How To Measure Large Diameters?

[sanjiv]

Recently i was asked to make 1100mm dia special flange with accuracy of plus minus 0.1 mm. How this could be measured as i have only a digital micrometer ranging 0 to 25 mm. i can not purchase a new micrometer for measuring 1100mm because iot is too expensive. can you give some idea?

Similar Threads:

• Solved Wrap on Multiple Diameters
• Depth Micrometer to measure outside diameters-good practice?
• Need Help!- Cutting diameters offcentre
• Different offsets for different diameters?
• large diameters

[Geof]

Just Google 'diameter tape'. These are tapes that are calibrated to measure diameter when they are wrapped around the circumference of the part. Each diameter inch on the tape is 3.1416 inches long.

[mrainey]

That's really interesting. I had never heard of them. My first thought - "How accurate could they be"?

This is the claim at www.pitape.com


Our precision diameter tapes are manufactured to exacting standards in temperature controlled facilities at 68°F (20°C). The lines and numerals on our standard tapes are engraved and acid etched on a ground surface. The tapes have fixed readings that do not require periodic adjustments.

Our standard Imperial tapes (inches) read to 0.001” diameter with an accuracy of ±0.001” up to 144 inches.

Our standard metric tapes (millimeters) read to 0.01mm diameter with an accuracy of ±0.03mm up to 3,600 millimeters.

Each Pi Tape brand precision measuring tape is delivered with a Calibration Report traceable to the National Institute of Standards and Technology and all calibrations are performed in accordance with ANSI/NCSL Z540-1-1994, ISO/IEC 17025, ISO 10012-1, MIL-STD 45662A and 10CFR, part 21.

[HuFlungDung]

1100mm? Heck, you should be able to measure across that with a laser and a stop watch

[RICHARD ZASTROW]

I used pi-tapes for years and they are as accurate as claimed IF you learn the proper technigues of using them. Temperature of workpieces and tape must be the same and calculate for a normal (20C 0r 68F) temp and compensate for different co-efficient of expansion.
If you're trying to hold .001" tolerances on large diameters this can be important. Also, keep the tape straight around the diameter or you will be measuring an elipse. I may be exagerating for 1100mm, but my workpieces were 12 (3600mm) to 15 ft.(4500mm) in diameter in the field. In the shop, workpieces were machined to 30 ft.(9100mm) in dia, and fabricated cylinders (gear blanks) to 50 ft. (15200mm)
Fortunately, I didn't have to do the measuring in the shop. There are other methods for large O.D. measuring. One is to measure from a known reference point outside of the swing of the lathe/boring mill to the O.D. of the workpiece with an inside mic.

[Caprirs]

Those are cool. Thanks for the link. I'd never heard of those before.

[Zumba]

Interesting... not so much the pi tape thing that's being discussed, but the job at hand.

If the machinist has a hard time measuring it, how will the customer? If the customer has the tools and/or know-how to measure it, how come he can't make the part himself? Also, what's the reasoning for the high accuracy on a "flange"? If there's a mating part, who made it and why can't that same person make the flange? How accurate is the mating part? How did the customer measure the mating part?

So many questions... not so many answers. :rainfro: :rainfro:

[lerman]

It measures the perimeter (and then divides by pi). If the object is round, that corresponds to the diameter. If you want to check the dimensions of your part, that should be fine if you can prove that it is round or otherwise check the roundness.

A micrometer can measure the part as it is rotated to a variety of positions. If you are turning the part on a lathe, you can pretty much guarantee that it will be round (within the accuracy of the bearings).

You can mount a micrometer head and anvil on a piece of channel and use it to check the roundness of your part. That, together with a pi tape should do a reasonable job.

[RICHARD ZASTROW]

Ken, A mic doesn't guarantee a round part either. Visualize a lobed part (like a 3 lobed cam). The distance between a low point and the opposite high point would be the same at three points but not round. If it was truly harmonic, I believe it would measure the same at any opposite point and not be round. The same assumption as lathe/mic/bearings would be true for lathe/pi tape/bearings. If roundness is critical, within reasonable tolerance, check the surface with a dial indicator. If that is insufficient, resort to Tallyrond type instument.

[Geof]

Try Googling 'constant diameter figure'.

When you make something on a lathe you more or less take it on faith that it is round.

[lerman]

Yeah,

I realized that as I was writing it, but didn't want to look up if the shape is called a trochoid. (Still haven't looked it up.) I should have figured that I'd get caught with this group. :-)

Ken

[martinw]

I may be missing something, but surely if you fix a pivot at the centre of the workpiece, fix a rigid rotatable bar to the pivot, and a vernier or micrometer to the other end of the bar, you should be able to measure the RADIUS at any point of the circumference??

On another point, how exactly do you read a pi-tape to an accuracy of 0.001 inches?

Best wishes

Martin

[RICHARD ZASTROW]

martinw, ref. second point. There is a widely spaced vernier scale at the measuring point of the tape. Google pitape, the pictures and info are self explanitory.

[martinw]

martinw, ref. second point. There is a widely spaced vernier scale at the measuring point of the tape. Google pitape, the pictures and info are self explanitory.

Dear Richard,

Thanks.

My mistake: I completely missed the vernier scale. Apologies to all.


Best wishes

Martin

[Geof]

I may be missing something, but surely if you fix a pivot at the centre of the workpiece, fix a rigid rotatable bar to the pivot, and a vernier or micrometer to the other end of the bar, you should be able to measure the RADIUS at any point of the circumference??

On another point, how exactly do you read a pi-tape to an accuracy of 0.001 inches?

Best wishes

Martin

It would not be necessary to fix the pivot at the center, yes? no? maybe?

Use a rotating bar with a dial indicator that was calibrated from the pivot point. When the pivot point was attached to the workpiece if it was slightly off center the radius would be the averagr of the largest and smallest readings.

[vger]

Another point about Pi tapes, each .001" change in diameter of the part will make a .0031416" change on the tape.

[JerryFlyGuy]

True Geof, but how would you know it's round in all the other quadrants? I'm sure a guy could make it as you went and figure it out. It might be easier to just get it centered up to start w/ I dunno. I've never done anything this big, but one of my first projects is going to be to grind a 48" disk so I might find out yet

Jerry

[Geof]

True Geof, but how would you know it's round in all the other quadrants? I'm sure a guy could make it as you went and figure it out. It might be easier to just get it centered up to start w/ I dunno. I've never done anything this big, but one of my first projects is going to be to grind a 48" disk so I might find out yet

Jerry

The sophisticated way would be to have a digital indicator with a readout to a computer and graph the output. If your part is a true circle the graph will be a sine curve and it would be possible to mirror it and superimpose it on itself perfectly.

The tedious way would be to simply note the dial reading at +/- the same number of degrees either side of a line joining the high and low spots. These will be the same for a circle.

[JerryFlyGuy]

Sorry, can you remind me which one was tedious? Wouldn't it just be easier to center the dang thing?

Jerry [Headin' your way on the 16th, gonna be around? I owe ya dinner don't I?]

[Geof]

Sorry, can you remind me which one was tedious? Wouldn't it just be easier to center the dang thing?

Jerry [Headin' your way on the 16th, gonna be around? I owe ya dinner don't I?]

Okay how are you going to center it? By sweeping your dial around the circumference until you get the same reading all the way round. But if it is not a true circle then you will never get the same reading all the way around. So how are you going to center it? By sweeping......whoops I'm repeating myself.