ZVW30 Transmission Oil Analysis

I am a great fan of analyzing transmission oil because unlike engine oil, it tends to stay in service at least 10 times longer. Furthermore, earlier studies of transmission (aka., transaxle) oils revealed the NHW11 service interval needed to be shorter, 30k miles instead of 60k. The NHW20 using Type WS looks to be working fine at 60k miles.

When I took my wife's ZVW30 in for the first change, I had the transmission oil changed and got a sample, about a cup. It was opaque but still smelled like oil. For sure, I wanted to remove the initial SI leached from the sealant as well as the early 'lapping' as the gears wear in. Also, I am very interested in seeing any evidence of oil oxidation (more about this later.)

Here is the initial sample report and there will be updates in the future. Oh and one last thing.

I am no longer in the business of collecting oil samples from others and sending them out to be tested. In the past that was the fastest way to collect a population of oil samples at different mileages to understand what was going on. But the ZVW30 is new, less than four months in North America.

If you want to participate in this study, feel free but please arrange for your own samples to be send to a quality lab for analysis. I have history with R&G and they provide 40C and 100C viscosity testing, which I think is an important metric equal to the wear and additive content.

Go here if you want to see a reference sample of Type WS and why transmission oil testing is important.

Enjoy,
Bob Wilson

 

Ok, this is the oil analysis that I'm interested in:


Ok, the reason I'm interested is to find out how the viscosity decreases over time. I spoke with James, the lab tech, and we're going to do some special tests to look for micro-dieseling. Micro-dieseling is a type of oil oxidation. We're also going to check the particle load.

ANALYSIS

The goal is to estimate the miles to a 15% loss of viscosity:

	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6	Column 7	Column 8
0	C	cSt 0mi	cSt 5kmi	%	miles	~15% mi
1	40	24.6	23.3	-5.3%	5	000	14	192
2	100	5.5	5.3	-3.6%	5	000	20	625

.
This suggests the next change needs to be after 15-20k miles of service. This initial rate of viscosity loss is similar to rates seen with the NHW11s.

Other experiments have shown the silicon is leached from the case sealant. The highest concentrations in the past, 227-258 ppm, suggests we've taken out half of the silicon and leached sealant elements.

The iron, aluminum, copper, nickel, and manganese appear to be wear materials but some of them could be leached from the sealant. We'll need another sample closer to 15-20k miles to gain insights. Looking at iron, this rate .0042 ppm/mi is consistent with the .0037 ppm/mi seen with Type WS in an NHW20 transmission. We're probably seeing local spike due to early wear. The next test will give a better idea of the long term wear.

I suspect the boron is boron nitride, an anti-friction additive, that is used up over time. Interesting, I see no evidence of anti-friction, molybdenum suggesting its other characteristics may cause material problems with these transmissions.

It took about four months to reach 5,000 miles. At this rate, it will take a little over a year before taking our second change and analysis.

Bob Wilson

 

I've gotten the microscopic examination report:

(see attached)

We also had a particle count:

4/ml - 99999
6/ml - 9959
14/ml - 18
20/ml - 4
30/ml - 2
40/ml - 0
​

Mostly we're seeing very small, very fine particles and not much in the larger sizes. These are not in the sizes that an ordinary filter would handle, ~14 u. The smaller particles would take a micro-filtration system and commercial ones are pretty expensive.

The acid number is up a little, 1.49 vs. 1.41, we'll watch it over time. This is our primary indicator of micro-dieseling.

Silicon leaching is evident but lets compare the sealant test, 8 months soaking Type WS in a jelly jar, values and a virgin Type WS sample with those from this change:

	Column 1	Column 2	Column 3	Column 4
0	element	virgin	ZVW30 5k	sealant
1	Silicon	3	103	1280
2	Boron	59	47	24
3	Phosphorus	245	247	258
4	Calcium	109	120	124
5	Zinc	0	10	69

.
My thoughts:

• silicon - sealant is the likely source being leached
• boron - an additive to reduce friction, the loss when exposed to sealant is curious.
• phosphorus - an additive, nothing remarkable, a formulation change?
• calcium - normally an additive, a formulation change?
• zinc - sealant is the likely source, one patent describes using zinc oxide to improve surface characteristics

The red oxide observed may be from the sealant. Dow reports using "iron oxide" in some of their products and the sealant is red in color. However, there was no iron found in the sealant dish test (8 months soaking in a jelly jar.) Speculation on my part, sealant under stress may shed small particles.

There are some reports from Amsoil sites claiming leaching of silicon sealants leads to foaming in oil. But I've not found anything specific.

BOTTOM LINE - IS A 5K TRANSAXLE OIL CHANGE WORTH IT?

This is what I'm after:

• reduced rolling resistance - oil debris increases friction and generates local heat, the enemy of oil
• viscosity tracking - oil wears and loses its ability to reduce drag and protect surfaces ... I want to understand the rate

My early change removed some wear material and sealant leached compounds. Also, there appears to be ~5% viscosity loss after 5k miles. So I feel comfortable going 15k before my next change but this satisfies my curiosity.

I'm not doing transaxle oil changes for extending the transaxle life. The ZVW30 transaxles have been out for less than six months and we have no failure history. The earlier NHW20 transaxles have been very robust with very few documented failures. The NHW11 had one go over 350k miles and the hand full of NHW11 failures have been electrical, not mechanical. But rolling drag is something I do worry about.

My earliest hill roll-down tests showed a measurable reduction after an early transaxle oil change. Given it is a 'free' drag reduction, keeping the transaxle oil clean seems like a small price to pay. Not to worry, the used oil is recycled.

Bob Wilson

 

Excellent! Your 11,000 miles and my 5,000 miles means we can start to plot rates. But a couple of questions about the report:

• OXID - how is this measured? The reason is I am concerned (Ok, paranoid) about micro-dieseling.
• KF - huh? What is this?
• FLASH - interesting, we need to send them a baseline sample.

This is great!

Bob Wilson

 

I'm staying with Toyota Type WS:

• lowest viscosity - I've checked several alternate sources and their specifications indicate a higher viscosity. If I were in a hotter climate, say the desert SouthWest, I might consider a higher viscosity oil. But in our temperate climate, OEM is not just good enough, it is better than anything else I've found.
• lessons from Amsoil - they claimed to have a Type T-IV compatible oil but the following Cu rates were excessive. Later I learned their additives are hostile to 'yellow metal', which include copper and bronze bushings.

As for our friends in Canada, Alaska and colder climates, I would strongly suggest a block heater for the ICE and pan heater for the transaxle. The reason is viscosity increases with colder temperatures and increases the mechanical stress not only on the parts, more drag, but also on the oil itself.

Bob Wilson

 

First we have to get some data and that means samples and testing. I preferred an early change to flush any residual manufacturing debris, which I had analyzed. It will take more samples to draw any conclusions. I don't think we'll really know for at least two years.

Bob Wilson

 

With three samples, we can see more patterns:

WEAR MATERIAL

	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6
0	service miles	Si	Al	Fe	Cu	Zn
1	0	2	1	1	0	0
2	5000	103	19	21	16	10
3	11520	106	54	30	19	9
4	19687	132	110	55	21	14

bwilson4web, dcoyne78, tumbleweed

In normal wear, we expect to see an initial jump followed by a longer period of low increases. The initial service 'laps' or 'polishes' the moving parts that come in contact. We are less happy to see a rate that remains constant and get concerned if we see a rate that increases over time.

The silicon is probably leaching out of the case sealant. Other testing involving just transaxle sealant in a sample jar confirmed this. The data suggests the rate has tapered off but we'll need samples at +20k miles to draw any conclusions.

Aluminum is unexpectedly high. In ordinary samples, aluminum and silicon are the signatures of grit getting into the transaxle. But again, we need follow up samples to see what is happening. Speculation, the transaxle oil pump may have wear surfaces. We'll need to watch this one.

Iron and copper shows an initial wear rate that appears to have tapered off. This looks normal.

OIL ADDITIVES

	Column 1	Column 2	Column 3	Column 4
0	service miles	P	Ca	B
1	0	245	109	59
2	5000	247	120	47
3	11520	253	85	49
4	19687	288	143	46

bwilson4web, dcoyne78, tumbleweed

With additives we expect to see high levels initially that decrease over time by either capture in deposits or have an initial decrease and remain at a constant level.

The dramatic changes in calcium suggests an oil formulation change. However, we can't rule out different lab sensitivities (WITHOUT CALLING EITHER INTO DOUBT!!!) We'll need more samples and possibly understanding the manufacture date to track this. The bwilson4web Prius was manufactured in April 2009. But we've seen an increase in calcium over service life with the earlier transaxle oil studies.

The boron, probably boron nitride, was probably initially deposited on rolling contact surfaces and then appears to hold constant ... a good thing.

Zinc is normally part of an additive. Given the change from the reference sample, it again suggests an oil formulation change.

VISCOSITY

	Column 1	Column 2	Column 3	Column 4
0	service miles	cSt 40C	cSt 100C	Vis. Index
1	0	24.6	5.5	171
2	5000	23.3	5.3	171
3	11520	23.5	5.1	152
4	19687		4.78

bwilson4web, dcoyne78, tumbleweed

We are seeing some viscosity loss but the rates appear modest. However, if there are changes being made in the transaxle oil, we may see it in the oil viscosity too. Again, we'll need additional samples to understand what is going on. However, if I had a new transaxle that was all 'gears' and no longer had the 'silent chain,' I would see a lighter viscosity at 100C as being a good thing.

Now we can start looking at one viscosity, 100C, as a function of miles:

	Column 1	Column 2	Column 3
0	service miles	cSt 100C	decrease %
1	0	5.5	0.0%
2	5000	5.3	3.6%
3	11520	5.1	7.2%
4	19687	4.78	13.1%

bwilson4web, dcoyne78, tumbleweed

The target is a 15% loss of viscosity means the oil needs to be changed. My earlier straight-line suggested 15,000 miles and I'm waiting to 20,000 miles before my second change (at 12.5k miles, I have about 7-9 months before the next change.) Prius friend 'tumbleweed' is showing something between 20-25k miles may be the optimum target. I may extend my next change to 20k service miles, 25k total.

CONCLUSION
Three samples were taken out early, as much a flush and baseline as anything else. However, based upon these samples, something interesting may be going on. Possibly an oil formulation change and an interesting aluminum wear pattern needs watching. We do have a better understanding of when viscosity changes would mandate a change, 20-25k miles.

Bob Wilson

 

I know but we didn't really have much guidance for the NHW11, NHW20 and now the ZVW30. But unlike Toyota, we own our cars and want them to last a long, long time. So we are conducting the experiment, actually sending oil samples for analysis and <ZUT ALORS> sharing the results!

Bob Wilson