Академический Документы
Профессиональный Документы
Культура Документы
Noel F Herbst
6 JOINT DESIGN 10
C
two. This will most usually be stresses could be sufficient to induce a crack
one of the brazing alloys. The either during cooling, after welding or in
melting ranges of some of the service
Rigid Constraints common brazing alloys are given The coefficients of thermal expansion for a
in Table 2. This process is known number of common materials are shown in
as buttering and is a common so- Table 3.
0.2% C mild steel S30400
lution for a lot of dissimilar metal Differential expansion can also produce a
welding problems, see Section problem during service. The following ex-
0.05m 0.05m 5.2. ample illustrates this:
· Metal A: S30400 stainless steel - expan-
5.2Expansion sion coefficient = 20.0 m m/m.°C
5.2.1 Fusion welds · Metal B: 0.2% carbon steel - expansion
coefficient = 13.4 m m/m.°C
Differential thermal expansion If an assembly containing these two materi-
over a dissimilar metal weld can als, Figure 3 is heated, the before and after
Figure 3 Expansion stresses in a welded joint introduce stresses additional to conditions would be:
those normally accompanying
welding. It is possible that these
NOTEBOOK
Completed weld
Thermal conductivity variations in the components of a dissimilar weld can give
problems with over-heating one component and/or under-heating the other.
Directing the arc to the lower conductivity component may assist to minimise this
Figure 4 Buttering with an intermediate expansion co- problem.
efficient alloy
1 To calculate expansion: L1 = L0(1 + at) where a is the coefficient of thermal expansion and t is the temperature rise. Assuming an
average temperature rise over the 5 cm of 500°C, L1=5(1+20.0*10-6*500)) = 5.05 cm
2 American Welding Society Welding Handbook 7th Ed Vol 4, Chapter 12 Dissimilar Metal Welding p523 (1982)
3 Avery R E Pay attention to dissimilar welds - Guidelines for welding dissimilar metals Chemical Engineering Progress May 1991.
Reprinted as NiDI publication 14018
NOTEBOOK whole of a dissimilar metal joint containing weld pool and the heat affected zone of the
Pre- or post-heating is often necessary in that alloy. carbon steel.
hardenable steels or alloys requiring an ageing Pre-heating is frequently important for As mentioned above, the effect a necessary
heat treatment. If one component of a weld re- higher carbon and/or restrained plain car- treatment on one side of the weld would have
quires either pre- or post-heating then the bon steels to prevent post-weld cracking. on the other side - or the weld pool - must
whole weld will probably have to have the same This will not present a serious problem with always be considered.
treatment. This could usually be expected to most dissimilar metal joints although in
generate considerable difficulties.
some cases where pre-assembly or jigging 5.6 Weld pool properties
is required, there may be some handling
difficulties. 5.6.1 Metal mixing
Post-weld heating is not as simple. It is Metal mixing is essentially a mechanical pro-
example since alloying with copper in- conceivable for example that a carbon steel cess and for any mixing to occur, the metals
creases its thermal expansion but iron, chro- welded to a UNS S30400 stainless steel must be wetted by the filler metal. This could
mium and molybdenum will reduce the may accentuate the possibility of sensitisa- require specialised fluxes.
expansion coefficient tion corrosion due to the combination of The normal considerations in mixing that
5.2.2 Brazing welding heat input plus the post weld heat- apply to all welding operations will also ap-
ing. Sensitisation is the decrease in aqueous ply to dissimilar metal joints.
When dissimilar metals with differing ex- corrosion resistance due to carbide precipi- Six zones are usually identified, Figure 54:
pansion coefficients are brazed, the clear- tation. · Composite zone where there has been
ance required for correct capillary action Heat treatment can be considered as a post complete mixing
during brazing must be calculated. For ex- weld heating operation.. If one side is to be · Unmixed zone where the parent plate has
ample if a tube with a high thermal expan- heat treated by, say ageing, then the effect melted but not mixed with the metal of
sion is a press fit at room temperature on the other side must be considered, eg two the composite zone. This zone can give
around another with low thermal expan- different age hardening alloys may have phases that might not be present in the
sion, it is probable that the clearance at the different ageing treatments. Clearly other overall structure and that could markedly
brazing temperature will be too much to types of heat treatment could cause con- alter the weld structure.
permit the correct capillary action. If the re- cern.
verse arrangement of tubing is used, the · Fusion line
clearance will be too small. · Partially melted zone: This can give cor-
5.5 Choice of Welding rosion problems due to dendritic solidifi-
If two solid components of differing ex- Process
pansion coefficients are being brazed, the cation on cooling but also can have
brazing alloy should have an intermediate The main points that must be considered penetration by the weld metal into the
brazing coefficient. when selecting the basic process for com- parent plate - ie liquid metal corrosion.
pleting a dissimilar metal weld is a need for · Heat affected zone. The normal area of
5.3 Thermal conductivity precision location of the arc to permit dif- the parent plate where heat from the weld
ferential heat transfer between either side of can affect the parent plate structure
The effect of thermal conductivity variation
is similar to both melting point and thermal
the weld. · Unaffected base material
Other factors relating to pre-heat, It is probably the composite zone and the
expansion problems. The problems arise
post-weld heat treatment, shielding gases partially mixed zone that can give unex-
when one half of a joint has a markedly dif-
etc depend on the most sensitive side of the pected results in a dissimilar metal weld. The
ferent coefficient of thermal conductivity
weld, eg in welding a hardenable carbon reasons for this are essentially related to the
compared to the other. Directing the weld-
steel to an austenitic stainless steel, a effect of mixing on the phases that will be
ing heat source can qualitatively allow for
pre-heat must be given to ensure there is present, see Section 5.6.3
this, preheating the high conductivity
control over martensite formation in the
metal can also assist this.
Thermal conductivity changes with tem-
perature. A tabulation of some metallic
thermal conductivities with the applicable
temperature range are given in Table 4 . FUSION LINE
It is interesting to note that conductivity Parent 1: 10%A
increases with increasing temperature for WELD POOL PARTIALLY Parent 2: 20%A
MELTED ZONE
some metals, eg UNS S30400, but de- (Composite Zone)
creases with others, eg carbon and low al-
Filler: 30%A
loy steels.
Components where distortion is critical Parent 1 Parent 2
may require procedures to counteract the
effect of a thermal conductivity that could
cause problems. This may require heat in- MELTED BUT
Weld Pool:
put on some occasions - or extraction on UNMIXED ZONE Parent 1: 35%
others. HEAT AFFECTED
ZONE
Parent 2: 40%
5.4 Pre- and post-heating Filler: 25%
If pre-heating or post-heating is required
Figure 6 Weld pool dilution calculation
Figure 5 Zones in a welded deposit
on one half of a joint for metallurgical rea-
sons, this must also be the case for the
4 Taylor JS The fusion welding of dissimilar metals The Welding Technology of Stainless Steels WTIA Seminar Melbourne 1995
NOTEBOOK als. There are a number of factors that will Apart from their widely differing
The composition of the weld metal in a dissimi- affect these proportions: melting points, each of these metals is
lar metal weld can be approximately calculated · Thin materials: The low heat input re- virtually insoluble in the other. In the
from the lever rule. quired to melt for thin materials to- molten state they react to form brittle
This process requires an estimate to be made of gether with the low cross sectional area intermetallic phases x,qh etc. These
the amount of each constituent that ends up in to conduct heat away will be expected will give the weld unsatisfactory
the weld pool. It is usually assumed that there to generate a higher proportion of these properties.
are no losses from oxidation during the weld. in the weld pool. A further example of the third type, per-
There are also some structural factors that can · Location relative to the weld face: The haps of more industrial significance, is the
closer the weld run is to the parent iron-copper system, Figure 10: Like iron-
affect dilution: component thickness, location
metal face, the greater will be the con- aluminium, iron has very limited solubility
relative to the weld face, weld run placement
tamination from the parent. Root runs for copper so that the two metals virtually
relative to the previous run, factors that alter
will have the most contamination. form a mechanical mixture. Unlike
penetration.
· Weld run placement: Placing a second aluminium-iron the phases formed are more
run on top, rather than between previ- ductile, but do have the disadvantage that
ous runs should produce less contami- there is the likelihood of corrosion in the
5.6.2 Dilution calculation mixture due to the galvanic effect between
nation from underlying runs.
The basic concept behind calculation of · Penetration: Factors that would nor- the copper- and iron-rich phases. There is
weld pool composition relies on the lever mally be expected to give more penetra- also the wide solidification range that
rule so familiar to metallurgical calcula- tion can be expected to give more would almost certainly give hot cracking
tions. dilution, eg GTAW, higher current, problems.
If it is assumed, as a simple situation, that slower travel rate. Where brittle phases are likely to be a
a weld pool between two dissimilar metals, problem, it may be that low temperature -
A and B, contains half of metal A and half 5.6.3 Microstucture determination non-fusion - brazing will have to be resorted
of metal B then the composition of the pool When metals are
must be an equal mixture of each alloy. mixed, they will ei-
In the case of a three component system, it ther mutually dis-
will be remembered that the composition of solve in each other,
the weld pool will depend on the ratio of form a mixture of
each metal. On the phase diagram the three phases or appear as
alloy lever rule is used for graphical calcu- a mechanical mix-
lation of the weld pool composition. ture of the two met-
Referring again to a binary alloy. If it con- als with virtually no
tains one quarter of metal A and three quar- mutual solubility.
ters of metal B, the composition must The structure after
reflect this. ‘mixing’ will de-
If there is a third alloy introduced as a pend on the actual
filler material then the composition of the zone in the weld,
weld pool will be controlled by the amount Figure 5. These
of filler present. As an example of this, the considerations are
effect of the ratio of parent metals and filler illustrated in the
metal for a mild steel/UNS S30400 weld phase diagrams for
with and without S30900 filler is shown in three systems:
Table 6. · Copper - nickel, Figure 7 Copper-nickel phase diagram
For example, assume that each compo- Figure 7: These
nent has the following composition of metal alloys have simi-
A. lar crystal struc-
Parent 1: 10% tures and each is
Parent 2: 20% soluble in the
Filler wire: 30% other.
Also assume that the weld pool contains · Lead-tin, Figure
the following proportion of each compo- 8: Each of
nent, (See Figure 6): these alloys has
Parent 1: 35% a limited solu-
Parent 2: 40% bility for the
Filler wire: 25% other but at
The amount of metal A in the weld pool room tempera-
will therefore be: ture it would be
(0.35 x 10%)+(0.4*x 20%)+ found that all
(0.25 x 30%)=19% alloys, other
than those at the
This calculation assumes the unlikely extreme ends,
situation that there has been no loss by oxi- would consist of
dation during welding. If necessary, an es- a mixture of the
timated correction could be made for this. two solid solu-
All of the above assumed relatively sim- tions. Figure 8 Lead tin phase diagram
ple ratios of each of the contributing materi- · Aluminium -
Iron, Figure 9:
to. Even in this case it could be that care
1600
1600
Liquid LIQUID
TEMPERATURE °C
1400
Fe
Temperature °C
1200
+LIQUID
1200
g
FeAl
1000
800 +
Al
Magnetic Fe3Al
will have to be taken to ensure that there is · Carbon and low alloy steel fillers:
no phase reactions. 5.6.4 Microstructure stability
These are not normally recommended
If melting of the base materials can be for dissimilar metal welding. A dissimilar metal weld could have unusual
eliminated, these reactions can only occur The addition of alloys from stainless or phase structures that may give problems, for
by the much slower solid state diffusion high nickel alloys increase the example if the weld is required to have long
and thus the problem is less likely to arise. hardenability with probable post weld term stability at elevated temperatures.
cracking The common problem of sensitisation with
5.6.3.1 Consumables stainless steels is an example. If a stainless
Taylor has summarised the effect of vari- Nickel, chromium and copper can in-
crease the probability of post weld steel is welded to a medium carbon steel and
ous elements in filler materials on the
cracking both are then subjected to elevated tempera-
microstructure:5
Aluminium, magnesium and titanium tures, it is possible that the stainless steel will
· Nickel fillers: These are tolerant to a have its carbon content increased by diffu-
alloys will generate brittle intermetallic
wide range of diluting elements. The sion from the carbon steel. This can then lead
compounds
main problem areas are lead, sulphur to sensitisation corrosion.
and phosphorus · Austenitic stainless steel fillers: These
have been studied extensively and there An extension to this is that there will also be
· High nickel (Monel®) type fillers: High a following stage where more extensive car-
is quite a lot of information on the toler-
nickel-copper type fillers, because of bides will form in the austenitic steel, thus
ance for many elements, see Section 6.1
their mutual solubility, can be deposited making the weld zone more brittle and likely
Aluminium, magnesium and titanium
over copper, nickel, Monel® or cupro to crack under the design stresses plus, in this
will again lead to brittle intermetallic
nickels case of elevated temperature service, the
compounds
Dilution with up to 10% iron or 5% thermal stresses that will be present.
chromium will lead to cracking al- · High nickel-chromium alloy fillers:
These are dealt with later, see Section There is a further potential problem in that
though this can depend on the process the lowering of the carbon content of the low
(see Table 8) 6.3.
In general chromium should not exceed alloy steel will decrease its strength. The
· Copper Nickel type fillers: Again these lower carbon steel is also more susceptible to
around 35%. This is a sufficiently high
can be diluted with any amount of grain growth with a further possibility of a
value that it would seldom lead to prob-
nickel and copper because of the mutual decrease in mechanical properties.
lems.
solubility There is also the possibility of martensite
Copper should not exceed 30%, al-
Dilution with more than around 5% of formation, see Section 4.2.
though some authorities limit this to
iron or 5% of chromium can lead to so- Other than lowering the amount of heat in-
15%
lidification cracking. Weldable grades put, there is not a great deal that can be done
of cupronickel also have close limits · Aluminium bronze fillers: These fillers
to avoid these problems - it is probably better
placed on carbon, phosphorus, silicon, can withstand dilution from both iron
to avoid the situation where this type of weld-
sulphur and zinc to minimise cracking and copper and are often recommended
ing is required.
These fillers also can form brittle inter- for the dissimilar metal welding of car-
metallic compounds with aluminium bon and low alloy steels, stainless steels 5.6.5 Corrosion
magnesium and titanium when solubil- and copper base alloys
The most likely problem with a dissimilar
ity limits for these alloys are exceeded. These items will be discussed further for
weld in a corrosive environment is the gen-
specific metal combinations in Section 6
· Aluminium, magnesium and titanium eration of a galvanic couple. This can occur
fillers: These fillers cannot tolerate iron, on the macro scale between one of the parent
chromium or copper without generating plates and the mixed weld pool as well as be-
an unacceptably brittle weld
5 Taylor JS The fusion welding of dissimilar metals The Welding Technology of Stainless Steels WTIA Seminar Melbourne October
1995
0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 OXIDE
NOTCH
NOTEBOOK
The type of alloy microstructure obtained in the
solidified weld pool will control the weld prop-
erties. The microstructure can be predicted to a
large extent from the phase diagram for the two
alloys.
The types of structure found are:
1. Complete mutual solubility
2. Mixtures of two solid solutions
3. A solid solution or solutions with a interme-
tallic phase or phases. These phases are gener-
ally brittle but there are many cases where they
are not.
Another problem that can arise with unusual al-
loy additions is a major alteration in the melting
range. Iron copper alloys show this.
Microstructure stability is also important, eg
carbides can form with prolonged heating at in-
termediate temperatures in stainless steels that
Figure 11 Galvanic series in sea water have been contaminated with a plain carbon
steel. These carbides can then result in unde-
sirable mechanical properties or accelerated
corrosion due to sensitisation.
Higher Cr
Liquid +Mo
NOTEBOOK
Galvanic corrosion is an ever present possibility
with dissimilar metal welds.
Area relationships become significant if the
weld is anodic, ie likely to corrode.
Carbon or low alloy steels welded to stainless
steels are a likely site for this type of corrosion.
A common, but not particularly successful, pre-
Cr ventative measure is to paint the carbon or low
Ni
alloy steel plus the weld zone and to continue
Mo the paint film over the stainless steel for about
Dendrite composition increases 1cm.
This procedure will not help the carbon steel if
in Mo and Cr as solidification the film is damaged - and may, in fact, acceler-
continues. Higher Ni ate corrosion if the damaged area is small and
also close to the stainless steel.
Figure 12 Segregated areas in the dendrites forming in a highly alloyed Ni-Cr-Mo
weld.
Stainless
Consumable Also known as WELD POOL RATIO WELD POOL COMPOSITION.
Steel
Parent Metal A Parent Metal B Filler Metal C C Cr Ni
S30400 S30900 309
0.00 1.00 0.00 0.06 19.00 10.00
S30403 S30983 309L 0.25 0.75 0.00 0.10 14.25 7.50
S31000 S30900 309 0.50 0.50 0.00 0.13 9.50 5.00
S31600 S30903 309L 0.75 0.25 0.00 0.17 4.75 2.50
1.00 0.00 0.00 0.20 0.00 0.00
S31603 S30986 309MoL
S32100 S30982 309Mo 0.50 0.50 0.00 0.13 9.50 5.00
S34700 S30900 309 0.45 0.45 0.10 0.13 10.85 5.85
0.40 0.40 0.20 0.12 12.20 6.70
S31803 W39209 2209
0.35 0.35 0.30 0.12 13.55 7.55
2510 or 0.30 0.30 0.40 0.12 14.90 8.40
S32750
25.10.4 0.25 0.25 0.50 0.12 16.25 9.25
S43000 S30900 309 0.20 0.20 0.60 0.11 17.60 10.10
S44400 S30982 309Mo 0.15 0.15 0.70 0.11 18.95 10.95
0.10 0.10 0.80 0.11 20.30 11.80
S41000 S30900 309
0.05 0.05 0.90 0.10 21.65 12.65
0.00 0.00 1.00 0.10 23.00 13.50
32
Nickel Equivalent: %Ni + 30x%C + 0.5%Mn
qualified fillers are listed for most stainless
steels in AS 1554.6 (p45) with 309 type
28 stainless steel being the most common. A
Austenite %
5% 10 listing of the grades suggested for some of
24 the more common steels adapted from this
%
20 Standard is given in Table 5 .
20 A+M Consumable selection can be understood
304 A+F 40% by reference to the Scaeffler DeLong dia-
16 gram, Figure 14
80% It is possible to predict the type of alloy
12 that will be obtained in the weld pool by re-
lating these to the approximate ranges of the
Martensite 100%
8 A+M+F
particular alloy groups superimposed on
MS this diagram.
4 Ferrite For example, Figure 15 shows what type
F+M M+F of alloy would be expected in a weld pool
0 with equal amounts of S30400 and mild
0 4 8 12 16 20 24 28 32 36 40 steel present in the pool, a condition that
would be expected in an autogenous weld. .
Chromium Equivalent: % Cr + %Mo + 1.5x%Si + 0.5x%Nb This condition would not usually be ac-
Figure 15 Probable weld pool composition from a mild steel-S30400 weld with no ceptable since it can be seen that a marten-
filler material, ie within the martensite range and hence there will be a high probability sitic structure with its attendant undesirable
of cracking properties would occur.
If a higher alloyed consumable, such as
UNS S30900, were used and it was as-
32
Nickel Equivalent: %Ni + 30x%C + 0.5%Mn
4
A
0
16 of this element in the diagram can allow pre-
2
22
diction of martensite regions. It is for this
14
12
6
reason that the original Schaeffler diagram
20
18 28
35
10
24
16 still finds application in dissimilar weld
26 30
14 40 50 structure prediction6.
One particular area of usefulness of the
FA 60 70
Schaeffler diagram is to indicate the type of
80
behaviour that can be expected with welds of
90
12 various compositions, Figure 187
F 100 6.1.2 High temperature applications
Because of the problems with microstructure
10
stability, it is usual to use one of the high
nickel alloys in joints expected to operate
over around 400°C. This is because of the
18 20 22 24 26 28 30 higher tolerance to carbon and their favour-
able coefficient of thermal expansion as dis-
cussed in Section 5.2
These alloys also have a higher inherent
Figure 17 WRC 1992 diagram
creep strength and oxidation resistance to as-
sist their survival at these higher tempera-
tures.
The consumables usually employed are
AWS 5.14 ERNiCr-3 or 5.11 ENiCrFe-3
Hot cracking
6.2 Ferritic/martensitic
28 above 1250° C stainless steels - carbon
Nickel Equivalent: %Ni + 30x%C + 0.5%Mn
24 steel
Austenite Martensitic
cracking The principal point of concern here is the
20 A+M below 400°C hardenability of the combined joint
A+F There are a large number of possible com-
16 Brittleness after binations but almost all will generate a
heating between
12 500- 900°C hardenable steel through the combination of
Martensite chromium and carbon that w in the weld pool
8 A+
M+
F High
4 Ferrite temperature Welded to
brittleness
0
F+M M+F High nickel Carbon or
Most suitable alloy Stainless
0 4 8 12 16 20 24 28 32 36 40 structure, low alloy
steel
austenite steel
with 3-8%
Chromium Equivalent: %Cr + %Mo + 1.5x%Si + 0.5x%Nb delta ferrite ERNi-1
ERNi-1
N02200 ERNiCr-3
ERNiCr-3
ERNiCrFe-6
ERNiCr-3
Figure 18 Modified Schaeffler diagram showing composition regions prone to brittle- N04400 ERNi-1
ERNiCrFe-6
ness and cracking
ERNiCr-3 ERNiCr-3
N06600
ERNiCrFe-6 ERNiCrFe-6
N08825 ERNiCrMo-3 ERNiCrMo-3
NOTEBOOK
In the welding of high nickel alloys the concern with maximum limits of iron and/or chromium means N10665 ERNiMo-7 ERNiMo-7
that special care must be taken in consumable selection. N10276 ERNiCrMo-4 ERNiCrMo-4
It is also important to eliminate the normal nickel contaminants, lead, sulphur, phosphorus and zinc
Tables have been prepared with suggestions on maximum values that can be accepted.
Again, as with all dissimilar metal welds, it is advisable to conduct test welds to check the ability of the
weld to meet specification requirements.
Table 7 Suggested TIG/GMAW filler ma-
terials for dissimilar nickel alloy welds to
carbon and stainless steels
This can be minimised by using a filler erable to use a less hardenable filler Copper and steel are virtually insoluble in
metal with the same composition as the car- metal. the solid state and a weld pool between the
bon/low alloy steel but it is still probable two will be a mixture of two phases, refer
that there will be sufficient chromium 6.3 High nickel alloys Figure 10. This diagram also shows the
pickup from the stainless steel to give a The principal problems here are associated wide freezing range that can occur with
martensitic weld pool with contamination of the nickel alloy. these alloys. This can point to the possibil-
If hardenability is a problem, then it may Nickel alloys are particularly sensitive to ity of hot cracking.
be better to use a buttering layer of high sulphur because of a low melting point Thermal conductivity also presents a
nickel stainless steel on both components. eutectic that gives cracks and later failure in problem with the copper alloy frequently
Types 309 or 310 can be used . These may high temperature service. requiring manipulation of the heat source to
then be heat treated to obtain the desired Other contaminants that must be avoided give a uniform temperature in the weld
properties. The weld can then be completed are phosphorus, lead and zinc. zone.
with an austenitic alloy such as type 308 The major alloying elements can also give Pre-heating, particularly of the copper, is
stainless steel problems and it is usual to use the dilution necessary to also help overcome conductiv-
The following general rules have been calculations explained in Section 5.6.2 to ity problems
proposed for joining the 4xx series stainless determine the weld pool composition and GTAW and MMAW are usually preferred
steels8 then relate this to tabulations of generally for this type of junction because of the bet-
· For welding one hardenable chromium acceptable impurity levels such as that ter control that can be achieved over heat in-
steel to another with a higher chromium shown in Table 8. put and placement. Oxy-acetylene welding
content, filler material with chromium An alternative way of presenting this in- would not normally be used because of the
content equal to that of either steel may formation for iron and chromium is given lack of control.
be used. Furthermore, any filler mate- by the American Welding Society9, Figures Where dilution can generate problems, eg
rial whose chromium content lies be- 19 and 20 They state that this information by iron pick up in the copper giving a wide
tween these limits is equally is based more on practical experience than solidification range and subsequent solidi-
satisfactory provide the weldment is fundamental metallurgy . fication cracks, buttering may be necessary,
properly heat treated. A listing of some suggested filler materi- particularly on thicker sections, ie greater
· A general rule for welding any chro- als for dissimilar nickel welds for bare wire than about 3 mm.
mium steel to any low alloy steel is to welding processes are given in Table 710. Iron can give hot cracking problems over
use a filler metal that has the same com- Readers are also referred to trade publica- a reasonably well defined composition
position as the low alloy steel, provided tions in this area11,12 and the appropriate range, Figure 21 .
that it meets the service requirements of AWS Standards13,14 Phosphorus is a problem with
the application. With any low alloy cupro-nickels because of the formation of
steel filler metal, the chromium that is 6.4 Copper alloys brittle nickel phosphides.
picked up by the dilution with the chro- Buttering can be done by a deposit of a
mium steel base metal must be consid- 6.4.1 Dissimilar fusion welds brazing material or by a weld deposit. A
ered. Copper and its alloys can be welded to car- common buttering material is nickel be-
· For welding any chromium steel to a bon and stainless steels as well as high cause of the total mutual solubility of cop-
carbon steel, carbon steel filler metal nickel alloys per and nickel, Figure .7 The weld can then
can alternatively be used, but it is pref-
0 10 20 30 40 50
0 10 20 30 40 50 60
MAXIMUM ALLOWABLE CHROMIUM CONTENT (%) MAXIMUM ALLOWABLE IRON CONTENT (%)
Figure 19 Tolerance for chromium in nickel alloy welds Figure 20 Tolerance for iron in nickel alloy welds
8 American Welding Society Welding Handbook 7th Ed Vol 4, Chapter 12 Dissimilar Metal Welding p528 (1982)
9 American Welding Society Welding Handbook 7th Ed Vol 4, Chapter 12 Dissimilar Metal Welding p531 (1982)
10 American Welding Society Welding Handbook 7th Ed Vol 4, Chapter 12 Dissimilar Metal Welding p532 (1982
11 INCO Alloys International Nickel based welding products p3 (1991)
12 VDM Australia Pty Ltd, Welding the VDM high nickel alloys (undated)
13 American Welding Society. Standard AWS 5.11 Specification for nickel and nickel alloy welding electrodes for shielded metal arc
welding
14 American Welding Society. Standard AWS 5.11 Specification for nickel and nickel alloy bare welding electrodes and rods
TYPICAL FILLER
ALLOYING ELEMENT
FILLER GROUP MATERIALS
AWS 5.11 AWS 5.14 Fe [see Note] Ni and Cu Cr C,Si and Mn
About 30-35%.
Cracking commences at Levels normally found in
NICKEL BASED Above this value
ENI-1 ERNI-1 about 25-40%Fe with ERNi1 No limits commercial practice can
FILLERS sigma phase can
being more susceptible usually be tolerated
form
ENiCu7 can take up to 30%
before hot cracking, C: Values above about 0.4
ERNiCu7 commences to can give graphitisation
6-8% is the
crack at 10-15%. Flux Si: values greater than about
acceptable upper
NICKEL-COPPER control available with SAW 1%give unacceptable weld
ENiCu-7 ERNiCu-7 No limits limit. Hot cracking
BASED FILLERS can allow even more iron. ductility.
can occur above
With the gas shielded Mn: Increases weld ductility.
this value
processes values between 5 Some fillers use up to 9% to
and 10% have been help prevent cracking
suggested
30% Cr is about the
maximum level.
Ni: satisfactory to This is the is the
Can accept up to 10-15% Apart from silicon, which
NICKEL-IRON- all levels approximate
but above that level becomes should be limited to 1%,
CHROMIUM- Cu: Can accept up composition of the
ENiCrMo-3 ERNiCrMo-3 similar to an austenitic values found in normal
MOLYBDENUM to about 15% copper filler. Problems
stainless steel and commercial products
FILLERS before hot cracking will therefore arise
susceptible to hot cracking should not be a problem
occurs if welding high
chromium
materials
Ni: satisfactory to all Maximum level is
Apart from silicon, which
Up to 50% iron can be levels 30-35% without
NICKEL- should be limited to 1%,
tolerated with MMAW but Cu: Can accept up to cracking or
CHROMIUM- ENiCrFe-3 ERNiCr-3 values found in normal
only 25-30% with non about 15% copper problems
IRON FILLERS commercial products
coated filler materials before hot cracking associated with
should not be a problem
occurs second phases
Cu can cause hot
shortness in the
The structure should
weld pool so that The structure should be
be calculated from
PLAIN CARBON All values of iron can be carbon and low calculated from the
the
OR LOW ALLOY VARIOUS VARIOUS tolerated since the filler is alloy steel fillers Schaeffler-DeLong diagram
Schaeffler-DeLong
FILLERS essentially iron should not be used to avoid martensidte
diagram to avoid
for welding high formation
martensite formation
copper nickel base
alloys
The structure should
Cu can cause hot
be calculated from
The structure should be shortness in the The structure should be
the
calculated from the weld pool so that calculated from the
Schaeffler-DeLong
AUSTENITIC Schaeffler-DeLong diagram. austenitic stainless Schaeffler-DeLong diagram.
diagram. The aim
STAINLESS VARIOUS VARIOUS The aim should be to avoid steel fillers should The aim should be to avoid
should be to avoid
STEEL FILLERS martensite and end up with not be used for martensite and end up with
martensite and end
about 4-10% ferrite to avoid welding high about 4-10% ferrite to avoid
up with about 4-10%
hot cracking copper nickel base hot cracking
ferrite to avoid hot
alloys
cracking
Not normally present in the
70-30 cupro-nickels have a 5% m ax i m u m usual applications where
COPPER NICKEL
5.6 - ECuNi 5.7 - ERCuNi limit o f 5-10% b efo r e No limits before hot cracking these alloys are welded but
FILLERS
cracking occurs occurs substantial quantities would
be harmful.
Table 8 Weld pool composition limits for some grades of nickel alloys and welding electrodes.
Note: Iron is often limited to 5% maximum in the surface layer of high alloy welds to minimise corrosion problems
per into the steel. This is sometimes re- This defect is also sometimes known as
ferred to as liquid metal corrosion. infiltration
Ni
Molten copper has a low surface tension
on iron and will quickly penetrate down 6.4.3 Dissimilar metal brazing
80 20 grain boundaries. Internal stress acceler- Because of the tendency of the copper
ates this type of corrosion. nickel alloys to hot cracking, Figure 21 and
60 40 Cu% Because of this, care must be taken when stress cracking (ie copper infiltration) the
Ni%
welding or brazing copper materials to steel silver brazing alloys are preferred for this
40 60 to ensure that the conditions are such that type of operation. Phosphorus is a particu-
liquid metal attack does not occur.
20
80
Hot short range
METAL B
METAL A Phosphor Aluminium Silicon Cupronickel
Fe Cu Copper
bronzes bronzes bronzes s
20 40 60 80
Fe% Low zinc brasses, eg ERCuSn-A
C23000 540°C
Phosphor bronzes eg ERCuSn-A
Figure 21 Hot short range in cupro C51000 540°C
nickel alloys caused by iron
contamination Aluminium bronzes, ERCuAl-A2 ERCuAl-A2
eg C61400 540°C 200°C
be completed with a filler suitable for the Silicon bronzes, ERCuSn-A ERCuSi-A ERCuAl-A2
nickel buttering layer. eg C65500 540°C 65°C max 65°C max
Brasses can be welded to steel if the zinc is
less than about 20% and the brass is not di- Cupronickels, eg ERCuAl-A2 ERCuSn-A ERCuAl-A2 ERCuAl-A2
rectly heated by the arc. It is usual to use a C70600 540°C 65°Cmax 65°Cmax 65°Cmax
copper tin buttering layer (ERCuSn-A) and Nickel, eg N02200 and ERCuNi or ERCuNi or
then use this same material as a filler. nickel-copper, eg ERCuNi-7 ERCuNi-7
There is a wide range of filler materials N04400 alloys 540°C These combinations 65°C max
specified for this type of junction. These not usually welded
High nickel alloys, eg ERNiCr-3 ERNiCr-3
will vary according to the type of copper al-
N08800, N06600 540°C 65°C max
loy and the welding procedure. Some ex-
amples for GTAW, with suggested ERCuAl-A2 ERCuSn-A ERCuAl-A2 ERCuAl-A2 ERCuAl-A2
Low carbon steels
pre-heats, are given in Table 915. Probably 540°C 200°C 150°C 65°C max 65°C max
the most common filler for steel junctions
ERCuAl-A2 ERCuSn-A ERCuAl-A2 ERCuAl-A2 ERCuAl-A2
is aluminium bronze (CuAl-A2). Silicon Low alloy steels
540°C 260°C 260°C 200°C 65°C max
bronze CuSi-A and phosphor bronze (Cu-
Sn-A) are also used for non-nickel bearing Stainless steels, eg ERCuAl-A2 ERCuSn-A ERCuAl-A2 ERCuAl-A2 ERCuAl-A2
materials. S30400 540°C 200°C 65°C max 65°C max 65°C max
NOTEBOOK
Dissimilar welds containing copper base alloys can be made
Aluminium plug weld between copper alloys as well as with high nickel alloys, car-
bon and low alloy steels and stainless steels.
Copper In many cases iron pick up can lead to hot cracking but but-
tering can be a way of minimising this problem.
Silicon bronze, aluminium bronze and phosphor bronze are
common filler/buttering alloys.
Copper infiltration is a potential danger with steels. This is
penetration of copper along the grain boundaries of the steel.
Internal stress in the steel can promote this type of liquid
metal embrittlement
Aluminium
15 ASM International ASM Handbook Vol 6 Welding Brazing and Soldering (1993), p769
16 Smith L M Engineering with clad steel Offshore Technology Conference, Houston, 1992. Reprinted as NiDI 10064
Check compositions against appropriate Standard before use February 15, 2002
19 Dissimilar metal welding Alloy Compositions (maximum unless specified)
Check compositions against appropriate Standard before use February 15, 2002
20 Dissimilar metal welding Alloy Compositions (maximum unless specified)
Check compositions against appropriate Standard before use February 15, 2002
21 Dissimilar metal welding Alloy Compositions (maximum unless specified)
Check compositions against appropriate Standard before use February 15, 2002
22 Dissimilar metal welding
INDEX
A carbon and low alloy steel fillers . . . . . . . 8 lead - effect of . . . . . . . . . . . . . . . 13
copper nickel fillers . . . . . . . . . . . . . . 8 weld pool composition limits . . . . . . . . 14
aluminium alloys to copper high nickel (Monel®) fillers. . . . . . . . . . 8 zinc - effect of. . . . . . . . . . . . . . . . 13
brittle intermetallic compounds . . . . . . . 16 high nickel-chromium alloy fillers . . . . . . . 8 nickel silver
high zinc solder . . . . . . . . . . . . . . . 16 nickel fillers . . . . . . . . . . . . . . . . . 8 brazing . . . . . . . . . . . . . . . . . . . 16
plug weld . . . . . . . . . . . . . . . . . . 16 Nickel fillers . . . . . . . . . . . . . . . . . 8
silver coating . . . . . . . . . . . . . . . . 16 flux
welding . . . . . . . . . . . . . . . . . . . 16
aluminium alloys to steel
selection for wetting . . . . . . . . . . . . . 6
friction welding · · · · · · 16
O
aluminium coating of the steel. . . . . . . . 16 aluminium to copper base alloys. . . . . . . 16 oxidation
diffusion problems · · · · · · · · · · · · · 16 aluminium to stainless steel . . . . . . . . . 16 oxide ‘notch’ . . . . . . . . . . . . . . . . 11
brittle intermetallic compounds . . . . . . . 16 diffusion limitation. . . . . . . . . . . . . . 16
thermal conductivity. . . . . . . . . . . . . 16 tool steel shanks . . . . . . . . . . . . . . 16
welding . . . . . . . . . . . . . . . . . . . 16
P
B G phase diagram · · · · · · · 7
copper-nickel . . . . . . . . . . . . . . . . . 7
galvanic series · · · · · · · 9 lead-tin . . . . . . . . . . . . . . . . . . . . 7
boiler tubes · · · · · · · 5 grain growth · · · · · · · 8 post-heating
brass - welding to steel · · · · · 15
sensitisation . . . . . . . . . . . . . . . . . 6
brazing
post-weld heating · · · · · · 6
effect of expansion on capilliary action . . . . 6
brittle phases H prequalification of dissimilar welds
pre-weld heating · · · · · · 6
· · 3
minimised by brazing . . . . . . . . . . . . . 7
high temperature oxidation · · · · 11
hydrogen embrittlement
C
cathodic reaction . . . . . . . . . . . . . . 10
Q
qualification of dissimilar welds · · · 3
clad plate · · · · · · · · 17
copper alloys I
copper penetration . . . . . . . . . . . . . 15
infiltration · · · · · · · · · · · · · · · · · 15
infiltration · · · · · · · · 15
R
surface tension · · · · · · · · · · · · · · 15
roll bonding · · · · · · · 17
cupro-nickel
problems with phosphorus · · · · · · · · · 13 L clad plate . . . . . . . . . . . . . . . . . . 17
solidification range - iron contamination . . . 13
lever rule
suggested dissimilar weld fillers. . . . . . . 15
thermal conductivity. . . . . . . . . . . . . 13
dilution calculation . . . .
lower melting point alloys
. .
·
. .
·
. . .
·
. .
·
. 7
4
S
corrosion · · · · · · · · 8
contraction stresses . . . . . . . . . . . . . 4 Scaeffler DeLong diagram · · · · 11
galvanic couple. . . . . . . . . . . . . . . . 8
effect of arc direction . . . . . . . . . . . . . 4 solidification range
area
hot cracking . . . . . . . . . . . . . . . . . 4 wide range and hot cracking . . . . . . . . . 7
advantage · · · · · · · · · · · · · · · · 10
use of intermediate alloys . . . . . . . . . . 4 specific heats · · · · · · · 3
stainless steel fastener 10 stainless - carbon steel
area affects · · · · · · · · · · · · · · · · 10
filler metals . . . . . . . . . . . . . . . . . 10
carbon steel - stainless stee
paint protection· · · · · · · · · · · · · · 10 M high temperature applications . . . . . . . . 12
low temperature applications . . . . . . . . 11
hydrogen evolution · · · · · · · · · · · · 10
magnetic effects Scaeffler DeLong diagram . . . . . . . . . . 11
micro scale
arc deflection . . . . . . . . . . . . . . . . 11 martnsite formation prediction · · · · · · · 11
molybdenum segregation· · · · · · · · · 10
melting ranges - table · · · · · 3 weld pool alloy prediction · · · · · · · · · 11
micro scale between phases · · · · · · · · 10
melting temperatures · · · · · 4 WRC 1992 diagram . . . . . . . . . . . . . 11
metal mixing · · · · · · · 6 martensite formation prediction · · · · · · 11
microstucture
D brittle intermetallic phases, eg Al-Fe . . . . . 7
dilution calculation · · · · · · 7
galvanic effect, eg Cu-Fe . . . . . . . . . . . 7
mixture of the two solid solutions, eg Pb-Sn . 7
T
lever rule . . . . . . . . . . . . . . . . . . . 7 mutual solubility, eg Cu-Ni . . . . . . . . . . 7 thermal conductivity · · · · · · 6
stability. . . . . . . . . . . . . . . . . . . . 8 distortion . . . . . . . . . . . . . . . . . . . 6
wide solidification range, eg Cu-Fe . . . . . . 7 effect of temperature. . . . . . . . . . . . . 6
E Microstucture
stability
thermal expansion · · · · · · 4
coefficients . . . . . . . . . . . . . . . . . . 4
explosion welding · · · · · · 16 carbide formation · · · · · · · · · · · · · · 8 intermediate expansion coefficient . . . . . . 5
clad plate . . . . . . . . . . . . . . . . . . 17 carbon depletion and grain size · · · · · · · 8 stress calculation. . . . . . . . . . . . . . . 5
molten 'jet' . . . . . . . . . . . . . . . . . 17 elevated tempeature · · · · · · · · · · · · · 8 stresses and cracking . . . . . . . . . . . . 4
sensitisation · · · · · · · · · · · · · · · · · 8 minimising · · · · · · · · · · · · · · · · · 5
microstuctures · · · · · · · 7 thermal cycling - cracking . . . . . . . . . . 5
F titanium welds· · · · · · · 16
transition layers . . . . . . . . . . . . . . . 16
ferritic/martensitic stainless steels - carbon steel
4xx series stainless steels . . . . . . . . . . 13 N
buttering . . . . . . . . . . . . . . . . . . 13
filler impurity limitations
nickel alloy
- effect of phosphorus . . . . . . . . . . . 13
U
aluminium bronze fillers . . . . . . . . . . . 8 - effect of sulphur . . . . . . . . . . . . . 13 unmixed zone · · · · · · · 6
aluminium, magnesium and titanium fillers . . 8 chromium tolerance . . . . . . . . . . . . . 13
austenitic stainless steel fillers . . . . . . . . 8 iron tolerance . . . . . . . . . . . . . . . . 13