardness value can be read-out directly.
All hardness test methods to Rockwell use a ball (diameter 1.5875 mm, equiv. to 1/16 Inch)
or a diamond sphero-conical penetrator (cone angle 120°) as the penetrating body. There are
differences in size of pre- and test load, so different test methods are scaled for different
hardness ranges. The most commonly used scale methods are Rockwell B and C. The most
considerable advantage of these test methods compared with Vickers and Brinell are the low
time duration and a possible fully-automatic measurement value recognition. The disadvantage is the reduced accuracy in contrast to the other methods. Measured hardness numbers
are only comparable under identical conditions and with the same test method. A comparison
of hardness values which were determined
with different methods can only be carried out
for similar materials. A conversion of hardness
values of different methods can be carried out
piston
for steel and cast steel according to a table in
DIN 50150. A relation of hardness and tensile
strength is also given in that table.
All the hardness test methods described above
require a coupon which must be taken from the
reference bar
workpiece and whose hardness is then determined in a test machine. If a workpiece on-site
is to be tested, a dynamical hardness test
specimen
method will be applied. The advantage of these
methods is that measurements can be taken
br-er10-11.cdr
on completed constructions with handheld
© ISF 2002
Poldi - Hammer
Figure 10.11
10. Testing of Welded Joints
133
units in any position. Figure 10.11 illustrates a hardness test using a Poldi-Hammer. With this
(out of date) method, the measurement is carried out by a comparison of the workpiece
hardness with a calibration piece. For this purpose a calibration bar of exactly determined
hardness is inserted into the unit, which is held by a spring force play-free between a piston
and a penetrator (steel ball, 10 mm diameter). The unit is put on the workpiece to be tested.
By a hammerblow to the piston, the penetrator penetrates the workpiece and the calibration
pin simultaneously. The size of both impressions is measured and with the known hardness
of the calibration bar the hardness of the workpiece can be determined. However, there are
many sources of errors with this method which may influence the test result, e.g. an inclined
resting of the unit on the surface or a hammerblow which is not in line with the device axis.
The major source of errors is the measurement of the ball impression on the workpiece. On
one hand, the edge of the impression is often unsharp because of the great ball diameter, on
the other hand the measurement of the impression using magnifying glasses is subjected to
serious errors.
Figure 10.12 shows a modern measurement method which works with ultrasound and combines a high flexibility with easy handling and high accuracy. Here a test tip is pressed manually against a workpiece. If a defined test load is passed, a spring mechanism inside the test
tip is triggered and the measurement starts.
The measurement principle is based on a
Test force
measurement of damping characteristics in
5 kp
5.0
the steel. The measurement tip is excited to
kp
emit ultrasonic oscillations by a piezoelectric
4.0
crystal. The test tip (diamond pyramid) pene3.0
trates the workpiece under the test pressure
2.0
caused by the spring force. With increasing
Federweg
penetration depth the damping of the ultrasonic oscillation changes and consequently
the frequency. This change is measured by
the device. The damping of the ultrasonic os-
- little work on surface preparation of specimens (test force 5 kp)
- Data Logger for storage of several thousands of measurement points
- interfaces for connection of computers or printers
- for hardness testing on site in confined locations
br-er10-12.cdr
© ISF 2002
cillation depends directly on penetration depth
thus being a measure for material hardness.
The display can be calibrated for all commonly used measurement methods, a meas-
Figure 10.12
10. Testing of Welded Joints
134
urement is carried out quickly and easily. Measurements can also be carried out in confined
pulsation range
(compression)
Application
Dye penetrant method
σm = σa
σm > σ a
crack is free, surface
is clean
σm < σa
compression -
+ tension
Description
σm = 0
σ m < σa
σm = σa
σ m > σa
spaces. This measurement method is not yet standardised.
time
crack and surface
with penetrant
liquid
cleaned surface,
dye penetrant
liquid in crack
pulsation range
(tension)
alternating range
all materials with
surface cracks
surface with
developer shows
the crack by coloring
Wöhler line
Magnetic particle testing
II
A workpiece is placed
between the poles of
a magnet or solenoid.
Defective parts disturb
the power flux. Iron
particles are collected.
I
III
σD
Stress σ
failure line
Surface cracks and
cracks up to 4 mm
below surface.
However:
Only magnetizable
materials and only
for cracks perpendicular
to power lines
0
1
10
102
103
104
105
106
Fatigue strength (endurance) number lg N
107
I area of overload with material damage
II area of overload without material damage
III area of load below fatigue strength limit
br-er10-13.cdr
© ISF 2002
br-er10-14.cdr
© ISF 2002
Fatigue Strength Testing
Figure 10.13
Figure 10.14
To test a workpiece under oscillating stress, the fatigue test is standardised in DIN 50100.
Mostly a fatigue strength is determined by the Wöhler procedure. Here some specimens
(normally 6 to 10) are exposed to an oscillating stress and the number of endured oscillations
until rupture is determined (endurance number, number of cycles to failure). Depending on
where the specimen is to be stressed in the range of pulsating tensile stresses, alternating
stresses, or pulsating compressive stresses, the mean stress (or sub stress) of a specimen
group is kept constant and the stress amplitude (or upper stress) is varied from specimen to
specimen, Figure 10.13. In this way, the stress amplitude can be determined with a given
medium stress (prestress) which can persist for infinite time without damage (in the test: 107
times). Test results are presented in fatigue strength diagrams (see also DIN 50 100). As an
example the extended Wöhler diagram is shown in Figure 10.13. The upper line, the Wöhler
line, indicates after how many cycles the specimen ruptures under tension amplitude σa. The
10. Testing of Welded Joints
135
Application
Description
X-ray or isotope radiation penetrate
a workpiece. The thicker the workpiece, the weaker the radiation
reaching the underside.
W ire diameter
Mainly for defects with orientation
in radiation direction.
Tolerated
deviation
mm
3,2
2,5
2
1,6
1,25
1
0,8
0,63
0,5
0,4
0,32
0,25
0,2
0,16
0,125
0,1
¬
-
W ire number
mm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
± 0,03
± 0,02
± 0,01
± 0,005
°
Abbreviation
®
W ire number to
Table 1
FE 1/7
1 to 7
FE 6/12
FE 10/16
CU 1/7
¬ radiation source
-
¯
CU 10/16
AL 1/7
AL 6/12
workpiece
® film (displayed in distance from workpiece)
¯ defect in radiation direction; difficult to identify (flank lack of fusion)
° defect in radiation direction; easy to identify
br-er10-15.cdr
AL 10/16
© ISF 2002
50
6 to 12
50 or 25
10 to 16
50 or 25
1 to 7
CU 6/12
W ire length
mm
6 to 12
10 to 16
Material groups
to be tested
mild
steel
iron materials
copper
copper, zink, tin
and its alloys
aluminium
aluminium
and its alloys
50
50
50 or 25
1 to 7
50
6 to 12
50
10 to 16
W ire material
50 or 25
br-er10-16.cdr
© ISF 2002
Determination of Picture Quality
Number to DIN 54105
Non-Destructive Test Methods
Radiographic Testing
Figure 10.16
Figure 10.15
damage line indicates analogously, when a
Description
US-head generates high-frequency sound
waves, which are transferred via oil coupling
to the workpiece. Sound waves are reflected
on interfaces (echo).
Application
Mainly for defects with an orientation
transverse to sound input direction.
damage to the material starts in form of
cracks. Below this line, a material damage
does not occur.
Ã
Test
À
methods
described
above
require
specimens taken out of the workpiece and a
Á
partly very accurate sample preparation. A
testing of completed welded constructions is
Â
impossible, because this would require a deÄ
À sound head
Á oil coupling
 workpiece
à defect
Ä ultrasonic test device
Å radiation pulse
Æ defect echo
³ backwall echo
Å
Æ
³
br-er10-17.cdr
© ISF 2002
Non-Destructive Test Methods
Ultrasonic Testing II
Figure 10.17
struction of the workpiece. This is the reason
why various non-destructive test methods
were developed, which are not used to determine technological properties but test the
workpiece for defects. Figure 10.14 shows
10. Testing of Welded Joints
136
two methods to test a workpiece for surface defects.
Figure 10.15 illustrates the principle of radiographic testing which allows to identify also defects in the middle of a weld. The size of the minimum detectable defects depends greatly on
the intensity of radiation, which must be
adapted to the thickness of the workpiece to
be radiated. As the film with documented defects does not permit an estimation of the
plate thickness, a scale bar must be shown for
estimation of the defect size.
For that purpose, a plastic template is put on
the workpiece before radiation which contains
metal wires with different thickness and incorporated metallic marks, Figure 10.16. The size
of the thinnest recognisable wire indicates the
size of the smallest visible defect. Radiation
Figure 10.18
testing provides information about the defect
position in the plate plane, but not about the
position within the thickness depth. A clear
advantage is the good documentation ability
of defects.
Figure 10.18
An information about the
depth of the defect is provided by testing the workpiece with ultrasound. The
principle is shown in Figures
10.17
and
10.18
(principle of a sonar).
The
display
of
original
br-er10-19.cdr
pulse, backwall and defect
Ultrasonic Testing of Fillet Welds
echo is carried out with an
oscilloscope.
Figure 10.19
10. Testing of Welded Joints
137
This method provides not only a perpendicular sound test, but also inaccessible regions can
be tested with the use of so called angle testing heads, Figure 10.19.
Pores between 10 and 20 mm
depth provide an unbroken
echo sequence across the entire
display starting from 10mm. The
backwall echo sequence of 30
mm is not yet visible.
30
Wall thickness is below 40 mm.
The roughness provides smaller
and wider echos.
Echo sequence of 20 mm depth.
The backwall is completely
screened.
The perpendicular crack
penetrating the material
does not provide a display
because the reflecting surface
(tip of crack) is too small.
40
The oblique and rough defect
from 20 to 30 mm provides a
wide echo of 20 to 30 mm.
Starting with SKW 4, an unbroken echo sequence follows.
The inclination of the reflector
is recornised by a change of
the 1st echo when shifting the
test head.
The oblique backwall reflects
the soundwaves against the
crack. this is the reason why
an ‘impossible’ depth of
65 mm is displayed.
Echo sequence of 10 mm depth.
The reflector in 30 mm depth is
completely screened.
br-er10-20.cdr
© ISF 2002
br-er10-21.cdr
Defect Identification with Ultrasound
© ISF 2002
Defect Identification With Ultrasound
Figure .10.20
Figure 10.21
Figures 10.20 and 10.21
show
macro section
base material
schematically
the
display of various defects
50 µ
ferrite
+ perlite
coarse grain zone
bainite
on an oscilloscope. A correct interpretation of all the
signals requires great experience,
2,5 mm
fine grain zone
ferrite
+ perlite
fusion line
weld metal
cast structure
Steel: S355N
(T StE 355)
bainite
because
the
shape of the displayed signals is often not so clear.
br-er10-22.cdr
Metallographic Examination of a Weld
Figure 10.22 illustrates the
potential of metallographic
Figure 10.22
examination. Grinding and
10. Testing of Welded Joints
138
etching with an acid makes the microstructure
visible. The reason is that depending on
structure and orientation, the individual grains
react very differently to the acid attack thus
100
25
Fe
% Fe
% Cr
80
Cr
20
15
20
10
a complete survey about the weld and fusion
line, size of the HAZ, and sequence of solidification. Under adequate magnification, these
0
areas can still not be distinguished precisely,
10
8
Ni
however, an assessment of the developed
6
4
macrosection, i.e. without magnification, gives
60
40
% Ni
reflecting the light in a different way. The
5
microstructure is possible.
2
0
0
200
mm
100
0
An assessment of the distribution of alloy
100
Distance from fusion line
br-er10-23.cdr
elements across the welded joint can be car-
© ISF 2002
Micro-Analysis of the Transition Zone
Base Material - Strip Cladding
ried out by the electron beam micro-analysis.
An example of such an analysis is shown in
Figure 10.23
Figure 10.23. If a solid body is exposed to a
focused electron beam of high energy, its atoms are excited to radiate X-rays. There is a
simple relation between the wave length of this radiation and the atomic number of the
chemical elements. As the intensity of the radiation depends on the concentration of the elements, the chemical composition of the solid body can be concluded from a survey of the
emitted
X-ray
qualitatively
and
spectrum
quantita-
tively. A detection limit is
50
50
20 20
about 0.01 mass % with this
20 20
50
0
10
method. Microstructure areas of a minimum diameter
50
1. weld
2. weld
weld
of about 5 µm can be ana-
axis of
bending former
weld
Agents:
- electrolytic copper in the form of chips (min. 50 g/l test solution)
- 100 ml H2SO4 diluted with 1 l water and then
.
110 g CuSO 5 H2O are added
lysed. If the electron beam is
Test:
The specimens remain for 15 h in the boiling test solution.
Then the specimens are bent across a former up to an angle
of 90° and finally examined for grain failure under a
6 to 10 times magnification.
moved across the specimen
(or the specimen under the
br-er-10-24.cdr
beam), the element distribu-
Strauß - Test
tion along a line across the
Figure 10.24
axis of
bending former
10. Testing of Welded Joints
139
solid body can be determined. Figure 10.23 presents the distribution of Ni, Cr, and Fe in the
transition zone of an austenitic plating in a ferritic base metal. The upper part shows the related microsection which belongs to the analysed part. This microanalysis was carried out
along a straight line between two impressions of a Vickers hardness test. The impressions
are also used as a mark to identify precisely the area to be analysed.
The so called Strauß test is
standardised in DIN 50
12
914. it serves to determine
web
80
the resistance of a weld
measurement points
tack welds
against intergranular corro-
base plate
weld1
40
40
20
sion. Figure 10.24 shows
the specimen shape which
is normally used for that
a
a a
20
aa
a
a
12
weld2
120
80
aa
test. In addition, some debr-er-10-25.cdr
Test of Crack Susceptibility of Welding
Filler Materials to DIN 50129
tails of the test method are
explained.
Figure 10.25
Figure 10.25 presents a specimen shape for testing the crack susceptibility of welding consumables. For this test, weld number 1 is welded first. The 2. weld is welded not later than 20
s in reversed direction after completion of the first weld. Throat thickness of weld 2 must be
20% below of weld 1. After
cooling down, the beads
are examined for cracks. If
tensioning bolt
hexagon nut
min. M12 DIN 934
guidance plates
a
tensioning plate
specimen
base body
cracks are found in weld 1,
the test is void. If weld 1 is
free from cracks, weld 2 is
examined for crack with
magnifying glasses. Then
weld 1 is machined off and
weld 2 is cracked by bend-
br-er-10-26.cdr
Tensioning Specimen for Crack
Susceptibility Test
Figure 10.26
ing the weld from the root.
Test results record any
10. Testing of Welded Joints
140
surface and root cracks together with information about position, orientation, number, and
length. The welding consumable is regarded as 'non-crack-susceptible' if the welds of this
test are free from cracks.
Figure 10.26 presents two proposals for self-stressing specimens for plate tests regarding
their hot crack tendency. Such tests are not yet standardised to DIN.
thermo couple
electrode
cross-section
groove shape
60°
60°
welding direction
weld
metal
support plate
Wd./2
H
Wd.
2
implant
Hc
Wd./2
2
load
temperature in °C
specimen shape
load in N
Tmax
start
end crater
150
crack coefficient
C=
c
x 100 (in %)
800
500
1
2 3
4 5
sections
60
anchor weld
80
test weld
150
100
60
anchor weld
br-er10-27.cdr
t8/5
© ISF 2002
rupture time
br-er10-28.cdr
Tekken Test
Figure 10.27
time in s
© ISF 2002
Implant Test
Figure 10.28
There are various tests to examine a cold crack tendency of welded joints. The most important ones are the self-stressing Tekken test and the Implant test where the stress comes from
an external source.
In the Tekken test which is standardised in Japan, two plates are coupled with anchor joints
at the ends as a step in joint preparation see Figure 10.27. Then a test bead is welded along
the centre line. After storing the specimen for 48 hours, it is examined for surface cracks. For
a more precise examination, various transverse sections are planned. The value to be determined is the minimum working temperature at which cracks no longer occur. The specimen shape simulates the conditions during welding of a root pass.
10. Testing of Welded Joints
141
The most commonly used cold crack test is the Implant test, Figure 10.28. A cylindrical body
(Implant) is inserted into the bore hole of a support plate and fixed by a surface bead. After
the bead has cooled down to 150°C the implant is exposed to a constant load. The time is
measured until a rupture or a crack occurs (depending on test criterion 'rupture' or 'crack').
Varying the load provides the possibility to determine the stress which can be born for 16
hours without appearance of a crack or rupture. If a stress is specified to be of the size of the
yield point as a requirement, a preheat temperature can be determined by varying the working temperature to the point at which cracks no longer appear.
As explained in chapter 'cold cracks' the hydrogen content plays an important role for cold
crack development. Figure 10.29 shows results of trials where the cold crack behaviour was
examined using the Tekken and Implant test. Variables of these tests were hydrogen content
of the weld metal and preheat temperature. The variation of the hydrogen content of the weld
metal was carried out by different exposure to humidity (or rebaking) of the used stick electrodes. Based on the hydrogen content, the preheat temperature was increased test by test.
Consequently, the curves of Figure 10.29 represent the limit curves for the related test.
Specimens above these
heat input: 12 kJ/cm
basic coated stick electrode
plate and support plate thickness: 38 mm
°C
cracks, below these curves
°C
Implant-Test
150
Tekken-Test
100
50
cracks are present. Evi-
150
Rcr = Rp0,2 = 358 N/mm²
Preheat temperature
Preheat temperature
curves remain free from
fractured
starting cracks
crack-free
20
dent for both graphs is that
with
100
temperature
50
starting cracks
crack-free
20
0
10
20
30
ml/
40
100 g
increased
0
10
Diffusible hydrogen content
br-er-10-29.cdr
Test Result Comparison of
Implant and Tekken Test
20
30
ml/
40
100 g
preheat
considerably
higher hydrogen contents
are tolerated without any
crack
development
be-
cause of the much better
hydrogen effusion.
Figure 10.29
If both graphs are compared it becomes obvious that the tests produce slightly different findings, i.e. with identical
hydrogen content, the determined preheat temperatures required for the avoidance of cracking, differ by about 20°C.
10. Testing of Welded Joints
142
Figure 10.30 illustrates a method to measure the diffusible hydrogen content in welds which
is standardised in DIN 8572. Figure a) shows the burette filled with mercury before a specimen is inserted. The coupons are inserted into the opened burette and drawn with a magnet
through the mercury to the capillary side (density of steel is lower than that of mercury, coupons surface). Then the burette is closed and evacuated. The hydrogen, which effuses of the
coupons but does not diffuse through the mercury, collects in the capillary. The samples remain in the evacuated burette 72 hours for degassing. To determine the hydrogen volume
the burette is ventilated and the coupons are removed from the capillary side. The volume of
the effused hydrogen can be read out from the capillary; the height difference of the two mercury menisci, the air pressure, and the temperature
provide the data to calculate
the
norm
volume
to pump
hydrogen
under reduced pressure
under
VT
air pressure B
evacuated
standard
conditions.
This
capillary side
volume and the coupons
M
meniskus1
weight are used to calculate,
as measured value, the hy-
meniskus2
mercury
coupons
drogen volume in ml/100 g
weld metal. This is the most
a) starting condition
commonly used method to
determine
the
c) ventilated after degassing
Burettes for Determination of
Diffusible Hydrogen Content
hydrogen
content in welded joints.
b) during degassing
br-er-10-30.cdr
Figure 10.30