water temp

[gpshumway]

Thanks for posting this info driver01. It's nice to see that the ECU monitors oil temperatures, it makes me think there's an engine-saving limp mode in there someplace in the event oil hits 225*C or more. Similar to the Nissan 370Z.

As to which oil to use for track vs winter, rice_classic got much of it right, but not all. Centistokes is a measure of viscosity, but does not translate directly to flow, especially in an oil system with a positive displacement oil pump. Thicker oil (higher cSt) will require more power to pump and will result in higher oil pressures, but it won't necessarily take longer to reach engine components.

"W" number requirements apply at very cold temperatures and are highly dependent on pour point modification additives which have little to no effect on viscosity above 0*C. It is therefore possible to have a 0w oil which is thicker at moderate temps (20*C) than an otherwise similar 5w oil. Oil viscosity is measured and reported at 100*C and 40*C. The 100*C number determines the grade (20 weight, 30 weight, etc) while the 40*C measurement determines the viscosity index, which can be used to interpolate the oil's viscosity at temperatures between 40*C and 0*C. Several online calculators can do this for you. I like the one from Widman's which is HERE.

The "W" number comprises two requirements, a Cold Cranking Simulator (CCS) viscosity and a Mini-Rotary Viscometer (MRV) viscosity. The CCS tells you how hard your starter has to work to move the oil, the MRV tells you if you're in danger of sucking air from the sump instead of oil (pump cavitation). The MRV is measured 5*C below the CCS to ensure a safety margin against cavitation. For a 5w oil CCS is measured at -30*C and MRV at -35*C for a 0w oil CCS is at -35*C and MRV at -40*C.

You can see from this Widman graph that Pennzoil Platinum 5w30 is thinner at all temperatures above 0*C than M1 0w30. Below that temperature the difference between base stocks and the pour point modifiers take over to the point that M1 is thinner at -40*C than PP is at -35*C. At -35*C M1 is likely half the viscosity of PP. You can see this in the MRV specs on their respective data sheets.



There can absolutely be good reason to use 5w oil instead of 0w oil in street cars, especially those which are particuarly hard on oil. The most obvious current examples are direct-injected engines which tend to introduce large amounts of raw fuel into the oil (fuel dilution). The viscosity index improvers and pour point modifiers which differentiate 5w from 0w oil are less thermally and mechanically (shear) stable than the base oil they're added to. Therefore trading some low temperature viscosity performance for greater shear stability and lower volatility can be a good decision in many engines and operating conditions. For example, Pennzoil Platinum 0w20 has superior cold temperature viscosity characteristics to Pennzoil Ultra 5w20, but Ultra has dramatically lower volatility. Platinum looses 14% of it's weight in the NOACK test, while Ultra looses only 5%.

Is the BRZ engine one which will benefit from a low volatility oil? We don't know yet, but many direct injected engines prefer low volatility oil. Someone could look up a bunch of UOAs on the Lexus IS350 for an idea.

[ahausheer]

Originally Posted by gpshumway

.

At proper operating temp it appears that all the oils are the same viscosity and the asymptotic trend suggests that track (oil) temps will not affect viscosity enough to justify a 30 or 40 weight oil. Also I imagine a peak track oil temp of 118 suggests the oil temp on average is pretty darn close to 100c, where it was designed to be reinforcing the idea of using the recommended oil weights. What are your thoughts on people that use a 5-30 oil for track days?

[gpshumway]

Originally Posted by ahausheer

At proper operating temp it appears that all the oils are the same viscosity and the asymptotic trend suggests that track (oil) temps will not affect viscosity enough to justify a 30 or 40 weight oil. Also I imagine a peak track oil temp of 118 suggests the oil temp on average is pretty darn close to 100c, where it was designed to be reinforcing the idea of using the recommended oil weights. What are your thoughts on people that use a 5-30 oil for track days?

The oils are not the same viscosity at operating temp. It's not asymptotic behavior, it's just a result of plotting a logarithmic trend on Cartesian linear scales. If you were to put 50*C as the minimum temperature into the calculator you would clearly see the difference at 100*C. Also, engines are very sensitive to small differences in viscosity at those temperatures, particularly the hydrodynamic bearings.

Does an oil temp of 118*C indicate use of a thicker oil on track? Why yes, it does. If 8.3 cSt is the optimal viscosity (Subaru 0w20 at normal 100*C operating temp), then you'd need a thick 30wt oil like Valvoline Maxlife to have 8.3 cSt at 118*C. In reality I'm sure there's some safety margin for higher temps with the 0w20, but Subaru themselves recommend thicker oil for high speeds and high temperatures, it's right in the manual.

As to "average" oil temp on track, the oil temperature will not change very fast, certainly not corner to corner. When racer01 says maximum, I'm sure thats' a slow rise over several laps. Sure there are some flat spots & bumps for esses and straights, but it doesn't look like a mountain range as an intake air temp graph in a turbo car does.

[ahausheer]

I see, I was looking at the chart and not the numbers. Interesting, I always thought all 30 weight oils would behave the same at the same temp hence the same rating. That calc seems to use only two data points for a non linear trend. How can it do that and be accurate?

[gpshumway]

Being that all motor oils are hydrocarbon liquids of one form or another, the non-linearity of their viscosity is predictable given only two data points. The graph is only accurate within a limited range, though and only for kinematic viscosity. It won't accurately predict high-shear viscosity or viscosity near the pour point. For instance ester based oils like Redline have much greater high-shear viscosity relative to their kinematic viscosity than Group III or PAO (Group IV) based oils. Redline 5w20 has a High Temperature High Shear Viscosity (HTHS) of 3.3 cSt, which is higher than many 5w30 oils. HTHS simulates a worst case engine bearing scenario, 300*F at a high rate of mechanical shear.

HTHS is actually much more important for oil selection than 100*C viscosity, but it's not neatly displayed on the bottle in the form of a XwXX grade. Each grade of oil is actually a viscosity range at 100*C and some minimum requirements for HTHS viscosity. 30wt oil must be between 9.3 and 12.5 cSt at 100*C, quite a range, thick 30wt oils like Maxlife and Syntec 0w30 are up to 30% thicker than thin 30wt oils like Edge with Titanium 5w30.

Over the years the original spec, SAE J300, has become a bit outdated and has been added to and modified by other specifications along the way. Today it's a bit of a mess.

[gpshumway]

Originally Posted by driver01

Techstream lite. Its spedified for dealers and such

https://techinfo.toyota.com/techInfo...ite&_nfpb=true

I also tried a generic reader (plx wifi and dashcommand) very little info out of the ecu (compared to all the data from techstream)

This is really interesting to know. I wonder if the ECU in a BRZ is more similar to a Lexus IS350 or a Subaru WRX. The open source products RomRaider (software) and Tactrix OpenPort (hardware) are very useful and powerful for logging a WRX. Hopefully they'll be easily adaptable to a BRZ and let us log on the cheap the way we can a WRX.

[rice_classic]

Originally Posted by gpshumway

There can absolutely be good reason to use 5w oil instead of 0w oil in street cars, especially those which are particuarly hard on oil. The most obvious current examples are direct-injected engines which tend to introduce large amounts of raw fuel into the oil (fuel dilution). The viscosity index improvers and pour point modifiers which differentiate 5w from 0w oil are less thermally and mechanically (shear) stable than the base oil they're added to. Therefore trading some low temperature viscosity performance for greater shear stability and lower volatility can be a good decision in many engines and operating conditions. For example, Pennzoil Platinum 0w20 has superior cold temperature viscosity characteristics to Pennzoil Ultra 5w20, but Ultra has dramatically lower volatility. Platinum looses 14% of it's weight in the NOACK test, while Ultra looses only 5%.

There's always a "well it could be" when you're thinking of all possibilities with an engineering perspective. Sure. Just like a philosophy major can answer every question with "it depends". Good data. Thank you for that especially that graph that was helpful.

Going back to the 5w vs 0w. What you state of above proves a couple things, 1: M1 isn't the greatest oil and 2: 0w will generally out perform a 5w in cold start wear protection, except in some circumstances.

If the 5w-xx is going to perform better in the NOACK test then I would still prefer to change the 0w-xx more often instead of sacrificing startup wear for the sake of 9% variance on the NOACK test over the life of my oil change. Please critique the validity of this approach for the benefit of others (or invalidity...)

Most folks tracking the car are typically smart enough to let the engine idle and "come up to temp" before going and beating on it so using a 5w-30 or 10-40 for the track is of little significance on startup wear where as in a street car with startup and hit the road cold day after day, multiple times a day maybe. Also, track junkies also tend to change their oil more frequently.

I've approached choosing oil like this: (please poke holes in this as well if necessary)

1: Know your oil pressure and temperature in order to properly choose oil grade.
2: Choose an oil that is the best compromise for how you intend to use it, in what environments and how often it will be changed.
3: Review the cSt and HTHS data, especially if oil is being used in a high heat/high load scenario for long periods of time.

Ok, it's late, I'm tired so please excuse spelling errors.

[gpshumway]

Originally Posted by rice_classic

There's always a "well it could be" when you're thinking of all possibilities with an engineering perspective. Sure. Just like a philosophy major can answer every question with "it depends". Good data. Thank you for that especially that graph that was helpful.

Going back to the 5w vs 0w. What you state of above proves a couple things, 1: M1 isn't the greatest oil and 2: 0w will generally out perform a 5w in cold start wear protection, except in some circumstances.

I think you're making a couple of assumptions people commonly make when thinking about engine wear and oil. The first is that a thicker oil takes longer to get to the far reaches of the engine (valvetrain), the second is that more oil flow translates directly to less wear. Both of these assumptions are true, but only under specific circumstances. So basically I would state your generalization in reverse, 5w and 0w oils generally perform the same, except in some circumstances.

It makes intuitive sense that thicker oil would take longer to flow through the engine, but it's not the way the physics of most modern engines work. The oil pump is positive displacement, meaning it moves the same volume of fluid per revolution regardless of viscosity, provided the circuit stays full. It's only when the pump starts to cavitate that the thinner oil will be delivered to the far reaches of the engine faster. Cavitation is a sudden phenomenon and as long as it doesn't happen the thick and thin oils will both be delivered to the far reaches of the engine at the same speed. The SAE J300 requirements are arranged so that oils should be safe from cavitation down to their CCS test temperature, which is -30*C for 5w and -35*C for a 0w. Additionally, modern synthetics beat the MRV (cavitation) requirements by a wide margin, such that most 5w synthetics would pass the 0w MRV test. In other words, from a "time to get to the cams" perspective 5w and 0w oils should perform the same at any temperature above -30*C (-22*F). The OP lives in Canada, so he actually does have to worry about the difference between 5w and 0w, but even here in frigid Minnesota I almost never see temperatures where 0w oil would be necessary to prevent pump cavitation.

As to preventing cold start wear, the primary reasons engines wear quickly at startup is the anti-wear additives in oil must be hot in order to work and the engine components haven't expanded to their optimal sizes. Flow and viscosity of oil has very little to do with it. It pays to remember that oil is both a lubricant and a coolant. The lubricant function is served by the mere presence of the oil, more is not proportionally better. The coolant function is what requires high flow rates, and we obviously don't care about oil being a good coolant when the engine is cold.

It's easy to look at the graph and say PP is better than M1, but remember, that graph has limited applicability and shows virgin oils. M1 has dramatically better cold start properties at -35*C, and some oils retain their properties during use, while others don't. The other thing you won't get from viscosity information is the effectiveness of the additive package, particularly as it relates to preventing wear, this problem is compounded because UOA doesn't really show wear well either. The best you can do is examine the performance specs met by each oil, and here again M1 looks better than initially presented. If you look at the Product Data Sheets (PDS) you'll see PP meets ACEA A5-02, while M1 meets the much tougher ACEA A5/B5-10.

If the 5w-xx is going to perform better in the NOACK test then I would still prefer to change the 0w-xx more often instead of sacrificing startup wear for the sake of 9% variance on the NOACK test over the life of my oil change. Please critique the validity of this approach for the benefit of others (or invalidity...)

Well, NOACK is a rate of evaporation loss, so no matter how frequently you change the oil, more of the higher NOACK oil will be evaporating. This is especially important in a track car where oil temps are high and that evaporating oil will end up in the intake through the PCV. Oil in the intake charge substantially reduces it's effective octane. On the other hand, evaporation is dependent on the specific engine, oil temp, and oil condition. I don't think we know about the FA20 yet, but other direct injected engines cause enough fuel dilution to substantially increase the volatility of oil during use, making low volatility oil a good idea. In other DI engines the problem is compounded by the evaporated oil depositing on the intake valves, hopefully the FA20's port injectors prevent this problem.

Most folks tracking the car are typically smart enough to let the engine idle and "come up to temp" before going and beating on it so using a 5w-30 or 10-40 for the track is of little significance on startup wear where as in a street car with startup and hit the road cold day after day, multiple times a day maybe. Also, track junkies also tend to change their oil more frequently.

I've approached choosing oil like this: (please poke holes in this as well if necessary)

1: Know your oil pressure and temperature in order to properly choose oil grade.
2: Choose an oil that is the best compromise for how you intend to use it, in what environments and how often it will be changed.
3: Review the cSt and HTHS data, especially if oil is being used in a high heat/high load scenario for long periods of time.

Ok, it's late, I'm tired so please excuse spelling errors.

I generally agree with you regarding track use and thick oil, but I'll caveat that by saying you should probably be careful about using really thick oil. My statements above relate to differences in "w" number, but differences in grade (operating viscosity) can be more important on track. In particular if you "peg" your oil presure on track at high RPM, your oil is probably too thick, causing the pressure relief valve to open, diverting flow back to the pan which could be helping cool hot engine parts. Given peak oil temps of 118*C I wouldn't use anything thicker than a "thick" 30wt oil in the FA20, and I'd probably stick with "thin" 30wt oils of the Resource Conserving variety, even for track use. If racer01 or others come back with more plots showing higher temperatures on hotter days or other tracks, I may change my mind.

Your approach to choosing an oil seems entirely reasonable to me, and is exactly how I do it when the relevent data is available. I put more emphasis on oil temperature than pressure, but the two together give great information. Unfortunately most modern cars have a pressure idiot light and no temperature measurement at all. It's great to see the FA20's ECU monitors oil temp, racer01's data is really useful and hopefully we'll see much more. It also indicates that Subaru did a good job on the FA20's thermal design with on-track oil temps stabilizing at a very reasonable 118*C. For reference, Corvettes sometimes see 150*C on-track, and 370Zs have been known to go into limp-mode on track, which is triggered at ~125*C.

It's late for me too, sorry to be so verbose, I'm sure I have my share of grammatical errors as well.

Take care-

[rice_classic]

Thanks for the NOACK info regarding DI engines, that helps direct my choice of oil.

From one source to another I'm getting some conflicting information, if you'd be so kind to help clarify:

Originally Posted by gpshumway

It makes intuitive sense that thicker oil would take longer to flow through the engine, but it's not the way the physics of most modern engines work. The oil pump is positive displacement, meaning it moves the same volume of fluid per revolution regardless of viscosity, provided the circuit stays full. It's only when the pump starts to cavitate that the thinner oil will be delivered to the far reaches of the engine faster. Cavitation is a sudden phenomenon and as long as it doesn't happen the thick and thin oils will both be delivered to the far reaches of the engine at the same speed.

As to preventing cold start wear, the primary reasons engines wear quickly at startup is the anti-wear additives in oil must be hot in order to work and the engine components haven't expanded to their optimal sizes. Flow and viscosity of oil has very little to do with it. It pays to remember that oil is both a lubricant and a coolant. The lubricant function is served by the mere presence of the oil, more is not proportionally better. The coolant function is what requires high flow rates, and we obviously don't care about oil being a good coolant when the engine is cold.

For cold-start wear you state that it's the additives that aren't hot being the reason for cold start wear as opposed to viscosity/flow and this directly conflicts with "Bob" (in quote blow).

You say the positive displacement oil pump provides oil at the same flow regardless, assuming cavitation isn't occurring and when providing equal flow at a range of of temperatures the variant would be the pressure. Example would be my race car; on a cold start the idle pressure is 95psi and when the engine comes off the race track the idle pressure is 15psi. My concern is that bypass valve while cold on a street car as mentioned on Bobistheoilguy.com

On bobistheoilguy.com it is stated that:

Originally Posted by bobistheoilguy

Remember that most engine wear occurs at startup when the oil is too thick to lubricate properly. It cannot flow and therefore cannot lubricate. Most of the thick oil at startup actually goes through the bypass valve back to the engine oil sump and not into your engine oil ways. This is especially true when you really step on that gas pedal. You really need more lubrication and you actually get less.

The part where he says: "Cannot flow and cannot lubricate, is that not a attribute of the oil and pump but due the fact it goes through the bypass valve first and that's why "In cannot flow and cannot lubricate" is a valid statement?

Also on that site it is mentioned several times that as an oil gets thinner the flow increases, yet you claim with positive displacement oil pumps this is not the case? Do Ferrari 355's not have positive displacement oil pumps? (like mentioned on that website).

Originally Posted by gpshumway

In particular if you "peg" your oil presure on track at high RPM, your oil is probably too thick, causing the pressure relief valve to open, diverting flow back to the pan which could be helping cool hot engine parts.

Here you mention the relief valve, wouldn't this come into play at cold start like mentioned by Bob?

Once again, thanks for enduring and humoring me.

[gpshumway]

This is where the system gets very complicated, and generalizations break down. Bob is correct that thicker oil won't flow as quickly through the small orifices in the engine, but the effect of this on wear is highly dependent on the design of a particular interface and RPM. It is my belief that most modern engines deliver sufficient flow to all parts of the engine even at reasonably high RPM such that the difference between similar oils (say synthetic 5w20 vs 0w20) is minimal at moderate temperatures. some older engine designs require the valvetrain to run "dry" for a period until the BPV closes. I believe they place a higher premium on viscosity than newer designs.

Bob and I are actually talking about slightly different phenomenon, the difference being the behavior of the PRV. I chose to simplify my previous post for the sake of brevity, but you've very astutely recognized that simplification. To illustrate how both Bob and I can be correct at the same time let's game out a typical cold start scenario:

You park the car in the garage and over night the engine cools to ambient temperature. The anti-drainback valve in the oil filter leaks a little bit, causing a small void volume to form between the oil gallery and the valvetrain. Also, the pump itself leaks some more, resulting in a small void between the pump and oil filter.

Now you start the engine. 100% of the pump's flow goes to the engine until the void volume is filled. Unless your oil is extraordinarily thick, the resistance of the large oil passages is not enough to raise pressure high enough for the PRV to open, but once the void is full, the pressure spikes to the PRV set point. Therefore, within reasonable limits, thick and thin oils take the same amount of time to reach the distant parts of the engine. The flow rate through the small orifices of the engine once the passages are full is another matter.

The oil pressure is now fixed at the PRV pressure until the oil warms up. A thicker oil will flow less through the engine's small orifices, requiring the PRV to open wider and divert more flow to the pan in order to control the pressure. Are the functional interfaces on the other side of those orifices designed such that additional flow of thinner oil will provide superior lubrication compared to a lower flow rate of thicker oil? Is there a "break point" where the wear increases non-linearly? Only OEMs have the resources to definitively answer those questions, but we can do a little informed speculating.

Contrary to popular belief, hydrodynamic bearings don't care very much about pressure, they only need enough flow such that the bearing stays full of oil, the rotational velocity of the shaft does the rest. When it comes to separating surfaces, the thicker the oil, the better. If the main and rod bearings were not receiving sufficient flow we'd expect to see spikes in Pb in the UOAs of engines run on thick oil, I've never seen such issues. That leaves the cam bearings as hydrodynamic interfaces. They're generally plated with Ni, and I haven't seen lots of Ni wear in engines run on thick oil either.

So the extra wear from thick oil must be coming from non-hydrodynamic interfaces, right? This is where the complication comes in. Some of these interfaces will be totally dependent on the oil's anti-wear package, but for the thin oil to be superior, the interface must somehow benefit from a higher flow rate. Even thick oil should be trickling through the orifices and providing a continuous supply of anti-wear additives. So that leaves boundary-lubricated interfaces, where we would expect a thick oil to provide greater surface separation. It's just too hard to generalize.

Again, as long as we're substituting oils within a reasonable range of viscosity, I think the fear of cold start wear is largely overblown. Yes, most wear does occur at startup, but a 0w20 with a 200 VI will not dramatically reduce that wear relative to a 155 VI 5w20 or even a 165 VI 5w30 at reasonable temperatures. 140 VI 10w40? Now you're asking for trouble. Starting the car at -20*F? Go for the 0w20!

I'll see if I can dig up some BITOG posts and other info to support this succinctly.

[gpshumway]

Check out post #4905 by MolaKule, a professional oil formulator:
http://www.bobistheoilguy.com/forums...at&Number=4905

How fast does oil get to the valvetrain? With SAE 30 at 60*F
http://www.bobistheoilguy.com/forums...ue#Post1435870

[NeedsmoreCowbell]

guess im draining the w40 :") and going to try some motul eco w20

[rice_classic]

gpshumway: thank you for your time.

I guess the philosophical response of "it depends" is adequate in regards to summing up this topic.

This has been the most worthwhile thread since I joined this forum.

[dsgerbc]

Originally Posted by neutron256

Ah that would explain it. I'm using Torque with OBDLink MX which pulls back a lot but oil temp is one that I'm missing and would very much like. I wish they would make PID info more available so we could manually add them in.

For what it's worth, techstream is available on teh torrentz. It even connects to my BRZ (via Tactrix 2.0). I can even get to menus with customizable features like headlight delay and such. Haven't poked around too much though, too hot in my garage.