High accuracy angle measurement

[robert123]

I'm looking at a part that is about 3/8" diameter with a cone on one end. The cone does not come to a point instead it is cut off with a flat. The angle between the cone and the flat nose is required to be +/- 10' (minutes). Is there a protractor or other device to measure this with enough accuracy or is an optical comparator needed?

 

[Booze Daily]

I wouldn't trust a comparator to that tolerance unless is was really high magnification.
I'd probably use a sine plate.

 

[michiganbuck]

likely looking for about .0015 in a 3/8 length so pretty tough to use a hand protractor.
Sometimes one can set the angle flat on something and then test the long aspect of a part with a sine bar or an optical comparator.
But a half degree is 30 minutes, 1/4 degree 15 minutes... and half is about the spec on a non-specified tool bit or cutter and is checked with the eyeball and a good optical comparator, protractor, indicator, and sine bar, template ..or a fish gauge.

if you made cutters 1/4 degree off you would lose some customers.

 

[robert123]

How would one go about measuring the angle using a sine plate?

 

[just Dave]

What are you making the cut with? CNC, taper attachment, compounding it. I have good answers for a couple of those.

 

[DanielG]

The other big question to consider is how is the customer going to measure it?

 

[modelmakerblue]

Just set up your compound slide using trigonometry.
Tony

 

[CarbideBob]

How would one go about measuring the angle using a sine plate?

Place the part in a know perfect vee block perfectly square on the plate. No dust or lint.
Set the sine and run a indicator/surface gauge across it. Just inside the tip and just inside the exit.
This tells you the cone is spot on or not.
Now you have to worry if the flat is indeed square to the shank. Here instead of a thou or more you have a tenth or less to play with.
We would use a 100x optical compartors or 500x microscopes but that is expensive shit.

More that once I get into long arguments when my customer's receiving inspection has a CMM checking the parts.
One certainly hesitates to ask the customer how they will check as this makes you seem unsure about what you make.
Regardless of which measuring, take it out, do it again three times. Are the numbers the same within some decent range?

I do not know the cone angle or the flat size, turned or ground part, surface finish.....
Bob

 

[michiganbuck]

You might make a master gauge using a sine bar and a surface grinder that should easily be within a tenth or two (likely <50 millionths), then set up a magnifying camera over a fixed V block, and with a simple computer and a monitor get your inspection down to a few millionths. Plenty close for 10 minutes.

Amazon.com : computer microscope usb

looking at a feature at 1000x you should be able to detect +/- 1 minute.

A description of the part might help PM guys to give answers. Is the cone area on the end of a shaft, is the cone about 3/8 long?

 

[robert123]

It would be cut on a CNC lathe (with live tooling for other features).

I drew the part in CAD and the tapered length is only about .100" (overall length is 3/8"). So it looks something like the head of a 3/8" countersink with the nose blunted and no flutes. The included angle is about 90 degrees. Since it would be cut on a lathe, I think it would be fine to measure either the included angle or just one side.

 

[michiganbuck]

You might have a gauge shop (or in-house) make a couple of gauges that would be like a lathe fish gauge having an included angle that one was plus, and the other minus 8 minutes per side (or +- 5M). for lathe adjustments, and the magnifying camera set up on the bench so you might make an exact read-out/certification.

90* +- 10 min?

 

[greif1]

I like the idea of making go and no go gauges. Trying to make direct measurement on such short angle leg lengths (to use trig) will result in huge error.

 

[michiganbuck]

I like the idea of making go and no go gauges. Trying to make direct measurement on such short angle leg lengths (to use trig) will result in huge error.

and 10m over .100 length is about .0017 so pretty easy to eyeball.

 

[guythatbrews]

If the include angle is about 90* and .1 long, the normal movement at one end of the taper is around .0004 for 10'. A sine bar will do it but holding the part is a problem. A 50x comparator will do it, 100x better. The "flat end" not being very, very flat is a problem.

If any measurements are shaky I always contact the customer and see how they inspect it. If they are worth their salt they'll work cooperatively to make sure they get good parts. Sometimes they have to scramble to answer, and not all are worth their salt! And they have also found errors on their drawing. Didn't have to be that close.

 

[CarbideBob]

Amazon.com : computer microscope usb

looking at a feature at 1000x you should be able to detect +/- 1 minute.

If only this would work at this price and this accuracy.
It is rise over run and number of pixels/3. This ignores the lens and those pin/barrel problems plus the part is round.
The plastic lens on these guys are not flat field by any stretch of the imagination. Yes I do use them and super handy.
Turns out this a tad hard to do.
High mag is of no help since you have poop for view size. Lower mag better and you do best fit line not two points.
I for sure get how it makes common sense in video or machine vision. But, but, but.... so many buts.
At this surface finish matters. A lathe will make cusp and forms even on the cut off end. Do we do tips, valleys or middles

Not to say this is hard to make or hard to check. Plus zero minus 15 is sort of standard to me so we set at 7.5 as close as we can and that alone is problem on a simple carbide insert.
Welcome to WTF in measuring. Not once but 6 times unbiased or diffenrt people on differnt days making the measurement.

The customer is always right but they may check it one way and you check it different.
Right, wrong or off .....Make those who send the check to be cashed so happy campers that they will come back for more.

 

[michiganbuck]

I agree with CarbideBob on the limited merits of anything other than proper measuring devices.
A jury rig measuring device would be just that, you would have to also jury rig the measuring with a gauge that could also come in focus while the part is in focus. For some parts, like with compound angles you have to roll the part in place when measuring. and the Ops part likely one would need to have the cone edge in focus.

TOMLOV DM201 Pro 7" Digital Microscope 1200X Magnifier Video Recorder 1080P 32GB | eBay

Easy to Set up and Use -- User-friendly interface design makes video microscope easy-to-use for all ages. Video Microscope 1pcs. HD Cable 1pcs ( optional ). USB Cable 1pcs. Video resolution: 1080P30 (clear), 720P60 (smooth).

www.ebay.com

* A 100x or greater optical comparator would be OK , the 100X would make .ooo1 look like .010 so would be ok.

*Being handy on a decent surface grinder one could make a fixture that would hold a part at the correct angle on a surface plate and measure the angle to .0001 or so, one might even make it adjustable for limited various parts and gauges. likely gauges would also have to be made or bought.

Fortunately, I have almost always had a comparator and/or a CMM at my disposal so getting tenths and better was not a problem.
In my shop I get to 12 millionths or so in flat work , and I avoid close angular and radius parts. I avoid tenths work when I can.

 

[michiganbuck]

How would one go about measuring the angle using a sine plate?

You would set the sine on a surface plate to the angel and have a precision V block to set on the sine..and strike an indicator across the part..
for a part with a compound angle, like a cutting tool, the part would bump a stop at the front or rear..and be rolled ti check the high place of the cutting edge.

That would be a plate check so not too good for setting the machine
With having a surface grinder best to make a V block sine.

You might make one part with a full-to-a-point precision angle to be used as a machine set-up gauge.

 

[Conrad Hoffman]

You said the angle between the cone and the flat. How good is the cone to the body? How good is the surface of the flat and what diameter is it? There are several things that can add up here and, IMO, it might not be so simple to measure. Comparator seems like the best bet.

 

[just Dave]

If the include angle is about 90* and .1 long, the normal movement at one end of the taper is around .0004 for 10'. A sine bar will do it but holding the part is a problem. A 50x comparator will do it, 100x better. The "flat end" not being very, very flat is a problem.

If any measurements are shaky I always contact the customer and see how they inspect it. If they are worth their salt they'll work cooperatively to make sure they get good parts. Sometimes they have to scramble to answer, and not all are worth their salt! And they have also found errors on their drawing. Didn't have to be that close.

Another thing Ive seen is a lack of surface finish to support an extreme tolerance requirement .

 

[eKretz]

The depth of field goes right to shit on very high magnifications, which is probably why they aren't often used. There are new video mag systems these days that automagically focus stack but they are, umm, not cheap, to put it mildly. Comparator maybe; might also be a good candidate for a toolmakers microscope.