Gear cutting.

[John Stevenson]

Here are two shots of a couple of gears we cut at the recent Harrogate show in the UK.




Is a helical gear cut with a D bit as a cutter.


Is a double helical, usually know as a herringbone, cut in one pass and then the centre part removed as a groove to prevent interference.

.

[oldtiffie]

Gearing

Hi John,

a good enjoyable lesson and demo in every way - as always.

As I can't see a compound dividing head (n)or any indication of a universal mill, I guess that it is a "multi-axis (dividing head included) CNC" as job as well.

You really do have a limitless talent for this sort of thing don't you?

I sometimes think you come down on that big dark cloud that I'm under.

Did you mill them in one go or multi-passes?

If "multi", I'd guess that it would have been a very interesting problem (and in your case, solution) as regards generating a smooth involute curve. The single tooth (as in "engraving tool") D-tooth cutter would have been an interesting tool and cutter grinding problem as well.

As I've been made aware, herringbone, like any helical gear, has at least one tooth in mesh at any one time and due to side-thrust will always push the gears toward the end of the shaft they are on every time - dependent upon direction of rotation, of course, and that size-for-size that they can carry bigger loads faster than any comparable "straight" gear.

I've been in some very interesting discussions regarding single and helical gears in high speed (read steam turbine and diesel-driven) high-load (read multi thousand shaft horse-power main engine reduction gear-boxes) in a marine environment (Engine Rooms et al). (Navy Destroyers and Aircraft Carrier).

For the record (and the Dick Tracy's out there) I was not an "Engine Room Tiffie" - but I had an interest in these things.

It seems that first of all, helical gears in a "herring-bone" set-up if exactly matched as regards side/end thrust if equal will cancel each other out in this regard.

There was some discussion on which direction the gearing should run under the primary load direction (these gear-boxes are reversible).

I seem to recall that the consensus of advice was that the "points" should go into mesh first and have the forced feed lubrication oil stream fed onto the back of the engaging teeth so that they acted as gear-pumps and forced the oil under load to be pumped from the centerer groove toward the outer faces to ensure full-face lubrication. Excess oil just fed into the groove and was returned to sump and pump on a continuous basis - as in an automotive engine if you like. (Or WAS it the other way?).

The oil, as is often the case was a coolant as well - HSM machine gear-boxes, automotive engines, diffs and gear-boxes for instance.

But in those pre-CNC/digital days in (pre) Jurassic Park times, it was a very highly skilled, involved and expensive task to generate helical gears - special pre-formed/profiled cutters on a horizontal universal mill with a lead screw-driven universal dividing head.

I'm bl**dy certain that Mr. Heath-Robinson Esq. - of some renown - "cut his teeth" (as it were) on helical gear milling, gear-hobbing, generating, grinding, lapping etc. before going on to more esoteric things.

Those who did these things in the shop I was in could only be approached through intermediaries who "spoke in tongues" and called on all sorts of invocations and burned incense. Well, may not quite - but it seemed like it. Can't recall if genuflection was required - possibly was.

In and still am.

I seem destined to never get out from under that bl**dy cloud.

[j king]

Neat John. You are really talented.

[NAMPeters]

Sir John, inventive as usual. What is the status of the electronic hobbing method for gears?

[John Stevenson]

Sorry I forgot to add that they were done on the X3 CNC mill we had on show.
Technically they were a fudge as the cutter doing to job was a straight sided cutter ground up to make a rack section on an earlier job, it was just handy.

No reason why this couldn't have been done with a D bit round to an involute. I did one ages ago for a small reduction gearbox on the big mill but I can't find the cutter used for that.
Having said that and being small, about 1-3/8", no one noticed the profile was wrong

Cut in brass, one pass, about 95 thou deep, hardly a taxing cut. The herringbone was also cut in one pass, skew thru the blank halfway, then reverse the rotary table, it leaves a tooth with a sharp leading vee but a radius on the back side hence removing the centre section.

It's very easy to do these on a CNC as the code is only about 20 - 30 lines long and much of it can be edited to do other gears.
We took a video of it cutting these two gears and a larger one in steel with a normal B&S cutter, 16 DP 22 teeth, 159 thou depth of cut in one pass.
We then had these showing on the machine screen when the machine was unattended at the show. Need to video more jobs.
I'll ask small son if we can't take a section out of the video and upload it to our 'dump'

NAMPeters,
The electronic hobbing is still at the same stage as it was. As far as I'm concerned it's a completed item and ready to use.
I use mine probably twice a month doing gears for Gert to sell on Ebay.
I have no idea on the number it has cut in total but it's a lot. I usually stay over in the workshop of a night and run some off, for the Myford 20 and 21's which actually use the same blank made from 30mm bar I can usually bore ream and gear cut about 40 in a night in sticks of 10 on a mandrel.
Keywaying usually has to wait until the next day as the arbor press is outside and it's dark when I get done

Brian Thompson wrote the article up for MEW in vol 108 I think it was and published the board details so anyone can build one now, nothing is with held.

.

[NAMPeters]

Thanks Sir John, just checking to see that I have not missed anything before I press forward and try my hand at making one.

p.s.
Went looking but could not find your MEW108.pdf file, got a file not found message.

[TGTool]

Brian Thompson wrote the article up for MEW in vol 108 I think it was and published the board details so anyone can build one now, nothing is with held.

.

A great article. Just the idea blew me away. Kind of like the time, years back, when we got our first NC machining center. I'd run Moore tools, read their books on mechanical accuracy and the lengths they went to getting it. I was VERY impressed. Then I read the service manual on the NC. Instead of going to the lengths Moore did to get an accurate leadscrew, they just made the leadscrew, measured the inaccuracies, fed them into the controller's compensation table, and when the table moved to a certain position the controller fed in the compensation between the inaccurate mechanical position and corrected location. Sheesh! Alexander and the Gordian Knot.

[Alistair Hosie]

huh show off anyone could do that Alistair I'll stick to playdough

[John Stevenson]

huh show off anyone could do that Alistair I'll stick to playdough

It shouldn't be hard for you.........you are a dentist

.

[oldtiffie]

Corrections for DRO and ball-screws

A great article. Just the idea blew me away. Kind of like the time, years back, when we got our first NC machining center. I'd run Moore tools, read their books on mechanical accuracy and the lengths they went to getting it. I was VERY impressed. Then I read the service manual on the NC. Instead of going to the lengths Moore did to get an accurate leadscrew, they just made the leadscrew, measured the inaccuracies, fed them into the controller's compensation table, and when the table moved to a certain position the controller fed in the compensation between the inaccurate mechanical position and corrected location. Sheesh! Alexander and the Gordian Knot.

Thanks TDT for the reminder and the "heads up".

Marv Klotz at his web site http://www.myvirtualnetwork.com/mklotz/ has a utility for measuring and correcting (for) such error on DRO's.

I'd guess the principle would be much the same for correcting errors in a ball screw. Whether it could be Incorporated in the software for CNC is another matter that you would need to address to Marv K (address on his web site)and John Stevenson (in this thread and on this Forum).

I guess that "having the machine but not the manual" occurs fairly often in this forum, and perhaps your need may be sorted out by looking at Marv K's item and getting further advice from John and others on this forum.

I don't have CNC or ball screws. But I have thought of applying the same principles to my lead screws. I have all the DRO's still on the shelf - but not yet fitted - in my workshop.

I am leaning more toward the ball screw-cum-DRO-cum-CNC "fit" all the time.

I hope Marv K reads this thread/post and responds to it.

[darryl]

Just thinking about those herringbone gears- it seems to me that arc shaped teeth would be fairly easy to produce, as it could be done with something equivalent to a hole saw, but with involute shaped teeth. You'd still need a rotary table or some method of indexing the blank, and you could just lower the mill spindle with this tubular shaped cutter to a preset depth to hog out the teeth from the major diameter to the gullet (maybe my terminology isn't right).

I'm not trying to take away from your elegant cnc method, just trying to see how someone could make pseudo-helical teeth on manual equipment without having more than a method of indexing the blank.

Seems to me the arc shaped teeth would behave to some extent like a helical tooth (smoother than a straight tooth) yet have the pressure balanced feature that herringbone gears give.

[oldtiffie]

Spiral gears

Just thinking about those herringbone gears- it seems to me that arc shaped teeth would be fairly easy to produce, as it could be done with something equivalent to a hole saw, but with involute shaped teeth. You'd still need a rotary table or some method of indexing the blank, and you could just lower the mill spindle with this tubular shaped cutter to a preset depth to hog out the teeth from the major diameter to the gullet (maybe my terminology isn't right).

I'm not trying to take away from your elegant cnc method, just trying to see how someone could make pseudo-helical teeth on manual equipment without having more than a method of indexing the blank.

Seems to me the arc shaped teeth would behave to some extent like a helical tooth (smoother than a straight tooth) yet have the pressure balanced feature that herringbone gears give.

Interesting and creative thought darryl

Set the processes going in the cess-pit that passes itself off as my brain/mind.

As you say, herringbone gears (called after the "herringbone" pattern in the materials that men's suits were made from long ago) are often just two opposite spiral(led) gears fitted together.

The basic problem is that the spiral gear is essentially a short length of a multi-start thread.

That's why/how John made them on a mandrel on his mill. The number of "starts" are the same as the number of "slots" (between the "teeth" (threads)) in the spiral gear.

Continuing on from that it almost seems that it could be "lathe screw-cutting job" - and it can - if you get the "lathe tool" correct (use a sample and a template) and then work out the "screwing gear ratio/s" for your lead-screw.

In fact, worm screws can often be made on a lathe if the spiral is "steep" enough. A worm is just another screw thread or spiral gear after all.

You can probably get get the "screw" "pitch" and "lead" from the "sample. But its going to be one helluva load on the gear-train. But, providing you have the gears and take it easy, you might just make it.

Now, one of the "nasties" that have plagued threads on this Forum" the dreaded "gear-chaser dial" and its use.

For a good source of info and software, go to Marv Klotz's web site at:
http://www.myvirtualnetwork.com/mklotz/ as he has utility called "stick" which makes this process a whole lot easier (less difficult??).

Remember, this is only multi-start screw-threading under another name.

I hope it helps.

[John Stevenson]

Herringbone gears are just two helical gears together, either put together or machined as a pair.
Some have the central clearance groove and some don't.

The solid ones are harder to make because of the clearance required by the cutter, they can't be hobbed or cut with Brown and Sharpe type cutters because of the run off required for the cutter.

They can be cut on a special Fellows type machine with helical slides but this is very rare as normally these are designed for large reduction units bigger than a fellows can take.

Most are done of Sunderland type gear cutters which are actually planers.
They use a section of rack as a cutting tool and plane 5 to 7 teeth at a time, then the wheel indexes round 5 or 7 teeth and the cutter backs off to do the next cut.

By inclining the cutting slide it's possible to do helical's. with the addition of a second slide it's possible to to do herringbone gears.

The second slide on these isn't unusual as they often do double blanks of gears on the same mandrel with two cutters.



Two pic's, the top one is cutting double helicals and you can see the inclined slide and the section of rack as a cutter.
The lower one shows the double method of cutting .

.

[oldtiffie]

Mr. Heath-Robinson - a mark of his presence.

Thanks John.

That's the sort of gear box I was talking about - and the planers are almost a patent taken out by Mr. Heath-Robinson (Esq) or is it (Bt)?.

[A.K. Boomer]

It seems that first of all, helical gears in a "herring-bone" set-up if exactly matched as regards side/end thrust if equal will cancel each other out in this regard.

.

Right on Tiff, as a mechanic the herring-bone gear is a thing of beauty, they do indeed cancel each other out --- An HB is the best of both worlds, zero side loads and no running noise, a straight cut has zero side loads but has running noise (just put your manual transmission car in reverse and accelerate)
A helical removes the running noise (due to having two or more teeth in contact at all times) but at the expence of creating side loads, For the record, I remember being impressed with Porsche for the reason that while into one of thier 1985 911 transmissions I found that thier 5th gear helical had needle bearing side thrust washers, A nice touch for those high speed rocket cruises on the autobaun...

Further more, An HB has the ability to operate without any added side thrust mechaniszms, the gear itself will take care of this, As long as one shaft is of standard design, it means that you can actually have the gears run on shafts that utilize roller bearing and no thrust washers at all, You cant even do that with a straight cut, HB's --------- you gotta luv em.

Nice work John.