What's new
What's new

Help with identifying metric involute spline

Das_Wookie

Aluminum
Joined
Jan 11, 2008
Location
Austin, TX
I have a shaft with an unknown metric spline, and need help with identifying what it is.

retainer-600.jpg
spline-1.jpg
spline-2.jpg

Any assistance here with figuring out what the spline is would be greatly appreciated!
 
Machinery's Handbook has everything you need to get started. You are going to have to take measurements and calculate the possibilities, comparing them to what you have. Module Involute Splines can be wildly different animals. We do a fair amount of them and are working on just such an example, right now. Sorry to be the bearer of disappointment, but there is no easy answer.
 
What I can figure out:
Diameter min: 18.75mm
Diameter max: 22mm
Across 2 teeth: 8.45mm
Across 3 teeth: 12.25mm
Across 4 teeth: 15.45mm

The teeth measurements are averages as it seems the initial shaft was machined with a good bit of variability / wide open tolerances. I'm completely inexperienced with trying to match up an unknown mod spline. I'm willing to do the work to determine what it is, but I don't know what that "work" is to know what steps to take to try and determine what the answer is. That's why I'm asking for help. If there is no easy answer, that's fine. I don't know what I don't know in this regard however, so I'm unsure of how to even go about determining what it will take to get there. I have to make 50 of these things, but there's no drawings. I can't send out for EDM of the center spline if I can't give them a drawing of what they need to cut.
 
All the calculators and calculations I can find, are for determining which spline to use, size of the splines, power the splines can handle, et cetera. All of that presumes you are making a new spline, I guess. I have to instead determine what the existing spline on the shaft is, and model that.

None of the online calculations I can find however have anything where I can say, for example: "Show me a 13T Mod 2 spline with a 22mm OD, a 18.75mm ID, and 30 degree pressure angle. Now show me the same thing with a Mod 1.5, and a 1.75, and a 2.5. Now with a 45 degree pressure angle" and etc etc etc.

If I could find a way that I could print out a wide array of splines and then try and see which one most closely matches, I would try my hand at doing that. I'm TRYING to help myself here, but not knowing what I don't know is my problem. I need some help from a FOG to kick me in the right direction from the knowledge I do have. If I have a print, I can make the part. If I have an old part in my hand, and need to make more than one of them, I can (sometimes) get into trouble. I'm a strong self learner, but Other than this being identified as a Mod spline, and somewhat stuck in the water here. :(
 
May sound like a stupid question, but I have to ask given the obvious variance present and your assertion that the tolerances were "wide open". Are you positive that they are Metric Module? Any chance in heck that they're Imperial? ( Diametral Pitch )

After that, you're going to need to figure out which fit they are supposed to be. ( there are a number of them - Flat root/Side Fit, Fillet Root/Side Fit, Flat Root/OD Fit, SAE, etc... ) They could even be JIS. Did you receive any data from the customer? Any background info, at all? Or are you trying to resurrect something obsolete? What does it need to accomplish?

You can try this, if it helps. ( but most of this is in MH, anyway )
Technical Gear Info | Omni Gear & Machine Corp.

Hate to say it, but "Mod Spline" is not really that helpful. We need more info.

Just looked from another phone. Hard to say for certain, but it appears like that could be an OD Fit, but can't see the rest of it under the lock, and the pic is still too poor ( on this device ) to tell what kind of Root it is.
 
Everything else on it is metric. It's metric. The ID and OD measurements are metric. I have no experience with designing or drafting involute splines. I met with a mechanical engineer yesterday for two hours trying to get help identifying it. He kept wanting to design the splines to match power requirements and spline length. He couldn't get out that I'm just trying to IDENTIFY the spline. It's a set part I need to match. His thoughts were I needed to build a new shaft with a known spline. I can't do that. That's not what's requested.

It might be a standard spline. I don't know. I don't know what I don't know. I can make a part to print. I can draw most parts from direct measurement when I have to mate to an existing part. I don't know mod splines.

I've tried drafting the silliness with my standard method of drawing ANSI gears. That's not working. From trial and error I've gotten close, but it's not close enough. I need to get this right.I'm just stuck as to how.
 
I've tried drafting the silliness with my standard method of drawing ANSI gears. That's not working. From trial and error I've gotten close, but it's not close enough. I need to get this right.I'm just stuck as to how.

Not trying to be evasive or annoy you, but that is precisely part of the issue. Without specifics, you are guessing. Period. I say that because while I'm not on a big screen with a clear picture, it looks like they're beat to shite. So, any information you can interrogate from it is likely to be "off" from what was originally specified. And that doesn't even address tolerancing.

That said, try running it through your app and use 1.5M using DIN 5480. I think you will get close to what you are looking for.
 
Your photos appear to be of three separate objects. Which piece do you need to identify, and which does it have to fit? That shaft in pic #2 doesn't look like an involute profile but the plates in #1 and #3 do. If all you have to do is make something to fit that existing shaft, just grind a slotting cutter and chunk it out with the BP quill on a dividing head.
 
Not trying to be evasive or annoy you, but that is precisely part of the issue. Without specifics, you are guessing. Period. I say that because while I'm not on a big screen with a clear picture, it looks like they're beat to shite. So, any information you can interrogate from it is likely to be "off" from what was originally specified. And that doesn't even address tolerancing.

That said, try running it through your app and use 1.5M using DIN 5480. I think you will get close to what you are looking for.

The shaft is what I'm trying to get the profile for. I have the old drive gear and retainer, which I'm also trying to use for modeling. It's a slip fit, and is fairly tight. There is supposed to be as little amount of backlash as possible, but obviously there would need to be some amount if it is going to be able to be put on without a press. The parts need to be able to be changed in the field so I can't make it TOO tight.

I hear the complaint about photo quality. I dug up a USB microscope, in hopes this might help.

WIN_20180731_20_30_50_Pro.jpg
WIN_20180731_20_32_37_Pro.jpg
WIN_20180731_20_33_20_Pro.jpg

I had previously included the photo of the gear on the shaft, as well as the gear, and the retainer clip. I wasn't sure which would be the most helpful and the clip was the one I was able to get the best image of as I was able to put that down on a flatbed scanner. This time the photos are all of the old gear. I tried getting good photos of the shaft itself, but I just can't get anything without terrible parallax error. The old gear seems to be the best of the available parts to try and use, I think, for matching the geometry, and shows it's a flat root with flat sidewalls.
 
Hi Das Wookie:
If this was my project, here's how I would find out what I need to know:
1) I'd ask the customer what it is for (assuming he hasn't already told me) so I could make some educated guesses as to how good it has to be to make a happy customer.
2) I'd ask if there are any new, or near new units available to measure or if all he has is the thrashed crap you've been given so far.
3) I'd look on both mating parts to identify the areas that are least damaged or worn.
4) I'd mount the parts, make a template and make a Polysiloxane cast that I can mount on the Shadowgraph.
5) I'd satisfy myself what the basic tooth form is: straight sided? Involute? Stub involute?
6) I'd take the best tooth form from my Polysiloxane cast and make a CAD reproduction, using all I've found out so far.
7) I'd see if I can make some basic determinations about pressure angle, tooth depth, pitch diameter etc etc from that CAD model.
8) I'd try to match it with a standard (assuming they're ready to hand).
9) I'd take a stab at the necessary tolerances based on what it does.
10) I'd make a final drawing for the customer to sign off before I lifted a single tool.

Oh yeah, before I even looked seriously at the job as a candidate for my shop to get involved in I'd ask two questions:
1) To my customer I'd ask:...How much are you (the customer) prepared to spend to find this all out and will you pay cost plus, COD and a thousand bucks up front for the finding out?
2) To myself I'd ask...is all this shit gonna be WORTH it? Do I REALLY want this job?

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Everything else on it is metric. It's metric. The ID and OD measurements are metric. I have no experience with designing or drafting involute splines.
This really is a nightmare. Metric splines are a cesspool filled with alligators.

First thing to find out is when was it made ? Until recently there were no metric involute splines. They were inch splines converted to metric. The straight-sided splines were metric but the invos were an abortion. So if it's older, it's probably really a bastardized inch spline.

If it's newer, they did finally create real metric invos starting from scratch. All the ones I have seen were 30* so if you can find the standard somewhere you can calculate from that. AGMA prices are retarded, forget those idiots. Machinery's Handbook didn't use to have it but they might now. Or try the Ash Gear catalog. The older Ash catalogs were more useful for gear stuff than the Handbook. I don't know about now, haven't been there since forever.

If you are still stuck after a while, shoot me a pm and a sketch, everything here is metric. I can ask the guys at the shaper factory to check what you've got, if you get kind of close.

Metric involutes suck. Metric teeth suck. Metric sucks :D

p.s. I see you are in Austin ? And you are going to make 50 ? Amarillo Gear is close to you. Closer than China anyhow. They have done this all before. Sometimes it's faster and cheaper to go to a guy that has experience :)
 
Your photos appear to be of three separate objects. Which piece do you need to identify, and which does it have to fit? That shaft in pic #2 doesn't look like an involute profile but the plates in #1 and #3 do. If all you have to do is make something to fit that existing shaft, just grind a slotting cutter and chunk it out with the BP quill on a dividing head.

I'm likely going to have to send it out for wire EDM. I'm not making the shaft, but the gears to go on it. If I were making a single replacement shaft, that's certainly what I would do. I have to make replacement gearing for the shaft, and need to make 50 of them. Not something I can just bang out with a dividing head and a single piece of HSS. Otherwise yeah, that's what I'd be doing. I am thinking of worst case using a dividing head and turning my VMC into a shaper... but if I can get the profile nailed, then I can sub out for EDM.
 
This really is a nightmare. Metric splines are a cesspool filled with alligators.

First thing to find out is when was it made ? Until recently there were no metric involute splines. They were inch splines converted to metric. The straight-sided splines were metric but the invos were an abortion. So if it's older, it's probably really a bastardized inch spline.

If it's newer, they did finally create real metric invos starting from scratch. All the ones I have seen were 30* so if you can find the standard somewhere you can calculate from that. AGMA prices are retarded, forget those idiots. Machinery's Handbook didn't use to have it but they might now. Or try the Ash Gear catalog. The older Ash catalogs were more useful for gear stuff than the Handbook. I don't know about now, haven't been there since forever.

If you are still stuck after a while, shoot me a pm and a sketch, everything here is metric. I can ask the guys at the shaper factory to check what you've got, if you get kind of close.

Metric involutes suck. Metric teeth suck. Metric sucks :D

p.s. I see you are in Austin ? And you are going to make 50 ? Amarillo Gear is close to you. They have done this all before. Sometimes it's faster and cheaper to go to a guy that has experience :)

It's about 3 years old, so recent-ish I suppose. That seems like it's something relative. Some would consider something from 2000 still recent while others would consider anything older than 2 years ancient. :) I've got a less than a year old VMC sitting next to a 1942 Monarch CBB. While I agree that Metric in general sucks, most of all the parts I've been making for the last 4-5 years are all metric. I design in metric. I might plug an imperial dimension into a dimension every now and again, especially if I'm going to be doing anything manual-ish in the manufacturing... but if I'm making it on my VMC I'm full metric. I ALSO can wrap my head around what 20cm is between thumb and finger... tho I admit I'm still doing the 25.4 math in my head before I do. LOL
 
I am thinking of worst case using a dividing head and turning my VMC into a shaper... but if I can get the profile nailed, then I can sub out for EDM.
Worst case is a guy with a gear shaper who will do this properly, in the minimum amount of time with the correct cutter that will give you the correct profile :)

Oakland is a good alternative also. Dan from Oakland comes to mind, he's done this many times before. Even when other people here said it couldn't be done ...

You are doing this the hard way. UPS shipping to Oakland of your sample parts is what, ten bucks ? Fifteen ?



Yes, 3 years old is recent. That would be real metric, not converted. That's easier, you can get cutters off the shelf.

Now all you need is a shaper to put them in :) A GEAR shaper, a different animal than an old Atlas clunker.
 
If I could find a way that I could print out a wide array of splines and then try and see which one most closely matches, I would try my hand at doing that.

Gearotic might be a start for that. - "Gearotic Motion Gear design Software"

There's good info here on PM, but it will take some searching. Someone posted a way to reverse engineer a spline/gear with a spreadhseet and measurements over pins tha ti found useful, I can't offer any help finding it right now.

If you start with SAE spline info and find something that is close you can sort of correlate it to a module spline, that's not accurate but it's more than you have now. For example a 1.25 mod 30° spline is so close to a 20/40 30° spline that it's hard to tell the difference without careful measurement.

Zahnrad is a gearmaker and good at it, consider sending it to him or someone like him. I usually have to ask for outside help and I'm not completely new to gears.
 
Convert 22mm to inches. Count the splines, 13. Comes out to .866 which is real close to the major for 7/8-13 16/32DP, common as dirt in Hub City catalog.

Ed.
 
I think your best choice is to send this to someone who has done this before, as mentioned above.

If you still want to go it alone, you could buy the closest options (inch and metric) at Stock Drive Products and see which one fits. Me, I would spend the money to get the expert to do it. If you think that's too expensive, wait until you try and do it cheaper.
 
Not trying to be evasive or annoy you, but that is precisely part of the issue. Without specifics, you are guessing. Period. I say that because while I'm not on a big screen with a clear picture, it looks like they're beat to shite. So, any information you can interrogate from it is likely to be "off" from what was originally specified. And that doesn't even address tolerancing.

That said, try running it through your app and use 1.5M using DIN 5480. I think you will get close to what you are looking for.

I've been tryingn to fight with Fusion360 to get it to try to spit out a DIN5480 1.5Mod, and so far, no luck. I found a PDF for the spec, and have been fumbling around as I try to get it to create the profile from that. So far, no joy. If you could kick out a DXF of that spline, I'd be very thankful!
 








 
Back
Top