Heat Treatment - Crystal Weaving Foundation

I took my time prepared test bars and leisurely shot a long video (22+ minutes *as a warning :p*)

Bending test ~4" length (corrected - video stated 3.125" length but from freezed frames, bending length is close to 4")
W2 0.120" thick, ~ 5/16" wide. HT 2.4 - 66rc and 64trc bars.

ZpJUcfd.jpg


 
Deduction/extrapolation of bending length to achieve 90 degrees bend angle - for W2 0.120" thick, 66rc above.
* keep in mind - this calculation under the assumption of fairly ductile steel, whereas this W2 bending is mostly elastic and very narrow plastic/ductile.

Radii below are for 69(left) and 75(right) degrees angle

5nyUh7u.jpg


Compare against a control/baseline W2 at 66rc. So if control thickness is 0.120", minimum bending length to achieve 90° bending angle is ~10.68" long.

yb0jFVK.jpg


* for 0.25" thick: BA90° (R:6.4") = 10.25", BA90°(R:5.6") = 9"

General calculation
https://en.wikipedia.org/wiki/Bending_(metalworking)

I found a cool drawing on grain roles in compressive side (inside arc) of bending. Easy to envision grains on the outside arc to be - elongate and bend along the arc line.
https://www.researchgate.net/profil...n-and-compression-layers-of-the-bend-zone.ppm

Figure-2-Grain-measurements-in-the-tension-and-compression-layers-of-the-bend-zone.ppm
 
Last edited:
Hey Luong:


You deserve to be in the BladeForums Hall of Fame. Makes me laugh. Amazing stuff.


The diagram of grain elongation doesn’t seem quite right to me. My thought is that the grains would be stretched parallel to the blade on the inside and outside of the bend— and that stretching would make the steel more vulnerable to permanent bending or breaking on a second bend.


I tried to show that with a simple experiment:


A little crude, but here are the round(ish) grains before bending.


UNADJUSTEDNONRAW_thumb_2fc_zpspjj5yibj.jpg




Not a perfect illustration, but after the bending, grains on both the inside and outside of the bend are elongated parallel to the blade.


]
UNADJUSTEDNONRAW_thumb_2fd_zpsgypfsy6i.jpg
 
Thanks! Great catch Twindog! I misread to graph (the darn outside circle threw me off). Neutral line is along those grain stay spherical/circular shape, thus grains on the outside part are stretch. Which agreed with general bending for hardened steel where K-factor is between 0.5 and 0.4 (t/T, depend on hardness - K is lower for higher hardness, I conservatively used 0.5). K-factor for your experiment look like closer to 1.0 (mostly stretch, almost no compress).

btw - (not really applicable but if) Diamond K-factor would be close to zero.

http://www.mathalino.com/sites/default/files/images/flexure-of-a-beam.jpg
flexure-of-a-beam.jpg


Addressing the bend-back prone to breakage issue - you pointed out:
When tensile yielded (permanent bent) - thickness (T) is reduced (thinner), so stretched part of the test bar is actually longer. Compressive strength is higher than elongation, so bending back now stretch/elongation the was compressed (inside arc) to match the outside (was stretched) length. IF elongation exceed allowable (per test piece), fracture ensure, otherwise 'necking' occurs (narrower than other non-affected section). High hardness blades (and or high alloy) often break because of this behavior. Easy to observe 'necking' and fracture by bend a nail back/forth a bunch of times.
 
Last edited:
I received a couple bars of D6 steel (along with Niolox/sb1 and O1), looking forward to making a few D6 serrated blades ~65rc with ~50 micron teeth (via SiC sub 220 grit sharpening with micro bevel +5dps using 3um diamond).

nLm1ofT.jpg
 
We are surrounded by limits - resisting is futile... Umm can we re-evaluate the one below? (please feel free to correct any of my newb in fatigue&yield fields)

https://en.wikipedia.org/wiki/Fatigue_limit
Typical values of the Endurance Limit (Se) for steels are 1/2 the ultimate tensile strength, to a maximum of 290 MPa (42 ksi)

350px-S-N_curves.PNG


Aah no wonder high-cycle spring (such as vehicle leaf spring) is thick & heavy ;)

5160 at 50rc has tensile strength around 1450 MPa but Se at 290MPa? Of course it supports higher load but will shorten life.

It seems correlated to - True Elastic Limit (TEL)

https://en.wikipedia.org/wiki/Yield_(engineering)
300px-Metal_yield.svg.png

Stress–strain curve showing typical nonferrous alloys.
*Industry/std Ht*
Take a martensite matrix with ~100% complete transformation (no RA and very little % of ferrite) at 66rc.

Se = when load cycles less than TEL.
2,3,4 are not far beyond TEL, since dislocations movement would lead to fast fatigue eventually fracture.

Why is TEL range so small/narrow for this matrix? Well high % of HRC reading came from as-quenched standing stressed/dislocation, so add a load on top would rapidly exceed crystal lattice yield point (slip). A metallurgy fact, right? Not really, this is just a poor initial condition. OK, tempered this matrix down a couple RC - well, it is just a less severe dislocations in neutral state (not flex) but quickly push/lean on each others with flex, so it will only extent TEL to Se max as HRC lower toward 49-50RC.

Existing metallurgical empirical data (various industries and academic) clearly confirmed the extreme brittleness of 66rc matrix. Elastic Limit (in term of bending angle for 4", 0.12" thick) would be less than 35 degrees, while TEL angle would (guess) be less than 5-6 degrees. Aha, that is why you don't see springs at 60+rc.

*BCMW Ht*
Bcmw test W2 66rc above has elastic limit around 70 degrees == TEL range should be at least twice relative to industry. Also Elastic Potential Energy is much higher (more bending/loading up) therefore matrix initial condition has much lower dislocation, contribute less% hrc at neutral state, translated to more resisting strength to displacement/move.

Good news: These properties are very useful for a blade overall structure and cutting edge ;)

My to do: I'll try to determine a high confidence TEL range of HT 2.4 and how it could translate/transfer to Endurance Limit.
 
Thanks wise & Chris!

At this point, proving is simple: measure test bar dimension(4" bending length, 0.12" thick) and hardness(65-67rc) and bending angle(50+°).

Figuring out implication/application would be more complex, so math+physics+engineering+etc is helpful. For example - if my TEL is 2x of industry TEL and at much higher tensile strength(hrc), then a high cycle spring can be only 1/2 thick or thinner.
 
Last edited:
Exactly, if there's thinner and lighter material, imagine the cost saving in fuel to transport the material and other design element involving steel use :thumbsup: ... (Offset by the cost increase if any for the HT I guess :eek:) ... Still the nett effect would be tremendously beneficial :D
 
Last edited:
Using evidence (my W2 micrograph above) to work one step backward. Se (endurance limit) = ~ TEL = ~σy
df526de99dfa26824e6aa0d04e21ef097c8e79b8

Hall–Petch Relation

https://en.wikipedia.org/wiki/Grain_boundary_strengthening

Let's use Ky and d^1/2 at micron/um for steel and hardness around 50rc
http://web.mit.edu/course/3/3.225/mech/solns/hw9.pdf
σ0= ~70, K y = ~700

So for example a steel(5160) with grain diameter ~10um (optimistic), σy = (70+700*(10^-1/2)) = 291 MPa. Which industry/metallurgical set practical max for Se (290MPa).

BCMW W2 1.5um grain, σy = (70 + 700*(1.5^-1/2)) = 641 MPa « at least 2.2 times stronger than 5160 with 10um grain.

Taking 62+rc into account, strength will increase by certain % more.

I am fine tuning grain refinement (aspect of ht 2.4) to get dia down to around 0.75um (σy = 878 MPa = 3x other steels at 10um grain dia).
 
Chopper with ht 2.4 in actions...

It's hard to show how edge of a 64-64.5rc W2 chopper perform in a chop-about. A 58-62rc would do the job too, except this 64rc W2 afterward edge remain more keen/sharp than one at lower hrc. Per high tensile strength, it should able to take on m4/3v/a2/5160/... blades with std ht at any hardness.

This edge geometry is on the verge of binding, so going thinner would be counter-productive. This chopper blade supports major bending (if one is strong enough to bend it).

W2 1/4" thick, 10" blade, 16" OAL
64-64.5rc edge via HT 2.4, spine is torched back to lower HRC.

15-16dps with 18dps microbevel
Thickness: bevel shoulder 0.023", 1/8" up 0.034", 1/4 up 0.067"

Chops: dried eucalytus, green oak & sycamore.


Edit: Received part to fix my ht equipment. In addition to on-hand 8, I will cut 24 test bars of W2. Try to put finish touches on few test & couple production FS knives - funding ht 2.4x.

When grain diameter is below 500nm - optical microscope will no longer able to discern grain boundary. SEM/BSED probably won't able to resolve grain dia less than 100nm. For 750nm dia grain, I will try prevent substantial loss of hardness - hope to stay ~64rc. Tempering could also result in rapid loss of hardness - it's the case when dislocation is excessive.
 
Last edited:
Result of 8 test sticks and 4 blades W2 steel 1/8" thick. Hardness between 65-65.5rc, much better than pessimistic projection of 64rc, which mean possibly there is room for one more level of grain refinement.

Grain diameter is getting very close to optical microscope resolution (~250nm - wave length spectrum of visible light)

DoElTAL.png
 

Attachments

  • upload_2017-6-9_16-58-15.png
    upload_2017-6-9_16-58-15.png
    2.1 KB · Views: 4
Luong,

You made me wanting that W2 you're holding :p :eek:
bluntcut said:
Chopper with ht 2.4 in actions...

It's hard to show how edge of a 64-64.5rc W2 chopper perform in a chop-about. A 58-62rc would do the job too, except this 64rc W2 afterward edge remain more keen/sharp than one at lower hrc. Per high tensile strength, it should able to take on m4/3v/a2/5160/... blades with std ht at any hardness.

This edge geometry is on the verge of binding, so going thinner would be counter-productive. This chopper blade supports major bending (if one is strong enough to bend it).

W2 1/4" thick, 10" blade, 16" OAL
64-64.5rc edge via HT 2.4, spine is torched back to lower HRC.

15-16dps with 18dps microbevel
Thickness: bevel shoulder 0.023", 1/8" up 0.034", 1/4 up 0.067"

Chops: dried eucalytus, green oak & sycamore.


Edit: Received part to fix my ht equipment. In addition to on-hand 8, I will cut 24 test bars of W2. Try to put finish touches on few test & couple production FS knives - funding ht 2.4x.

When grain diameter is below 500nm - optical microscope will no longer able to discern grain boundary. SEM/BSED probably won't able to resolve grain dia less than 100nm. For 750nm dia grain, I will try prevent substantial loss of hardness - hope to stay ~64rc. Tempering could also result in rapid loss of hardness - it's the case when dislocation is excessive.
Click to expand...
 
Got ht 2.42 done on a few W2 blades ~4.25" blade length, 65rc. 14dps/16dps-micro edge passed whittled ebony, African blackwood(AB), frozen cooked beef rib bone. Chopped AB: mostly fine, 1 visible riple, 2 micro rolled (visible with 10x loupe). I will try tune ht params for grain diameter toward 200nm.

Micrograph of W2 blade

s3wd7Xg.png
 
These 2 are thick and too thick for splitting, while works fine for capping a coconut. 12" blade, 1/8" thick, 1.7" wide, 1/2-3/4" sabre grind, 1 side micro bevel, would be an efficient all-day-long for a street vendor selling young coconuts :D
Chris "Anagarika" said:
Nice!!!
... though I'm not sure it's long enough for halving a coconut :D
Click to expand...

edit to add:

Tip and bending tests.
W2 0.13" thick, 4" blade
65rc ht 2.42 (nano grain)

Result of ~3" bending fractured angle ~44 degrees (or about 63.5degrees if bend 4" at 0.12" thick)

 
Last edited:
bluntcut said:
These 2 are thick and too thick for splitting, while works fine for capping a coconut. 12" blade, 1/8" thick, 1.7" wide, 1/2-3/4" sabre grind, 1 side micro bevel, would be an efficient all-day-long for a street vendor selling young coconuts :D


edit to add:

Tip and bending tests.
W2 0.13" thick, 4" blade
65rc ht 2.42 (nano grain)

Result of ~3" bending fractured angle ~44 degrees (or about 63.5degrees if bend 4" at 0.12" thick)

Click to expand...
I think you're right. A street coconuts vendor can't afford BCMW IMhO :D
 
I was going to do destruction bending test a bushcraft profile W2 1/8" 64rc blade but pain intervention happened. And then a friend asked for a blade to dig a 3" dia hole through 2x4 in less than 5 minutes. Hard to resist this type of fun. After a 10+ minutes trial dig, I put on a temporary handle...

kcPyTSnl.jpg


FmfyNmM.jpg


 
Last edited:
You must log in or register to reply here.
Back
Top