Much more oil data (a lot more)

Mitch, you and I must have too much time to burn.

This is great stuff. Thanks for the effort.

I ran the Murcielago hard yesterday and got the oil temp up to about 240F. I forgot to remember the driving pressure/RPM but noticed that at idle I was 15 PSI at 950 RPM. I wonder if you still think I should try 20wt oil. Maybe the temp. will not get that high if I used the 20. What do you think.

Soon I will post the data for Amsoils. It is very interesting.

ali

 

My guess is that if your oil doesnt get above 250dF you are OK, however, wear will be taking place at a lower rate with 30W oil than with 20W oil. But this is just straight extrapolation of W20 oil being as far under W3 as W30 is under W40. Note that if you are not near redline constantly and not with hot oil, the wear rate is limited by something not lubricated by motor oil.

However, the quoted article before I went of on this modeling of oil versus wear indicates that cars using the lighter oil regimine are not expected to live as long. (C.F Ford and Honda).

In effect, at some point you are caught between startup wear and operational wear. For me and the application I put my car through, I think sticking towards the thicker end is better. For my application, it does look like RedLine is good oil due to its special viscocity curve that thins slower with temperature. However for a car that seldom sees 250dF 5W-30 will be just fine and be a better balance between startup and operational wears.

The other thing I noticed is that for those companies marketing a 5W-30 and a 10W-30. It looks like they target a mixture right on the boundary between specifications. Those batches that are a little on the thick side get called 10W-30 and those on the thin side are called 5W-30.

Ali: what we need is more data in the 150dC range to complete the model.

 

Gentlemen, (are there no hard core women gear heads??)
I digress before I begin.
Mitch, I am awstruck by the clarity of your graphic analysis and coment.
Ali, I have followed and appreciated your substantial wisdom is matters slippery.
In looking at the fine graphs, ( I am a physical and visual kind of guy), I am struck by the aparent need to cool the oil at high temps. Ali has, many times over, (since I re-read a lot), convinced me that we need low viscosity at start up for flow and some measure of lubrication, (5-6 centistokes at operating temp) for adequate lubrication
- Can we have a larger than necessary oil heat exchanger, (for flexibility and capacity) and bring the temp to 100 degree C. AND keep it there with a thermostat??? in exactly the same manner as we bring the coolant temp up to operating spec.???

With the utmost reverence, may I genuflect in your general direction??
chris

 

Most oil coolers have thermostats so that the oil is not overcooled. I think the Maranello one starts to open at 185F. I could check on it if you need the exact temperature. The transmission gear oil cooler also has a thermostat.


Mitch, in reviewing this guys statement I find a big flaw:

"<snip>
However unlike the exact scientific value of Poise for Absolute Viscosity, the SAE viscosity numbers are "staircase" approximations for KINEMATIC Viscosity.
For example Motor Oil that is measured to have viscosity of 9.5 cSt @ 100 C will be rated as SAE 30, while another Motor Oil that is measured to have viscosity of 12 cSt ( or 26% more viscous ) will also be rated as SAE 30 Motor Oil.
Yet in real life operation 26% difference in viscosity may make difference between engine that will run forever and one that will wear out prematurely.
<snip>"

An oil that has a viscosity of 10 is not twice as viscous as an oil with a viscosity of 5. The scale is not absolute. My favorite analogy is for temperature.
Is a brick of aluminum at 80F twice as hot as one at 40F? No. Is zero Celsius the temperature of no molecular activity? No. We need to use the correct scale. For temperature that is the Kelvin scale.
40F is 278K. If you want something that is double the temperature it would have to be 556K or 541F. The true doubling of 40F is to 541F.

The true oil viscosity may be more like 110 and the less one may be 105. One is not twice the other but only about 5 percent more or less.

aehaas

 

Ali:

I was hoping to get your comments about that exact paragraph. So, both Poise and Stokes are like centegrade and Ferinheight, and what we need is some absolute metric like Kelvin or Rankin. Do you know of a function that takes Stokes and creates someting in an absolute manner?

 

The dry sump in the F355 has a thermostat, and the car comes wiht an oil temperature guage. Cruising along the il stays right at the thermostat opening point (185).

It is the cars that are taken to the track without a temperature guage that worry me. No way to know how hot the (engine) oil is or transmission oil,... Even with my Vette, the gugage is not obvious unless you select it in particular, and even agressive backroad driving I've seen upwards of 300dF. Now seeing this oil data, I will not be taking this car to those oil temps. <blank> only known how hot the tranny and diff are.

One other tidbit of info that ran under my nose: highly viscous oil is so stiff that the metal parts will distort before the oil is forced out. In effect, the oil apears to be more solid than the metal supporting the bearings. The oil is so stiff, that a journal rotating in a bearings will cleave the solidified oil in order to allow rotation. This is why you don't want to use high revs with cold oil.

 

I went from 30 to 20 wt. oil in my Maranello. The pressure (at 2,000 RPM and 185F oil temperature) dropped from 80 to 73 PSI. The oil flow had to have gone up, hopefully 10 percent or more. This can only be a benefit as I see it.

I do not know of any absolute viscosity scale. Rheology is a very difficult topic. I cannot even imagine such a scale that would work in all fluid states.

ali

 

Ali,

I read your research into oils and I wondered if I can pick your brains.

In my 328 I have used in the past 0 - 30w oils but I was told that the engine was not designed for such thin oils and there was a higher probability of an oil leak. I did infact have to replace a crank seal.

I now use 15 - 40w The car is mainly used on the road but I also attend a couple of track days a year.

My car has covered 40000 miles. I complete around 800 miles a year and change the oil and filter every year.

What do you recommend?

Martin.

 

What is your oil pressure at 6,000 RPM while driving in town, and what is your pressure at 6,000 RPM when driving on the track?

If your seals leak it may be that your car was a garage queen at one time. This will ruin any seals. The newer synthetic oils supposedly do not leak through even with older style seals.

My oil recommendations are on page 2 and based on the pressures your car requires be that 20, 30, 40 or 50 wt.

One should NEVER use a thicker oil just to prevent leaks, replace the bad seals.

aehaas

 

Amsoil:

I find this data interesting. If you study the bearing wear figures one sees that it is about the same regardless of the thickness of the oil. Look first at all the XL-7500 series.
Also looking at other 20W-50 and the straight 60 wt oil, they all are about the same. One would think the thicker oils would give better protection. In fact, the 60 wt oil (AHR) is tested at lower RPM and is therefore one of the worst performers.
From this perspective we should all be using the 5W-20 XL-7500 in all conditions as the thinner oil would give better flow (more cooling) and the same metal to metal protection (or better) as anything thicker in their entire lineup.
In addition, the 5W-20 is only around $5 a qt. while the highly rated series 2000 0W-30 is around $8 per qt.
They show data that Mobil 1 gives about 2 mm of ball bearing wear in the same test. It was the worst of several synthetics they showed.
They quoted the API J300 documents that gives the SJ and SL test criteria yet I did not find any ASTM D 4172 4 ball test in that document.

This stuff makes no sense to me. Otherwise the data reminds me of Redline motor oil products. A product line I like.


aehaas

 

The wife is out with the big white MB. Yes I do have plenty of oil in stock.

 

I did some research, and until about a year ago, many oils were in the 0.60 wear range, and most oils in the 1.0+ wear range. So, based on this test, oils are doing 3 times better now than they were a mear 2 years ago.

I am suspicious as to how this test relates to real automobile engine wear. Since all oils are giving similar wear indications (0.35 to 0.4).

However, I did spend a few minutes surveying this database (RedLine). The first chart shows all of the xW-30 weight oils in Redlines catalog and the viscocities at various temperatures.

A couple of things catch my attention. There are two groups of oils, those that try to be as thin as possible and still be 30 weight oils, and those that try to be as thick as possible and remain 30 weight oils. I will also note that the replacement for XL7500 oils are the S2000 oils and in this grade the modern oil is thicker than the earlier oil. Notice you basically cannot buy an oil in this catagory without knowing whether you want the thicker 30s or the thinner 30s. As long as your oil temp guage stays below 250dF all of these oils provide adequate lubrication. However, I do worry about the 250+ region.

 

In the 40 weight catagories, we find a tighter pattern. Almost as if the oil companies put a third grade {20->27->33->40} instead of the expected.{20->30->40}. Now, all the oils pass the 4 cSt at 285dF

 

And for completeness the 50 weight and 60 weight oils.

 

Now, this data is from a single manufacture. I found negligible differences in motorcycle and automobile oils.

Go back to the 40 weight oils and look at 5W40E. This is the oil RedLine recommends for "european" sports cars. If has rather thick 83.6 startup viscocity, but a superior hot viscocity of 4.2 and a hot viscocity of 14.5.

One could infer from this that Ferrari might want 14 cSt for the 100dC oil, whereas the modern industry is using 10 cSt as the target hot viscocity. Or one could infer that Ferrari was really targeting the performance at higher temperature and extrapolating backwards using oil technology at the time of manufacture. I don't know but I am looking into it.

At this point: I disrecommend 30 weight oils for operation above 285dF however if used below 250dF the startup performance is likely to outweigh (sic) the hot temperature performance.

 

Ali:

Take a look here:

Dynamic viscocity: http://scienceworld.wolfram.com/physics/DynamicViscosity.html
http://scienceworld.wolfram.com/physics/Viscosity.html

This justifies using a log axis on the plots, and gives a general formula for viscocity versus temperature.

 

Just curious, why wouldn't you use something like Mobil 1 0W-40 in your Maranello? Won't it provide the same amount of protection upon startup (the 0 viscosity) and better protection at higher temps (40 viscosity)?

 

Mitch, that reference is very interesting. 'A lot there to go over many different ways. Most significant is about half way down the page. I could not copy the formula but it shows that the force between layers is directly proportional to the velocity. This would go along with the thinner oil at higher velocity being a good thing. That is in addition to additional cooling from additional flow.

Look again at all brands of racing oils. There are an increasing number of thinner and thinner oils coming out. Amsoil has 3 different 0W-10 racing oils for eg.

Art, I published this oil data because those numbers on the can do not tell you what is going on inside. The start up viscosity varies considerably among the 0W-X oils. Note that start up means 40 C, 104 F. The real thickness at real start up, say 75 F is even thicker. The top number of 0W-X, is dependent on you and your car's needs. My Maranello's owners manual states the oil pressure should be between 75 to 85 PSI at 6,000 RPM. It should not be more or less. I get 80 PSI at only 2,000 RPM with around town driving therefore I still need a thinner oil if I am to comply with the requirements in the manual.

ali

One more thing Art - thicker is not better. This is the biggest misconception. Oil should be the appropriate thickness, otherwise we should all be using 50 or 60 or 90 wt oils. The "appropriate" thickness is generally defined as -that thickness needed to give your engine 10 PSI for every 1,000 RPM.

 

I found a reference for Jet engines with viscosity versus temperature. They produced a chart showing those pressure/temperature/viscisity relationship that leaves the lubricant liquid, and those where the lubricant solidifies under hydrodynamic pressure. A surprisingly small transition region. In addition to the thinness at atartup argument, it is possible to have an oil that is so thick (even warm) that it solidifies under hydrodynamic pressure and then is sheared by the requirement that movement continue. This does nothing good for the bearings, of for the oil and generates great distortions in the bearing surfaces (for an instant) and great heat (through shear forces) into the oil.

I believe this is true, however, their web site has no xW-10 weight motor oils of any kind.

I am fully on board with the lowest possible startup viscosity restrained only by the requirements for high temperature viscosity to be sufficient. Unfortunately, nobody is willing to volenteer a number that is sufficient and where that number should be on the temperature scale.

To Art's question: xW-40 oil seems on the out. Either the manufacture want xW-30 or thinner, or the "performance orrented" guy wants the thickest he can find (ergo xW-50). leaving teh xW-40 guys holding the empty bucket.

But even in the 0W-30 weight grades, there is a 50% spread from the thin ones to the thick ones in viscosity. This implies to me, that if you want to be careful with your oil (like I do due to track days) you actually have to do the research. On the other hand, if you don't take the car to the track, and don't get above 250 dF almost any oil will protect the engine from wear, so pick the lightest oil to prevent the most startup wear possible. But even here you have to do your research as some 0W-30 oils are thinner than 0W-20 oils (go figure).

 

My mistake. It is Royal Purple's Racing oils.

Go to this page:
http://www.royalpurple.com/prodsa/rpro.html

download the PDF:
http://www.royalpurple.com/pdfsa/rprops.pdf


Ali

 

I found this link: http://birfield.com/archives/html/80scool/2003-05/msg00725.html

I quote several interesting tidbits:

"4 Ball: This is a test where 3 ball bearings are arranged in a triangle
and a fourth is placed on top with a applied downward load of 40kgf
while the top ball is spun at 1800rpm for 1 hour while the balls are
flooded with oil heated to 150°C for 1 hour. This simulates boundary
lubrication areas in the engine such as the rings on the bore or the
valvetrain. The dimension quoted is the width of the wear scar at the
end of the test."

"Japanese versus European versus American engine design effects on oil.
(See Willem's webpage and caltex article for more detail)
Basically the american engine has a high top ring and a large piston
crown to bore clearance.
European engine low top ring and a small piston crown to bore clearance.
Japanese engine low top ring and a small piston crown to bore clearance
plus high engine oil operating temp (by purposeful design).

These three internal designs require 3 different types of engine oil to
function correctly, in very crude terms:
American cheap simple additive package.
European simple additive package.
Japanese complex expensive additive package."

" >From the above you can see that an oil for a Japanese diesel engine
should have the following properties: 10w30, low or no magnesium, at
least 3000ppm calcium (if you ask for this most oil companies will tell
you that this information is commercially sensitive, yet it can be
easily tested for), relatively high levels of ZDDP ie zinc & phosphorous.
We would prefer something with a low volatility and small wear scar on
the 4-ball test as well."

So, we see a story where the 10W-30 designation is not sufficient to choose an oil with acceptable additive packages (in Japanese diesel engines).

 

In here: http://www.imperialoil.com/Canada-English/Files/Products_Lubes/IOCAENINDESUnivis_n.pdf

We see the basic info we want for automobile engines except that this dataset pertains to hydraulic oils.

note: the optimal viscosity is in the 20 Cst to 50 Cst range.
Note: nobody recommends operation at lower than 10 cSt.
Note: recommendation for viscosity before going to full power.

Just food for thought about this complicated area.

 

Yes but if we review the Amsoil data again, the 0W-20 gives the best while the 60 wt oil gives the worst ball bearing test results. Some of the worst results were only at 1,200 RPM and 75 degrees instead of 1,800 RPM and 150 Degrees.

Also, racing engines and regular car manufacturers are still going to thinner oils while engines continue to last longer.

Mitch, how many miles do you have on your 355?

Also, no oils are more than 4 or 5 at race engine temperatures while the 20 wt oils at 185 F are probably at 14 or 15 cS because the temps are kept lower with the higher flow rate.

Comment on the equation that directly relates flow to separation force. It looks like a 1:1 correlation rate. Oil viscosity is one thing but the rheology of it shows that it is not the only thing. Fluid velocity is perhaps more important. Note that even pure water has a good viscosity at the 30 to 40 C temp. At a good flow rate it might lubricate quite well if there was no rust problem.

ali

 

I am completely on board with this.

Completely agree. However in the catagory of race cars, the builders have the option of putting in a big enough oil cooler to keep the temperature where they want it, In effect, we are trying to find "where they want it" and why.

I'm just over 40,000.

[QUTOE]Also, no oils are more than 4 or 5 at race engine temperatures while the 20 wt oils at 185 F are probably at 14 or 15 cS because the temps are kept lower with the higher flow rate.[/QUOTE]

Agreed, see oil cooler size postulate above.

[QUTOE]Comment on the equation that directly relates flow to separation force. It looks like a 1:1 correlation rate. Oil viscosity is one thing but the rheology of it shows that it is not the only thing. Fluid velocity is perhaps more important. Note that even pure water has a good viscosity at the 30 to 40 C temp. At a good flow rate it might lubricate quite well if there was no rust problem.

ali[/QUOTE]
Mitch

 

Mitch, you have a total of 40,000 miles on your 355, about half on the track at an oil temp. of between 250 and 285 F. Your oil pressures will allow you to run a 30 wt oil on the track despite the miles (age) of the engine. There is clearly no wear in there.

Many prospective Ferrari buyers take a car to be inspected only to find that the garage queen has 2 bad cylinders with low pressure. Running the car at the highest of engine temperatures, in the heat of Texas and at highest RPM seems to have no adverse effects on a well maintained car.

I still think you should try the 30 qt oil, not the thicker one as Redline but a thinner one as Mobil 1. I am sticking with "thinner equals better flow" and therefore higher separation force and better cooling of critical parts.

Tell me please, what is better than that?

ali