although there are great replies coming from Darren, Phil and of course John Wright, I truly hope you are furthermore interested in some details of what Manganese steels are and some specifics of welding these interesting materials. I hope I will find the right words for my thoughts.
By knowing a little more about austenitic manganese steel in regard to its manufacturing and utilization of the components been made of this base material one can derivate alittle the specifics of welding these materials.
In Germany we use the term "Manganhartstahl" (hardly to translate but comparable to "Hard Manganese Steel"?) for the material you want to weld. Historically this material has been named firstly by R.A. Hadfield (Journal of the Iron Steel Institute Vol. 34; 1888/II; page 41-82) in 1888. In term of metallurgical aspects this iron-based "steel" is actually a "cast" alloy, containing 1.2% Carbon (C) mean and 12% Manganese (Mn) mean and is being rolled or forged after casting.
The solubility of carbon in iron rises with an increasing manganese content. In case of austenitic manganese steel therefore the mechanical properties of the material do not vary significantly - when holding a C/Mn-ratio of 1:10 constantly - up to a C-content of 2% and Mn-content of 20%(!). Due to this increased carbon-solubility at higher temperatures (> 1000°C) the materials microstructure is nearly entirely free of very hard and thus brittle metal-carbon-compounds - called carbides. I suppose herein a little knowledge of the iron-carbon-diagram. Manganese works here as an austenite-area extending element. Therefore one is able to achieve no hard and thus low ductile martensitic microstructure after quenching the steel, but a very stable and ductile austenitic microstructure. This is also called "homogeneous" austenite. Mentioned by the way, the martensite starting point (Ms) of 12% austenitic Mn-steel, i.e. the temperature martensite is firstly being created, lies at approx. -180°C(!). The austenite on the other hand is, as already mentioned, very ductile basing on metallurgical reasons not to be treated further herein. But it has been reported that the austenitic Mn-steel could achieve strengths of 700... 1050 Mpa at elongations of higher 40%(!). In exceptions (very accurate specimen preparation) values for elongation of > 80%(!) could be achieved - from my point of view austenitic Mn-steel is even one of the most impressive materials can be welded. Hardness after quenching is ~ 220 HB (Brinell) what is interesting, since with this relatively low value of hardness it should be difficult to believe the term "Hard Manganese Steel" or "Manganhartstahl", respectively. Actually the material impresses (me) also in other views. For instance, there is normally a numerical coherence between the Hardness in Brinell (HB) and the materials strength (lets call it "Sigma") which can be expressed by:
Sigma = 0.35 x HB
This expression does not count for austenitic Manganese steel and it could be proved that the values been measured in tensile testing do not correspond to the theoretical values been calculated for the material. Now I am where I wanted to go when I started writing this reply - at the most outstanding property of the austenitic Mn-steel - its »strain hardening ability«. "Normal" ferritic-perlitic steel materials showing a defined yield-strength followed by a defined peak load in tensile-testing. Austenitic Mn-steels do not show this behaviour. Instead of this, the material is elongated uniformly over its entire length(!) with increasing the tension, what is equal to an increase in its hardness. The point of time of fracture of the specimen is not calculable and depends on stochastical factors, what is also the reason for the impossibility of using the expression of Sigma and HB, mentioned above. But what is important is, that due to the fact of strain hardening one can achieve a high hardness of the material - although the material is basically very ductile - only by charging the material with tensions. It is also important that the carbon must be resolved within the solid solution of the material and the austenite has to be homogenous. Furthermore to achieve a hardness increase of the material the tensions to be charged have to be positive or compressive stress, respectively. And where can compressive stresses be found? Everywhere where the material is being charged promptly. Thus it can be understood what kinds of components are being fabricated from austenitic Mn-steel:
· Cone crushers
· Excavator parts (teeth etc.)
· Cross pieces (railway tracks)
· Stone compactors
Such parts and components are being delivered - as far as I know - mostly in forged or cast form. By charging now the materials surface by compressive strengths (rolling, pushing or beating) the surface hardness can rise up to 550 HB(!). Thus you achieve a real impressive material, ductile in its core and hard on its surface. But caution! Do not use the material for grinding strain applications - it will fail, since the surface will not be hardened.
Now after the elementary issues regarding the material properties let me please make a short skip to the most interesting - the welding of the steel. I hope very much that some information mentioned above may show the first direction for how to handle the material in a right manner. We know that the carbon has to be dissolved within the solid solution of the base material to avoid the precipitation of hard and brittle carbides. This carbon's solution is being achieved by rapidly cooling or quenching the material from high temperatures down to room temperature. If the base material is being warmed up subsequently a precipitation of carbon from the solid solution with subsequent generation of carbides can occur what means that the material will loose its fantastic properties. Those precipitation sequences can be described in so called "Time-Temperature-Transformation-(TTT)Diagrams", since they are - from the standpoint of physics - functions of temperature over time, also for the 12%-Mn-steel for instance can thus be determined the critical temperature the solved carbon starts to be separated from the solid solution (austenite) of the base material. This critical temperature can be recognized at >300°C. Above this temperature and appropriate cooling-times one must expect the segregation of carbon, diffusing from the austenitic solid solution, creating carbides and lead to an embrittlement of the base material.
What occurs over the welding sequence? Yes, the base material is being liquefied and thus the temperature mentioned above is exceeded widely. All areas of the base material warmed up in a too high amount while welding sequence have to be expected to be brittle and are thus cracking- and fracture susceptible. By using the 12%Mn-steel mostly for components as mentioned above I - from my personal standpoint - do not really suppose that one may experience a situation where the base material should has to be joined by welding. That was the reason for my question when I asked what you would like to perform, Joining or Surfacing. Mostly from my personal experience, components as described above have to be surfaced for reconditioning worn out areas on the component. Therefore the subsequent basics can be stated when 12%Mn-steel has to be handled by welding:
· Use small electrode diameters in Manual Shielded Metal Arc Welding
· Use Stringer Bead technique with only 2 x electrode diameter (max.) and avoid large weaving
· Take care for high cooling rates (low cooling times). Quench the welding seams with compressed air or even water.
· If possible weld in a water quench for avoiding a warming up of the seam adjacent areas and zones.
· Weld short seam lengths
· When using GMAW use low arc power (approx. = amperage x voltage)
· Do not use the components at temperature levels larger 300°C
· Slight hammering of the still warm weld-seams induces compression strengths. Take care of direction of hammering (always continuously in only one direction)
Particularly the latter can be important besides the "cold" welding sequence. Hammering can reduce significantly the shrinkage tensions being induced by the materials relatively large thermal expansion coefficient (austenite) and can furthermore lead to a very first compression of the weld-seam. The large thermal expansion of this material in combination with its relatively small coefficient of thermal conductivity is the main reason for problems in joint-welding austenitic manganese-steel. I have been very interested in welding physics of this material because I had to SMA-weld large cone crushers for a flint mill many years ago. And we welded these parts which have lain in a water quench, by only having access to those particular areas of the jaws of the crusher which had to be welded. And thus I had tried to discover at that time if it might be possible to joint-weld this interesting steel also. In 1940 some researchers (K.L. Zeyen: "Die Schweißung von Manganhartstahl" in »Elektroschweißung« (1940) Vol.5 page 78...81) found out that it is also possible to joint-weld 12% Mn-steel by using austenitic stick-electrodes and additional slight preheating of the part to be welded. For the field of surfacing, these researchers recommended a deposition of buffer layers by using a filler-wire comparable ER 307. That kind of wire electrode or filler metal, respectively, should also be usable for joint-welding, what I personally unfortunately con not confirm due to lack of experience. For surfacing operations in the field of SMAW also high-chromium and high-manganese containing stick-electrodes should be usable for weld performance. These consumables have a composition like following:
· C = 0.6 % (Carbon)
· Si = 0.8 % (Silicon)
· Cr = 13.5 % (Chromium
· Mn = 16.5 % (Manganese)
Electrodes of these compositions are being melted in Germany as special-alloys, have an appropriate high price but should have very good mechanical properties hundred percent comparable to the base material properties. As standard alloys also electrodes of following composition can be used for welding 12% Mn-steel:
· C = 0.7 %
· Si = 0.1 %
· Mn = 13.0 %
· Ni = 2.8 % (Nickel)
I case of GMA-welding for the finishing surface layer (after buffering with a ER 307) also wire-electrodes comparable to AISI HNV 3 (X45CrSi9-3) having a composition of:
· C = 0.45 %
· Si = 3.0 %
· Mn = 0.4 %
· Cr = 9.5 %
are recommended to be used.
But regarding choosing filler-wires for welding 12% Mn-steel there have fortunately been already made so great suggestions of the appreciated colleagues who having already answered on your topic, that it would be a sin from my side to add here something more! The only issue I would like to recommend is to consider also the shielding-gas choice, since the CO2-content of the shielding gas used in GMAW may also influence the hardness of the weld-deposit.
Well, likewise here, strongly reduced thermal input into, and fast cooling or quenching of, the base material should be the headline for a proper weld-sequence and sound welds.