Thread: Saw File Hardness Measurements

Originally Posted by FenceFurniture

I was looking for a pattern there but three are up and three are down which doesn't really prove much. Surely the Japanese file (citric) is an aberration - I can't see the graph whilst replying but the drop in hardness is something like 30 Rc - can't see an acid bath doing that.

If the acid treatment roughened the surface of the file significantly the measured hardness will decrease. I suspect that the alloy of that file may have been more eroded, unfortunately I don't have a surface roughness instrument (yet).

Originally Posted by rob streeper

I'm thinking of processing of new files too if it's okay with you. Given that nobody is satisfied with the majority of newly manufactured files the cryotreatment process may offer an economical and accessible way of making them perform somewhat better.

Yebbut you only have one of each, so nothing to compare it against.

Maybe wait until I get the nest order of Liogier files in (2 months?) and I'll get some excess to be treated - enough so that the comparison can be done 4 times over, and perhaps three people using one corner each of the same file.

Originally Posted by FenceFurniture

Yebbut you only have one of each, so nothing to compare it against.

Maybe wait until I get the nest order of Liogier files in (2 months?) and I'll get some excess to be treated - enough so that the comparison can be done 4 times over, and perhaps three people using one corner each of the same file.

I'll do it. I should have a big batch of saw plates ready for treatment by then so all can be done together. I'll also make up some test swatches of 1095 steel for distribution with the files.

Originally Posted by rob streeper

If the acid treatment roughened the surface of the file significantly the measured hardness will decrease. I suspect that the alloy of that file may have been more eroded, unfortunately I don't have a surface roughness instrument (yet).

Ok, in that case then testing a ground area (that will have the same roughness) should sort that out. It may have been just one really crap file in the batch, but that's kinda counter-intuitive. I'd back Japanese manufacturing consistency over Indian any day of the week.

Originally Posted by rob streeper

Here's a hardness vs. user rating plot with each file identified.

Attachment 345148

Suggestions for improving readability.
List all makers in groups and then use the same colour or symbol shape for the same maker and then use different colours or symbol to identify specific files of each make
eg list all Pferd files one after the other on the list, use all squares to identify these and the use different colours to identify the specific Pferd files.

I reckon you are trying to show to much on the same graphs. e.g. I suggest not putting the Citric acid data on this graph - use a separate graph for these.

The citric acid data looks like it's highly problematic and like you say going to affect the surface. The only area of the files that should be measure is the working area and requires the removal of the teeth and this is likely to remove the effect of the citric acid - anyway you can test this ?

Originally Posted by rob streeper

I think that in the cases of files 4A, 7, 7A and 12 the annealing of the tang softened the steel quite far up the bodies of the files - this could be considered a measure of manufacturing consistency or lack thereof. Note that the tip measurements for those files are higher. The plates I've prepared thus far are based on the average of all measurements for all files and thus the files with softer tangs will lower apparent hardness. I'll work up the figures with the tip data alone later.

Interesting findings Rob, most people do not use the file at the tang end very much, nor tip for that matter (because they are hanging onto it), and with a used file, there is often still good teeth towards the tang.

Cheers
Peter

Originally Posted by BobL

Suggestions for improving readability.
List all makers in groups and then use the same colour or symbol shape for the same maker and then use different colours or symbol to identify specific files of each make
eg list all Pferd files one after the other on the list, use all squares to identify these and the use different colours to identify the specific Pferd files.

I reckon you are trying to show to much on the same graphs. e.g. I suggest not putting the Citric acid data on this graph - use a separate graph for these.

The citric acid data looks like it's highly problematic and like you say going to affect the surface. The only area of the files that should be measure is the working area and requires the removal of the teeth and this is likely to remove the effect of the citric acid - anyway you can test this ?

Yeah I know the graphics can be improved, lot's of work though. I'm also thinking about coding the datapoints according to the buy/maybe/don't buy suggestions of the original analysis.

One of the conclusions I can draw from these data is that hardness is a necessary but not sufficient condition for user acceptance/usability. Some of the least acceptable files were also hard. Variability in hardness tends to decrease with increasing hardness. I think that the uniformity of the toothing of the files is probably the primary determinant of acceptability. Low hardness and high variability in hardness are bad signs but hardness is not a guarantee of usability.

Originally Posted by Heavansabove

Interesting findings Rob, most people do not use the file at the tang end very much, nor tip for that matter (because they are hanging onto it), and with a used file, there is often still good teeth towards the tang.

Cheers
Peter

For this go round I decided to test the files as they arrived. Next step is to grind off the teeth and test again. W-1 (if that's what these files are made of) is a 'shallow hardening' tool steel but by shallow they mean about 8mm apparently thus not a problem for this work. Not much in the way of info on the alloys used for files but W-1 is mentioned.

Cheers,
Rob

I wouldn't spend any time at all with the Citric Acid files Rob. It was a uniform failure and so any data is irrelevant anyway (IIRC the only treated file that showed any improvement was still terrible). The only time trying to unravel a failure would be worthwhile is if it can lead to improvement but I doubt anyone in their right mind would CA treat a new file.

Originally Posted by rob streeper

One of the conclusions I can draw from these data is that hardness is a necessary but not sufficient condition for user acceptance/usability. Some of the least acceptable files were also hard. Variability in hardness tends to decrease with increasing hardness. I think that the uniformity of the toothing of the files is probably the primary determinant of acceptability. Low hardness and high variability in hardness are bad signs but hardness is not a guarantee of usability.

I agree that hardness not likely to be the only factor in file acceptance but given the current data is not from the working area of the files I don't think any conclusions relevant to working files can be drawn from this data

The amount of effort that has gone into this experiment is indeed impressive and you guys are to be congratulated for this. To reduce your workload it would have been better to just target the business area of files i.e. don't worry about tang and heel measurements. The citric acid effect on hardness is immeasurable so I would have just left this out

I dislike creating more work for you and I'm sure you also know this, but testing just one file from each brand is also fraught with problems. It's like consumer testing " one of" a product from a manufacturer, every now and then, even quality manufacturers produce a lemon or indeed several.

The first thing I would have suggested is a pilot with just say 4-5 files of just two brands as this would have uncovered most of the experimental problems etc. If the hardness variability of files from the same manufacturer turns out to be greater than the difference between the manufacturers then that changes things significantly, even to the point where it may not be worth pursuing the rest of the experiment.

A number of times I have suggested that forum members about to embark on experiments should consider posting their design before they create a lot of work for themselves. Some of us have extensive experience in experimental design and reviewing of project proposals for national and international experiments.

Originally Posted by BobL

.... but testing just one file from each brand is also fraught with problems. It's like consumer testing " one of" a product from a manufacturer, every now and then, even quality manufacturers produce a lemon or indeed several.

The funny thing is Bob - I deeply suspect we have actually tested more than one file from the same manufacturer. Quite a number of the brands are made in India and by crikey they look the same. Same crappy black ink for the logo, same looking steel, same almost nil taper, same coarse grinding of the blank, same lousy performance except for one or two (musta been having a exceptional day). I know that's nowhere near conclusive, but you get my drift.

This goes for Grobet Swiss, Grobet USA, F.Dick, and Pferd although theirs may be from a different factory. Apparently Pferd used to actually own their own factory in India but it got too hard and they subbied it out.

There are quite a number of other brands in the same kettle but they weren't tested (When researching I got to the point where I could spot an Indian file a mile away and figured I had already collated enough of them.....)

Here's an example of two of the files:


Spot the odd file out.....

The top and bottom are two "different" brands. The name was not shown so as not to cause any bias to the testers (the pic was originally taken for the monster thread two years ago, which just preceded the testing of the 23 files).

Originally Posted by BobL

The citric acid effect on hardness is immeasurable so I would have just left this out

It was originally done to see if there is any improvement or otherwise to the file's performance - not specifically hardness but just "cutting power". There have been many claims that acid sharpening of new (and old) files improves the performance, but I think we pretty clearly mythbusted that (at least for new ones).

Maybe there was a bit of hydrogen embrittlement happening - who knows - but I do know the CA treatment is nil use and causes files to perform even more poorly (with one or two aberrations probably due to file inconsistency). Just not worth pursuing.

Originally Posted by FenceFurniture

It was originally done to see if there is any improvement or otherwise to the file's performance - not specifically hardness but just "cutting power". There have been many claims that acid sharpening of new (and old) files improves the performance, but I think we pretty clearly mythbusted that (at least for new ones).

Maybe there was a bit of hydrogen embrittlement happening - who knows - but I do know the CA treatment is nil use and causes files to perform even more poorly (with one or two aberrations probably due to file inconsistency). Just not worth pursuing.

Crikey I would never have thought of doing it to new files.

Originally Posted by FenceFurniture

Rob, I guess you're trying to see if you can increase the performance (hardness?) of them? Don't forget that these files are knackered now and performance increase (or not) won't be able to be measured.

EDIT: However, it would be very interesting to treat about 4-5 new files and compare them against untreated.

That's the idea. The hardening effect of cryotreatment is apparently more modest in magnitude than is the improvement in scratch resistance or toughness. The older dogma was that the cryotreatment was only effective for martensitic alloys. However more recent data and opinion is that most ferrous alloys can gain some benefit either in hardness or in scratch resistance. For instance, disk brake rotors that are cryotreated apparently last about twice as long as untreated parts. Springs that are cryotreated retain their load carrying capacity longer. Why not for files? Technically they're a cutting tool but there is also a foreseeable benefit in increasing the scratch resistance of the files used in saw making. Some sources state that for some alloys cryotreatment can double the scratch resistance. Is a similar effect possible with saw files? We have the technology...