I have tryed the 3 plates method described by J. Whitworth to make flat surfaces from scratch, i.e., without the help of any other flat surface to use as a reference. Awesome and supertedious. Since there is no much online that give information on how to actually carry out this process, I have shared my experiment in this video, with the hope it will be interesting and helpful to others.
This is the link to the video:
Thanks for sharing! I can surely attest that the process is tedious!
Though you're working with straightedges, I have a few tips for people making their own surface plates in
how to make your own spherical comparator/ spherometer at:
Workshop Metrology - Meccano Gallery
but also how I made my own "Cheap-o-meter" at:
Repeat Reading Gage - DIYSwede's "Cheap-O-Meter"
-Now, I might be barking up the wrong tree here,
but given the stains near the edges of your laminated chipboard (?) working table,
this seems to be out of flat itself, which just might give you false readings.
When getting down into tenths of thousands of an inch (or microns for us metricated patients),
I've learned (thru bitter experience) that the Devil also certainly lies in these seemingly insignificant details.
AFAIK: Straightedges could/ should be supported at the "0.5536 times the length" for minimum sag
(i.e. 0.2232 times the length from each edge):
Airy Points, Bessel Points, Minimum Gravity Sag, and Vibration Nodal Points of Uniform Beams | Mechanics and Machines
Just my 2 cents
Motto: "-This might seem simple, but just wait 'til I've explained"
Hey Johan, glad to have found an other crazy man who endeavored into the three plates method to gain its own flatness
I've been humble and what I aimed was something in the order of 10-30 microns, so to have a final precision within/less than .01 mm.
About the laminated chipboard, the stains are there for other reasons (spilled dye), but besides that board is fairly flat in its kind, you are definitely right about the fact that it is not flat by itself as every piece of wood is. So, was it a source of troubles in providing a good support of the pieces? Well, yes it could have been, because the deflexion caused by the weight of the part itself (plus the second part put on top when rubbing against the dye) in the worst case would be 22 microns.
Ouch, it didn't went in my mind to check this thing.
However, in a way or an other, that must have not influenced much the work as I checked the final result against my surface plate (which is AA grade) and it seemed ok, but as it is smaller than the part to be measured I did the job in three steps (see picture).
Then I had the opportunity to checked it against a larger surface plate, confirming it went nicely flat.
I was lucky.
So thank you for having pointed it out, I will add a pinned comment to my video mentioning this point and crediting you, if you don't mind.
Also I liked your instruments.
Last edited by Claudio HG; 12-02-2019 at 05:09 AM.
Thanks, Claudio! I'm sure crazy enough to go down that route - only to find when I was "Semi-finished"
a discarded surface plate in a dumpster behind a tech lab...
Found out it was hard to check with my spherical comparator/ spherometer - I made a "cheap-o meter" myself.
Figured it was within the AA specs (albeit not thick enuff) and I was more than satisfied,
and have thus ditched the three rectangular ones.
Snag is in grinding plates, the top one will become concave and the bottom one convex,
so you have to flip them often when your getting near completion (as if that would ever happen...).
Then - square plates are better than rectangular, as you can twist the top one to 90, 180, 270 deg.
Rectanguar plates will worsen the bow as the weight of the top plate's overhang
will exaggerate the grinding at the edges of the lower one.
When getting down to such small differences as a couple of mu-meter everything counts:
Temperature gradient of the room, support of the surface plate, even lighting matters-
making a flat plate somewhat convex from the sun/ worklight.
not to mention your hand heating the straightedge top.
I saw a video somewhere, where a guy just breathed on a biggie cast iron camelback supported at its ends,
measuring with an indicator at the middle...
The fun never ends:
-Guess it was Secretary of Defence Donald Rumsfeld* who mentioned the "unknown unknowns"?
It isn't what you think, but just isn't so, that's gonna bite you in the a** -
it's the things you didn't even concieve you were ignorant about that'll hit you hard...
i.e. Lack of meta-cognition, a k a "cock-sure confidence not backed up by neither knowledge nor insight...
I suffer from this too, but try my best to improve.
Go ahead and reference me - I'm not picky in being credited.
*neither being a fan, nor a critic of him - guess it just was his lucky day mentioning that...
The unknown unknowns are subtle back stabbing that can be mitigated only by being humble in confidence with what we think to know.
Who's the philosopher who said "In know that I know nothing"? ...ah, yes, Plato.
Well, we have some limited knowledge and experience, so for example my long experience with electronics carved in me the attitude to be wary with what I see on the instruments in relation with what likely the reality could be instead. Nevertheless it is impossible not to forget in one occasion or an other some "unknowns", even when they are known.
That said, this forum is cool as I found interesting people like you, or that anyway like to share their experiences/experiments and knowledge, without being fussy.
On Reddit I've posted the same post on the "metalworking" sub, but it seems that there, there are hypersensitive people that don't grasp the concept of rethorical question as some have criticized the fact I've put a quesion in the title. Gosh.
I will look better at those instruments you made, now I am quite in hurry.
Post your reply!
Join 42,532 of us and get 200+ tool plans, tool eBooks, build guides, and much more.
There are currently 5 users browsing this thread. (0 members and 5 guests)