PDA

View Full Version : Building a Cylindrical Square



projectnut
02-08-2013, 06:20 PM
I would like to build one or more cylindrical squares. I've seen several different types and styles and would like some input as to which are more useful. I would also like some opinions as to what material would work best for this application. The ones I've seen range from a solid cylinder to those made from precision thick wall tube. They range in height from 4" to well over a foot and are generally 3" - 4" in diameter. Some include features like one end at a fixed angle to the sidewall and a series of hash marks to determine the exact degree of squareness of the object being measured.

Here's an example of a basic one offered by MSC:
http://www1.mscdirect.com/cgi/NNSRIT2?PMAKA=06504302&PMPXNO=1757448&cm_re=ItemDetail-_-ResultListing-_-SearchResults

Thanks, projectnut

Hopefuldave
02-08-2013, 07:02 PM
Make a bunch of 'em! The overall accuracy depends on the squareness of the base to the cylindrical surface and the degree of taper, so a good method would be to turn the cylindrical OD, mic' to ensure it's parallel, then *without taking it out of the chuck* face the "bottom" end - if made from solid bar rather than tube all of this can be done with tailstock centre support, as you'll want to recess the base to leave a narrow ring contact (to ensure it doesn't rock) once you're finished. They're precision tools, so if they turn out good, make a nice wooden box to keep 'em safe too :)

I have a few of 'em ranging from tiny (a 2-stroke engine's gudgeon/wrist pin 1/2" diameter by 1-1/2") up to a scrapped lathe tailstock barrel (12" tall x 3" diameter, trued up in the lathe to eliminate wear and taper) - it's handy to have the small ones as they can be used in awkward places.

Checking how far out of square a workpiece is is pretty easy if you slip a feeler gauge between work and square, see how far from the base and do a little trig...

Dave H. (the other one)

alanganes
02-08-2013, 07:40 PM
Rudy Kouhoupt wrote an article about how to make these in the Sept-Oct 1997 Home Shop Machinist magazine. The ones in the article are smallish, 1 inch diameter I think, but the method applies to any size you care to make. Reprints may be available or it is likely published in one of the Village press compilation books.

mikey553
02-08-2013, 07:42 PM
The good material is tool steel, either thru hardening or carburized grade. The goal is to achieve high hardness (not less than HRC 60) on the cylindrical and face surfaces. After hardening it needs cylindrical grinding, followed by stress relief and lapping. This is how I believe the professional tools are made.

Any soft material would next to useless since it would not be able to maintain its accuracy over time.

Mike

RWO
02-08-2013, 08:00 PM
In Guy Lautard's Second Bedside Reader, he describes the process a friend used to make cylindrical squares from heavy wall stainless steel tubing (surplus submarine periscope standard) that had been precision ground on the OD.

RWO

Mcgyver
02-08-2013, 08:19 PM
Any soft material would next to useless since it would not be able to maintain its accuracy over time.

Mike

you described the text book and preferred method, but only a small percentage of home shop guys will have cylindrical grinding capabilities. How long soft material lasts is a function of how much its used and with what level of care. Since a home shop guy's works isn't subject to ham fisted knuckle dagger treatment and its used less often, soft materials can work very well.

The thing to make a cylindrical square on the lathe is super consistent diameters. tweak things until its to a tenth over its length and you've got a high class square. Of course undercut the end that is held by centre so you can face a 1/4 land or so that becomes the base. Now turning to a 10th ( or bloody close there to) over any length is not simple, but that's the exercise and what you need for good square.

I've got one that won't budge a tenths indicator needle. just dead on. I got another that is out a tenth and half and another that is out several tenths. All are hardened ground and made by tool and die makers, one good and and two not so good. I don't think I could replicate the good one the lathe, but I could probably do as well as two hacks :D

darryl
02-08-2013, 09:12 PM
I've considered this a few times. For a working grade cylindrical square, you could use some precision shafting. Mount a short section of aluminum tubing in the chuck and bore it until the shafting will make a very close slip fit into it. Without moving it from the chuck, insert the end of a suitable length of the shafting. The other end of the shafting gets held by a steady. With the piece in place, tighten the chuck jaws. This should get that end about as well centered as possible. Now face the end and turn a recess also. You should be left with a lip that will seat flat against a surface plate. Reverse the piece and machine the other end the same. Make two. You can test them against each other on the surface plate. I know from experience that they will be better than many items that are supposed to be square.

I would use something that's at least 1-1/2 inches in diameter to give it a wide enough stance to be usable. It will be heavy enough to stand some feeling around with shim stock or paper. You can polish them up with nevr-dull or something like that, and as dave suggested you can keep them in a protective box. Throw a silica gel packet in the box to help keep moisture at bay.

Rich Carlstedt
02-08-2013, 10:52 PM
I have a Wrist Pin from a Catapillar Diesel Engine
It is perfectly round, and square and ground all over and it is hard.
Couldn't be better

Rich

Paul Alciatore
02-08-2013, 11:17 PM
Another thing I remember from reading about cylindrical squares is the working ones should have a central hole so they can be bolted to a fixture plate or mill table or a face plate. This hole does not need to be bored or reamed or a precise diameter, just a clearance hole for a bolt.

Forrest Addy
02-08-2013, 11:44 PM
Screw the version of cylinder square with the "calibrated" out of square end and the scribed etc. It's a PITA. Make a universal comparator square you can zero with the cylinder square at the preset "reckoning height" you wish to verify. The virture of a comparator square is you can check features and quantify errors to the observable resolution of the dial indicator and do so over protuberant features.

The fixed comparator square has the heavy base whose R is coaxial with the mast. An adjustable dial indicaor attached to a clamp travels vertically on the mast. The base is bumped against the cylinder square and the indicator zeroed. The comparator square is then set against the work under test and the out of square is registed directly on the dial indicator. Dial indicating comparator squares were quite common. Taft Pierce made a dandy in their day.

http://www.ebay.com/itm/Taft-Peirce-Comparator-Square-Indicator-/290231149039?pt=LH_DefaultDomain_0&hash=item43931fe1ef

The universal comparator square is similar but the base has no reference feature except the face on which it rests. It has an adjustable "fixed" jaw that slides vertically on the mast. The reference jaw and the dial indicator may be positioned at any height and separation to validate squareness of any vertically oriented feature once set with the cylinder square. I've jury rigged many an iuniversal ndicating comparator square from a transfer stand with a hunk of keystick dressed to a blunt point and clamped to the mast.

You don't see universal transfer squares. They are rare. The few I've seen were shop made or extemporized with shop made hardware clamped to a regular transfer stand.

You can construct a universal comparator square from a surface gage using the ball on the mast, a DTI and /or other dial indicator hardware and snugs but they are very sensitive to temperature and tiny bumps. If you are forced to assemble one be sure to use "repeat zero" technique to validate before and after zero settings.

I'll see if I can find some links.

danlb
02-08-2013, 11:57 PM
I've considered this a few times. For a working grade cylindrical square, you could use some precision shafting. Mount a short section of aluminum tubing in the chuck and bore it until the shafting will make a very close slip fit into it.

I thought that the idea of the cylindrical square was that once it was parallel to the lathe's spindle and faced, it was pretty much guaranteed to be 90 degrees.

What am I missing?
Dan

Mcgyver
02-09-2013, 12:07 AM
I thought that the idea of the cylindrical square was that once it was parallel to the lathe's spindle and faced, it was pretty much guaranteed to be 90 degrees.



that would be true if there was zero twist in the bed, no wear, the axis of motion of the carriage perfectly straight and the tailstock dead on the spindle axis. I reality it takes some doing to turn say a 6 inch length and have it mic the same to a tenth at each end.....and if its not the same dia, its a cone and it's surface wont be square when sitting on the surface plate

darryl
02-09-2013, 12:25 AM
What I'm talking about is not doing anything to the OD- you start with what might be called superior, TGP, or whatever it is that's already finished. You cut a length from it and face the ends.

The idea of the aluminum ring in the chuck is that it lets you bore a perfectly concentric home for one end of the bar chunk you start with. When you further tighten the jaws after the bar has been inserted, it grips it without throwing it off center. With a steady at the other end, the piece of bar is going to rotate with no wobble. When you then face the end, it comes out perfectly square.

The aluminum ring cannot be used again if it's removed from the chuck, as it won't remount with perfect concentricity. When you loosen the jaws to turn the bar around, you loosen only enough to relax the grip on the bar, not enough to let go of the ring. The ring is a jig, and goes in the scrap once you're done making your parts.

dian
02-09-2013, 01:33 AM
how do you get the lathe to face perfectly "square"? get a precision square, attach it to the shaft and adjust the compound, maybe? otherwise you will have a quare standing on an edge, that will wear rapidly.

by the way, did anybody (rich?) with the appropriate equipment really measure wrist pins, as i use them too.

Forrest Addy
02-09-2013, 01:52 AM
Maybe I'm jumping to conclusions. Making a cylindirical square can be accomplished on a lathe of modest size and middling condition and it's not that tricky. However final finishing and certification is done post machining and calibration has to be carefully done.

Another point is the error in the square's cylindricity has to be 1/4 to 1/10 of the accuracy in the work to be tested with it. Centerless ground or turned and polished barstock may or may not have the cylindricity you may expect of it. Be sure you check the starting material by rolling it in a V block under a 0.0001" indicator before you commit.

Factory cylindrical squares are made of stablized hardened steel cylindrical ground between dead centers (not live) and lapped to finished cylindricity of 0.0001" or better. BTW cylindricity is a metrology term having a specific definition. It's calculated from the measured errors in taper, roundness, rainbow, and other deviations from a true Euclidean cylinder.

A pretty good cylindrical square can be made in the home shop on basic equipment but the work has to be carefully done. Chrome rod and TGP bar stock can be used provided you satisfy yourself that it's roundness etc is up to snuff. One can make a good cylinder on a shaky home lathe provided the finish cut is taken between dead centers. Think of locking the spindle and driving the work directly via a belt from a separate motor. Use a power drill or something variable to run the saddle feed. Take light cuts with a heavy feed and a flatted nose radius. Spend the money and buy a single Sunnen external honing stone to refine the finish, hone out residual taper, etc until your work is within 0.0001" of a common dia for the full length. Then face the registration end flat and square relieving the center to 60% of the outside diameter. The registration face should be a narrow rim, easily corrected by abrasive sheet lapping (400 grit!) should it be necessary.

Set your new square on end on a surface plate and check the squareness of the cylindrical body at 8 points on the diameter. If the square is tipped a trifle, determine the points of maximum acute and obtuse error and mark the upper surface. The square's diameter at 90 degrees to the plane of max error will be square.

Squares qualified to read 0.0002" in a foot can be made following this recipe. Remember that a dial indicating comparator square is an inescapable necessity for a cylindrical square. These are also easy to make. Maybe you chould make the comparator square first because you will need it at the end.

Relax the accuracy if my suggestions seem to be persnickity for your purposes but rememer that if errors are present they may loom up and bite you where you're least prepared. Know your equipment.

darryl
02-09-2013, 02:31 AM
Checking of wrist pins, etc is pretty easy. After checking to see that there aren't any burrs, etc, just stand them side by side on a surface plate. Slide them sideways until they barely touch. Hold one steady and rotate the other one, making sure to keep both firmly in contact with the surface plate. If there's error, you'll easily see it in the gap between them.

Forrest has gone more deeply into it than I did, but basically a lathe will automatically make an end square to the body of a cylindrical section, as long as that section doesn't wobble as it's being spun. There's no need to make adjustments to the compound or whatever. If you machine the recess in the end, that takes care of it possibly rocking when placed on a flat surface since only the outer, full length area will be touching. If it was a tube you were turning, it would basically be the area defined by the end of the tube that would be the contact face- recessing the central portion of the solid bar gives the same contact area. Unless your lathe spindle is sliding in and out as it rotates, there's virtually no chance that a skewed facing on the end would result- provided the workpiece is rotating without wobble.

That is the premise of my method- the OD of the bar you use does have to be well and accurately machined from factory, and your main job is to ensure that the lathe is turning it with no runout- the lathe then automatically faces the end square.

I suppose to clear up (or possibly confuse) another issue- it has been said that a lathe is set up to turn a very slight concave when facing. This is to allow a faced part to be able to sit flat without rocking on a flat surface. If this action happened, and all the way to the exact center, then it's acting as the recess and giving the desired result. Only the outer edge of the face would contact the flat- thus the stability is maximized. In reality, because the angle would be so slight, the contact will be over more of an area than just the extreme outer edge. Recessing the solid bar just ensures that no material inside this area is capable of touching the surface to possibly cause a problem.

dian
02-09-2013, 02:47 AM
daryll, so were they square? i still use a kitchen top for my granite plate.

"Think of locking the spindle and driving the work directly via a belt from a separate motor."

does this take care of the slop in the bearings, just by the weight of the shaft? is it better than turning a center in the chuck and driving by that?

MichaelP
02-09-2013, 03:05 AM
The good material is tool steel, either thru hardening or carburized grade.

Mike,

I'm afraid this is not an accurate statement. All TOOL steels can be hardened through since they have relatively high carbon content. And this ability to be hardened through is what makes them TOOL steels. You may harden them through or case harden, but they don't require carburizing.

Mild (low carbon) steels would require carburizing (or a similar process) to be hardened. And you can only achieve case hardening with those.

Jaakko Fagerlund
02-09-2013, 03:15 AM
And if one does harden, the usual hardness is 58-60 HRC for measuring tools, depending on what the tool is and what properties are wanted. For a home shop tool the hardening isn't necessary as pointed out previously.

If one puts a through hole for a bolt in the square, I suggest putting a hole and a tapped end to it. For example, drill 8.5 mm hole for an M8 bolt (or even 9 mm), but tap the other (or both) ends then M10. This adds some flexibility to attaching it and is quite stndard method in some parts that have a through hole for a bolt.

darryl
02-09-2013, 06:01 AM
Dian, I've turned many things on the lathe that needed to have a long dimension perfectly perpendicular to an end. This usually means turning the diameter as well as facing the end. There is a risk of turning a taper, so in this regard the lathe has to be carefully set up so it turns true. There is much more to this of course, beyond this discussion, but suffice to say that if you turn a diameter on something, it should be the same diameter all along its length. If your lathe can do that, and you face an end, the end will be square to the side- simple as that.

You're using a kitchen top as a flat- is this a piece of granite top, or your typical countertop laminate material? Either might be fine, depending on your application, but at some point you might need a known accurate reference surface. There is little point in trying to test a square for accuracy on a surface which is itself an unknown.

The method which Forrest brought up about having the workpiece between centers and separately driven takes much of the possible inaccuracies out of the picture when turning something to be as true as it can be. It basically eliminates the headstock bearings from the picture, and the bearings in the tailstock center, if it's 'live', or in other words has bearings of its own. If the two centers don't spin and don't move around, then the workpiece with its dimpled ends can't move around either- only spin. But you have to spin it, thus the separate motor and a belt. You can turn the workpiece around and it will run just as true. This is perhaps the most used 'feature' of turning between centers.

You mentioned turning a center, then driving the workpiece by that- I presume you mean by using a dog. This is probably the most usual way of using centers. It will not be more accurate than having both centers non-rotating, but if your spindle bearings are good and the spindle has no slop, then a turned center will be pretty darn close to not wobbling around as well. By good I mean it should not have any random off-axis motion beyond a couple of tenths, which is about the borderline for making a cylindrical square able to show deviation of about 1 thou over 4 inches or so.

My method of using already precisely machined rod means that you don't have to turn the OD and go through the process of getting it precise all along it and beautiful. You still have to ensure that it spins with as little runout as possible, and because it's being turned by the spindle it's subject to the spindle runout. It's going to be virtually impossible to set it up between centers without having to turn the OD, so a steady rest is also needed.

This is interesting- suppose you mount the aluminum ring, bore it, then fit the workpiece and tighten the jaws down. That end should be very concentric, and if you knock the other end around bit by bit until it also has no detectable runout, then set the steady fingers to keep it there, you will have a true running setup. But suppose your steady doesn't precisely locate the outboard end of the workpiece exactly on the spindle axis. You will still be able to face it perfectly square! The faced end will either be coned or concaved, but if you recess the center portion you will still have a reference end that is square to the sides. Of course, if the steady is off by much at all, the workpiece will walk right out of the chuck. You should be using an indicator as an aid to getting it all working properly.

Forrest Addy
02-09-2013, 07:27 AM
...

My method of using already precisely machined rod means that you don't have to turn the OD and go through the process of getting it precise all along it and beautiful. You still have to ensure that it spins with as little runout as possible, and because it's being turned by the spindle it's subject to the spindle runout. It's going to be virtually impossible to set it up between centers without having to turn the OD, so a steady rest is also needed.

This is interesting- suppose you mount the aluminum ring, bore it, then fit the workpiece and tighten the jaws down. That end should be very concentric, and if you knock the other end around bit by bit until it also has no detectable runout, then set the steady fingers to keep it there, you will have a true running setup. But suppose your steady doesn't precisely locate the outboard end of the workpiece exactly on the spindle axis. You will still be able to face it perfectly square! The faced end will either be coned or concaved, but if you recess the center portion you will still have a reference end that is square to the sides. Of course, if the steady is off by much at all, the workpiece will walk right out of the chuck. You should be using an indicator as an aid to getting it all working properly.

Wel-l-l kinda.

My experiences in using chrome rod, drill drod, TPG etc for precision pins and quick and dirty metrology has not always been a walk in the park. These materials are often to-size, free of taper, and smooth but often lobed and rainbowed. Odd-number lobed error is indetectible with a mike and tough to detect in a 90 degree V block. Use pre-finished material if it attracts you but remember it's a "trust but verify" thing.

The ring? If the work is inserted in a bored aluminum ring, what is the ring's L/D ratio? Less than 1 and there's hazard of the piece being cocked. You mention knocking the unsupported end into concentricity which is good practice. The questions are has the ring been carefully fitted to the jaws? Is the ring "death-gripped" to ensure stability regardless of minor jaw adjustment? How do you go about adjusting the jaws to grip the work after finishing the ring? Three jaw or four jaw? If its a three jaw chuck whose grip concentricity cannot be adjusted, you stand a good chance of disturbing the TIR. Four jaw, no sweat. That's what four jaw chucks do best: tweak TIR to near zero.

Additional point. Anyone know what a half center is? It's like a regular lathe dead center with one side of the 60 degree cone cut away somewhat short of the apex. They're commonly used on cylindrical grinders but they're handy on lathes too. The cut-away allows you to turn/face work very close to the centerhole.

Look here; http://www.ebay.com/itm/MT-2-HALF-DEAD-CENTER-8-29-12-/180963066342?pt=LH_DefaultDomain_0&hash=item2a223d41e6

RLWP
02-09-2013, 07:34 AM
{snip}

lobed and rainbowed. Odd-number lobed error is indetectible with a mike and tough to detect in a 90 V block.

{snip}

Addy, I understand lobed, what is rainbowed please?

Richard

Forrest Addy
02-09-2013, 07:45 AM
Addy, I understand lobed, what is rainbowed please?

Richard

Very slightly bent or bowed. Centerless grinders running miles of bar are not fonts of cylindrical perfection.

RLWP
02-09-2013, 08:08 AM
Ahh, that makes sense, thank you

Richard

Forrest Addy
02-09-2013, 08:22 AM
Crap! I just looked at the time. 5:21AM PST and haven't been to bed yet. I gotta get a life.

Euph0ny
02-09-2013, 09:40 AM
There is a nice description of how to turn a (small) cylindrical square on a home lathe at pages 35-44 of Harold Hall's book, Lathework, a complete course (http://www.bookdepository.com/Lathework-Harold-Hall/9781854862303), which is no. 34 in the Workshop Practice Series.

J Tiers
02-09-2013, 11:33 AM
I suppose a cylindrical square is one way...... but turning it is unlikely to give a true surface. The issue is tool marks..... grooves, burrs, "fuzz" etc.

You probably need to do it between centers, for the ability to remove and replace.

I'd turn it as well as you can, then start to polish it. When you have the 'fuzz" off of it, you can check the straightness by trying it against a granite flat (or other good reference).

Work a bit harder on places that show up as high, checking diameter as well, although that should be OK aside from possible taper.

When you have the print as an even marking up the length of the "larval square" in several places, and still have good diameter, you can finish cut the end. I'd replace it in the centers, and check with an indicator for wobble, and when it is right, cut the ring. With care and a shaving cut you can get that good and smooth, but will probably still need to polish it, then check with indicator and the flat for flatness and trueness.

A comparison against a good square, or a "comparison square is also a good plan

The WIDEST possible square is likely best, as you have the least squareness error for a given amount of inaccuracy cutting the end. The problem is maybe not having larger mics reading in tenths.

I thought about it, but bagged the idea, instead scraping three squares together to get my reference.

Mcgyver
02-09-2013, 11:35 AM
Crap! I just looked at the time. 5:21AM PST and haven't been to bed yet. I gotta get a life.

Forrest you're behaving like one of my teenagers....sure there is no video games involved :D :D

can you elaborate on the indicating square? is this essentially the function (enhanced perhaps?) of the indicator on a surface gauge and using the ball end of the spindle?

Someone trashed that approach in the shop made tools thread (but also iirc admited that hadn't tried it lol). imo that method is one of the most accurate methods to check for squareness so long as the stand is solid on the plate and you are using a good quality indicator. With a starrett grade A plate and interapid tenths indicator and my best cylindrical square the needle consistently goes exactly to the same position suggesting its indicating squareness to a 1/4 of a tenth or better. I've a 1/100,000 indicator I should try just as an experiment, but its a big cumbersome thing....not that we care about squareness to that degree but it might quiet the doubters who haven't experience with it

Forrest Addy
02-09-2013, 12:07 PM
McGyver. Yeah, some people have difficulty with the concept that squares are self checking. You need no comparison standard (ie, another more accurate square) to qualify it.

The indicating square test I illustrated some time ago is difinitive and adapted from NIST procedures for checking squares. If a square is defined as a tool used to establish a perfect right angle from a reference plane and whose exteror and interior angles are equal whose stock and beam are flat and parallel then the reverse angle and repeat zero test is valid to the limits of the testing apparatus.

Look in the link at the eBay auction for the Taft Pierce comparator square I provided in my post #11. Substitute the 0.0001" graduated dial indicator with an LVDT and gage head and you can prove squares to 0.000005" in 10" (1 part in 2 million slope) - provided the surface plate is FLAT. A local undulation of 0.000010" in 10" (not unhead of even in AA plates) will wreck such refined readings. And temperature. I don't know how long stuff has to soak in a 0.1F climate controlled room to attain thermal equalibrium within 0.1 degree F of its environment. Better work with remote manipulators.

But all this high falutin excess, probably more than a guy with a 9" lathe in a basement shop wants to contend with. I do get obsessed; ask me the time and I'll tell you how to make a clock.

Mcgyver
02-09-2013, 12:20 PM
right, that would make a good project. The critical thing is the base is flat; a good beginner scraping project. Careful work with the surface gauge works so well that I may not make one, but do see that it would be easier to use, is heavier, more solid and could be left set up.....maybe I will make one :)

danlb
02-09-2013, 06:18 PM
That's a good point. A really precise square is overkill for most of us.

I don't think I have a surface in my house that is flat within .0001 over 6 inches.

Dan

J Tiers
02-09-2013, 08:56 PM
I suspect the attraction of cylinder squares is a sort of "something for nothing" deal, where it seems that the process will "generate" the correct 90 deg angle with less work than any other method.

I'm not sure it;s really true, although it could be nearly true, particularly if you finish the thing with a tool post grinder, both on cylinder portion and also (and especially) on the end.

Ultimately it won't be any rounder than your machine can make it, and won't have any less error than your centers produce. But at least finishing it with a grinder will take care of the "machining fuzz" issues.

Carefully done, it can make a pretty darn good square.

projectnut
02-09-2013, 10:05 PM
My questions to all of this are:

- do you really need a cylinder square because there really are are no alternatives to having and using a cylinder square?; or

- do you just want one for the sake of having one without any real known use or pupose for it?;

- do you just want to make one for the sake of doing it?

I made several "near enough" cylinder squares and found that I only ever needed them for squaring up angle plates on my mill (turned out very well as the angle plates are very square) and I haven't needed or used the cylinder squares since.

I can do just as well or better with a machine square or a granite square - both of which I bought and use.

http://i200.photobucket.com/albums/aa294/oldtiffie/Surface%20plate/630SPandsq1.jpg

http://i200.photobucket.com/albums/aa294/oldtiffie/Surface%20plate/Precgransqsheet1.jpg

I can't really see the need for a "square" with accuracy exceeding 0.0001" per 6" - as these precision granite squares are:
http://www.shars.com/product_categories/search/?search=granite+square

Or this machine square with an accuracy of 0.02mm/Mtr = 1 in 50,000 or 0.00012" per 6"

http://www.machineryhouse.com.au/Q208

I've used them in the past and liked the simplicity and accuracy. We had several in the shop where I worked. It might be a little old school, but there's something to the challenge of making a precision tool. While I probably don't need accuracy to less than a tenth per 6 inches I like the fact that it can be done. I don't have either a surface grinder or a cylindrical grinder in the shop at this time. Both are on the list, but so far the ones I've found take up way too much real estate.

I'm able to go to the local technical college 1 evening a week to use machines I don't have in my shop. Since the time and equipment are available I thought no better time than now to make 1 or more. I could just buy them on e bay or from a supplier but why pay several hundred dollars for a tool that will get limited use when I can make it myself at little cost. Plus I can have the pride in the accomplishment (assuming it turns out as expected).

Also I'm a big believer in shop made tools. I would rather fulfill a need by creating a solution rather than throwing money at it.

I understand there are a number of alternatives, but the object is to make a quality tool that can be usefull

wierdscience
02-09-2013, 11:13 PM
I've got two I made from some new 3" chrome cylinder rod I had left over from a job both 6" long.

Dialed it in chucked in a four jaw to less than .0001" T.I.R.,faced the free end and recessed it leaving a 3/8" wide rim.With the workpiece in the same position I used a parting tool to plunge in .500" after cranking the carriage back the finished length plus .002" using jo blocs and a feeler gauge.I used the band saw to finish off the cut after which they were re-chucked to recess the cut end.

Once out of the lathe the length was checked with a mic at four positions and found to be .0015 over length.I taped down a sheet of 600grit to my surface plate,set the first end of one square on it and gave it a push across.Rotated it 90* and pushed it back,rotated it again and pushed rinsed and repeated as needed until 6.0000" was the reading on the mic.

At that point I had two 3" round chrome plated blanks as exactly 6" long as anything in my shop could tell.I set the two blanks next to each other on the surface plate and brought them together until they touched.Put light on the backside and looked at the opposite side of the "gap".I rotated one blank 90* at a time and didn't see any signs of light getting through.Last time I looked photons were pretty small so at that point I called them cylinder squares and haven't looked back.

Are they accurate to .000001?I don't know,but they are accurate to .0001" I am sure,which is good enough for me.

danlb
02-09-2013, 11:19 PM
A piece of grit is a piece of grit, no matter what type of square it's under.

You can tilt a square sideways to establish a line of contact on one edge to avoid the grit, and since it would be 3 points of contact you might not even know.

By the same token, a cylinder sitting on a piece of grit will tend to rock since it is supported on two points.

In my mind the cylinder is "magic" in that the sides are a single face that can be rotated to confirm that the face is 90 degrees to the base. I dropped

Dan

.RC.
02-09-2013, 11:19 PM
I was quite pleased with the results as they turned out quite well.

How square were they?

Pleased with the results is a inconclusive statement as they could be out 0.1mm but that is as good as you were aiming for....

Forrest Addy
02-10-2013, 02:43 AM
I really do wonder why people make or need a cylinder square as a precision granite square will do most jobs. But if you buy one from say a Dealer they are expensive:

Curiousity and desire to expand capabilities are sufficient motivation to pursue info on cylinder squares and their uses and maybe buy/make one. No-one needs approval whether cylinder squares are expensive or not.


Master precision machinist's squares take some beating for price and accuracy too:

But they do not serve all purposes.


I would never buy or make a cylinder square unless it was a last resort - which it was to "square up" my angle plates (see previous posts).

So why are you posting here is all you have to contribute is uninformed skepticism?


A good machinist's square usually does all I need a square or otherwise its my machine square or granite square.

Your needs do not necessarily suit all home shop and small commercial shop users.


One thing that has not been mentioned to date is the requirement for a steady consistent ambient temperatue of shop machine and job as the temperature co-efficient of steel is reasonabley high and any differences in temperature may well cause different movement/sizes.

Mentioned several times. Read over the previous posts. Coefficients of thermal expansion for metals are widely available and citing actual numbers may be more illustrative than non-quantitive descriptors like "reasonabley high" Coefficients of linear expansions for common metals vary widely from 2.7E-6 (some grades of tungsten carbide) to 16.6E-6 (copper). Take a look at zinc, lead, cadmium, magnesium,...

http://www.engineeringtoolbox.com/linear-expansion-coefficients-d_95.html


Sop that almost suggests that a reasonable metrology environment is needed.

Mearurement and the need to measure to finer resolution is a continuing need to any shop not commited to defensive nihilsm.

Jaakko Fagerlund
02-10-2013, 03:25 AM
Making the cylindrical square (cylindrical..square? talk about a glitch in the universe :D) is quite easy as it doesn't need any accurate measures, just consistent measurements when measuring the OD.

BigJohnT
02-10-2013, 08:06 AM
Additional point. Anyone know what a half center is? It's like a regular lathe dead center with one side of the 60 degree cone cut away somewhat short of the apex. They're commonly used on cylindrical grinders but they're handy on lathes too. The cut-away allows you to turn/face work very close to the center hole.

Look here; http://www.ebay.com/itm/MT-2-HALF-DEAD-CENTER-8-29-12-/180963066342?pt=LH_DefaultDomain_0&hash=item2a223d41e6

I knew what a 1/2 center was but didn't know what it was used for. At least I learned something from all the babble on this thread about being within millionths or your no good LOL.

Seems to be no middle ground thinkers on this forum...

John

Mcgyver
02-10-2013, 09:07 AM
Curiousity and desire to expand capabilities are sufficient motivation to pursue info on cylinder squares and their uses and maybe buy/make one. No-one needs approval....


Well said! Nothing is more tiresome than the rhetoric BS "you don't need that" or "good enough for home use". Its indicative of, well, something.... that one assumes their paradigm apples to all. Newsflash, because I don't have to do any of this, it's for kick and giggles, I can do or need or buy any damn thing I want! :D


So why are you posting here is all you have to contribute is uninformed skepticism?

Better program that line into a function key

J Tiers
02-10-2013, 08:57 PM
I don't really care what anyone else does or thinks about cylinder squares.....

But, if you are gonna make one, you may as well do as good a job as possible at it. I'm not sure what the "within millionths or don't bother" deal is..... as if there was something WRONG with that attitude..... It's as good (at least) as the attitude that "it's good enough for me if it's pretty close, I ain't workin fer NASA"..... Accept one, you have to accept the other unless you want to be a bigot.

A cylinder square is normally a precision comparison tool. (Versions that will *measure* are also made.) Since it IS a precision tool, as usually produced or made, there would seem to be little use in making one that was NOT as good as you could possibly get it....

If all you really need is a framing square that isn't too badly bent out of shape, then why bother with a cylinder square? Of course, if you want to make a cylinder square, and really don't care if it is a few degrees "out", well, you are free to do that, for sure, but others may be forgiven for wondering why you would bother ......

John Stevenson
02-11-2013, 05:19 PM
How square is the head on that machine and what was it referenced to ?

J Tiers
02-11-2013, 09:38 PM
The nominal length is not important but both ends must be very flat and highly finished and parallel to each other.



Why that?

ONE end is all that goes on the granite flat, etc, so the other end is not very important except for looks and to make sure the center runs true.

In fact, it is quite common to make the top with a goodly "spigot" on it so there is a place for the dog to clamp on when machining between centers. IIRC the designs listed in the HSM article of maybe 25 years ago had that.

Mebbe if you want to be able to use either end, but other than that it seems that one very flat end is plenty good.

lazlo
02-11-2013, 09:39 PM
http://www.ebay.com/itm/Taft-Peirce-Comparator-Square-Indicator-/290231149039

Tool porn! How many here made offers? :D

TGTool
02-11-2013, 10:08 PM
No offers but I noted and memorized features I can duplicate at home for a somewhat lower price. :D And I see where I can use that tenths reading AGD indicator sitting in the drawer. Hmmm.

TGTool
02-11-2013, 11:21 PM
Good question.

Machining both ends is a very good secondary check of the square - and it works very well as a sort of largish "slip gauge" as well.

If it is dropped and spoiled you have a spare reference (bottom) that is the basis for the accuracy of the cylinder with reference to the surface plate or what ever it refers to. Resurfacing the end of the square with only one accurate end that has been damaged is quite a task.

Given the cost of the second face against the cost of the square and the risks involved I'd be more than happy to pay extra (in the unlikely even that I'd buy one anyway).

Give it a rest. That's pretty thin arguing. How does a secondary check come into play? For an ordinary check of squareness, once assured that the sides are parallel, a turnaround (rotation) test with an indicator tells you it's true. If it's square turning it upside down doesn't help because this is now testing new geometry and a totally different surface. If it's dropped and damaged it either needs to be repaired until tests show it to be once again correct or not used.

If both ends are trued square that's all well and good, and one could use it either way up, but if there's any doubt about one end being less accurate, suddenly you've introduced a new doubt and variable into the shop. Are you sure the craftsmen are using the accurate end or might some be mistakenly using the inaccurate one? With a single ended cylinder, if in doubt it's simply taken out of use and not left to muddy the waters.

danlb
02-12-2013, 12:39 AM
Give it a rest. That's pretty thin arguing. How does a secondary check come into play? For an ordinary check of squareness, once assured that the sides are parallel, a turnaround (rotation) test with an indicator tells you it's true.

The explanation I heard was for the case where you don't have a precision square to measure it against.

Checking end 1: Place against a vertical surface and rotate. Any gap should be consistent all the way around. If there is no gap you are good to go.
---- but ----
If there is a gap, and it is not consistent the cylinder is out of round. If the gap is consistent there might be a taper. To eliminate taper as suspect turn it upside down. If the gap remains the same the reference surface is not vertical. If the gap changes from high to low there is a taper.

That's how I recall it was described.

As noted earlier it's easy to face both ends by using a parting tool for the end nearest the headstock. You only need a small ring to do the trick.

Dan

J Tiers
02-12-2013, 01:01 AM
If you can face both ends, then you must have a hole through in order to put it on a mandrel for machining.....

Naturally quite doable..... but it's extra steps and makes it harder to re-finish later if dinged or worn.

danlb
02-12-2013, 01:26 AM
If you can face both ends, then you must have a hole through in order to put it on a mandrel for machining.....

Naturally quite doable..... but it's extra steps and makes it harder to re-finish later if dinged or worn.

No need for a mandrel. Between centers was mentioned as the best way to turn it, so it's easy to turn that outer 1/4 inch band that is used as a base. The center of the base need only be indented a few 10s of thousandths.

Dan

.RC.
02-12-2013, 04:54 AM
The angle plates turned out very well too - so far as I could see - and more than adequate for my needs.



To be brutally frank, I do not know why you did not either bin the crappy out of square angle plates or take them back for a refund.... You often say if it does not do what you want to bin it and get another one... You do not spent time fixing it...

.RC.
02-12-2013, 07:10 AM
Your being "brutally frank" seems to indicate that you are getting wound up about it for no real good reason as it will do you no good- and you are none the wiser.

Oh no, I was just wondering why someone who always said they bin equipment they no longer need, or is not suitable for their needs did not actually bin something that was not suitable for their needs... I am not getting wound up about it at all.... You are free to do as you wish....

J Tiers
02-12-2013, 08:33 AM
No need for a mandrel. Between centers was mentioned as the best way to turn it, so it's easy to turn that outer 1/4 inch band that is used as a base. The center of the base need only be indented a few 10s of thousandths.

Dan

But there is the issue of driving it. No place for a standard driving dog. Tiffie alluded to it, but you'd have to make a special pin setup to drive down in the recessed area.

never said you can't do it, just that the benefits don't seem to really be there. if you want to do that it's fine.

It only came up when I asked Tiffie why there MUST be a perfect surface on both ends, and he turned the "must be" into a "preference", which is what I figured was the real truth of the matter.

I have not made one, and probably won't, but I am always interested in techniques and wondered why in the world he was saying you HAD TO do that.

Mcgyver
02-12-2013, 09:56 AM
But there is the issue of driving it. No place for a standard driving dog. Tiffie alluded to it, but you'd have to make a special pin setup to drive down in the recessed area.

.

The ones I have are are far too large in diameter for any lathe dog I'd have lying around. A mandrel is actually a good idea imo, leaves the entire surface open for machining and is easy to fix a dog to. Mine have centre holes so didn't use a mandrel, however all have
axial holes for lightening....so it wouldn't be to difficult to make one hole centred and then use a mandrel.

Seeing the different uses a good reminder that these things can come in all shapes and sizes and are used in varying ways..... and one guy will want it to be the squarest of square as an inspection device (for setting indicators etc) whereas another will use it as a milling fixture.

JCHannum
02-12-2013, 10:50 AM
I believe either Frank McClean or Rudy Kouhoupt used a spigot on the end of the cylinder square for clamping the dog while turning between centers. I seem to recall that each had articles, but can't remember the exact details. It is one method of accomplishing the job and leaves a convenient handle.

Rudy Kouhoupt also demonstrated the use of PT&G shafting or hydraulic cylinder shafting, carefully faced square and undercut, tapped on the end and attached to the faceplate as a method of fixturing to face an angle block square ala Tiffie's fixturing to true his blocks. I have a couple residing with my lathe tooling for such purposes. I would refer to such tooling as fixtures, rather than cylinder squares as it would be of little value for inspection purposes. Not to sound pedantic, but there is a difference in use, application and necessary accuracy.

TGTool
02-12-2013, 11:12 AM
The explanation I heard was for the case where you don't have a precision square to measure it against.

Checking end 1: Place against a vertical surface and rotate. Any gap should be consistent all the way around. If there is no gap you are good to go.
---- but ----
If there is a gap, and it is not consistent the cylinder is out of round. If the gap is consistent there might be a taper. To eliminate taper as suspect turn it upside down. If the gap remains the same the reference surface is not vertical. If the gap changes from high to low there is a taper.

That's how I recall it was described.

As noted earlier it's easy to face both ends by using a parting tool for the end nearest the headstock. You only need a small ring to do the trick.

Dan

Ah, perfectly good point. I was still thinking in terms of having an indicator to quantify the condition and you're showing that it can be done just by having SOMETHING else as a static reference. That's useful to keep in mind and probably the only way it could have been done initially.

Steve Seebold
02-12-2013, 02:05 PM
I would use 1018 steel, put a center hole in each end along with a screw hole to be used for a driver in one end. Have them case harden .030 to .050 deep, then have the OD's ground, and then kiss the ends to minimum clean up. Size doesn't matter as long as the grinder does the diameter and the ends without removing the part. You could probably get a dozen of them ground for a couple of hundred bucks.

I made one like this about 45 years ago and I still use it.

Paul Alciatore
02-12-2013, 06:24 PM
Since I am not a diesel mechanic, I am not familiar with these pins. How big are they and do they have a axial hole? And what do they cost?



I have a Wrist Pin from a Catapillar Diesel Engine
It is perfectly round, and square and ground all over and it is hard.
Couldn't be better

Rich

J Tiers
02-12-2013, 07:39 PM
Sir Tiff:

Methinks thou dost protest too much.....I have not rated thee about thy measures and thy bins so recently, nor in the Rialto nor out of it, nor else're 'thin our common bounds upon this mortal coil.

I did'st but inquire, that thou might dispense thy thoughts as the dew from heaven, freighted with thine own good wisdom as the most excellent gift of knowledge and elucidation, here but touching the higher planes and the lower, so relating them that all might see the means and sense of their consanguinity.

Thought I naught but to ask, nay, to encourage, thy exposition of these points and surfaces, having due regard to the hounds of rotation with their attendant dens and holes, that a higher understanding might be set at liberty upon the world, that benefits be sow'n most widely an' perhaps take hold to grow, to in their turn produce increase according to their kind. May I set forth my thanks ere I exit, and leave no tumult in my train.

John Stevenson
02-12-2013, 08:07 PM
If Tiffie mentions binning his gear once more I'm going to donate $1,000 to Village press just for the pleasure of being moderator for one day just so I can ban the troll.

uncle pete
02-12-2013, 10:13 PM
Paul,
Wrist pins? There's a variety of sizes. Right from something out of a lawn mower or chain saw up to the worlds largest ships engines. I got mine from a commercial truck repair and a couple of larger ones from a V-12 Cummins mining truck engine. Yes used will have some wear, but there hardened and ground to very tight tolerances with very accurate ground ends, much tighter than most could do in a normal home shop. And the largest area covering the central portion of the pins gets no wear at all. They work very well as a very true 90 degree surface to a mills table, or lathes faceplate while fixtureing work. They already come with a through hole so it's dead easy to fasten them down. Any large truck repair / engine rebuilder's should be able to provide you with a couple @ no charge since if there at all worn, they don't get reused and there judged as scrap. Explain what your looking for and why and I'm sure you can get a couple pretty easy. A box of donuts might help though.

Pete

JCHannum
02-14-2013, 09:59 AM
When someone asks how to make a particular item, any response that suggests the purchase of that item is worthless and repeated responses to do same become tiresome and add nothing to the discussion.

Yes, you can purchase anything you might need in various qualities for a range of prices. However, the aim of the HSM often is to make it himself for no other reason than the experience gained in the process. A HSM with reasonable tooling and abilities can duplicate many items with at least the quality of low priced imports. With a bit if skill and care, he can certainly exceed that low bar.

There are many choices and types of squares available. Each has its own attributes. One is not better than the other, just different. The OP dealt with cylinder squares. Suggestion of other types is useless and repeated mention also becomes tiresome.

Black_Moons
02-14-2013, 12:21 PM
If Tiffie mentions binning his gear once more I'm going to donate $1,000 to Village press just for the pleasure of being moderator for one day just so I can ban the troll.

I would recommend he give them away to someone he likes. Then I realised there must be nobody he likes, hence why he bins everything. He likes the trash can best and wants to be forgotten when hes gone.

Mcgyver
02-14-2013, 12:44 PM
I've never seen a need for me to have - or make - a "cylinder" square as there is so much available with "flat" faces.

Yet in a thread with 81 posts you have 19. The next most prolific poster is Forrest at 7 who contributed much and knows what he's doing.... probably more than any here....yet on a project you view as not worthwhile you have almost 3x as much to say as Forrest. You do this all the time, have you no sense of how ridiculous it looks? and its painful for the rest of us, well speaking for myself anyway.

uncle pete
02-14-2013, 01:33 PM
I thought the thread title and the OP first comments about wanting to build a CYLINDRICAL square were fairly easy to understand. Building or buying any type of cylindrical square should obviously be about the cheapest bang for the buck as far as the degree of accuracy verses cost and just how simple they are to check compared to any form of precision flat type square I know of. You could I think say that a good cylindrical square forms one of the basic trusted reference surfaces that are sooner or later needed in any shop. And if I did buy or build any flat type squares? I'd certainly be checking them against either of the two types of cylindrical squares I did buy. I don't necessarily blindly trust any manufacturer, and I have and will continue to double check even Starrett or Mitutoyo equipment. So far there equipment has always agreed within factory specifications, but I know it does because I checked. If I hadn't bought my first cylindrical square, I would have built one because of the need for that reference surface.

Pete