General advice on broken blade and heat treatment

[SamB]

Hi all

I recently made my first san mai knife and besides a few issues with the bevels was really happy with how it came out. It was my first attempt at forge welding and as far as I can tell it worked beautifully.

Unfortunately I was a bit rough when fitting the handle and snapped the tang off. I could have TIG welded it back on, but combined with the bevel issues I decided to break the blade itself as well and see what the weld and grain looked like.

I suspect my heat treatment process was not optimal, since I shouldn't really have been able to snap the tang off (at least, not easily) if it had been tempered properly. Additionally, the blade only bent about 15-20 degrees before snapping (though that's a guess - I should have filmed it and used a protractor to confirm).

My question for you all is whether that sounds like the blade was under-tempered (or had some other issue), or whether it's within the normal range of toughness you'd expect from a knife. Vague question, I know. I'm mostly just looking for advice to improve future knives.

The blade had a 15N20 core and 1084 cladding. The heat treatment process was:
1) Normalise by heating to 870C, cooling, 840C, cooling, 800C, cooling, something <800C, cooling
2) Heated to 815C, soaked for about a minute, then quenched in canola oil preheated to 50C
3) Waited for it to cool in the oil, then cooled to room temperature on the bench and brushed the crap off
4) Tempered twice in the oven at 200C for two hours, cooling on the bench in between

Here's a couple of pics of the unbroken blade:



And here's a couple showing the grain after breaking the blade:



Any suggestions? Next time I'll make the notches in the tang semicircular, to avoid those stress risers, and probably try a differential temper to further reduce the hardness of the spine.

I also won't bash it with a hammer to try and fit the handle

 

[HSC ///]

“I'm mostly just looking for advice to improve future knives”.
Here you go - Don’t use a hardenable steel for the outer layers next time. Use mild steel or stainless.

 

[Stacy E. Apelt - Bladesmith]

Hardened steel breaks when bent. A 15-20° bend isn't out of the norm before it breaks. Tangs are thinner, and yours has notches to cause stress risers, which makes the tang easier to break. So, nothing seems out of the norm in that part. The grain looks fine from what I see in the photo.

I would have tempered at 220-230°C, but your temper wasn't all that much too low if the blade was a slicer.

Question - How did you determine the thermal cycling and austenitization temperatures? By eye or in an oven?

 

[weo]

Any suggestions? Next . . .I also won't bash it with a hammer to try and fit the handle

This would be my main suggestion.

 

[imill3567]

I'd say the cutouts were probably stress risers. In my opinion even a thinner tang would bend pretty far without breaking further than you'd bend it during handle fitting. Also yes as a general rule if you avoid using a hammer during fitting that's probably for the best Did it break at a cutout or at the junction with the blade?

 

[Horsewright]

Excellent photos by the way. Really illustrate what your showing.

 

[Ken H>]

If the bevel on the other side looks anything like the bevel shown you did really good at keeping the core centered. Good job on the San Mai, and on the grinding. Other folks have made good suggestions so no need to repeat here.

 

[Natlek]

Hi all

I recently made my first san mai knife and besides a few issues with the bevels was really happy with how it came out. It was my first attempt at forge welding and as far as I can tell it worked beautifully.

Unfortunately I was a bit rough when fitting the handle and snapped the tang off. I could have TIG welded it back on, but combined with the bevel issues I decided to break the blade itself as well and see what the weld and grain looked like.

I suspect my heat treatment process was not optimal, since I shouldn't really have been able to snap the tang off (at least, not easily) if it had been tempered properly. Additionally, the blade only bent about 15-20 degrees before snapping (though that's a guess - I should have filmed it and used a protractor to confirm).

My question for you all is whether that sounds like the blade was under-tempered (or had some other issue), or whether it's within the normal range of toughness you'd expect from a knife. Vague question, I know. I'm mostly just looking for advice to improve future knives.

The blade had a 15N20 core and 1084 cladding. The heat treatment process was:
1) Normalise by heating to 870C, cooling, 840C, cooling, 800C, cooling, something <800C, cooling
2) Heated to 815C, soaked for about a minute, then quenched in canola oil preheated to 50C
3) Waited for it to cool in the oil, then cooled to room temperature on the bench and brushed the crap off
4) Tempered twice in the oven at 200C for two hours, cooling on the bench in between

Here's a couple of pics of the unbroken blade:



And here's a couple showing the grain after breaking the blade:



Any suggestions? Next time I'll make the notches in the tang semicircular, to avoid those stress risers, and probably try a differential temper to further reduce the hardness of the spine.

I also won't bash it with a hammer to try and fit the handle

You don t need that notches in tang .......it is knife ,not Excalibur and you need to draw the sword out of a stone .Epoxy is more then enough to do job .These notches I guess should prevent pulling tang from handle , right ? I really can’t imagine any situation for that to happen . If you really want to put notches , cut them at the end of the tang .....one from side is more then enough .
Next time taper the tang and make some radius on ricasso side of tang . Knife will break on that place not inside handle ....It is much easier to make slot in wood for tapered tang then for one like yours .
Nice blade , I like shape

 

[FredyCro]

If you want to do notches anyway you can do them on the corner of the platten with the belt overhanging so they are rounded. Grind the notches only on one corner of the tang and alternate corners, sides and height of the notches.

 

[SamB]

Wow, thanks for all the quick replies and advice. I will respond below

Here you go - Don’t use a hardenable steel for the outer layers next time. Use mild steel or stainless.

I'll try mild steel at some point when I can get some 1008/1018/1020. I tried 1084 since it's supposed to be easy to weld. I assume that if I use low-carbon cladding, it won't etch dark and I should use 1084 or similar for the core if I want contrast?

Hardened steel breaks when bent. A 15-20° bend isn't out of the norm before it breaks. Tangs are thinner, and yours has notches to cause stress risers, which makes the tang easier to break. So, nothing seems out of the norm in that part. The grain looks fine from what I see in the photo.

I would have tempered at 220-230°C, but your temper wasn't all that much too low if the blade was a slicer.

Question - How did you determine the thermal cycling and austenitization temperatures? By eye or in an oven?

Thanks, glad to hear it's roughly what you'd expect from a knife. The notches definitely added unnecessary stress risers.

I determined the temperatures by eye. I have a fairly new thermocouple in the forge, so I adjust the burner to get the desired temperature on the thermocouple and then match the blade and thermocouple temperature by eye. I've just started building a heat treat oven so hopefully in only 3-5 years (I shouldn't joke) I can do it a bit more precisely.

I'd say the cutouts were probably stress risers. ... Did it break at a cutout or at the junction with the blade?

Yeah, they definitely were and that's exactly where it broke - at a cutout.

If the bevel on the other side looks anything like the bevel shown you did really good at keeping the core centered. Good job on the San Mai, and on the grinding.

Thanks The core was centered, but one of the bevels ended up curved outwards and extended further towards the spine than I wanted. It looked fine at first glance but you could definitely see it if you looked at the blade end-on...

You don t need that notches in tang .......it is knife ,not Excalibur and you need to draw the sword out of a stone .Epoxy is more then enough to do job .These notches I guess should prevent pulling tang from handle , right ? I really can’t imagine any situation for that to happen . If you really want to put notches , cut them at the end of the tang .....one from side is more then enough .

On reflection I think you are right. The last knife I made had a full tang and when I clamped it up pretty much all the epoxy was squeezed out, except for where I drilled holes in the tang. I thought the notches would give the epoxy more purchase this time, but after seeing how messy the hole I made in the handle was I don't think there's any need to add more notches...

If you want to do notches anyway you can do them on the corner of the platten with the belt overhanging so they are rounded. Grind the notches only on one corner of the tang and alternate corners, sides and height of the notches.

Yeah, that's another good suggestion thanks.

Thanks again for all the great advice! Hopefully round two will be more successful

 

[Randydb]

Dang, that was a nice looking blade! I prefer making my learning mistakes on stuff that doesn't look that nice.

 

[Natlek]

Hi all

I recently made my first san mai knife and besides a few issues with the bevels was really happy with how it came out. It was my first attempt at forge welding and as far as I can tell it worked beautifully.

Unfortunately I was a bit rough when fitting the handle and snapped the tang off. I could have TIG welded it back on, but combined with the bevel issues I decided to break the blade itself as well and see what the weld and grain looked like.

I suspect my heat treatment process was not optimal, since I shouldn't really have been able to snap the tang off (at least, not easily) if it had been tempered properly. Additionally, the blade only bent about 15-20 degrees before snapping (though that's a guess - I should have filmed it and used a protractor to confirm).

My question for you all is whether that sounds like the blade was under-tempered (or had some other issue), or whether it's within the normal range of toughness you'd expect from a knife. Vague question, I know. I'm mostly just looking for advice to improve future knives.

The blade had a 15N20 core and 1084 cladding. The heat treatment process was:
1) Normalise by heating to 870C, cooling, 840C, cooling, 800C, cooling, something <800C, cooling
2) Heated to 815C, soaked for about a minute, then quenched in canola oil preheated to 50C
3) Waited for it to cool in the oil, then cooled to room temperature on the bench and brushed the crap off
4) Tempered twice in the oven at 200C for two hours, cooling on the bench in between

Here's a couple of pics of the unbroken blade:



And here's a couple showing the grain after breaking the blade:



Any suggestions? Next time I'll make the notches in the tang semicircular, to avoid those stress risers, and probably try a differential temper to further reduce the hardness of the spine.

I also won't bash it with a hammer to try and fit the handle

You can still save that knife ...... I guess it broke in one of these places ? Weld it and use Micarta or G10 for handle , the weld will be deep enough in the handle .You will not lose any strength compared to the tang that is one part , I mean not welded.

 

[Stacy E. Apelt - Bladesmith]

I agree with Natlek. I would weld a new tang on that blade and have no worries. The tang does not need to be hardened anyway.

 

[Jrmysell]

On reflection I think you are right. The last knife I made had a full tang and when I clamped it up pretty much all the epoxy was squeezed out, except for where I drilled holes in the tang. I thought the notches would give the epoxy more purchase this time, but after seeing how messy the hole I made in the handle was I don't think there's any need to add more notches...

To prevent that without the notches if you have a round wheel attachment you can do a hollow grind down the middle of the tang on each side where the epoxy can stay

 

[Alex Topfer]

The last knife I made had a full tang and when I clamped it up pretty much all the epoxy was squeezed out, except for where I drilled holes in the tang. I thought the notches would give the epoxy more purchase this time, but after seeing how messy the hole I made in the handle was I don't think there's any need to add more notches...

You only need enough epoxy to fill the gaps between the two surfaces. You can still get a strong join with very little glue, if everything is properly prepared.