Precision Measuring "Long" Distances

[LOTT]

Here's a dumb question- How do you measure +/-.005" on lengths over what any of your calipers or machines (or CMM) can travel? Say 6' to 10'. Most of the stuff we make that large is part of a welding project and tape measure tolerance is fine, but what do you do when it needs to be tighter?

Thanks.

 

[Bakafish]

Professional laser interferometer measurement systems like from Keysight are likely the entry point for that kind of precision and distance. There are even more advanced systems that I've seen that use point cloud scans and special datum point units, but that's hundreds of thousands of dollars. If it's something you are making a lot of, a robust jig and strict environmental control could allow you to just contract out the jig's certification rather than buying the equipment yourself.

 

[john.k]

I ve got a set of inside micrometers with extension rods s to 20 feet ,and outside jaws to measure either internals or externals Id forgotten all about them ,until a month or so ago ,the damn cat jumped up onto the cupboard where they were and knocked down a stack of wood micrometer boxes ,breaking some ,others still in one piece.

 

[plastikdreams]

Get yourself one of the laser tape measures from harbor freight.

 

[LOTT]

I think we just created the Goldilocks thread. BekaFish is more high end then we can do.

Plastikdreams is too low end (Horrible Fright for measuring tools... I wouldn't be able to look in the mirror...).

And John.K is just right. A long reach micrometer set would do it, and allow us to set up a quick jig like what BekaFish referenced; we aren't doing enough to warrant sending it out but we could knock something together ourselves as long as we can measure it accurately.

Thanks for the help.

 

[Conrad Hoffman]

The build-up of standards is common in metrology, be it electrical, weights or measures. What's the longest you can measure to under a thou? Maybe 24", more I hope. Make up a set of rods as close to 24" as possible, measured as accurately as possible. You're essentially making long shop-made gauge blocks. Rack 'em, stack 'em and compare to the part that needs to be measured. If you could get three rods or bars to better than a half thou, you might be able to measure within two thou in six feet. Watch temperature over those distances!

 

[DanielG]

.005 isn't too bad. For a medium sized part like yours, look at an arm, like Faro or Romer.

Laser trackers are very nice for fab work, with a larger envelope but much more expensive. You can stick an SMR to a part and tap it around with a hammer until the readout tells you you're in the correct place. We also have a centerpunch nest which allows us to lay out holes very accurately. Used, you're probably looking at $60-$70k with software; the arms are a lot less expensive.

 

[Bakafish]

I think we just created the Goldilocks thread. BekaFish is more high end then we can do.

Bakafish

Plastikdreams is too low end (Horrible Fright for measuring tools... I wouldn't be able to look in the mirror...).

And John.K is just right. A long reach micrometer set would do it, and allow us to set up a quick jig like what BekaFish referenced; we aren't doing enough to warrant sending it out but we could knock something together ourselves as long as we can measure it accurately.

I thought about suggesting this, I've seen some beautiful old sets of rod micrometers that screw together, but honestly these sets are uncommon these days and I don't actually think this is the right solution. Although I'm sure that they were very well made and as accurate as they could manage at the time, you'd be putting a lot of faith in some pretty old gear to get the tolerance you specified over a span of 10 feet (assuming you could find a working set. ) There's a lot of potential stacking error going on and hiring someone to come out with a modern trustworthy method (that is certified and traceable, something customers usually expect) seems like a better business expense. But that's just my take, Japanese engineering culture is as conservative as it gets.

Thanks for the help.

 

[eKretz]

We had a set of rail mics in the big shop I apprenticed in. Long I-beams of various lengths around 4" x 4" with movable micrometer head and anvil that mounted on the beam. We had very long standards and a stand for setting them in good alignment with v-blocks while checking/setting the mic. You could use multiple standards aligned in the v-blocks for setting up the mic heads and zeroing. Both the setting stand and the mic beam were planed nice and flat. Worked really good, but very important to be sure that the mic was held exactly the same way in use as it was when checking the standard. Also had insulated grips that could be stuck on the rail to help with holding the mic in the exact same places as when checking the standard. We had up to 12' long rails so they were pretty flexy. Another shop I worked in bought the rail mics when that bigger shop folded on my advice.

Edit: mic heads could be flipped around also to measure in between two inside facing surfaces.

 

[LOTT]

Bakafish



I thought about suggesting this, I've seen some beautiful old sets of rod micrometers that screw together, but honestly these sets are uncommon these days and I don't actually think this is the right solution. Although I'm sure that they were very well made and as accurate as they could manage at the time, you'd be putting a lot of faith in some pretty old gear to get the tolerance you specified over a span of 10 feet (assuming you could find a working set. ) There's a lot of potential stacking error going on and hiring someone to come out with a modern trustworthy method (that is certified and traceable, something customers usually expect) seems like a better business expense. But that's just my take, Japanese engineering culture is as conservative as it gets.

Sorry, can't spell. .005 at 10 feet is the worst case scenario, +/-.010 at 6' is more likely. There are several sets available online, maybe picking up a set and then checking it against @Conrad Hoffman method of shop made rods. This isn't something we are doing all the time so I justify putting too much money into it. And this was partly curiosity because everything I know is either small and precise (machined) or big and rough (welding/fabrication), no experience with big and precise.

 

[memphisjed]

Egrats is on track how I would do it. Screw a track(s) on the inside flange of a beam, two linear rails into the web (thin cheap rails would work in this case) and make a giant tramel. An encoder and simple dro on the movable end.
Giant calibers.
You would need a one time calibration to get counts correct.

 

[triumph406]

I have 3 6' Starret scales that i put end to end and use them to calibrate the steel tape if it that's what I 'm using as a measuring tool

I used to cut a lot of 8020 extrusion, and that type of precision with a tape was usually sufficient.

 

[plastikdreams]

Starrett does have calibrated tape measures...

 

[LOTT]

I have been amazed at how accurate it's possible to be with a steel rule, but it's also easy to double check those distances with some calipers or a mic. Using one of those calibrated tapes would probably hit our numbers in all honesty, but I'd still want something else to check it against.

I appreciate all the responses and have a direction to go, if anyone wants to hijack the thread go for it.

 

[gbent]

I have a set of Starrett inside mics that go from 5-460", iirc. Extensions are 2' each, no recent calibration.

 

[john.k]

The one I got has the extension rods in fibre tubes ,so hand heat doesnt affect it........It isnt in the stack the cat knocked down ,what I thought was it ,is actually a Davis Thompson Tubular Micrometer which measures to 24"........the search goes on.

 

[Peter S]

When setting up welding jigs, I used 2x 1000mm (39") steel rules. These rules (Toledo) had one end rounded but they still worked fine. I could measure up to 2000mm within approx. 0.25mm (0.010") and could have used longer rules if necessary. Using a loupe helps.

For this type of work (setting flat internal faces of jig), 2x shorter rules are better than 1x long rule. The square end of each rule is placed against a jig face and the rounded ends meet nearest the centre. Measurements are taken from one of the 1000mm lines, one rule on top of the other. Easy to add 1000mm to whatever the other rule says.

External jig faces can be made "internal" by clamping flat steel to them. Or you can buy or make clamp-on caliper-type jaws to fit steel rules.

BTW, 1000mm rules are quite reasonably priced, longer not so much. Good for checking tape measures too.

Keep an eye out for long vernier calipers at auctions. I recently bought a 1000mm Mitutoyo and have seen 1500mm/60" and even 2000mm/80" sizes for reasonable prices. Not always Mitutoyo though.