Wootz/Pulad Old Recipes

[Peter T. Swarz-Burt]

Jan,

I have used W-2 as an ingredient in several wootz ingots for its carbon and vanadium content, and I have also forged it out just by itself as bar stock. Even just forging it out and thermal cycling it to accentuate carbide formation will yield a definite pattern. It doesn't look the same as wootz, but that is simply because the starting ingot was probably over a ton as opposed to a few pounds for a traditional wootz ingot. My understanding is that OLD nicholson files were W-2, but the only source that I know of these days is Cincinatti Tool Steel, and they don't carry it in less than 1" thickness.

It looks like the sulfur levels in your iron are quite low, but what are you going to use to bring up the carbon content? The carbon source is very often also the sulfur/phosphorous source. If you are going to use charcoal or other non-metallic carbon sources, make sure your crucible is very well sealed, since otherwise the carbon will simply burn off and escape. A lot of folks are using cast iron for the carbon source, and it can be very hard to find cast iron with low sulfur and phosphorous levels. It is out there, but it is tricky to find.

It sounds like you weren't getting quite enough temperature to achieve a true molton state with those files. Files these days are generally W-1 or something similar, which is about 1% carbon. I don't have my chart in front of me, but I think that you need to hit around 2600F to melt that. One thing that I have heard about homemade crucibles is that sometimes the wall are too thick, and in some cases thick enough to prevent the interior from reaching melting temp even when the exterior refractory is starting to slag. I know that Ric Furrer has done a lot of work developing crucible shapes, so it might be worth asking him for some tips.

Peter
http://www.fallinghammerproductions.com
http;//www.dragonsbreathforge.com/wootz.htm

[Jan Ysselstein]

Peter:
I see Greg has just started a clay crucible thread today. This evening I will respond to your comments above, but I will do it on the clay thread.

Jan

[Manouchehr M.]

Excellent thread Greg and thanks gentlemen for your contributions. I will make this thread sticky.

Kind regards

Manouchehr Moshtagh Khorasani

[Greg T. Obach]

thank you Manoucher

along with the recipes... I forgot to mention that we should start a post of the Roast times for the ingot... this is an area that is often neglected ... and its so important..... a raw ingot can be a beast to forge....

In the " Voysey" account.... it says that the ingot is very hard and requires to be annealed 3 or 4 times, covered in clay, and exposed to a red heat for 12 to 16 hours.

I believe this does make the ingot much more forgiving and I also think it enlarges the pattern...

Greg

[Ann Feuerbach]

Hi Greg,
I have a gut feeling that the roasting time won't enlarge the pattern because the dendritic structure is already in place, but I could be wrong. I think it is important to note the roasting/annealing time in conjunction with the fast or slow cooled ingots. This is because of the different cementite phases and locations. For examples, are the very slowly cooled ingots "self annealed"?

[Greg T. Obach]

Hi Ann

maybe i should have worded it abit different... the roasting does affect the pattern ... i've noticed that it makes waterings more flowing rather that the spotty dendritic look...

if you forge an ingot with very low roast...its very dendritic looking.... but if you take that same ingot and give it a good roasting ... the pattern flows better..
-- at least this is my experience... i could be wrong

and you are correct... it won't make a small pattern ingot into a big pattern slow cooled ingot..

what is going on in the roast... ?
are some of the matrix elements going into solution.. some boundaries being dissolved

even a long cooled ingot... i would still give it a roast.... as i believe the ingot would still be troublesome ...

I did try one ingot this way ... looking at my notes..... it was when i first started... i melted it .. then left in in the furnace to slowly cool over night...
- then cut it in half with a chop saw... and tried to forge one half...(without success)... the other other half i arc welded a mild iron capsule around it and it forged out abit... only got a pocket knife out of a 5lb ingot... ughh
-- and the pattern was very dendritic... but nice

Greg

[Robert C]

Yes the higher the roasting [annealing ]temperature the more carbides are being dissolved. That also means less dendrite .The dendrite is formed on initial solidification .You can't increase it later ,only decrease it .Forging of course changes the pattern .Wrapping with mild steel helps prevent cracking and they did the same with a surface that was decarburized.

[Ann Feuerbach]

Thanks Greg for the info, VERY important observations when trying to figure out provenance of blades and technology. I think I would like to analyses some ingots before and after annealing to see exactly what is "moving and changing".

[Peter T. Swarz-Burt]

I have trouble seeing how the roasting would have any notable effect on the pattern, since the only way to alter the pattern is to get the vanadium to move. If the vanadium starts moving, I would think it will go in random directions, rather than move in some organized fashion that would make the pattern better. My understanding of the roasting is that it

1) decarburizes the surface to aid in forging

2) lowers the overall carbon content slightly, which was probably a good thing in most cases

3) starts growing a new grain structure not based around the dendritic structure

4) if the ingot is heat-cycled above and below critical, the grains would also be refined and reformed a number of times, again a great aid in making the ingot less crystalline in nature

5) if the ingot is heat-cycled above and below critical, the carbides would do a nice job sucking up additional carbon, in effect making the rest of the ingot a lower carbon steel...and therefore more forgeable

I've never had any great success with roasting, but I do heat-cycle every ingot 10-12 times for the reasons stated above. I have still been having trouble with cracking in the upper portion of many of my ingots, but I am pretty sure that this is the result of an impurity problem that I still haven't gotten sorted out yet rather than the lack of long roasting.

[Jan Ysselstein]

"I have trouble seeing how the roasting would have any notable effect on the pattern, since the only way to alter the pattern is to get the vanadium to move. If the vanadium starts moving, I would think it will go in random directions, rather than move in some organized fashion that would make the pattern better. "


Peter:
I can imagine a scenario where some slight random movement of Vanadium would be desirable, allowing more sites for the formation of spheroidal cementite. If things go too well during solidification it could get a little crowded for those Vanadiumcarbide sites.

As a user of W-2 steel I am hoping you will do a spark test comparing W-1 to W-2, and post your observations. Thanks.

"Jan

[Peter T. Swarz-Burt]

Jan,

I did a spark test on a piece of W-2 vs a file I had sitting around (all out of official W-1) and they looked basically identical. In theory the only difference between the steels is a very small amount of vanadium, and I doubt that I would be able to pick up on that in a spark test. One important thing to note about W-2 is that there are actually several grades with substantial differences in carbon content. Cincinnati Tool Steel's info indicates that it is available from .70%C all the way up to 1.30%C. The only way to know what you have is to check with your supplier. Anything noticeably above or below the eutectoid point should get you a pattern, although hypoeutectoid "wootz" is not officially "wootz" in my mind.

On the roasting, I guess what you are saying is that, if the segregation went TOO well, the vanadium would end up in "pockets" rather than "ribbons" or "strands". My feeling is still that the random diffusion of vanadium would be more likely to fuzz or erase the pattern than make it better. Looking back at what has been said, I also have my doubts about how much the vanadium would even be able to move if the roast is at a dull red heat. I think the only way we are going to sort this out is to cut an ingot in half vertically, roast one part as Greg advocates, give the other part a quicker treatment of thermal cycling say 10 times (I think we all agree that some degree of treatment pre-forging is good), forge them down into bar stock and see what the result is. Here are the possibilities that I see:

1) There is no difference in ease of forging or pattern appearance
2) One ingot is easier to forge but the patterns are the same
3) No change in forging, but pattern in one looks better
4) Both forging and pattern are better in one than the other

Anyone feel like making a test ingot?

[Greg T. Obach]

Hi

yes... I agree that the heat cycling above and bellow non-mag does the trick for the pattern...

but if you think of it... why do the carbides arrange themselves in is such patterns..?.. there are other elements at play that make up a matrix/network.
-- when the ingot is fresh... i believe there is a very dendritic matrix/rigid

now... when you roast at a red temp... ( ignoring Carbon completely ) some of the matrix starts to move/dissolve ??? ... now the trick is to get some of the matrix to go and make the waterings nice and flowing as opposed to the rigid dendritic look..

ofcourse... if you roast at too high a temp.... too much matrix will go, and you will end up with a more homogenized steel...

i know, i know... its a balancing act... but i do believe it does make the pattern flow abit better..

or maybe i'm just crazy...

Robert is right on the money... the size of the dendrites at solidification is all you got !!

also... remember as you forge... some of the matrix is changing abit aswell...and the pattern does open up with many forge cycles... but not as much..

if you look at this pic.... this is a quick cool ingot... couple min to solidify so the pattern is small..... but i did roast it for a good length of time... and the waterings are small but more flowing



have you read the study by Verhoeven on destruction and reformation of the pattern... ?
its very neat.... i'll see if i can dig it up.

Greg

[Greg T. Obach]

got the study

here are the conclusions



continued:

the planes of segregated third elements. Additional experiments are needed to identify the role played by the different 3rd elements, Mn, Si, P, and S, in the mechanism for the cluster sheet formation during forging..


Greg

sorry bout the sideways pic's.... doh

ps... there is an excellent chart on their thermo cycling experiments that says alot aswell

[Greg T. Obach]

Hi

the chart aswell... beautiful work.... I really recommend that people get the studies done by Verhoeven... very good work !




Greg

ps just to break down the chart abit... the Tmax and Tmin are the temperatures the wootz was cycled at... .... the initial treatment is what the sample was exposed to.... and the clarity of sheet morphology is the visible pattern that is left at the end.....

obviously the " strong " clarity is the goal !

[Ann Feuerbach]

Great discussions, very usefull! I must dig out my Verhoeven papers again.
Greg, do I have your permission to use your pic (with reference to you of course) to illustrate how a quick cooled ingot can make a fine (in all meanings of the word) pattern, in an upcomming lecture?

[Greg T. Obach]

Hi Ann

yes, ofcourse.. you can use the pictures.. .. i've got other wootz stuff on that photobucket album..

the pattern that i talked about above... i observed this phenomena after making several ingots with a longer roasts... I'd like to set up some iron clad tests and measurements ... to make sure my eyes aren't fooling me...

yes... that was a quick cool ingot... if you can believe it.... I melted the charge... held it for abit..... then removed the crucible from the furnace to aircool... ( which i call a quick cool )....the dendrites were very small..... my longer cooled ingots stay in the furnace and the burner is turned down bit by bit...

take care
Greg

[Peter T. Swarz-Burt]

It's been a long time since I looked at that article. If I recall, this article was the one that didn't make much mention of vanadium, which indicates to me that the association between vanadium content and the cementite bands hadn't been shown yet...I could be wrong in my recollection. I am pretty sure that of the various other elements that you listed, manganese is the only one which will be noticeably heterogeneous within the ingot.

I also want to note that they found that a short austenization at 1100C (around 2000F) did not permanently affect the pattern, while a long soak at 1200C (approx. 2200F) erased it permanently. This indicates to me that in a soak at a dull red heat, say 850C or so, perhaps no amount of time would be sufficient to noticeably move the vanadium. Vanadium is a smaller atom than iron, but I am pretty sure that it takes the place of an iron atom within the lattice. The only way for it to move is for it to trade places with an iron atom.

the knife you picture has an excellent pattern, but I still argue that the roasting and the nature of the pattern are not related. A quickly solidified ingot will have a finer dendrite pattern, which will in turn make for a "tighter" wootz pattern in the end product. I see no reason why a finer dendrite shouldn't result in a finer, but flowing, wootz. As I said previously, I would want to see physical proof, or at least a good explanation, of how a long soak at "low" temperature could cause the vanadium to "flow" in an organized manner such that the pattern was accentuated rather than erased.

[Greg T. Obach]

although... I think that vanadium is an important component...but its not the only thing working here...

otherwise why would we need a slow cool steel... you could make wootz from a chill cast steel... as Vanadium carbides will form in that aswell..

there has to be some kind of matrix that is formed when the steel is slow cooled... some of the elements come out of solution ...as their solidifying temps are hit...

the V/carbides are important as noted on the online article... key role of impurities by Verhoeven

"The formation mechanism of the carbides clustered selectively along the IRs during the cyclic heating of the forging process is not resolved. It seems likely, however, that it involves a selective coarsening process, whereby cementite particles lying on the IRs slowly become larger than their neighbors lying on dendrite regions and crowd them out. A model for such a selective coarsening process has been presented.17 During the heat-up stage of each thermal cycle, the smaller cementite particles will dissolve, and only the larger particles will remain at the forging temperature, which lies just below the Acm temperature. The model requires the segregated impurity atoms lying in the IRs to selectively reduce the mobility of the cementite/austenite interfaces in those regions. Larger particles would then occur in the IRs at the forging temperature. They probably maintain their dominance on cool down because one would not expect the small particles that had dissolved to renucleate on cool down in the presence of the nearby cementite particles. These near-by particles would provide sites for cementite growth prior to adequate local supercooling sufficient to nucleate new particles. "

---its the roasting that i think changes the boundaries, abit..... or maybe an entirely different mechanism ?

--... how to prove it.... well that is the problem itself !
- i'm only stating this of my observations... ... hardly scientific !!!


-- lets look at the opposite.... the very large flowing patterns such as the Persian style
... how did the Persians make such large patterns... ... did they forge the ingot out as is.... and if they did would the pattern be very dendritic...
-- or could it be all pattern manipulation ( laddering, holes, folding, or other)

how about an very long slow cool ingot.... you'll have very large dendrites... but does this necessarily equal a long large flowing pattern

these are all interesting questions... lots of work has to be done

Greg

[Peter T. Swarz-Burt]

Greg,

I do not believe that the other elements have a substantial role to play based on one simple observation...W-2 will form a very nice wootz pattern, and as a tool steel the level of everything except carbon, manganese and vanadium is kept to very low levels. Even the manganese content is not very high. Pretty much any steel (non-CPM) with sufficient carbon and a small amount of vanadium will form such patterns. I do not believe that chill cast steel would form any kind of pattern, simply because there is no opportunity for any segregation during solidifaction.

My reading of the the quote is that the impurities he is talking about are the vanadium atoms. Vanadium is known to be a very strong carbide former, and therefore it will form carbides at higher temperatures and those carbides will also dissolve at higher temperatures than will iron. The vanadium carbides then form a nucleus around which further cementite growth will preferentially occur.

If I were going to make a guess about the factors which most affect the end pattern, I would say these:

-speed of initial cooling and solidification...extremely slow cooling will result in an ingot with such organized dendrite growth that in industry they call this "ingotism" and consider the ingots unforgeable.

-size of ingot and therefore the amount of forging required to bring it final size...my guess is that a larger ingot will actually produce a larger, more flowing pattern in the final product...this is just a guess...a larger ingot will also have an inherently slower cooling time, which may make a difference.

-the shape of the crucible...it is my feeling that different shapes of crucible will affect the way in which the ingot solidifies and perhaps the orientation of the dendrites...Ann can speak more to the traditional crucibl shapes, I would guess.

If you checked out my website

http://www.dragonsbreathforge.com/wootz.htm

you will see that the two blades I have pictured have totally different patterns. The one that looks closer to antique blades is forged out in the traditional manner in which the top and bottom of the ingot become the spine and edge of the blade. The other blade is forged out vertically, which is to say that the top and bottom of the ingot became the point and tang of the blade. If you look closely at the pattern on this second blade, you will see that the carbides have essentially formed strands. I have had this same effect in other blades forged out in a similar fashion, which makes me think that this is pretty typical dendrite formation...at least in my ingots. What it means, I am not quite sure.

I guess one way that we could test the effect of roasting without cutting an ingot in half would be to do the following:

1) make ingot in usual fashion
2) grind, polish and etch several spots on the ingot and get good photographs, preferably at some magnification
3) roast for a period of time
4) grind, polish, etch and photograph the same spots again being sure to get down below the decarb layer...you would be deeper in the ingot, so the same dendrites wouldn't be showing, but the general appearance of the dendrites should show whether a change is taking place.
5) repeat 3 and 4 several times

If a change seemed to be taking place, and the ingot forged out to have they type of pattern desired, it would at least be evidence of a sort, although not exactly proof. If only we could get government funding for such studies, maybe there would be more incentive....

[Robert C]

Peter, I think you've stated it well .Many times I've tried to explain the importance of proper temperature and soak time in heat treating a high vanadium steel such as S30V. Tungsten and molybdenum are the two other strong carbide formers .Crucible size and cooling rate certainly control dendrite size . Having worked with large cross sections of tool steel and the difficulties of dealing with "massive carbides, too large a crucible or too slow a cooling rate would not be good !... Ann ,can you tell us of different size and height to diameter ratios and their effects ?? ...Keep on experimenting !

[Greg T. Obach]

Hi
wootz is a complex steel..
I do understand what your saying about the Vanadium carbides.. and how they work... but i disagree with you that the system is so simple... its not

W2 will never be wootz.... its not a slow cool crucible steel... i've done the 52100 and file steel, A2 thing... heat cycled them... got some nice lite pattern to show when etched.... and when compared to the tulwars, and shamshirs i've restored, and my crucible steel.. its not even close ...

theres a fellow on the net that does this kind of heat cycled steel but can't obviously drop the name here..

the ingot is obviously influenced by crucible size... egg shaped, flat etc and yes the ingot size does effect cooling... i've made 8lbs ingots... and the pattern is similar as my 3lbs ingots

Amount of forging.... well.. I got this one nailed... i work my ingots down by hand... I have no apprentice... no powerhammer... no forge press (cept for one time i used Als press).. so, i hand hammer them starting with a 12lbs hammer.... then to 8 and to 4lbs till its barstock size... so there is many many heat cycles involved.... and many more for the larger ingots.... --- and i haven't noticed too much a difference between a 2 lbs ingot and a 6lbs

its not one simple observation.. its made over many ingots and many many forgings

... why did the ancient smiths think the roast was important... if a couple normalize cycles were all thats needed..... thats alot simpler to do... save alot of expensive fuel aswell..

- one thing forsure... roasting does reducing cracking alot... .. at first i had a hard time to get a sword length come out... without cracks... but now with a better roast... it much much easier to do

its a tough road to walk down ... when trying to characterize a steel and its particular behaviors ....

even the definition of wootz is uncertain...... some people discount the low carb wootz.. ... which is odd? the ancient smiths didn't discount the low carb wootz ... it was there and it was included.... but for some reason the modern smiths have only accepted the high carb..... they've changed the word's meaning

look at the Zschokke blades samples... one was low carb.... but its not wootz... just cause it doesn't fit their nice tidy definition... doesn't mean it doesn't exist.. ...


I agree with you completely.. that tests have to be done... I'd like to observe strict measurements and have good science... my roasting observations are just that... casual observations... and are no replacement at all for a good empirical study !
-- but until the study come out.. all opinions and theory's are valid, with a grain of salt... and pinch of imagination....


Greg

[Peter T. Swarz-Burt]

Greg,

I guess I would tend to agree with you that what makes the pattern in wootz look the way it does is complicated and the result of many factors. What I don't see is how roasting could cause the vanadium to move. Regardless of what causes the pattern to develop in the way that it does, the vanadium is the source of the pattern. Excluding the involvement of another strong carbide former, none of the other impurity elements will be visible in the pattern. I can't argue with what you have observed in your own work, but I don't see how the mechanism would work. As to why the ancient smiths did the roasting, I think we aren't actually in a position to say how widespread it was. It has been noted by many people over time that many of the accounts of wootz-making have been less than factual, either through lack of knowledge on the part of the person writing the account or through purposeful misleading by the smith. Assuming that roasting was widespread, we then come up against the question of why they did it. My guess is that, since they had much less control over, and understanding of, what actually became part of the ingot, it was especially important for them to reduce the carbon content overall and decarburize the surface before forging.

I wonder whether the look of your pattern vs some of the other smiths making wootz comes down to the fact that you work entirely by hand. I know that it is possible to see the difference between PW steels made by hand vs power hammer vs hydraulic press. Perhaps the same is true of wootz. I don't know why it would be true, but it can't be thrown out as a theory.

I am split on whether to include the hypoeutectoid wootz in the trm "wootz" as we generally use it. Where hypereutectoid wootz forms a cementite/pearlite pattern, while hypoeutectoid wootz forms a ferrite/pearlite pattern. They look essentially the same, and the underlying reasons for the pattern forming are basically the same, but the pattern itself and the characteristics of the materials are scientifically quite different. I don't know whether the ancient smiths would have even known that the two were different, since they end up looking essentially the same. I guess another complication is that not many people would understand the difference between the two without a lengthy explanation.

[Greg T. Obach]

Hi Peter

you maybe right... I could be the outlying factor by hand forging ... that could weigh in alot...
- i realize that my roasting observation is hard to grasp.. and to go after it further i'll have to organize some tests to isolate that... .. propose a good thesis question.. and see what happens to the null..

I think as a crucible steel making community... our numbers are limited... so the more each of us contributes, the quicker the body of knowledge grows..
it's a changing field... and it helps to be abit plastic when it come down to wootz theory... lol

whether the V carbides are in the interdendritic areas and grow by milking carb from the dendritic areas.... through thermo cycles ...... and whether those boundaries between the IR and dendrites are dissolved abit during low temp roasting, some what linking the IR areas to make longer or by aligning the IR areas by some bizarre mechanism...... who really knows at this point..

it maybe ... that the roast just simply makes the carbon more available ... and the rigid inner structure more compliant to forging.... so it can flow together rather than have a more rigid dendritic matrix.... and fight the flow of forging..

at the moment... i think its helpful to observe and note a phenomena ... then see if it chimes in with past research... and go from there..


Peter... you got it, on the hypo wootz.... there is obvious element segregation here... ( don't know about the V carbides now? )
-- with a good etch... it does look so much like the high carb...
- how could the ancient smiths know... they didn't have any modern spectrograph analysis of steel...
-- just a slow cooled crucible steel ... that forged and etched like wootz
-- but i'm sure they noted the type of steel as they had many catagories for wootz steel...
-- since alot of the current research is put out by a few... i feel it is in their interest to keep it focused on the spec's of the patent.. ( keeping wootz in the high carb catagory )
- also denying that any prior recipes exist aswell..
- i don't see it this way... i'm trying to look at the ancient material, and in its proper context...



Greg

-- maybe i was too confident in my roast... now i'll have to pay more attention and see what it going on...

[jeff Pringle]

Where hypereutectoid wootz forms a cementite/pearlite pattern, while hypoeutectoid wootz forms a ferrite/pearlite pattern.

Not exactly, it's more high alloy/low alloy, 'cause you can get good patterns right around the eutectoid where it's pearlite/pearlite. The carbides just hang out in the high-alloy metal in hypereutectoid steel. I don't know of anyone who has made crucible steel of low enough carbon content for free ferrite to show up as a visible pattern element...

[Peter T. Swarz-Burt]

Jeff,

If it is a eutectoid steel, and the carbides are in the "high alloy" section rather than the "low alloy" section, then I must be missing something in your explanation. By definition, pearlite is an exact mixture of ferrite and cementite...carbide. In a eutectoid steel there is just enough carbon to make that perfect mixture...so if one area has more cementite than the other area, then neither area is actually all pearlite. I guess I could have been more exact in my statement and said that the pattern was caused by banding of "a mix of ferrite and pearlite vs a mix of pearlite and cementite", but that seems kind-of silly. If we come right down to it, all wootz and wootz-like steels derive their patterns from differences in alloy levels, so the carbon level is in many ways a moot point. In theory, a pattern should be visible down to quite low carbon levels, since the carbides will always preferentially form in the regions with the higher vanadium (or a few other alloys) level. I have never seen it, but it should work. Saying that the pattern is caused by differences in alloy levels INSTEAD of differences in ferrite, pearlite, and cementite levels seems odd to me, since in the steels that we have been talking about, the differences in alloy levels would not be visible except for the resulting effect on where the cementite preferentially forms. I can't speak about the higher alloy modern steels which exhibit alloy banding and whether the visible pattern is caused by the alloys directly or indirectly through their effect on carbide formation. In wootz and the lower-carbon versions of wootz, I think I can say pretty confidently that the trace amounts of vanadium and other carbide formers would be utterly invisible if not for their ability to form carbides. Proof is as easy as heating a wootz blade to 1800F for a couple minutes and quenching in oil. Voila, no pattern.

My guess is that while I was referring to examples of antique swords which show wootz patterns while not being hypereutectoid, perhaps you were referring to some of the higher alloy modern steels which also exhibit similar patterns. Seems to me we are probably both right...although I still disagree with your assessment of the steels as being pearlite/pearlite...one alloy or another is getting the lion's share of the carbon and pushing the rest of the steel into pearlite/ferrite territory.