Need help designing keyway in gear

[sanddrag]

Will the oversize (1/8") keyway in the picture below be okay to make? I want to go oversize because I'm worried that a 3/32 key is too small for where this gear is in our box. It needs to be able to handle some torque, maybe somewhere around 13 Newton-meters. But in the same way I'm worried that 3/32 keyway is too small, I'm worried that 1/8" is to large and does not give us enough "wall" between the top of the keyway and the dedendum. This gear is Small Parts number gss-2014 and is getting bored out to .375 Although we would like to, design contraints keep us from moving to a bigger gear.

So, the question is, 1/8" keyway and thin wall, or 3/32 keyway and thicker wall?

Or would cross drilling through the gear hub and shaft and pinning it be better? If so what size pin? Should we key AND pin?

EDIT: One more idea, maybe go with the 1/8" key but set it a little deeper in the shaft and don't cut the keyway so deep in the gear, but I believe it gets broached once without a shim and then again with the shim. Would anyone know the size of this shim?

EDIT2: I found this very good pdf document but I do not know how to do the calculations myself. Hopefully someone can give me a definite answer and reason of how to make this gear reliably spin with the shaft, at the torque around the number listed above.

Here's the pic:

[Max Lobovsky]

It'd be a pretty fair assumption to say that ANSI keyways (3/32 in this case) are designed to for nearly maxmimum torque transferring power (not maximum because they are designed so the key breaks first when everything is the same material). You can do some pretty simple shear strength calculations to figure out exactly how much torque you can transfer. I did half of it for you just because I was curious

13 Nm = 9.6 ft-lb
9.6 * 12 = 115 in-lb
115 / ((3/8)/2) = 614 lb of force shearing the key

[Matt Adams]

Originally Posted by sanddrag

Will the oversize (1/8") keyway in the picture below be okay to make? I want to go oversize because I'm worried that a 3/32 key is too small for where this gear is in our box. It needs to be able to handle some torque, maybe somewhere around 13 Newton-meters.

Originally Posted by Max Lobovsky

19 Nm = 9.6 ft-lb
9.6 * 12 = 115 in-lb
115 / ((3/8)/2) = 614 lb of force shearing the key

You need one more piece of information: Length of the gear and hub. This gear has a face width of .50 and hub of .46, for a total of .96 inches.

You will have an overall width of either .125 or .09375.

Steel has a tensile strength of 40,000 lbs / in², and a pure shear strength, (using Mohr's circle), of roughly half that. This is conservative, because most keys will be of a higher grade material with a higher strength.

You also have 614 lbs of shear, correctly calculated earlier.

Area = .96 * .09375 = .09 in²

614 lbs / .09 in² = 6,822 lbs / in²

Giving you a factor of safety around 2.93 with your 3/32" keyway.

If your feel pretty comfortable (or even within 100% of) the expected torque, I would say that your 3/32 key should be acceptable.

Matt

[sanddrag]

One thing Max, where did the 19 Nm number come from. Orginally I had said 13. Did I miss something or did you add a safety factor?

Thanks for the calculations Matt, this is good news. We will go ahead and give it a try, then I guess if it does't work we'll try 1/8, or pin.

Also, I forgot to mension that the key will be only under the face width of the gear, (.5) not under the hub. At motors stalled, the axis of this gear could have up to 20 Nm of torque on it, but I don't think our wheels or cicruit breakers will allow that. Are you still confident that it can take the load? What about when slamming between forward and reverse (which we'll try not to do but it would be good to know if we should worry or not). Thanks again.

[Max Lobovsky]

19 Nm was a typo, I used 13Nm (post edited). With the reduced length of key engagement, you are down to a safety factor of 1.53. That definitley sounds like its cutting it too close. A relatively simple solution that I know is pretty common is to use a double keyway, a key on opposite sides of the gear. It wouldn't be too hard to broach the gear like this (aligning the second cut may be difficult), but I don't think I have seen any stock shafts that come double keyed.

[sanddrag]

Originally Posted by Max Lobovsky

19 Nm was a typo, I used 13Nm (post edited). With the reduced length of key engagement, you are down to a safety factor of 1.53. That definitley sounds like its cutting it too close. A relatively simple solution that I know is pretty common is to use a double keyway, a key on opposite sides of the gear. It wouldn't be too hard to broach the gear like this (aligning the second cut may be difficult), but I don't think I have seen any stock shafts that come double keyed.

That is a really good idea. I know it is done in many real world commercial parts, but I hadn't even thought of that for this. However, like you said we may have trouble "timing" it to be exactly 180 degrees apart. In my current design, the hub of the gear is turned down so it has a very thin wall, and then 4 slits are cut in it at 90 degrees apart, and a split shaft collar is used to clamp the gear to the shaft to keep it from moving laterally because the gear is not "captured" on the shaft. There is a bearing a little ways away from the tooth side of the gear but there is only air on the hub side of the gear. The shaft ends right where the hub ends and the gear "hangs out." I may just redesign a couple things to leave the entire hub the way it comes, put a spacer between the tooth side of the gear and the bearing (making sure it is thin enough to ride only on the inner race), and then make the shaft longer and cut a groove to put an e-clip on the end so it doesn't fall off. I'm liking this idea better. One 3/32 keyway through the entire bore, and leave the hub the way it comes. Space it and clip it. Alittle less machining too.

Also, everyone you may consider the above picture a teaser. This gear appears in our HexaMax R2 system, and by the torques we discussed, you may be able to infer some other details.

[Max Lobovsky]

If you clamp the gear on like that (incidentally, www.sdp-si.com sells stock gears with that type of hub for clamping), that may very well transfer enough torque. It will definitley be pretty hard to analyze the amount of torque it should transfer (though I'm sure the clamp and they keyway would be plenty), so you may not feel comfortable with it, but its just another thing to consider.

[Matt Adams]

Originally Posted by sanddrag

Are you still confident that it can take the load? What about when slamming between forward and reverse (which we'll try not to do but it would be good to know if we should worry or not). Thanks again.

Use part number 98535A125 from McMaster. ($1.50)

It's a high carbon steel with a tensile strength of 95,000 PSI. That'll boost your factor of safety above 3.

No worries.

Matt

[sanddrag]

One last quesiton. It seems as if my brain has malfunctioned once again and I didn't order those gears when I thought I had. Anyway, I can't fully understand all the calculations but since you guys do, can I just ask will a .688 length through bore be okay, or do we really need the .940? Because all I can get locally is .688 through bore. Also, for that key stock you posted, is that a safety factor of 3 on the .5 length through or .94? Thanks so much. You have no idea how much I appreciate this.

[Max Lobovsky]

Originally Posted by Matt Adams

Use part number 98535A125 from McMaster. ($1.50)

It's a high carbon steel with a tensile strength of 95,000 PSI. That'll boost your factor of safety above 3.

No worries.

Matt

I was sure steels were commonly over 70k psi, I should have pointed that out.


Sanddrag, with a good steel key like Matt pointed out, you will have a safety factor of over 4 or 5. You should be fine.

[Travis Covington]

Hex broach? Haha, people these days will broach a hex in anything! No key to lose!

[sanddrag]

Originally Posted by Travis Covington

Hex broach? Haha, people these days will broach a hex in anything! No key to lose!

I think I know what that wink is for. And likewise. Let's just say we both get influenced by the same guy whose intials are AB, if you know what I mean. (And travis, I hope you don't stick another stage on the end of last year's tranny for six speeds, that would be absolutely insane.)

Anyhow, a hex broach would be great, but the shaft this gear is on is .500" diameter up the line a bit, so machining down to a 3/8" hex on this round shaft wouldn't be too fun. Nice idea though.

[Travis Covington]

another stage would make 8 speeds! haha!

Either way, it looks like the 3/32 will work fine for you. What function is this gear train for?

Goodluck, can't wait to see what it is!

[Andy Brockway]

Originally Posted by sanddrag

EDIT: One more idea, maybe go with the 1/8" key but set it a little deeper in the shaft and don't cut the keyway so deep in the gear, but I believe it gets broached once without a shim and then again with the shim. Would anyone know the size of this shim?

Dave,

I do this on the 5/16" shaft with 12 tooth gears in my gearboxes. The shim is 1/32". Cut the keyway 1/32" deeper in the shaft and single pass broach.

Btw, I do not use setscrews with this method. I now use green loctite with the key with excellent results. I also used the loctite without a key on the dual motor DeWalt I built this past fall with no problems.

If you are planning on this gear sliding along the shaft, then obviously this information will not help

[Paul Copioli]

Let me tell you about that "green Loctite" Andy talked about. It is Loctite 680 and is meant to carry significant loads. FANUC Robotics uses it on some of our paint robot's gear to shaft interfaces. If applied properly, it is meant to be permanent and is awesome. You can, however, remove the pieces by applying heat to the components (a heat gun works best). We have used Loctite 680 on our robot in the past where we just couldn't get a key or some other locking mechanism.

-Paul