[Tropi]

Here's a slightly off-beat enquiry!

Are there any engineering people on board who could advise on how to measure the Inside Diameter (ID) of a cylinder the size of a full-size car tyre ranging up to a large dump-truck tyre, to an accuracy of one mm?

Outside Diameter (OD) is easy, using a tightly wrapped tape-measure to obtain circumference and dividing that figure by Pi to obtain diameter, very accurately.

But ID is a different ball game altogether.
It's far too difficult, well-nigh impossible, to run a tape accurately round the inside circumference as per OD above.

Trying to obtain a direct measurement across Internal Diameter is full of pitfalls.
A tape is too difficult to position and read accurately - useless!
We have used a £50 laser, which was not even accurate to 10mm - useless!
We ae now experimenting with a telescopic 'pole' (Nedo Messfix) which is accurate enough, but rather heavy and hard to handle in awkward confines - It has promise. It would be a LOT better if it were spring loaded and with a small ball or wheel on its end.
We are also experimenting with a £300 Leica laser, accurate to 3mm but need to improve on that - this also has promise and is by FAR the easiest of all to handle.

My gut feeling is that an accurate 'measuring wheel', to obtain Internal Circumference would be best of all.
Tiny ones are commonly used to measure scale distances on maps, but these are really TOO small for my needs.
Larger ones are used to measure fields and roads etc, but are way too unwieldy and inaccurate for my purpose.
I found a small/medium sized wheel, used for measuring carpets, but it its accuracy is no better than 10mm.
All measuring wheels discovered to date are impractical for the reasons listed.
But I suspect there will be something suitable, if only I knew where to look . . .

So, suggestions would be VERY welcome!

Oh, the range of diameters required runs from 300mm to 2300mm.

 

[David Lawson]

I'm no engineer but is there a way you could draw a tangent on the circle and bisect it at right angles to ensure the line runs through the centre of the circle which will give you an accurate diameter measurement to work with. That's the theory I don't know how you would do this in practice.

David

 

[MK2-Escort]

I know in engineering at college they have tools to measure inside diameter, but don't know about tools to measure that great of size.

 

[McLaren]

Just got a couple of suggestions:

1 - Make your own measuring wheel.

2 - Take a piece of masking tape and lay it round the inside of the tyre. Overlap it by an inch or two and cut it with a sharp knife somewhere in the overlap. Now peel it off and measure the length when it's out flat on the floor.

3 - You could find the thickness of the tyre, use some cunning calculations, with the OD and Robert is your mother's brother.

But I guess you might have tried these already...

McLaren

 

[cheater]

Tropi, construct an telescoping pole, in the manner of an adjustable curtain rod, with a locking screw to set the length. Then remove the pole from your cylinder and measure the length of the pole.

This is analagous to the telescoping hole gages used in the machinist trade to measure diameters of small holes.

 

[CoolScientist]

A set of large Vernier Calipers would do it....

I likt the tape idea..... so simple

CoolS

 

[Harry Porsche]

Perhaps you could measure the thickness of the wall, double it, and subtract that amount fron the O.D. measurement? Just a thought. I've used this method with good results before (on smaller objects).

 

[a Bill]

Why don't you just measure the inside diameter and put your L337 math skilZ to work?

They don't teach math in schools anymore? You're looking for circumference. It's something like Pi times Diameter and is as accurate as you want to be. You can take it out ten places past the decimal or two places, whatever floats your boat

The elaborate measuring tool you would use is called a yard stick here in the States but you can use anything that measures as accurately as you need.

I've been out of school for, holy smoke!, longer than some of you have been alive! and I can remember some of it. Me, the same guy that can't remember where he left a beverage or if he even fixed a beverage, remembers some stuff from school

 

[cheater]

Umm, Bill, you might want to re-read Tropi's original post a little more carefully. He says he's wanting to measure the ID, not inner circumference, and is having difficulty doing that to the desired accuracy with equipment at hand.

Do you remember just how many of those beverages you've had tonight?

 

[a Bill]

I edited my post and added the miracle measuring tool

He was saying he's wrapping a tape around the outside of the tire, that is grossly inaccurate and that's circumference. Sounds like he's going about it backwards.

OK, how about this? Take the tire, lay if flat on the ground. Place one framing square on the floor in a 'fixed' position. Take a second square and place it loosely on the opposite side. Hook your inaccurate tape measure on the fixed square and read what it says on the opposite side. Now slowly move the square off center each way until you can determine where actual center is. This is determined by reading off the largest measurement. So if you read 65mm and swing it to the right a bit and it reads 66mm then you are now closer to center of the circle. It's hard as blazes to explain but anyone can show you in seconds how it's done. It's a standard construction technique used by carpenters.

Anything using a tape measure won't be very accurate. Measuring instruments that are accurate are terribly expensive. When they print the characters on the tape it gets shifted a bit because it's a rotary print job on the fly and not a stamping or press type print. Besides, it's just printed. Maybe you've seen that engineers rules cost a billion euros compared to the same rule your Aunt Martha buys. This is because the engineers rule is correct. If it says 33mm, you can bet your lunch it's 33mm while Aunt Marthas rule might say 33.5mm or 32.9mm, who knows?

I'm in a typing mood tonight. No clue what I'm typing but there are lots of letters up there

 

[Swissracer]

OK so who ever said it was even a tyre? where did they come into it?

Might seem silly but isnt' the easiest solution to pay someone else to do it?

Actually the measuring OD and doing the maths is the easiest and most accurate .

Tape wrapped round, though a neat idea would involve alot of care and chance that it didn't stretch. ALL tape stretches and the tolerance Tropi asks for is fairly high over the biggest diameter mentioned.

 

[Xlot]

mmm.... no practical experience, just speculating.... even Vernier calipers on an ID can be a bitch

First, do you want the 1 mm accuracy over 300 mm ID (that's circa 0.3%, feasible) or over 2,300 mm - one order of magnitude more ?

A possible system could be :
- mark a point on the circumference
- place a ruler in any direction on that point and mark the second point where it intersects the circumference
- run a line from the second point, perpendicular to the ruler - you thus get a third point on the circumference
- the first and third points being on a diameter, you can either measure their distance directly or use Pithagoras

 

[McLaren]

Ah, the problem is, I know loads of ways to do it mathmatically... but I think Tropi is asking for a way of doing it practically.

I like the idea of an sprung loaded, extending pole with little rollers on each end. That way all you have to do is compress it, put it in and let it do the rest of the work.

Unfortunatly, without spending big bucks or making something yourself, I think you're buggered.

McLaren

EDIT: Sorry Fergy!

 

[Fergy]

Tropi, perhaps you could be more specific about the actual items being measured? Are they tires?

Something like the spring loaded extension pole (which I like) will only work if the object is quite solid. For something more flexible, the pole may deform the item (to some extent) thereby giving false readings.

McLaren.... stop posting just in front of me!

 

[cheater]

I still say my suggestion of a telecoping rod, whether spring-loaded or not, is the cheapest solution that will give the required accuracy, assuming Tropi has a way to measure the length of the pole.

For greater accuracy, there are products made for just this sort of problem:

caliper extenders

 

[TSRF]

I use a simple device: two sticks of wood with a bolt retaining one to the other and working as a compass. Then it is easy to measure across with a tape.

 

[Tropi]

Cor Blimey chaps!
I thought I'd leave it for a day or two and let one or two build up - what a magnificent response and some of them rather good!
ALL deserve my thanks AND a response, though responding to all is going to take some time. Please, no one feel left out if you are left out!

First of all, I am impressed by the more eagle eyed noticing that I didn't actually say these thing ARE tyres. You can get away with SFA here! Must confess, I cheated ever so slightly but with good intent. That good intent was obviously to pump all your superb brains dry and thus exercise your mental facilities so that they don't atrophy.
(OY! Who added "like yours have, Tropi!"?

I picked 'tyres', partly to legitimately qualify for inclusion in this board and partly because there seemed a fair chance that one of you might have real world experience with tyres that was applicable to my little problem. And partly because it seemed there might be spin offs for other people on the board too.

I'll make a start:
QUOTE isnt' the easiest solution to pay someone else to do it?
Yep! But expensive and I know this because I AM that someone and I am charging a LOT!
Well, THEY think so anyway.

A word on tape measures.
A CALIBRATED steel tape measure actually does have good enough accuracy in itself - IF it is checked against certificated instrumentation, a correction table can be legitimately applied. The practical problem is in one person holding both ends in exactly the correct positions AND being able to read one end to 1mm. Even with 20/20 vision, which most people don't actually have, reading 1mm accurately is slightly problematic, in addition to which, parallax error is a real problem and can, alone, very easily add a 2mm error, even with an accurate tape measure. Holding 'in the correct position' is another problem in practice - you must ensure that both ends lie on a true circle and then move one end back and forth in an arc to obtain max distance without losing the true circular line, while continually reading the scale to find that max length. Two people can manage it, but that is expensive and it always remains problematical to read that damned tape to 1mm, no matter what. Maybe a gorilla could do it?
Flippant remark maybe, but the ideal solution actually would be one that a trained monkey could carry out to the required accuracy.

While I remember, yes, 1mm accuracy is needed right up to 2300mm diameter.

The sticky tape sounds an EXCELLENT idea, but, as pointed out, pretty well all tape other than steel and fibreglass is a little stretchy and I don't believe would give that 1mm accuracy. Maybe someone knows of suitable sticky tape?

I have to look at Beppe Xlot's mathematical thoughts when it's not so late at night - it looks interesting.

The beauty of actually measuring Internal Circumference and then calculating Internal Diameter is that it increases accuracy. Circumference errors can be slightly over 3 mm before registering as a 1mm Diameter error. Other advantages are that only a single measurement is needed, whereas at least 3 direct diameters would be required to give anything like the same consistency. So Circumference is both faster and considerably more accurate in every way.

Phew! I can't manage any more this evening. <Audience heaves a collective sigh of relief>
But, I'll be back!

 

[Xlot]

Hi Tropi,

it's even later here but I'm not letting you go to sleep before you come clean !!

WTF are you measuring and why do you need a 1 mm accuracy over 2,300 mm ??? I mean, if the "object" is that round, it must have been made with precision machinery, and the ID would be known from the start - if not (say centrifugal casting), ID isn't going to be that constant and 1mm accuracy is meaningless - think own weight deformation, just for one

That said, I believe so far the best method is to use my suggestion to determine the diameter points, and then cheater's caliper extension to measure it - but you would also need temperature compensation

Ciao
Beppe

 

[Tropi]

Oh, did I forget to say what the things were?

OK, they are bloody great concrete water pipes and brand new European Standards (ENs) require that 1mm accuracy in measurement.
Strangely enough, although the accuracy of the joints is absolutely critical (hence the 1mm accuracy) there is NO requirement for them to be ROUND! Only that they match and are consistent. So egg shapes are quite permissible and are made by some companies, but I could never get away with pretending that eggs were tyres, could I?

So, yes, slump under the massive dead wight of 'green' (uncured) concrete can be a real problem, but the measurements must still be to 1mm in all dimensions.

And now I must slink off to the pit.
Night all. :

 

[McLaren]

Just make up a measurement. If anyone can prove that it's wrong then they must have a fairly accurate measuring system, so copy them. Otherwise, no problem!

Mclaren