Kitchen knife san mai forging tips?

[lotfong]

Hi everybody, newbie smith and first post here.
I'm trying to make a san mai kitchen knife, without success so far.
Some knowledgeable help would be greatly appreciated.
To keep things simple I will simply state what's going on.

I've previously successfully made many good monosteel 80crv2 blades without much hassle with my limited tools. Now I tried several attempts in san mai, 26c3 core with 1018 cladding. I do my welding, my forging and my ht at the best of my knowledge gathered from my own experience and also as a reference litterally HUNDREDS of hours lurking on this forum and many others in the last year or so.

*BTW If somebody wants to have more details about my whole process I can provide them by message or through specific answers to the comments here.

the issues I encounter are:
- things seemed well welded together only to blow up and fall apart during quench, at worst plain snapping of the core, at best partial delamination of the clad in all but one of my 18 attempts in the last month. I must add that I forge by hand, alone, no power hammer or hydraulic press.
- heavy twisting and warping of the blades post quench. ALL THE TIME. I expected this tho. the issue is whenever I try to fix it either it doesn't bulge (clamp +penny straightening jig during second and third temper) or it shatters (soft hammering with a wooden mallet to straighten one tap at a time, after 2x400°f temper).
- this one I aint so sure but I feel the carbon migration is more severe than I planned, as mono 26c3 (except warping) had no issues of ht and got a really great edge rather easily, the only san mai blade not so much.

The thing is that I feel like I might have all the answers to everything that happens here but I definitely need validation.

My hypothetic solutions include :

-Do longer soaks at a little lower temperature. I aim for bright yellow where the borax is liquid and bubbling but below the temp where sparks begin to appear. I hold roughly 2 min at that temp before starting the hammer welding process. maybe I should aim a darker shade of yellow for 5 to 10min instead. I always prepare my pieces by surface grinding them on my flat disc grinder and I assemble them with steel wire tightened around it multiple times. I make sure they stay without oxyde by brushing them while hot before borax'ing them.

- working with thicker material, thus switching for another steel than 26c3 as I've got the thickest available at 0.150" and I want to do big blades and I feel that hammering a lot of area during welding makes it less efficient. I also prefer avoiding stacking the 26c3 altogether. Probably that a piece of 0.250" thick w2 steel of much smaller surface with correspondingly sized mild steel cladding would be easier to weld than what I have right now, as the contact surface is smaller.

- instead of strictly doing normalization steps before soaking and quenching/temper, I should probably also do a proper sub critical anneal. I feel my breaking and warping issues are from the stressed up steel. I seriously thought normalization and grain refinement were taking care of it. I don't have vermiculite but I plan on getting some, as I have no reliable way to make steel cool down slowly right now. I do my ht withouf a kiln , just a coal forge, a magnet and my eyes that watch for a good color of the steel and decalescence in darkness. My ht on monostesl 26c3 I guess is not bad in therms of hardness and grain refinement, I stick as much as I can to Larrin Thomas's article about 26c3 heat treat and it does get crazy sharp really easy and holds it at a hardness that almost completely skids my 65hrc hardness test file post temper at 350°f.

- getting nickel layers for insulating the core and inhibiting carbon migration. This I feel isn't an absolute necessity, althought preferable, as if I use a thicker core material the carbon migration is especially fast at very high welding temps, and if I get things welded faster I can then go on with lower forging temps and mitigate the issue. Also the sheer improvement in size of the piece mean that at the core the steel has a broader zone where carbon migration hasn't happened much and therefore during reducing temps stays almost intact (I GUESS). Last but not least, IF THERE WAS A THICKER STEEL WITH SIMILAR SPECS, even thought I lose maybe 0.15-0.2% I still have a very clean hypereutectoid normalized steel to bring into austhenitization after doing proper procedures. This might not work as well with W2 as it has 0.96% carbon and I've heard about the HT issues with it lately. I avoided doing it with 80crv2 for the carbon migration concern, maybe I should give it a go, as I've got 0.250"x2"x48" of it. Also if there is another hypereutectoid steel interesting in thick size for it let me know. I've looked at cruforgeV, 1095, 52100, O1, 1.2519, Silver steel (bohler k510) and all the japanese pre clad steels as options.
So in perspective I was wondering if I was maybe on the good path with my hypothetic answers and if you guys had some more points you felt needed to be adjusted.
thanks all for reading me,
Ludvig

 

[Ken H>]

Two points you might consider, #1: warping can be caused by use of two metals with vastly different expansion rates, and #2: you might try doing your grinding bevels jpost HT'ing.

I've made 2 or 3 San Mai billets with mild steel cladding, but used 1095 for the core. I grind post HT. Make the billet, anneal, normalize, etc are usual for a forged billet. Do the surface grind to have a nice clean billet of desired thickness - I'll normally be less than 1/8" for the billet, and profile the blade. Then do Quench 'n Temper with last step grinding bevels of blade.

 

[lotfong]

Two points you might consider, #1: warping can be caused by use of two metals with vastly different expansion rates, and #2: you might try doing your grinding bevels jpost HT'ing.

I've made 2 or 3 San Mai billets with mild steel cladding, but used 1095 for the core. I grind post HT. Make the billet, anneal, normalize, etc are usual for a forged billet. Do the surface grind to have a nice clean billet of desired thickness - I'll normally be less than 1/8" for the billet, and profile the blade. Then do Quench 'n Temper with last step grinding bevels of blade.

Yes I will do all the proper preparative steps every time before HT from now on, just got vermiculite this morning. I don't grind any bevels prior to HT thought. I'll stick to that as you recommand.
I thought that 1018 as a cladding material was perfectly suitable as I have seen it used a lot by other makers with success and I've read it isn't as tricky to forge (let alone forge weld) than wrough iron. I will keep using it but I will use a thinner piece of core steel so that the expansion rate of the core steel doesn't impact much the overall blade. Thank you very much sir to take the time to read and answer me it is greatlt appreciated!

 

[Willie71]

I have only made about 10 successful san Mai billets so far, with five failures, so take this with a grain of salt. My failures have been when doing this by hand, and success has been near perfect when done with a press. I welded at bright orange/yellow heat. I have a Venturi burner, and this is how hot it gets. After the weld is set, I draw it out at bright orange heat. I started with either 1/4” or 0.150” stock for the core, and slightly thicker cladding for the jacket. None have crumbled upon heat treat.

I normalize in my over twice for 10 min at 1650f, cool to magnetic (black) then 1500f for 10 min, cool to black x2. Then I heat treat using the core steel protocol. I Sabre grind the edge until it is evenly exposed on both sides prior to heat treat, so I know where my center is. I have used a full quench on each blade, and they all survived. I have found that poor welds show up during the normalization/thermal cycle steps.

I get a lot of carbon migration, with a press, so the extra time at heat may be part of your issue. You might be welding too hot, but I can’t say that for sure. My forge doesn’t get to white heat, so I have no frame of reference. Once the steel is overheated at a certain point, no amount of normalizing and cycling will fix it. I suspect that is where your issues lie, or secondly, normalizing and cycling in a coal forge would be difficult. You need even heat and 10 min soaks to get the structures refined.

I have used W2, good stock from 5 years ago, 26c3, 1084, and cruforgev successfully. I failed with one M4 core, and 2 z-wear cores. My last z-wear core seems to have worked.

 

[lotfong]

I have only made about 10 successful san Mai billets so far, with five failures, so take this with a grain of salt. My failures have been when doing this by hand, and success has been near perfect when done with a press. I welded at bright orange/yellow heat. I have a Venturi burner, and this is how hot it gets. After the weld is set, I draw it out at bright orange heat. I started with either 1/4” or 0.150” stock for the core, and slightly thicker cladding for the jacket. None have crumbled upon heat treat.

I normalize in my over twice for 10 min at 1650f, cool to magnetic (black) then 1500f for 10 min, cool to black x2. Then I heat treat using the core steel protocol. I Sabre grind the edge until it is evenly exposed on both sides prior to heat treat, so I know where my center is. I have used a full quench on each blade, and they all survived. I have found that poor welds show up during the normalization/thermal cycle steps.

I get a lot of carbon migration, with a press, so the extra time at heat may be part of your issue. You might be welding too hot, but I can’t say that for sure. My forge doesn’t get to white heat, so I have no frame of reference. Once the steel is overheated at a certain point, no amount of normalizing and cycling will fix it. I suspect that is where your issues lie, or secondly, normalizing and cycling in a coal forge would be difficult. You need even heat and 10 min soaks to get the structures refined.

I have used W2, good stock from 5 years ago, 26c3, 1084, and cruforgev successfully. I failed with one M4 core, and 2 z-wear cores. My last z-wear core seems to have worked.

Thanks warren. You were one of the bladesmiths I was hoping would get to answer this post. I definitely feel that my initial welds are probably done at a too hot temp and that I should also do a longer soak at said lower temp. 2min is most definitely not enough to make a good weld according to the information I've just gathered on this forum. I will make it more like 15-20min. For the HT (except for the anneal process) I must say I am not too worried as I have over 10 years of experience with a coal forge and I do manage my temps not too far off if I put myself to it. I am used to do cycling by eye with a couple years of doing it with a thermocouple in a muffle pipe to validate temperature color and decalescence as a reference. I've had good results on monosteel 1084, 80crv2, 26c3 and white paper steel (pre-clad stock from dictum). I actually thought for a while to restrain myself to monosteel and pre-clad untill I get better tools but I've seen some people do san mai purely by hand and I wanted to give it a shot.
Here are photos of my WORST CASE grain in monosteel 26c3 after 350°f temper. a little under grc65 according to my file test. You be the judge. I don't want to appear snobby when mentionning my past experience as it can definitely be part of the reasons of my mistakes, and I really appreciate your advice and I will definitely invest in a kiln as soon as I can! For the peace of mind at least. for now on I've got to find a (hopefully not too shitty) thermocouple on amazon for my k type thermometer to make sure things are ok.

 

[lotfong]

 

[lotfong]

26c3 grain post temper https://imgur.com/gallery/XFyHYw1

 

[lotfong]

Well leave it be for now with the photos. I'll do when i will be on my computer and not from my phone. I guess the Imgur app on mobile doesn't work with sharing here.

 

[Willie71]

Thanks warren. You were one of the bladesmiths I was hoping would get to answer this post. I definitely feel that my initial welds are probably done at a too hot temp and that I should also do a longer soak at said lower temp. 2min is most definitely not enough to make a good weld according to the information I've just gathered on this forum. I will make it more like 15-20min. For the HT (except for the anneal process) I must say I am not too worried as I have over 10 years of experience with a coal forge and I do manage my temps not too far off if I put myself to it. I am used to do cycling by eye with a couple years of doing it with a thermocouple in a muffle pipe to validate temperature color and decalescence as a reference. I've had good results on monosteel 1084, 80crv2, 26c3 and white paper steel (pre-clad stock from dictum). I actually thought for a while to restrain myself to monosteel and pre-clad untill I get better tools but I've seen some people do san mai purely by hand and I wanted to give it a shot.
Here are photos of my WORST CASE grain in monosteel 26c3 after 350°f temper. a little under grc65 according to my file test. You be the judge. I don't want to appear snobby when mentionning my past experience as it can definitely be part of the reasons of my mistakes, and I really appreciate your advice and I will definitely invest in a kiln as soon as I can! For the peace of mind at least. for now on I've got to find a (hopefully not too shitty) thermocouple on amazon for my k type thermometer to make sure things are ok.

try auberins for thermocouples and such.

 

[jake pogg]

You're getting some very good,informed advice on the HT and related issues already,and the possible Compatability issues have also been mentioned.Those include at least two aspects,the relative plasticity,and of course the similarity of HT sequences between the dissimilar alloys.
The only thing i'd caution against is seeing the designation AISI 1018 as something absolute.It is not necessarily That consistent.If not too much of a headache,try a different source of mild(not infrequently that is where the problem lies).

A couple of thoughts on your welding process.I'm not a knifemaker,i seldom work in such fine sections,and so please take it for what it's worth.I do weld a lot,and do so under conditions similar to yours,solid fuel+not much technology.

-Do longer soaks at a little lower temperature. I aim for bright yellow where the borax is liquid and bubbling but below the temp where sparks begin to appear. I hold roughly 2 min at that temp before starting the hammer welding process. maybe I should aim a darker shade of yellow for 5 to 10min instead. I always prepare my pieces by surface grinding them on my flat disc grinder and I assemble them with steel wire tightened around it multiple times. I make sure they stay without oxyde by brushing them while hot before borax'ing them.

If i read you right...You grind the corresponding faces of laminate clean and Then you bring them to Nth degree of heat and Then brush them(and Then apply borax)?
And only then assemble the parts?

1.If i read that correctly,i'd change this to pre-assembling+wiring the entire laminate,corresponding faces ground clean.Then as the whole is coming up to heat applying borax to perimeter of all seams...It Should capillary,sweat,all the way into the laminate.
2.Once borax has become molten,surface covered with tiny bubbles in places,the forging is At heat;holding at this T is detrimental.
Common forging practice is to cut the blast(allowing the heat to rise just a bit more,as it does momentarily),welding on a count of 2 or 3 hence.
3.When hammer-closing the weld-seam Initially there's always some degree of back-and-forth when each blow makes some bonds while ruining a % of previously made ones.To help ameliorate this it helps to use as heavy of a hammer as you can control(i use a 10#),but doing it Ever so gently.The idea is to dampen the harmonic bounce,it's very destructive to have the forging bounce and jangle.Ideally,the hammer momentarily Rests on material dampening all vibrations.
(a horn,being a dead-blow type surface would almost be an ideal bottom-tool,if it wasn't for the unhandy fullering effect).
All described above is why the roll-mill works so good for welding of parallel-sided laminate.It eliminates vibration.
Long time ago those large self-contained air-hammers had a special valve setting,a regime for welding(it momentarily held the blow down to dampen vibration).
So it's the pressure that brings the parts together that is important initially,to start the bonding that is then further "refined" by working the fresh weld at (high,welding-,or near there)T's.
I had a friend that used a foot-operated heading jig(just a large flat clamp)very successfully for initial setting of a weld;and i've used tongs before on some very awkward situations,without removing the composite from the fire,to also-Set the weld.Brought to heat again the composite now held together enough to be now held by grasping only one part of it,so sufficiently.

Just a quick note on C diffusion(though i know naught of it's science,alas...).
I'm not sure if you're not exaggerating both it's speed And the detriment of it's effect...Would not the increasing homogeneity of composite make it work together better?Plastically and HT-wise?...I may well be way off on that,just intuition...

VERY best of luck in any case,all of this is quite a challenge,but Very satisfying once you'll succeed.And you will.

 

[lotfong]

You're getting some very good,informed advice on the HT and related issues already,and the possible Compatability issues have also been mentioned.Those include at least two aspects,the relative plasticity,and of course the similarity of HT sequences between the dissimilar alloys.
The only thing i'd caution against is seeing the designation AISI 1018 as something absolute.It is not necessarily That consistent.If not too much of a headache,try a different source of mild(not infrequently that is where the problem lies).

A couple of thoughts on your welding process.I'm not a knifemaker,i seldom work in such fine sections,and so please take it for what it's worth.I do weld a lot,and do so under conditions similar to yours,solid fuel+not much technology.



If i read you right...You grind the corresponding faces of laminate clean and Then you bring them to Nth degree of heat and Then brush them(and Then apply borax)?
And only then assemble the parts?

1.If i read that correctly,i'd change this to pre-assembling+wiring the entire laminate,corresponding faces ground clean.Then as the whole is coming up to heat applying borax to perimeter of all seams...It Should capillary,sweat,all the way into the laminate.
2.Once borax has become molten,surface covered with tiny bubbles in places,the forging is At heat;holding at this T is detrimental.
Common forging practice is to cut the blast(allowing the heat to rise just a bit more,as it does momentarily),welding on a count of 2 or 3 hence.
3.When hammer-closing the weld-seam Initially there's always some degree of back-and-forth when each blow makes some bonds while ruining a % of previously made ones.To help ameliorate this it helps to use as heavy of a hammer as you can control(i use a 10#),but doing it Ever so gently.The idea is to dampen the harmonic bounce,it's very destructive to have the forging bounce and jangle.Ideally,the hammer momentarily Rests on material dampening all vibrations.
(a horn,being a dead-blow type surface would almost be an ideal bottom-tool,if it wasn't for the unhandy fullering effect).
All described above is why the roll-mill works so good for welding of parallel-sided laminate.It eliminates vibration.
Long time ago those large self-contained air-hammers had a special valve setting,a regime for welding(it momentarily held the blow down to dampen vibration).
So it's the pressure that brings the parts together that is important initially,to start the bonding that is then further "refined" by working the fresh weld at (high,welding-,or near there)T's.
I had a friend that used a foot-operated heading jig(just a large flat clamp)very successfully for initial setting of a weld;and i've used tongs before on some very awkward situations,without removing the composite from the fire,to also-Set the weld.Brought to heat again the composite now held together enough to be now held by grasping only one part of it,so sufficiently.

Just a quick note on C diffusion(though i know naught of it's science,alas...).
I'm not sure if you're not exaggerating both it's speed And the detriment of it's effect...Would not the increasing homogeneity of composite make it work together better?Plastically and HT-wise?...I may well be way off on that,just intuition...

VERY best of luck in any case,all of this is quite a challenge,but Very satisfying once you'll succeed.And you will.

Wow. This is seriously explicit in information and I can't thank you enough. I am very grateful for all I get from you guys. It will save me quite a bit of trial and error , althought not too much I hope!! it is needed for a healthy way of learning thingsI believe! this being said I think it does explain my issues about welding really well. I will try the specific assembly with the wiring firsthand and letting the borax flow instead of how I used to do. Also I do machine-gun- hammer my weld and I don't go gentle with each blow and it must induce unecessary vibration. I might try setting the initial welds with a 20lbs hammer on my railroad track anvil instead of my flat regular 200lbs one, as the rounded surface of it might make a more "squishing" action. As for the carbon migration, the thing is thay I get to reduce the thickness so much while welding that it does spend many occasions at welding heat. Thats the reason I wanted to use thicker stock to start with, but now I will go gentler with tge hammering instead as a start. But seroously thanks a lot to you and everybody here, I feel more confident to get to work tomorrow!! I'll post pics when I get something done!

 

[jake pogg]

railroad track anvil instead of my flat regular 200lbs one, as the rounded surface of it might make a more "squishing" action.

I'm sorry-i must've expressed my thought poorly-i'd Not use RR rail anvil.
What i meant about the conventional anvil-horn is that it's a Dead-blow area,and has no rebound.
Conversely,an RR job would be a very poor choice...
And again,the fuller effect of either is probably not something you'd want either.
So a large heavy anvil would be best,with gentle but firm blows.
A pass down the center of billet,then however many passes from center out toward edges.

 

[lotfong]

Ohh okk I get it. no worries this isn't on you it's my english that is limited at times . I thought that the fuller effect by squishing might do more pronounciated welds but I guess it will also split open by side effect... Well then I'll stay on the flat anvil but as you say I'll go with a heavier hammer with more gentle blows.

 

[jake pogg]

P.S.
Just as an aside,in case some of the problem has to do with Atmosphere,there is a way to judge the Atm. by looking at the molten borax.
(it's a very old method,and will sound a bit occult,but it helped me,and several other people i know).
When borax melts at a properly-reducing atmosphere for whatever chemical reason it's surface will be of a Warm tint...It'll be that golden,honey-colored liquid,with areas of very active small bubbles,but the main deal is that overall the glow appears as Warm.
Conversely,if the Atm. contains excess O,the surface of borax although just as liquid will appear Cold in color,and at times scaly,slightly rough...
It's a constant challenge to keep track of Atm in a solid-fuel fire.
A rule of thumb is to have 4"-5" of burning fuel between the blast and your work(that is a pretty deep fire).But the amount of blast can still be excessive even at such ratio.
Atmosphere is crucially important to welding,more critical than flux or temperature.
Many folks weld fluxless,many-in surprisingly low T ranges,but oxidising A is a killer...

 

[lotfong]

Seriously this is great stuff thanks again. I hadn't dug nearly deep enough beforehand to find those informations about forge welding. I will do my best to apply this sucessfully. Can't wait!!

 

[Stacy E. Apelt - Bladesmith]

I would suggest welding the pieces together on the ends and in the middle with a MIG, TIG, or stick welder. Wiring has lots of problems.

Stress relieve the billet thoroughly before hardening.

Quench plates after the oil will help a lot to avoid warp.

 

[lotfong]

I would suggest welding the pieces together on the ends and in the middle with a MIG, TIG, or stick welder. Wiring has lots of problems.

Stress relieve the billet thoroughly before hardening.

Quench plates after the oil will help a lot to avoid warp.

I will make a plate quench jig to put in my vise. Aluminium is best for that? or I guess just plain steel does it well?

If I could do the welds I would gladly do it... The day I get a welder and an hydraulic press things will be much easier that's for sure!! Thanks Stacy I appreciate your advice.

 

[Stacy E. Apelt - Bladesmith]

Even a cheap welder from Princess Auto will do the job. $250 for stick and $450 for MIG wire.

A welder is almost a must if you want to do damascus and san-mai.

 

[Natlek]

All described above is why the roll-mill works so good for welding of parallel-sided laminate.It eliminates vibration.
Long time ago those large self-contained air-hammers had a special valve setting,a regime for welding(it momentarily held the blow down to dampen vibration).
So it's the pressure that brings the parts together that is important initially,to start the bonding that is then further "refined" by working the fresh weld at (high,welding-,or near there)T's.

I would really appreciate if you look in this thread and I would like to hear what do you think about that ? Thanks in advance
Sorry for of topic OP ....
https://www.bladeforums.com/threads/san-mai-on-small-rolling-mill.1699098/