PDA

View Full Version : Cutting hypoid gears



gbritnell
12-14-2012, 08:10 AM
Has anyone ever had any experience cutting hypoid gears in the home shop? I have an article in my old Live Steam magazines by Kozo Hiroaka on how to cut skew bevel gears. The reason I ask is because I saw a short video on Youtube that showed a fellow cutting what I assume was a hypoid gear with some type of compounding attachment for a lathe.
I am planning the next stage in my 302 engine, T-5 transmission project by making a scale 9 inch Ford differential.
Thanks,
gbritnell
http://img.photobucket.com/albums/v43/gbritnell/302%20FORD%20V8/302e.jpg
http://img.photobucket.com/albums/v43/gbritnell/Borg-Warner%20T-5%20transmission/TRANSFIN1.jpg

willmac
12-14-2012, 12:05 PM
There have been some videos of 'home' setups which the contributor claimed were for cutting hypoid gears. I have not yet seen one that would produce a correct hypoid or even a reasonable approximation. The contributors did not understand what the generating process for a hypoid actually looks like. This is an example. https://www.youtube.com/watch?v=8aUQM-VtElo
I don't recall one using a lathe attachment and I can't find it on Youtube, so maybe that setup does work, but I am doubtful.

I think that doing this properly will be a very big challenge indeed, but if anyone can do it it will be you! I would start by seeing if you could make something that mimics the way a Gleason hypoid cutter works. It might be possible to simplify this enough to make it feasible, but I haven't thought this through at all. I suspect that there would be as much work involved in that as your original engine project, but if successful you would have established a method that others would certainly want to follow. I for one hope that you have a go at this and will be watching progress with a lot of interest.

JoeCB
12-14-2012, 12:45 PM
My first real job after Engineering school was production foreman in the hypoid gear department at GM Cadillac Motor Car in Detroit. If you have ever seen the compound action of a Gleason Hypoid generator as it rolls both the gear blank and cutter you pretty much would give up on trying recreate that action on your own. To the best of my knowledge ,The Gleason Works in Rochester NY remains the only source for these machines in the world. I think that the gear being cut in u-tube video is a spirial bevel gear.
Joe B

PS, beautiful work on the engine and transmission !!

willmac
12-14-2012, 01:55 PM
I think there are other makers of hypoid gear cutters these days, but the Gleason process is certainly going to be the most accessible in the US at least. My experience was something like JoeCB's, but Chrysler rather than GM. I had a look for a Gleason machine making a crown wheel. This one is short, but it does give you an impression of the generating action:
https://www.youtube.com/watch?v=8wpZJ75g3Wo This is going to be really tough to replicate, but I wouldn't give up on this project for several reasons:

1: I think that a less general machine than the Gleason could be made in small scale. There are a number of different settings and capabilities that you wouldn't need if you aimed at producing just one size of gear.

2: The Gleason machines that I am familiar with were totally mechanical and complex. I think you could eliminate a lot of that complexity with direct drives and some electronics to synchronise rotations.

3: You are presumably looking at a producing a hypoid gear set that works in the sense that it transmits rotation. Noise, life, efficiency etc are not a concern. This means that you can get away with a result that would be totally unacceptable in automotive production, where they would be expected to be near silent about as efficient as theoretically possible and have a lifetime under the worst possible conditions that should exceed the rest of the vehicle. That means you could take some liberties.

Having said all that it is still a massive undertaking - just making a cutter (on the left in the video) that would work would be truly amazing.

Jaakko Fagerlund
12-14-2012, 03:12 PM
Thanks, now my head hurts when I looked at those Gleasons making hypoid gears and just trying to picture it in my mind...

saltmine
12-14-2012, 05:05 PM
Just off the top of my head.... Possibly having the gear blank mounted on an indexer, attached to a pair of ROTABS, one on top of the other, would generate a hypoid gear pattern. Adding CNC controls would prevent a nervous breakdown or going crazy on the part of the operator.
I've seen a similar setup to cut turbine blades.

kitno455
12-14-2012, 05:16 PM
Instead of making one, how about repurposing the spiral bevel gearset from an angle grinder? they are about 3.5:1

allan

Mark Rand
12-14-2012, 06:42 PM
Probably because with a hypoid gear set the axis of the pinion does not intersect the axis of the gear. Whereas it does with a spiral bevel gear set. Presumably the prototype that is being modelled requires that specific feature.

willmac
12-14-2012, 06:44 PM
Spiral bevels would be a bit easier. The problem is that for a proper old fashioned back axle you need hypoid gears. The difference between the two would be obvious at a glance from the appearance of the back axle casing. Now if you were making a front wheel drive transmission it would be a different matter, but the gearbox already built is obviously intended for a conventional rear wheel drive back axle, and given that it is a traditional American power train that would almost certainly be a live axle.

mike4
12-14-2012, 07:04 PM
Is there any where that the gear train calculations can be found as that looks very useful , I often have to buy (big$) little gears for pump or tacho drives , making your own would be cheaper and often quicker.
Michael

willmac
12-14-2012, 07:25 PM
Is there any where that the gear train calculations can be found as that looks very useful , I often have to buy (big$) little gears for pump or tacho drives , making your own would be cheaper and often quicker.
Michael

Yes, Gleason has lots of documentation and work sheets and software that you use to design a hypoid gear set and set up the machines. Having said that, you probably would not make any progress with them unless you were a real gear guru and in addition had some specific training. Even then, you would probably need support from Gleason or Klingelnberg or whoever's systems you were using. Then you would need the cutters for your gear; I'm not sure how you would even start on that. This is not a trivial exercise and nothing like setting up to cut an ordinary spur gear.

mike4
12-15-2012, 12:20 AM
The gear train in the video where the guy was cutting gears similar to pinions was what I was asking about , difficult maybe impossible not.
People managed to cut reasonable gears on manual machines before CNC and specially designed tools , so what if it takes a bit of time to set up , may be quicker than three months from an overseas manufacturer who asks over $400 for a 30mm dia gear with a slight taper .

The more that someone treats another as an untrained hobbyist the more determined I become.
Michael

tyrone shewlaces
12-15-2012, 01:29 AM
... you probably would not make any progress with them unless you were a real gear guru and in addition had some specific training. Even then, you would probably need support from Gleason or Klingelnberg or whoever's systems you were using. Then you would need the cutters for your gear; I'm not sure how you would even start on that. This is not a trivial exercise and nothing like setting up to cut an ordinary spur gear.

"Piffle !!!
Gimme a screwdriver and a hacksaw 'n get ready fer some learnin'."
-Granpa

mike4
12-15-2012, 02:10 AM
The first video was the method that I was interested in (https://www.youtube.com/watch?v=8aUQM-VtElo ) that showed the type of set that I am very interested in , Manual and done in a normal shop.
The only difference would be my finished products would be hollow.
Michael

Paul Alciatore
12-15-2012, 05:04 AM
If power transmission and silent operation at high speeds are not requirements, perhaps a method could be worked out where each tooth space were cut with just two, three, or four different settings instead of the precise, continuous generation action of the Gleason machines. This would be somewhat similar to generating a bevel gear, which has a constantly changing cross section, with a single pitch cutter by using it at two different angles. The bevel gears produced this way are not ideal, but they could work and the model would be animated. Likewise, a less than perfect "hypoid" may be possible.

Since this would utilize several static setups to produce each tooth instead of the continuous generating action of the Gleason, the cutter could be made with a single insert instead of many. This would simplify the design of that cutter. Of course, the cutting process would take quite a lot of time with multiple passes for each tooth.

Just a thought.

Another thought: Perhaps it may be possible to make multiple generations of each gear in the set. Start with the smaller (pinion?) and generate it as best as possible using the above method. Then convert that gear into a cutter (by gashing and relieving it) to cut the larger ring gear by driving them at the final speed ratio (like a hobbing setup) and bringing them slowly together. The figure of the large gear would be better than that of the small cutter due to averaging effects. Then reverse the procedure and convert the large one to a cutter and generate a second generation small gear. Continue for several generations until no further improvement is achieved. Use the last two tools generated in this manner to cut your actual, final gears.

Again, just a thought. I have no evidence that the figure of each generation would actually be better than the previous. But it seems to me that it would. Yes, I know this would be a very laborious process but so would any other that you may attempt.

mike4
12-15-2012, 05:15 AM
Please watch the video in the link , those are the gears which i need to make, in tha video there is a set of gears going from the milling table to the dividing head. those are what I need to set up . No interest in cnc at the moment as I dont have a spare$100k for a basic useful machine installed on a ready to run basis.
Michael

willmac
12-15-2012, 07:51 AM
mike4-

We seem to be at cross purposes. The OP, gbritnell, specifically wants to make hypoid gears for a good model of a back axle. Most of this thread is discussing how that might be done, difficult though it is.

The video that you reference does NOT show the making of a hypoid gear. Basically, the video shows spiral milling, and there are plenty of references for that. To mimic the video You need a universal dividing head geared up to the table leadscrew with the correct ratios, or some electronic equivalent. Unfortunately,spiral milling on its own does not make hypoid gears. At best, it is a crude form of spiral bevel, but I do not think that the geometry is close even for that. . As Paul A points out above, the cross section of the teeth of any bevel gear need to change continuously from the OD to the ID. The setup in the video does not appear to satisfy that requirement. Perhaps you could file up the gears after machining them like that and get something that would work for very low speeds - I have no idea how well this would work, but I suspect roughly and with a very short life.

You are quite correct that people managed to make good gears way before CNC was common. However, special purpose machine tools were used for gear cutting more than a century ago. The Gleason machine in the video clip is definitely not CNC, nor are ordinary hobbing machines. The breakthrough that allowed the use of hypoid gears in rear axles was just that - it revolutionised automotive design and production and it relied on a whole range of special purpose machine tools and what was for the time and even now, rather complex technology.

Nobody is doubting your skills or experience - I would be really interested and extremely impressed by anyone who can come up with a workable way of making even poor hypoid gears on typical home shop machinery.For what its worth, I think that Paul Alciatore's ideas may be on the right track, but there is a way to go to get it working.

willmac
12-15-2012, 08:17 AM
This software looks interesting, but for spiral bevel gears only unfortunately: http://spiralbevel.com/yahoo_site_admin/assets/docs/SPIRAL_BEVEL_CORPORATION.361220306.pdf
This allows you to define a spiral bevel gear via a spreadsheet worksheet and get a generated model in SolidWorks. From this you can generate Gcode via a CAM package for a 3 axis CNC mill, with or without an additional indexing axis. 5 axis would be better, but it looks feasible to make at least crown wheels with a 3 axis mill. Extending this to hypoid gears might well be possible, but it would need a real gearhead to do make the software for it. These are not generated gears, but if you took sufficient time with a good setup the results should be workable.

ideally Gearotic would do this type of gear, but as far as I know it can't yet - perhaps if John S is reading he could comment?

Machining pinions this way might be a little more difficult; I'm not sure that a 3 axis mill would be able to do it, but I could easily be wrong.

PJRitz
12-15-2012, 10:07 AM
Instead of making the scale Ford 9", how about considering a scale race style quick change rear end, and make it so it would work with standard bevel gears for the ring and pinion

The Artful Bodger
12-15-2012, 03:13 PM
I have been looking at a diagram of a hypoid gear set and it appears to me that the shape of the ring gear is much less complex than that of the pinion so my suggestion is to cut a ring gear in steel that can be hardened and use that as a hob to cut the pinion.

JoeCB
12-15-2012, 04:11 PM
Many great creative ideas here ! Assuming that the OP really wants a Hypoid gear set (pinion axis off set from that of the ring gear) and not simply a bevel gear set. I like the idea of using a "faked" hardened ring gear as a hob to cut the pinion. Perhaps a pinion in brass might serve the purpose. That would be an interesting project.
Joe B

willmac
12-15-2012, 05:48 PM
That is an interesting idea and I can't immediately think of a reason why it would not work. However that still does leave the problem of making the ring gear. Perhaps two ring gears; one to use as a hob and the other less gashing for cutting edges to use for real. The CAD/CAM CNC route would allow you to make the ring gear with just a 3 axis CNC mill if you could find someone to model it for you, but that is maybe not what the OP wants. The hobbing arrangement would need to be set up from scratch - I don't think a conventional hobber could be set up for this.

The Artful Bodger
12-15-2012, 06:27 PM
On the other hand, I assume the teeth of the pinion enter at the edge of the crown wheel and slide through to the cente?

If this is so then every point along the pinion tooth passes every point along the crown wheel groove that it engages with?

Perhaps then if a flat disk was substituted for the crown wheel and a rotating (shaped) milling cutter was fitted though the flat disk (in the position otherwise occupied by a single point along a crown wheen tooth) and the disk rotated while a pinion blank was brought into the relative position the cutter would cut the required hypoidal grooves in the pinion blank.

If you can make a pinion you can make a hob for cutting the crown wheel.

Maybe?

oldtiffie
12-15-2012, 08:25 PM
http://en.wikipedia.org/wiki/Spiral_bevel_gear

http://www.google.com.au/search?q=making+hypoid+gears&hl=en&tbo=u&rlz=1W1IRFC_enAU360&tbm=isch&source=univ&sa=X&ei=bBTNUNjHI-iaiAeC9IHwDg&sqi=2&ved=0CDAQsAQ&biw=1920&bih=846

http://www.google.com.au/#hl=en&sugexp=les%3B&gs_nf=3&gs_rn=1&gs_ri=hp&cp=19&gs_id=24&xhr=t&q=making+hypoid+gears&pf=p&tbo=d&output=search&sclient=psy-ab&rlz=1W1IRFC_enAU360&oq=making+hypoid+gears&gs_l=&pbx=1&fp=1&bpcl=39967673&biw=1920&bih=846&cad=b&bav=on.2,or.r_gc.r_pw.r_qf.

JoeCB
12-15-2012, 10:55 PM
Thanks for posting the three links. I espically liked watching the Gleason machine working on the hypoid pinion... brings back memories. The actual cutting process involves three distinct steps, each with it's own specific machine and cutter. The first step is cutting the rough tooth shape in the blank, then roughed blank is moved to one finisher where one side of the tooth profile is finished, either drive side of coast side. the semi-finished blank is then moved to another finisher that does the opposite face of each tooth. The finished gear is then heat treated. The final step (at least at Cadillac) was to abrasive lap the ring gear and pinion, producing a mated set. After a through washing the matched sets came to a temperature controlled room where the gears were run together under load, skilled operators adjusted the clearances for best running, pinion and gear were then marked for proper shimming at assembly into the differential carrier.
I don't know how other auto makers processed their "rear ends" but at Cadillac back in the day, quiet differentials were of paramont importance.
Joe B

oldtiffie
12-15-2012, 11:28 PM
Is there any where that the gear train calculations can be found as that looks very useful , I often have to buy (big$) little gears for pump or tacho drives , making your own would be cheaper and often quicker.
Michael

See Machiery's Hand Book 27 - pages 2026 - for all typesof gearing.

See pages 1964 > for "Milling, change gears for different leads".

See pages 2081>2095 for "Bevel gearing".

See pages 1962> for "Thread milling - includes "Helical Milling".

willmac
12-16-2012, 08:26 AM
I don't know how other auto makers processed their "rear ends" but at Cadillac back in the day, quiet differentials were of paramont importance.
Joe B

Yes, that is how I remember things as well. I think that most auto makers were similar because typically Gleason were involved heavily and supplied many of the key machine tools. I particularly remember the heat treatment process for crown wheels, which was very important for life and accuracy if not done right. We used carbo-nitriding, followed by press quenching for this part. For anyone not familiar with press quenching, it is a way of minimising distortion that would otherwise occur when quenching steel parts. For crown wheels it used a Gleason special purpose hydraulic press with anvils which were specifically designed for the specific wheel, with a complex pattern of oil channels in their surface. You loaded the hot crown wheel (by hand!) with a steel hook, then the press closed and oil was flushed through the press anvils under pressure. Not light work in a high volume plant.

gbritnell
12-16-2012, 08:54 AM
I would like to thank everyone for their input on this thread. Over the years I have read about and cut many types of gears, for my model work, with much success. I have watched videos relating to the Gleason machines and understand the complexities in creating hypoid gearing. With this wonderful world of information available at the tap of a key I thought that there might be someone lurking out there that had performed a satisfactory construction of this type of gearing. As I stated when I started this topic I have the series of articles by Kozo Hiraoka published in Live Steam regarding the construction of skew bevel gears. Since this project will be a reasonable facsimile of a 9 inch Ford differential I'm sure the skew bevel gears will work just fine. Mr. Hiraoka gives 3 pages of formulas to machine a set of gears and I am going through his numbers and plugging in my numbers but so far don't seem to be getting the same pictorial comparison in AutoCad.

I do have one question for anyone who had knowledge of these type of gears. In the calculations there is a formula for calculating the tooth offset number, not the shaft offset number. If the shaft offsets is already known what is the reason for the tooth offset number?

Thanks,
gbritnell

J Harp
12-16-2012, 09:27 AM
You might find this of interest.

http://ansol.us/Publications/MartinoVimercati_05FTM05.pdf

Black Forest
12-16-2012, 02:24 PM
Has anyone ever had any experience cutting hypoid gears in the home shop? I have an article in my old Live Steam magazines by Kozo Hiroaka on how to cut skew bevel gears. The reason I ask is because I saw a short video on Youtube that showed a fellow cutting what I assume was a hypoid gear with some type of compounding attachment for a lathe.
I am planning the next stage in my 302 engine, T-5 transmission project by making a scale 9 inch Ford differential.
Thanks,
gbritnell
http://img.photobucket.com/albums/v43/gbritnell/302%20FORD%20V8/302e.jpg
http://img.photobucket.com/albums/v43/gbritnell/Borg-Warner%20T-5%20transmission/TRANSFIN1.jpg

Wow, I have never seen a micrometer that big!