Making a double enveloping (globoid) worm and gear

has anyone got any ideas on making a double enveloping (globoid) worm and gear, in the home workshop, i have a lathe and a mill with diving equipment to do the job, i have a tool grinder and optical profile grinder to make the cutters, i could make a machine to do the job but want to avoid this if i can, your advice would be very much apreciated
Cheers

Wow globoid, they sound complicated things,you say you have the mill,the dividing head with gears, the grinder to make the cutter,i dont understand why you need to make a special machine,unless you are going to bo making them quantity.i,am sure you going to some good info here.

alwyn, note he said "diving" equipment....this is way more complex when done under water !

struth i got it wrong,sorry,yes you are right it would more difficult under water

Ignorance rules.

What is it please?

Jim

Here's a search I did to find out what it is.
http://www.google.com/search?hl=en&q=globoid+worm+gear&btnG=Google+Search

If you download the pdf file for the first item on the list ( 0301 MSD BOGE.qxd), and look at the fourth page, there's a picture of one.

It's a way to get more tooth area in contact to increase the torque capacity.

Roger

The trick will be to get the cutting tool to make a radial sweep thru the worm blank. Sounds like very special attachment for lathe with its rotation geared to spindle. Sort of a reverse of the idea of cutting a worm wheel with a worm type hob, or simulation thereof, like metalmuncher used in his CNC lathe set up. In this case you need to cut a worm with a worm wheel shaped cutter, even if it only has one "tooth".

John

Thanks Roger - not familiar with the terminology

Get on ebay and get a Textron Cone Drive gearbox. The worm is already split and the sides preloaded against each other for zero backlash.

-Matt

He asked for ideas, so here goes

It goes without saying that a special toolpost would have to be built. The toolbit would mount on the end of a lever. The lever has a hole in the middle, spaced properly from the toolpoint, to match the radius of the worm wheel.

The opposite end of the lever also has a hole in it. This hole has a tie rod fastened between it, and a solid mount block affixed to the lathe bed. The tie rod should be parallel to the lathe bed, when the tool is at 90° to the worm blank, and when the toolbit is at full depth in the worm.

I'm not sure of the exact radius between the tie rod hole and the pivot hole in the lever. They should be "the same" as the toolbit swing radius, whatever that is. Best make this adjustable

In operation, you would use the normal halfnut threading procedure on the lathe carriage. However, the ratio of carriage travel per thread would have to be reduced by a factor of Pi. This is because, if a wheel rolls one revolution along a surface, its circumference rolls a distance of Pi*D, but the center of the wheel moves only a linear distance of D

In this setup, the pivot point of the lever occupies the center of the worm wheel position, therefore its rate of travel must be reduced as described. This could be very inconvenient to calculate change gears for, using a manual lathe. Not so bad doing this on a cnc: you can have whatever you want for a feedrate.

Glub glug glug burble burble, its awfully wet over here, gawsh me andd mi spelin, i told you i was a lixdexick illiterate haven't I?

John, this is the way it should be done, or so i am led to belive, with controlled relationship between the cutting "wheel" and the worm, then a globoid worm is used to cut the wheel, but for a prototype job it seems a lot of trouble to set up, but i guess i may have to

Matt, the cone drive wont fit my requirements, so i'll have to make my own

HuFlungDung, one of the research professors at Nottingham Trent University has done it this way, i think? i reckon this is what i'll probably end up doing, but i would like to modify the worm geometery for experimental purposes, just to compicate thing you know

thanks you guys, please keep em comming

This site may be of help. Click on the picture below:

On second thought, maybe you would not have to worry about special change gearing on your lathe. Genuine diametral pitches, such as 10 DP amounts to 10 teeth, in Pi inches of circumference, so in this case, you want that ratio to revert back to 10 teeth in 1 inch of travel of the worm wheel center. Therefore, to cut a 10 DP worm, simply feed the carriage along at 10 tpi. The rotation of the lever around the pivot will take care of the Pi factor automatically, I think.

I'm open for rebuttal. I've not done any of this in real life to discover what the kinks are. Like that article referred to above says, I don't think this worm is going to mesh with a standard worm wheel, without cutting some additional clearance on the worm wheel teeth. This is because the helix angle of the globoid worm is variable, whereas on the worm wheel, it can only be a fixed angle. By definition, the helix angle is a ratio of the pitch circumference divided by the helix lead, so, since the globoid worm has to increase in diameter as it wraps around the gear, this causes a continuous change in its lead angle.

HuFlungDung:

I think the lever idea would create non linear rate of travel around arc due to changes of angle of lever.

Now a rod clamped to the bed with a rack gear attached to its end driving a pinion would do some good.

John

John,
I suppose you are right if the tie rod is very short. Solution? Make tie rod very long

Maybe a "pantograph" style of tie rod would work better.

Making a globoid worm in the home workshop

deckelnut said:

has anyone got any ideas on making a double enveloping (globoid) worm and gear, in the home workshop, i have a lathe and a mill with diving equipment to do the job, i have a tool grinder and optical profile grinder to make the cutters, i could make a machine to do the job but want to avoid this if i can, your advice would be very much apreciated
Cheers

Click to expand...

Hello,
This thread may well have gone cold but we clearly have a common interest. I have built a Jacobs gear hobbing machine which starts as a set of castings from College Engineering Supply. I have adapted it to make helical and worm gears as well as spur gears. One of the items I have made is a double enveloping (or globoid)worm and wheel. See photo, the globoid item is in the foreground. As chance would have it I have an article published in Model Engineers' Workshop which covers this. It will be published in parts with the first part in the April issue, number 253.
Best regards, Chris

clrob3437 said:

This thread may well have gone cold but we clearly have a common interest. I have built a Jacobs gear hobbing machine which starts as a set of castings from College Engineering Supply. I have adapted it to make helical and worm gears as well as spur gears. One of the items I have made is a double enveloping (or globoid)worm and wheel.

Click to expand...

Nice

In case anyone is interested, the way these were/are made in production is a variation on the hobber theme. Michigan Tool built gear machinery and also made and sold the Cone Drive gearboxes. So they made their own machines to build them. Textron bought Michigan Tool when they went tits-up, and saved the double-enveloping Cone Drive product line.

Imagine a standard gear hobber. Put a large (same size as the wormgear) shaper cutter on the table. The teeth are thinner than standard. Put the worm blank where the hob normally goes. Feed the hob slide down onto the cutter mounted on the table to get your threads on the worm. Now the cutter on the table is advanced a ways, then retarded a ways to get the correct toothspace. Then the hobslide is moved up, hourglass worm removed, viola.

To cut the wormgear, similar idea. An hourglass-worm-shaped thin-tooth cutter placed in hob slide and fed into the wormgear blank radially, just like normal infeed wormgear cutting. Then the blank is advanced a ways, then retarded a ways, to get the correct geometry. Then the tooth space is big enough to withdraw the cutter.

They claim an advantage of these is higher loading but I don't see it. In order to assemble them the wormgear has to be split, assembled, then the teeth spread to contact the worm on both sides. So you only have half as much wormgear in contact, so what's the advantage ?

For low-backlash tho, they seem to work well. The Michigan hobbers used double-enveloping worm sets to drive the table. And at least one high-power gear gasher uses them to drive the cutter, so no chatter on some pretty big teeth, but the advertised "higher load" thing doesn't seem right.

They do make more heat. That gear gasher has a separate chiller just to cool the wormgear and the Michigan hobber had an enormous lube oil tank and humongous feed to the wormgear for the same reason. For an 8" machine, the wormgear on that thing was enormous. It must have been 12" in diameter ? And more than two inches wide ? the change gears for that thing weighed fifty pounds ...

Oh, history ... in this case, Cone Drive is not just a marketing term. John (?) Cone invented the things. He either sold it, or worked for, or was connected in some way to Michigan Tool who developed the idea. So they really owned the whole concept. Cone Drive isn't a later add-on name, it's more like Fellows, who invented the gear shaper.

Hello SeaMoss,

The process you describe is pretty much how I did it. A gear was first hobbed with backed off and sharpened teeth the same DP and tooth count as the worm teeth, in other words a gear shaper like the Fellows company pioneered. This was used to cut the globoid worm by reversing the hobbing process using plunge feed. The worm wheel was hobbed in the normal way with the same hob as that used for cutting the "shaper"., Now I realise that this will not give perfect tooth geometry but it's fine as a demonstrator of the process.

I think Michigan Tool's claim for higher load capacity is valid, even if only half the worm has contact. The basis for this is that a single enveloping worm will have just one (or perhaps two fleetingly) line contact between the worm and the wheel. In the double enveloping design it will have multiple line contact. So if the worm envelopes say 6 teeth there will normally be 6 line contacts at least theoretically. This spreading of the load over multiple teeth is what gives the design more load capacity. In the case of the split anti-backlash design you mention it will be less with say three teeth having line contact but still more than a single enveloping design. Best regards, Chris.

clrob3437 said:

I think Michigan Tool's claim for higher load capacity is valid, even if only half the worm has contact. ... So if the worm envelopes say 6 teeth there will normally be 6 line contacts at least theoretically.

Click to expand...

You know, you are probably right. Normally you'd have about two teeth in contact, so if only half the face width of a split double-enveloping are in contact, that's still 6/2=3 which > 2 ...

Thanks. Good arithmetic lesson