Calibrating Hardness Tester, need known hardness material

[rattletrap1970]

Hello all,

I'm working on a hardness tester and it's going pretty well. I'm looking for some samples of material that are (as accurately as you can possible determine) known hardnesses.

It works much like the Cabine with some refinements. Mostly I'm bored and wanted something to play with.

So.. if somone can help me out i'd really appreciate it. The more known different hardnesses the better. Then I can generate a curve that fits the tester.

Thanks!

[Calamity Jake]

If you can find a bench rest shooter in your area that makes his own bullets he will have lead wire for making cores
it should be certified as pure and be BHN 5.
Or maybe see if rotometals can provide some certified pure lead or other known hardness material.

[dbosman]

Does the University of Connecticut-Torrington campus have an engineering building/program?
If so, someone on campus has a calibrated hardness tester.

[dikman]

rattletrap, I'll be very interested to see what you come up with, design-wise (seeing that I cobbled together my own version of it).

[rattletrap1970]

I don't really know about the university up here. I just figured if someone had maybe 10 samples of known hardness that they just tested in little labeled bags, I could test it in mine and that would establish the calibration curve. If I get a chance today I'll put a rendering off the CAD system.

[dikman]

I look forward to it. Mine, of course, isn't calibrated and without samples of known hardness to test it's virtually impossible to do so. I've been doing quite a bit of testing with stuff that I do have, and the results are quite interesting. Testing thick sheet lead is pretty consistent, giving a reading of 36-37 on my scale, and it's generally agreed that this will give a BNH of 5, so that is the bottom of the scale. From there on it becomes a bit murky. I got hold of a few WW's today and melted them into five small round ingots (I'm using this as a fixed size so that all my samples have the same diameter/thickness). These also tested pretty consistently, reading 48-50. As to the BNH, who knows? It could be anywhere between 12-15, I'm guessing. Not that it matters, as I only made it to use to test for comparative hardness between different samples (primarily range scrap).

I made up samples of range scrap (which tests harder than the WW's) mixed with pure lead in ratios of 1:1 - 1:5. I took seven readings on each of several samples of each ratio, as I found that there could be variations in the readings even within the same ingot being tested. This wasn't what I expected. I then averaged them out to give me a final figure for each ratio. I'm not completely happy because of the variations I was getting, but for now it will have to do until I can figure out why they varied.

What it has shown is that I can identify range scrap and pure lead (not surprising, I guess). 1:1 and 1:2 test very close together, only a little below the "pure" range scrap. 1:3, 1:4 and 1:5 also test very close (I'm not sure I could distinguish between them if I was given unmarked samples to test), significantly lower than the other two samples and noticeably higher than pure lead. My main aim was to try and find out how much lead I needed to mix with the range scrap to make it soft enough to use in my Remmie 1858 - it looks like at least 5:1 lead:rs, possibly even 6:1, which means I need a lot more lead than I thought for the amount of rs I have!

While this is hardly a scientific test, it's the best I can do with what's at hand, which is why I'm interested to see what you come up with.

[rattletrap1970]

Well, I have (at this point) no hardness tester of any kind. So. If I were to have some nice .45 caliber bullets (In varying hardnesses, as many hardnesses as someone could come up with) donated from someone who DID have a good quality tester, and if they did multiple readings on the same bullet, and sent said samples to me in a baggie with the hardness written on said bag I could then test them on my rig.

[rattletrap1970]

Here are the renderings from the one I'm making.

http://i885.photobucket.com/albums/ac51/rattletrap1970/Render-01_zps60b2af53.jpg (http://s885.photobucket.com/user/rattletrap1970/media/Render-01_zps60b2af53.jpg.html)

http://i885.photobucket.com/albums/ac51/rattletrap1970/Render-02_zps54a4f76b.jpg (http://s885.photobucket.com/user/rattletrap1970/media/Render-02_zps54a4f76b.jpg.html)

[btroj]

Sure looks a lot like a Cabin Tree........

[rattletrap1970]

Like I said, same basis, some refinements, but the overall premise has been around a loooong time. I aint sellin mine, I'm just doing it cause I like makin stuff.

[bangerjim]

Just get some roof lead and lino. Those give you a very good ZERO and SPAN for the tester.

That is how I checked/tweaked my Cabine. Was pretty dead on out of the box.

You do not need 10 samples to calibrate something. Mabe a midrange (12) for just verification.

banger

[rattletrap1970]

Assuming it's function is linear. I figured a high and low (but then using the known ones in between for verification), I mean, if it isn't accurate then what's the point.

[bhn22]

You'll want certified samples for testing. The hardness of alloys can, and likely will change, depending of how fast your sample cools. This is why water quenching works. Also, smaller samples cool faster than larger samples, so the hardness of a cast bullet will likely be slightly higher than that of a 1 lb ingot.

[rattletrap1970]

Yeah, I was thinking of looking up calibration standards.

[dikman]

Sure looks a lot like a Cabin Tree........
As does mine. Which is not surprising if you consider the operational requirements (a screw exerting pressure against a spring, which in turn allows pressure against a dial gauge). I doubt if there's too many ways of meeting these requirements and keeping things simple, construction-wise, at the same time.

BHN, I hadn't considered the cooling effect on my test samples. That may explain the variations I was getting. Sounds like I'll need to make a simple mold to cast multiple bullet-size samples for testing.

I haven't graphed my results, but looking at them it doesn't look like it's a linear response as I add more lead to the range scrap. Lots more testing required.

[bangerjim]

BHN is a relative moving target and is not easily calibrated to. In the olde days, we worried about the hardness a lot. With lead and it's alloys we use, it is constantly changing over time for an unknown period.

Today, after much research and trial/error, we know from many documents that softer lead is perfectly OK. hardness is not really a big factor anymore.

And............with the advent and success or powder coating, the hardness becomes even more of an unimportant aspect of casting. I have a Cabine and use it all the time, but I sure do not worry about it being NIST traceable to certified samples!!!!!!! CEGW....close enough for government work....is fine with me. HA.....ha!

With PC on all my boolits, the hardness can be anything from 10 to 15 and I do not care anymore.

I have a professional bhn tester in my lab, but it starts at 150 and goes up from there to over 600. It uses NIST traceable samples for cal checking, but those are NOT lead! They are various VERY stable alloys of steel and SST that do NOT change over time.

Good luck on your endeavor!

bangerjim

[rattletrap1970]

Since you have a lab tester would you be able to send me a few boolits from softest to hardest with your reading if what they are? Since the tester itself doesn't really change in design, the only thing that changes is the cross reference sheet based on my numbers vs. yours.

[dikman]

Rattletrap, I forgot to compliment you on the design, it's very neat and tidy. To make the body, though, will require either casting it or a lot of milling, I would think. I also toyed with the idea of a knurled knob, but decided against it. I've found that some samples require a bit of effort to turn the screw and the brass rod through the nut on the end actually works well (even if it does look a bit agricultural). Your sliding indicator is a nice idea, but may be a little fiddly to make and might be vulnerable to being knocked/bent. It might need to be made thicker (wider) than in the drawing.

[rattletrap1970]

If I were to mass produce, cast wouldn't be a bad idea, with all the coring underneath it. I have a bunch of Aluminum stock kicking around and I do Type II anodizing at home (in 18 or so colors) so I was just going to machine it. The indicator I think would be fine, although depending on the performace could be tirned into a single pin that slides through the housing. The knurled knob... Well... Depending on the diameter might be workable, or a hex could be added to the top to use a hex key, or again if could have a sliding cross-pin. The indicator I'm showing is an analog one (cause that's the nicest CAD model I found, but I like digital so that's most likely what it would be). I was also going make a provision for the indenting tip to be on a bearing so as you turned the load inward the tip isn't "drilling" into the test piece.

1. I added a "Hex" to the knob for turning in though materials.
2. Incorporated a thrust bearing and rotating tip so as not to drill through test material.
3. Indenting tip is now hardened A2
4. Included a Hex key.
5. Put a "Vee" notch in test area to hold bullets on center.

http://i885.photobucket.com/albums/ac51/rattletrap1970/Render-03_zps30fa5017.jpg (http://s885.photobucket.com/user/rattletrap1970/media/Render-03_zps30fa5017.jpg.html)

[Larry Gibson]

BHN is a relative moving target and is not easily calibrated to.

Very good statement and quite true. One does need a sufficient sample size (number of tests for each alloy). I usually do 3 test on a bullets/ingots of an alloy which gives a fairly good idea but if I really, really want to know I use 10 tests per alloy. The average gives a very, very good idea of the BHN then. Doing just one test on a single bullet/ingot might put you in the ballpark but that's about it.

I got a couple pounds of certified pure lead and used it for the "reference" offset for my Lee BHN tester, works for me. How precise you want to be depends on what you want to know.

Larry Gibson

[cbrick]

BHN is a relative moving target and is not easily calibrated to.

Very good statement and quite true. One does need a sufficient sample size (number of tests for each alloy). I usually do 3 test on a bullets/ingots of an alloy which gives a fairly good idea but if I really, really want to know I use 10 tests per alloy. The average gives a very, very good idea of the BHN then. Doing just one test on a single bullet/ingot might put you in the ballpark but that's about it. Larry Gibson

Exactly! Just as with a chrono test, one shot over the screens tells you from very little to nothing about either velocity or consistency. A 10 shot chrono test averaged will give you a reasonably accurate idea of the particular lot, same with BHN testing.

Rick

[rattletrap1970]

Took out the hex wrench at the suggestion of Dusty.. Now it uses the ratchet wrench from the Lube sizer (which are avaiable from Lyman for $8)
I figure the best sample for testing (on my end) would be actual bullets that someone cast from "Dead soft" up to the hardest thing they could make. Then bagged and on the bag written their hardness based on an average of ten tests per bullet. If I then did the same 10 tests and averaged MY results I think I'd have a pretty good tester. Like I said, this more of a design project than anything else.

http://i885.photobucket.com/albums/ac51/rattletrap1970/Render-04_zpsd958f7d2.jpg (http://s885.photobucket.com/user/rattletrap1970/media/Render-04_zpsd958f7d2.jpg.html)

[dikman]

I like the ratchet handle, but will the "stop-start" of the ratchet action affect the penetration of the point when screwing it in? I have no idea if it's critical to have a continuous pressure applied during this part of the process. As to the point, the idea of a bearing is nice. I already have something that might work, although it's a little oversize (it's off a wood-turning lathe). I might have to look into it, as my point is slightly off-centre. It works, but it annoys me. I drilled and tapped the end of the threaded rod to take a screw-in hardened point, but no matter how much I tweaked it I just couldn't get it exactly centered in the 4-jaw chuck. Also, the angle of the point will make a difference, as I feel that mine may need a bit more of an angle on it. I tried a blunter, rounded-off point, but wasn't happy with the readings, as it gave less spread to the figures.

This is all most interesting.

[rattletrap1970]

I was thinking of the stop-start issue. The only other way that I can figure is to have a sliding spring loaded indenter. So, you'd slide up to sample, then roll a lever over loading a spring. Kinda like an actual Rockwell hardness tester. Personally for best results I think just having the knob at the edge of a bench or table and giving it one nice complete turn might be easier.

[dikman]

Yep, I reckon that'll work. A great idea about the bearing mounted point, btw :2_high5:. I just modified mine to try it out and it works great. Mine is, of course, considerably oversize compared to your drawing, but that doesn't really matter. It's smoother to wind in, and I think your large knurled knob should be fine, as I could give it one turn just by gripping and turning the threaded rod. You'll need clearance to turn the knob, either by mounting the device on a block of some sort or, as you say, using it at the edge of the bench.

I might make a knurled knob for mine sometime (out of brass, maybe - I like turning brass).

[rattletrap1970]

McMaster Carr sells them. In different materials.



Yep, I reckon that'll work. A great idea about the bearing mounted point, btw :2_high5:. I just modified mine to try it out and it works great. Mine is, of course, considerably oversize compared to your drawing, but that doesn't really matter. It's smoother to wind in, and I think your large knurled knob should be fine, as I could give it one turn just by gripping and turning the threaded rod. You'll need clearance to turn the knob, either by mounting the device on a block of some sort or, as you say, using it at the edge of the bench.

I might make a knurled knob for mine sometime (out of brass, maybe - I like turning brass).

[dikman]

The postage to Oz would probably cost more than the knob!! Besides, I like making things......

[bangerjim]

Since you have a lab tester would you be able to send me a few boolits from softest to hardest with your reading if what they are? Since the tester itself doesn't really change in design, the only thing that changes is the cross reference sheet based on my numbers vs. yours.

I use a Cabine for lead. My other unit does NOT go down that soft. That is why I bought the Cabine!

banger

[dikman]

I found a knurled wheel in my lathe bits drawer, turned out to be a perfect fit on the threaded rod! And it turns real easy. I think I might replace the nut with a circular wheel with indent markings on it next.

Then one day I may even paint it! [smilie=l:90467

Just had a thought - a collar in front of the knurled wheel, with an extended locking screw. Set it so that when the point is just touching the sample under test the extended screw is touching the far side of the indicator rod. Then turn the wheel until the screw touches the other side of the rod. No need to bother with lining up the indent mark with the pointer, just do it by feel. It's not quite a complete turn, obviously, but that shouldn't matter as long as it's consistent from one test to the next.

Thoughts?

[rattletrap1970]

Thinking something like this. Threaded collar with a pointer on it. Turn knob in till it touches one side of a finger or pin that juts out from the body of the unit. When you turn the indentor load the collar makes one revolution and touches the finger again (on the other side). The finger on the unit and the finger on the collar can be machined in such a way to correspond with the travel of one turn (2 steps on both fingers) so they line up on the same plane.

[dikman]

I just tried the "rotating pointer" idea. I tapped a hole in the nut in front of the knurled wheel, and fitted a machine screw with a wingnut loc-tited at the head end (to use as a "handle"). I screwed the nut up to the mounting, then backed it off about 1 1/2 turns. Put sample in place, screw the point in 'til it just contacts the sample, rotate the nut and screw against the right of the pointer rod, tighten screw and then turn knurled wheel until the set screw contacts the pointer rod on the other side. Easy.

Using your idea, I should be able to replace the pointer rod with something with an off-set in it, so that as the nut rotates the set-screw slots into a recess and allows a full rotation of the main screw.

A little more fiddling, perhaps, than just looking at the markings on something, but a lot more consistent.

Next thing is a very light coil spring in front of the nut to hold it against the pointer rod when it's not tightened up.

[rattletrap1970]

I'd love to see a photo of that.

[dikman]

Ok, how's this? Even Though the main screw is a fairly coarse thread, one rotation isn't enough to allow for an offset, so I filed a groove in each side of the brass rod so that I could get as close as I could to full turn (between 350 - 355 degrees, I reckon). Doesn't really matter, as the main thing is repeatable consistency in each test. My idea of a coil spring in front, to hold the set-screw upright didn't work - yet).9066390664

[Grump]

Back to the calibration question...

How reliable are solder bars?

At the extreme, you could consider solid tin. I would shy away from pewter because it's not that "standard" an alloy.

You also seem ambitious enough to perhaps rig up some electroplating--with the correct solutes, you could get pure lead to calibrate the bottom end of the scale, but that is a lot of trouble compared to buying some.

I could send you some of my stuff with notes of where it falls with the SAECO tester...

[rattletrap1970]

That's what I mean, if someone has a tester the trust and could test samples of their own (like 5 hits or so on each sample and average them), then send me the tested samples with YOUR readings on the Baggie(s). Then I could test them and compare my dial numbers to your numbers. Once I had a bunch of samples, then I'd make a custom dial.