CVT Transmission 2

[enginesrus]

Does anyone know what the hardness is for the steel balls used as drive keys for the variators, on Jatco CVT Transmissions?

 

[enginesrus]

I suppose a guess is needed, I think they would be in the Rc 58 to 62 range.
Now any ideas of the failure mode of the balls and grooves for the variators? I have some thoughts on it.

 

[enginesrus]

What material are the CVT belt ring packs usually made from?

 

[TugboatEng]

Can you explain the failure you're seeing? Are the balls leaving indentations in the grooves?

 

[TugboatEng]

It looks like the current material used for push belts in CVT transmissions is a grade of 18% nickel maraging steel.

https://nickelinstitute.org/media/1598/18_nickelmaragingsteel_engineeringproperties_4419_.pdf

18Ni 16.5Co 5Mo appears to be the current alloy. I think that is a development after the linked article. Older CVT transmission used the 18Ni1900 alloy.

 

[enginesrus]

Yes sometimes the balls will indent the grooves especially on the softer shaft. I feel as that progresses, then the movement of the balls is restricted during the axial variating movement , then the next thing to happen is cracking the surface of the balls, and then total destruction of the grooves especially the shaft, in some cases the deformed balls then wedge and split the sliding hub. There are some videos on youtube showing the destruction.
The belt rings have been known to crack and cause major failure, probably most always caused by belt degradation from slipping and then the rings contact the sheave thus causing a stress riser on the edge of a band or or bands (there are 12 per side). I have a CVT torn down that has over 120K miles and am impressed how good of shape the pulleys and belt is, it likely never had a fluid change either. It had low fluid pressure.
Its too bad the JATCO CVT's have the bad reputation, there is some room for improvement in many areas but the basic architecture is pretty spot on. Having a torque converter is to me a better way to go than having a start clutch like the Honda's have. I'd like to know how well that works in stop and go slow traffic and lot of hilly starts.

 

[TugboatEng]

If I had to guess a cause for the indentations is the is false Brinelling. If the variator isn't varying constantly, the balls will squeeze the lube out and go metal to metal. No amount of hardness is going to fix this problem. A redesign that uses larger contact patches, materials that don't fret against each other, and programming that keeps the keys moving ever so slightly are going to be possible solutions if this is ever considered again. Mimicking shift points likely exacerbated the problem as it always put the balls in the same locations.

 

[enginesrus]

One fix is the use of pins to replace the balls. For loading the extra surface area is nice. There maybe a small issue with sliding friction from that addition though. Yes there are many areas in these that could stand much improvement, and along with some parts design, using the proper materials would be a huge help for longevity, as well as a very refined filtration system. I'm always for the addition of some weight to components that are important in any vehicle we travel in. There are other areas that weight can be saved.

 

[EdStainless]

Programing to prevent the unit from sitting in the same position is a big step.
My wife's Toyota has mock shift points, but they are slightly randomized, not the exact same every time.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[GregLocock]

When I started work JaguarRoverTriumph were very excited about the Perbury CVT,

http://autoindustrynewsletter.blogspot.com/2017/05/torotrak-signals-end-of-era.html,

which is a toroidal CVT using discs as the transfer wheels. The pressures involved were huge and it was basically a metallurgy/lubrication experiment.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[enginesrus]

The multi segment push belt seems to work very good. It also is or was a metallurgy/ lubrication experiment, and after witnessing a 120,000 plus mile CVT variator pulley faces, I am very impressed how well
that system as a whole lasts. I suppose the fluid plays a huge part in the operation of that variator system, then of course the material and heat treat selection. I have not seen any of the chain type variator arrangements yet, it would be interesting to see how they compare. I have not studied the control system for the variators yet, I know another main failure mode of these transmissions is loss of fluid pressure, mainly from worn aluminum valve bodies, sticking valves and worn flow control aluminum valve. In my experience cast iron worked good as a valve body, only thing that ever stuck valves in the old days was gummy goo.

 

[enginesrus]

I have heard some trans shops will sand the variator pulleys in a lathe if they are blemished, does anyone know the surface finish requirements are for multi segment push belt and what grit is used during that process?

 

[TugboatEng]

These questions are asking proprietary information. You are attempting to reverse engineer a process so you're going to have to come up with these numbers on your own. There are a couple of ways:

Find a lab that can make a surface roughness measurement on the good parts of the variator. Work with your abrasive manufacturer to determine an abrasive that can produce that finish.

Or, take some scrap parts and experiment until you get the same surface finish.

In these cases, you generally can't go too fine on the finish so as long as you can produce a smoother part you'll be good to go.

 

[enginesrus]

I just recently found some information, the one particular article mentioned using 320 grit. Then another mentioned scotchbrite, but that was not to remove belt deposits.