Enjoyed! How much time passed between your two measurements? A consistent 0.0001" all over sounds suspicious. Wonder if there was a temperature change? I think a 5 degree change, in steel, = 0.0001" over an inch. Nice square, by the way.
Trust me Tony, I'm not a square......bad joke? Glad you enjoyed and thanks for stopping by! I don't recall the temps being much different during the two days but 5 degrees difference it could have very well been. I think what I have is a "very close" square, good enough for government work right?
Carbon steels have a thermal coefficient of expansion about 11ppm/deg C or 6ppm/F. So 1/100,000 inch per degree C per inch; 0.000 006 inch per inch per F psec.uchicago.edu/thermal_coefficients/cte_metals_05517-90143.pdf The biggest effect would be warpage of the surface plate due to an internal thermal gradient if the shop temp fluctuated over just a few hours (overnight). The larger thicker plates are stabilized for several days before lapping or calibration, depending on the desired rating of course, AAA lab grade has less wiggle room in the specs than a class B shop plate (though they both warp the same amount)
Love learning with you Abom! You have inspired me to start machining, bought myself an American Pacemaker! I really like hearing stories about your Dad & Granddad, keep them coming! Thanks!
Stefan Gotteswinter has made a rather nice squareness comparator, one I plan to use for inspiration for exactly this kind of measuring. Looks like metrology is something all the cool kids are doing these days! Thanks heaps for sharing
I plan to build a reference square like that soon. Very cool! A good method for checking square on that cylinder is as follows: Use a 1" or 3/4" OD bearing-ball in the V-notch of a surface gauge with your .0001" indicator to create a squareness comparator. Place the cylindrical square on the surface plate and push the comparator-ball up against the large OD of the cylinder. Adjust the test indicator so that it is testing the top of the smaller OD of the cylinder. Zero the indicator to the "high" spot as you pivot the surface gauge at the bearing ball. Rotate the cylindrical square around and keep finding/checking the high spot (should repeat to zero if it's concentric). Use a sharpie to mark the difference from your zero on the square. Now repeat the same procedure with the indicator testing the lowest point of the smaller OD of the cylinder. Finally, use your best set of .0001" mics to verify that the small diameter is consistent (not tapered). These three tests will eliminate the V-block setup and the errors it can introduce. Also, you'll get more from your granite because you're confining your test to a small area with very little movement, just rotation and minor pivots. You can check several points along the height of the cylinder, but the most accuracy will be from very near the top and bottom ends. There are better ways to build a squareness comparator, that's just the low effort way and it sure works well.
Excellent post, cwtoyota. Another thing is that a cylinder with a slightly tilted but flat base has very near perfect accuracy at two clock positions. Test it as suggested above, and mark the two places that are vertically parallel on the cylinder square. For fussy work, just use those two positions only. The base can also be lapped to get the cylinder perfectly vertical, but flatness can be compromised (or repaired) in the process, that is a fussy job in itself...
Thanks bob. That's a good point and a good way to use an imperfect tool to do higher quality work. Those two clock positions should be exactly 180° apart if tilt is the cause. On a cylinder of those proportions (height is greater than the width of the base) a small dust particle, or small grinding imperfection in the base makes a large indicator reading at the top. If the height is 2 times the width of the base (2:1 ratio) .00005" of tilt in the base surface will result in about .0002" of total indicator reading, meaning that the top of the cylinder square is off of center by about .0001".
It sure took a while, but now i have watched all your videos, so now i have started subscribing so i wont miss anything in the future. Keep up the good work, kind regards from Sweden.
Adam, I count five elements in your measuring setup. That means you have to be able to 'trust' five devices to be true... Reducing the amount of elements would help. It would lessen the amount of potential faults you introduce Paddy
Good video Adam, thanks. Dimensional Metrology is endlessly fascinating, isn't it? What I enjoy most about it is that eventually you need to account for the material properties of the tools and standards that you are dealing with be it steel, cast iron or granite.
I see you doing measurements with the tenth indicator and I'm instantly thinking of your surface grinder. I would like to see this in action in the near future.
Ours at work (tool and die shop) is within .0001". Years ago one of my co-workers got tired of cleaning metal particles off the magnetic end and grabbed the de-magnetizer. A long period of profanity from all those who worked in the shop ensued. We had to de-embed the magnet, order a new (and stronger) magnet from McMastrr-Carr, re-embed and re-squarify the thing.Good times!
As a sanity check spin it in Vee blocks with a good end stop to set the longitudinal position or between centers. Then indicate the base. Doing it this way will minimise the impact of surface plate flatness and the need to move the indicator around. Keeping the square as close as possible to the surface plate will also minimise any magnification of the surface error by the height gauge. ie Abbe's error.
Hey Adam: Just for your information, I've had a number (maybe 4 or 5) of what appear to be the exact same brand of resin spring clamp fail on me. They don't give any warning at all, they just snap. I don't use them anymore for much more than holding up a painting tarp.
Adam. Love your video's. Ive wached all of them and always look forward to the next. To measure this cylindrical square, eliminate the need to move the height gauge and this will eliminate the error you may have in your surface plate. Turn your entire setup 90 degrees so the narow shaft is pointing up perpendicular to your surface plate. Remove your angle plate and use a static point on your surface plate. Use your vee block on its side to reference the large diameter of your cylindrical square and palce your height gauge and dial in place and dont move the height gauge or the vee block. Especially that you suspect error on your plate. To take a measurement, simply rotate the cylingrical square keeping it referenced against the vee block. Hope this helps. Chich2
Well, as you said, I think that you have proven your suspicions about the stone not being plat is correct. But look on the bright side, you are learning different ways to check this thing. Rich
Plain (vertical standing) cylinder squares are used to measure perpendicularity of a vertical planar or cylindrical surface, tied to a specific height from a horizontal reference plane, such as a surface plate. Flanged versions, as Adam has, will do the same operation, indirectly, but the type of optimal indicator used is not the same. Also, flanged cylinder squares often have a magnetic base (the larger diameter) flange. Adam's flanged unit doesn't appear to be a magnetic version. Stefan's home-shop unit works best with an absolute-measurement drop-spindle DTI, spindle being mounted parallel to the surface plate, calibrated using the 180-degree rotation method against a plain vertical squaring cylinder. The classic version of these setups is from Taft-Pierce, available through Suburban Tools. (Don Bailey has shown a good video on this subject, using both types of setup equipment.) The flanged version is secured to a vertical plane (the 90 degree angle block), then, usually, a test comparator such as the Interapid is run over both peak ends of the skinny cylinder, noting the differences in height at each end. A good reading would be 0.0001" deviation per 6" of horizontal 'skinny' cylinder length. Also, the Interapid probe should be used at 12 degrees to the surface plate to eliminate cosine error corrections. Other comparator-DTI units would be used at zero degrees (parallel) to the surface plate. There is no requirement that a flanged cylinder square be built with both cylinders having the same centerline. Only the skinny cylinder and how perpendicular it is to the larger flanged base counts.
Hi Adam. I like those squares. I used a magnetic cylindrical square for years and loved it. One suggestion I have for your setup is to use 2 clamps. One on either side of the base as oppose to just one on top. It may yield better results.
I was hoping to get a magnetic square but this one come up for a real good price so I jumped on it. One problem I had the first time inspecting it was the clamps were moving it around on the sides, I think it's because of the angle plate though. . It's not exactly a perfect set-up for checking it but we'll improve as we move forward.
Looks like you will have plenty of grinding and scraping projects to keep you busy with all the older setup blocks and such around there. Lots of "apprentice" marks on them, for sure.
as its consistentl high towards the end, if you check the stem for diameter with your best micrometer and its the same both ends, it can only be your mounting bracket that is out, not the square, they all show about 10 positive at the end
Your cylindrical Square should be self checking without any need for any complicated setup using angle plate and clamp etc. That is, if you check with just the square itself, and using a Surface Gage with a radial "foot" stop like a ball bearing or radius of some kind, in order to get a single point of contact up against the cylindrical square`s own radial surface and checked with a mounted indicator at that "stop point" just touching the square, set your zero and rotate the square to the opposite side 180 degrees apart. If you get the exact same zero reference as before, then your cylindrical square is indeed SQUARE or perpendicular to the surface plate it has been checked and rotated on! Any deviation on your indicator will tell you how much out of perpendicular your cylindrical square is compared to the first side checked. Divide by 1/2 and that is the tolerance + or - your cylindrical square is over the actual height distance measured.
If you don't already, you should use plate conditioner, it just seals and lightly lubes the surface to reduce wear, a dab 'll do. You might also check around for on-site plate calibration, it may be about the same cost for them to come to you as it is for you to ship the plate and then it is lapped in on your support points and at your normal shop temperature. (maintain temp steady for 40+ hours before the appointment) Plate makers will calibrate other brands so if you live near a Starrett shop they will calibrate a Mitutoyo plate and vice versa. They use the same tools in the field and lab; optical auto-collimator for the overall flatness, a widget with a dial indicator for small dips bumps and waves, and an iron hand lapping plate for adjustments.
Adam, something you can try is to put the square on the table with its base against the base of the test indicator stand and rotate it against the test indicator probe, if that makes sense
A better way to check that particular type of square, is the same way it was made. Put it between centers.. check to make sure there''s no run-out on the shaft, then check for run-out on the face (bottom)
very interesting video Adam . I enjoy this depth of information. for a time I had worked as a precious metals assayer at a one of a kind lab. we had a granite pedestal that was perfectly level and required for the micro gram scales we used. they had to be certified yearly by a specialist that brought very precise standards. for our analytical gear we would buy standards. with all that machinery and precision expectations for your work, how do you know for certain whether your gear is in spec ?
From your numbers and measurements in the video, it looks that your angle plate is rather off square (they all have a bias roughly in the same direction / amount). To check a cyl square, you should use a vertical comparator, base against a fixed point, dial gauge at the top, take a reading, rotate (they cylinder) 180 deg, if identical readings, then you're close to the mark.
Maybe also try a high-end meter gauge, e.g. a new and certified one from Mitutoyo. I had some issues with repeatability in the gauge itself with other brands, which also had some age to them already.
G'day Adam, nice job on the setup. You may not have covered it in your video, but you should first check the square is both round and has no taper. You should be able to do away with that setup however and self-prove the square. That's one of the advantages of cylindrical squares. It should be a more accurate and repeatable way to test this.
wouldnt it be best checking it with a square comparator. it would eliminate the angle plate and v block and remove them from any error. most height gages can be used by using a ball bearing with it i think.
So a question....seems to me standing it up and measuring in vertical would be a better indication of if it is consistent. Feel like the set up is measuring the angle plate and vee block.
Maybe you could have done that on the lathe: 1) check the squareness of the cross slide to the chuck using the "rollie dad" method 2) chuck the cylindrical square and move the indicator along the its flange using the cross slide
I can understand playing with the tools, I do it myself in my garage. However, you are using at least 4 tools of unknown accuracy. You can't have any idea the accuracy of your cylindrical square if 1 tool is known to be off. "Pretty dainged square" just doesn't cut it. Still, though, I learned a few things. I look forward to you getting the granite plate, the height gauge foot & the angle plate ground & calibrated & the test redone. What fun! Keep it up.
Hi Adam thanks for the video. Just a few thoughts. Is that cyl' square unusual in having a flange on the end? As the base has holes in it do you think it may have been torqued down and slightly distorted? It looks to be a difficult item to check for square unless you have some granite plates and squares like Stan Z. I would have thought the plastic clamp was not be man enough to hold the weight tight against the angle plate. Or it could be dead on square but your old cast angle plate is a little tired. regards from the UK
Adam interesting. Did I see a center in the end of the cyndrical square? is there a corresponding center on the bottom. If so could you set it between dead centers and measurer run our on the opposing 90 Dec surface. ? liked the vid Tks Will
Hi Adam, fascinating stuff, even if it is way over my head. Do you have to take into account variations in air temperature, or doesn't that matter? cheers peter.
So since the 1 and 4 quadrants are adjacent and .0003 to .0005 different at the top of the shaft relative to the more consistent base measurements, this would indicate a very slight lean in the shaft and that the angle plate setup gives it a roughly .0013 upwards tilt?
Yet another great vid Adam. OK I have a question for ya, maybe considered personal so if you don't answer no worries, but why the shaved arms? I have another friend that does this too and he is also a really big guy like yourself so I'm pretty sure you guys aren't doing to achieve faster laps at the pool lol . I've asked him many times before why he does it, not making fun or in a condescending way and he will never give me a straight answer. I'm honestly just curious, maybe there's a secret benefit and I'll have to try to figure it out.
Adam, there's a couple things that could be contributing to your repeatability issues: One is the granite flatness as you mentioned Angle plate condition is another Also the fact that you are clocking the square by hand means that you're probably not returning to the same exact position every time. You're also relying on the OD of the flange to be 1) round and 2) concentric to the shaft of the square since that is what is riding on the v-block as your rotate it. I would also suggest readjusting your indicator tip to be as parallel to the surface you're measuring as possible. Having it at an angle like that introduces cosine error and falsely increases the value of the readings compared to what they actually are. www.mmsonline.com/columns/correcting-for-cosine-error
The Interapid brand indicator is designed to be used at 12 deg, rather than parallel like most other indicators, so you can get over obstacles. In either case though, this was a comparative measurement to see how consistent the cylinder was, so it probably wouldn't matter much.
They are for checking square on a surface plate with an indicator (i.e. to actually measure how out of square you are.) I don't think you would ever use one as a test bar in a lathe. They are self proving insofar as if they are truly square to the base then you can clock them 180 degrees and the readings should not change when measuring at a specific location or distance from the base, even if the angle plate was not square.
+J Berkey It's used for inspection work, to check squareness of parts. You can stick it on an angle plate and use an indicator to sweep across it at different points to see how square the angle plate is to the bottom surface.
I had to chuckle, Adam. I see you measuring to thousandths or ten thousandths in all of your videos, then you say you haven't really done the "precision" measuring.
Also remember that you weren't completely square to the work with your indicator and the indicator needle wasn't as perpendicular to the piece as it could be so you've introduced a bit or error there. You may be closer than you think.
2 microns = .7 "tenths", and he was using the terms "half-tenth's" and "tenths" in the video. So actually you are using the same scale of measurements as he was, just in the metric system instead of inches. Written another way, .002 mm = .000070 inch, or 2 microns = 70 micro-inches.
Abom, keep in mind that you are doubling your error by rotating the square. It is really only off half of the total indicator reading. That would be within 50/millionths or half a "tenth". Not bad at all. There is no such thing as perfect.
This is absolutely NOT how to check that type of cylinder square. Its a simple check between centers. Diameter and face should run true spun between centers. If true then the square is as square as 1/2 the taper in the diameter. If the diameter has a .0001 taper then its square within 50 millionths.
Until the surface plate is checked all of this isn't really useful. You could try repeating the measurements at various different location around the plate but even that is of limited value. Thanks for making the effort and posting the video.
![[UNCUT] ลูกหมี-ลิลลี่ เหงียน เปิดใจฟาด ปู มัณฑนา I คนดังนั่งเคลียร์ I 15 ก.ค.67](http://i.ytimg.com/vi/FKQ60Y7bJ6s/mqdefault.jpg)
![fellow fellow - Proud [OFFICIAL MV]](http://i.ytimg.com/vi/9uZF7BlUQPE/mqdefault.jpg)
Enjoyed! How much time passed between your two measurements? A consistent 0.0001" all over sounds suspicious. Wonder if there was a temperature change? I think a 5 degree change, in steel, = 0.0001" over an inch.
Nice square, by the way.
Trust me Tony, I'm not a square......bad joke?
Glad you enjoyed and thanks for stopping by! I don't recall the temps being much different during the two days but 5 degrees difference it could have very well been. I think what I have is a "very close" square, good enough for government work right?
not bad, not bad.. I wouldn't close the shop up just yet.. but not bad. ;)
Yes. Just taking a stab at that constant 0.0001". Though he did mention the zeroing thing.. likely the more probably cause.
Carbon steels have a thermal coefficient of expansion about 11ppm/deg C or 6ppm/F. So 1/100,000 inch per degree C per inch; 0.000 006 inch per inch per F
psec.uchicago.edu/thermal_coefficients/cte_metals_05517-90143.pdf
The biggest effect would be warpage of the surface plate due to an internal thermal gradient if the shop temp fluctuated over just a few hours (overnight). The larger thicker plates are stabilized for several days before lapping or calibration, depending on the desired rating of course, AAA lab grade has less wiggle room in the specs than a class B shop plate (though they both warp the same amount)
I was going to ask if it was cold!
Love learning with you Abom! You have inspired me to start machining, bought myself an American Pacemaker! I really like hearing stories about your Dad & Granddad, keep them coming! Thanks!
I just love it when you talk metrology! Thanks again, Adam.
Stefan Gotteswinter has made a rather nice squareness comparator, one I plan to use for inspiration for exactly this kind of measuring. Looks like metrology is something all the cool kids are doing these days! Thanks heaps for sharing
I plan to build a reference square like that soon. Very cool!
A good method for checking square on that cylinder is as follows:
Use a 1" or 3/4" OD bearing-ball in the V-notch of a surface gauge with your .0001" indicator to create a squareness comparator.
Place the cylindrical square on the surface plate and push the comparator-ball up against the large OD of the cylinder.
Adjust the test indicator so that it is testing the top of the smaller OD of the cylinder.
Zero the indicator to the "high" spot as you pivot the surface gauge at the bearing ball.
Rotate the cylindrical square around and keep finding/checking the high spot (should repeat to zero if it's concentric).
Use a sharpie to mark the difference from your zero on the square.
Now repeat the same procedure with the indicator testing the lowest point of the smaller OD of the cylinder.
Finally, use your best set of .0001" mics to verify that the small diameter is consistent (not tapered).
These three tests will eliminate the V-block setup and the errors it can introduce. Also, you'll get more from your granite because you're confining your test to a small area with very little movement, just rotation and minor pivots. You can check several points along the height of the cylinder, but the most accuracy will be from very near the top and bottom ends.
There are better ways to build a squareness comparator, that's just the low effort way and it sure works well.
Excellent post, cwtoyota. Another thing is that a cylinder with a slightly tilted but flat base has very near perfect accuracy at two clock positions. Test it as suggested above, and mark the two places that are vertically parallel on the cylinder square. For fussy work, just use those two positions only. The base can also be lapped to get the cylinder perfectly vertical, but flatness can be compromised (or repaired) in the process, that is a fussy job in itself...
Thanks bob. That's a good point and a good way to use an imperfect tool to do higher quality work.
Those two clock positions should be exactly 180° apart if tilt is the cause.
On a cylinder of those proportions (height is greater than the width of the base) a small dust particle, or small grinding imperfection in the base makes a large indicator reading at the top. If the height is 2 times the width of the base (2:1 ratio) .00005" of tilt in the base surface will result in about .0002" of total indicator reading, meaning that the top of the cylinder square is off of center by about .0001".
Ultra fine measurement, of other measurement devices is always interesting. Must have when accuracy counts. Thanks Adam.
It sure took a while, but now i have watched all your videos, so now i have started subscribing so i wont miss anything in the future.
Keep up the good work, kind regards from Sweden.
Adam,
I count five elements in your measuring setup. That means you have to be able to 'trust' five devices to be true...
Reducing the amount of elements would help. It would lessen the amount of potential faults you introduce
Paddy
In early makes a happy TH-camr! Nice job Adam! I was watching your videos with my ex-Navy-machinist father. He enjoys it.
Good video Adam, thanks. Dimensional Metrology is endlessly fascinating, isn't it? What I enjoy most about it is that eventually you need to account for the material properties of the tools and standards that you are dealing with be it steel, cast iron or granite.
I see you doing measurements with the tenth indicator and I'm instantly thinking of your surface grinder. I would like to see this in action in the near future.
Ours at work (tool and die shop) is within .0001". Years ago one of my co-workers got tired of cleaning metal particles off the magnetic end and grabbed the de-magnetizer. A long period of profanity from all those who worked in the shop ensued. We had to de-embed the magnet, order a new (and stronger) magnet from McMastrr-Carr, re-embed and re-squarify the thing.Good times!
Great job Abom
Most interesting Adam -- not something I had seen before.
As a sanity check spin it in Vee blocks with a good end stop to set the longitudinal position or between centers. Then indicate the base. Doing it this way will minimise the impact of surface plate flatness and the need to move the indicator around. Keeping the square as close as possible to the surface plate will also minimise any magnification of the surface error by the height gauge. ie Abbe's error.
Hey Adam: Just for your information, I've had a number (maybe 4 or 5) of what appear to be the exact same brand of resin spring clamp fail on me. They don't give any warning at all, they just snap. I don't use them anymore for much more than holding up a painting tarp.
Adam. Love your video's. Ive wached all of them and always look forward to the next. To measure this cylindrical square, eliminate the need to move the height gauge and this will eliminate the error you may have in your surface plate. Turn your entire setup 90 degrees so the narow shaft is pointing up perpendicular to your surface plate. Remove your angle plate and use a static point on your surface plate. Use your vee block on its side to reference the large diameter of your cylindrical square and palce your height gauge and dial in place and dont move the height gauge or the vee block. Especially that you suspect error on your plate. To take a measurement, simply rotate the cylingrical square keeping it referenced against the vee block. Hope this helps. Chich2
Well, as you said, I think that you have proven your suspicions about the stone not being plat is correct. But look on the bright side, you are learning different ways to check this thing.
Rich
Plain (vertical standing) cylinder squares are used to measure perpendicularity of a vertical planar or cylindrical surface, tied to a specific height from a horizontal reference plane, such as a surface plate. Flanged versions, as Adam has, will do the same operation, indirectly, but the type of optimal indicator used is not the same. Also, flanged cylinder squares often have a magnetic base (the larger diameter) flange. Adam's flanged unit doesn't appear to be a magnetic version.
Stefan's home-shop unit works best with an absolute-measurement drop-spindle DTI, spindle being mounted parallel to the surface plate, calibrated using the 180-degree rotation method against a plain vertical squaring cylinder. The classic version of these setups is from Taft-Pierce, available through Suburban Tools. (Don Bailey has shown a good video on this subject, using both types of setup equipment.)
The flanged version is secured to a vertical plane (the 90 degree angle block), then, usually, a test comparator such as the Interapid is run over both peak ends of the skinny cylinder, noting the differences in height at each end. A good reading would be 0.0001" deviation per 6" of horizontal 'skinny' cylinder length. Also, the Interapid probe should be used at 12 degrees to the surface plate to eliminate cosine error corrections. Other comparator-DTI units would be used at zero degrees (parallel) to the surface plate.
There is no requirement that a flanged cylinder square be built with both cylinders having the same centerline. Only the skinny cylinder and how perpendicular it is to the larger flanged base counts.
Good stuff sir!
would be interesting to check every inch or so and see if you can isolate a specific area that set things off
Thank You Adam
Hi Adam. I like those squares. I used a magnetic cylindrical square for years and loved it. One suggestion I have for your setup is to use 2 clamps. One on either side of the base as oppose to just one on top. It may yield better results.
I was hoping to get a magnetic square but this one come up for a real good price so I jumped on it. One problem I had the first time inspecting it was the clamps were moving it around on the sides, I think it's because of the angle plate though. . It's not exactly a perfect set-up for checking it but we'll improve as we move forward.
Looks like you will have plenty of grinding and scraping projects to keep you busy with all the older setup blocks and such around there. Lots of "apprentice" marks on them, for sure.
Thanks, Abom!
as its consistentl high towards the end, if you check the stem for diameter with your best micrometer and its the same both ends, it can only be your mounting bracket that is out, not the square, they all show about 10 positive at the end
Hey Adam. mount your super spacer to your doall and chuck that part up and run your indicator on the shaft see what kinda readings you get.
Your cylindrical Square should be self checking without any need for any complicated setup using angle plate and clamp etc.
That is, if you check with just the square itself, and using a Surface Gage with a radial "foot" stop like a ball bearing or radius of some kind, in order to get a single point of contact up against the cylindrical square`s own radial surface and checked with a mounted indicator at that "stop point" just touching the square, set your zero and rotate the square to the opposite side 180 degrees apart.
If you get the exact same zero reference as before, then your cylindrical square is indeed SQUARE or perpendicular to the surface plate it has been checked and rotated on! Any deviation on your indicator will tell you how much out of perpendicular your cylindrical square is compared to the first side checked. Divide by 1/2 and that is the tolerance + or - your cylindrical square is over the actual height distance measured.
Adam. I'm thinking sweep the v-block first. If it's perpendicular to the surface plate, then place the cyl square in the v-block, then test.
Thank you
If you don't already, you should use plate conditioner, it just seals and lightly lubes the surface to reduce wear, a dab 'll do.
You might also check around for on-site plate calibration, it may be about the same cost for them to come to you as it is for you to ship the plate and then it is lapped in on your support points and at your normal shop temperature. (maintain temp steady for 40+ hours before the appointment) Plate makers will calibrate other brands so if you live near a Starrett shop they will calibrate a Mitutoyo plate and vice versa.
They use the same tools in the field and lab; optical auto-collimator for the overall flatness, a widget with a dial indicator for small dips bumps and waves, and an iron hand lapping plate for adjustments.
hey adam this stuff will make you nuts. chasing millionths is a job for the qc dept in their temp and humidity controlled rooms.
Adam, something you can try is to put the square on the table with its base against the base of the test indicator stand and rotate it against the test indicator probe, if that makes sense
A better way to check that particular type of square, is the same way it was made.
Put it between centers.. check to make sure there''s no run-out on the shaft, then check for run-out on the face (bottom)
I'd like to see how the cylindrical square is used. We didn't come near one in bricklaying. Greg
After having watched Tom Liptons lapped vee block yours almost made me cry ;-;
Great video though ^^
very interesting video Adam . I enjoy this depth of information. for a time I had worked as a precious metals assayer at a one of a kind lab. we had a granite pedestal that was perfectly level and required for the micro gram scales we used. they had to be certified yearly by a specialist that brought very precise standards. for our analytical gear we would buy standards. with all that machinery and precision expectations for your work, how do you know for certain whether your gear is in spec ?
From your numbers and measurements in the video, it looks that your angle plate is rather off square (they all have a bias roughly in the same direction / amount). To check a cyl square, you should use a vertical comparator, base against a fixed point, dial gauge at the top, take a reading, rotate (they cylinder) 180 deg, if identical readings, then you're close to the mark.
Maybe also try a high-end meter gauge, e.g. a new and certified one from Mitutoyo. I had some issues with repeatability in the gauge itself with other brands, which also had some age to them already.
G'day Adam, nice job on the setup. You may not have covered it in your video, but you should first check the square is both round and has no taper.
You should be able to do away with that setup however and self-prove the square. That's one of the advantages of cylindrical squares. It should be a more accurate and repeatable way to test this.
wouldnt it be best checking it with a square comparator. it would eliminate the angle plate and v block and remove them from any error. most height gages can be used by using a ball bearing with it i think.
Very interesting video however what would you build in your shop and with your equipment and need to be that precise?
So a question....seems to me standing it up and measuring in vertical would be a better indication of if it is consistent. Feel like the set up is measuring the angle plate and vee block.
Maybe you could have done that on the lathe:
1) check the squareness of the cross slide to the chuck using the "rollie dad" method
2) chuck the cylindrical square and move the indicator along the its flange using the cross slide
yes, if you turn it between centres, you use your dial guage on the stand face, any runout will be your 'out of squareness'
I can understand playing with the tools, I do it myself in my garage. However, you are using at least 4 tools of unknown accuracy. You can't have any idea the accuracy of your cylindrical square if 1 tool is known to be off. "Pretty dainged square" just doesn't cut it. Still, though, I learned a few things. I look forward to you getting the granite plate, the height gauge foot & the angle plate ground & calibrated & the test redone. What fun! Keep it up.
Hi Adam thanks for the video. Just a few thoughts. Is that cyl' square unusual in having a flange on the end? As the base has holes in it do you think it may have been torqued down and slightly distorted? It looks to be a difficult item to check for square unless you have some granite plates and squares like Stan Z. I would have thought the plastic clamp was not be man enough to hold the weight tight against the angle plate. Or it could be dead on square but your old cast angle plate is a little tired. regards from the UK
Adam interesting. Did I see a center in the end of the cyndrical square? is there a corresponding center on the bottom. If so could you set it between dead centers and measurer run our on the opposing 90 Dec surface. ?
liked the vid
Tks Will
or have a jig made with a laser on it to sight the long axis for true ? then you could slowly rotate the cylinder and check the entire circumference ?
Hi Adam, fascinating stuff, even if it is way over my head. Do you have to take into account variations in air temperature, or doesn't that matter? cheers peter.
Where do you send that granite to get it flattened? Ive got a pool table and i need one of the slates ground flat
Remember the gravity also!
So since the 1 and 4 quadrants are adjacent and .0003 to .0005 different at the top of the shaft relative to the more consistent base measurements, this would indicate a very slight lean in the shaft and that the angle plate setup gives it a roughly .0013 upwards tilt?
Couldn't you dial in 0 on your 4 jaw on the small diameter and then sweep the larger outer?
Just a thought...
I have no idea.
Yet another great vid Adam.
OK I have a question for ya, maybe considered personal so if you don't answer no worries, but why the shaved arms? I have another friend that does this too and he is also a really big guy like yourself so I'm pretty sure you guys aren't doing to achieve faster laps at the pool lol . I've asked him many times before why he does it, not making fun or in a condescending way and he will never give me a straight answer. I'm honestly just curious, maybe there's a secret benefit and I'll have to try to figure it out.
Do you think that the difference is caused by temperature? I know a lot of inspection rooms where .0001 is critical are temperature controlled.
I think you should have the Stannet calibration llabs check it when you are out here.
RichardTheDick7 No, Stannet, check out the banner in the back of Bar-Z's shop...
What if you take the cylinder and put it in between sentences in your lathe and take out the possibility of error in the setup.
Curious if this was recorded after scraping in your angle plate from the class?
How do you use a square that has a hub on the bottom?
Are you taking temperature into allowance? As you have seen temperature can effect things. Temperature MUST be a constant.
Adam, there's a couple things that could be contributing to your repeatability issues:
One is the granite flatness as you mentioned
Angle plate condition is another
Also the fact that you are clocking the square by hand means that you're probably not returning to the same exact position every time. You're also relying on the OD of the flange to be 1) round and 2) concentric to the shaft of the square since that is what is riding on the v-block as your rotate it.
I would also suggest readjusting your indicator tip to be as parallel to the surface you're measuring as possible. Having it at an angle like that introduces cosine error and falsely increases the value of the readings compared to what they actually are. www.mmsonline.com/columns/correcting-for-cosine-error
The Interapid brand indicator is designed to be used at 12 deg, rather than parallel like most other indicators, so you can get over obstacles. In either case though, this was a comparative measurement to see how consistent the cylinder was, so it probably wouldn't matter much.
checking a circle for squareness. now this is precision machining.
i've never seen one of these before, what are they used for?
checking a collect lathe , if it has centers in it, u can use it to set ur tailstock , blah blah blah :)
They are for checking square on a surface plate with an indicator (i.e. to actually measure how out of square you are.) I don't think you would ever use one as a test bar in a lathe. They are self proving insofar as if they are truly square to the base then you can clock them 180 degrees and the readings should not change when measuring at a specific location or distance from the base, even if the angle plate was not square.
I am not understanding how you can get repeatability when the cylinder is being rotated and re clamped when it is in a horizontal position?
looking at your numbers it appears that the square is tapered.did you check the od at both ends
New subscriber to your channel and the discipline. What is a cylindrical square used for? Do you have a video in which you put it to use?
+J Berkey It's used for inspection work, to check squareness of parts. You can stick it on an angle plate and use an indicator to sweep across it at different points to see how square the angle plate is to the bottom surface.
I had to chuckle, Adam. I see you measuring to thousandths or ten thousandths in all of your videos, then you say you haven't really done the "precision" measuring.
have you considered curvature of the earth? that changes everything.
waiting to see how flat stone is just curious
Also remember that you weren't completely square to the work with your indicator and the indicator needle wasn't as perpendicular to the piece as it could be so you've introduced a bit or error there. You may be closer than you think.
haha we just though one of them away in the chip bin because it was .00002 Out of Tolerance should have sent it your way.
What is that square used for?
It's funny to see in witch Tolerances he's woking, when you work in like 2-8 1/1000 mm ^^
2 microns = .7 "tenths", and he was using the terms "half-tenth's" and "tenths" in the video. So actually you are using the same scale of measurements as he was, just in the metric system instead of inches. Written another way, .002 mm = .000070 inch, or 2 microns = 70 micro-inches.
unless that square u bought or acquired was a 1500$ , it has a tolerance of out of round and out of square too no need fer a tenths indicator. :P
new subscriber who doesn't know anything. But, consider the uses of laser lighting.
perhaps the cylinder sq was dropped?
Abom, keep in mind that you are doubling your error by rotating the square. It is really only off half of the total indicator reading. That would be within 50/millionths or half a "tenth". Not bad at all. There is no such thing as perfect.
FIFTH!
Abom
Adam Booth, I want to have your babies....metaphorically speaking.
RichardTheDick7 Not so. I find comments like yours quite abhorrent, whilst the first comment was humorous and respectful.
This is absolutely NOT how to check that type of cylinder square. Its a simple check between centers. Diameter and face should run true spun between centers. If true then the square is as square as 1/2 the taper in the diameter. If the diameter has a .0001 taper then its square within 50 millionths.
Until the surface plate is checked all of this isn't really useful. You could try repeating the measurements at various different location around the plate but even that is of limited value. Thanks for making the effort and posting the video.