Everyone needs a hardness tester.

[Toolmaker51]

The missing link to the PDF in the main post has now been added. Here it is:
https://diqn32j8nouaz.cloudfront.net...ess_tester.pdf

Searched (Ctrl+F) your PDF for which version Arduino Nano chosen best suited for this tester. There are many choices, bewildering to newbs......such as 67AH8870. Remaining bill of materials very simple, resistors etc are clearly self identified. The video, and schematics make this a slam-dunk to re-create, not intimidating at all.
Alike with at least 2 previous replies, I have minimal electronics background design-wise, more confident in repairs.
There is a vast amount of resources now that just didn't exist, and this is an outstanding entry point. Here it's 5°F out. Sitting in the basement creating is superior.
I'll advise web-surfers, our US distributors will not be the most economical source to fill the BOM. Right now looking for an odd "D" battery holder; $4.60 one place, $28 another + 100 piece minimum. Caveat emptor. YMMV. IYKWIM. DAMHIKT.

[tonyfoale]

Searched (Ctrl+F) your PDF for which version Arduino Nano chosen best suited for this tester. There are many choices, bewildering to newbs......such as 67AH8870. Remaining bill of materials very simple, resistors etc are clearly self identified. The video, and schematics make this a slam-dunk to re-create, not intimidating at all.

Almost any Arduino will do. I use the Nano because it is so small and easy to fit in boxes and on to circuit boards. As shown in the video I use an Arduino Uno (which as the name suggests was the first) for development or proof of concept work, its larger size is better for that. Otherwise the Nano is the same.
I think that all the Arduinos fit into one of two classes.
1. Genuine
2. Clones.

The clones are legit because the Arduino hardware is open source, however there is a difference. The genuine ones use a comms chip for which they have to pay royalties. To keep the price down the clones use a royalty free chip which needs a software driver. This is a free download from many sources, look for the English version (unless you understand chinese) of the CH340 driver. This driver is to allow the inter-connection of PC and Arduino via a USB cable, which is simpler than an alternative method of serial communication.
On the net there are many complaints by people who cannot get a clone to work. All or the vast majority that I have seen are due to the complainer not following instructions to get the driver. It is just like loading any other hardware driver into a PC.
This driver is only needed for communications between a PC and Arduino for either programming the Arduino or for applications where you run the Arduino in conjunction with a PC.

As for which Nano to use, I would say ANY. The cheapest require you to solder on the supplied header pins so if you want to avoid that, look for those that come pre-fitted. You may find it useful to look up adafruit.com, they are a company which specialise in electronic hobby stuff and have their own range of many variants of the Arduino and they have loads of free download help and tutorial files. I do not recall ever buying from them but they have built a good reputation. They are in NYC.
https://www.arduino.cc is the official Arduino site and is full of help files, programme examples and forums etc.

Just to make the process clear.
An Arduino can operate autonomously or it can run connected to other stuff like a PC or even other Arduinos.
The programming is usually done on a personal computer and then downloaded to the Arduino via a USB cable.
You need to install the Arduino IDE (free programming software from the Arduino web site) then you write your programme (for some reason they call a programme a sketch, I don't) or in this case you copy and paste the programme that I listed into the IDE and then download to the Arduino.
If it is a stand alone application you can then disconnect from the PC and feed the Arduino from a battery or wall wart etc.
Some applications like data acquisition will need to keep the PC connection to download data. The connection in some cases might be two way, you might use PC software to send commands to the Arduino.

[DIYer]

Thanks tonyfoale! We've added your Hardness Tester to our Measuring and Marking category,
as well as to your builder page: tonyfoale's Homemade Tools. Your receipt:

Hardness Tester by tonyfoale

tags: tester, measurement

[jdurand]

An arduino also can be programmed using traditional C programming instead of "Wire" (the language Arduinos use). The development software is free and the debug pod to program, single step, examine memory is under $15. You just have to find Nano clones with the debug header, some have it, some don't. Only a couple of $ each for them.

[Claudio HG]

Wow interesting and well done project. Maybe you arelady said this somewhere and I missed it, but why impact bounce relates to hardness?

[tonyfoale]

Wow interesting and well done project. Maybe you arelady said this somewhere and I missed it, but why impact bounce relates to hardness?

Drop a tennis ball on concrete, then go and drop the same ball on some soft ground.
Softer material deflects more and more energy is lost when impacted.

[CanBeDone]

Drop a tennis ball on concrete, then go and drop the same ball on some soft ground.
Softer material deflects more and more energy is lost when impacted.

Tony Foale is only partly correct with his claim to measure hardness with his (ingenious) device. What it does measure is the energy loss an elastic ball experiences when rebounding from a surface, since

KINETIC ENERGY = mass times VELOCITY squared

and velocity is what Tony's device measures. Genuine hardness testers measure, however, only the energy consumed by moving material sideways and upwards, the energy consumed moving the test piece downwards is ignored, because reading takes place only AFTER all dynamic effects of doing the measurement have died away. This has consequences for what parts can be hardness tested.

Firstly, the test piece needs to be thick enough that elastic or inelastic bending does not consume appreciable energy. If Tony were to measure the hardness of a THIN knife, with supports way away from the point of impact, he would obtain a different "hardness' than what he would if the knife blade were supported by a sandbag, or, worse, by a heavy and smooth steel anvil.
Secondly, the tested material must be uniform, i.e. the orientation under which the piece is tested does not make any difference to the outcome of the test. This demand arises from doing hardness conversions, such as to Brinell or Rockwell. Thus, materials with directional properties, such as fibre re-enforced materials, but also cold deformed steel, will produce ridiculous hardness results.
Thirdly, if you are old enough to have been exposed to Silly Putty, you'll know that its response to deformation (what we commonly call hardness), seriously depends upon the rate of deformation, to such a degree that it may be claimed with perfect validity it has hardness zero, by just watching a steel ball slowly being swallowed by no more force than its own weight. But, when doing the deformation fast enough, e.g. by dropping a silly putty ball onto concrete, its rebound is higher than what you can achieve with a hardened steel ball, which has a hardness of more than 50 Rockwell C.

Thus, be careful when you do measurements with Tony's device. It may be "hardness" what you get, but it may also be a load of bull. It all depends on your own setup and understanding of what you are doing.

Lastly, a word on data conversions, since Tony's device relies on them. There is a reason behind reporting hardness as a number such as 50 Rockwell C, and not as, say 1000 Newton per square millimetre (the stress to produce a 1mm² indentation, or 1000 Newton millimetre per cubic millimetre (the energy to produce a 1mm³ depression). Rockwell C specifies the load that your testing machine applies to be 100 kilogram, i.e. a mass of 100kg is lowered onto your sample, making an impression, via its diamond cone front end, a fraction of a millimetre in diameter. This diameter is then converted into a number, using a standard table. This table assumes, for example, that the impression is a perfect circle. But visual inspection with the measuring microscope shows that it is not . . .
Why? Take a Knoop hardness tester, which leaves impressions only a few thousands of a millimeter wide, and repeat your hardness measurement on 100 positions on the same test piece. Result? 100 different results, with a standard deviation of easily 20 percent, and I have observed 50% standard deviation.

Again, be careful with what you conclude from using Tony's device. If you do Quality Assurance, and your customer has told you to use it, then do use it. But if you do process development, and you need to optimize both wear resistance ("my knife gets blunt so quickly") and fracture resistance ("the cutting edge chips"), I at least would stay far away from the gadget, as in that situation it is just a waste of time.

As I have said before, it is a brilliant design, but it also totally ignores the reality of the physical metallurgy of a real piece of metal, its composition and its internal structure from grain size and orientation, inclusion content, via dislocation density and orientation, down to chemical bonding. They all matter, as only when interacting together they produce hardness.
Take your pick. I won't build it. For my purposes, and with my background both in process design and quality assurance, a file is just as trustworthy a hardness measuring tool, and I don't need to build it.

And if you truly need a portable hardness tester, ebay sells a portable Brinell one for USD1580.00. And of all the hardness testers I have used, Brinell is the most repeatable one, because it samples a larger area than any other tester I have used. Vickers comes second, but that is clearly a research tool, not for Production use.

[Toolmaker51]

True; but limited. At the commercial level certified instruments, de rigueur, in a private setting excellent.
Grassroot demystification of a scientific principal.
A collection of known coupons [ie Rockwell discs] would easily bracket a sample, even somewhat out of the lower range of a Leeb device. Immediate velocity at the part seems quite reliable compared to discerning bounce height, even though they are related. That a sample must have full support is a given. Demonstration of bearing ball on HSS bounced as one would expect. The need to interpret softer yet metallics virtually non existent, but camera and standard rebound instrument cover most remaining conceivable work. Safe wager says not 1 in 1000 private environments have test apparatus, while a good number are heat treating simple work often; oil and water varieties, probably casehardening most of all. I'll lay down funds now to build a portable tester, over a bench bound item requiring destruction of large part, 10:1.
And it wont be a chinesium imposter.
I realize this could be taken as pugnacious, but experience sourcing and building a $10 kit soundly beats 1.6k for incidental use.

[tonyfoale]

The name of this forum is HOME MADE TOOLS.

[clavius]

Tony Foale is only partly correct with his claim to measure hardness with his (ingenious) device. What it does measure is the energy loss an elastic ball experiences when rebounding from a surface, since

KINETIC ENERGY = mass times VELOCITY squared

and velocity is what Tony's device measures.

<Other stuff snipped>

Wow, that is quite a reply. Holy cow.

So much to look at here. For starters, this device does not measure velocity as you assert, it measures the ratio of the initial and rebound velocities. That is a very different thing. That measure is a accepted by ASTM and similar standards bodies as a valid hardness test. Like any test, it has to be done properly and within the intended use or it will as you say report a "load of bull."

A Rockwell tester will not give useful measurements if used on "materials with directional properties, such as fibre re-enforced materials" as will a Vickers, Brinell or other similar tools, That is not what they were intended to do, and neither is this.

If you "measure the hardness of a THIN knife, with supports way away from the point of impact" or that "were supported by a sandbag" with a Rockwell tester you will not get valid results either. That does not make a Rockwell tester suspect, it means you are using it improperly.

As for the "speed of impact" example, Silly putty and steel are entirely different things that behave entirely differently. A steel ball placed on a steel sample will never be "swallowed by no more force than its own weight" no matter how long you wait. Silly putty is designed to behave exactly as you describe, and does, but that has zero bearing on what is being discussed here. The use of the ratio of velocities is specifically to take the absolute speed of impact out of the equation.

The design in no way "totally ignores the reality of the physical metallurgy of a real piece of metal, its composition and its internal structure from grain size and orientation, inclusion content, via dislocation density and orientation, down to chemical bonding." That test was defined near 50 years ago by a Swiss company that develops tools for materials testing. Google it. Tony did not define the test, he rather created a homebrew version of a tool to perform the test and it's a rather clever one at that.

"Again, be careful with what you conclude from using Tony's device. If you do Quality Assurance, and your customer has told you to use it, then do use it. But if you do process development, and you need to optimize both wear resistance ("my knife gets blunt so quickly") and fracture resistance ("the cutting edge chips"), I at least would stay far away from the gadget, as in that situation it is just a waste of time."

Did anyone suggest that this was an acceptable substitute for qualified testing in a commercial setting or for process development? I didn't see that. As for wear or fracture resistance, no hardness tester tells you that. Neither will a file. One needs to be careful what one concludes from ANY test or measurement. If one misapplies any test in the ways you describe for this one, the results will be suspect at the very best.

To recommend a $1580.00 tester from ebay as an alternative for a bunch of home shop folks on a homemade tools forum (It's right in the name) is a bit over the top, really. Now that I think of it, if one is doing process development or quality assurance work for customers, I would probably steer them away from buying their development and QA tools from ebay.