What are the driving characteristics of using too (1) low and (2) high of a viscosity ATF in an automatic transmission?

[JasonATaylor]

If you are driving a car with an automatic transmission and the ATF fluid has the wrong viscosity, what will you be able to notice from the driver's seat? Please reply for any cases you have knowledge of. Obviously, the two cases are in which the viscosity is (1) too low and (2) too high.

Please don't warn me (or readers) to use non-oem fluid. Viscosities lower with time, climate, and other factors anyways, so it's not even a correct answer, let alone relevant to my actual question. If you absolutely are unable to answer the q but feel a need to post, or you must know of why I'm even asking, feel free to pretend you took your car in to a shade-tree mechanic to do a transmission flush and you wanted to know if he or she used the wrong fluid or not and what exactly was really wrong and how it could be fixed without necessarily needing to buy the OEM fluid.

 

[bullwinkle]

Too high might mean difficulty getting into gear or shifting in cold weather (an issue the 10R80 in the Transit 250 in my sig has even with the correct Mercon LV in freezing weather). Too thin could mean low/insufficient line pressure from a (worn?) front pump, as I’ve heard can happen using Dex VI in place of thicker Dex Ii-erratic operation (slipping clutches, etc.). My personal experience is Type A in place of Dex/Mercon in my MGM’s power steering, much better warm operation (higher fluid pressure).

 

[joegreen]

May notice the shift speed from one gear to the next increases or decreases. I can’t say though which way it would go though. Especially with modern computer controlled transmissions.

 

[Tkonrr]

I've noticed quicker firmer shifts with thinner atf. I prefer it for good performance in extremely cold conditions. There is a fan on my aftermarket atf cooler so the atf doesn't get too thin when under heavy loads for long periods.

 

[MolaKule]

If you are driving a car with an automatic transmission and the ATF fluid has the wrong viscosity, what will you be able to notice from the driver's seat? Please reply for any cases you have knowledge of. Obviously, the two cases are in which the viscosity is (1) too low and (2) too high.

Please don't warn me (or readers) to use non-oem fluid. Viscosities lower with time, climate, and other factors anyways, so it's not even a correct answer, let alone relevant to my actual question. If you absolutely are unable to answer the q but feel a need to post, or you must know of why I'm even asking, feel free to pretend you took your car in to a shade-tree mechanic to do a transmission flush and you wanted to know if he or she used the wrong fluid or not and what exactly was really wrong and how it could be fixed without necessarily needing to buy the OEM fluid.

What is the viscosity variation you're asking about? A 7.5 cSt ATF may not shift as well in cold weather as a 6.3 cSt fluid but assuming your oil pump is good and you have no valve varnishing that delays shifts, I would say not much.

The friction modifier in the ATF additive package plays a major part in shift quality.

BTW, please tone down your rhetoric as parts of it border on the offensive.

 

[JasonATaylor]

I've noticed quicker firmer shifts with thinner atf. I prefer it for good performance in extremely cold conditions. There is a fan on my aftermarket atf cooler so the atf doesn't get too thin when under heavy loads for long periods.

Thanks Tkonrr! This was the kind of answer I was looking for. So, to summarize, at least for the older automatic transmissions (not dual clutch), with the planetary gears and the bell housing torque converter with lockup features (no clue about ZF's 8HP), for case (1), the viscosity is too low, the car will shift too quickly causing the car to lurch during gear changes. By process of elimination this would imply that case (2), the car will shift too slowly and slide too long.

Can anyone confirm these two conclusions?

 

[mva]

Our old 2001 Civic tranny would slip like crazy, especially in cold weather, until the fluid heated up and got thin enough to flow through the plugged up internal filter screens. My solution was to use Dex 6 and thin it even more with seafoam which kept the car running for a few more years.

Unplugging the internal screens was only possible by removing and dismantling the transmission - poor design by Honda.

 

[dumkid]

Thicker fluid gives firmer shifts when your valve body seals are dying. I've never tried thinner fluid but it makes sense that thin fluid shifts faster if your transmission can build enough pressure on thin stuff

 

[mdocod]

I put Valvoline Maxlife ATF in our XC70 right after we bought the car used (like 7-8 years ago). Was "flushing" out the factory fill "lifetime fluid" that had 95K miles on it and was in terrible shape (I think I pumped 3 gallons through it to dilute out most of the old fluid). The Valvoline bottle claimed cross compatibility with the specification called out in our car manual, so I "assumed" it would be a good fluid for this application. Also, at the time, BITOG really had nothing but great things to say about Maxlife ATF, it was like, the "best" thing since hydrocracked bread.

One of the "behaviors" I sometimes noticed in this transmission over the last many years, is that when I would stab to about the half-throttle mark when already moving about 20mph, like coming out of a sweeping corner to merge onto a major thoroughfare, it would grab 3rd gear and do this weird thing with the converter unlocked that felt like a CVT... engine RPM's would just hold steady around 3500RPM as the vehicle sped up, eventually it would grab another gear or lock the converter, but this weird "CVT" controlled-slip like behavior always struck me as odd for a 5 speed auto, and I wasn't sure if it was working as intended. Almost felt like the converter had a variable geometry. Otherwise the shifting felt reasonable but a little harsher than I would have expected a Volvo to have been programmed to do, but I never had a baseline for that normal was supposed to be so we just lived with it. When manually shifted I also always noticed shuddering on larger downshifts, like pulling down to 2nd gear at 40mph.

At the 145K mile mark I decided to swap the fluid again. I did some more research before deciding on what fluid to buy this time. I discovered that the OE fluid specified for this car is in fact, a grade thicker than Valvoline Maxlife. I don't know why Valvoline claims cross compatibility here but they probably shouldn't. I ordered 12 qts of Mobile 3309 from Amazon for this fluid change. The weird CVT-like shift-flare-miss-shift-something behavior seems to have gone away. Now when I stab half throttle in that type of situation at around 20mph it just grabs 2nd gear unlocked for a moment, RPMs climb in 2nd to about 4000rpm, then it shifts to 3rd, engine RPMs climb normally again, then locks the converter smoothly, then climbs RPM's again; the way I would expect it to behave. From stopped, during normal acceleration, the shifting between all gears feels more like I would have expected Volvo to program the shifting to feel in a car like this. The transmission feels more confident and smooth in general. It finds the right gear both more quickly and more gracefully than before. Also, the shuddering on downshifts is less pronounced.

Thanks Valvoline. If you could claim a few more cross compatibilities that are not true, that would be great.... Oh, I forgot, you did this to my H3 too with your "MTF" that claims cross compatibility with a GM spec that is 2 grades thicker than your product. Luckily I only ran that one for about 10K miles before swapping to some MT-90 from Redline, which has been perfect in that transmission.

 

[JasonATaylor]

Thicker fluid gives firmer shifts when your valve body seals are dying. I've never tried thinner fluid but it makes sense that thin fluid shifts faster if your transmission can build enough pressure on thin stuff

So, in the spirit of maximizing the value of this thread, let's attempt to really nail what's going on in detail.

Correct me if I am wrong, but what you are implying, including other contributions to this thread thus far, is that there are four factors in the torque "jerkiness" equation during automatic shifting:

bell_housing_torque(viscosity)
valve_body_torque(viscosity)
fluid_pressure(viscosity)
friction_modifiers(brand)

and that in most transmissions, the valve_body_torque factor dominates (i.e., has the highest slope), and that it has a positive derivative. This means that as ATF viscosity increases, so does shift jerkiness. Is that correct?

My intuition suggests: the bell_housing_torque slope should be positive, but, apparently, based on the posts thus far, it just isn't important. I also would postulate that the fluid_pressure slope should be negative because as the fluid becomes thicker the pressure upon the lock-up plates should be lower. At least, this is how I am trying to process mva's datum.

It is a rather complex question, so I really appreciate all of the posts thus far helping me sort this all out. My own tranny needs a software upgrade to be less jerky, but the dealer's estimate exceeds what I paid for the car, so hopefully I can get a handle on the options.

 

[meep]

I have experienced this twice with two different transmissions.

first was a gen 1 tundra 5AT. Dealer had flushed with BG universal before I bought it. That had the same VI as a standard dex3, but it wanted toyota WS, which was a low viscosity. I bought it that way. It shift firmly which was good and I would not have thought anything was wrong.

when I found the records from the servicing dealer after I bought it (inquiring about timing belt history), they told me about the ATF. I soon did 3 D/F to the proper VI with an amsoil fluid. The shifting was notably improved. they became quicker - probably by as much as a 40-50% reduction in time spent. While the engagement was still positive, if not even more positive, the perceived ‘bump’ into gear was much less. It felt more like a well-executed clutchless heal-toe. Going to the thinner fluid was a much better move.

ford f150 with the 10 speed. mine was having big problems and the dealer wasn’t helpful. It’s well documented on this site. I like the truck and so thought I would experiment. Amsoils recommended fluid is LV, while the truck is ULV. I did not like it at all with the thicker fluid. the difference was exactly the same description as with the tundra. Re-D/F’ed at home with ULV and at least the parts that shifted well returned to a better place. Soon after discovered that a small splash of LG red resolved whatever was sticking in the VB.

bonus - went from dex3 to dex6 (as was recommended) in an AW4 in a very high mileage 1993 JGC. I thought the thinner fluid would be a disaster. On the contrary, it was excellent. It did seem to quicken the shift process with nothing undesirable that i observed.

 

[dumkid]

So, in the spirit of maximizing the value of this thread, let's attempt to really nail what's going on in detail.

Correct me if I am wrong, but what you are implying, including other contributions to this thread thus far, is that there are four factors in the torque "jerkiness" equation during automatic shifting:

bell_housing_torque(viscosity)
valve_body_torque(viscosity)
fluid_pressure(viscosity)
friction_modifiers(brand)

and that in most transmissions, the valve_body_torque factor dominates (i.e., has the highest slope), and that it has a positive derivative. This means that as ATF viscosity increases, so does shift jerkiness. Is that correct?

My intuition suggests: the bell_housing_torque slope should be positive, but, apparently, based on the posts thus far, it just isn't important. I also would postulate that the fluid_pressure slope should be negative because as the fluid becomes thicker the pressure upon the lock-up plates should be lower. At least, this is how I am trying to process mva's datum.

It is a rather complex question, so I really appreciate all of the posts thus far helping me sort this all out. My own tranny needs a software upgrade to be less jerky, but the dealer's estimate exceeds what I paid for the car, so hopefully I can get a handle on the options.

Thicker fluid generally increases pressure (including clutch apply pressure) while decreasing flow. So with a thicker fluid it takes longer for the fluid to get to the clutches, but once it gets there the clutches lock up tight. Thin fluid gets to the clutches faster, resulting in faster shifts, but it also flows through worn bushings/cracked seals easier, which means a worn transmission won't be able to keep the clutches from slipping because it can't build enough pressure to lock the clutch plates together. Keep in mind that most automatic transmissions have multiple clutch packs and these clutch packs can engage in different patterns to make different gear ratios.

Im no expert on automatic transmissions, i just watch videos about them so I could be totally wrong

 

[nascarnation]

Most all modern automatics are looking at shift slip time (via the speed sensors) and will adapt the solenoid pressures toward getting to the desired profile. That's why with an Allison for example, you can run 0W20 engine oil, Dexron, or 15W40 engine oil without changing electronic calibrations. Having said that, at 40 deg below and using 15W40 you may well see some pump priming issues but that's another aspect.

 

[Dave1027]

My experience has been transmissions act differently with different fluids. My old BMW's transmission (which was designed to use a Dex3 type fluid) hated Dex VI even though many people swore it should have worked fine. They said Dex VI is backwards compatible with Dex3. I don't think that is true. After experimenting with different Dex3 type fluids I found that Pennzoil high mileage ATF hit a sweet spot and provided the best shifting and torque converter traction.

 

[clinebarger]

The friction modifier in the ATF additive package plays a major part in shift quality.

^^This!

 

[MolaKule]

So, in the spirit of maximizing the value of this thread, let's attempt to really nail what's going on in detail.

Correct me if I am wrong, but what you are implying, including other contributions to this thread thus far, is that there are four factors in the torque "jerkiness" equation during automatic shifting:

bell_housing_torque(viscosity)
valve_body_torque(viscosity)
fluid_pressure(viscosity)
friction_modifiers(brand)

and that in most transmissions, the valve_body_torque factor dominates (i.e., has the highest slope), and that it has a positive derivative. This means that as ATF viscosity increases, so does shift jerkiness. Is that correct?

My intuition suggests: the bell_housing_torque slope should be positive, but, apparently, based on the posts thus far, it just isn't important. I also would postulate that the fluid_pressure slope should be negative because as the fluid becomes thicker the pressure upon the lock-up plates should be lower. At least, this is how I am trying to process mva's datum.

It is a rather complex question, so I really appreciate all of the posts thus far helping me sort this all out. My own tranny needs a software upgrade to be less jerky, but the dealer's estimate exceeds what I paid for the car, so hopefully I can get a handle on the options.

"bell_housing_torque
valve_body_torque"

You might to define those and show a graph of same.

"friction_modifiers(brand)" I am not sure what you are asking here. In ATF formulations, a myriad of commercial friction modifier chemistry is used.

Lastly, why are you using underscores between words?

 

[JasonATaylor]

They said Dex VI is backwards compatible with Dex3. I don't think that is true.

Well, it all depends upon how you look at it. Using a thinner oil in a machine is always 100% compatible if your real goal is to get it into a yard. And it will get you to where you need to go until it fails and you need to purchase a new unit. Sales are inverse of longevities.

 

[JasonATaylor]

"bell_housing_torque
valve_body_torque"

You might to define those and show a graph of same.

"friction_modifiers(brand)" I am not sure what you are asking here. In ATF formulations, a myriad of commercial friction modifier chemistry is used.

Lastly, why are you using underscores between words?

Sorry if you are lost. The Q is asking about an end effect that is actually a function of several different components all operating in series. So, one has to look at all of the parts of the system to figure out what's going on but the Q itself is just about how to detect too low or too high of an ATF viscosity. The _ is because these are functions which in most systems cannot have spaces, such as C++ or LaTex, a mathematical formatting system. I think each function should be multiplied. It's possible that two effects might cancel out another effect. brand is actually a variable while friction_modifiers is a function although I probably should have written friction_modifiers_torque(brand) to denote that it is to be a torque change function since a poster wrote that it is also really important.

 

[JasonATaylor]

"Most all modern automatics are looking at shift slip time (via the speed sensors) and will adapt the solenoid pressures toward getting to the desired profile."

Doesn't this mean that both valve_body_torque(viscosity) and fluid_pressure(viscosity) are one, so they have no effect?

 

[MolaKule]

Sorry if you are lost. The Q is asking about an end effect that is actually a function of several different components all operating in series. So, one has to look at all of the parts of the system to figure out what's going on but the Q itself is just about how to detect too low or too high of an ATF viscosity. The _ is because these are functions which in most systems cannot have spaces, such as C++ or LaTex, a mathematical formatting system. I think each function should be multiplied. It's possible that two effects might cancel out another effect. brand is actually a variable while friction_modifiers is a function although I probably should have written friction_modifiers_torque(brand) to denote that it is to be a torque change function since a poster wrote that it is also really important.

I asked some simple questions

"bell_housing_torque
valve_body_torque"

You might to define those and show a graph of same.

"friction_modifiers(brand)" I am not sure what you are asking here. In ATF formulations, a myriad of commercial friction modifier chemistry is used."

and you come back with.....this?

I am not lost and I can truthfully say I have had more higher mathematics and programming courses than you will ever see.

So if you are truly serious about a technical discussion then define your terms and answer my questions.

You may find some of your answers here as well:

Automatic Transmissions - A Study - Bob is the Oil Guy