This is a very interesting topic. When considering the topic, Steve emailed each moderator to get some thoughts on the subject. I responded and, while some may not agree, below is what I wrote back. Maybe this will break the ice.

I forge weld a lot. But what is it? What is happening? If I had to say in a nut shell, it's the temporary liquefying of the two surfaces along with applied pressure, allowing the two surfaces to mix molecules. The mixing makes them one piece now. Actually, I think it is as simple as that.

It's called a weld. Perhaps a surface fusion. Other than that, I think we're over thinking it.

The molecules are flowing withing each other and when solidifying, they are one. The mixing is greatly enhanced by the excitement of the molecules which happens at welding temperatures.

I do think the base metals mix, yes. Especially the base metals. Otherwise, I don't believe it would hold. How deep? Deep enough to consider it one piece but not too deep to blur the distinction after etched. It may be measurable in that context. Any other seems irrelevant to me.

For carbon to migrate, there has to be a conveyance. That means the metals were compatible and mixed sufficiently for the bridge to carry the carbon.

Time is another factor. Since it is only the surface that has temporarily liquefied and quickly re solidified there is a distinct zone where the two pieces bond. I believe any longer in the heats required for welding, the whole billet would liquefy which, if not contained will slump. If contained, it will mix completely making a new alloy.

In my view, this (welding) just another exercise where we have to balance three factors, Time Temperature, and Technique. Fortunately we are using a material that lends itself well to the operation in a shop setting, albeit in a very controlled setting.

Above are some simple ramblings using your questions as a rough guide.

Excellent explanations Lin! I would have said pretty much the same things. As with many things in Bladesmithing, and especially forge work, I sometimes have to smile when folks beg to know all the scientific explanations, thinking it will give them all they need to perform the act successfully. I think with most, it comes down to learning HOW to do it, before they can actaully understand WHY they can do it. <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//smile.gi f' class='bbc_emoticon' alt=':)' />

I have a question on making san mai. What is the ideal thickness ratios of mild to high carbon steel? We just made some and we used 1/4 inch mild steel on each side and 1/4 inch 5160. It welded together very well. I just wondered if there was a best ratio.

Thanks,

Cal

Hi Cal!

When it comes to San-Mai, I always try to think of the thickness I want for the particular blade I'm planning to build, and base thicknesses on that. In the case of the materials you mentioned, it's not a big deal, as you can simply draw that combo down to the desired thickness. When it comes to other mixes such as Damascus cores with 410 or 416 laminates, then you have to work with pieces/thicknesses that are much closer. That's because types of mixes won't tolerate nearly as much drawing/shaping without problems.

Personally, I wouldn't use 5160 as a core....in my experience, it just doesn't play well with others, and there's no reason to make things any harder on yourself then you have to .... I would go with 1080/1084 or something similar. I suspect that drawing that billet you mentioned out to knife thickness is likely going to cause you some grief because of the 5160. Often time it will delaminate somewhere along the line.

OK, all that being said, I always try to start with pieces that will give me about 1/8" to 1/4" total thickness more, then what I want the finished blade....that gives me extra for grinding off forge scale, pits, etc. Say I want to end up with 1/4" blade (.250") I'll use a core that is .125-.155" and outside laminates of aprrox. .100". I know those numbers don't add up to .250", but you have to take into consideration the compression that occurs during welding, and the clean up after annealed. Each individual is going to be a bit different based on their experience and skill level. My motto in the shop is..... if it's too thick, you can always grind it off later, but if it's too thin, you can NEVER put it back. <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//smile.gi f' class='bbc_emoticon' alt=':)' />

One of the major question I have is, how does one minimize the amount of carbon migration after welding? I have a PF-200 forge and I have lined it with ITC-100 and it holds heat like mad now. However I am having an issue running it below 7 psi. I typacilly weld at 10 psi and have had great results. Thing is I did a san-mai of 1018 and Hitachi blue #2 and the migration line was very bad. When I do 52100 and 410 there is very little migration. I am new to using stainless in this method. Ed I have seen your work and it is truly amazing and your insight has helped me greatly.

I am really worried that this much carbon migration is going to be a mayor issue in the performance or my blades.I am thinking that my forging heat is just too high.

Thanks for the information Ed. I will stick to a simple steel like 1084 for a core. I appreciate the info on the ratio of mild steel to high carbon steel. I know you have a lot of experience making san mai.

One billet we made had a small delamination on the very end. I am cutting that off. Other than that it looked ok. When we get blades made and heat treated I will post pictures. Hopefully in a few days.

Thanks to all the fine bladesmiths that contribute their hard learned information to this forum.

Cal

I just presented about an hours-worth of power point on this very topic in Ontario Canada about three and half weeks ago. Metals, like iron, do not involve molecules. Molecules form with covalent type bonding of compounds, metals are a lattice structure of atoms held together by metallic bonding. The vast majority of steel is iron (metallically bonded) with a few scattered compounds (like Fe3C) that we can hope will not get involved in forge welding since they would only get in the way. In fact, the absence of compounds, especially oxides, is rather important to good bonding. At welding temp carbon will not be involved in carbide making since it will be in solid solution (a mixture not a compound) in the iron and thus not be a problem with the bonding.

Fusion welding is what happens with a MIG, TIG or arc welder, where the liquidus is actually reached. In solid state welding the liquidus is not reached, and without good shielding the steel will burn before getting there anyhow. In solid state welding one relies on temperature and pressure to get the metal surfaces close enough for metallic bonding to commence with the iron atoms, this allows the individual layers to maintain the autonomy of their chemistry. Decarb, oxides and boron infiltration account for the different effects in the weld zones. Examples of how this happens can be seen in incredibly high polished precision steel blocks that will grab each other quite well at room temperature as the metallic atoms begin to flirt with each other. One can also bond metals with enough pressure, very unique metals welds have been made using explosives to provide the pressure to make the bonds. I read in a very old metallurgy text about tests that were done a long time ago where gold plates were clamped to other metals for a few years and subsequent analysis showed gold atoms had diffused into the other metals.

Once metallic bonding is achieved, diffusion of interstitial atoms (like carbon) will readily occur between the two parts, as will diffusion of substitutional (alloying- Cr, V, Mo Ni etc…) atoms but at a much slower pace since they need to actually swap positions with iron atoms rather than skitter between them. Carbon diffusion can be slowed or even stopped by using alloy barriers that do not form readily form carbides, i.e. nickel and silicon to some extent. Bladesmiths should be careful not to wish diffusion away too much at it has saved more than a few damascus mixes from being a disaster. With san-mai the idea is to stick the welds and get done before too much diffusion can occur, or slip a thin sheet of pure nickel between the low carb and high carb.

Carbide formers will accelerate the diffusional effects. Here is a 1000X image of a 416/W-2 san mai weld zone from a series of metallographs I did for Burt Foster several few years ago:

To the left is the carbon rich W-1 that becomes carbon depleted white ferrite adjacent to the weld. To the right of the weld the low carb stainless is already forming heavy carbides and carbon solutions making a brown zone beside the weld. The chromium is having a striking effect on the properties of the diffusion causing the contrast to be much more pronounced.

Damascus loses quite of bit of its mystery if you spend a few years dissecting it under the microscope.

|quoted:

One of the major question I have is, how does one minimize the amount of carbon migration after welding? I have a PF-200 forge and I have lined it with ITC-100 and it holds heat like mad now. However I am having an issue running it below 7 psi. I typacilly weld at 10 psi and have had great results. Thing is I did a san-mai of 1018 and Hitachi blue #2 and the migration line was very bad. When I do 52100 and 410 there is very little migration. I am new to using stainless in this method. Ed I have seen your work and it is truly amazing and your insight has helped me greatly.

I am really worried that this much carbon migration is going to be a mayor issue in the performance or my blades.I am thinking that my forging heat is just too high.

Don't confuse carbon migration with carbon loss due to decarburization.

Carbon migration within the billet will result in a homogeneous block of steel, and that is a good thing.

We often think of carbon as just running rampant through the steel, but in fact, you need to consider this at the atomic level.

From our perspective, you would really have to over-heat for a considerable amount of time your steel before huge loss of performance will take place.

But once you do - it's gone.

I'm not suggesting that we don't lose carbon, we do, but with proper forge atmosphere, this loss zone is mostly ground away when shaping our knives.

|quoted:

This is that same concept only from the outside:

If I were to ask Mr. Peabody and Sherman to set the Way Back Machine to 1979/1980 and my college chemistry class, I would probably find that what happens in a forge weld is a form of Ionic Bonding. In layman's terms, this is a sharing of electrons between two or more atoms causing the atoms to "stick" to one another. I will assume for a moment that everyone understands atomic theory at a very basic level: Matter is made up of Atoms, Atoms are made up of two parts: the nucleus and the electron shield. The electrons orbit the nucleus in the same way the planets orbit the sun.

When you add significant energy (in the form of heat) to the steels, the electrons begin to get "excited" and their orbits expand. Eventually, they expand their orbits enough to become attracted to the nucleus of the atom next to it, and orbit sharing happens. The electrons begin orbiting more than one nucleus forming what is basically called an ionic bond. Now you may think that this would not be very strong, being that electrons are so miniscule, but ionic bonding is what holds all matter together. Basically in a forge weld, you are bonding plates of different metals to one another by getting the atoms to share electrons.

Disclaimer: It has been a very long time since I took chemistry. This is the best my memory has to offer.

It took a while to dig this back up... I must be getting old.

Two versions of the same thing. Suitability dependent on viewers view... =]

Not really specific to the original topic, but it fits with the wander.

http://www.lehigh.edu/engineering/research/undergraduate/symposium/archive/2009/posters/ugs-2009-nizolek.pdf

http://www.asminternational.org/documents/10192/1895783/amp16702p24.pdf/815e5afe-300f-457e-9a2d-9057a4bb538b

Mike

to help my students understand what all the science means in a real application, I use a Venn diagram of three interconnecting circles. One represents heat/energy state, one the pressure or distance of the two parts, and the last how clean the surfaces are. An increase in any one of these factors can make up for deficiencies in another area (ie higher heat or a cleaner surface can make up for less pressure in the joint) up to a point. our job as bladesmiths is to be consistent and use a methodology that consistently gives the results we want, so we do not end up with blade that has welding flaws after 30-50 hours of work (I hate that...) so we need and want to do everything we can to stay in the center of those interconnecting circles.

Another way to think of it is that three factor need to be present to get a good weld, a clean surface , an inert atmosphere in the weld area and perfect contact of the two surfaces. notice that I don't mention heat .. it is possible to get forge welds at room temp given the proper conditions. I remember reading something years ago about wrenches welding to nuts after galling the surface in space, forcing the use of dissimilar materials for wrenches in our space program.

lets take our classic flux weld, the joint is fitted close,(scarfed what have you) heated and covered in flux, the flux will do two things , it will chemically strip the oxides from the surface and bind with them, and it coats the surfaces of the joint, protecting it from the atmosphere. when the joint is pressed or hammered together the flux is pushed out of the joint allowing the two surfaces to come into contact. of course the down sides of a flux weld are first that not all of the flux is carried out of the joint, and second boron can alloy into the steel. (the white line that can show up in a flux weld when etched is most likely boron)

a similar thing is happening in a atmospheric or Oil weld, in this case it the heat is reducing the surface of the steel in the weld joint back to iron with a very shallow decarb line. this forms co2 as if reduces leaving the joint inert, sort of a micro atmosphere between the layers. pressing or hammering the joint together forms the weld, with nothing to stop the bonds form forming very strong clean welds are formed. the edges of the stack around 0.010" or so do not seem to weld I think this is due to a sort of eddy current near the edge pulling in O2 and never quite reducing the surface.

MP

|quoted:

I just presented about an hours-worth of power point on this very topic in Ontario Canada about three and half weeks ago. Metals, like iron, do not involve molecules. Molecules form with covalent type bonding of compounds, metals are a lattice structure of atoms held together by metallic bonding. The vast majority of steel is iron (metallically bonded) with a few scattered compounds (like Fe3C) that we can hope will not get involved in forge welding since they would only get in the way. In fact, the absence of compounds, especially oxides, is rather important to good bonding. At welding temp carbon will not be involved in carbide making since it will be in solid solution (a mixture not a compound) in the iron and thus not be a problem with the bonding.

Fusion welding is what happens with a MIG, TIG or arc welder, where the liquidus is actually reached. In solid state welding the liquidus is not reached, and without good shielding the steel will burn before getting there anyhow. In solid state welding one relies on temperature and pressure to get the metal surfaces close enough for metallic bonding to commence with the iron atoms, this allows the individual layers to maintain the autonomy of their chemistry. Decarb, oxides and boron infiltration account for the different effects in the weld zones. Examples of how this happens can be seen in incredibly high polished precision steel blocks that will grab each other quite well at room temperature as the metallic atoms begin to flirt with each other. One can also bond metals with enough pressure, very unique metals welds have been made using explosives to provide the pressure to make the bonds. I read in a very old metallurgy text about tests that were done a long time ago where gold plates were clamped to other metals for a few years and subsequent analysis showed gold atoms had diffused into the other metals.

Once metallic bonding is achieved, diffusion of interstitial atoms (like carbon) will readily occur between the two parts, as will diffusion of substitutional (alloying- Cr, V, Mo Ni etc…) atoms but at a much slower pace since they need to actually swap positions with iron atoms rather than skitter between them. Carbon diffusion can be slowed or even stopped by using alloy barriers that do not form readily form carbides, i.e. nickel and silicon to some extent. Bladesmiths should be careful not to wish diffusion away too much at it has saved more than a few damascus mixes from being a disaster. With san-mai the idea is to stick the welds and get done before too much diffusion can occur, or slip a thin sheet of pure nickel between the low carb and high carb.

Carbide formers will accelerate the diffusional effects. Here is a 1000X image of a 416/W-2 san mai weld zone from a series of metallographs I did for Burt Foster several few years ago:

To the left is the carbon rich W-1 that becomes carbon depleted white ferrite adjacent to the weld. To the right of the weld the low carb stainless is already forming heavy carbides and carbon solutions making a brown zone beside the weld. The chromium is having a striking effect on the properties of the diffusion causing the contrast to be much more pronounced.

Damascus loses quite of bit of its mystery if you spend a few years dissecting it under the microscope.

I would like to add that there are currently some interesting test being done through ABANA, utilizing the same numbered run of mild steel and various fluxes and temperatures as well as various smiths doing the tests. I believe once the data and research has been done it will be written up in the hammers Blow. I had several long conversations with some research metallurgist in Texas while demonstrating last year, there was specific agreement that the process of forge welding is actually considered a "cohesion" not fusion as Kevin pointed out, due to the fact we do not need to reach liquidous on the surface to create the cohesive properties for the atoms to begin to knit together, as well as Matt points out the given the perfect conditions, they begin to bond. I am really looking forward to the test showing us more.

|quoted:

If I were to ask Mr. Peabody and Sherman to set the Way Back Machine to 1979/1980 and my college chemistry class, I would probably find that what happens in a forge weld is a form of Ionic Bonding. In layman's terms, this is a sharing of electrons between two or more atoms causing the atoms to "stick" to one another. I will assume for a moment that everyone understands atomic theory at a very basic level: Matter is made up of Atoms, Atoms are made up of two parts: the nucleus and the electron shield. The electrons orbit the nucleus in the same way the planets orbit the sun.

When you add significant energy (in the form of heat) to the steels, the electrons begin to get "excited" and their orbits expand. Eventually, they expand their orbits enough to become attracted to the nucleus of the atom next to it, and orbit sharing happens. The electrons begin orbiting more than one nucleus forming what is basically called an ionic bond. Now you may think that this would not be very strong, being that electrons are so miniscule, but ionic bonding is what holds all matter together. Basically in a forge weld, you are bonding plates of different metals to one another by getting the atoms to share electrons.

Disclaimer: It has been a very long time since I took chemistry. This is the best my memory has to offer.

I am trying to stay safely with ferrous metallurgy, where I am fairly confident, and not wander too far into chemistry, which I barely squeaked by with a passing grade (way too many numbers <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//ohmy.gi f' class='bbc_emoticon' alt=':o' /> ). But the main difference between ionic and metallic bonding is the nature of that sharing you describe. In my limited understanding, with ionic type bonding the electrons are bound much tighter in a polar nature, e.g. negative to positive, but this results in a more rigid system that is brittle and not malleable and not as naturally conductive; two defining properties of a metal. Metallic bonding is more an ionic lattice surrounded by a multitude of free electrons that is free to allow malleability and be readily conductive. As for other bonds, I have gotten away with simply avoided the word "molecule' altogether when discussing metallurgy as it rarely applies with metals due to these covalent, ionic, metallic differences, I know it comes close to hair splitting, which I myself hate, but these small differences are huge when brought together for the properties of the materials.

Also it should be emphasized that while there can be diffusion of substitutional atoms (Cr, Ni, Mo etc..) it requires much more heat and time to be anything more than incredibly superficial. This is why keeping the heat lower will allow for a nice clean and crisp layer look, but when we overheat the billet repeatedly we start to get that fuzzy look to the layers.

I agree with Matt on the components of a successful weld, and experience has showed me that the most powerful of the factors is surface preparation. A good clean surface that is flat with no voids or pits to catch crud ,and free of oxides, will go far on compensating for any shortcomings in heat and pressure.

This has been a good chat, but I know I am not the only one that is now shutting down my home PC to get on the road to Atlanta GA. See you all there!