The Mechanisms of Formation of
W e l d Defects in High-Frequency
Electric Resistance W e l d i n g
The two kinds of defects associated with previously defined welding
phenomena are cold welds comprised of oxide colonies and
penetrators consisting of slag inclusions produced when molten slag
is drawn into a gapped zone
BY H. HAGA, K. AOKI A N D T. SATO
ABSTRACT. In order to clarify the
mechanisms by which weld defects
are formed, welding phenomena were
observed using high speed cameras,
and weld defects produced in relation
to corresponding phenomena were
examined. It is found that the possible
occurrence of certain kinds of weld
defects is determined by the types of
phenomena
encountered
during
welding: A possible weld defect
accompanied by the 1st type phenomenon is a cold weld; a penetrator is
accompanied by 2nd and 3rd types of
phenomena.
The cold weld is composed of a
colony of oxides which are formed on
the edge surfaces and remain because
they are not squeezed out from
between them. The penetrator by
nature is a slag inclusion produced
when molten slag is drawn into a
gapped zone by the action of surface
tension and capillary effects in a
returning process of the molten
bead.
Introduction
Weld defects of ERW are perpendicular to the plate surface and have
planar shapes caused by upsetting in a
process of welding. It is reported that
these defects have occasionally caused
a burst accident in field tests of ERW
pipelines. Because of their shape and
orientation, it is very hard to find them
w i t h NDT, and many efforts have been
made to develop an effective NDT
method for locating such defects.
Very few investigations, however,
104-s I JUNE 1981
Table 1—Chemical Composition of the
Materials Used, Wt-%
C
0.11
Si
Mn
Nb
0.25
0.70
0.013
P
S
AI
0.01
0.01
0.015
have been reported in regard to the
cause of the weld defects in ERW; the
defects have been simply conceived as
certain oxides that form on the surface
of strip edges and remain without
being squeezed out between them.
Macro- and microscopic observations of the weld defects showed that
there are t w o kinds of the defects: a
cold w e l d and a penetrator. In most
cases the former is found over a wide
range along the weld line and the
latter only intermittently. It is difficult
to understand by the past simple
explanation why t w o different kinds of
defect are generated.
The welding phenomena encountered in ERW, as mentioned in a previous report, 1 can be classified into
three types according to the periodicity of the welding process and the edge
profile between the V-convergency
Paper presented at the AWS 60th
Meeting held in Detroit, Michigan,
April 2-6, 1979.
H. HAGA, K. AOKI and
ciated with the Products
opment Laboratories of
Corporation, Sagamihara
Annual
during
T. SATO are assoResearch & Develthe Nippon Steel
Kanagawa, lapan.
point and the weld point. It is inferred
that the type of welding phenomena
might determine which kind of the
defects will be generated.
Experimental results given below
show that the inference is almost true.
Moreover, observations through high
speed photography help identify a
phenomenon accompanying the formation of a penetrator. This gives a
definite base for correlating the kind
of defects to the type of welding phenomena.
Experimental Procedure
Materials
Materials welded in previous work 1
were examined. Chemical compositions of the materials are given in
Table 1.
Identification of Actual Weld with
Photographed One
In order to find welding processes
accompanying weld defects, it is
essential to find how each point along
the weld was joined. Each external
bead on the weld was identified w i t h
the one observed in the high speed
photograph as shown in Fig. 7. This
method gave an accuracy w i t h i n 1 mm
(0.04 in.) to the corresponding position.
Examination of Weld Defects
Welded pipes were cut into semicylindrical shells and flattened. From
the flattened shell dynamic tear (DT)
test pieces were prepared. External
(a)
t e s t 17
(b>
t e s t 15
r
:c)
test 1
1'0
A
2'0
'
3'0
4>0
'
5'0
Fig. 1-DT fracture surfaces of the weld representative
are contoured
beads o n t h e test p i e c e s w e r e p h o t o graphed together w i t h measurements
before they were removed. Charpy
i m p a c t test p i e c e s w e r e also p r e p a r e d .
T h e D T test p i e c e s w e r e s u b j e c t e d t o
brittle fracture and Charpy testing.
W e l d defects were examined on
f r a c t u r e surfaces o f D T a n d C h a r p y test
p i e c e s by m i c r o s c o p e a n d s c a n n i n g
e l e c t r o n m i c r o s c o p e ( S E M ) , a n d classified according to their m o r p h o l o g y .
Figure 1 s h o w s D T f r a c t u r e surfaces
of t h e w e l d r e p r e s e n t a t i v e o f t h e 1st,
2nd and 3rd types of w e l d i n g p h e n o m e n a . W e l d d e f e c t s are c o n t o u r e d in t h e
p h o t o g r a p h s . E l o n g a t e d b l a c k parts
o b s e r v e d near edges o f t h e f r a c t u r e
surface are cracks f o r m e d o n e x t e r n a l
or
internal
bead, and
not
weld
defects.
In t h e case o f t h e 1st t y p e p h e n o m e n o n , no w e l d defect was observed o n
a b r i t t l e f r a c t u r e surface as s h o w n in
Fig. 1(a). Large w e l d d e f e c t s , h o w e v e r ,
are r e v e a l e d o n a d u c t i l e f r a c t u r e surf a c e of t h e C h a r p y test p i e c e p r e p a r e d
as s h o w n i n Fig. 2 ( a ) . T h e w e l d d e f e c t
o f Fig. 2(a) is c o n s i d e r e d t o b e l o n g t o a
cold weld.
T h e w e l d d e f e c t s o f Figs. 1 ( b ) , 1 ( c ) ,
2 ( b ) a n d 2 ( c ) are c o n s i d e r e d t o b e l o n g
to a penetrator. The penetrators were
q u i t e s i m i l a r in a p p e a r a n c e b o t h o n
the brittle fracture surface and o n t h e
ductile one.
T h e results o f e x a m i n a t i o n o f w e l d
d e f e c t s are s u m m a r i z e d in T a b l e 2. In
this e x p e r i m e n t t h e 1st t y p e p h e n o m e n o n always a c c o m p a n i e d a c o l d w e l d
w h i c h is f o u n d o v e r a w i d e range
along the w e l d line. The c o l d w e l d
£fc
mm
7'0
&Q A
.8?0
'
POO
weld
defects
? jr.-:?--L • • VA
A.A-A^.-AA
(a)
Results
Correlation between the Kind of Weld
Defects and the Type of W e l d i n g
Phenomena
'
of (a) the 1st, (b) the 2nd and (c) the 3rd type phenomena;
(b)
(c)
Fig. 2—Charpy fracture surfaces of the weld representative ot (a) the 1st, (b) the 2nd and (c)
the 3rd type phenomena; weld defects are contoured. Test pieces were prepared from the
same tests as those in Fig. 7
h a p p e n e d t o b e a c c o m p a n i e d also b y
the 2nd type p h e n o m e n o n .
Penetrators w e r e a c c o m p a n i e d by
2 n d and 3rd t y p e p h e n o m e n a and never by t h e 1st t y p e . P e n e t r a t o r s w e r e
d i s t r i b u t e d n o t u n i f o r m l y , b u t in f r a g ments along the w e l d line. A penetrat o r g e n e r a l l y e x t e n d s in t h e t h r o u g h thickness direction.
In t h e 2 n d t y p e p h e n o m e n o n , t h e
l e n g t h o f m o s t p e n e t r a t o r s is less t h a n
2 m m (0.08 in.) a n d t h e i r f r e q u e n c y is
g e n e r a l l y l o w . In t h e 3 r d t y p e p h e n o m e n o n , p e n e t r a t o r s of v a r i o u s sizes w e r e
f o u n d . In tests 1 , 2, 4 a n d 5, in w h i c h
the gapped zone had developed long
e n o u g h , large p e n e t r a t o r s t h a t w e r e
l o n g e r t h a n 5 m m (0.20 in.) w e r e o f t e n
f o u n d . The frequency of the penetrat o r is h i g h e r i n t h e 3 r d t y p e t h a n in t h e
2nd type.
Fractographic Observation of Weld
Defects
Weld
defects
assume
different
shapes, m i c r o s c o p i c a l l y as w e l l as
macroscopically, according to the type
of w e l d i n g
phenomena.
Figure 3
shows a SEM image of cold w e l d on a
Table 2—Results of the Examination of the Type and Frequency of Weld Defects
Defect
Frequency, %*
Cold w e l d
10 :£ F
1 S F < 10
F< 1
Penetrator
5 < F < 10
1 S F< 5
F< 1
Test
11,17
14
15
1,2,3,4,5,6
7,9,10,13,16
8,12,15
*Percentage of w e l d defect area to total w e l d area.
WELDING RESEARCH SUPPLEMENT 1105-s
SEM
Mn
Fe
.
•;-y:?yyy?y::
V
.'AAA::
ifs m
0.
t
A
o
>
UJ
a
o
<
UJ
50 M
t/s
UJ
cn
-*-',.„
iSM ,
Si
0
Al
Fig. 3—SEM and characteristic X-ray images ol cold weld on a tracture surface
fracture surface. The cold weld consists of a colony of numerous microdimples, arranged on a flat plane, most
of them being less than 1/j in diameter.
Inclusions are found in most microdimples. XMA results show that comparatively large inclusions are rich in
M n and Si, but most of them are too
small to analyze by XMA.
Micro-inclusions in the cold weld
are found on the cross section of the
weld as shown in Fig. 4. Their size is
extremely small compared with the
diameter of the grains that have grown
across the weld. In this case a cleavage
fracture is expected to propagate
legardless of the presence of cold
weld. A ductile fracture, on the other
hand, propagates by creating a dimple
around an inclusion; consequently, it
is considered that ductile fracture can
better reveal the cold weld.
Figure 5(a) shows a SEM image of
the penetrator generally accompanied
by the 2nd type phenomenon. D i m ples and inclusions associated with it
are much larger than those associated
with cold weld. Brittle fracture is
thought to propagate the tearing off of
inclusions from the base metal. XMA
results show that the inclusions are
106-sl JUNE 1981
richer in M n , Si, Al and O and poorer
in Fe than is the base metal.
Figures 5(b) and (c) show other
types of the penetrator. They are
accompanied ordinarily by the 3rd
type phenomenon and occasionally by
the 2nd type. Inclusions associated
w i t h these penetrators are larger than
those found in Fig. 5(a). XMA results
show that the inclusions are also richer
in M n , Si, Al and O, and poorer in Fe,
than the base metal.
••'UAJ-' A,v-'
J
^ A
weld
Fig. 4—Optical micrograph oi cross section
of the cold weld region
In the 2nd and the 3rd type phenomena, which did not accompany the
cold weld, such extrusions were not
observed between the V-convergency
point and the weld point. The extrusions disappeared when molten beads
were produced before both the edges
met.
It is considered that the extrusions
consist mostly of oxides. The reason
for this is that it should not be possible
to form such extrusions against strong
electromagnetic pressure if the extrusions consist of metal. Thus, the cold
weld is a colony of oxides formed on
the edge surfaces that remain because
they were not squeezed out from
between them; this has been conceived to be a mechanism of producing all weld defects in ERW.2-3 If a
sufficient amount of molten metal is
removed from the edge surfaces by the
electromagnetic
pressure,
oxides
formed on the surfaces are also
removed and a cold weld will not be
produced. But such a condition leads
to the 2nd or the 3rd type phenomenon in this experiment.*
Penetrators
M e c h a n i s m of Formation of W e l d
Defects
Cold Weld
A cold weld occurs uniformly over a
wide range along the weld line, and
any accidental phenomena are not
considered to account for the cold
weld. Figure 6 shows the phenomenon
(of the 1st type) which accompanied
the cold weld. A number of extrusions
are observed to be formed on both the
edge surfaces approaching to the Vconvergency point. These extrusions
were sandwiched between the edges
when they were welded.
A penetrator is accompanied by 2nd
and 3rd type phenomena. Particularly
with the 3rd type, a large amount of
molten metal is removed, so a penetrator is not considered an oxide residual
but a defect different from the cold
*According to the theory given in the 1st
report, the 1st type phenomenon can be
realized depending on the welding rate and
V-apex angle even if sufficient molten metal is removed. Therefore, the 1st type phenomenon is not always considered to
accompany a cold weld. Inversely, the 2nd
type phenomenon will accompany the cold
weld ii sufficient molten metal is not
removed.
SEM
Fe
Mn
Si
Al
z
UJ
Q.
o
>
m&
UJ
o
I
o
a:
<
LJ
Fig. 5-SEM and characteristic X-ray images of penetrator accompanied by (a) the 2nd and (b), (c) the 3rd type phenome
weld; the cause producing the penetrator must be found in a certain process peculiar to the 2nd and 3rd
types.
In the 3rd type phenomenon, a gapped zone develops, removing molten
metal out of the gap. The removed
molten metal takes the shape of a
raindrop resulting from a balance
between electromagnetic force and
surface tension of molten metal.
When both the edges contact at the
V-convergency point in stage 3 mentioned previously, 1 all of the welding
current flows through the contacted
V-convergency point. Naturally, the
electromagnetic force suddenly vanishes in the gapped zone. Only the
surface tension acts on the molten
bead, and the balance of force is lost.
The molten bead then is expected to
begin to return to the gapped zone by
the action of surface tension.
High speed photography revealed
that the molten bead in fact returns
and fills the gapped zone as shown
typically in Fig. 7. Furthermore, it was
made clear that the penetrator is produced in such a zone as that filled by
the molten beads.
In the 3rd type phenomenon, molten beads fill most of the weld. Only a
small part of the weld is filled by a
molten metal bridge produced at the
V-convergency point in stage 3.1
In the 2nd type of phenomenon,
molten metal bridges fill most of the
weld. But occasionally the returning of
molten beads also occurs. A penetrator
was found in such a place.
In the 1st type phenomenon molten
beads cannot return, because the w e l d
point coincides with the V-convergency point and the electromagnetic force
does not vanish until both the edges
are welded; this is the reason why 1st
type phenomena are not accompanied
by a penetrator. Although a penetrator
may be found in the weld filled with
molten metal, observation showed
that the reverse is not true—that is, the
filling of the gap does not always
accompany a penetrator. Some accidental process accompanied by the
filling process is considered to be the
cause of producing the penetrator.
Surface Tension Acting on Molten Bead
There remains a question to be
answered: Is it not surface tension but
gravity that forces the molten bead to
return to the gapped zone? Gravity,
however, is not considered to take any
part in the returning process. This is
Fig. 6-Phenomenon accompanying cold weld
W E L D I N G RESEARCH S U P P L E M E N T 1107-s
time
(msec)
I-
zUJ
s
Q.
o
—1
UJ
0
2
>
UJ
Q
X
ce
1
<
UJ
</>
Ui
cc
10
12
Fig. 7—High speed photographs of molten bead action: (a)
returning process of molten beads observed in test 1,
showing: V-the V-convergency point, M-the molten metal
bridge and R-returning bead; A and B of (a) correspond to
the A and B of (b), external beads of DT test piece, and to
those of (c), DT fracture surface, in which the penetrators
are contoured
14
(a)
because the process finishes within 10
msec at most, and the gravity-induced
falling distance during that time is 0.5
mm (0.02 in.), which is much smaller
than the plate thickness.
It is possible to estimate the time
necessary to fill the gapped zone by
capillary effect, if viscosity and internal
pressure of the molten bead can be
neglected. Let the surface tension be
denoted by y and the angle of contact
be 0 deg; then the force (per unit
length) which draws molten metal is
2y. The equation of motion for the
filling process is expressed as follows:
Molten bead
2y
dt
(pbss)
2y
•
pb
(D
where p is the density of molten metal,
b the w i d t h of the gapped zone and s
the height (or the depth) of drawn
molten metal, as illustrated in Fig. 8.
Equation (1) reduces to:
1
d2
2Y = - P b - ( S 2 )
(2)
Solving equation (2) under the initial condition s = S and s = 0 at t = 0,
let S approach to 0. Then the solution
is given by:
The values of v and p are respectively about 1500 d y n / c m and 7.3 g / c m 3
for molten steel. W h e n b is 0.2 mm.:
s(mm) = 1.4t(msec)
This calculation shows that molten
metal fills the gapped zone throughthickness w i t h i n few milliseconds.
As is evident from the above, internal as well as external molten beads
should be able to return and fill the
gapped zone. In fact, occasionally it is
possible to observe the trace of return-
Table 3—Chemical Compositi on of Penetrator Analyzed by XMA
Penetrator, wt-%
Base Metal
Test
-Edge s u r f a c e
15
1
1
3
3
Yokoyama, et a/.4
Fig. 8—Cross-sectional view of the returning
process of molten bead
108-s I JUNE 1981
t
Mn
Si
0.7
0.25
1.4
1.5
1.1
0.17
0.41
0.06
MnO
SiO,
FeO
ALO,
46
44
44
41
42
43
41
42
40
40
7
4
5
5
4
4
3
3
4
3
55
51
82
33
41
3
6
4
13
6
4
2
SIO,
Fig. 9—End portion of weld showing the trace of returning of
internal bead
Fig. 10 (right)—Ternary equilibrium system FeO-MnO-Si02
ing of the internal bead as shown in
Fig. 9.
Metallurgical Considerations in Penetrator
Formation
The molten bead contains the oxide
removed from the edge surface. New
oxide is further produced on the surface of the molten bead exposed to
the air. At the interface between the
oxide layer and the molten bead, the
following reactions should occur:
(FeO) + [ M n ] ^ ± [Fe] + ( M n O )
2(FeO) + [Si] ^ ± 2[Fe] + (Si0 2 )
where [ ] is the concentration in the
molten iron and ( ) the free oxide
content of the slag. A part of M n O ,
Si0 2 and FeO is free or forms silicate,
and exists as a liquid phase. The rest
exist as a solid phase. The proportion
of free oxides depends on the solubility of the oxides.
In order to examine w h i c h phase of
slag becomes a penetrator, the chemical composition of the penetrator is
analyzed quantitatively by XMA. Results of the present analysis are given
in Table 3 together w i t h results of
available past analyses.4 In this experiment, the contents of M n O , S i 0 2 and
FeO in penetrators are 35 ~ 50%,
35 ~ 50% and 3 ~ 10%, respectively,
regardless of the type of welding phenomena. A ternary phase diagram (Fig.
10) shows this component of M n O Si0 2 -FeO exists as a liquid phase at
1500°C (2732°F). Other results also
show that most of the penetrators exist
as a liquid phase at 1500°C (2732°F).
The temperature of the molten bead
cannot exceed the melting point of
FeO
MnO
steel too far; this is because most of
welding current flows along the edge
surfaces due to a high frequencyinduced proximity effect. Therefore, it
is considered that a part of the molten
slag on the molten bead changes into a
penetrator. The molten slag covers the
surface of the molten bead, because
the density of the former is about half
that of the latter.
It is concluded that the penetrators
are produced when molten slag is drawn
into the gapped zone in the returning
process of the molten bead.
convergency point and the electromagnetic force vanishes in the gapped
zone.
5. It is the surface tension and the
resultant capillary effect that force the
molten beads to return into the gapped zone. The internal bead as well as
the external one is able to return.
6. The penetrator by nature is a slag
inclusion that is produced when molten slag is drawn into the gapped zone
in the returning process of the molten
beads.
Ac/cnow/ec/gmen(
Conclusions
1. There are two kinds of weld
defects in ERW: the cold weld and the
penetrator.
2. The type of welding phenomena
determine which kinds of weld defects
are produced. A possible weld defect
accompanied by the 1st type phenomenon is the cold weld. The 2nd type
phenomena accompanies a small penetrator, or the cold weld if sufficient
molten metal is not removed from the
edge surfaces. The 3rd type phenomenon frequently accompanies the penetrator.
3. The cold weld is composed of a
colony of oxides which are formed on
the edge surfaces and remain because
they are not squeezed out from
between them.
4. In 2nd and 3rd type phenomena,
molten beads, which have been
removed from the edge surfaces,
return and fill the gapped zone when
both the edges contact at the V-
The authors wish to express their
appreciation to Dr. T. Ikeno, the director, and Dr. S. Kanazawa, the deputy
director of the laboratories, for their
proposal of these successive studies,
and to Mr. H. Imai, Dr. H. Mimura and
Mr. M. lino for their helpful discussions. The contributions of Mr. K.
Sakurai and Mr. M. Yamada are also
appreciated.
References
1. Haga, H., Aoki, K„ and Sato, T. 1980.
Welding phenomena and welding mechanisms in high-frequency electric resistance
welding—1st report. Welding lournal 59(7):
pp. 208-s to 212-s.
2. Martin,
D.C.
High-frequency
resistance welding. Welding Research
Council Bulletin 160.
3. Hasebe, S., Kyogoku, T., Takahashi, S.,
Yamura, T. and Okazawa, T. 1972. Development of ERW high-test line pipe for arctic
service. Sumitomo Metals 24(2): pp. 67-90.
4. Yokoyama, E., Yamagata, M., Kanou, T.
and Watanabe, S. 1977. ). Iron & Steel Inst.
lapan 63(11): p. S650.
WELDING RESEARCH SUPPLEMENT 1109-s
O
tx
<
Ul
CO
Ul
cc