I think I’ll take a lubrication engineer’s advice who worked for NASCAR who repeatedly shows the data behind his claims over advice from a faceless dude on the internet, but thanks.
Determining the Proper Oil for Your Engine - Email from Lake Speed Jr
- Thread starter rstcso
- Start date
dnewton3
Staff member
Did you bother to read the link to the normalcy article??? That would clearly show why his methodology is flawed. And speak to your example.
Last edited:
dnewton3
Staff member
So, I do have a face; I'm not AI.
And I wrote this article; published in Machinery Lubrication magazine from Noria
https://www.machinerylubrication.com/Read/30383/engines-oil-analysis
That article is a condensed version of the one here on BITOG (already linked in my previous post #14).
And my data is in that article for all to see; apparently you didn't bother to read it, (or did, but didn't understand it?)
I was a Statistical Process Quality Control Senior Engineer, and I did everything from manage a quality lab (including all equipment certifications/validations) to writing FMEAs, DOEs, etc. By the time I left, I was the SME for TXV selection and application; I rewrote decades old testing methodology and brought it into the modern era.
I have always been up-front that I'm not a chemist nor a tribologist. What I did for many years was process quality data with a focus on statistical analysis. I'm fairly sure that I am qualified to know how/why LSJr's methodology is flawed. Most anyone can collect data. Few know how to properly collect it; fewer still how to properly analyze it. LSJr is clearly out of his element when it comes to discerning and employing the proper methods.
And for the record, I was offered an opportunity to work with a lubrication supplier to a prominent NASCAR team to help with their oil analysis, but I turned it down as I was looking forward to retirement after having multiple concurrent careers. Just because I didn't take the job, doesn't mean I wasn't qualified to do the job.
Last edited:
formulating a engine oil is like one size fits all, well not in many cases as the different specifics & types of engines, the condition of the engine, the climate it is operated in, and driving habits and on and on/ and of course the proper licenses,approvals,viscositys, cafe requirements etc. to pick a oil from the never ending list of options , you may have to learn what is best and available published information (U-TUBE) and other sources etc. ( and use your judgement ) for your ride,,,,,, by using what is available to the consumer , also in the terms of uoa and the manufacture of the oil less than transparent pds, ,and finding the best testing company that you feel is good,,,,,, seems to me and probably others, Lake is making a good effort to attempt to help people learn about engine oil and its applications in general, plus he is not out to criticize or intimidate . or put readers down,
dnewton3
Staff member
IMO, LSJr has become better at producing his videos. Maybe (whether directly or indirectly) he's got feedback about his noobian attempts in the beginning. I, for one of many, have criticized not his content per se, but the presentation (lack of polish; run-on sentences; repeating himself; tangents; poor organization). I think he's getting better at the craft of YT video production. I used to cringe watching them; I find them tolerable now.
But I'm not going to let him off the hook in regards to poor analytical methods and haphazard conclusions. Maybe he actually knows how to do proper analysis, but if so, he's perverting the presentation by putting out such irresponsible stuff. Or, more likely, he simply assumes his knowledge of chemistry and tribology port over to data processing, but it doesn't in reality.
But I'm not going to let him off the hook in regards to poor analytical methods and haphazard conclusions. Maybe he actually knows how to do proper analysis, but if so, he's perverting the presentation by putting out such irresponsible stuff. Or, more likely, he simply assumes his knowledge of chemistry and tribology port over to data processing, but it doesn't in reality.
His recommendations seem reasonable to me and are on-par with a lot of what is posted/discussed here on BITOG w/r to this topic. It's "average person" advice and easy for folks to understand that are seeking oil advice and aren't interested in a high level of detail or research. It's actionable advice without a bunch of time to generate massive amounts of more statistically significant data. I've adopted his 5ppm/1K mile limit for my UOA data on the Sportwagen...it's the red line on my graphs.
Last edited:
What are thoughts on his advice to change oil at 500-1,000 miles initially & then again in 3-4k miles?
dnewton3 - are you disagreeing with his thinking here as well? I don't think you are saying either way on that but wanted to ask.
dnewton3 - are you disagreeing with his thinking here as well? I don't think you are saying either way on that but wanted to ask.
dnewton3
Staff member
I am a fan of a couple early, short-duration OCIs in new equipment; yes. Mainly to rid any residual machining remnants, break-in, etc out of the sump.
I personally haven't done this on any new car I've owned and not seen any adverse effects over long ownership runs. However, I understand why folks recommend and do this.
dnewton3
Staff member
I counter with this, not to be argumentative with you personally, but to bring another perspective ...
I do understand why some of his videos are perhaps geared (dumbed down) to an "average person" level. He's trying to broaden his YT base by lowering the bar, so to speak. The more hits, the more his brand grows, which drives monetization not only from YT but also his company sales, etc.
But ... for $60, I'd expect "expert" advice. Not white-washed info; or worse, poorly executed data analysis.
https://www.speediagnostix.com/shop/p/standard-drain-sample-kit
I would expect an email response to be very detailed; well above the "average person" answer.
(This reminds me of the Dyson UOAs from years past ... IYKYK .... ).
Even bad advice is actionable advice. Just because you can put something into motion doesn't automatically make it a good thing to do. Being "actionable without a bunch of time ... " could also be interpreted as taking the short cut because "TLDR" fits one's mantra. His approach being "on par" with the mainstream BITOG approach doesn't make it right; just echoes what is popular.
Singular samples have no basis for averages, and small sample sets have no trustworthy basis for understanding variation. It's hard enough to control inputs and parameters in a lab, let alone one's daily-driver. Without a sufficient quantity of samples, the Stdev is subject to wildly inaccurate valuation, and if you don't know Stdev, then you have zero ability to understand what is "normal" to any process. Absurdity best describes this approach.
Sometimes, perhaps often, the devil is in the details. You gotta be willing to put time, money and effort into really understanding something to make solid conclusions. You can make haphazard conclusions and perhaps have a chance of getting it right; swags will get you there every once in a while. But to really make an informed decision, something more than "average" effort is required.
To be fair, LSJr's email recommendations in the OPs post certainly are not harmful to the engine; we're smart enough to realize that. Even though they violate any sensible measure of good data analysis, they won't result in destruction of the engine.
But this is BITOG, not Reddit. We should be shooting for much more than "average" answers. And I'd certainly be disappointed if I paid $60 to an "expert" to give me "average" advice.
Again - I apologize; not meant as a taunt, but a ying to offset the yang, as it were.
Zee09
$200 Site Donor 2023
- Joined
- May 5, 2018
- Messages
- 20,701
I won't be following his advice...
My daily concern is not motor oil but crazy in a big hurry drivers I deal with daily... I still see him dumping a fresh quart of oil through a new motor as a flush and his facial expression was as if he gave away a big secret...
My daily concern is not motor oil but crazy in a big hurry drivers I deal with daily... I still see him dumping a fresh quart of oil through a new motor as a flush and his facial expression was as if he gave away a big secret...
In this thread, 2024 Tahoe LM2 Duramax 27.3k mi; HPL Dexos D 5w-30 12.2k mi; excessive Cu and Tn, high praise is given (and rightfully so) for the OP to have used a total of six UOAs (as shown) to determine his low-mileage engine was having bearing wear. He traded in before the engine failed. Although not statistically sufficient, it was enough to show something was going south prior to total failure, which @dnewton3 helps explain in post #10.
Last edited:
Mr. @dnewton3 was questioning his knowledge of statistical analysis. Mr. Newton did statistical analysis for a living and knows what he's talking about.
Agreed. Also, in addition to engine types, driving conditions and fuel dilution can have a significant impact on results. If you drive a Honda 1.5T in cold weather with predominantly short trips, you will NEVER achieve the wear rates LSJr suggests. Same engine, driven in mild weather with plenty of highway miles, yes.
all wear metals ppm added up.
I look at Al, Fe, Cu, and Si individually for his 5ppm/1K recommendation.
but he doesn't in his videos
One of the good things about this forum and some feedback from subject matter experts here is that it makes me re-think some (only days) old information I've been operating on. It *seems* to me that the only real things to note on UOAs - without doing dozens and dozens of them back to back - is how much fuel dilution is shown via GC, viscosity loss/gain, and maybe an absurd spike in wear metals or silicon contamination (50+ iron from the last time, etc). With that being said, it seems most UOAs will appear pretty boring.
I realize that.
100%.
Similar threads
- Locked