Amsoil ATF test in NHW11 01-03 Prius Transaxle

The following chart shows how the oil sample assay changed over time
and covers the Amsoil ATF tests:

http://hiwaay.net/~bzwilson/prius/pri_tt_100.jpg




Some general observations:
1) Amsoil has nearly twice the levels of calcium and phosphorous
additives. In the first change, without cleaning the pan, they dropped
significantly. I suspect they found a lot of material to combine with.

2) After the initial Amsoil flush, it looks like the Ca and P
additives 'came back.' This suggests whatever they had combined with
released them back into solution.

3) Copper wear went through the roof suggesting that the Ca and P
additives may have chemically combined with exposed copper.

4) Rate of silicon contamination significantly reduced after
installation of the vent relief tube. The slope of Si is lower than
the first transaxle oil load. Now it is possible that latent Si from
uncleaned areas brought the load up quickly. However, the chart
suggests the Si rate was significantly lower.

--- In [email protected], "Robert J. Wilson"
<bwilson4web@...> wrote:
>
> Hi folks,
>
> This is just a quick summary. I now have three Amsoil ATF samples plus
> one other to be covered in a follow-up posting. I'll try to get all of
> these reports into one of the "files" areas, probably SAT1, with a
> more detailed analysis. But starting with my most recent data:
>
> 235219 - Amsoil ATF inventory (virgin)
> 244497 - Amsoil ATF, 7k miles, w/o prior pan cleaning
> 278168 - Amsoil ATF, 27k miles, with prior pan cleaning
> 278169 - Amsoil ATF, 27k miles, sludge report
> 278169 - Amsoil ATF, 27k miles, sludge microscopic analysis report
> 278170 - Amsoil ATF, 27k miles, paper towel wipe

The PdMA oil testing results can be found in the "files" section of
"toyota-prius-sat2", a Yahoogroup.

>
> VISCOSITY
>
> Amsoil ATF appears to preserve viscosity better than Type T-IV. After
> 27k mile, the 33.7/6.6 viscosities are significantly higher than the
> 31.5/6.6 viscosity of a Type T-IV, 20k sample. If preservation of Type
> T-IV viscosity is a primary goal, Amsoil ATF appears to be a better
> choice. However, I know of no Type WS equivalent.
>
> CHECK FOR SOURCES OF ABRASIVES ENTRY
>
> This was one of PdMA's earliest recommendations but I wasn't prepared
> to disassemble the transaxle to find the source.
>
> I have been running an experiment to eliminate the negative transaxle
> air pressure by replacing the transaxle, one-way vent plug with a hose
> to the air cleaner. Unless I discover of some latent defect (like some
> leak not observed in my installation,) this experiment did not work.

Now I'm not sure. There are significant areas that can not be reached
and cleaning the pan only gets what is at the bottom. There is a
gallery to the rear that looks to hold about a cup, ~100 ml, of oil
and some unknown amount of sludge. Above the pan is a structural piece
that probably holds some unknown amount of sludge. As follow-up, I'll
PM Florian to see if he can inspect his transaxle for insights on
holding areas for sludge, oil and approaches to a through cleaning.

> The amount and quality of sludge found in these latest test results
> were disappointing. From the sludge microscopie analysis:
>
> "Ferrogram Interpretation:
> There were significant amounts of abrasive/dust/dirt particles found
> covering the slide. Moderate amounts of nonferrous white metal rubbing
> wear and copper alloy rubbing wear particles were present. A few
> ferrous rubbing wear, ferrous cutting wear, nonferrous white metal
> cutting wear, copper alloy cutting wear, tempered metal, and large
> fiber particles were also seen. The morphology of the abrasive / dust
> / dirt particles suggests possible contamination of the sample and/or
> reservoir from an external source. The presence of cutting wear
> particles suggests possible misalignment and/or abrasive
> contamination. Tempered metal particles may be indicative of possible
> overheating."
>
> This report pretty well describes the problem.
>
> CONCLUSIONS
>
> Transaxle Flush Needed
>
> My first conclusion is dropping and wiping down the pan is not enough!
> To properly clean out any latent dirt or grit we need to develop an
> effective transaxle cleaning system or methodology, which doesn't
> exist today.
>
> My current thinking is after removing the pan and wiping down all
> surfaces, we need to put the pan back and connect the drain to a
> filter and pump and then use some spray mechanism to go in via the top
> plug Hobbit identified in the NHW20 transaxle. Pressurized oil from
> the pump needs to be 'sprayed' on all internal surfaces with some
> degree of force to try and remove any latent abrasive material. It may
> make sense to use a lighter oil, say kerosene. One key element is
> having a filter or 'trap' that can be monitored to capture any
> dirt/grit and tell us when the flush is done.
>
> In truth, I have no direct experience with transaxle or transmission
> flush systems. I know it is somewhat common with regular automatic
> transmissions. If anyone has any suggestions, I'm all ears.
>
> Find The Source
>
> The only holes I know of are:
>
> 1) wheel power shafts
> 2) ICE power shaft
> 3) drain and fill plugs
> 3) top plug (NHW20)
> 4) MG1/MG2 power inlets
> 5) parking paw shaft
>
> I plan to take off the rubber boots covering the wheel power shafts
> and look for loose dirt. I'll also do a close inspection of the
> external seals.

Based upon Patrick's descriptions, this may require some way to clean
away the CV grease and fresh grease to repack the CVs. If I do this, I
will check with PdMA on what sort of analysis makes sense. In an
extreme case, it may require removing the shafts, which means changing
the transaxle oil.

> . . . If it makes sense, I may try to find a dry, powdered
> chemical that could be detected by PdMA and put some in the boots
> after cleaning. If these trace elements show up in a future test,
> we'll have confirmation of the source.
>
> I would like to inspect the space between the ICE and the transaxle.
> However, I don't know of any ports or holes to put in a bore scope. If
> anyone has a suggestion, I'd be interested.
>
> I don't see the plugs, power interfaces or parking paw as being likely
> sources of dirt/grit. Given the negative pressure maintained by the
> vent plug, it would be a stretch. However, if some of our friends with
> disassembled transaxles could take a peek?
>
> Correlation Oil Test to Sludge Report
>
> We need to quantify sludge and oil test numbers but until we have a
> way to collect all of the sludge, this remains a problem. The NHW11
> transaxle has an internal 'roof' above the pan that blocks direct
> observation of the PSD, gears and MG1/MG2. It also serves as another
> collection surface for sludge than can not easily be reached.
> Furthermore, the NHW20 does not have a removable oil pan and the
> internal structure remains somewhat unknown.
>
> In-pan Oil Filter
>
> The current screen protects the oil pump but does nothing for fine
> material. We need a filter-paper assembly that can slip over the oil
> pickup screen. Think of it as an oil filter without a can, just the
> paper assembly. This will entrap the fine grit that the pump picks up
> and passes to the rest of the transaxle. It should have an
> over-pressure relief valve to protect the filter in cold weather.

I've ordered an oil pickup tube and screen to prototype an in-pan
filter. It makes sense to have the prototype ready when I do my next
transaxle oil change.

>
> Last TYPE T-IV / AMSOIL ATF test
>
> The only advantage of Type T-IV and Amsoil ATF has is significantly
> higher viscosity at 40C (compared to Type WS.)
> . . . But the bulk of operation is going to be at
> significantly higher temperatures, about 50F / 10C. Other than the
> dirt/grit problem, they look a lot alike at operating temperatures.
>
> I have switched to Type WS in my transaxle and unless testing reveals
> some unexpected wear problem, will remain with it. In fact, I'll be
> drawing my 1k mile sample this afternoon.

I can not recommend Amsoil ATF in our NHW11 transmissions because of
the high levels of copper found in the first and second changes. The
data suggests there may be a chemical interaction between the copper
and the additives BUT this begs testing. Before using Amsoil again, I
would need to conduct a test involving copper in an agitated Amsoil
ATF bath to see if it is attacked.

Amsoil ATF appears to have excellent viscosity wear characteristics
but this is not enough if it is chemically incompatible with the
copper windings of MG1 and MG2.

Bob Wilson

 

Hi David,
<div class='quotetop'>QUOTE(David Beale @ Jul 31 2007, 10:35 AM) [snapback]488044[/snapback]</div>

There is merit in your suggestion but I think the gasoline vapor risk is fairly low. For example, the transaxle has nearly 5 quarts of oil versus gasoline vapor dilution. Also, the transaxle oil heating would tend to drive off volatile hydrocarbons like gasoline. I am more concerned about water vapor from the PVC return.

Do you know of any such filters sold at AutoZone or other auto supply stores?

Bob Wilson