epicycloidal gears

Seeing as you are replacing both gears, why not update the profile?

Tom

The conservators(sp?) have decreed that we must stay as close as possible of original specifications. I agree it would be a lot easier but THEY have spoken! I suppose there is some merit in it as it was the last stem roller built in Australia.

Tom Miller

How old is this machine? late 20s or 30s?

Tom

'Can't help you with standards, but there is one thing to watch for with cycloidal gear forms: the between-axes distance of the meshing gears is part of the equation. In other words, unlike involutes, which give you a little slack in this department, conjugate action of cycloidal gears depends on getting the between-centers distance of the shafts dead right. If you re-cut or refine epicycloidal gears, thus changing their effective diameter, you effectively change the required between-centers distance. It's tricky.

Good luck. You might have to dig into the heavy gear literature. There is a gear society around that publishes a sort-of magazine. Check with their editors.

-- Ed Huntress

========================== Good place to start is "Gears & Gear Cutting" (Workshop Practice Series) (Paperback) by Ivan Law ISBN 0852429118. He discusses both cycloid and involute tooth form and suggests ways to produce a "one-off" gear. see

I note that several Australian suppliers stock this book.

Good luck on identifying the profile. Do you have a shaper available, or will this be a lathe/mill-flycut job? Let us know how things progress.

Uncle George

1939 is on the nameplate

Tom Miller

tom, not that I understand it all, but:

is, truly, interesting if one wants to know more about the epicycloid gear forms.

Mike in BC

Tom,

Are your museum conservators technical types who KNOW the difference between the two gear profiles, or are they political appointees throwing their weight around?

If the latter case just make the appropriate drawings and tell the gear cutting outfit verbally that alternative, better, cheaper, profiles are acceptable. (The "cheaper" is to CYA)

Unless, of course, your museum has sooooo much money to throw around on this that one could make a career out of this one job. (I'm available as a consultant on this! :))

Let us know what finally happens.

Regards,

Wolfgang

I think you should order a set of involute gears as a "stopgap" measure, so that restoration can proceed, pending receipt of the proper cycloidal gears. In the meantime, the steps in acquiring the "permanent" set won't have to be pushed very hard, and perhaps will grow complete cold.

That's a standard way to handle many such difficult restorations, especially those that intend to operate the device. Fit new parts, being careful (in so far as possible) to make all necessary modifications to the NEW parts only, keeping the original parts unmodified, so the 'restoration' is completely reversible. Then keep the old parts for those that need the historical information they represent, and run the device using the new (and expendable) parts.

Dan Mitchell ============

According to Ed Huntress :

I think that it was spelled out in either _A Catechism of Steam_, or _Modern Machine Shop Practice_ -- both of which are available in scanned PDF format on the web.

O.K. Here is where to get _Modern Machine Shop Practice_, written by Rose, and published back in 1887-1888 ("Modern" is obviously relative. :-). The first three chapters are the ones dealing with gears, and each chapter is downloaded separately. They are about 150-160K per chapter, at least for the first three.

I'm looking through it now (on the computer, so it is very slow turning pages), and so far, the first chapter is defining terms and concepts, including how to apply gearing to accomplish various speed ratios.

I would recommend printing out the chapters, so you can study them at your own rate for page turning, instead of being slowed down by the PDF rendering speed of your computer.

O.K. On page 8, we start getting into the description of the cycloid gears.

And the description continues for quite a few pages, including information on how to generate the shapes.

There are even details here on how to generate asymmetrical gear teeth, for greater strength running in one direction compared to the other.

BTW The original dimensions of the gear were given in mm, but I suspect that the original was designed and constructed using inches instead, unless Australia converted to metric before I thought that they did.

The old _Modern Machine Shop Practice_ may be just what he needs, as it was written before the involute tooth form came into use.

Good Luck, DoN.

"DoN. Nichols" wrote in message news:fa7d$43ebf5ff$48f4e632$ snipped-for-privacy@msgid.meganewsservers.com...

Thanks for all the details Don. I'll download it tomorrow and check it out over the weekend. I too was tempted to resort to devious methods with regard to the gears. Even worse, we spent all day today trying to get the brake shaft out of the miserable thing. A sixty year accumulation of rust is nearly as solid as a weld. We finally got it off this afternoon.. In my previous incarnations I would have just had one of my fitters cut it off and press the rusty remains of the shaft out to send to the scrap merchant. We spent all day and half a bottle of oxygen getting it off. We will have to replace the shaft anyway,but that's what they want, so that's what they get.

We still have a lot of work to do before we have to cut the gears but our target is to have it runable by the 12 Jun as there is a rally we would like to display it at.

regards

Tom Miller

Thanks Mike > All information helps.

By the way, is that BC as in British Columbia? I spent my first 28 years in Alberta and Saskatchewan,before I discovered that not everyone in the world had to shovel snow off their doorsteps every morning for six months of the year..

In thg original post it was stated that the wear was not discovered until the gear covers were removed. This makes me ask will the public even see the gears? If they are not visiable and its a static display why spend the time and money doing anything to them?

Actually it was discovered when the covers were removed. The machine will be a working machine and not a static display. I hope to drive it at a rally later this year.

Tom Miller

Both the English and the Swiss have standards for cycloidal gear profiles, mostly for their (now almost vanished) horological industries. However, if you examine them, you will find that they actually specify an approximation to the true cycloidal curve; this consists of an ogival arch (like a church window) constructed of segments of circular curves, with the radii specified in a table. And even at that, most users modified the form beyond recognition- at the Waltham Watch Co., for example, after experimenting with the cycloidal form for years, they finally threw it out and used a form of their own invention. The problems of constructing a cycloid curve in metal are not easy to solve, nor are they new. In addition, there is the problem of center distances Ed Huntress mentioned. What he didn't say is that the form of the gear tooth also depends on the tooth count of its mating pinion. So, to cut a true cycloidal form with form cutters, one needs a set of cutters, for each tooth size, which includes a range of forms for the range of gear tooth counts (just as for involute) but for each pinion count as well. By the end of the

19thC, Brown and Sharp had gotten it down to only 64 cutters for each pitch size, vs. 8 for involute (though they offered sets of 32 for non-critical work.) You can see why involute took over. Another reason might be that involute is better suited to gearing down (e.g. the speed of a gasoline engine to the speed of a tire, or an electric motor to a loom) while cycloidal is better suited to gearing up (e.g. the once/day rotation of a grandfather clock winding barrel to the once per minute of the escape wheel). So you need to ask yourself, not only, do you really need to reproduce cycloidal teeth, but also, are the old teeth really cycloidal? Most guys think they can tell at a glance, but they're not looking at the curve- they're looking at the whole tooth. Though clock wheel teeth are usually sharp-pointed, and slenderer than regular gear teeth, those features have nothing to do with the actual curve. In other words, cycloidal teeth can be stout and truncated, and involute teeth slender and pointed. These features have more to do with the overall wheel design (e.g., how deep the mating wheel is gashed, or the ratio of tooth width to intertooth space, or the required tooth strength) than with the mathematical curve the tooth profile follows. And of course, the tooth root being rounded or square is completely inconsequential to this. And don't forget, most horologists, let alone gearheads, have never seen a true cycloidal tooth, or even an image of one (as opposed to what I call an ogival tooth, i.e. what the English & Swiss standards define as "cycloidal".) On top of al this, I'd like to point out that the involute form occupies the middle ground, or boundary, between the epicycloidal and hypocycloidal curves. That's why any involute gear will mate with any other of the same tooth size and pressure angle, whereas an epicycloidal wheel will only mate with a hypocycloidal pinion. But if you're still hell-bent on cycloidal, the book you need is Willis's 19thC "Treatise on the Teeth of Wheels". It contains a handy cardboard calculator he calls the "Odontograph" for generating cycloidal teeth. There's really nothing practical on cycloidal tooth generation in any of the 20thC books on gears (at least that I've seen, and that's a lot), except that the textbook for the Swiss watchmakers training course (given free to selected candidates in the U.S., but whose German acronym I can't seem to extrude from my tired brain) has the Swiss "cycloidal" standard. Good luck!

"Tom Miller" wrote in message news:43ed0fe3$ snipped-for-privacy@quokka.wn.com.au...

If you run into a dead end the folks at this place may be able to supply some practical advice. I had a nice talk with the owner about what they go through in order to cut the large epicycloidal gears for tower clocks.

Ned Simmons

Thanks Ned I'm making a bit of progress I think. I'll email them if my ideas don't work out.

Tom Miller