CA1091306A - Electrode combination for electroslag welding of base metal sections - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE

Sections of cast manganese steel are joined by electroslag welding. Metallurgical limitations, oper-ating parameters and physical attributes of the guide tube are imposed to retard embrittlement of the base metal and weld metal as well, to prevent tearing, to assure satisfactory ductility, to preserve austenitic integrity of the weld metal and to assure the mechanical properties of the weld acceptably match those of the base metal. This divisional application is directed to an embodiment wherein the weld wire is centered inside a copper guide tube which in turn is supported inside a sleeve of stainless steel.

Description

lV~
~.
This invention relates to a method for joining sections o~ manganese steel.
Manganese steel is used extensively in castings ; subjected to severe abrasion or impact: earth moving equipment, grinding mills, railroad trackwork and so on, principally because the material embodies both ductility and wear resistance.
; Large parts are economically manufactured by casting several sections and joining them by welding. Proportions æ e becoming enormous: in metallic mining dipper buckets ~f 25 cubic yard capacity are employed; 30 cubic yard capacities are being planned.
The labor cost for welding becomes severe as the castings beco~.e larger: more man hours are invol~ed, less quality can be expected and delivery dates are retarded. This is so in spite of the fact that it is customary to join the parts by semiautomatic welding techniques The cost problem of joining heavy sections of manganese steel castings by welding ~the heavier the section the more heat in-put) is exacerbated by the embrittlement phenomonon encountered when cast, heat treatea manganese steel is reheated, as will now be explained.
; Austenitic manganese steel, which is.also called Hadfield's manganese steel after its inventor, is an extremely .tough non-magnetic alloy in which the usual hardening trans-formation has been suppressed by a combination of high mansanese content and rapid cooling from a high heat treatment temperature~
It is characteri~ed by high strength, high ductility and excellent wear resistance and is extensively used in severe gougingj crushing, impact and grinding wear applications because the material actually gets harder the more it is worked. ~ -109130~i The nominal composition contains 1.2% carbon and 12% or 13% manganese as essential elements. Commercial products will vary within the 1.0 - 1.4% carbon and 10 - 14%
manganese ranges establlshed by AST~I Designation A128.
The as-cast structure of manganese steel contains carbides and other transformation products that produce marked brittleness by their continuity. The standard toughening heat treatment involves austenitizing above 1832F to place all the ` carbides in solution, followed by rapid cooling in water to prevent re-precipitation of the carbides.
Subsequent reheating of standard manganese steel parts is potentially more serious than for ordinary structural steels. Instead of the usual softening and increase in ductility, manganese steel will become embrittled if heated enough to induce partial transformation of the metastable austenite. As stated I -in the Metals ~andbook (1961): "As a general rule manganese , steel should never be heated above 500F, either by acci~ent or plan, unless the standard toughening treatment is to be appliedn.
Both time and temperature are involved, lower temperatures requir-ing longer for impairment to develop. Only a few minutes are required at the dull red heat of 1000 to 1200F to begin embrittlement of this steel.
Since prolonged reheating of toughened manganese steel results in embrittlement, only arc welding is currently recommended for welding manganese steel. With a covered electrode or semiautomatic welding, the welder can usually control heat in-put in such a way that no area is seriously overheated.
The problem, then, is essentially two fold. Larger sections with iong weld seams entail high labor cost; thick _ _ _ ___ _ ~

``-` `' ' 1091;~0~

~ `
sections tsay two to three inches or more) involve a great deal of heat in-put likely to produce embrittlement.
Electroslag welding is known to be more economical than semiautomatic or manual arc welding fro~ the standpoint of time re~uired. However, the thermal cycles involved are discouraging to the idea of applying the process to joining sections of manganese steel.
The foregoing explains the problems we faced in - ¦
recognizing the need to find a more acceptable way to jo~n sections of manganese steel, particularly thick sections. The 1 objects of the invention are: to use electroslag principles to ; supplant semiautomatic arc welding (and`the other forms as well) as a method of joining cast manganese steel sections; to produce a wel~ of high integrity and satisfactory properties in manganese ~ -steel parts using electroslag principles; and to attain such a i' weld by constantly maintaining a reserve of austenite stabilize~s during progression o the weld. Other objects of the in~ention are to incorporate austenite stabilizers in the weld wire and~or . , I
consummable guide; to reduce hot tearing in the weld metal and heat affected zone of the base metal; to enable the highest possible welding current to be used, thereby accelerating the process so that the base metal is exposed to high temperature for as little time as possible; an~ to reduce the likelihood ; ~of an unacceptable e~brittlement of the base metal.
', ` 1~J9130 .

In the drawin~:
Fig. l is a dye checked section of B electroslag weld, showing severe cracking in the weld metal and heat affected zone of the manganese steel base metal;
'` Fig. 2 is a macro-etched section o~ an electroslag weld showing internal cracks in the base metal;
Fig. 3 is a macro-etched section through an electroslag weld showing crack;ng in the heat affected zone of the base mëtal with superimposed cross weld tensile test bars;
' ~ I .
Fig. 4 is a macro-etched section through an electroslag t weld showing hot tears at the interface between the weldment and base metal;
j Fig. 5 shows the results of two bend tests on electro- 1 i~ slag welds, exhibiting the effect of high aluminum in ~he base , .-, :
metal on hot tearing; J
Fig. 6 shows the results o~ two bend tests on electro- I
slag welas where the base metal was low in aluminum;
Fig. 7 shows damage in the base metal (bend test) caused by welding with a current of 600 amps, relying on lo~
20 phosphorus and low aluminum in the base metal castings;
~; Fig. 8 is a perspective vie~ of a dipper bucket which can be fabricated in accordanc~ with the present invention; and Fig. 9 is a schematic vie-~ of an elfectroslag wel~ing system as it may be used to practice the present invention.

~ J

`

~-- .
~(191~0~

Reference may be made to Fig. 8 for a consideration of practices involved when joining sections of manganese steel by a wela. Fig. 8 is a perspective of a manganese steel dipper !~
bucket. The middle and lower sections 10 and 12 are separate ~-castings of manganese steel joined by a long butt weld 15. In -use, any part of the bucket may be in tension. One advantage of manganese steel is its inherent ductility; it will stretch when tensioned and at the same time the tensioned area work hardens in the localized area. As a consequence the original yield strength of the pristine metal is increased. The adjacent pristine metal, not stretched, is relatively weaker. On ~he ` ~ext occurence of tension, the adjacent areas of pristine ~etal ~ yield, work harden and increase in yield strength (the same -~ pattern as before) which is to say the increase in strength is progressive throughout a section, progressively as that section is tensioned ~rom time to time. There is, then, a reserve of ductility in austenitic manganese steel. In field servioe this reserve is important in order that there will be no failure due to unexpected, abnor~al tensioning. For this reason, high temperature embrittlement which depreciates the reserve in ductility cannot be tolerated.
Eowever, by our reasoning, the reserves are larse enough to tolerate some embrittlement in the heat affected zone, if controlled to an acceptable degree. But an additional factor is involved, namely, to obtain a substantially unifor~ profile of mechanical properties across the weld zone, taking into account the indisputable fact that the base metal has to melt as the weld metal is being deposited. ~Je fo~nd the problem of attaining substantially matched yield strength could be resolved in principle during electroslag ~elding by employing a weld wire .

30~ -` .
of a particular alloy content, assuming of course proper control over weld par~meters.
Electroslag welding is a welding technique based on - the generation of heat by passing an electrical current through molten slag. Copper shoes, normally water cooled, are used to bridge the gap (joint) of the components to be welded, thus -j -forming a cavity to hold the molten fIux. Filler metal obtained from a welding wire is fed into the molten flux and the Il -resistance of the slag bath to the current flow provides the heat to melt ~he wire and the aajacent sections of the base metal. A guide tube is ordinarily used to feed the electrode wire into the molten flux and this guide tube also melts and contributes metal to the weld.
!
- The major obstacle to successful electroslag welding -of austenitic manganese steel is the ~astly different time~
temperature relationship of an electroslag we~d as compared to a shielded metal-arc weld.
; With a metal arcr the temperature of the fusion~
zone is relatively high instantaneously but it cools rapialy and only a very small a~ount of metal is at a high temperature relative to the weld. In electroslag welding, however, the ~1 temperature of the flux pool (3,000 - 4,000F) is much lower than the temperature of a welding arc but the ~ass of the slag !
pool and molten weld metal at a high temperature is relati~ely - large. Since a larger area of the base metal is heated in electroslag welding, both the heating and cooling rates of - the metal in the heat affected zone (HAZ) are much slower in co~parison to arc welding. This thermal feature of electroslag welding can be very beneficial when welding carbon and alloy 3~ s~eel$~ In these steels, the slow cooling rate considerably . ~ ~

1(~91;~0~
reduces the risk of cracks developing in the heat affected zone of the weldment. However, this characteristic thermal cycle associated with electroslag welding adversely affects the properties of austenitic manganese steel for reasons explained above.
In the first attempt to join manganese steel I~
sections by the electroslag process, three weldments were made, one 2" section and two 4" sections, using an experimental welding wire. Otherw~ise, welding parameters were mainly b~sed on experience with other steels. The base metal for two o~
the weldments cons;isted of standard manganese steel ana for the third weld a grade of manga~ese steel con~aining molybaenum was used (AST~-A-128, Grade E-l). This particular grade is krown to offer better resistance to heat embrittlement than the regular grade of manganese steel.
Nonetheless, all three welding tests were unsuccessul d7~e to severe cracking in both the weld deposit and the heat affected zone of the base metal - see Fig. 1. Microstruc~ural .
examination revealed severe embrittlement of the weld and base metal and evidence of incipient melting in the base metal. In addition, large metallic inclusions were found in the weld, suspected as being unfused portions of the carbon steel suide tube normally recommended for electroslag wel2ing.
. The abnormal structure resulted in spite of the fact that the welding wire contained a relatively large amoun~ of nic~el, normally considered helpful in avoidins e~brittlement of manganese steel. The nominal analysis for the wire was 0.92C, 20.8Mn, and 3.2Mi.
As will be evident from Fig. 1, extensive cracks ,~- -are revealed at the ~nterface between the weld metal and ~he 10~13~

weld. CracXs were persistent throughout the cross-sectiOn and were not confined to the exposed end surfaces. Heat damage, as evidenced by a continuous grain boundary carbide network, was observed in the base metal up to an inch from ~ interface. -¦
-; Analysis showed that melting of the base metal contri~uted nearly fifty percent of the weld metal, a considerable dilution. Realization of this large dilution factor, coupled with the immense heat input, coula be viewed as causing catastrophic instability of austenite in the - critical area. It was therefore reasoned that modifications in both the weld wire and guide tube conceivably co~ld be relied on to preserve austenitic stability, provided heat input could be reduced.
', The heat input was reduced by:
a. ~educing the root gap from 1-1~4"
to 3~4"; -b. Liniting the electrical parameters to 400 amps and 38 volts; and c. Using a smaller diameter wire ~1/16 instead of 3/32~) since a thinner wire would provide increased deposition rate for a gi~en amperase.
The base metal was further modified to provide improved heat resisting properties, the nominal chemical analysis being:
C% Mn~O Mo~ Si~ P%
0.80 14.00 1.20 0.5 .05 max.
This chemical analysis still ~alls within ASTM
specification A-128, Grade E-l, 1~9.130 . .
To compensate for the tremendous dilution by the melting base metal, to introduce austenite stabilizers and -in a further effort to reduce heat input, a sta;nless steel guide tube (SAE 304: 18Cr, 9Ni3 was com~ined with a welding .
wire having the following nominal analysis: O.9C, 18Mn, 7Cr, .
. 6Ni. The guide tube, it was reasoned, would melt at a - - --temperature lower than the carbon steel guiae initially used;
the nickel-chromium content in both the guida tube and weld.. , .- ., wire would impart heat resistance (resistance to embrittlement ~ :

of the weld metal) and would continuously contribute austenite . . - . ( . .
stabilizers in the form-of nicXel, manganese and carbo~ during . the progression o' the weld.
These modifications in the ~uide tube and wel~ wire were determined as responsible for establishîng mechanical proper.ies across the weld satisfactorily matching those of a standard manganese steel tY.S. 50-55000; El. 30-34~ as will . be evident from data obtained from this successful experimental wela set forth in Ta~le 1:
-.
,:' . , ~:
, ' ' ' l(J9i3~6 .
: TABLE 1 "` , I , Automatic ~elding of Man~anese Steel ; 2" Section Test Weld . Base Metal Heat C~ Mn% Si% Mo~ P%

2-001 0.8214.02` 0.44 1.25 0.028 ~ ,~

2-020 0.8114.10 0.51 1.22 0.020 J

: Wire Composition Experimental formulation AN 4 cal~ulated ~ composition:

: C~ Mn~ Cr% - Ni%

0.932 18.77 7.67 6.39 (Calculated to provide a weld composition of 0.80%C, 14.04%~, 4.01~Cr, 3.52%Ni) .
~ , . . I
Actual Weld Analysis: ¦

C% ~ Mn% Si~ Cr% Mo% Ni% P%

Burn 1 0.81 15.70 0.36 4.12 0.52 3.40 0.022 Burn 2 0.82 15.80 0.40 4.13 0.50 3.46 0.021 Burn 3 0.79 15.70 0.36 3.92 0.55 3.45 0.022 ''~
Cross-~eld Tensile Properties . , I
Sample . -No. Y.S T.S. El.~ R.A.%
AT-484-A 54,999102,000 33.5 44.9 AT-484-E 50,000104,000 34.0 35.0 When givin~ the analysis (chemistrv) of the base metal, wire ana weld it is understood the remainder or balance (percent by weight) is substantially iron, that is, iron diminished by incidental impurities.

--lO9i3Q~i . ~ .

Based on the successful trial weld, pilot production was instituted. ~owever, upon sectioning the initial pilot welds internal cracks were found in the heat affected zone of ". , the base metal (HAZ zone); a typical example is shown in Fig. 2.

Cross weld tensile tests (.505" dia. bars) on the weld shown in Fig. 2 exhibited zero ductility, Fig. 3 and TabIe 2, but microstructural examination did not reveal any -., ., - .-. - ~
obvious structural embrittlement i~ the base metal ~AZ which would account for cracking. These microstructural obser~atlons ¦--were confirmed by taking smaller tensile samples (.242" di~. bars) ¦
from crack-free regions of the heat affected zone and ~he weld J
metal. The tensile tests exhibited very good ductility values ~see Table 2) which clearly ruled out the possibility of a heat embrittled microstructure as the cause of the cracking, ~er~fying that we were stabilizing the austenite in the weld by means of ~ ;
nickel in ~oth the weld wire and guide tube, and by molybaenum ; in the base metal. I

,. ., . - ~
Cross Weld Tensile Tests (.505" Dia. Bars) Lab. No. Y~S. (PSI) T.S. (PSI) El.% - R.A.

AU-665-1 37,200 37,900 0 2.3 AU-605-2 36,600 38,~00 0 2.3 Results of Tensile Tests Wlth Small Dia. Test Bars (.252" Dia.~ ~

~ab.No. Location Y.S. (PSI) T.S. (PSI) El.~ R.A.~ i ., ~ I
AU-665-A2 Weld* 59,160 112,500 38.0 37.0 AU-665-Bl H.A.Z.** 57,000 107,800 36.0 36.4 *Tensile bar taken from all-weld metal.

**Tensile bar taken through fusion zone on the fine grained side of the weld.

!, .

1(~9i.30~;

- Since the cracks were always located in the coarser grains, further test blocks of base metal were produced with a controlled fine grain structure to determine whether grain size and cracking were related. Additional weldments ~rere produced with these fine grain blocks. Upon sectioning, internal cracks were again revealed, except now the cracks ... - . I
were present on both sides of the fusion zone; an example is shown in Fig. 4.
It will be noted from Table 1 that the amount of phosphorus in the base metal involving the successful experimental weld was 0.028 and 0.020. On the other hand it was found that pilot welds, Fig. 3, was performed on a base having 0.~38 phosphorus.
Based on the above, it was suspected that ~he poor high te~perature strength of manganese steel was the problem - and high temperature tearing was occuring.
Mhen an electroslag weld solidifies, there is considerable contraction and both the weld ~eposit and super- !
heated base metal must possess sufficient strength to withstand the high strain generated by the hindered contraction, other~ise hot tears will result. Contraction is hindered by the mass of the parts being joined. .
In thP case of manganese steel, higher than norma} -strains can be generated during solidification of the electroslag weld due to the high coefficient of thermal expansion of manganese steel ~deemed to be 1-1~2 times that of ferritic steels). In addition, much steeper temperature gradients will exist causing a higher strain concentration because of the low thermal conductivity of manganese steel (about one-sixth that of pure iron).

~` 109i30~i This situation is further aggravated due to the relatively poor high temperature strengths of manganese steel especially as the phosphorus content increases. In manganese steel castings with phosphorus contents above 0.06%, phosphide eutectic envelopes can be observea at the grain boundaries which drastically reduce the high temperature ductility of the steel.
~- It has been postulated that at elevated temperatures the eutectic is either soft or completely molten ana if a stress i5 applied whi~e in this temperature range the grains can easily separate wherever the envelop~ exists. Even below .06% phosphorus, .
where the phosphide eutectic is not visible with an optical microscope, the propertie~ o~ ~anganese steel are adversely affected by phosphorus.
Since the aegree of hindexed contraction and the te~perature gradients are more severe with an electroslag weld, increased susceptibility to hot tearing (and a lower tolerance for phosphorus) can be e~pected in the heat affected zone.
Therefore, after specifying an upper limit for phosphorus of 0.025 to 0.035 in the base metal, three additional 20 welds were produced on a pilot scale, using the stainless steel guide tube and the weld wire identi~ied in Table 1. All three welds were tear-free and excellent ductility values were - -obtained as shown by Table 3:

., . I

-` lO9i3Q~;

~:: . TABLE 3 ' `~ ~UTO!~ATIC ~LDING OF MANGANESE STEEL

Test Welds No's 49, 50 and 51 I. Base Metal:

Heat No. C% Mn% Si% Cr~ Mo% P~ Al~

74-018 0.79 13.20 0.52 0.64 1.02 ~029 .OgO
._'.;''J- -' - II. ~Jeldin Parameters: -Thickness of Test Plate: 2"
Number of Electrodes : 1 -Ocillation Distance : 1.5n Root Gap : 3/4"
~lectrode Wire : 1/16~ Dia.
Electrode Guide Type : 304 SS
Current : 400 Amps Voltage -: 37.5 Volts III. Cross-Weld Tensile Properties t.505" Dia.~ -Yield Tensile SampleStrength Strength No. ~PSI) ~PSI) El.% R.A.~ -49-A S6,640 101,300 25.0 32.9 -B 56,400 105,700 26.0 26.5 50-A 54,600 113,200 35.0 37.0 ~, -B 58,200 110,800 31.0 28.9 51-A 56,750 117,740 37.5 38.2 -B S6,880 117,200 37.0 35.0 -C 56,400 114,500 37.0 37.7 Two additional welds.were produce~ at a hi~her ; amperage ~450 instead of 400 amps) but both exhibited tears in the ~AZ of the base metal. Thus, it was concluded that 2"
section manganese steel plates could be successfully joined ~ith the electroslag process by operating at 400 amps using a base metal containing a restricted amount of phosphorus.

, , ~ 0913~6 A small test sample of S/64" diameter weld wire was produced with a lower carbon content to provide 0.60% C
instead of 0.80% in the weld. This change was made in order to lower the yield strength of the weld deposit to more closely match the yield strength of the base metal; also a lower yield strength would help to reduce the tearins susceptibility by allowing easier deformation of the weld during cooling and -thus promoting a better distri~ution of induced strain across the total weld joint.
For the initial weld trial with the lower strength ire r low phosphorus tand low aluminum for reason~ explained below) were used and the electrica~ parameters were maintained at 400 amps and 38 volts. A vexy good wela was produced with excellent cross-weld tensile properties.
! In order to determine whether higher amperages could be tolerated with the combination of lo~Jer strength wire, low phosphorus and low aluminum, a weld was produced at 500 amps.
This weld again provided excellent results and it was thereore concluded the lower strength wire was also acceptable. ~ater - 20 tests with a 3/32- diameter wire gave equally good results,

3/32a diameter wire being easier to produce than a 5/64" wire.
, Further experiments established it woula not be .
; ~ necessary to oscillate the weld wire while producing the weld, but these s~me experiments revealed that high residual aluminum in the base metal (0.100%~ was producing an effect.
Thus, when some tears were found in the HAZ of the base metal, the only noticeable difference between this and previous tear-fxee welds was a high residual aluminum content .100%?. This fact, together with other inaications that ' ' , . :, .

9130~
aluminum could be affecting results, prompted production of a heat of test blocks with both low phosphorus (.026~) ana ` low aluminum (.020~).
Seve~al weldments were producea using these low aluminum test blocks ~no oscillation) and these welds were found to be tear-free and exhibited good cross-weld tensile - properties.
In order to study the influence of aluminum in ~ore detail, a series of test block (base metal) heats was produce~
at three different aluminum levels: high aluminum (analyzed at ~.14%), low aluminum (.031~) and zero aluminum (~.01%).
All the test heats contained less than .03~ phosphorus and the 3/32" diameter wire at 400 amps and 38 volts was used for each weld.
! Upon sectioning the ~Jelds; tears were observed in the ~Z of the welds produced with high aluminum test blocks.
~Jo tears could be observed in the ~Jelds produced with the low or zero aluminum test bloc~s. Bend tests were performed on ` each wela and the results are sho~m in Figs. 5 and 6. Tensile data on low and zero aluminum welds are shown in Table 4:

Cross-Weld Tensile Test Results ~leld No. Y.S. (PSI) T.S. (PSI) El.~ R.A.
71-A 53,160 99,800 30.5 36.3 -B 51,960 103,300 35.0 35.0 -C 5~,~80 100,600 30.0 35.4 72-A 56,760 110,700 39.0 39.8 -B 56,400 115,000 43.0 42.2 ~ ' , ' . .

1~J9130~ -Another weld was produced with the low aluminum, low phosphorus test blocks but in this case the current was raised to 600 amps to determine if higher amperages could-be tolerated. Reasonably good bend test results were obtained but some small tears were evident in the base ~see Fig. 7).
The above data confirms that aluminum does influence the hot tearing tendency.
Good manganese electroslag wel~s can be produced under the foll~wing conditions:
a. Not more than about 0.035% phosphorua in the base metal;

b. Using only small amounts of aluminum for deoxidation of the castings,`so the residual aluminum after deoxidation ' is not more than about 0.05 - O.OÇ~;

., ; c. Using the lower strength wire; and ~ ~
.. - .
d. Maintaining a maximum current of 5Q0 amps.

" The process in practice is shown schematically in Fig. 9~ The two sections of base metal are separated and the sides are closed by a pair of mQlds such as graphite or water cooled copper shoes. The guide and weld wire are disposed in ., .
the gap betwe2n the base metal sections. The weld wire andjor guide tube is used as an electrode! while grounding the base metal sections to establish an electrical couple. In the course of perfecting the weld the weld wire is fed at an appro~riate rate. ~ continuous slag cover is maintainea by adding flux from time to time~

109130~;

P~ATE OF WELD ~ETAL DEPOSIT; CO~POSITE GUIDE TUBE

Having determined the effect of phosphorus and aluminum, the preference for molybdenum in the base metal and the chemistry for the weld were, welds were tried on a production scale. Two conclusions emerged: (l? embrittlement and tearing are a function of the rate at which weld metal is -deposited; and t2) a composite guide, characterized by a copper guide tube inside a stainless steel sleeve, is re~uired fo~
long welds.
Considering first the requirement of a composite guide it was found that in the instance of long seams of the character shown ~t lS in Fig. 8 the stainless steel guide tube distorted.
A bent guide tube results in erratic arc behavior and uncontrolled weld metal deposit. It was reasoned that distortion was caused by prolonged exposure of the guide tube to high te~perature. The problem was further complicated by ~he fact that stainless steel was desirable as a guide not only because of its compatability with the chemistry of the weld wire, developed after considerable thought and experiments, but also because of its role in maintaining the austent~c character of the weld deposit.
Nonetheless the in~tial use of a carbon steel guide~
was reevaluated. It was ruled out because a minimum diameter of 1/2" was required for an eisht foot weld seam ~nd a diameter of that size would virtually monopolize a 3t4" gap bet~een ` sections to be welded, causing arcing.
If the stainless steel guide could be insulate~ from the effect of the electrical current this would diminish t~e 1~
, lO9i30f;

distortion problem. This ~as achieved by a decision to use a stainless steel tube as an external supporting sleeve for a copper guide tube, the weld wire in turn bein~ centered in co p,~e r the ee~er guide tube. Copper is much weaker structurally than steel but its electrical conductivity and thermal conductivity are vastly superior. Also, copper does not adversely effect the achievement of austenite; indeed copper encourages austenitic stability. '-The composite guiae thus developea (e.g. a 1/4"

; 10 diameter copper guide tube inside a 3/8" stainless steel support sleeve) performed admirably. The copper tube employe~
as the electrode easily carried the current and its structural~
. .
~eakness ~Jas obviated by the steel sheath.

The proposition of controlling the rate of deposit ithin limits, and verification of the effect of phosphorus " and aluminum, emerged during a trial run of ~elds on a : production basis~ The data are set forth in Table 5.
, '.
;' ' ' . ' -' ' . - -' ~

.. . .

.19 '' lOgl3~

,. N N q _I L~) L~ Lq ~ I~ I~ Q L'~ e~ D l~J 1~ 1( --I
q e~ ~ q ~ 1~ Ul ~ q N N N N N .~ I ~I C N
~P O O U~ 1 u7 IJ~) Lq a~ q ~ ~ O N L'~ I q ~ Q
e~ ~r ~ ~ ~ c~ 1~ q N N N N N N N N _I ~1 ~1 ~ q 1l~ lo ,1 ~ N N Lq ~r ~ 1~ O
q rq I ~') Q ~ ~ N q ~ N N ~ 1 N L~ Lq ~ N
~ I~IIItIIIIIIIIIIIItIo d?~ a~ ~ O C o ,1 ul ~ o N C~ ~q o ~ ~r N _( ,I L~ N o q, ~ ~q ~ N N ~ ~
O Lq ~ rq ~ ~ rl ~ ~ ~D CO ~ N
o ~ ~ _I ,J C _I O G _I o o _I O Lq Lq C~ N ~1 ~ ~ Lq N O Q
tq o ~ Q C~ D Lq Q t~ U~ Lq 1`
? t,7 ~,? ~ ? ~1 X ~ ~ J O O G o _i O a~ C~ C Q r~ Q ~D ~ ~ L'~ L'') ~o ~ e ~ ~ 15 C r~ C
': '' O Q~ ~ ~ C) V ~ O C~ G) ~
:. O 1~ U~ `J ~ Cq L'~ Q U~ Lq C~ O Lr) ~ C~ ~r N Cl C`. ~ e .~ Q ~ ~ Q Q
V~ O Lq Lq Lq ~ L'~ Lq In L'l Lq L") Lq Lq ~ U~ L'` Lq Lq L~l C' Lq L'~ L'l ~ L~) ~1 '~1 ~1 ~ ~1 _I -rl '.1 ~1 ~rl I L'l L') Lq ~D Lq L~l Lq ~ q N ~ ~ `J ~ t~ ~.D C~ I Ll tq ~ ~ L L7 L ~ L7 U~ . .

i~ ~ ~ t U r~ L~ O n tq G~ O ,,~ CN _I ~r ~ ~N1 n ~ t~ Q Q _I Lq Lq ~ I` t-- 1` 1` t ~ tn tn ~ ¦ u7 In U~l Lq 1~
--~ 1~ 1~ 0 t~ 1~ t~ r-- q 1` 1~ L' l S~ Q I-- L") ~ Q L'~ Lq Lq tD 1~ ~-- 1~ Lq C7 1~ r- 1~ ~ 1~ tD D ~
~ ~q tq tq tq tq tq q tq tq tq q tq tq tq tq q q r~ t~ q tq tq tq t~ ~q tq tq tq tq tq tq tq tq tq .
0 0 0 0 0 0 0 0 0 0 0 0 0 Lq L'7 0 Lq O O O O O O O O O Lq O O O O O O O
~ Lq ~ ~ U~ ~ ~ ~ e' rq e- ~ q ~ tq ~ q Lq Lq C, r~ N O O O O O O O C
, _1 I o _I ~D o o o I e ~D O ~D ~ O Q O O O O O e' 1~ 0 0 N tD O ~ ~ N N O
51 o o o o o o o o o o o o o o o o ~ o o o o o o o o ~I c _I I o o o o o o .. .
G¦ t N t7 N ~ N O O O O to Oq oN <o O O O O O oq o o o o o o o o o o o o o o o . .
--I /!1 C ~ ~D t ~ D L"~ t~ tD tD O O e~ q o ,-~ ~~ D O N O ~ t~ tn L'~ ~ ~D O _l tq _I
. ~ t~ 0 tD D U~ D W Ln Lq L'l tn Lq L'~ tq. Lq O t~ tq t~ tD t~ D ~ t` t~ t~ ~ Lq Cl C O O O O O O O O O O O O O O O O O ~ i 0 0 0 0 t; ~i o O o~ O _i _i O o O o :
109i3~

. .
. .
.
U~
.. . . 0 ...
. ...

~" o - ~ ~ Gl h 0 01 ~: - ~ 3 `. . - ~ ".............. ~
'~ ' ~ 0 Q, 0 .. - ~ . . " ~ . .

a~ H ~ D

"
3 :i: 3 O O O ~ ~

~1 ~ ~
., , ~, . p, ~ S
R ,, a p, s '~O ''I O o O O ''~ ;~ 30 S 3 -~
' ' ` ~ O ' ~ ~
3 ~ :~ 3 3 0 3 o. ~ 0 ~ 0 .~ O C O O O O C O O ~ ~ ~ 1.1 ~ ~ h ~ v O ~ ~ ~ ~:10 ~ 0000013 ' G ' ` ` ` CJ ~ S S ~.C S
Il~ ~ JJ ~ v ~ v ~ ~ v ~ O o o o o a~
~ ~ S S S
:: ~ ~ $ ~o ~ ~ ~ v v v v ~ ~ ~ s s s s s ~ s ~ ~ s s s s E ~ b E ~ ~ ~ ~ 0 ~ 0 CJ t I I I I I I I ~ ~ a ~ 0 0 S S S S
S 3 3 ~ ~ 3 ~ ~ ~ S S S S S S S S S S
l O O 0 0 0 0 0 ~
O O O O ;~ E~. O _ ~ ~ ~ ~ ~ ~ ~ ~ -- 3 ~ ~ _ S

.
~ ~ W r ~ u~ O cr~ o ~ o ~ o ~ I~ ~r co c~ o ~r u~ ~D ~ ~t .

osl~nt,; ., Based on the d~ta in Table S, the upner limit of 0.03S% phos~horus is verified and aluminum should be lLmi~ed to 0.05 to 0.06%, nominal. The electrical parameters m~y ~e ~ariea and the gap adjusted for optimum conditions as long as the rate of deposit is in the range of 0.6 to 0.9 inches per ;~ minute. A slo~Jer speed incxeases the chance ~or embrittlement;
a higher speed encourages hot tears because of the ther~al gradient being too steep, even though embrittle~ent may not ; occur. The smaller the gap the faster the rate o~ aepos;tr and vice versa, other conditions being equal. Therefore, tlith a given electrical rate, the gap is adjusted and a feed rate ~or the wire is selected, which will result in a weld metal de~osit t?ithin ~he limit o about 0.6 to 0.9 inches per minute~
Evidence shows that moly~den~m when incor orated in the base metal may be as low as a nomiIIal v~lue OL 0. ~ to 0.6 ; This application is a division of copending Canadian application Serial No. 256,363, filed July 6, 1976.

.. . .
'' ' - ' , ' ';

Claims (3)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. In the method of joining base metal sections by electroslag welding, in which a weld wire is fed into the gap between the sections, the steps of centering the weld wire in a copper guide tube employed as the electrode and supporting the guide tube inside a sleeve of stainless steel.

2. The method of claim 1 including the step of adjusting the gap and feeding the weld wire at a rate such that weld metal deposit progresses at the rate of about 0.6 to 0.9 inches per minute.

3. The method of claim 2 in which the base metal is manganese steel containing not more than about 0.035%
phosphorus and not more than about 0.05 to 0.06% aluminum.