View Full Version : How do I measure this? (small round stuff with angles)

11-19-2012, 05:03 PM
A coupler in our icemaker broke and I'm not paying $60 for this tiny piece of metal. I have a small CNC mill so making the part is easy. Getting the dimensions is the hard part. Most of the projects that I've worked on had linear features that were relatively easy to measure with calipers. I've run into a few other parts with similar issues but this one is so simple that I thought this would be a good opportunity for me to learn the right way to dimension it. I have digital calipers, micrometer, surface plate and a height gage. I don't have any gage pins, radius gages or angle gages. Are there some techniques that I could use to get the dimensions that I need to draw this up in CAD?

The angle of the lugs and the radius on the through hole are the most difficult measurements for me.

The coupler is 0.70" in diameter and 0.50" tall.





11-19-2012, 05:45 PM
First thing to do is super glue the thing together.

Then start at it. If you have a cad program, it's going to be easy. Having the mating piece will also be a plus.

The only dicey thing I see on it it the inside hole. That will need some handwork to fit the shaft.

BTW, we have a store here in KC that is called MarBeck.


Just sayin'

Bob Fisher
11-19-2012, 06:15 PM
I would not worry too much about the angles, they are there for release from the mold. I don't think there has to be a precision fit to the mating piece since it only drives in one direction. Don't over engineer it. Bob.

11-19-2012, 06:21 PM
I'd like to take a "wild" guess...I don't have anything that you list either nor a surface plate, height gage. Could you not have a probe of some sort, like for a DTI, held in the spindle of the mill and just go
"touch, touch" and get measurements that way?
Also an opportunity, legitimate, to purchase more machining stuff ;). I am also not convinced some of those small radii are very specific, just square edges knocked down (?)

11-19-2012, 06:35 PM
for the angles you could take some close up pics with a digi camera on macro, make sure you're as square on as poss to the feature. print them out and draw lines on them so you can break out a protractor measure the angles, you can even scale some of the features from the print and measure the remnants of the original for comparison. to be fair i doubt most of it needs to be that accurate

edit to add: the width of the dogs can be measured and then used to work out the angle of each dog, input the diameter and chord length into something like the calculator in the link below and that will give you the dog angle. the chamfers on the top you could probably file by hand if needed.

have to assume the dogs are on a basic 120degree equal spacing, though you could measure the gaps between them in the same way (and input into calculator) to confirm once you've glued it back together

chord calculator (scroll down to the input chart):


also the shaft fits on will give you the nominal diameter of the bore, and the flats look to be equal

you can always print out the cad drawing 1:1 and overlay it to the real thing to see if it's in the right ball-park


11-19-2012, 07:07 PM
Seems to me all you need is the caliper.
You can measure everything except the actual angle of the 3 drive pins. Offhand it looks like the gaps between the pins is twice the width of the pin itself. Also it looks like all the angled sides converge on the centerline of the piece.

Like someone else already mentioned, you probably have room for a fair amount of slop on the whole thing. No need to make it within .001" unless you want to.

Besides with a CNC machine, if it comes out wrong, just tweak that dimension in the program and let it make another one. I could probably freehand it in an hour with some hand tools but that wouldn't be any fun. Hardest part would be making the hole for the shaft.

11-19-2012, 07:34 PM
That looks like a coupling from an old Scotsman flaker or IceOMatic flaker ice machine. Just be sure to account for the spider insert that goes between the driving lugs.

11-19-2012, 07:43 PM
That looks like a coupling from an old Scotsman flaker or IceOMatic flaker ice machine. Just be sure to account for the spider insert that goes between the driving lugs.

Funny thing is that it doesn't have a spider and the repair manual doesn't show one either. The two couplers just drop together so it's easy to attach the ice bin.

11-19-2012, 07:48 PM
I assumed that if you handed this part to a real machinist that there was a standard way, or two, to measure angles on a broken part.
What I probably do is take a macro picture, trace over all of the features in cad, dimension the angles and then average them.

11-19-2012, 07:51 PM
Besides with a CNC machine, if it comes out wrong, just tweak that dimension in the program and let it make another one.

yea, but the old one was chrome plated to resist corrosion so I'm going to make it out of stainless. Something that I usually avoid because I start breaking tools.

11-19-2012, 08:04 PM
I'm pretty sure that the standard ways are to use all the tools that you mention but that you do not have. An angle gage between any two legs will give the angle. A protractor will also work, of course. A bore gage will give you the radius for the inside of the center hole. A pin gage set inside the legs will give you that dimension.


11-19-2012, 08:31 PM
Figure out the radius on the center hole by setting the piece down on a ruler and playing with a compass...keep walking the other end of the compass down the ruler until you get the same sweep...then read the number where the pointy end of the compass is sitting on the ruler.

If it were me, I'd just take a slice of aluminum round, drill 3 holes in it and press 3 pins through it. Machine a steel dummy shaft and sharpen it to press-punch the center hole accuratly after roughing it out close with round tools.

It's just an icemaker.

Now, if I was bored...I'd glue it back together, stick it in my rotary table chuck and spend about 5 minutes with an indicator in the quill figuring out my toolpaths, then machine it all...but I'm rarely bored anymore...I have a 2 year old kid.

11-19-2012, 08:40 PM
Assuming you first glue the part back together, would the design of the mating components allow for you to just make a ring and press fit it over the OD to keep it in one piece?

11-19-2012, 08:58 PM

Hold it in a 3-jaw chuck on a rotary table set vertical. Use the hand-wheel graduations and a good dial indicator on a furface gauge - all on a good flat surface (ie mill table or surface plate).

Get one face/surface parallel to the base and measure the rest of the angles from there by rotating the rotary table until ach flat in turn is horizontal.

Then add or subtract each angle for each flat.

Bob Fisher
11-19-2012, 09:32 PM
I like Chris's idea ! Saves a lot of work if it can fit. Probably will not break again. Maybe ought to do both ends! Bob.

11-19-2012, 09:43 PM
Funny thing is that it doesn't have a spider and the repair manual doesn't show one either. The two couplers just drop together so it's easy to attach the ice bin. That may be, but I doubt it. The space between the lugs you are showing is a bit larger than the lugs shown and the lugs on mating half of the coupling are usually the same size as those shown, therefore there is room there for a spider. I have seen them chewed to nearly nothing on a few machines. What brand machine are you working on, anyway? If that is off of an ice machine, it will have an insert to prevent the metal to metal contact and wearing the lugs off.

11-19-2012, 10:39 PM
It looks like a small Lovejoy coupling to me. Google it and find the info for the lovejoy of the same diameter. I would Google for you but I'm on my phone and have a crappy signal out here in the boonies.

Mike Nash
11-19-2012, 10:49 PM
I agree with Mr Syil. I just used the last of your pictures and can make pretty good guesses on all the dimensions using your ruler to set the scale in Autocad. I'm getting 45 for the lugs with a line from the faces extending through the center as Dave mentioned. For actual OD, ID and flats, just measure those. The ruler being at the bottom of the piece throws off the dims somewhat using the image, but I'm seeing maybe 3/8" on the ID and 1/4" between the flats.

11-19-2012, 10:56 PM
I just went to the Lovejoy site and they even have the 2d and 3d cad drawings in a download available. You could just measure the OD and find the matching one.

11-19-2012, 11:38 PM
Much like this on page 1177 at McMaster-Carr:

Jaakko Fagerlund
11-20-2012, 12:45 AM
You only need calipers to measure that thing, there is nothing special in that piece that would require anything else. The diameters, length, the hole, the lugs, all with basic calipers.

What I mean is that you are not shooting with it or sending it to space.

11-20-2012, 10:41 AM
You only need calipers.

For the internal radius- Measure the wall thickness from the od to the id of the radius and subtract that from .350 (.700 od/2).

The angle of the lugs- You have two legs of a right triangle. One being a .350 hypotanuse(the radius of the part) and the side opposite the angle (1/2 the chord as measured from a lug). Get the sine of the triangle by dividing the op side by the hyp.; look that number up, it is the angle of the triangle. Double this for the angle of the lug.

The rest of it is pretty straight forward.


Paul Alciatore
11-20-2012, 02:11 PM
My take on this:


The center hole does not have to be the double D shape of the original. The arc ends of this hole are not functional except to keep it centered on the shaft. It is the two flats that do all the work. So a rectangular hole with round corners is just as good and easier to cut with a small milling cutter.

The driving studs are a bit harder. First, their sides are 99% likely to be radial lines: so this is a good first assumption. This means that all you need to measure is the OD of the part and the chordal width of the studs (m in my drawing). A simple trig formula gives the angle of the sides. A complication is that the corners of the studs are probably rounded so a fudge factor is in order. You can either increase the measurement, m with a good guess or decrease the OD to the point where the calipers rest on the rounded corners. This will make it a cut and try proposition.

PS: I am betting the angle a is about 40 degrees. That is simply 1/9 of 360. 3 studs on each end coupling and six rubber fingers between them at half the width of the studs. 6 x 40 = 240 6 x 20 = 120 120 + 240 = 360.

11-20-2012, 05:00 PM
Since this is going to be done in a CNC mill the lugs are easy, as is the radius of the double d. The hard part is the corners of the double d. I would relieve the corners by drilling their locations first with a small drill on the locations where the radii and flats intersect.* (Look at a quality 6 point mechanic's socket to get the idea.) This will leave you at most, with just a little bit of clean up with a file depending on the diameter of the endmill.

Unfortunately, I can't upload pictures here and I don't have a host site.


*When you draw it up in CAD play with the location of the drill to get the best results.