esk8.news esk8 calc donate now

Gear Drives - Hunting Tooth Frequency

Hello,

This is my first post here, and itw kinda scary doing that.

But i wondered if you know that with gears, and especially high rpm, its wayy better to use odd numbers to decrease wear?
Flash Edit:
You have probably gotten this a lot over the course of time

8 Likes

Welcome dude, I Have no clue if what you say is true, but no doubt one of those mad scientist will be able to discuss this with you. I am just glad people like that make stuff we can ride.

2 Likes

Welcome, Luuk! Mostly good guys here, so no worries.

2 Likes

Never had someone ask that but every ratio presented to you has been already tested

4 Likes

Haha yeah thanks me too, it has to do with the same teeth hitting eachother with for example 10/40. This means each in 4 turns its thr same teeth hitting eachother i think. This doesnt happen with odd nunbers because the commas are way llonger such as ,33333333
Edit: link to this explained here
https://power-mi.com/content/troubleshooting-gears

I am studying industrial product design, this is one of the first things i learned i think. But im not an expert like the mechanical engineers here

13 Likes

Can you share the thinking/research behind this? Always interesting to get different viewpoints.

1 Like

I edited the post with a link to the topic, look up the HTF (hunting tooth frequency) part.

2 Likes

It makes sense that because the same teeth are running over each other, an imbalance (ie a teeth wearing more suddenly than the rest) will be made worse quicker over time. That happens in an evenly divisible teeth configuration I believe like 10/40. But in an odd configuration, say 11/40, the teeth that are running over each are rotated around. This helps spread the wear if there is any.

Pulley
1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
Motor
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9

In this even configuration 3/9, 1 will always mesh with 1,4,7. 2 will always mesh with 2,5,8, and 3 will always mesh with 3,6,9. Say 1 is a bad teeth, it would cause even wear only on 1,4 ,7.

Pulley
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Motor
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7

In this configuration 4/9, you can see 1 getting meshed with every other teeth so if 1 is a bad teeth, it would evenly wear with the rest of the teeth.

I understand the concern for this in applications with high rpm/load but I am curious to see if it applies to eskate gear drives.

edit: for grammar and wording

6 Likes

Hi guys nice discussion but it’s not the right thread for it.

A lot of gear drives in esk8 have gear ratios that are integer numbers. Including mine,kaly, etoxx,3dservisas,trampa

5 Likes

its a similar idea to a skid patch calculator with fixed gear bicycles… with certain ratios every time you whipskid it will be the same tire position against the ground, which will wear out the tire faster with certain ratios

^ 44/11, 36/12, 13/52, 14/42, 15/45, 16/48, 17/51, 18/54, 20/40 all only have one patch, for example

5 Likes

For a precision gear pair is good practice to do that if the design allows, or even alter the gear ratios slightly to make it happen

But I don’t see as a must in our boards, starting that usually one of the gears are made from polymers, and also the overall tolerances of the whole assembly is not as tight, and second that the expected life is way lower

5 Likes

@Kug3lis

1 Like

Interesting. Sounds like the grease would spread better if one of the gears had a prime number of teeth.

3 Likes

There are non-integer amounts of teeth?

7 Likes

How so?

1 Like

How so?

He stated gear ratios, not teeth.

3 Likes

So maybe i should clarify (as far as i can remember) , if for example you want to run a 1/4 gear ratio, you would take 10/40. In this case, the same teeth would keep touching eachother as @annihil8ted stated. It would be better to make te gear ratio 11/40 or 13/40. This way the teeth will encounter more different teeth and degrade less, especially at very high rpm as many people run here.

6 Likes

Let’s get another smart gear dude in on this:

@Avio

3 Likes

@b264

5 Likes

Nice highlight, I think I well see the point, but I never think about this (also, I’m a low kV guy…but definitely a GD rider now).
Seeing all the reports of the different gear out there from esk8de, e-toxx, Kaly, Idea, Jeff, 3Dsas, moon acboard etc. from +2500 some +5000 to +7500 miles on them without real maintenance (some just to check or that could only be change grease) …

It could be a nice upgrade to take in mind for future GD, but also, did we really need this high degree of accuracy? (@ all Gear Designer)
I agree that we could definitely accommodate a not rounded ratio number too …

I would state the same.

Arcboard maybe could help us being an automotive engineers team (right ?) that directly start production eboard with gear (sorry if Im wrong, and don’t know their name here to tag them…).
I would like to hear also the thinking of @Duffman, @3DServisas / Audrius and @IDEA on this :relaxed:

Now, how to know more ? How could we have a reallife report about this ?? Make this experience ? how many kms of testing ? what is this high rpm in our application ?

4 Likes