PDA

View Full Version : Perpendicular Scraping- Another Attempt

Teenage_Machinist
12-27-2008, 09:15 PM
I know how things would be scraped flat. I know how they are scraped paralel. However, nobody has really, plainly explained how to scrape perpendicular without a reference, despite several threads (if somebody has despite searches, please link, and I will apologize:o ). One method is to make cylindrical master squares out of whatever you have, but they are not too useful for marking the bearings. Another would be to indicate a parallel piece clamped to the cylinder squares, which would show height, but not bearing. Without a way to hold them to the plate that method might not work very well. Can somebody explain how they do it---p- I am not looking to get things to 0.0001, but would like to split the half-thousandths.

dp
12-27-2008, 09:35 PM
You can use a granite angle plate on top of your granite surface plate. They're not cheap. You can also use an iron angle plate that you know is true. In Forrest's scraping class we used a granite angle plate.

lazlo
12-27-2008, 10:11 PM
To add to DP's suggestion: take a right-angle granite indicator block, blue one side, sit it on your surface plate. Then slide the piece you're trying to get perpendicular down the right angle piece 'till it contacts the surface plate. That will spot the end.

Or sit the workpiece on it's end, slide it up to the angle block, and shimmy it around, that will spot the long end.

You can also use a DTI: measure the four corners of the piece, decide which two are "high", mentally interopolate between the four corners, and scrape in sections until the top surface is flat and planar to the surface plate you're measuring off of.

lazlo
12-27-2008, 10:21 PM
By the way, here's a thread I posted awhile ago with Nick Mueller's Youtube scraping videos.

The "Step Cut" process of correcting geometry like Nick Mueller is doing, by scraping one or more plateaus on the high side, is described in "The Art of Hand Scraping" -- a booklet prepared in 1981 for Lawrence Livermore Laboratory. Step Cutting (to achieve perpendicular surfaces) is described on page 23 of the second .PDF:

http://www.practicalmachinist.com/vb/showpost.php?p=681916&postcount=38

oldtiffie
12-27-2008, 11:51 PM
I know how things would be scraped flat. I know how they are scraped paralel. However, nobody has really, plainly explained how to scrape perpendicular without a reference, despite several threads (if somebody has despite searches, please link, and I will apologize:o ). One method is to make cylindrical master squares out of whatever you have, but they are not too useful for marking the bearings. Another would be to indicate a parallel piece clamped to the cylinder squares, which would show height, but not bearing. Without a way to hold them to the plate that method might not work very well. Can somebody explain how they do it---p- I am not looking to get things to 0.0001, but would like to split the half-thousandths.

OK.

First of all - and I don't see that too many have acknowledged it yet - your range of accuracy limit/s is 0.0005" - and not to 0.0001" - as you say.

If you have a pair of cylindrical master squares (relatively easy to make) - preferably 3 or 4 to test against each other - and a sizable, preferably old and "seasoned" angle plates (two minimum to test against each other) and a vertical mill that has a good flat table, good quill and has been accurately "trammed in", you are set to go.

Clamp the master cylinder guages to the table and then clamp the angle plate to the squares and fly-cut all faces (common sense will dictate the order).

Test angle plates (on the mill table) against each other and the master squares - all combinations. If all is OK to the limits you require you can use the faces of the angle plates as your references.

Might seem "as rough as guts" - which it isn't - and "not according to numerous books by sainted authors" - which it is - but it works.

I did it some time ago - posted here - and got all the "faithful" all upset - but it neither mattered nor made any difference to a very satisfactory result.

This is a good machining project and should not cost much at all.

A machine level is very handy to have as well as it has an accuracy of 0.02mm/metre = 0.02/1,000 = 1/50,000 = 12/50,000 = 0.0002" per foot.
https://www.machineryhouse.com.au/Products?stockCode=Q208

http://www.measumax.com/ViewPDF.php?ID=35-208

I have one and it is my reference square when I don't use my master squares.

J Tiers
12-28-2008, 01:02 AM
With NO reference you are out of luck.....

You can make one.

Three angle plates can be scraped in rotation to zero in on perfect perpendicularity. I suppose one of them could actually be your target surface..... which would leave you with two references and a finished surface.

I'd have to look up the best sequence, but you can see the method, since if two things equal each other, and both equal a third item, then all three must be equal.

if A,B, and C are "right angle irons", and A leans back just a bit, while B leans forward just enough to fit flat against A, you won't know it. But if C fits against A, then it cannot also fit against B, so you detect the error.

Thus if all three fit flat against each other in any combination when on a flat surface, they must all have "perfect" 90 degree angles.

Fasttrack
12-28-2008, 01:54 AM
With NO reference you are out of luck.....

You can make one.

Three angle plates can be scraped in rotation to zero in on perfect perpendicularity. I suppose one of them could actually be your target surface..... which would leave you with two references and a finished surface.

I'd have to look up the best sequence, but you can see the method, since if two things equal each other, and both equal a third item, then all three must be equal.

if A,B, and C are "right angle irons", and A leans back just a bit, while B leans forward just enough to fit flat against A, you won't know it. But if C fits against A, then it cannot also fit against B, so you detect the error.

Thus if all three fit flat against each other in any combination when on a flat surface, they must all have "perfect" 90 degree angles.

Thats a slick trick. I think that was mentioned in "Machine Tool Reconditioning", iirc. I think its the route I would go in this situation. Like you say, you get whatever you wanted scraped perpindicular in the first place, plus two good refrences (or one could be used as an angle plate in a mill)

Teenage_Machinist
12-28-2008, 02:08 AM
God bless the giver of these links...

Though I do not quite understand the one with the DTI. It seems like the method for getting things paralell, not perpendicular.

12-28-2008, 02:29 AM
A DTI mounted on surface gage, in various configurations, on a surface plate (or reference surface), can be used to check both perpendicularity and parallelism.

However, it can quantify these things with only (comparatively) "rough" accuracy. For reasonably accurate quantification you would need a plunger "DI" set carefully perpindicular, and for serious accuracy (re More), you would need other techniques like referencing slip blocks or feeler gages, and so on...

12-28-2008, 02:47 AM
Squareness is a surface gage and a dial test indicator away and somewhere in the old posts is a verbal description how to set it up and make the squareness test. Squareness is one geometrical test you can check by reversal error. With it your can determine squareness to any degree you wish dependeing on the sensitivity of the DTI and the care you take with the readings. Squarness determinations of 0.0001" is possible with a 0.001" graduated DTI and 0.0005" with a 0.0001" reading DTI. The error is independent of the span within reasonable limits of apparatus rigidity and environmental and teperature considerations.

You can set the surface gage and DTI against a square of known error or prove a square edge on something parallel sided like a 1-2-3 block. Tomorrow I'll try to find my post. It shows a way to quantify squareness error.

dp
12-28-2008, 02:57 AM
It can certainly be done with simple tools but it's like trisecting an angle with a straight edge and compass. It can be done but there's more practical ways to do it :)

I can fly an airplane with the 5 basic instruments but I far prefer to just look out the window. -- Dennis Peterson, 1974

oldtiffie
12-28-2008, 04:43 AM
You will get some pretty impressive results using a bit of thick float glass plate from the scrap bin of your friendly local Glazing contractor. I have several and have never needed a surface plate for any of the work I do in my shop.

If I am really "fussy" I use the top faces of my magnetic chucks from the grinders - making sure there is no or very little residual magnetism.

If anyone is really fussy about a surface plate or similar, they should insist on seeing a NIST certificate that is "in date" and has not expired, and tabulated errors as well as classification, and that the surface plate was suspended/supported as per the NIST standards.

From my perspective for what I need to do what I do in my shop a surface plate is only an up-market easily damaged marking-out table/plate.

I'd have trouble remembering when I last needed a certified surface plate. I have quite a selection of good marking-out surfaces that do pretty well as a reference surface though. My master squares and my "true-ed up" angle plates do very well.

If 2-3-4 or 4-5-6 blocks and my Tool-makers vises are as good as said in the adverts as regards flatness, parallel and square they would be very hard to beat as regards all of those criteria.

Here are my "Tool-makers Vices:
http://i200.photobucket.com/albums/aa294/oldtiffie/Vise/VIse3.jpg

and here is the "inspection" sheet that came with one of them. Others may "rubbish" the vises and the sheets as "pi\$\$-poor China stuff" but I have no reason to doubt them and some very good reasons to think they are correct!!
http://i200.photobucket.com/albums/aa294/oldtiffie/Vise/Vise1.jpg

The allowable tolerance of 0.005/100 = 5/100,000 = 0.5/10,000 which is 0.5 "tenths" per inch which is arc sine 0.0029 arc degree = 0.1719 arc minute = 10.3 arc seconds.

So, if you stood those vices up on either their end or sides you would have a pretty good vertical reference plane. And a damn side more useful than a granite square.

A look at the web sites of CDCOMachinery.com and LittleMachineShop.com (both in the US - and yes the stuff is mainly Chinese) you will see a whole lot of stuff such as angle plates etc. with similar accuracy.

The machine square, the details of which I posted earlier, is a very good option as well.

Now I have nothing against having or making surface plates or squares or scraping per se', but it needs to be said that they are not or may not necessarily be the best or only option/s.

12-28-2008, 08:14 AM
Part of being a machinist is knowing your tools and whether they are accurate enough for a particular set of circumstances.

Those of you having no need for calibrated squares or making verifiable squareness determinations should read no further nor need they scoff at the following text's applicability to their particular line of work. It's intended to illustrate calibration technique for those having an interest or a need for the topic.

Checking square against square only ensures that their stocks are parallel when their beams are butted together. Scribing a perpendicular, then reversing a square to eyeball the error is adequate for a framing square but laughably crude when applied to even rudimentary squareness determinations used in the machine shop. OTH, you need three squares to check in rotation. While this method can be used to verify absolute squareness it lacks the ability to quantify the squareness of any particular item needed to be square.

Quantify means ",,,express as a number or measure or quantity..." My definition of quantify as applied to squareness error means a measure of the error and its direction (ie acute or obtuse stated in terms of the interior or exterior angle.

A square is self checking, that is no standard is required to assess its accuracy: it can be used to check itself.

Quantitatively checking a square to very close tolerances is bone simple with a minimum of equipment. All you need is a granite flat, a surface gage, a pair of good 1-2-3 blocks, and a dial test indicator.

Clean and de-bur the square to be tested. Verify the parallelism and straightness of the stock and beam independently. Verify parallelism and equal size of the 1-2-3 blocks in the 2" dimension.

Reverse the surface gage mast so its ball end extends down through the notch of the surface gage base. Use the surface plate as a flat reference:

Mount the dial test indicator on the upper part of the mast.

Set the pair of 1-2-3 blocks on the 2" edge or on end so the stock of the square can pass between them

Set the surface gage base on the 1-2-3 blocks. Ensure the stock of the square can pass under the base and between the 1-2-3 blocks.

Position the square so the ball of the mast contacts the beam close to the stock.

Adjust the DTI so its ball contacts the beam towards the upper end. Note: work carefully and gently minimizing heat input to the square and the surface gage. The DTI contact has to be on a perpendicular over the tangency of the ball end on the square. This is necessarily a finicky adjustment. The DTI cannot be properly nulled unless this adjustment is within close limits.

Slide the square back and gently re-contact the ball of the mast rotating the square slightly in the vertical plane to null the reading. Zero the indicator. Re-test a few times to ensure consistency.

Slide the square out and reverse it 180 degrees on the surface plate. Contact the ball and the indicator again. Note the reading.

Reverse the square and run the stock back under the surface gage base. Note the repeat zero reading. If the reading reading is not zero adjust and tweak until it is.

The indicator reading is DOUBLE the actual square error. Note the algebraic sign of the error. Calculate the raw readings to convert them to a slope. Apply a calibration sticker to the square or note in the calibration records (you meticulously keep). Place a certificate od calibration in the square's box. I suggest you use wording like "Error 0.000X per ft - interior angle ACUTE (or OBTUSE)." This provides enduring notice of the amount of square's error and its direction. Anyone using it can use the error information to compensate its known error. Thus it's possible to make squareness determinations that are more accurate than the square itself.

I discussed this procedure in some detail in my Home Shop Machinist article in Jan/Feb 2005 and there are some pictures to see. I didn't use a surface gage for the article. Instead I used a gadget I built. A surface gage is quite tricky to adjust for this evolution and, once set, its adjustments can easily slip.

The same procedure may be used to check other items. Angle plates may be checked by using adapting this technique but a precision parallel must be added to the equipment.

Wish I was smart enough to post pictures.

lazlo
12-28-2008, 09:47 AM
Wish I was smart enough to post pictures.

Forrest, email me the pictures and I'll post them.

J Tiers
12-28-2008, 09:53 AM
I see that you are actually measuring the distance off square by a sort of "self transfer standard".....

You must know the front and back surfaces of the beam are parallel. Therefore ONE surface cannot be checked that way without slightly different equipment.*

And you might need added equipment, for the "independent verification" of stock and beam. Although, you can use the DTI to do that, checking against the flat.

I think, though, that you cannot quite verify totally, since ultimately you can only have a "go/no-go" verification that the stock and beam are straight.... they either blue up on the plate or they don't. That is reasonably equivalent to not being able to do more than a "go/no-go" verification that two squares butt up parallel. of course in either case you can scrape them flat against the granite flat, and verify parallelism with the DTI.

* Two equal sized balls on a beam, with a dti and some means of hinging the beam could verify a single square surface, by contacting them on one side, moving to the back side and contacting them again, checking the movement of the balls with the DTI.

lazlo
12-28-2008, 09:55 AM
You can set the surface gage and DTI against a square of known error or prove a square edge on something parallel sided like a 1-2-3 block. Tomorrow I'll try to find my post. It shows a way to quantify squareness error.

Forrest, the way I've done that, which I think you're describing, is that you place the base of the square against the edge of a know flat edge, and then scribe the length of the beam. Then flip the square around, and scribe the beam again. The divergence angle between the two lines is twice the square's error.

lazlo
12-28-2008, 10:02 AM
* Two equal sized balls on a beam, with a dti and some means of hinging the beam could verify a single square surface, by contacting them on one side, moving to the back side and contacting them again, checking the movement of the balls with the DTI.

One textbook method of checking a square, as shown in Audels Handybook, for example, is to place 4 equally sized discs in a grid between the beam and the base, and measuring the length the diagonals.

Teenage_Machinist
12-28-2008, 12:47 PM
Toolmaker's Buttons?

Forest, I am glad for your post, but I have a hard time understanding it. It sounds like you elevate a DTI holder off the plate and then put the square on the plate, so it touches the DTI at the top and the bottom of the gage at the bottom, and then flipping it.

Sounds a bit like a go-nogo setup I saw that I had doubts about, but quantitative.

12-28-2008, 01:05 PM
If the angle of the contact of your DTI to the work is right (about 20 degrees for most DTI's) the dial reading may be taken with some confidence. Thus the dial reading gives you quantitative data of the squareness error. In a few hours I will send Lazlo photos he will post. A picture is worth a thousands woedsa of tecknical prolxitiy.

lazlo
12-28-2008, 04:15 PM
If the angle of the contact of your DTI to the work is right (about 20 degrees for most DTI's) the dial reading may be taken with some confidence. Thus the dial reading gives you quantitative data of the squareness error. In a few hours I will send Lazlo photos he will post.

Forrest's pictures and explanation deserve their own thread, so I've posted them here:

J Tiers
12-28-2008, 05:16 PM
One textbook method of checking a square, as shown in Audels Handybook, for example, is to place 4 equally sized discs in a grid between the beam and the base, and measuring the length the diagonals.

That is yet another way, although it sounds a bit hard to actually DO without moving things out of position as you measure.

The two balls scheme is really just a way of getting the DTI to read from both sides..... You read the difference in position of the balls when contacted from "front" vs "back". The balls could as well be disks.

There shouldn't be any difference.

darryl
12-28-2008, 06:40 PM
Somebody mentioned the granite angle plate- I've been considering making one from some slabs of countertop material that I scrounged. These are about 3/4 inch thick and seem to be very flat on at least one side. I have verified this on my surface plate. What remains is to epoxy a few pieces together, using a cylindrical square clamped to the surface plate to keep one slab piece perfectly perpendicular to the plate, while another piece lays flat against the plate. A couple of webs would be epoxied in at the same time. I would use a good grade of epoxy- my choice would be PC7, or you could use JB weld.

I just may go ahead and do this soon. It won't look pretty since my pieces are not perfect, but the two important surfaces would be at 90 degrees to each other, and that's what counts. The cylindrical square I can make on the lathe. Instead of relying on the lathe to make this completely without a taper, I'll carefully measure the diameter of two 'contact rings' that I turn onto the cylinder. The accuracy of this square is going to depend mostly on how consistently I can read the two diameters and make the truing cuts. All pertinent things will be taken into account, such as distortion due to clamping forces, heat buildup during machining, stress relief of the material, etc.

The angle plate could easily be made from float glass, in which case it's much easier to cut the pieces to the desired size and shape beforehand. If I go with the granite pieces I have, I either have them cut by someone to my specs, or use them as is which will work fine but won't look that good. Another option is to use suitably sized tiles which can be found in building supply stores. Marble or granite- as they say, these materials have been aging for millions of years, and should be free from warping, though the flatness of the surface is dependent on how well machined they are.

No scraping or finishing to perfect squareness required- just care in making a cylindrical square to start with, and care to equalize temperatures of the pieces and positioning them on the surface plate while epoxy sets.

Anybody is welcome to shoot this idea down, or add to it. I think I'll take a quick trip the store and see what materials there might be to work with.

oldtiffie
12-28-2008, 06:53 PM
Whoa there!!

I have no problems with the discussion between Forrest Addy, lazlo and JT.

It is all "good stuff" and "grist for the mill".

It is also excellence for the sake of excellence unless it has a practical use and purpose for the reader in general and for the machinist in his day-to-day needs in his shop.

It may well be regarded by some - me included - as "science/knowledge for the sake of science/knowledge" and perhaps "nice to know about" and perhaps "nice to be able to do if I need it". The same can be said of mathematics.

These items - scraping, squaring, science and mathematics - can essentially be reduced to two categories - depending on the needs and circumstances in individual cases - "pure" and "applied".

"Pure" is essentially theoretical or under very high levels of control and/or precision. Some may see this as a combination of "Class-room" and "Laboratory" (and quite esoteric - "up in the clouds") type "stuff".

"Applied" is essentially whether and how it is relevant to and usable in the circumstance/s in which the user/reader finds himself and in which he has to equate or reconcile his needs, his ability and capacity of his shop, machines etc. with/against the theory and scientific results of the items in/under "Pure".

If someone wants to "test the/his limits" in what-ever environment or circumstances he finds himself in or elects to use, I have no problem with that at all either.

But whether it is practical or is or can be or needs to be applied in the every-day circumstance/s in the HSM shop or else-where is another matter entirely.

It is entirely up to the individual to assess his real needs and the capabilities of himself, his machines and tools and his measurement level and repeatability that count.

The OP made a very good job of stating his required limits and performance objectives within the context of his use for the finished job. For those that have forgotten or ignored the OP it is:

I know how things would be scraped flat. I know how they are scraped paralel. However, nobody has really, plainly explained how to scrape perpendicular without a reference, despite several threads (if somebody has despite searches, please link, and I will apologize:o ). One method is to make cylindrical master squares out of whatever you have, but they are not too useful for marking the bearings. Another would be to indicate a parallel piece clamped to the cylinder squares, which would show height, but not bearing. Without a way to hold them to the plate that method might not work very well. Can somebody explain how they do it---p- I am not looking to get things to 0.0001, but would like to split the half-thousandths.

12-28-2008, 06:58 PM
Can't shoot it down, Darryl; it's pretty a good plan. Since you're going to do all this work, I do suggest you cut the pieces and make them to a good net shape with a diamond chop saw or a larger sized tile cutter. I also suggest bead blasting the ares to be epoxyed to ensure cleanliness and to provide tooth for a bond.

I assume you have a webbed angle plate in mind as a model for final configuration. Contact the stone countertop people for what epoxy to use. There's lots of different flavors of epoxy each with its own application.

I would have two concerns. one is stability of the epoxy joint. The other is the (alleged) flatness of the countertop granite. You can't tell by eye aloine flatness by reflection even from mirror finishes - at least even to relaxed metrology standards.

Excellent rationalization Tiffie of the rankings and necessity for precision over a spctrum of human endevor. You can carry this squareness thing as far to either extreme as you wish and there will still be valid application. On the one hand squareing up concrete froms with a string 3.4.5 triangle and, on the other, 0.01 arc second optics used for research and lab applications. I contend that people in home shops should be aware of the next step in their accuracy even if there is no immediate need for it.

12-28-2008, 07:27 PM
I contend that people in home shops should be aware of the next step in their accuracy even if there is no immediate need for it.

Well put and very true. I may never take that next step in my machining projects but If I don't know what the next step is I'm not going to learn very much.

J Tiers
12-28-2008, 07:28 PM
The OP made a very good job of stating his required limits and performance objectives within the context of his use for the finished job. For those that have forgotten or ignored the OP.....

Well he wanted to do perpendiculars with NO reference.

THAT cannot be done.

You need to make some form of reference or transfer means.

Even the OP alluded to making references, either cylinder squares or whatever. So the clear actual question posed is "how to get a reference".... since obviously the OP understands that a reference is needed.

The rest of the items seem to be directly on-topic for that.

Forrest's method works fine for a square, but cannot check a single surface. You need to check the back as well (assuming it is parallel), or possibly use an auxiliary flat to '"reverse" the front of the square so it can be checked when reversed.

I have proposed a way to accurately check a single surface without referencing the back. *

Whatever method you use, you will need two checks for the surface. One that it is flat, and one that it is perpendicular. If it is flat, it is sufficient to check perpendicular one place, since any other locations must be coplanar, and will also be perpendicular.

*
This is the geometry. A and B are identical diameter disks or balls. Some suitable swinging support for B, and a spring etc to assure contact with the square is required, and A must remain in position.

A frame not terribly different from Forrest's in the other thread would work.

Zeroing from one side, and then checking with square on the other side will show up a difference in indicator reading if the square is not at 90 degrees

http://img.photobucket.com/albums/0803/jstanley/sqchek.jpg

lazlo
12-28-2008, 07:52 PM
Jerry, I'm missing something obvious -- you're using the two discs like toolmaker's balls -- you can indicate off of any tangent.

But what's the advantage over using the DTI directly on the surface, like Forrest's method?

J Tiers
12-28-2008, 08:17 PM
Jerry, I'm missing something obvious -- you're using the two discs like toolmaker's balls -- you can indicate off of any tangent.

But what's the advantage over using the DTI directly on the surface, like Forrest's method?

You do NOT need to try the back side of the square....... only ONE surface is required. You try it on the right side of the disks/balls, zero the DTI, then move it around to the right side and get a direct indication of any out-of-true.

You never use the back side of the blade. The disks/balls form the transfer means, just as an auxiliary flat could, so you indicate off the one side of teh square.

With Forrest's method, which is otherwise exactly the same, you need both sides of the blade parallel, because you first use one side, then use the other as a "transfer means" to get at the complement of the angle. if one side is at an angle to the other, "your error will be erroneous".... there may be one, but you don't know what it is, nor which side has it.

oldtiffie
12-28-2008, 09:20 PM
.................................................. ...

You never use the back side of the blade. The disks/balls form the transfer means, just as an auxiliary flat could, so you indicate off the one side of teh square.

........................................

It might be different in the US as I use the "back/out-side" of the square/s to set stuff up square from a table or to use the reference square that came with my digital and vernier protractors.

I have "backed/back-plated" squares that are used for precisely that purpose. Another is "bevel-edged" for use of the "inner" and "outer" squares.

I think that as many here seem to either be pulling numbers and/or "standards" "out of the air" that it would be instructive and add to the credibility if the standards and quantities and acceptable limits of or within those "standards" were posted.

This is particularly so given that these "Standards" specify different "classes" of items such as surface plates and squares - to mention just a few.

The US and ISO standards are pretty well the same as one is usually a "hard" and/or suitably "rounded" copy of the other - the UK (British - aka BS) Standards are similar.

So how about all you knowledgeable gents who are "calling up" or "quoting" or "citing" those "Standards" posting them for the edification of any who either haven't seen them or don't have access to them?

J Tiers
12-28-2008, 09:30 PM
It might be different in the US as I use the "back/out-side" of the square/s to set stuff up square from a table or to use the reference square that came with my digital and vernier protractors.

No fear... we use it also.

But for purposes of checking, they are two separate surfaces..... or lines..... A square might have a perfect frontside, and the backside could be 'out".

The whole point per laslo's question was that there is a way to check one single surface (line) without needing the other one to verify it. So both can be checked directly.

Otherwise you must first be sure they are parallel by some means, before checking squareness, because the other posted method uses the back to verify the front.

Not all angle plates have usable backsides. Not all square types have backsides that are relevant to the frontside...... they may be several inches away, as on a toolmaker's square, which is a square block (if I have the right term for it).

Teenage_Machinist
12-28-2008, 10:06 PM
AWESOME.

JT you are a genius. This resembles a method I had seen to check machinist's squares with a non-accurate angle plate, toolmaker's buttons, and shim stock. It seemed to me like getting better than about 0.0005 would be extremely difficult and it would not give quantitative measures nor work for scraping..

dp
12-29-2008, 02:06 PM
I have proposed a way to accurately check a single surface without referencing the back. *

*
This is the geometry. A and B are identical diameter disks or balls. Some suitable swinging support for B, and a spring etc to assure contact with the square is required, and A must remain in position.

A frame not terribly different from Forrest's in the other thread would work.

Zeroing from one side, and then checking with square on the other side will show up a difference in indicator reading if the square is not at 90 degrees.

Nicely done, Jerry. I had the same concern regarding the parallelism of the blade. Certainly on the blades of the HF squares I have I would make no such assumption. I have tossed out at least two of them that I have dropped - didn't even check for error. I used the stock of one for a project :)

In practical terms it is in the best interest of the home shop machinist to keep a decent granite surface block and an equally decent reference angle, iron or granite, in the tool kit as it will save a good deal of time checking perpendicularity. Particularly helpful if one is bringing perpendicularity to the work through scraping as was the apparent case for the OP.

Scraping a iron 1-2-3 block to flatness, perpendicularity, and dimension all at the same time has to be a kind of nirvana when successful.