Compass, or alternatives?

Hi,
Many of our measurement jobs are of cylindrical shape with drillings at both ends, standing on CMM table on either end. Does anyone use a compass, pushed against pins in the drill holes, or something similar to get the rotational alignment right? We want to use a fixed initial coordinate system without having to reset it by hand every time. How accurate are they?

316head said:

Hi,
Many of our measurement jobs are of cylindrical shape with drillings at both ends, standing on CMM table on either end. Does anyone use a compass, pushed against pins in the drill holes, or something similar to get the rotational alignment right? We want to use a fixed initial coordinate system without having to reset it by hand every time. How accurate are they?

Click to expand...

I'd suggest a digital but I've no idea as to how accurate they are. Depends I suppose on how accurate you need.

digital compass - Google Search

digital compass for car - Google Search

General claims for digital compass accuracy are within 0.5°, output resolution 0.1°. Given the nature and construction of the devices I imagine there are some fairly fundamental reasons for the commonality of such claims. Probably a combination hard to do much better without getting too expensive and whats good enough for the general market. I can think of some fancy ways of improving resolution assuming the sensor is well behaved, that you have a suitable display and are willing futz about during set-up but, frankly, such messing about will be too much trouble.

I'd have bought that the Argumented Reality mavens would have come up with something exploiting fixed markers for set-up and alignment of, well, pretty much anything. Mobile phone camera resolution and processing capability has gotten pretty ridiculous over the past few years and would certainly be up for the job. If I were looking to start up a niche business I'd certainly be taking a very cool look a what could be done with repurposed mobile phone hardware and software. I guess its hard to see past the "What does new Ikea furniture look like in my room" question.

Clive

0.5 degrees would be good enough. I was hoping someone would have actual experience with them; not in geocaching, but inside a machine shop full of interference...

A few years back I was doing a layout on my property using a reasonable well made compass. I was having a hard time getting my results to add up. If I checked in one direction, then checked in the opposite direction, my measurements were way off. I had my truck parked about 50 feet away and that was the culprit. In a shop full of iron, I would expect your measurements to be greatly affected, however it may not affect your repeatability

316head said:

... Does anyone use a compass, pushed against pins in the drill holes, or something similar to get the rotational alignment right?

Click to expand...

Seems unrealistic to me, what kind of angular resolution are you looking for?

I would use jo-plugs or gage pins in the holes, square the part up and clamp in a vee block or fixture. Then flip it over and square the vee block or fixture against an angle plate to check the other end.

I've had cylindrical parts that had to have features timed to tapped holes in the ends. I make a plate that bolts to the end of the part using shoulder bolts, then use the plate to orient the part.

Lots of ways to skin that cat, but magnetic compasses are not that accurate for splitting degrees.

Seems like a pretty sketchy way to align the part for metrology, considering the environment. If you have threaded inserts in your CMM table, just make up a simple fixture to align against a hard reference. Like Jancollc notes, many simple ways to do it. A compass is something I would never have considered for aligning a part on a CMM.

I get it. Aligning to bolt caps is what we currently do - it's slow, considering the model is different every time. I really thought someone would have had some experience, maybe I'll be the first to try?

316head said:

Hi,
Many of our measurement jobs are of cylindrical shape with drillings at both ends, standing on CMM table on either end. Does anyone use a compass, pushed against pins in the drill holes, or something similar to get the rotational alignment right? We want to use a fixed initial coordinate system without having to reset it by hand every time. How accurate are they?

Click to expand...

A drawing (schematic is OK) might help to see exactly what it s you're trying to achieve. I've got an idea that would involve making a "universal" template.