What's new
What's new

Are these Bevel Gears DP or Metric Modules Gears, How Do I Figure This Out?

A simple way of cutting lathe apron bevels is to narrow the tooth face down to say 1/2" at the max dia and cut one pass ....replacing the plastic gear with a steel one would give same strength .....I never used a file,but rather a 4" angle grinder with a narrow cutoff wheel to touch up .
Thus converting the large bevel closer to a contrate wheel ..good one John.
 
Just out of curiosity, what is the serial number of your lathe. Mine is 568810.
Serial# 530024 with Mfr date of 1983-1. Cannot remember if I gave them incorrect date of 1981, I did not have a part#, they did ask for a pic of the machine, and diameter and tooth count of gear, but something is not adding up if they say it was discontinued in 82, and machine was made in 83??? Parts manual arrived a couple days ago so I now have the part#, I'm going to call back on Monday and see if that changes parts availability and price.

Using a protractor, face angle measured 16 degrees, cut angle 23 degrees. Looking at chart in MH (page 760 in 1946 copy) it shows 20T pinion with a 56T gear, the face angle of gear should be 17 degree 44', that is within 2 degrees of what I measured, and is probably correct. On page 766 chart (Number of cutters), it shows #1 cutter for the gear. Does the equation for diametral pitch (number of teeth + 2, divided by diameter = DP) work for an almost flat gear like this? If so, 58 divided by 4.5 = 12.8888, do I round down to 12, or up to 13? Or do I need a different equation?
 
I have had to cut replacement plastic gears for various finishing machines where I work. The majority of them were blue Nyloil ,Nylon with oil. I have used Delrin also. Nyatron is Nylon with graphite is a very good gear material.I have been stuck on Nylon with oil and molysulfide lately. They all machine nicely. As far as I know all bevels are measured from the large end, don't see how you could do it any other way.
When installing them the back cones should match evenly.
In the video EmGo posted the related video's listed at the end have one by Kieth Rucker which is exactly what you are trying to do.
 
Hmm. This sounds like something one could make with a rotary table fixture on an old-school horizontal shaper. Slowly.

How do the modern generator tools work? Edit: @EmGo answered this question in posting #16. It's a glorified shaper.
Guys, no. Bevel gears are cones. The teeth and the tooth spaces both get smaller as you go from outside to inside.

Yes, a straight bevel gleason is a "glorified shaper" but part of the glorification is that the tools are at an angle. The other part, which you didn't see in that video because it was just roughing, is that the cutting tools move on an arc simultaneous with the part rotating. It is similar to a rack shaping teeth but it's not a rack, it's an angled rack. That's why there's two tools instead of just one, each side of a cutting tool is one side of the cutting-rack tooth, except they are not parallel.

As far as I know, there's no version of high speed steel that can get big and small as it is stroked. Which is why

dalmatiangirl61 said:
On page 766 chart (Number of cutters), it shows #1 cutter for the gear. Does the equation for diametral pitch (number of teeth + 2, divided by diameter = DP) work for an almost flat gear like this? If so, 58 divided by 4.5 = 12.8888, do I round down to 12, or up to 13? Or do I need a different equation?
this will not work either. The cutter has to get smaller (but the same shape) as it moves towards the inner end of the blank. Which, at this point in our time-space continuum, is not yet possible.

Therefore, when cutting a bevel gear using a space cutter method

1) it is not possible for the end product to be correct

2) you get to choose which incorrect cutter you want to use then hand-file the other end of the tooth to shape.

There is no single angle to a bevel tooth. The root is at one angle, the pitch cone diameter is at another, and the face is at another. These aren't cylinders, they are cones, so all your diametral measurements are ... not exactly useless, but they won't give you the answers that a cylindrical gear does, either.

If you absolutely have to do the space cutter method, I'd go from outside to inside with the bigegr cutter because it's easier to file the addendum on the inner end than the dedendum on the outer end. On both sides, 56 times, yum. At least it's plastic and forgiving. The slitting saw rack-cutter method would work if you made a really big slitting saw so that the curve on the outer end was big enough to make the arc in the root unimportant, and you went around twice at the relevant angles. Easiest would be using an nc milling machine, the tooth faces are just ruled surfaces, pretty easy to model. And I guess if the part wasn't plastic you could wire edm it.

In Spot's case, I'd do a different job with the time, make the money and buy it. Usually I am a dedicated diy-er but in this case, way too much work for the return.
 
As far as I know, there's no version of high speed steel that can get big and small as it is stroked. Which is why


this will not work either. The cutter has to get smaller (but the same shape) as it moves towards the inner end of the blank. Which, at this point in our time-space continuum, is not yet possible.

Therefore, when cutting a bevel gear using a space cutter method

1) it is not possible for the end product to be correct

2) you get to choose which incorrect cutter you want to use then hand-file the other end of the tooth to shape.

There is no single angle to a bevel tooth. The root is at one angle, the pitch cone diameter is at another, and the face is at another. These aren't cylinders, they are cones, so all your diametral measurements are ... not exactly useless, but they won't give you the answers that a cylindrical gear does, either.

If you absolutely have to do the space cutter method, I'd go from outside to inside with the bigegr cutter because it's easier to file the addendum on the inner end than the dedendum on the outer end. On both sides, 56 times, yum. At least it's plastic and forgiving. The slitting saw rack-cutter method would work if you made a really big slitting saw so that the curve on the outer end was big enough to make the arc in the root unimportant, and you went around twice at the relevant angles. Easiest would be using an nc milling machine, the tooth faces are just ruled surfaces, pretty easy to model. And I guess if the part wasn't plastic you could wire edm it.
Looking at the gear, it appears that the cut angle is such that the inner edge is just shallower than the outer edge which creates the diverging angles. It does not need to be "correct", just "good enough". If I had a drawer full of cutters I'd pick one out that was "close enough" and run with it and this thread would never have been started. My christmas present to myself this year was a dividing head, just cause I always wanted one, and I don't have any cutters at all, yet. It's winter here still, probably for 2 more months, work is slow, I've got time to make some mistakes, and learn something new, and it keeps me from going stir crazy(er).

So what dimensions/calculations do I need to figure out what set of cutters to order?
Edit: Looking at the offerings of online sellers of involute cutters I see them in straight cut and bevel gear styles.
 
Last edited:
In Spot's case, I'd do a different job with the time, make the money and buy it. Usually I am a dedicated diy-er but in this case, way too much work for the return.

Having done this exactly ONCE - to match the pinion gear remaining on a lathe leadscrew - I second the 'buy it' recommendation.

I made mine out of cast iron because it had a cone clutch as an integral part. It did work, backlash was pretty bad but it was good enough. And there was a fair amount of hand filing needed.

I'd not do it again and I've far more tooling & knowledge now than I had then.

PDW
 
Having done this exactly ONCE - to match the pinion gear remaining on a lathe leadscrew - I second the 'buy it' recommendation.

I made mine out of cast iron because it had a cone clutch as an integral part. It did work, backlash was pretty bad but it was good enough. And there was a fair amount of hand filing needed.

I'd not do it again and I've far more tooling & knowledge now than I had then.

PDW
Great, I like to learn the hard way too, those types of lessons stick better. So how do I figure out which cutters to order?
 
Guys, no. Bevel gears are cones. The teeth and the tooth spaces both get smaller as you go from outside to inside.

Yes, a straight bevel gleason is a "glorified shaper" but part of the glorification is that the tools are at an angle. The other part, which you didn't see in that video because it was just roughing, is that the cutting tools move on an arc simultaneous with the part rotating. It is similar to a rack shaping teeth but it's not a rack, it's an angled rack. That's why there's two tools instead of just one, each side of a cutting tool is one side of the cutting-rack tooth, except they are not parallel.

Yes, I was being loose, but in the video you posted, one can see that the upper cutter assembly is not quite parallel with the horizontal item just above. I had looked at other videos, and it is clear that the upper and lower cutter assemblies need not be parallel. It's a very cute design.
 
Since you are in a DIY mood you could do the job in the manner I outlined earlier. Eyeball the 'radius' of the outer teeth, having Ivan Law's words in your head .. something like 'a simple radius is closer to correct than the outlying sizes of involute cutters'. Having an idea of the radius knock up some radius gauges by drilling some holes in sheet aluminium, steel or whatever the cut your gauges down with tinsnips to allow you to insert into the tooth space. Thus verifying your estimate of the radius 'by gauge' modify a suitable size centre drill to fit the tooth space nicely. Back in the day I did it by offhand grinding on the bench grinder then touching up/truing the radius ..radii .. with a mounted point in a Dremel style (homemade) tool. Set up your blank by the face/root angle of your original gear & have at it, end milling .. I too thought 'constant depthing' would be consistent with your original, but you can confirm that by measuring. Just emulate the original close as you can.
I'd post pics of my results, possibly even the tool (bet I still have it) but I (80 yo) just can't crack posting pics on this forum.
You'd have this job done before you could find the approximate involute cutter.
As a word of encouragement, one writer, a Prof W A Tuplin, ex David Brown, ex University of Sheffield, says in his book Gear Design, in relation to milling bevels "no bevel gear cutter has a specific 'pitch' .." & goes on to say there is quite a bit of overlap enabling any one of several sized cutters that will produce an acceptable result .. with subsequent hand working (filing). Then passes on to generation as per Gleason et al as the preferred/practical method ..
 
A friend has a whole Gleason hypoid bevel cutting line ......ring gear and ,pinion roughers ,shavers,oven ,hardening tank ..(the ring gear is clamped slightly sprung for quenching ,so that its flat when cold.)......all out in the open under plastic tarps .....the ring cutter has a rotary cutter head ,with a gap for the tooth to index past .....there are only 7(I think) cutting edges that cut the whole tooth space in one revolution of the head..........yes ,he has had it running out in the open..........He s a machine nut with acres of machines in the open ,and big plans of building a shed ,which the council will not allow in these days of big corporations ,bigger bribes ,and town planners beyond number.
 
I watched the video and it is fascinating. It is possible to generate a true bevel gear with involute tooth form with it on a simple spur gear.

But I think, with manual machining, it ends there. Even a helical would be very difficult because you would need to rotate the blank as the cutter passes through it. OK, you could CNC it.

But for a true bevel gear the cutter would need to travel on a line from the center of the cone to the involute shape on the outer edge of the teeth. You could not just set the RT or indexing head at the cone angle because that angle changes from the top of the tooth to the bottom. So, you would need to rotate (on a vertical axis) the RT or indexing head for every new cut and somehow re-establish the distance from the edge of the saw to the outer edge of the teeth. Just deciding on the location of the axis for that rotation would be difficult. But, again, perhaps CNC.

I wonder if wire EDM could do it. One end of the wire is fixed and the other moves in a path following the outline of the gear tooth, but larger because it would need to be further from the center of the cone. The blank could be still for each tooth and be indexed for the next one. Either the wire or the axis of the indexing device would need to be at the cone angle - probably the cone angle at the pitch circle. All CNC controlled, of course.



Here is a method using a slitting saw that should work as well:

 
Great, I like to learn the hard way too, those types of lessons stick better. So how do I figure out which cutters to order?

I used the info in Advanced Machine Work starting at pafe 11-53 on bevel gearing. Honestly can't remember how I determined the cutter size, I may have tried a bunch in the pinion to get one 'good enough'.

My cone angle was 73 deg 20 min - the sketch is still there as a bookmark.

I used a DH set so the face I wanted to cut was vertical and used the power up on a horizontal mill to do the cuts. IIRC I did 3 passes rolling the blank as needed (and fucked one tooth up but only partially).

It was entertaining but like cutting a 73 tooth gear (prime number) on one of those astronomical(ly complex) dividing heads once was definitely enough.

Personally if possible I'd buy a matched pair that will fit the space and piss off both the originals, but I understand the 'learn something' bit too.

PDW
 
I have only cut a bevel gear once in my life, and that was under supervision 50 years ago!
Here’s what I remember, oh dear.
It was on a horizontal mill,
Blank on chuck
DP cutter for tooth count of a spur gear was used
Head was inclined to angle
It was more or less cut like a spur except the teeth were generated by two cuts
First cut was done then the blank was reset in the Chuck and around again to generate the trailing flank
Daunting at the start, but after doing it quite satisfying!
Keith of vintage machinery fame does the same oddly
Look for his K&T video of helical attachment I beleive
I remember checking over pins in the gullets
Mark
 
On the hypoid Gleason,both the blank and the cutter head rock back and forth ,and rotate at the same time ..........far as I know ,the Gleason hypoid is a unique form and different in design from other machine makers hypoids.
 
Why not 3D model the gear and 3D print it?

That would allow you to make prototypes locally to you to check fit and dimensions.

Once you have worked out what it is, including giving it a 20000 revolutions test meshed with the steel one to assess contact pattern/wear, then send it to one of the online people (JLC PCB) to print as they have a wide choice of materials and processes that will produce the best engineered solution for you. Print a few spares as the extra-over cost to a single item is tiny.

Some useful references here if you want to do your own thing - the two references from the internet archive are the ones from whom Ivan Law's work was derived.

 
You guys come up with more ways to do bullshit than any ten normal people. Is there something really attractive about bevels ?

If all you care is that it turns, just make a spur gear and a face gear. Teeth look kind of weird on a face gear but it's not hard to make.

3D printer is not that terrible an idea here, but going to take a fair amount of experience with cadcam ....
 
FWIW, I have done the bevel ring gear on my Marvel 8 saw, twice, in 40 years. Horizontal mill, rotary table tilted a couple of degrees less than crown of tooth, involute cutter that fit outer end of tooth. Lasted 15+ years first time. Second time is about 10 years in. Bought replacement pinion gear each time.
 
Lasted 15+ years first time. Second time is about 10 years in.
You should see the shit I used to do to the tires on my Eldo, too* ... but can't say I'd recommend that to any innocent bystanders :D

* We're not even going to mention the brakes, they may repeal the statute of limitations just to put me behind bars.

I might do this, instead of spending the rest of my life profiling both sides of 56 teeth with a hand file ...

 
Last edited:








 
Back
Top