gear creation on CNC !! (from dwg) 2

[Tmoose]

A supplier quoted making a replacement pinion gear from a drawing we provided in a quote package. The drawing had some basic info about the metric module, pitch diameter, helix and pressure angle.

In the quote he said "We generated our own tooth form from the drawings we were given. Since we will generate the tooth profile by CNC machining, we can make any changes at the time we would machine the gear, based on the sample form you currently have."

Is CNC gear manufacturing common?

This is not a fast or fancy pinion gear, but the original was designed by others, so the info we have is thin. I can imagine the "right" CNC machine could carve a pretty accurate profile. I guess I just had not considered making gears using anything but gear machines.

I have a tracing of the tooth profile made off the real gear (which is functioning well), which in bygone days big mill manufacturers said they could use by itself to make a similar replacement pinion.

My future issue will be I certainly would not trust the info pulled off a drawing was the real tooth profile, especially these hastily cobbed drawings.

 

[israelkk]

Gear drawing doesn't mean to give the real profile. It only gives the gear module, number of teeth, gear profile shift (if exists), helix and pressure angle. the gear manufacturer uses this information to hob the gear on a gear machine. The CNC manufacturer usually has no knowledge of gear manufacturing and checking processes. If the gear needs to be heat treated the case is even more complicated. If you want your gear to perform well and last, you should use the right tooling and processes.

 

[drawoh]

Tmoose,

How well was the involute tooth form modeled in CAD? I have never modeled an involute form. I have always faked it with a straight line or with a couple of strategically placed radii. A lot of old drafting books explain how to do this.

If your involute form was modeled perfectly, the vendor's CAM system would have a chance.

JHG

 

[mfgenggear]

Tmoose

don't do it send it to a gear mfg house.
the shop probably does not have the correct gear measuring equipment.

a CNC House can make a gear from the actual geometry. so can an EDM house but they can not check it.

Take Care
MfgEngGear

 

[Bugie]

I have seen a few places CNC machining gears. It is actually a resonably fast process. All you need is the GEAR data

 

[israelkk]

Bugie

The so called "GEAR data" is limited to standard gears. If you need a profile or rack shift this gear data is useless. What about gear tolerances and accuracy per AGMA, DIN, ISO, etc. How the CNC operator will assure the gear accuracy?

 

[Bugie]

I guess we all have our opinions

http://www.spiralbevel.com/

 

[israelkk]

Bugie

Those guys are well known gear experts in gear theory and manufacturing from the Russia and I will trust them to manufacture a gear by CNC. What I meant is that a common CNC machine shop is not the place to go. However, show me how even those gear expert will CNC machine a fine pitch gear with small modules?

 

[shawnpeter]

they make CNC hobbers. plus there are CNC mill/turn hobber packages.

owning a horse doesn't mean you can ride a horse...

 

[dinjin]

Shawnpeter.
The same can be said of any shop equipment.
Glad to see the post by Bugie to see that it
can be done and done right by those who understand
gears and how to cut them. They are not rocket science.
I assume they would still be checked by the standard
methods of TCE or individual checks.

 

[Bugie]

Maybe the average shop cannot CNC gears accuratly but there are a few who can

Also Most modern Multitasking CNC lathes can HOB simple gears

There are some pretty wild gears here with CAD models

http://www.zakgear.com/Wormoid.html

Another CNC shop

http://www.nevodevelopments.com/index.html

and

http://www.bishopsteering.com/steer_mot.htm

 

[shawnpeter]

...But making the gear and be able to measure it for accuracy is another thing. I can have our plant make all kinds of stuff if given a model for them to derive tool paths from. But if they can't measure it to see if it is accurate to original design than what is the point?

Go to the right people

 

[mfgenggear]

Bugie

Yes cnc houses can make some very interesting products.
but there is more to this than just mfg.

Making soft simple gears is no issue. yes lathes & mills can cut gears. but what about the ISO or AGMA certifications.
must gear shops have specialized equipment, CNC Gear Checking equipment, or even the good old fellows Involute & Lead checker, Total Composite Inspection,& Indexing Inspection. OK then you have to the skilled people with a least 5 years experience to know what she or he is doing.
the engineers or experience person to know how to process this stuff.

I know of a company that tried making their own gears in house, purchased all the necessary equipment but fell flat.
why? because they lacked the skill to make case harden gears to the Engineering Drawings.

simple gear OK, Aircraft or Aerospace gears come a again.
I get aggravated when non gear shops are awarded gears because they are cheaper.

Cheers
Leesr

 

[changedsoul]

I work at a machine shop that specializes in cutting gears and one particular gear was cut by me on a cnc. We actually had a attachment though that could hold a hob. So I guess in a way we just turned the cnc into a gear machine.

The last place i worked they tried to cut a 15 inch O.D. aprox gear on the cnc using a couple of ball nose end mills. All though it looked cool while machining, it took forever long. I cant imagine machining a gear one tooth at a time would be or could be cost effective way to go. I think it was something like 6.5 hours to cut, and we only got two teeth done. haha

 

[israelkk]

To my opinion the real advantage of gear making by hob is that the cutting process is a continuous meshing process between the hob which is a "true" involute cutting gear and the hobbed part. Therefore, the process guaranty a gear that will smoothly mesh with any mated gear that was hobbed from the same or similar hob. While cutting the gear by any other mean such as wire EDM or common CNC machining by trying to follow a theoretical involute and trochoid, will never give the same meshing quality for the same cost.

 

[shawnpeter]

hobbers form the gear... cnc/edm will shape/profile the gear. as israelkk said, you won't be following the true involute shape, and instead you will be going off a theoretical.

 

[gearcutter]

Hobbing or shaping an involute profile does not create a "true" involute curve. What is being generated is a series of straight lines that approximate the profile.......this becomes very obvious when generating low numbers of teeth.
The most accurate ways I know of machining an involute curve is either on a worm type gear grinder or a dressed profile gear grinder.
It's interesting to note that the dressing of the profiled grinding wheel (on modern gear grinders) is CNC and is derived from the theoretical/calculated involute.


Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[tbuelna]

Actually, with the current gear design and solid modelling softwares available, creating highly accurate digital gear tooth surface topologies is not too difficult. As for that topology being a true involute form, it all depends on what the gear will be used for. Just because this pinion gear doesn't outwardly appear to be "fast or fancy" doesn't mean it wasn't carefully engineered.

High PLV, low tooth count, high stage ratio, or high L/D ratio gears always have all sorts of profile, tip, face, addendum and lead modifications that deviate from a true conjugate involute geometry. But for the average sized tooth profile these geometry mods are very small dimensionally (typically less than .001 or .002 inch) and must be tightly controlled from tooth to tooth. And if the gear contact is designed to be hydrodynamic in nature, surface roughness is also important. Making enough cutter passes with conventional CNC to limit the resulting cusps to less than the necessary micro-inch-level surface roughness limits for most gears would be prohibitive time-wise.

If the gear teeth must be case hardened, then there is the additional problem of CNC finish machining the case hardened tooth surfaces with a small diameter ball end mill. The ball end mill would logically need to have a radius equal/smaller than the root fillet dimension. The high localized cutter speeds and forces required with this small diameter ball end mill cutting hard steel could possibly generate enough heat (<500degF for most carburized alloy steels) at the tooth surface to detemper the case structure. The only practical techniques for finishing hard gears are hard shaping, grinding and honing. If the gears must be accurate, grinding and honing are the only practical options.

As gearcutter noted, generated or form ground teeth are the most accurate. A modern CNC gear grinder can easily cost more than $1 million dollars. To achieve maximum precision, these CNC gear grinders have software that compensates for grinding wheel wear, they have CMM's that constantly probe the gear teeth to check when the grinding wheel needs to be dressed, they have an integral CNC dressing tool, and they have auxiliary OD/ID grinding wheels so that the gear's bearing journals can be finish ground within the same set-up. In short, making a quality gear is all about accuracy.

Hope that was helpful.
Terry