Not how I would have stated it but the only 0w20 is M1 which is known to have higher Fe #s where as everbody makes a 5w20. I don't think it's the viscosity that's the issue there. i think when we're comparing viscosities, it should be relative to a single product in a specific engine. The the upper cylinder film thickness of a 5w20 may be thicker in a tight engine by not getting whiped of by the rings but in a BMW it may just increase consumption and not support the bearings as well. In a car that calls for it I don't see a problem with 5w20 under most and my conditions. When I was in ST Thomas, I asked the cab driver (Ford Van) about his oil use. 5w20 was what the factory recommends. It's 90* and you go verticle to get anywhere, always with a load. I suspect another viscosity would have served him better.
5w-20 and engine longevity
- Thread starter z917990
- Start date
- Status
I agree that a sample of 11 UOAs for each viscosity is too small and consequently inconclusive. I guess I was just trying to illustrate that a single wear indicator (level of iron) can vary depending upon the sample. If I had the time to gather more meaningful stats, I would get between 200 and 300 UOAs for each viscosity and then subdivide the stats based on engine type, dino vs. synthetic etc. But, I have a job and I spend way too much time thinking about my cars anyhow (at least according to my wife
). LOL.
I think that your car will get to that magical 200,000 platiue with 5w-20 if you dont abuse the h@!! out of it. So I would not worry.
Sorry z917990, No intent to degrade your efforts. Just pointung out that the 0w20 brand is an oil that shows higher iron regardless of viscosity.
Correctomundo. You and TomJones76 illustrate how a poor understanding of statistics can lead to false conclusions. Even a large sample size doesn't guarantee meaningful results if the other variables aren't constrained. Using darkdan's method, I can "prove" that smoking cigars or a pipe will extend your life.
I thought someone figured out that there's a compound in the 0w20 that doesn't show up in VOA as iron, but breaks down with useage and then does. Wasn't there a big thread on it?
Using my method I PROVED that on average people EXPERIENCED better results with 5w30.
Use my method and prove modular motors wear better with 5w20.
Only the 5w20 fans try to say my "study" isn't scientific. It's not. It's not a study. It's just a compliation of REAL RESULTS.
I'm not saying my spreadsheet is the holy grail of 5w20 vs 5w30. But I haven't seen anyone else come up with real world results proving otherwise. So I enjoy my warm and fuzzy feeling that I get from using 5w30.
Using my method I PROVED that on average people EXPERIENCED better results with 5w30.
Use my method and prove modular motors wear better with 5w20.
Only the 5w20 fans try to say my "study" isn't scientific. It's not. It's not a study. It's just a compliation of REAL RESULTS.
I'm not saying my spreadsheet is the holy grail of 5w20 vs 5w30. But I haven't seen anyone else come up with real world results proving otherwise. So I enjoy my warm and fuzzy feeling that I get from using 5w30.
Looks like the 10w30 shows a wider margin of improvement over the 5w30 than the 5w30 does over the 5w20.
Yes, in the same manner that if you start smoking cigars or a pipe you'll live longer.
"Yes, in the same manner that if you start smoking cigars or a pipe you'll live longer."
I may not live longer, but I have statistical proof that I will be far more relaxed. I'm also planning on smoking my spread sheet to see what effect this will have on my viscosity.
I may not live longer, but I have statistical proof that I will be far more relaxed. I'm also planning on smoking my spread sheet to see what effect this will have on my viscosity.
Whether or not the rich people smoke the cigars, they still lived longer.
So until you show that my 5w20 wear trend gets reversed when we bring in a second factor of the 5w30 users being poor......
So until you show that my 5w20 wear trend gets reversed when we bring in a second factor of the 5w30 users being poor......
I guess a cigars viscosity isn't the determining factor.
I see you still don't understand. May I suggest a good book on statistics?
And your constant "cigars and pipes, cigars and pipes!" has no bearing on the spreadsheet until you can show the "wealth factor" that effects the spreadsheets results.
My humble observation is as follows.....the thinner I run the faster I go...The thinner I run the more engine wear I experianced. This statement was true a few years back...now days with the products/packs available I can run 0-30 in a firebreathing V8 with no worries and excellent wear results...and still get the performance..my racing buddies all run 20-50 etc and tell me I am screwing up...they quit doing that after awhile
..I do run dual filters and a oil cooler...oil temps never see above 200-210....My Dad's Lexus calls for 0-20 and yes I had a knee jerk reaction to up the viscosity....but he demanded and received his Amsoil 0-20....thing purrs like a kitten for the past 40-50K miles and gets excellent MPG....engine is clean as a whistle and very tight.....bottom line is I now am convinced these modern engines call for 0-20 and should get a top quality 0-20 with reasonable OCI's..........
..I do run dual filters and a oil cooler...oil temps never see above 200-210....My Dad's Lexus calls for 0-20 and yes I had a knee jerk reaction to up the viscosity....but he demanded and received his Amsoil 0-20....thing purrs like a kitten for the past 40-50K miles and gets excellent MPG....engine is clean as a whistle and very tight.....bottom line is I now am convinced these modern engines call for 0-20 and should get a top quality 0-20 with reasonable OCI's..........
OK, guys, go to your neutral corners and put cold towels on your heads. At its core, this is an argument about whether stats "prove" anything or not, and here is where natural scientists part from social scientists. In the social science biz (I'm a history prof and I use stats in my research), stats tell us general trends and lead in directions of human behavior. This is why I argue that a larger sample might be more illuminary, but not absolutely conclusive i.e. "proof." As several posters have mentioned, factors other than viscosity can alter wear conditions in a UOA and this includes human behavior. So, if hypothetically, I were to do a sample and measure wear indicators from all of the different viscosities, and it turns out that engines with 5w-20 have higher wear indicators, I could conclude several things that are not necessarily compatible: 1. 5w-20 tends to cause engines to wear faster; 2. People who use 5w-20 may be harder on their cars than those who do not; 3. Engines that use 5w-20 naturally wear more than those that use another viscosity, not because of the oil, but because of the design of the engine (and the list goes on). Therefore, if I were writing an article and stipulated that the first conclusion was correct, I would have to show that numbers 2 and 3 are probably not true. I would then conclude that number 1 was more likely, not absolutely true i.e. "proof." Anyway, this is how some of us think in the history biz.
I believe I listed a bunch of them for you once before, but you continue to just ignore them.
Very well put, z917990. The best that can be said here is the "data indicates" (given the large number of unconstrained variables and limitations in sampling), not that anything was "proven".
The data collection method makes such an analysis extremely difficult, if not impossible. If the values for other variables aren't collected, you can't see if there are any correlations.quote:
Originally posted by Drew99GT:
How can you deny that darkdans results show that 5w-20 produces more wear? If your such a stats wiz, analyze the numbers for us.
The data for 5w-20 would be limited mostly to Ford and Honda engines. A significant potion of that sample are Honda 3.0, 3.2, and 3.5 V6 engines. These throw out wear metals regardless of the oil used. If these engines were excluded, I am sure the results would look quite different since they are a larger percentage of the 5W-20 population than the 5w30 or 10w30. For a test such as this to be valid, a consistent engine sample would be required and that is practically impossible.
- Status
