Bevel Gear Machining.

This is a bit of a random query. For an off-season project our team is needing to produce some rather large bevel gears on our 3-axis Tormach CNC machine. The specs of these gears would be 16 DP and 45 teeth. Anyone have any insight into the process of bevel gear machining? Any tips or advice would be appreciated.

1 Like

For gear making you are going to need at least an lathe and 4th axis for the tormach. We have cut out regular gears with a 4axis on a CNC mill before but with bevel gears I assume it is a more complex operation?

Here’s a cool vid of bevel gears being made on manual machine:

The hardest part of machining a bevel gear - or any gear - is the spacing between teeth. You can make the bevel portion easily with a lathe (watch the video) - but you need a 4-axis mill to make the spacing, or at least to do it “the right way.”

Of course, it can still be manufactured with a 3-axis mill if the vise can be rotated precisely: put the gear on a shaft in the vise, cut a slot, rotate it the correct number of degrees, cut the next slot, and so on. The tricky part is when you run out of vise freedom - you have to align the gear again to continue cutting.

You can make it even better if your mill’s head can be rotated - this will allow a consistent depth with the bevel gear grooves, rather than being far deeper at the wider portion (which would reduce strength).

Or, if your vise doesn’t turn, there’s always trig - with a CNC machine, shouldn’t be too hard. However, a rotating head won’t help, so you will have a gear with deeper grooves at the bottom than the top.

Another method would be to simply lathe a blank and send it out to a gear shape/hobbing shop.

Its a whole lot less work, but CNC milling your own is pretty fun!


The issue with bevel gears is that the teeth (and spaces between them) are not constant - they taper. Some careful clamping and indexing can help get the blank aligned with the cutting tool properly, but you need to use a tool that is narrower than the narrowest space and let it run its path. A specially-shaped cutter is best, tapered so it cuts the flanks of the teeth properly.

Get a bevel gear - any - and study it to plan out your buts. Make a few from wax or aluminum to get the feel and plan out the holding and indexing.

That is a really odd size of gear and sounds like WAY too much work to make it. May I ask what exactly you’re doing and how you determined you need this size of gear?

Do you actually want a Miter Gear? 1:1 Ratio?

If yes, ignore rest of post.

Martin Sprocket Catalog shows a Bevel Gear with 16P 48 Teeth. But bevel gears work in sets. This particular one is a 3:1 ratio.

Part # BS1648-3

Thanks for all the responses guys. In particular the last one. As it turns out this bevel gear does need to have a 3:1 ratio. The small pinion bevel gear (16 tooth) will most likely be a purchased item, However for a variety of reason it would be preferable to machine the large bevel gear ourselves. In particular I was looking into using a three axis process like this which is often used to make large bevel gears. Is this process a feasible one ?

In 2011 we used the bevel gears from some power tools from Harbor Freight. We needed some angle gear boxes at the last minute. They came with bearings and shafts that we could mount things too and are still going strong.

Which particular tool?

We used a 18 volt angle drill. One of the kop motors was a perfect match and we were able to use the reduction gears. You can look on line at the parts lists and diagram for most tools. If you search you can many different size of gear sets. Also winch and air tools are good sources. In 6 the hours the team had bought and made 2 angle gear boxes for 25$ each.:slight_smile:

There are several different tooth forms for bevel gears; if you’re planning to build something using a stock gear and a custom one, make sure you’ve matched the tooth forms carefully.

I’d say it’s usually a lot simpler just to buy an off-the-shelf set of gears.

Not as shown, because the bevel gear you are proposing has an angled face. In the video the tool is perpendicular to the teeth. To use a process like this, you’d need to elevate the workpiece at a precise angle, then turn it a precise arc for each tooth - i.e. have the tool perpendicular to the tooth root.

A spiral bevel gear would work OK (duh - they did it!), maybe better, with just a small extra effort in design. But you’d likely have to cut both gears yourself.

Alright then, That makes a bit more sense seeing as there’s a good amount on the internet related to 3 axis CNC work on Spiral bevel gears.,but not on straight tooth bevel gears. Having to manufacture both gears would be a trade-off we could make. (We have a 4th axis attachment for our Tormach we could use for the pinion.) Does anyone have any knowledge related to the advantages and disadvantages of using spiral bevel gears?

I’ve debated doing this myself a few times, but never really had to pull the trigger. I did decide i’d probably take the CNC and a really long runtime route, as you definitely couldn’t buy what I was after.

In the car world, the big advantage of spiral bevel gears are their much quieter operation and their somewhat longer life. The disadvantage is their higher production cost.

In your circumstance, I think ease of manufacturing would trump all that.

It is possible to make a flat straight-tooth bevel gear, but strength, durability and efficiency suffer. Really, the big gear is flat, and the pinion looks like a spur gear but with tapering teeth - the tooth profile is not constant, but gets thinner towards the far end. Generally the pinion is not very wide, maybe 1/2" - 3/4" for the kind of sizes I think you’re working with.

Maybe try it with some machining wax, machining foam, or maybe soft brass?

Hi FRC teams! As a team with little experience, we are trying to move away from imprecise tools. What would be a good and cheap 3/4 axis mill (~$2000)? Are used ones ok? Are complex parts manufacturable with better manual tools (ie manual lathe/mill)?

Many Thanks

There are good inexpensive mills that can do the job just fine for most of the work an FRC robot needs.

We have access to a HAAS four axis once in a blue moon, but most milling last year was done using a Grizzly 1007 manual mill with an autofeed table. It goes for ~$1700. Harbor Freight has a slightly lower model as well (different paint and without autofeed) for $1100 (#33686).

Grizzly has some inexpensive endmills (~$90 for a set in R8 collet size). Spend the money on a precision machining vise. Grizzly sells some as well.

I would budget about $2500-3000 minimum for a good size setup considering the clamps, collets, endmills, cutters.

You can also retrofit one of the Harbor Freight mini mills (#44991) with a X2 CNC kit for <$2000. They sell the mini-mill for $600 and they may accept the 20% off coupons that are always floating around.

I’m sure there are plenty of other options out there as well.

I would post links, but I think that sends my posts to pergatory for now while I get my post count up.

Used tools are OK, and can often get you much more machine for your money.
Search ChiefDelphi for many, many mill recommendations. I remember Dave Lavery offering a particularly good choice.

Far more important than the quality of the machine is the quality of the operator. An excellent operator can turn out world-class items from a poor machine, and a poor operator* can fail to turn out usable items from the finest machine on the planet.

Many schools / teams decide that CNC is the way to go, as a way of compensating for their inability to actually operate a milling machine. They let the computer make the decisions, and the result is just fine. I am of the opinion that having one or two kids on the team become excellent operators is a far better method.

So, find a local machinist or three, and see if you can get them to teach someone - ideally a Sophomore (who will be with the team a while) how to use a manual mill. And/or maybe introduce you to someone in the used machinery business who can find a nice machine with tooling (which costs as much as the machine!) that’s within your budget.

My 2 cents.

*=almost any high school student who hasn’t been taught.

Get someone who knows what to look for and find a used Bridgeport Series I/clone (Webb, Acer, Sharp, Lagun, Wells Index, many others) that’s in good condition.

You can likely find an older one for about $2000-2500. This is something that’s worth waiting and saving more if it means jumping from a Chinese benchtop mill to a US/Taiwanese/Spanish (Lagun) made knee mill. There’s no comparison.

I would probably start with a lathe first if you have no machine tools.