Thread: Hardening of sawplates

Originally Posted by Maple71

I find it interesting that someone is suggesting cryogenic treatment. What grade of steel are you going to do this to? Does this grade of steel respond at all or are you just doing this for fun and to have another variable. Cryogenic treatment does not work on all grades of steel and is only used on some grades where it is needed. For instance, it greatly helps grades such as A2 where there is a significant amount of retained austenite after normal quenching. This is true of quite a few grades of the higher alloy steels. The cryogenic treatment causes the retained austenite to transform to martensite and then needs to be tempered properly. There are many grades of steel where cryogenic treatment makes no sense at all. BUT...in order to know if one will benefit, you need to know the grade and/or chemistry. Besides, the transformation of austenite, the cryogenic treatment can also help produce some smaller carbides. The results of such a trial are a little meaningless unless you know the grade of steel and the current microstructure before treatment. If this is just a fun thing to do and you do it without knowing what you have, it really does not apply to anything else except the test sample. If you are serious about this type of thing, I would suggest that you learn a lot more about heat treatment and the effects of different grades of steel.

AISI1095

Oil quenched 1095 has a fair amount of austentite structure. http://link.springer.com/article/10.1007/BF02656400

http://vacaero.com/information-resou...austenite.html

Here's a few references on cryotreatment of various types of steel.

This isn't copyrighted so here's the full text.

Verhoeven Metallurgy of Steel for Bladesmiths and Others who Heat Treat and Forge Steel Part 1.pdfVerhoeven Metallurgy of Steel for Bladesmiths and Others who Heat Treat and Forge Steel Part 2.pdf

Not the same alloys but interesting and also not copyrighted.

The Effect of Cryogenic Treatment on the Fatigue Life of Chrome S.pdf

Another thesis, no copyright.

Thornton Investigating the Effects of Cryogenic Processing on the Wear Performance and Microstru.pdfThornton Investigating the Effects of Cryogenic Processing on the Wear Performance and Microstru.pdfThornton Investigating the Effects of Cryogenic Processing on the Wear Performance and Microstru.pdfThornton Investigating the Effects of Cryogenic Processing on the Wear Performance and Microstru.pdfThornton Investigating the Effects of Cryogenic Processing on the Wear Performance and Microstru.pdf

Behind a paywall but it's not just austentite...

http://www.sciencedirect.com/science...43164813003979

Copyright applies but it is open access.

http://www.seipub.org/fwr/paperInfo.aspx?ID=2989

Working...

http://www.calculatoredge.com/metallurgy/hardness.htm

Preliminary observations:

1) Rockwell N scale testing is particularly sensitive to the surface roughness of the test article, especially the 15kgf range.

2) The results of HRN30 and HRN45 testing both correspond within the bounds of instrumental error with each other and with 150 kgf HRC testing for these specimens (0.035" thickness 1095).

3) Differences between neat, ground and ground + hammered specimens persist through all stages of cryogenic treatment at -80 ^oC.

4) For the ground + hammered specimen, apparent hardness decreases initially during cryogenic treatment, likely due to cryogenic treatment induced stress relief, a well known phenomenon. The final hardness of the ground + hammered specimen remains higher than both the neat and ground specimens.

5) The ground + hammered specimen seems to become more uniform in hardness with increasing cryogenic soak time.

6) Cryogenic treatment generally increases 1095 ground and ground + hammered steel hardness as measured using HRN45.

7) The effect of -80 ^oC cryogenic treatment is essentially complete at 4 hours of soaking.

8) Neat 1095 carries some stress induced hardness which is relieved by cryogenic treatment at -80 ^oC.

Is five test points enough?

PS logging enabled 2/1/2015 6:47:12 AM
Version 3.1.2

Suggested citation:
Dupont WD, Plummer WD: 'Power and Sample Size Calculations: A Review and Computer Program', Controlled Clinical Trials 1990; 11:116-28.
or
Dupont WD, Plummer WD: 'Power and Sample Size Calculations for Studies Involving Linear Regression', Controlled Clinical Trials 1998; 19:589-601.

--------------------------------------------------------
Type of study: T-test
Requested output: Sample size
Design: Paired
alpha=0.05 power=0.8 DIFF=2.07 SIGMA=1.13 M=0
Sample size=5
t-test confidence interval width=3.122618

We are planning a study of a continuous response variable from matched pairs of study subjects. Prior data indicate that the difference in the response of matched pairs is normally distributed with standard deviation 1.13. If the true difference in the mean response of matched pairs is 2.07, we will need to study 5 pairs of subjects to be able to reject the null hypothesis that this response difference is zero with probability (power) 0.8. The Type I error probability associated with this test of this null hypothesis is 0.05.

Yes, five paired samples is enough to reliably detect the differences observed given the measured error levels.

While doing all of these other measurements I thought to look into the issue of how the method of cutting 1095 alters the hardness in the area of the cut edge.

I took a scrap piece of 0.035" 1095 that was still in the blue, clamped it between two blocks of soft pine in a vise and cut it off flush with the top of the holder blocks with a nearly new 6" diameter 0.065" abrasive cutting wheel running at 9000 R.P.M. (235.5 f.p.s.), the cutting time was just a few seconds and the edge discoloration was minimal. I dressed the edges manually with a file and measured hardness with the 150 kgf Rockwell C tester as close as I could get to the edge.

Here are the results.








The values for measurements 1 mm from the edge were not plotted as it was clear that the effect of plastic flow at the edge was throwing off the readings.

It is however apparent that abrasive cutting wheels cause statistically significant reductions in hardness up to 4 to 5 mm in from the cut edge. This effect will likely be more pronounced in thinner gauge 1095 because it has less thermal mass in which to dissipate the heat generated by the abrasive wheel.

Nice work Rob, I cut all my sawplate with an angle grinder and thin cut-off wheels, the only thing different to your test, is I clamp the plate between 1/4" mild steel angles, mostly to provide a guide for a straighter cut, but also for a bit of heatsinking.... now you have me wondering how effective my heatsinking is...

Ray

Originally Posted by RayG

Nice work Rob, I cut all my sawplate with an angle grinder and thin cut-off wheels, the only thing different to your test, is I clamp the plate between 1/4" mild steel angles, mostly to provide a guide for a straighter cut, but also for a bit of heatsinking.... now you have me wondering how effective my heatsinking is...

Ray

Thanks Ray,

1095 seems to be kind of touchy when it comes to heat. There was no visible discoloration extending more than .5 mm or so away from the cut line yet the decrement in hardness extends pretty deep into the body of the steel.

I've been lurking in the knife makers forums and the opinions about 1095 and cryo treatment are somewhat divided. The consensus among those that apparently have more knowledge of the science seems to be that it gains some hardness (remember their target hardness is 65 is or so). Their understanding of metallurgy is often quite advanced. The opinions and experience of those posters who sound most credible is congruent with the conclusions related in the scientific literature.
Some of the knife makers are also of the opinion that 1095 can be quite a variable product as far as the hardness goes.

The next step for my work here is cryo processing at -196 ^oC with liquid nitrogen.

Cheers,
Rob

Knife people are funny folks. I guess we're all a little off center.

As a non knife fanatic, 1095 hardened to the high 50s is my favorite pocket knife steel. They (the knife makers) are worried about its toughness, which has never been an issue that I've seen (though it would be at 65 hardness).

They (the knife makers) make a lot of quarter inch thick knives out of very hard to sharpen steels, steels that tend to fail with tiny microchips uniformly across the edge (which gives them the reputation of holding a mediocre edge a long time). I could never figure that group out, and what they do with all of those ultra fat and expensive knives.

When you say is five test points being enough, do you mean five strikes on the same area or five different spots on a plate?

Originally Posted by D.W.

When you say is five test points being enough, do you mean five strikes on the same area or five different spots on a plate?

No, only that I took enough measurements of the areas sampled to achieve the reported statistical power. This is a nice description: http://en.wikipedia.org/wiki/Statistical_power
The measurements points were spread across the test articles with each point separated from any others by at least 5 indentation diameters. These samples were the test swatches pictured at the first post discussing hammering of the 1095 steel, see post #103 on this thread.