Making a transmission more efficient: film strength versus thickness

[redwraith94]

What is the relationship (generally speaking) between film strength, film thickness, and viscosity:

If I polish / superfinish the gear teeth, and bearings on trans, can I go with a lighter viscosity oil, to improve the efficiency of a manual transmission, while still keeping reasonable safety margins?

The gains:
http://www.remchem.com/cms/index.php?page_id=2

The process:
http://www.remchem.com/cms/index.php?id=9

I keep seeing advertisements stating things like 'synthetic oils have 5x the films strength of conventional oils...' I know this doesn't mean I can run an oil that is 1/5 the weight (assuming that amsoil's claims are correct), but how much lighter of an oil could I run, using a quality synthetic? Especially if I go through the trouble of superfinishing the gear teeth, and bearing races.

The trans this is going in calls for ATF fluid. It is for an old Ford Taurus SHO.

 

[marcusl]

Since ATF must be balanced for both lubrication and hydraulic properties (ie., good operation in the fluid pump and torque converter in an automatic transmission), it is conceivable that a specialty oil intended for manual trans. lubrication would perform better. Polishing gear teeth might give tiny gains, messing with the bearings could potentially make things worse (some bearings are slightly porous to hold oil).

If you are racing, then you should talk to the manufacturers directly about oils since many support racing programs. If you aren't a professional racer, it's unlikely that any changes you make will be noticeable or even remotely worth the money.

 

[chemisttree]

I would think that a manual transmission would suffer from more parasitic losses as you lowered the viscosity. That loss would come from the oil splash (enhanced at low viscosity) and oil aerosol in the gearcase "fouling" the geartrain. There is a name for this type of loss but I'm at a 'loss' to remember it.

 

[redwraith94]

Thanks for the replies.

I was assuming the lower viscosity would be easier to splash around, and would have less friction on the gear teeth themselves.

http://www.remchem.com/cms/index.php?id=9

This site claims an extra 4hp of efficiency in a nascar trans (3rd gear, 6400 rpm ~378hp)

I don't know if it is true, or not. Regardless I am polishing all of the gears in the trans. I also don't know how the viscosity of the oil will effect the different components, but if anyone has any ideas...

 

[markw]

I am afraid that I'm probably not going to give you the direct answer that you crave, but:

• superfinishing looks nice and may or may not help much
• I don't think that there is a simple relationship between film strength, film thickness, and viscosity (by simple here, I mean a relationship that I can summarise in an equation of arbitrary complexity - this may not be your definition of simple)

Superfinishing of any kind (and there are a number of vendors offering superfinishing - hadn't heard of this particular one and so it isn't a in any way a comment on them) offers a smoother finish 'out of the box' than standard finishing. That much is in the name, more or less.

The problem comes when it comes to the exact alignment of that finished area. In the conventional situation, running in provides the 'finishing' and using that the finished area is automatically aligned in the corrrect way; in a separately superfinished part this is not necessarily the case and in something like an extended contact patch, maintaining that alignment at a submicron level over a a significant length may just not be possible or there may be enough float for it to occur. In this case, supefinishing just isn't a wear advantage (although it may be a stress raiser advantage - lower level of stress - at, say, tooth roots, where running in doesn't help).

So, whether superfinishing helps depends on part design and tolerancing and the wear regime. On the other hand, it is only in rare cases that it makes matters worse.

Now coming to you question about film strength: film strength is largely a phenomenon that can be measured/defined at low film thickness. Thicker oils can certainly help in pressurised lubrication, but if you try one of those testers that grind the balls into the bearing surface (to measure film strength), you are in a situation very unlike pressurised lubrication. Also, in this situation, anything that uses polar molecules that have a tendency to 'stick' electrostatically to the bearing surface, will have a worthwhile advantage over a non-polar lubricant of similar viscosity.

So a thin ester lubricant can have a lubrication advantage over a thicker mineral one. (And there seems to be an anamolous effect in mixing ester and pao lubricants that I don't understand; film strength seems to peak at about 15% ester and I'm not sure about the mechanism for that). But you also have to be careful with a manual transmission in that synchronisation depends on visous drag and if you overcome viscous drag, you don't get synchronisation (assuming that you wanted that).

but how much lighter of an oil could I run, using a quality synthetic?

By now, you'll have realized I'm not going to give you a number. If you ask an oil supplier, you'll probably get an answer like 'one grade lighter ought to be fine'. This is probably correct, but you will probably not get a guarantee (certainly not from me, probably not from the supplier...).

Depends how many transmissions you are prepared to wreck doing the development work. If you really wanted to make things more efficient, you could used sprayed lubrication and reduce the loss in the gearbox (but increase it in the spray system). This really would take development, though and I suspect its not worthwhile to you (although it can be in certain racing applications in which conventional means of raising engine power are ruled out).

 

[redwraith94]

markw, thank you very much for the informative reply.

I have spent a total of about 150 hours so far polishing the ring for the differential. I believe it has the largest tooth surface area, and sees the most force of any gear in the transmission. I used 3 different grits of hand stones, and I have moved on to diamond paste. I used four different tubes total, the average particle diameters are 32, 16, 4, and 1/2 micron. and after this I want to use some .05 micron sapphire powder that I purchased from a website (rockhound.com)

I realize that this is an art form, but perhaps for now it would be best to run normally weight fully synthetic oil, with the addition of some submicron Molybdenum disulphide.

Two quick questions for you, do you think that adding Moly to the trans fluid (if I use the recommended weight) would interfere with the synchros? If they do, that is ok. One of the biggest reasons I am building this car / modifying it the way that I am, is to learn. If I break stuff in the process that is ok. Would making the surface rougher on the synchro sets / blocking rings help them engage faster? They do have upgraded blocking rings with a fiber backing, unfortunately these are out of production, and the stock finish is pretty darn smooth.

If I finish the gear teeth down to .05 microns, which I calculated to be 2 microinches. Would it interfere with the oil retention for them? I read a 'patent' that stated 3 microinches Ra was the finest finish before oil retention became a concern, but I can't find another source that confirms that. It isn't a big deal for the ring gear, since it sits in the fluid, but for the other gears, that I will hopefully start polishing this weekend, I am not sure that such a smooth finish is good for them.

I also have debated switching the oiling system to a pump / spray setup, as you mentioned. I am not sure yet about it though. Again, thanks for all the feedback,

Mike

 

[markw]

Two quick questions for you, do you think that adding Moly to the trans fluid (if I use the recommended weight) would interfere with the synchros?

I think it could: syncro depends on the drag pulling the speeds of the rotating parts together as they close. If moly affects that drag, then syncro would be affected (OTOH, moly is effectively a surface treatment and by the time the surfaces have got that close, it might all be over. I can't be sure.)

I have certainly seen things like aktiv8 and prolong claim that they don't affect syncro, and the same general argument should apply to moly, but I don't know.

I realize that this is an art form, but perhaps for now it would be best to run normally weight fully synthetic oil, with the addition of some submicron Molybdenum disulphide.

I think I would be relying more on the normal running in process to produce the desired surface finish. So my preferred approach would be to run a normal lubricant (whatever that means - not a 'super' lubricant, in any case) for a short distance and during this running in period apply increasing loads for a short period (so a cadence of quick blasts of load, with a cooling off period between). This should polish the wear surfaces. Then change to the super lubricant.

Relying on the polishing process depends on getting both the surface finish and the alignment of the surfaces (with bearing float and transmission distrtion) and the latter is not easy.

Would making the surface rougher on the synchro sets / blocking rings help them engage faster?

This is an interesting idea, but my guess is that it won't work, at least in the obvious implimentation. My feeling is that once you are out at the thickness of the oil film, it won't have that much effect, and it may make wear on the syncro rings (down to the point at which they are smooth again) worse.

If I finish the gear teeth down to .05 microns, which I calculated to be 2 microinches. Would it interfere with the oil retention for them?

Again, my guess is that oil retention is not a factor here as the teeth run submerged in oil. Another view is that, given that the wear surfaces 'polish up' after a short period of time, there never was any chance of having good oil retention on the wear surfaces by surface roughness.

You will note that my answers have included the word guess a lot; I don't know and these are my best guesses.

 

[redwraith94]

It has been awhile, and I finally got around to polishing the ring gear, and some other parts in the transmission.

Here is a pic for your viewing pleasure:

http://img171.imageshack.us/img171/1462/0524082141pk7.jpg

http://img171.imageshack.us/img171/3417/0524082144cf5.jpg

http://img171.imageshack.us/img171/4438/0524082143ei4.jpg

As installed on the differential (a quaife torque biasing unit).

http://img364.imageshack.us/img364/2553/1104081604iq3.jpg

Unfortunately I have not gotten the opportunity to run the car yet, or dyno test anything, as there is a lot more wrong with it than was originally disclosed to me.

The trans is getting there though, and I used the following to polish that gear:
3 different Norton mold / die polishing stones.
4 grits of diamond paste 16 / 8 / 4 / 2 microns
Final polish was done with Linde A (.3 micron Sapphire dust). Which is what you see in the first picture.
Lots of elbow grease, and about 1,000 qtips!