Motor Oil 103

[AEHaas]

Go to SAE.org and purchase the compendiums mentioned here:
http://theoildrop.server101.com/ubb/ultimatebb.php?ubb=get_topic;f=42;t=001265
SAE SP-1894 and SP-1550

aehaas

 

[Gary Allan]

Oh

So, except for this small taste here:

Effect of Break-In and Operating Conditions on Piston Ring and Cylinder Bore Wear in SI (Spark-Ignition) Engines, Schneider et al:
The rate of wear is much higher within 15-20 minutes of start-up than after reaching normal operating temperature. There was a lot of data but I conclude that the initial start-up time period (first 20 minutes) result is 100 nanometers of wear whereas the steady state wear rate was only 4 nanometers per hour thereafter. (Hence we should be concerned about start-up oil thickness more than running thickness. This justifies the statement that 95 percent of engine wear occurs just after start-up).

I can purchase the entire article if I so desire? I can otherwise take your word for it?

Thank you!

 

[AEHaas]

JonS,
Boats and airplane engines are generally thought to run at or near full output and not intermittently as cars between shopping, picking up the kids and going out to eat. It is considered Heavy Duty use and the discussion is different than for cars.
You can reference the others to this site but wait a while as the chapters will be better organized into a single resource soon.

Gary Allan,
If I asked myself the question, I would say that the engine would do best by having a sump temperature of 212 F and the oil viscosity should be 10 cS.

aehaas

 

[Gary Allan]

quote:

---

There was a lot of data but I conclude that the initial start-up time period (first 20 minutes) result is 100 nanometers of wear whereas the steady state wear rate was only 4 nanometers per hour thereafter. (Hence we should be concerned about start-up oil thickness more than running thickness. This justifies the statement that 95 percent of engine wear occurs just after start-up).

---

I in no way dispute the assertions by the esteemed Schneider and concede that more wear indeed occurs during "warm up" than beyond that time. This in no way indicates that "start up" wear is not considered the first 30-45 seconds of operation. Nor does the esteemed Schneider compare the nanograms of the first 0-45 seconds of operation with the first 20 minutes (unless I missed something).

Suppose the wear occuring in the first 45 seconds of operation were 10X of the wear between 45 seconds and 20 minutes?

Did the esteemed Schneider (and apparently you too) assume a linear wear profile for the entire 20 minutes of operation? Common(er) sense would lead most to assume that the first 10 minutes of the 20 minute event would have more wear than the second 10 minutes of the event. We could then assume either a decreasing or increasing wear rate depending on which way to explore to the ends of the graph. This would further lead us to believe that the HIGHEST rate of wear would occur in the very first few moments of operation.

Have you plotted this wear from zero to 20 minutes? Wouldn't you expect to see the wear taper off as you approach the end of the event? Would you not expect the wear to be MUCH higher near the beginning of the event?

This has nothing to do with your conclusion ..merely the breadth and weight (range) that you assign it. Your, and Schneider's, line is drawn pretty far out there.

 

[Gary Allan]

quote:

---

If I asked myself the question, I would say that the engine would do best by having a sump temperature of 212 F and the oil viscosity should be 10 cS.

---

But why not 9 down to 2 (which is a good thing in your infamous 302°F obsession)? Suppose the spec of a 30 weight was chosen because of the infamous "Schneider standard"? The oil would be assumed to never reach either the 10Cst nor 212. If the same engine was spec'd for that output under those conditions ..would the spec be a 30 weight?

You didn't answer your own question. You qualified the test.

 

[AEHaas]

If you read and understand the article I think you will be enlightened. I used to think that the early wear was within a short time period as well. Now I know better.

Another person on this board suggested that I read : SAE paper #980702. I ordered it. Knowledge is a great thing. While most spend there time watching TV, I read. I think that if I was independently wealthy I would become a professional college student.

My profession in medicine demands this type of dedication.

aehaas

 

[Gary Allan]

quote:

---

The rate of wear is much higher within 15-20 minutes of start-up than after reaching normal operating temperature. There was a lot of data but I conclude that the initial start-up time period (first 20 minutes) result is 100 nanometers of wear whereas the steady state wear rate was only 4 nanometers per hour thereafter.

---

Okay ..let's take a different approach. I'm not yet willing to purchase the document from SAE.org ..but..

How do we determine that this is a function of oil ..and not just a side effect independant of oil??

That is, suppose we infuse the engine with 212 degree oil and then started it ...without regard to (or at various) ambient temperatures. Would we see less wear in that 20 minute time frame. Suppose the wear is from ill fitting parts that are distorted until they are fully expanded, putting overbearing stress on some surfaces while others are not in full contact ..and the oil temp just coincidentally tracks this event (which it would since as the parts warm and expand ..the oil would be subject to this warming as well)?

Since the assertion of the first 20 minutes of operation evidencing higher wear is independant of viscosity (that is, there was no distinction of any oil reducing this event duration or amplitude) ..then we must, or rather can, conclude that the oil itself has no impact upon it.

Any thoughts?

 

[AEHaas]

Gary,
..............................................................................--.................
......................................................
.......................
............
........
......
....
..
..
.
.

Connect the right side of the dots.
Call the total time an hour. The dash is when the engine was turned off for 10 minutes. Note the curve picks up where it left off. The steep part of the curve is the first 20 min and the right side does, very slowly go up. (See if this comes out OK). They did the test this way and many times over. It was very impressive.

Going to the right is time. Vertical is total wear and the slope of the curve is wear rate. The article indicated a wear rate of 100 for the first 20 minutes and 4 per hour thereafter for many hours. They also increased the load and the curve tilted up. Instead of 4 per hour it would be a steeper curve (less level looking) and a higher wear rate.

aehaas

[ January 27, 2005, 04:07 PM: Message edited by: AEHaas ]

 

[XS650]

quote:

---

Originally posted by Gary Allan:

That is, suppose we infuse the engine with 212 degree oil and then started it ...without regard to (or at various) ambient temperatures. Would we see less wear in that 20 minute time frame. Suppose the wear is from ill fitting parts that are distorted until they are fully expanded, putting overbearing stress on some surfaces while others are not in full contact ..and the oil temp just coincidentally tracks this event (which it would since as the parts warm and expand ..the oil would be subject to this warming as well)?

---

Starting with fully warmed up oil might make the problem worse. At startup many clearances are bigger and ideally would take a more viscous lube than when at operating temperature.

There is also the seldom mentioned issue that all else being equal, oil that is less viscous when it's cool drains more completely off the parts leaving them less protected at startup than those lubed with oil that doesn't run off so thoroughly.

Does it make a difference you could measure? Beats me, but it shows why obsessing about 10cSt oil under all conditions is not only pointless, it's wrong.

 

[Gary Allan]

I have no dispute with the test results (nice use of the keyboard btw). My only question is "is it the result of poor lubrication due to too high a viscosity"? That is, if the first 5 minutes of the test were performed with a fluid that was between 9.3 and 12.49 Cst @ 75°F (for the sake of the exercize - ambient temp) ..I would venture to say that the wear results would be within enough of a range to be essentially identical. That is, regardless of the wear rate ..the warming of the components lessened it ..not the viscosity of the oil.

Therefore, although I do not dispute the assertions that more wear occurs in the first 20 minutes of operation AND that this coincides with the oil temp and its reaching its 212°F (+/-) visocity, I don't think that the wear is caused by the viscosity of the oil nor can it be prevented by it.

Perhaps this is my misinterpretation of the lesson. It appeared to (and this is perhaps just me) use this data to show that the oil can not lubricate adaquately due to its thickness at that temp. My assertion is that no viscosity would adaquately lubricate the engine in that transitional state to emulate the steady state wear. That is, the 100 nano grams may be reduced to 90 or 80 ..or perhaps 40 ..but regardless of what lubricant was used (indexed for 10 Cst for each incremental upramp in system temp)it would result in this elevated wear rate.

Now I somewhat figure that Schneider wasn't exploring the full boat of "what ifs". That is, he probably didn't prewarm the coolant jacket and then run the same test through the oil warm up period ..nor prewarmed the oil ..or both thorought the sequence of 5-10-15-20 minute tests. At least I would expect not. So, although the wear event tracks the oil temperature and the reaching of its 10 Cst visocity ..it may be totally indenpendant of it.

If this wear was a byproduct of viscosity ..then a subsequent test sequence with a 20 weight oil would have shown lower wear numbers(or the curve would lengthen earlier) IF viscosity had a role in reducing it.

 

[AEHaas]

...Agree. It would certainly be interesting to do the same testing with 5, 10 and 20 wt. oils. It seems that many questions would then be answered. Or, would it just bring out a new set of questions? The more we know the more we ask.

aehaas