Heat Treatment - Crystal Weaving Foundation

[Chris "Anagarika"]

I just finished gathering my scattered brain reading your older post and recent exchanges with daberti.

If you didn't persist, there wouldn't be any CWF blade today. :thumbup:

 

[BluntCut MetalWorks]

Engineer brain so you, Chris!

LOL - I didn't want continue OT-ing Jim Ankerson's steel ranking thread. In this thread I try to balance shows with tells.

I am working on a batch of kitchen knives in S90V with ht 2.0 - aiming for 64rc. So far, my 2 thin s90V 64rc (ht 1.5) kitchen knives perform very well. Hopefully s90v ht 2.0 will be as good or better than 1.5.

I just finished gathering my scattered brain reading your older post and recent exchanges with daberti.

If you didn't persist, there wouldn't be any CWF blade today. :thumbup:

 

[foofie]

I am working on a batch of kitchen knives in S90V with ht 2.0 - aiming for 64rc. So far, my 2 thin s90V 64rc (ht 1.5) kitchen knives perform very well. Hopefully s90v ht 2.0 will be as good or better than 1.5.

This sounds perfect! I have been patiently waiting for kitchen knives, in a good stainless like elmax or s90v (my first choice).
Please save me at least one (or two):thumbup:

 

[BluntCut MetalWorks]

This ht 2.0 s90v batch is small (2 full tang parer; 2 stick tang parer; 1 hunter/util; 2 full tang 7" petty). I quick test a paring (DMT E 14-15dps, 0.010" BET) against AfricanBlackwood(AB); bamboo; cardboard - passed. Micro-chipped whittle dried beef rib bone. I will thin this parer and take edge down to 0.005"BET. If edge still passes whittle AB and light tap into bone - stay at 65-65.5rc.

This sounds perfect! I have been patiently waiting for kitchen knives, in a good stainless like elmax or s90v (my first choice).
Please save me at least one (or two):thumbup:

 

[BluntCut MetalWorks]

I've poured over my test data try to draw and project highest over all performance ht steel in 4 buckets of corrosion resistance (free Cr% after ht): B1) less than [0-2.5}, B2) [2.5-6}, B3) [6-10}, B4) [10+}. Performance covers wide range of tasks - includes high compressive and lateral impact load; cut abrasive materials such as rope+sand+etc... Thus must be very stable and has excellent wear resistance.

* peak performance projected be 64rc or higher hardness. However if spec required hardness be 61rc - high likely a different list.

B1 CruforgeV
B2 M2
B3 xwear (cruwear, zwear, etc..)
B4 Niolox

 

[BluntCut MetalWorks]

I sandwiched a test cruforgeV blade between anvil and 12 lbs hammer with one end of the blade dangling out. A hard sheer strike with a 2 lbs hammer - oh hey, this break look quite interesting ... at the same time - nonlinear break - which supports (1 sample statistic) with my assertion about CruV potential (hahaha better lucky than good)

 

[BluntCut MetalWorks]

With very limited testing data:

Tested Becut, RWL34 and Niolox. http://www.zknives.com/knives/steels/steelgraph.php?nm=BECUT,Niolox,RWL-34&ni=,525,49&hrn=1&gm=0
Although Becut only has 0.72%C but overall alloy volume is almost at RWL34 level (~18+%). Even at 62.5rc Becut isn't as ductile as niolox at 64rc. RWL 64.5rc performed great in kitchen & around the yard; however edge fail-mode is ~20um ripple then chip when yielded.

Wear resistant of becut and niolox are similar, but niolox edge easily out-last becut since niolox has much higher edge stability at (62, 63, 64, 65rc). I will lower RWL to 64rc and see it has any notable/saving-grace advantage over CPM 154 at 64rc. RWL/cpm154 is a nice mid wear resistant between niolox and cts-xhp.

Niolox

Edit to add: 20170419 ht 2.0 5 blades

CruforgeV 65rc - excellent keen thin edge with high stability. Fail mode: ripple and roll but no chip even with chopping Cumaru. Repeat careless whittle dried beef rib bone produced barely visible micro-chips.

W2 66-66.5rc - similar to cruV but less wear resistant.

52100 64rc - good but too eager to ripple and roll. I might re-ht the 2nd blade.

For low Cr% steels, it's quite challenging to perform interface-optimization(IO). 52100 blades probably received incomplete IO.

Edit 20170420:

W2 finished on 5K ws, a few strokes on balanced-strop at sharpening angle minus 5 dps ... aahhh smooth dry (actually oily face) shave

Edit 20170421:

Snapped a small coupon attached to cruV 65rc test blade. I couldn't break it with a small plier and barely did with a big 10" (oal) plier. It bent and straightened then finally broke. Quite a surprise - how a 65rc has this much plasticity - so I kept grind and re-check hardness - 65rc validated 5+ times.

Currently in weaving - 9 blades (1xO1, 2xW2, 6x52100). Target hrc: 01/65, W2/67, 52100/65. * result: broke 1 W2 blade (very fine grain). W2/67.5+, 52100/65.5+. Will lower hardness by 0.5-1rc. ** W2/67, 52100/65, O1/66 - on target: micro ripple prior to mini chips.

 

[Chris "Anagarika"]

Luong,

Nice dry oily shave is great!

Seems the CfV really hard & tough at the same time.
Didn't see these posts earlier . BF new S/W shows only limited posts from those I follow (used to be 'friends'). I need to familiarize again my forum habits.

 

[BluntCut MetalWorks]

Thanks.

65rc CfV is slightly stronger than 52100 and almost same as W2/67rc. However edge of these blades with 15dps keen bevel failed (rolled mode) the whittle test against dried cooked beef rib bone (the inside part). They passed whittle test against cooked pork rib bone. 64rc M4 passed both tests - yeah carbides add quite a bit of strength to apex - preventing apex from steer/deflect into roll.

To take advantage of CfV 0.75%V at high keenness, you need to finish sharpening with diamond/cbn abrasive. Otherwise carbides falled off and many remain on edges are protruding too much, which cause micro-chips. When whittle African Blackwood, CfV performs better than 52100 & W2 & O1. O1 is a good substitute for CfV if/when sharpening with non-diamond/cbn abrasive. LOL - W2, 52100, 10xx are great shaver.

This afternoon, I cut & profiled 8 CfV blanks. Resuming/re-activing my nano-grain (suspended ~ 2yrs ago due to nano-cracks). Ht 2.0 fixed weak links in the nano-grain ht. This is way more fun than finishing s90v, niolox, 10v, etc knives

Luong,

Nice dry oily shave is great!

Seems the CfV really hard & tough at the same time.
Didn't see these posts earlier . BF new S/W shows only limited posts from those I follow (used to be 'friends'). I need to familiarize again my forum habits.

edit to add comparison of broken surfaces:

Brittle - virtually no plasticity.

Vs BCMW

 

[Chris "Anagarika"]

It does look like dough (elastic)

 

[BluntCut MetalWorks]

HT 2.0 with nano-grain refinement was long and slow process through the weekend. Oh why not have some fun doing long preparation process for nano-grain... So, I mod a W2 0.370" thick chopper blank into a chomper (hence a slightly narrow tang).

Done about 20 minutes of chopping test(2x4, twisty pine, oak, cumaru, cooked pork rib bone) with the chomper = easy passed. Umm - calls for more fun...

I will buy a couple riped coconuts (black shell) and lop their top off - similar to image below, except use 1 chop with the chomper. At 66+rc, edge will damage - how big/bad?

 

[BluntCut MetalWorks]

Lop top of young and ripened coconut using w2 chomper and cruforgeV. didn't conducted well.

Result: no visible damage but catch when slicing phonebook paper. Under 22x magnification - There are W2 micro rolls and chips, and CruforgeV micro rolls.

 

[Chris "Anagarika"]

I'm confused. 66HRC only microchipping is good in my experience.
Remember the D2 went through younger coconuts?

 

[BluntCut MetalWorks]

Re-hrc-test W2 is about 66.5-67rc. Yeah, micro rolls+chips (under 22x mag) is very good for hacked through black-shell coconut.

Slow cook coconut chicken for dinner. CfgV blades hacked through all bones (including leg & thigh).

I'm confused. 66HRC only microchipping is good in my experience.
Remember the D2 went through younger coconuts?

edit 1: 1084 67rc 7.5" bread knife

 

[BluntCut MetalWorks]

Eyeball guesstimated grain diameter between 0.7um and 1.5um (ASTM size between 16 and 18). For context/comparison - standard/industry ht cfv or Cr% steels grain dia 8um-10um-coarser (best-avg-poor).

Edit to add what nano grain martensite matrix looks like (EBSD image):

Edit 2: Oh curse - I think (a big MAYBE) of a way to simulate cryo at LH2 temperature −423.17 °F/−252.87 °C. If result matrix is not destructive (nano-cracks), it should gains additional wear resistance and elasticity.

Edit 20170427.10: HRC(avg 3 reads) before and after SimLH2 - CfV 65.4 vs 65.9. Insufficient data -- due to +- 0.5rc hrc error margin, more larger sample data is needed. Grind went faster but again could be reduced(@30%) belt speed allows 50grit belt to dig deeper. Soon, I will test edge stability by whittle African Blackwood and dried cooked rib bones (pig & cow).

*11: 14-15dps DMT F passed whittle bamboo, AB, dried cooked pork rib bone. Micro-chips and ripples whittled dried cooked beef rib bone.

Resharpened to 15dps finished with 2K waterstone. Hard chopped cumaru wood and thick aluminum tube - edge under 22x mag is mostly rolled + rippled and didn't any chips.

Next will try SimLH2 for W2 & 52100 blades... If positive results keep coming - will apply SimLH2 to high alloy steels (10V, s90v, etc..)

 

[BluntCut MetalWorks]

I simulated cryogenic ht step at liquid hydrogen temperature (-253C), which is colder than LN2. Seeking a small gain in edge retention, plus some extra elasticity.

Blades (all LN2 quenched, so simLH2 is an extra step):
1 - W2 0.125" thick, 67rc HT 2.0 + simLH2
2- 52100 0.130" thick, 65rc HT 2.0
3 - 52100 0.130" thick, 65rc HT 2.0 + simLH2
4 - CruforgeV 0.098" thick, 65.8rc HT 2.1 + simLH2

W2 is close to CfV in all test aspects, except it suffer a bunch of 1mm chips when hacking a stout aluminum tubing. 52100 edge seem rolling too easily - I grinded the edge back a couple times and a 63rc blade skated against it, so not sure why its edge behaved this way (when whittle AB+bone and chop cumaru). To be fair, I need to apply ht 2.1 to W2 and 52100 (was using old blades w/o prepared for nano-grain). The W2 chomper as ht 2.1 and it closely matched CfV; however chomper is not in video below

 

[BluntCut MetalWorks]

Cryogenic isn't a magic-box, once given enough thoughts on conceptual and consequential, it's simple. If LHe2 is cheap and readily available as LN2, it would be a cryo method of choice. *note LHe2 is colder and safer than LH2 when boil/evap in gas form.

Crystal is atomic suspension-in-place by EM Field. Although weaker than other bonds with shared outer radius electrons, lattice structure allows larger working unit thereby more compressive strength per unit.

In order for lattice to ever possibly cool to near 0K - stress/dislocation must be resolve/gone. Stress/dislocation differential between orderly neutral lattice vs colliding lattices greatly amplified as temperature heads toward EMF disappearance. Naturally stress units will fracture and transform to more compatible form and shape of which resultant structures has lower potential energy. My simLH2 (or simLHe2 or near BEC) tries to generate this type of reaction/transformation without the required extreme cooling in the neighborhood of BEC(https://en.wikipedia.org/wiki/Bose–Einstein_condensate).

Tempering is thermal expanding side of stress relieving. Non-diffusive precipitate is cementite. In contrast, SimLH2 is compressive, where precipitate is pre-eta carbide (hcp). Akin to making diamond dust with explosive. Is it risky? You bet! It is why I break the CfV blade to re-assure matrix microstructure didn't compromised by simLH2 step. A globally compromised matrix would has low elasticity and strength because myriad internal nano/micro-cracks, which initiate cracks propagation under stretch/compress vs rotation+elongation when not compromised.

Applicability in cutlery? Almost bupkis in 100m stroll (lower than 62rc and gentle usage). Lower hardness of CfV above to 62rc, it would be hard to tell the difference vs std ht CfV 62rc. Lower to 60rc (bcp probably changed to cementite at this temnperature) - it probably be a draw.

* Submarine dive - surface when ping *

 

[BluntCut MetalWorks]

Bending Test: BCMW HT vs BF/Std HT

W2 0.130" thick, ~ 3/8" wide, 8" long

Bending test ~6" length: Elasticity, plasticity ranges and fracture point. Each set tempered with at 275F and 475F.

All samples(wrapped in stainless steel foil to prevent decarb) normalized at 1650F 30 minutes soak.

Industry Standard HT for Control1 and Control2:
Grain refine thermal cycles (1550F, 1500F, 1450F). Control2 with extra Sub-critical Annealed. Then both aust to 1460F 10 minutes soak, quenched in Parks 50.

Control1: 65.5rc@275F tempered, 60rc@475F tempered
Control2: 66rc@275F, 60rc@475F
vs
BCMW
HT 2.2: 67rc@275F, 61.5rc@475F
HT 2.3: 67rc@275F, 61.5rc@475F

Since the video is 29 minute long, so here is a summary (snapshots at particular time in video)

Edit to add ~max bending angle (using protractor) before fracture/break for 275F tempered samples:
Control1 - 42°, Control2 - 51°, HT 2.2 - 70°, HT 2.3 - 69°

 

[BluntCut MetalWorks]

From the image above - HT 2.x improved elasticity range over controls by more than 30% ((69-51)/51). From the video above - HT 2.x samples fractured more explosively than controls. When a load of stored/potential energy suddenly release, a higher load = higher explosiveness.

Implications/Applications? For edge-_tools/knives - HT 2.x brings x% higher edge stability+strength+toughness. For industrial (such as structural, spring) - possibly using x% less steel per specs. x% = tbd or to-be-calculate.

Can a knife user tells the difference? Probably not, unless in extensive usage/testing.

*** Summarize HT 2.3 Components - list by order of ht procedures ***

1. Normalize/reset - aim to attain highest uniformity of elements distribution
2. Grain refinement - smallest as possible, working in conjunction with #3
3. Grain Boundary(GB) cleanliness - avoid unwanted elements/particles/formation/deposition in GB
4. Pre-Nano grain setup before final hardening (for low Cr steels)
5. Hardening austenite - temperature&time critical. Additional procedures for subset of steels - Grain refinement + interface optimization
6. CWF - create optimal condition INSIDE A GRAIN
7. GB, Interface and HCP optimization. From now on - I will call this optimization as BEC

'A chain is only as strong as its weakest link'. Definitely, there are plenty of more room to strengthen bottom-feeder weak links.

 

[BluntCut MetalWorks]

Another ht improvement idea popped in my head, sort of CWF' (prime), hence ht 2.4. I squeezed time to ht 2.4 4 w2 test bar identical to above. In a hurry, I messed up one (and possibly compromised part of another step). Anyway, goal is to find out HRC point enable to hit 90° bend. Results are far below my expectation/projection:

CWF' step didn't negatively affected results (if it has, certain characteristics of failure would show up) and IDK whether it has any benefit. Hopefully time permits for another test pursue at 90° bend (60° flex) at 62.5+rc.

** Correction Angle Bias for previous 2 posts **
My setup has ~5° vise tilt bias toward pulling force, so stated angles should be 5° lower

Edit 20170518:
1. Identified cold tong/plier jaw handling/pickup spots localize weakened/enbrittle test bar, leading to pre-matured fracture/snap degrees, esp at 62+rc hardness.

2. Later today, I will cut at least 16 bars - same dimension as prev tests, except only 5/16" wide so I can bend easier than 3/8. I will add an extension to the clamp to increase leverage for better bending control. Keeping same length for consistency for now but future test bars will be 9.5" length - allows 1.5" for ht handling. Yeah bending length still be 6".

Edit 20170522:
Tested ht 2.4 w2 62rc & 64rc bar. 62rc bar messed-up in blank & tempered, so didn't break properly. 64rc bar showed good elasticity to at least 64° - unfortunately, I forgot to incremental bending to figured out max elasticity angle. Nevertheless, this bar exploded at 74° and shown very little elasticity - therefore I would guess elasticity angle is probably around 70°.

Also ht 2.4 a 1084 test blade - 63rc but with avg grain dia = failed. *note: combined ht steps didn't worked well for carbon below 0.9%

Edit 20170523 AM: Sparsely write down some of my thoughts - reckon: speaking out loud to myself

Edge stability reflects tensile yield strength and elasticity potential energy. It's mostly a side/lateral load the edge can sustain/support. Bending a test bar provides macro view for intergrain affect from lateral load. More grain boundaries = higher potential energy before grain dislocation forced to move/displacement.

Grains are interlocking and have limited rotation before displacement (yield). Beyond gb yield is fracture (99% so, unless gb is thick layer of ferrite(umm not good)). So for elasticity, this gb rotation contribute is (projecting) a 10% factor.

Individual grain plays 90% factor in elastic. Microstructure constituents(martensite, ferrite, ra, particles, elements) are interlocking, where elasticity limits by largest unit elongation; rotation and proportional elastic (where elastic is stronger than yielded, thus some % of back/forth yielding within microstructure). D2 for an example - 20+um carbides are largest unit within grain, which doesn't elongate nor rotate well, thereby easily yield. However when combination of ferrite and RA exceed percolation threshold, it becomes largest unit which has high plasticiy (substantially lower tensile yield strength but extensively add range to plasticity before fracture).

When largest unit is martensite unit block. Tempering reduces unit-length and increase ferrite (precipitate carbon from mart matrix to form cementite and free fe). Shortened unit has less defects, therefore it elongates more (theoretical elongation upper limits is ~25%) + supports more rotation however ferrite+cementite slightly lowered tensile yield strength (obviously Fe has very low yield strength). When units moved/displaced (a piled-up against other units aka work hardening) and will fracture after that. Mostly referred to less than 20% of ferrite/ra, whereas grain level ductility percolation threshold near 40%.

In order to bend 6" 0.130" thick 90° without fracturing (E° + P°)>90°:
a) 1-2um grain diameter, E° upper limit ~65°, so P° be at least 25°. P25+° requires drastic changes in grain shape, so ferrite% need to high ~15%
b) 0.6um grain dia (doubted a mart grain can get this small), E° can gain ~12°. so 10% ferrite maybe sufficient

Ties back to my long time assertion on - micro ripple/roll and macro chip edge - as desirable. Taking a) at 4-8% ferrite in mart matrix - about 65rc hardness. Edge cross section about 0.005" thick would easily supports 90° bend/ripple. While cross section thickness > 0.01" will has very little plasticity - therefore excessive force will end up with macro chip.

For wear resistant - Ferrite% on apex is the weakest/softest link. High carbide volume steels can help shields apex Fe wimps from some uses (smaller % at high keenness and shielding increase at wider apex and depth) - downside of carbide is lowered elasticity. So carbide is useful when dulling via wear and actually counter productive when wear mode is fracture.

Referencing
https://en.wikipedia.org/wiki/Yield_(engineering)

Taken BCMW HT 2.4 W2 at 64+rc, where ferrite+ra % is below 3%. Elastic(3) and Proportional(2) Limits are fairly simple to detect. Plasticity range is very narrow (between ref 3 and 4) - which look as if 50/50 from ref 2 to 4. Proportional means elongation+rotation is much stronger than displacement - dislocation move back/forth. However low ferrite% severely limit crystal slippage (forced by stretch+compress of dislocations). If this is actually the case - True Elastic Limit has much wider range. When operating/using below True Elastic Limit, there is no (statistically speaking) matrix fatigue. Consider this as no-mechanical movement system. Between TEL/1 to PEL/2 maps to life span of tools/blades based on cycles of crystal mechanical movements (slip back/forth). Just an obvious - we want TEL as large as possible - incredible long service life span. Longevity wise - Yep, it is more useful in a car leaf spring (thinner+shorter) than edge tools.

Edit 20170523 PM:
With grain dia within target range, I will do more bend test for 67rc, 63rc ,62rc bars. Afterward make a 1/4" thick W2 chopper for abuse - maybe 65rc edge, below 60rc spine.

2 big bars of niolox are coming ... more relaxing to ht but a lot harder to grind than W2.

1084 test blade above has inconsistent grain size: very large parent grain with thick boundary, and sub grain range from 1-8um. Confirmed my previous assessment on 1084 has insufficient carbon(use cementite to nucleate/pin grain) to achieve nano-grain.

Edit 20170524:
Tested 63rc, 63rc and 67rc bars - Elastic angle about 65° and fractured between 70-75°. At 0.130" thick, lowering hrc doesn't helped (blade/bar wise) until ferrite% beyond certain threshold. For W2 (and similar steels) I think, this ferrite% threshold is between 61-62rc. Thinner cross section (edge bevel) requires less ferrite% for supporting rotation displacement (micro-ripple & nano-roll).
HRC intended uses:
A - only edge interaction (no prying, side-whack, etc)
Slicer: 65+rc
Chopper: 64rc edge, softer spine
B - physics involves blade (flex, whack, twist, etc.)
Slicer: 65+rc, softer spine
Chopper: 61rc, softer spine

Edit 20170525 AM:
C - Edge steering activities (deep chop into dense materials, poor technique batoning, etc..)
58-60rc highly elastic and good range of plasticity (stay bend)
49-54rc allow peen/pry-back correction on bent part *note: work-hardened upon correction

Ht-ed 2.4: 4x w2 test bars, w2 chopper, 52100 slicer, cfv hunter. Will bend a ~58 and 60rc bars until fracture - to extrapolate fracture angle for cross section thicker than 0.130". Apply A) to chopper - except will test flex/whack/twist.

Separating 2x 8" 60-60.3 rc w2 test bars connected together .157" length. observed 5.7° angle at unconnected end (not very accurate measurements) - it elongated 33% (0.4/1.02") before teared. Theoretically elongation is about 25%. So conservative assessed elongation for this w2 bars is close to 25% theoretical elongation limits of mart+fe matrix.

Edit 20170525 PM:
Absent minded bend test 60 & 58.5rc bars w/o removed bandsaw edge and pre-ht 60 grit belt lines. In spite of lousiness - 58.5rc 4" bend length managed 76 degrees. I've 2 more bars to temper down to 60 & 58rc and try again - will prepare better next time.

W2 1/4" thick 16" oal chopper (more like a long heavy knife not very wide) 64+rc edge soft spine is a fun hard/abuse-use with confidence... end up chopped around the f/b yard for a couple hrs. will field f&F grind and put a handle on this one.

Edit 20170526:
W2 chopper (16.7 oz) chopping edge stability for 0.031" 1/8" up from apex, 0.070" 1/4" up from apex (starting to bind). 1-2K waterstone finished grit.
10 dps - 2x4, oak, green woods
12 dps - at threshold of passing rosewood
15 dps - threshold for katalox wood
17-18dps - threshold for Lignum Vitea Argentine and bones.