How to Measure and Model a Curved Gear Rack

[Lotaxi]

Hello,
I have a gear rack to reproduce that has a few broken teeth. I have next to no information on the original, so it's going to take a fair bit of measurement and figuring out in order to get it right. It's a little unusual, however, in that it's curved like a section of a ring gear. I've looked around for guidelines on trying to measure profiles on ring gears, but I haven't found much information. Does anybody have any tips on how to do this correctly?

Photos of the pieces in question

 

[Milland]

That's just molded plastic, so likely not super-precise. I'd be tempted to scan it, then import the scan and fiddle with it in CAD to mock up the profile.

What's the part do? What sort of loads and speeds?

 

[Lotaxi]

That's just molded plastic, so likely not super-precise. I'd be tempted to scan it, then import the scan and fiddle with it in CAD to mock up the profile.
What's the part do? What sort of loads and speeds?

It's a gear that meshes with a potentiometer for the purpose of measuring how far a foot pedal is depressed in an race arcade cabinet. People stomp on the gas, and the 30 year old plastic gets brittle and fails. Nothing that's super critical to get exactly right, but I'd rather have a clean profile that I can reproduce at will, and it's probably best to keep wear away because of excess slop in the system, especially if I use something less compliant than injection molded plastic. I'm planning on making it out of something a little harder and with a longer life so that it won't need replacing again. OEM doesn't make these anymore, and the primary source seems to be cannibalizing other cabinets in worse repair. I'll likely have to replace the gear on the potentiometer at the same time if I go with steel like my brain wants me to, but I can reproduce the outside gear profile just fine.

I'll likely start out by scanning it in with our laser table, but I typically confirm measurements where I can to use as a guide for my models.

 

[Milland]

I'd not use steel, how about Delrin? Steel may just mangle the mating part, and the extra weight may throw off the "pedal action".

 

[Nmbmxer]

Measure the spur gear, then it should be fairy simple to infer the ring profile.

I’ve been having stuff printed out of nylon for some prototypes and it is nearly as good mechanically as the same part machined from 6/6 and takes no work to make. Im positive that gear would last nearly forever if 3D printed in nylon. With proper stops on the pedal I don’t know why the gear would see much load.

 

[Lotaxi]

I'd not use steel, how about Delrin? Steel may just mangle the mating part, and the extra weight may throw off the "pedal action".

Only issue with steel I can think of is that it'll wear on the mating part as it exists, yeah. That's why I mentioned that if we go with something that hard we'll need to also replace the mating gear at the same time. Client stated that he prefers metal, but I told him I'd look into what the right material might be. Looking at the teeth that have been torn off the existing part, I think there's a good chance they're made of delrin or nylon as is. There's a fair bit of plastic deformation surrounding the shear area. As for weight, I don't think that parts this small are likely to mess with inertia that much. If it's a problem and he's dead set on metal pieces, there's always the jump to titanium...

Measure the spur gear, then it should be fairy simple to infer the ring profile.

Already called him to see if he could pull it off the machine. We'll see what he can do.

Im positive that gear would last nearly forever if 3D printed in nylon. With proper stops on the pedal I don’t know why the gear would see much load.

I mentioned 3d prints and he kinda balked, so I'll do my due diligence and see what he thinks. First thing I'm giving him is the CAD models I make out of the existing parts, so I'll suggest it again when those are done as a stopgap measure to just put the things back in service and see how the design functions. Otherwise I guess I get to play with the EDM and charge him a fair bit of money.

 

[SIP6A]

Hello,
I have a gear rack to reproduce that has a few broken teeth. I have next to no information on the original, so it's going to take a fair bit of measurement and figuring out in order to get it right. It's a little unusual, however, in that it's curved like a section of a ring gear. I've looked around for guidelines on trying to measure profiles on ring gears, but I haven't found much information. Does anybody have any tips on how to do this correctly?

Photos of the pieces in question

Well can you measure the center to center distance of the Two gears? from that and the pitch of the gear you should be able to calculate everything else you need to make the parts.

 

[Nmbmxer]

I’m a huge proponent for 3D printing in the right application. Reverse engineering is one of those applications, I’d print one whether you were going to make one from steel or not just to confirm it fits correctly before you spend a ton of time on the edm. The accuracy of newer printers is quite good and the engineered plastics are getting really good.

 

[APD]

Hello,
I have a gear rack to reproduce that has a few broken teeth. I have next to no information on the original, so it's going to take a fair bit of measurement and figuring out in order to get it right. It's a little unusual, however, in that it's curved like a section of a ring gear. I've looked around for guidelines on trying to measure profiles on ring gears, but I haven't found much information. Does anybody have any tips on how to do this correctly?

Photos of the pieces in question

Take a hi-res photo of it with a measuring scale in the photo. Bring it into your cad at full resolution and trace the perimeters. Then scale the drawing until dimensions match the actual part and the scale in the photo.
Then 3D print. This is a perfect application for 3D printing

 

[goldenfab]

Take a hi-res photo of it with a measuring scale in the photo. Bring it into your cad at full resolution and trace the perimeters. Then scale the drawing until dimensions match the actual part and the scale in the photo.
Then 3D print. This is a perfect application for 3D printing

That's what I would do as well. I'd scan it in a flatbed office paper scanner along with a good ruler to confirm the correct scaling.

 

[EmGo]

If it is curved, it's not a rack.

 

[michiganbuck]

I used to grind a part like that in hard steel on a one-up basis for a matching set of two with the teeth set on a radius arc. They were straight across teeth in three different angles 3, 5 and 7*, yours would be easier because the teeth are straight across, and yours seem to have a radius on the opposite, mine were square on the opposite side.
I dressed the wheel to the V angle and then with a magnifying glass aimed each V straight up and down-ground each V individually to a witness of the existing V notch on the teeth that still had a little V left, those that did not have enough to witness I used a depth gauge the I had made. It was a slow job that likely took an hour to do each one because each tooth V had to be re set and clamped there, and they came in matching upper and lower halves. likely could only hold about .001 but being difficult to inspect the inspector could not check them much closer than .oo1ave made a gauge at the radius arc and used feeling a wire like from a 3 wite thread checking set to get closer than .001, and because when the job came to me it was boiling hot with a machining line down untill I finished the set.
Agree plastic can be a bugger with loading a suface grinder wheel. Milling the part might need a cutter with the included V notch angle.

60 Degree Angle Cutter - Search

Find high-quality images, photos, and animated GIFS with Bing Images

www.bing.com

 

[atex57]

EG is right. That is simply a section of internal gear.

 

[Dan from Oakland]

or a section of a regular old spur gear- not sure the OP ever stated what side of the curve the teeth were on.........

 

[EPAIII]

That was precisely my first thought. They are segments of internal gears. Knowing the correct terminology will make the process easier.

Those more experienced with making gears can correct me, but if you are going to cut them, then you can not use standard gear cutters. The wire EDM with a good 3D model may be the way to go.

But when you create a 3D model, you may be able to turn the 3D CAD tooth curve inside out if the software allows that. Or do any 3D CAD programs allow internal gears to be modeled? Again, others may have more on this.

If it is curved, it's not a rack.

 

[implmex]

Hi All:
As some have pointed out, this is not a rack, and the photo in post #1 clearly shows, it is indeed a segment of an internal gear.
It's also pretty small from the looks of it.



If you wanted to make some of these, first you'd need to determine exactly what the specs of the gear are.
If you want to print it or wire it you need a CAD model first.
Measuring it accurately may be your biggest problem.

If you know the center to center distance of the pinion to ring gear, that will be helpful.
Calculating how many teeth in a complete internal gear and knowing how many teeth on the pinion is also helpful.
Determining the pressure angle is good too, and you may have a better chance picking it off the pinion.
Calculating the DP (or module) from the pinion is necessary too.

If you want to make it in metal, and can make the CAD model accurately, wire EDM is the obvious choice.
Alternatively you could shape a bunch of blanks with a Fellows gear shaper and then cut out your segments if you plan to make lots of these.
Doing it that way needs a proper Fellows gear cutter, so if it's not a standard, you have a problem you don't have if you wire it.

If you plan to make them in plastic and you cannot get the resolution or the strength you need from a 3D printed part, ideally, you'd shape them on a Fellows, and as before, you need a cutter, and you need to know exactly what to make.
For a little job like this you could wire a cutter.

If you don't have access to a Fellows, you can fake it on the Bridgeport with a rotary table and a single point cutter, either wired or ground...even freehand ground if you're good with your hands and don't need perfect.
If you Bridgeport it, you may get away without all the gear calculating first...you can just nibble out an approximate copy of what you already have in your hands.
Bolt the plastic bit to your brass (or whatever ) blank and go to town with a good magnifier and a bright coloured Sharpie so you can tell when your cutter touches the plastic sample.
Get your tooth spacing with the Rotary Table.

Don't plan to make many this way...it'll make you old before your time.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[michiganbuck]

A hack way of making it might be to carefully clean up any burrs and bugs from the small part that looks pretty good, put on some release and make a cast of it to be used as a gauge to literally bench by hand make both parts.

With a straight edge C clamped to the side and looking over joBlocks or a caliper one could get the arc radius, and apply that to the teeth height.
and find center-line between the bolt hole to make a 90* line to find the center tooth location.

 

[Lotaxi]

As some have pointed out, this is not a rack, and the photo in post #1 clearly shows, it is indeed a segment of an internal gear.

Mentioned that it looked like a section of an internal ring gear in the OP, but I didn't have a word for a small section of one. Guess it's just a small section of an internal ring gear.

It's also pretty small from the looks of it.

The larger of the two is about an inch across.

If you wanted to make some of these, first you'd need to determine exactly what the specs of the gear are.
If you want to print it or wire it you need a CAD model first.
Measuring it accurately may be your biggest problem.

Current plan I've been working with has been using an image I pulled from a flatbed scanner. Parallax is annoying, so we'll see how close I get with that. Other option is converting from our 3D scanner.

If you know the center to center distance of the pinion to ring gear, that will be helpful.
Calculating how many teeth in a complete internal gear and knowing how many teeth on the pinion is also helpful.
Determining the pressure angle is good too, and you may have a better chance picking it off the pinion.
Calculating the DP (or module) from the pinion is necessary too.

Don't think I'm going to get a good read on center to center, but I've asked for access to the mating pinion already. Will be considerably easier to model the pinion, and I think I should do that anyway. If it's plastic and they ignore my warning, they're going to come back to me when someone stomps the pedal and strips the teeth of that. It'll be best if I model the thing while it's clean. We'll see what I can get from the client. Hopefully I'll be able to just take it from him like the other pieces.

If you want to make it in metal, and can make the CAD model accurately, wire EDM is the obvious choice.

That's the plan. Prep a blank in the 3 axis, then into our W-EDM. I'm going to print a few of them in plastic to test fitup and function before I bother with fixturing, though. Also gets the guy back up and running until I can make the conversion kit all the way through.

Don't plan to make many this way...it'll make you old before your time.

W-EDM would be the best way I got for these, so that's what I'll do for the final product. Theoretically there's a niche market for these, so the work to make the replacement kits should be worth it.

 

[trevj]

It's a gear that meshes with a potentiometer for the purpose of measuring how far a foot pedal is depressed in an race arcade cabinet. People stomp on the gas, and the 30 year old plastic gets brittle and fails. Nothing that's super critical to get exactly right, but I'd rather have a clean profile that I can reproduce at will, and it's probably best to keep wear away because of excess slop in the system, especially if I use something less compliant than injection molded plastic. I'm planning on making it out of something a little harder and with a longer life so that it won't need replacing again. OEM doesn't make these anymore, and the primary source seems to be cannibalizing other cabinets in worse repair. I'll likely have to replace the gear on the potentiometer at the same time if I go with steel like my brain wants me to, but I can reproduce the outside gear profile just fine.

I'll likely start out by scanning it in with our laser table, but I typically confirm measurements where I can to use as a guide for my models.

Can you replace it with, for instance, a cylinder, with a wire cable wrapped around it, to actuate the pot? Would be pretty much stomp proof!

IIRC, there are a couple freeware gear profile programs that you may be able to use to calculate the pitch diameter with. What you are dealing with IS a bent rack, but is otherwise also known, as a sector of an Inside Tooth Gear. With some careful scanning and a graphics program, you can likely build up a full circle of overlapping parts and use the dimensions to calculate the DP or module.

Another option, if you are in or near a place that has an RC car related hobby shop, is to take your parts there, and see if they mesh up with any of the standard DP gears they stock. With that in hand, you may be able to source sectors or an inside gear to cut them off of.

As a last resort, I would look at making a single tooth cutter, some brass sector blanks, and use a lathe, with an indexer attachment or dividing head to cut the new teeth by hand. See any number of videos online about shaping keyways on the lathe, as well as gear teeth. Slow, but quiet work...

 

[implmex]

Hi again Lotaxi:
I was unaware you have access to proper resources.
My apologies for invoking the Bridgeport kludge...I foolishly got the notion you were a hobby guy playing in his basement.
Mea Culpa.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com