Gerald,
thanks for mentioning me (I hope it was me - Stephan - who was meant?)
Believe me, it's hard for me to currently not finding sufficient time to participate in the forum! So much interesting topics but the job requires his tribute twice at the moment.
However, this particular topic is a toffee. I would like to give a little comment on it - although everything that should to be treated has already been treated. What else been left to say than: Wow!
Basically from my rather limited practical experience in processing stainless steels and from what I have tried to learn on the theoretical coherences of those interesting materials, I would like to agree with Jeff et al who say that any kind of generalization is hardly to obtain since the tremendously variance of the different alloys + that "infinite" one of the peripheral variations makes such a kind of effort - from my point of view - impossible.
Therefore it is surely good that a limitation on specific materials and processes should be executed, just as you have said e.g. 304 + SMAW/TIG.
Well, when I have read the initial post coming from firstpass I have honestly asked myself what could basically be the reason for "cooling" stainless steels with water. A precious kind of clarification and understanding I could gain when I have read the case, Al has stated. Using water as a heat sink for saving sensitive areas of the part... Once again very impressive, once again very "known-how" - even Al (803056)!
But however, although I have reconsidered the topic several times now, it stays unclear for me why water should be used for cooling the material while it's being welded or between welding the multiple layers. What could in general be the reason for doing so? Affecting the morphology of the weld metal deposit? Or influencing the morphology of the heat affected zone? Whereas the first could be realized "easier" (at least from my standpoint) by using specific shielding gases, the latter - namely a wanted influencing of the heat affected zone - is more imaginable for me as for being the reason. But, what could in detail being wanted when the «Heat Affected Zone» should be the reason for cooling stainless steels with water? The Ferrite -- > Austenite transformation? And if yes, what could be the reason then therefore? Improving the corrosion resistance by rising the ratio of austenite? Or decreasing it for e.g. improving the material's hot cracking susceptibility? Hmmm, some questions which haven't been responded up to now, at least as far as I interpret the answers on firstpass' question. Perhaps firstpass may kindly provide some more information for the actual background of his topic? This would be very interesting - at least for me.
Forgive me but what interests me first in general when I read something new - and thus being allowed to learn something new - is, "What might the reason be for that the author has asked even a particular question or for having investigated something particular?" It's even me, Stephan, and you fellows know me, to "understanding" something I must "penetrate" it...
Well, then I have read what Jeff has stated and in one sentence showing in the direction of increased cooling rates by using High Energy Density processes (LBW or EBW). That was the point in time when I have remembered something I have read a time ago and what I felt could contain something in a direct coherence with what has been said on "Solidification" a/o "Delta Ferrite Content" a/o "Corrosion",...
You may see, it shows a very huge agreement to many statements been made in the thread up to now. I personally mean that it should be sensitively considered what kind of "stainless steel parameter" should be the subject of interest. Corrosion, Strength, Fatigue,... Each of those is depending on very particular and intricate mechanisms taking care for the base materials microstructure. Just as already been stated here several times. Quite little changes or variations in only one of the base material's, -process, consumable-parameters are able to change those properties i.e. "Corrosion Resistance", "Mechanical Properties",... again, depending to the mentioned above. I guess this is what it makes it so hard to make even a generalized "Yes" or "No" predication. And as I could see by reading what Marty Sims et al have stated, cooling stainless steels (at least specific alloys) can also be accomplished without any negative effects (under specific conditions of the later on material usage).
Thus the only little - and humble - addendum I would like to carry out is that I would like to report about a scientific survey accomplished in Germany and coming from our Federal Institute for Materials Research and Testing (Bundesanstalt für Materialforschung = BAM) located in Berlin, and which I remembered when I have read what you all have stated.
The - relatively current - investigation (finished in 10/2006) I am speaking of, was headlined (translated) "Survey on Solidification Behaviour and Weldability of Austenitic Steels Using Laser- and Hybrid Welding". It was executed basically to clarify the influence on high energy densities while welding (similar with increased welding speeds) on the mechanisms of solidification and an emphasized view on the "Hot Cracking" Susceptibility of austenitic (metastable and stable) steels. The survey can be seen as a conclusion of what even Jeff has mentioned (processes), welding base materials as "firstpass" has stated and having a specific view on the morphology of what you, Gerald, have mentioned. By having that fine combination I hope to be permitted to add the following information and thus giving you a small summary of what Germany's highest administration in materials research has found out.
They have surveyed 4 different base materials of what I would like to list the German Standards designations subsequently and referring those to the US-American AISI designations. These were:
· X15CrNiSi 20-12 ~ AISI 309
· X2CrNiMo17-12-2 ~ AISI 316L
· X2CrNiMo 18-14-3 ~ AISI 316L
· X2CrNi 19-11 ~ AISI 304L
The wall thicknesses were
· 4 mm ~ .15"
· 5 mm ~ .19"
· 6 mm ~ .23"
They have used:
· Laser Welding
· Laser GMA Hybrid Welding
· GMA Welding
as the welding processes, to compare the broadly used "low energy" process "GMAW" with "high energy" processes as "Laser" and "Laser Hybrid Welding".
The Joint geometries were prepared for:
· LASER: Square Groove Butt Joint
· Hybrid: V-Groove Butt Joint (Bevel Angle 10°)
· GMAW: V-Groove Butt Joint (Bevel Angle 50°)
In the following I would like to "extract" what the people in Berlin have investigated. I mean that only the pure Laser Welding shows a kind of comparability to the effects of a rapidly increased cooling rate by using water, just as Jeff has mentioned already. Thus the emphasize on pure Laser Welding should be accomplished and not having a further view on the other welding processes.
Nonetheless for a better and more basically understanding of how the different mentioned and investigated processes do vary in both "Heat Input" (you know my very critical attitude on this technical term) and "Welding Speed", I would like to consider the following. Considering that higher welding speeds lead to steeper temperature gradients. What does this mean actually? A specific amount of energy affects a specific amount of material for a specific amount of time. The higher the welding speed at a given amount of energy the less the time the amount of energy is affecting the material. This should correlate physically to a reduction of heat input. Let's have a closer look if this consideration can be proven, see also the Figure Heat_Input_over_Welding_Speed. I would say "Yes" higher welding speeds lead to lower heat inputs. But what occurs when increasing both, energy density and welding speed - as being executed in the beam processes. Here a relatively high energy density leads to relative limited isotherms which means that small sized welding beads can be generated having a relatively high Depth/Width ratio. This again depends of course on the thermodynamical properties of the base material to be welded. I.e. to achieve a similarity in the isotherm's form in welding different materials, having different values of thermal conductivity, one has to change either the energy source performance or the welding speed. Regarding and founded on the specific and well-known thermodynamical properties of "stainless steels" (in particular "low" heat conductivity) this means, that the heat is dammed up in a relative narrow area around the energy source. When now having a view on the "Heat_Input_over..." Figure one can see that - for comparing what firstpass has stated - the lowest "Heat Input" values were reached by using the pure Laser with high welding speed values. Presumed I am right by thinking that the value of heat input must be low when cooling the stainless steel by using water while welding and thus generating a steep temperature gradient between the limited area the heat source couples in its energy into the base material and its adjacent heat affected zone, this should correlate to the conditions when using a high energy density heat source (Laser Beam) and having a high welding speed. Therefore the cooling conditions should be comparable - in general principle. A small but from my point of view important detail whereas is, that e.g. an arc as an energy source is quite different to a Laser Beam, due to having a quite reduced energy density + a lowered welding speed as well what yields a quite different - since lower - D/W-ratio. But this detail should be neglected subsequently, since I mean to have understood that herein should be considered only the condition of a steep temperature gradient by using water for cooling the material while it's being welded.
Well, we should now talk about using this pure Laser for welding a particularly base material. Since what Jeff et al have mentioned it is hard to give a general prediction if a steep temperature gradient, i.e. drastically increased cooling rate, can negatively influence the base material. But however what could then be the reason for that there is being asked a question as firstpass did? What could be - if even - the reason that there could be a significant negative effect by using highly accelerated cooling rates by cooling the steel with water? Or perhaps water has on the contrary a positive influence on the material's properties which ever those were. Hmmm....
But now finally I would like to start to try to explain what the German people have found out.
First they have made sure to have a variation of different base materials, as to be seen above. What we can see there are different mostly metastable stainless steel grades. What they wanted to find out was the answer on the question how the welding process and its peculiarities can affect the mechanical-, corrosion- and hot cracking behaviour of the welded base material. For a kind of separation between the base materials listed above they used basically the specific Chromium equivalent to Nickel equivalent (Creq/Nieq) ratio, which is often used as a criterion for estimating the hot cracking susceptibility of the base material. Furthermore a separation between the different modes of solidification was carried out by using the WRC-1992 Diagram, see also the WRC_Diagram.jpeg. It is well-known that the basis of the WRC Diagram is the mode of primary precipitated micro structure's constituents. This means the field of primary ferritic solidification are abbreviated with "F". These materials are e.g. ferritic stainless steels or metastable stainless steels. The Primary_Deltaferrite.jpeg shows by the way such a primary ferritic solidified microstructure (white vermicular areas within the dendrites).
Those materials, solidifying primarily as austenite, are abbreviated with "A". These ones are e.g. fully austenitic or "stable" austenitic steels, containing very low amounts of Deltaferrite. A primary austenitic solidified micro structure besides an ferritic solidified area shows the figure Primary_Austenite+Ferrite.jpeg. Both ranges "A" as "F" are more or less and strongly simplified "clear" in regard on their properties and thus more or less relatively safe predictions can be made for their behaviour after welding.
But there are some additionally areas recognizable in the WRC-1992 Diagram being abbreviated with "AF" and "FA". Those areas are marginal with respect to their primarily solidification behaviour.
What does this term "marginal" means actually?
Well, let's have a look upon the mentioned areas "AF" or "FA". Both areas have a specific "preferred" primarily micro structure solidifications. Whereas the materials falling under the range of "AF" having a primarily austenitic solidification, the "FA" materials have a primarily ferritic solidification, i.e. these ones solidify as "Deltaferrite". The main common property between both whereas is, that after the primarily solidification has taken place an additionally eutectic solidification occurs. Here both phases - Ferrite and Austenite - solidify simultaneously -even eutectic - in between the interstices of the weld metals primary dendritic microstructure. And now it comes, when having materials belonging to the "FA" range and thus solidifying primarily ferritic under "normal" conditions, it can occur that by increasing the cooling rate a contrary solidification, i.e. austenitic, can take place. This is called in German language "konstitutionelle Unterkühlung" and if I would be demanded to translate this I would write "constitutional undercooling"(?). This mechanism makes sure that there is a kind of decomposition of alloying elements at the front of solidification reducing the liquid temperature of the alloy and making sure that an eutectic composition is obtained to solidify simultaneously as Austenite + Ferrite. This again reduces the amount of Ferrite which is being well-known for its positive influence in regard to reducing hot-cracking susceptibility. This means in fact that increased cooling rates at specific base- or filler materials can have a negative influence by reducing the Deltaferrite content and thus can increase the hot cracking susceptibility.
Subsequently we have to talk thus about the pure Laser Welding, since this is the only process - of the surveyed ones - reaching high cooling rates and having no kind of affecting the primary microstructure by influencing it by using an additional filler metal. Furthermore no view should be put on the other investigated issues just as Corrosion resistance or mechanical properties (Impact Strength). Only the influence of rapid cooling should be considered afterwards as having a negative influence on the austenitic stainless steel. To evaluate the range of where a "critical" Creq/Nieq-ratio can be expected when welding austenitic stainless steels and achieving high cooling rates (by using Laser Welding), specific tests were conducted by using the base material X2CrNiMo 18-14-3 (~ AISI 316L). By having a composition of:
- .02 C
- 17.08 Cr
- 12.76 Ni
- 2.585 Mo
- 1.035 Mn
- .397 Si
- .021 P
- .002 S
- .034 N (all contents in %)
it does have a Creq/Nieq-ratio of 1.39 (according to the WRC-1992 Diagram). Therefore a primary austenitic solidification can be expected from the liquid phase. This was also the case as having been found out. But on the other hand this composition of steel lays adjacent to the borderline of austenitic-ferritic solidification ("AF"). When following the mechanisms as treated above, one now should watch out, when welding e.g. this specific base material (or comparable to that) using high energy density processes, obtaining thus high welding speeds and... low heat input values, which - as we have heard - are at least theoretically comparable to increased cooling rates by using water. Then they have investigated what kind of variations can deteriorate the hot cracking resistance of a base material quite similar in chemical composition to the first named. Therefore they have used a base material (same batch) but having the following composition:
- .021 C
- 17.6 Cr
- 12.75 Ni
- 2.513 Mo
- 1.251 Mn
- .553 Si
- .021 P
- .005 S
- .037 N
Only slight differences in composition but yielding a different Creq/Nieq ratio, which was: 1.41. What they could then find out was, that by using the Laser as a heat source and holding the other parameters (Welding speed etc.) constant, a ferritic primary solidification could be observed, but, showing a constitutional undercooling and thus leading to partial austenitic areas where the thermal conduction was at its highest values. And now it comes somewhat interesting. By having just a slight increased sulphur content, which should normally be no significant drawback in using "low energy heat sources" as an arc (enabling a primary ferritic solidification), an increased hot cracking susceptibility could be observed when using a "high energy heat source" as a Laser beam inducing high cooling rates and leading to constitutional undercooling sequences. Hereby one can see, how critical some specific stainless steels can react on specific variations with particular regard to increased cooling rates. Only slight changes in the chemical composition in particular respect to the hot cracking susceptibility increasing element sulphur can under specific conditions induce inhomogeneities. Please see here also the figure Hot_Crack.jpeg, showing a liquation hot crack in a base material similar to that mentioned above. This shows that increased cooling rates in coherence with the primary solidification mechanisms can have negative influence on the material's properties. Of course these can likewise change and vary by using e.g. filler material to influence the composition of the weld metal deposit and thus "compensate" the negative effects again by increasing the primary Deltaferrite content, but by using only the pure beam as the heat source and thus having "only" the base material as itself, can create problems. By having found out that there are strong interactions of some main parameters affecting the steel, the German scientists have tried to find an in general coherence between the parameters:
· Creq/Nieq-ratio
· Welding Speed and
· Content of primary solidified Ferrite
Therefore they have calculated or evaluated respectively, a kind of "significance-diagram" showing the coherences between those three parameters as listed above. Please see also the Deltaferrite_Dependence.jpeg.
Well, although I mean to have understood that firstpass' topic was meant by cooling the stainless steel between the different passes were welded and thus the descriptions I have listed above were not quite transferable since they treat the period where the liquid metal solidifies to its primary microstructure I however hope that it could be proven by that what you, Jeff, Al, Marty Sims and all the other appreciated fellows have mentioned as well.
Thus I would like to finish this reply by concluding as following:
· High cooling rates (e.g. in Laser- or Laser Hybrid Welding) and Creq/Nieq-ratios of < 1.55 can form larger amounts of primary solidified austenite.
· Simultaneously presence of higher amounts of hot crack sensitive elements as Sulphur or Phosphorus can increase the risk of achieving hot cracking. Therefore it is recommended to use only very low limitations in these element contents when welding austenitic stainless steel by using Laser Welding or welding processes generating high cooling rates.
· It has always to be considered that both joint- and constructional design of the part to be welded do have a significant influence on the hot cracking susceptibility.
· Appropriate filler metals can improve the conditions due to having the ability to affect the primary microstructure solidification in a positive way by displacing the microstructure from austenite to increased ferrite contents.
That far my humble contribution on that interesting topic!
Best regards,
Stephan
P.S. For further interesting information please visit the BAM Homepage under
http://www.bam.de/index_en.htm