CNC Gear Cutting

[JasonBModerator @jasonb]

With HPC no longer carrying MOD1.75 gears as stock items and not having much luck finding them as off the shelf items my thoughts turned to other methods of cutting.

A bit of CAD revealed that I could just get a 2mm dia cutter into the root of the cuts so I did a quick test which turned out well. I started with a 3mm cutter doing an adaptive path to remove most of the waste. Then a 2mm cutter again doing an adaptive to get in where the 3mm would not fit and finally a few time saround the contour with the 2mm cutter to finish the profile.Just waiting on a 9mm long flute length 2mm dia cutter to arrive so I can make one to the required thickness.

Feed rate and depth of cut were kept quite small as there is quite a lot of cutter engagement on what will be an extended length cutter, no signs of chatter on the 6mm flute length one I used for the test

[Dave SParticipant @daves59043]

Nice.

The late JS use to advocate this method.

I think he worked with Arty on a gearotic tool to create the paths

Dave

[JasonBModerator @jasonb]

I think quite a few of JS’s gears were done with the 4th axis but he may well have also done the basic spur gears flat. I know he used Gearotec which would spit out the profiles.

I did think about doing it with the gear horizontal and just indexing it around 12 times (no 4th axis) but settled on this method as I could walk away and leave the machine to it once the 2mm tool was fitted.

[John HaineParticipant @johnhaine32865]

It’s a good method provided the cutters will fit the teeth roots.  I’ve used it to make pinions and wheels for a clock train with a minimum module of just over 1mm – even then I found that I had to use a cycloidal profile though I’d have preferred to used involute, just to allow a 1mm end mill.  Profile generated as a dxf using an online tool and imported into CamBam.

Do you form the teeth on the end of a bar and part the gear off Jason?

[blowlampParticipant @blowlamp]

Nice one Jason.

I modelled a gear cutter for a friend, who had broken the back gear on his Myford lathe.

I took the gear tooth profile generated from Gearotic into MoI, where I made it into a solid model. From there, a dense STL mesh was created and exported into DeskProto for g-code generation.

After machining, a light lick on the front of each tooth to sharpen them up was all it needed to let him cut a new gear in steel, rather than cast iron like the the original.

 

Martin.

 

[JasonBModerator @jasonb]

I think Julie has shown cutters made like that on the CNC in the past complete with relief. Allows you to get into smaller valleys and do wider gears as well as custom shapes like splines.

I’ve a need for a pair of equal diameter 2:1 skew timing gears and need to have a play to see if they can be done without a 4th axis just by treating the top surface as a 3D surface when the gear blank is held horizontally in a rotary table and then just index round and repeat the cuts. Though hoping a set of these rusty ones may find their way to me first

[JasonB]

On JasonB Said:

I think Julie has shown cutters made like that on the CNC in the past complete with relief. Allows you to get into smaller valleys and do wider gears as well as custom shapes like splines.

….

Good memory, I did make a cutter on the CNC mill for cutting splines on a crankshaft:

I’ve only made one spur gear on the CNC mill, using a 4th axis, as part of an internal gear set:

Cutting the pinion:

The first gears I made on the CNC mill were 6DP bevel gears:

The bevel pinions were machined using the 4th axis and a 4mm ballnose cutter:

The bevel gears were machined using radial toolpaths on the flat using 4mm, 3mm and finally 2mm ballnose cutters:

I’ve made helical gears the traditional way on my manual mills, but if making more I’d use the CNC mill. I would use the 4th axis and hand write the code as my CAM program seems to have trouble following a helical path.

Despite correspondence with the late JS I never got on with Gearotic. So all my CAD models for the gears have tooth profiles generated directly from the involute equations. I design the gears as 1DP and then do a final scaling to the DP I need so that any errors in the profile decrease in proportion.

Julie

[HuubParticipant @huub]

Julie, nicely done, I am jealous.

[Huub]

On Huub Said:

Julie, nicely done, I am jealous.

I would totally agree, I am in complete awe of your talent.

The last time I saw true straight bevel gears cut it was done on a Gleason machine that was about 75 years old at the time. (I believe that machine was designed by a lady in the US).

You have shown how it can be done in the 21st century, well done.

Andy

[Julie AnnParticipant @julieann]

I’ve seen a Gleason bevel gear planer in action which was mesmerising, many years ago in the main workshop at RAE Farnborough when I was a student engineer. The workshop offered to make the gears I needed but I wanted to make them myself. So I borrowed the appropriate involute cutter and cut them on a horizontal mill and dividing head. The involute cutter was one of the special BEVEL ones, which are thinner than the standard cutters of the same size. They don’t seem to be available now.

I thought the same about the design of the planer, but having looked it up I think it was developed in 1874 by William Gleason. He did have a daughter, Kate, who was an engineer in the company and was integral to its success, but she would have been nine in 1974.

I have a background task to design and make spiral bevel gears, just out of idle curiosity. I understand the basics of the design but need to translate that into a 3D CAD model and then program the CNC mill.

Julie

[JasonBModerator @jasonb]

Well my cutters from APT turned up so I know what I’ll be getting upto this weekend, stay tuned.

I also found a good cure for the hard areas in some of the castings😉😉😉

 

[JasonBModerator @jasonb]

Having part machined the gear blank yesterday, I set the CNC running this morning while I got on with other things, just popping in to change tools and record a bit of Video.

I’m happy with how it turned out and will be using the method again.

As I said earlier feeds were quite slow and small depths of cuts both vertically and sideways due to the cutter just fitting into the root of the teeth which meant a lot of cutter engagement. You can see this towards the end of the video where the contour cuts are being made, as the cutter gets to the bottom of the root there is an increased amount of swarf but no sign or sound of chatter which you can get if pushing it a bit harder. And no broken cutters.

 

[Andrew CrowParticipant @andrewcrow91475]

Nicely done Jason, it certainly seems the way forward for cutting gears and it will eliminate the need for a vast array of very expensive cutters.

Not sure I will ever get there, a bit too long in the tooth now for going down the cnc route even though I did manual cnc programming about 35 years ago.

Andy.

[Julie AnnParticipant @julieann]

Looking good; the finish in the video is a lot better than apparent in the picture. So must be a trick of the light in the picture.

Julie

[JasonBModerator @jasonb]

I think it was the sun coming in through the workshop window that was on the far side and making it hard to get a clear image.

[Michael GilliganParticipant @michaelgilligan61133]

A spectacularly good ‘proof of concept’ Jason !

I would barely have thunk it possible !!

MichaelG.

 

[DiogenesParticipant @diogenes]

Look good – A-ha; profiles from F360, yes? ..think I answered my own question..

[JasonBModerator @jasonb]

Thanks both

 

The profile actually came from Alibre which has a “script” for spur gears, at least it does in my “Professional” version. F360 will also do helical and internal ones as well a sracks that Alibre can’t.

 

[David JuppParticipant @davidjupp51506]

Jason,

There is a free ‘Add-on’ available for Alibre which will produce those extra gear types (plus bevels) – it even works with Atom3D.

[JasonBModerator @jasonb]

Thanks David, I’ve found it.

[JasonBModerator @jasonb]

One problem. It will only do helical gears upto 45deg helix. No good for the matching gear to the one above which needs an angle of 63.4deg to give gears with matching pcd but 2:1 gear count.

[Author]

Posts