Gear Making Information

[Alan Ing]

Hi all! I guess its time to brush off my old engineering books and start thinking about the 2002 competition.

Last year I had a chance to go to the Kickoff and sat in on a presentation by Mr. Joe Johnson and Mr. Michael Civaglia (Sorry if I didn't spell that right, thanks for having that second class after the Kickoff in the hotel lobby). Anyway I found their presentation the most useful with a lot of practical information about motor sizing, gear ratios and suggestions on how to use them. After the presentation, I was all set to find sources of premanufactured gears when I discovered that the 2001 rules limited gears to what you could purchase through SPI or what you could make with the approved materials list. Does anyone know if the same rules apply?

Fortunately, this gear limitation didn't hinder us too much, as we were still able to construct a fairly competitive robot, but I know that we would have had more options if we had the capability to cut our own gears.

With that in mind, I was wondering if the teams that cut their own gears could share with us how they did it and with what approved materials?

I'm assuming that it can be done with a standard Milling Machine a gear cutter, and a quality divider head. Any thoughts about that?

How about the Diametrical Pitch and Pressure Angle of the Drill Motor and Fisher Price motor pinion gear? I also understand that the gears may be metric so the Module system may be used (Is the Fisher Price motor Module .8 and the Drill Motor .7 ?)

What materials would you suggest, 1/4" plate steel and round shafts?

Do you have to Heat Treat the gears once you cut them? Can you do that with a torch?

Sorry, these questions may be elementary for some, but I am an E.E. who acts as the lead design engineer for our team.

Oh, one more question directed to the Technokats, in your dual motor assembly, how does one go about lightly welding that hex coupler to the back of the Drill Motor without destroying it (TIG, MIG?), any precautions?

Well, I guess I asked enough for my first post. Any information would be greatly appreciated.

Alan,

Since nobody else has stepped up yet I'll try to answer your questions. The change on gears last year really threw us for a loop. I had a drive system all designed but it was not feasible to construct with the gears available. We considered making our own but decided that we didn't have the resources to figure out how in the time available. Consequently I've been developing sources for "cut your own" gears just in case.

Our machine shop frankly refuses to even attempt cutting any gear teeth. (Sprockets are a different story) They just aren't equiped for it. BUT they do have the gear design module for MasterCam. I have checked with several EDM (Electro-Discharge Machining) houses in our area and they have indicated that they can burn gears out of any common material directly from the MasterCam file. Many have similar software themselves and can create their own pattern if you give them the proper parameters (pitch, face angle, bottom radius, etc).

It ain't cheap, but it's not terribly expensive either, especially for 2-4 gears. Expect to pay a premium of 10-40% over a catalog gear. But less than a custom gear made conventionally as most of the cost for just a few custom gears comes from the setup.

How about the Diametrical Pitch and Pressure Angle of the Drill Motor and Fisher Price motor pinion gear? I also understand that the gears may be metric so the Module system may be used (Is the Fisher Price motor Module .8 and the Drill Motor .7 ?)

Dr Joe has posted this information a couple of times. They might be in the old forum. I don't remember the numbers off hand.

What materials would you suggest, 1/4" plate steel and round shafts?

The shaft depends on how you plan to connect to the rest of the system. I like D shafts myself. The material thickness depends on the face width you require for your load.

Do you have to Heat Treat the gears once you cut them? Can you do that with a torch?

I wasn't planning on heat treating, I try to design stuff like that out. While a torch might work in the hands of somebody who knew what they were doing, I'd much rather use a heat treat oven. Since gears would be such a small load you might get a local heat treater to run them for free in the same load with a paying customer.

One note about published gear design guides. They add factor to factor to factor so your gear teeth wind up much stronger than they need to be if you add everything on. Remember that those design guides are intended to be used for things that are going to operate for millions of cycles and both the designer and operator need to worry about getting sued if something breaks. So they tend to give answers that are over designed for our application

[Joe Johnson]

Okay, here we go...

As to the Drill -- it is a standard full depth 20 degree pressure angle, 12 tooth, 0.7 module gear. It mates wonderfully with standard gear, but I put a few thousandths of an inch extra in the the center distance because typically the gear make (and mount to a shaft) is not often as concentric and true running as I would like -- add even MORE if you are going to heat treat the gears or weld them to the shafts.

As to the Fisher Price -- it is essentially a standard full depth, 20 degree pressure angle 16 tooth 32 Diametral Pitch gear. The tooth from is not quite standard because it is made to the plastic tooth profile standard, but as a practical matter it mates will with standard gears.

As to heat treat, I have given up the practice if there is any way to help it. I also have tried to limit welding the shafts to the gears unless I can get a reasonable shaft diameter. It is a bother that can be avoided with some fore thought.

As to materials -- almost all fall under the 1/4 inch thick plate rule. Essentially if it is made from 1/4 inch plate you need not use SPI stock, nor do you have to bother with getting SPI stock to the supplier so that they can begin cutting.

You need to find a Wire EDM source. In my opinion, the are the only way to go. We have some sources here in Detroit that will make a gear for about 150% of what SPI costs -- add that included getting them in about 2-3 days! Find local source if you can. If you cannot, contact me & I will try to give you sources that you can more or less send an e-mail to with the gear data and your Master Card Number and they will send you a gear in a few days. If you have CAD files (typically IGES files) with custom gear profiles, you can get those burned in to steel or aluminum as well.

As to gear tooth strength, I use a VERY simple approximation. I calculate the max stress by modeling the tooth as cantilevered beam with a load at the end. The load is the tooth loading. The height of the beam is the height of the tooth. The cross-section of the beam is the thinnest section of the tooth. There are pretty simple formulas for max stress for this loading case.

For steel, you want to stay below 60-70,000 psi (assuming you and get a fairly good steel stock that has about 90,000 psi in its unheat-treated state).

The Lewis formulas are VERY conservative for our typical applications.

I have forgotten the rest of your questions...

All for now.

Joe J.

[Andy Baker]

Originally posted by Alan Ing
Oh, one more question directed to the Technokats, in your dual motor assembly, how does one go about lightly welding that hex coupler to the back of the Drill Motor without destroying it (TIG, MIG?), any precautions?

Alan,

I believe that we used a TIG weld to do this. We first started with a partial weld around 1/2 of the shaft... and that didn't hold too well, so we put on a more complete weld to the shaft.

Also, before we welded the shaft, we tried our best to dry out the oilite bearing on the back end of the motor. After the weld was done and the shaft cooled down, we re-oiled the bearing. We didn't want the oil in the bearing to bubble up and deform the bearing.

There are other ways to skin this cat. There are other types of welds I've seen that may work, such as "press welds", that use high pressure between the materials while heating the metal.

I'm no welding expert. I suggest you do what I did... take it to someone who is very good at doing this precise, but quick job.

I suggest that you try some things out and have plenty of spare motors to use. We always had two spare drill motors ready to go with hex shafts welded in case of a broken weld on this coupling.

Good luck,
Andy B.

[Alan Ing]

Chris, Joe, & Andy,

Thank you very much for your responses. Our team is from Hawaii and while that may be great when I need to find a Coconut, Pineapple, a Mango, or a Japanese Tourist, it isn't so great when I need to get parts made up for the robot. We only have a handful of Machine Shops and most just repair industrial construction equipment. We really don't have much of a manufacturing industry here.

This year, my good friend the Professor said he was planning to make some gears for us out of two Coconuts and assorted Bamboo stalks. This would have been great, but he later found out that Gilligan had already given the Coconuts to Mary Ann who made them into a Coconut cream pie. He tried making the gears from the pie, but found that the finished gear did not have the tensile strength we required. Well, so much for making gears out of commonly available products in Hawaii. I guess it wouldn't have been legal anyway as I don't recall Coconuts being on the approved materials list.

Kidding aside, unfortunately, we don't have an EDM source. Thats why I was thinking about cutting our gears on a Mill with a good divider head. Is this a bad idea? Are we asking for a lot of problems? Our machinist has a standard Brigdeport type Mill.

Andy, thanks for the welding advice, I'd like to play with that motor coupling design as it is a truly unique solution.

Joe, are you going to have that motor class again at the Kickoff? I personally thought it was the most useful of all seminars that was offered (They should give you guys a real room!)

[Joe Johnson]

You can (and we have) make sprockets pretty easily on a mill.

Gears are another thing entirely. Unless your gear size is huge, you really can't get a mill cutter to make the proper involute shape that a gear requires (at least standard gear us involute shapes -- there are other profiles but, for a number of very good reasons, most popular is the involute).

I recommend that you find a wire edm source from the Mainland. Why not?

Joe J.

P.S. As to a repeat of my teammate, Mike Ciavaglia's motor presentation in Manchester in 2002, this seems unlikely at the present for a number of reasons: We are probably going to our local kickoff here in MI in order to save money, Mike C. is on a job that is about as hot as jobs get around here, and finally, Bob Hammond gave me too much grief about extorting extra invitations to the kickoff in trade for running the various Chief Delphi Seminars. All in all, I doubt it will happen. JJ

[Alan Ing]

Joe,

Sorry to hear that the Seminars won't be happening again. Its too bad as I know a lot of teams could benefit from them.

Let me see if I get this. Even though I can purchase a 20 degree involute gear cutter in 32 DP from MSC, using one to make gears on a mill will be very difficult or unsatisfactory. This is dissapointing as I was hoping to save time.

EDM sounds great, and I'll try to find a source, but what I'm worried about, is that if it takes a EDM source 3 days to ship the gear out, it could take up to 4 more days to receive them in Hawaii. UPS 2nd day air is more like 3-4 day service to Hawaii. If I ordered a gear on Monday, and it is ready by Wednesday and shipped UPS 2nd day air to Hawaii, it will most likely arrive on the following Tuesday. 8 days during the competition build time is a long time to wait. Can't afford to make mistakes! Just one of those sacrifices you have to make to live in paradise.

Alan

[Andy Baker]

Alan,

Although Joe is correct by saying that it is difficult to cut an involute on a mill, I suggest you try anyway... hey, it's worth a shot.

From my experience, steel gears can still get a decent mesh, even with inprecise involutes or while being slightly stripped. We've cut gears with not-so-good meshes before, and just wore them into meshing better. A few thousand RPM and some grease... voila... an OK gear mesh. We did this in '99, and it seemed to work OK. These semi-stripped OK meshing gears are still on our TKO robot.

Last year, we cut a pair of delrin gears (32 pitch) on a mill using a right angle indexing head. Although these were not steel, they were made fairly easily. Of course, steel gears would be harder to make, but I'm sure that the Professor could figure it out.

Also... since you're stuck on that island, you've got plenty of time on your hands for tinkering with steel gears.

One thing I'm wondering... why don't you just use some of Mr. Howell's money and buy yourself an EDM?

Good luck,
Andy B.

[Matt Reiland]

Maybe one of the teams out there that has used a reliable source could ask them if they are looking for more small jobs. Similar to Chief Delphi we have the luxury in MI of having quite a bit of machining here. However with the recent changes to the economy some of the smaller shops may be able to take on small quantity jobs with a rapid turnaround with credit card orders shipping UPS or FedEX. Just a thought.

Anyone who has a good CNC milling shop or Wire EDM source should ask them if they are interested in becoming a supplier for this type of competition. (This also leads to work in the companies that are supporting the teams, we use shops for projects in GM after seeing the work/speed they provide on US First)

[dlavery]

Another option is you could use all this time during the fall to make your own Wire EDM machine! Then you can cut all the gears you want in January-February! No, it is not as far out as it sounds. I think either "Home Shop Machinist" or "Machinists Workshop" published a set of articles a few years ago on how to do it. You should be able to get back issues from them (see http://www.homeshopmachinist.net/ for details). As an aside, I have found both of these publications to be very useful - as a non-professional machinist, they have been great to have around for reference during robot-building season! (I have no affiliation with either publication, but have become an avid subscriber!).

Anyway, if I remember correctly, the Wire EDM plans didn't even call for any coconut husks!

-dave

[dlavery]

Here is a follow-up: the Wire EDM construction articles were published in a series from May 1995-March 1996 in "Home Shop Machinist". The articles were written by Robert Langlois.

Build yer own, plug it in, and set the sparks a-flyin'!

-dave

[Jim Meyer]

The HOT team has successfully cut steel gears on a mill. It was the year before I joined the team so I don't know all of the details.

In '99 they didn't have the SPI budget to buy the gears for their custom drive train so they made them. They made 20 pitch, 20 degree pressure angle gears in a variety of diameters. They were made on a CNC bridgeport with an index head. I'm not sure whether an index head is the same as a divider head.

They started with the proper diameter cold rolled steel round stock from SPI and a full set of gear cutters (you need different cutters for different sized gears). The gear cutters they used are like fly cutters except with involute profile shaped teeth. They cut several gearwiths at a time and then parted the desired facewidths off in a lathe. The cut gears were nearly indistinguishible from purchased ones. Even though the gears were not heat treated, there were no problems with durability. (They appeared to be oversized for the task though.)

If you have more questioins, I can get better information from the machinist who made them. Just send me an e-mail at James.L.Meyer@gm.com or post here.

It is my impression that with today's small parts budget you are better off buying SPI gears than trying to machine them this way.

Another solution to the Bosh motor gear problem is to press your own gear on the motor shaft. (Something the HOT team also did in '99)

Hope this helps,

Jim Meyer

[Ken Leung]

Well, now that someone mentioned Heat-treating metals... I am wondering what process do teams use when they are building their robot beside using power tools and welding?

I am totally unfamiliar with any other methods besides using tools, welding, and heat-treating, so I can't really list any examples...

Are there a lot of unconventional ways of machining or processing metals these days?

A lot of teams with limited fundings won't even try to imagine beyond using mills and lathes, but it's still nice to know what's going on out there...

[jasoni]

If you are going to mill your own gears, a nice way to break them in is to run them with a lapping compound for a period of time. Lapping compound is really just a grittier polish, but it works wonders in helping gears mesh a little better, and you get the added bonus of a quieter drive train.

We at 550 are in a Technical High School, so we have our own machine and weld shops to go to whenever we please. Now, in our machine shop, we have all the standard lathes, presses, mechanical scribers, etc., all with CNC retrofits or new stuff. So, we're pretty equipped for the big stuff. But little stuff, like gears, as somebody said, many machine shops balk at. So, we pleaded and pleaded, and we eventually got them to make gears on a CNC Micro-Lathe, circa 1982. These units are great because they literally do all of the work themselves, including cutting gear teeth. Our robo-machinist, Joe, did all the programming for the CNC, and the gears came out great. We made them out of low-carbon steel and cut them with carbide chisels. The machine took it from round stock to gear within ten minutes. Nice job, too. These machines come on the cheap too. About $200 each, if you can find them. Then, of course, there's the DOS computer with the microcassette drive...