deviant gearing

[Delce]

Hi gear heads
I am now a fully paid up member of this gallant band and my quest now is to determine how far you can deviate realistically from ideal gear meshing assuming light duty only

the criteria
1) Gears mesh easily without tight spots - ie can fit existing centres
2) little excess back lash to fit duty
3) Little vibration to suit duty
4) reasonable wear and life expectancy to suit duty

I was hoping to generate a figure of merit scale 0-9 ( no way - good fit)
eg FOM = 4:5:6

I am hoping that Gearotic may help determine the FOM of say 30T Mod 1.25 vs 1.27 vs 20 DP also Mod 0.8 vs 32DP (31.75) anyone any ideas?


This is intended for the benefit of those simple souls like myself who dont have access to semi pro tools such as universal dividing heads, geared heads for hobbing, motorised spin centres, CNC etc

For example
a) is it practicable to shape out the flanks of a 20DP 30T 20PA wheel to match the faces of same wheel in 14.5PA - yes the centre distance may be too large to mesh on an existing set up but additional profiling may suit - given that 14.5PA hobs are almost impossible to obtain. Yes there will be other ways to solve the problem but would a simple square file or chisel ground down to suit?. This is only for light duty not power xmission eg lathe change gears

I was faced with this problem for my Myford ML7 change gears which are 20DP 14.5PA which are expensive or unobtainable

OBTW getting DP stock gears in the uk is very expensive, only one supplier and ca $30 for a Delrin spur gear as per example, a metric equivalent is ca $5

Delce

[ArtF]

Hi Delce:

Hmm. That sounds right to me, but I'm not sure how I can help you profile it
to prove it or quantify it. Gearotic can certainly generate all the required
profiles, and will allow you to see the mesh on various mixtures, but
you can only see basically identical gears on the design screen. You can only
see other pressure angles in simulation.


Art

[Delce]

Hi Art
Maybe I didnt explain myself properly. I wasnt expecting GearoticM to provide answers and a simulation would be way too complex involving collisions etc - been there and its not very instructive, no substitute for the real thing

I will be using GM to provide me with the data and shapes which I can export dxf to a cad modeller I use, so I can examine interference overlaps and what material I need to remove from where

I have an idea that maybe a simple needle file suitably dimensioned my do the job of widening the gulley between 20PAteeth to accept a 14.5PA tooth when you are stuck with a problem of having a Mod gear1.25 (20PA) having to mesh with a 20DP gear 14.5 PA. Gear pundits only tell me to do it properly

I have tried fitting it with my Hob and a gear and it seems so very close to the naked eye but if I wasnt so confused by all the gear pundits and gear science who say YOU CANT MIX PAs. that it cant explain why what I see in front of me looks like it will work

Yes I accept its not prefect and may wear out quickly (in my lifetime) but if it gets you out of a hole then surely its worth sharing

Well thats my aim so I hope GM will give me the dimensions I need to analyse the problem on CAD

If I can find a way I would post pix on here to show CAD results using GM

Cheers

Delce

[ArtF]

Hi Delce:

Cool, let me know how it works out. Never mind the pundits, their often wrong. :)

Art

[Delce]

Hi Art
thanx for the encouragement, more often I got slapped down by gear snobs who thought my notions were heresy = positively bestial

Am I right in saying that on the design page, I cant change the centre distances between wheels eg increase it a bit for some slack

Also I cant impose any profile shifting on the teeth, ie changing the ratio of addendum to dedendum from 50:50 to say 55:45

I have my gear models on CAD and hope to post pix soonest

I will also machine up some blanks using my unspeakable gashing method as I only have simple tools

but if you are stuck with trying to make some missing change gears for your lathe - well its got to be worth a try

[Delce]

Well here's my CAD attempt to superimpose a 14.5pa on a 20 pa (20dp 30t)

as you can see it involve a broach type chisel to shape out the flank area of the tooth gulley

It doesnt appear to remove much mtl but should allow the meshing tooth full depth

next step is to make some

It wont satisfy purists but will it do the job, maybe need some running in with a tray of fine abrasive fluid?

what do you guys thinks

All data thanx to GM btw

I might try animating by stop motion by swivelling gears by 1 deg but its time consuming and tedious making a gif

[ArtF]

I dont think there is anything wrong about this at all. Trochoidal's are one are with little or
no specification. Gearotic uses an incredibly complex algorithm to trace out the actual
tooth curve but its so complex I have long forgotten how it works.
It was a sticky point in development, as were most specs that turn out to be non specific
so I implemented as math pure a solution as I could at the time based on the laws of
meshing more than any spec from gear makers or published standards.

Gears have more modifications that most are aware and dozens of special
tooth profiles for specific situations. I dont think anyone can look down at
a gear that works. :)

Art

[Delce]

I appreciate what you say Art
Tooth profiles have challenged some of the top minds worldwide for decades

Way out of my depth

I am busy making up some gears and mounting them on a test rig

I dont know yet how to make a judgement on whether a pair meshes well or badly other that by eye. I wondered if I put some load on the driven and measured the noise/vibration whether this would be an indicator

I rather get the impression that there's a lot of dark info kept secret by gear makers

One of the impressive examples I came across was gearboxes for wind turbines. They have a speed up ratio to 1:100 and have to xmit 5MW under large varying load conditions (which is a killer for them). They have a life span of 3-5 years and weigh 20 tonnes