Difference between different oil weights with similar HTHS?

Joined
Jan 19, 2012
Messages
458
Location
UT
I am having trouble wrapping my head around the idea of oil weight vs HTHS. I understand that basic difference, but what difference would I get from different weight oils that have similar HTHS.

Lets just say, for example, that I could go buy a regular 5W-40 and a stout 5W-30 Euro oil that have the same HTHS. One is 40 weight at operating temp while the other is 30 weight at op temp. What will the difference be in performance or protection if any?
 
All else being equal, I would select the oil that gives you the most HTHS for the least KV100. This would probably give you the least polymer VII load and better shear stability, lower volatility etc.

It gets difficult when comparing a Group IV based PAO 0W40 to a Group III based 5W30, that are both A3/B4 rated (high HTHS). Easy when it’s a Group III 5W30 A3/B4 to a Group III 5W40 A3/B4, (I go the 5W30).

Look for other OEMs like MB 229.5 which means Noack volatility is below 10%, which is a very good sign.
 
BTW check out M1 5W30 ESP

Properties and Specifications​

Property
GradeSAE 5W-30
Viscosity, cSt @ 100°C, mm2/s, ASTM D44511.8
Sulfated Ash, mass%, ASTM D8740.8
Hi-Temp Hi-Shear Viscosity @ 150 C, mPa.s, ASTM D46833.5
Pour Point, °C, ASTM D97-48
Density @ 15 C, g/cm3, ASTM D40520.8458
 
Back in the day, we used to have this thing called the Harrier jump-jet. It could take off vertically, hover...and get this...'bow' to the crowds at airshows! Of course we had other aircraft too... Tornadoes, Jaguars, venerable Vulcans, etc but the Harrier was 'what made us great'.

Which is odd because objectively, as a fighter aircraft, it really wasn't that great. Too expensive, too complicated, too slow, too thirsty, inadequate range, too little load carrying capacity plus a marked propensity for killing its pilots in accidents but hey, those bows. Oh we clapped & cheered!

As the decades have all too quickly ticked by, I have grown to regard the HTHS test as something not unlike the Harrier; an interesting & clever technical oddity which sounds important but actually isn't.

It's technically clever because it measures fresh oil viscosity at conditions of relatively high temperature (150°C), high speed & high rotating shear (like what you get in bearings). Under these conditions, the long chain polymeric molecules that make up the oils VII can 'align' & cause a partial but TEMPORARY drop in the oil's viscosity.

To which I say, so what? First off, unless you're hammering it at Daytona, your bulk oil temperature should never get anywhere near 150°C. Typically, bulk oil temperature tracks your coolant temperature so should top out somewhere around 100°C. I'd hazard a guess that over the last five years, my car's oil hasn't even hit 100°C because moderate speed/light load driving doesn't overstress the cooling system.

Let's also talk about the relative importance of HTHS as a test. I never kept track but as a calibrated stab in the dark, over the course of a typical crankcase oil development, I'd estimate I'd run 5,000 simple, drop tube kinematic viscosities for every ONE HTHS test. Say you were doing a Euro 5W30 program. Typically, at the start of the program, you'd check if you could make a commercial finished oil with 3.5 min HTHS without falling foul of the upper KV100 limit of 12.5 max. If yes, you might then ignore HTHS until you write up the Candidate Data Package at the end of the program. That's at the development stage. When it comes to the commercial blending of oils, I doubt if HTHS ever gets checked as part of anyone's regular QC regime.

And let's also acknowledge that the importance of HTHS is regularly & systematically undermined by the same people who insist that it's important & essential! Historically, the European OEMs were fixated on the importance of having oils which have an HTHS of 3.5 min. This figured prominently in the early ACEA specs & the CCMC specs which preceded ACEA. But then A1/B1 appeared allowing 2.9 min HTHS followed by A5/B5 which demanded you meet all the wear requirements of A3/B4 but yet allowed 2.9 HTHS. And while all this is going on, the Yanks have moved en masse to 5W20 with 2.6 min HTHS, demonstrating that it is realistically do-able!

My advice is try not to get too overly fixated on HTHS. I used to be 'that person' but I'm better now...
 
Last edited:
And let's also acknowledge that the importance of HTHS is regularly & systematically undermined by the same people who insist that it's important & essential! Historically, the European OEMs were fixated on the importance of having oils which have an HTHS of 3.5 min. This figured prominently in the early ACEA specs & the CCMC specs which preceded ACEA. But then A1/B1 appeared allowing 2.9 min HTHS followed by A5/B5 which demanded you meet all the wear requirements of A3/B4 but yet allowed 2.9 HTHS. And while all this is going on, the Yanks have moved en masse to 5W20 with 2.6 min HTHS, demonstrating that it is realistically do-able!
The Yank's drive to 5W-20 is strongarmed by CAFE. The Yank automakers are willing to sacrifice some engine protection for CAFE credit $$$.
 
The Yank's drive to 5W-20 is strongarmed by CAFE. The Yank automakers are willing to sacrifice some engine protection for CAFE credit $$$.
If you can actually prove that moving to 20 weight oils 'sacrifices some engine protection', then I would be interested to see the results. These oils have to meet exactly the same demonstrable wear targets as the thickest 20W50. If anything, with timing chain wear now thrown into the mix, standards for wear are higher than they've ever been.
It's been a while now but I for one looked long and hard at the wear data I was generating. Deducing any kind of trend in the numbers was nigh on impossible. You sort of come to the conclusion that it's the additive system your using & not the oil's viscosity that's the main determinant of wear but it's not just a question of how much Zinc or Moly; there are other factors in play such as the levels of oil acidity you end up with. Remember, this was actual engine test data, not the silly bench friction co-efficient numbers that suggest more wear for thinner oils, which everyone in the industry knows are completely & utterly meaningless!
I've been using 0W20s since 2006. No one 'forced' me to do it or profited on the back of it. With my oil formulator's hat on, it just struck me as 'the smart play' to drop the viscosity of the oil. With what I know now, I might have given a Group II 10W20 a go instead but early retirement beckoned & I never got the chance to check it out. Pity that...

PS when I met my wife, I was a smoker. Perhaps 'strongarmed' is too strong a word to describe what she did but she 'powerfully encouraged' me to give up the cigs & I stopped after about 18 months. 26 years on & importantly still alive at 65, I'm rather glad she did. Sometimes a bit of external influence isn't such a bad thing.
 
Last edited:
If you can actually prove that moving to 20 weight oils 'sacrifices some engine protection', then I would be interested to see the results. These oils have to meet exactly the same demonstrable wear targets as the thickest 20W50. If anything, with timing chain wear now thrown into the mix, standards for wear are higher than they've ever been.
It's been a while now but I for one looked long and hard at the wear data I was generating. Deducing any kind of trend in the numbers was nigh on impossible. You sort of come to the conclusion that it's the additive system your using & not the oil's viscosity that's the main determinant of wear but it's not just a question of how much Zinc or Moly; there are other factors in play such as the levels of oil acidity you end up with. Remember, this was actual engine test data, not the silly bench friction co-efficient numbers that suggest more wear for thinner oils, which everyone in the industry knows are completely & utterly meaningless!
I've been using 0W20s since 2006. No one 'forced' me to do it or profited on the back of it. With my oil formulator's hat on, it just struck me as 'the smart play' to drop the viscosity of the oil. With what I know now, I might have given a Group II 10W20 a go instead but early retirement beckoned & I never got the change to check it out. Pity that...
I'd have to go search my computer for the link to the wear study paper this figure came out of, but in basic terms it looks like around HTHS of 2.6 (about what most xW-20 is) is where accelerated piston ring wear starts kicking in. There are other wear studies out there showing that lower viscosity oil can cause more wear. I think there was one study that used different viscosity oils that were blended with the same additive package so they could see just the viscosity effect.

Piston-Ring-Wear-vs-HTHS-at-130C-Oil-Temp(2).JPG



I agree that with more modern oils, the AW/AF additives can take to place of viscosity to some degree to help mitigate wear, but I don't think viscosity alone will ever be fully replaced with additives to prevent all wear because the viscosity is the main factor that keeps moving parts separated and from contacting each other in the first place. Film thickness (generated by viscosity) can't be totally replaced with film strength (provided by AW/AF additives).

Are you or were you an oil formulator? ... sounds like it.
 
I'd have to go search my computer for the link to the wear study paper this figure came out of, but in basic terms it looks like around HTHS of 2.6 (about what most xW-20 is) is where accelerated piston ring wear starts kicking in. There are other wear studies out there showing that lower viscosity oil can cause more wear. I think there was one study that used different viscosity oils that were blended with the same additive package so they could see just the viscosity effect.

View attachment 45644


I agree that with more modern oils, the AW/AF additives can take to place of viscosity to some degree to help mitigate wear, but I don't think viscosity alone will ever be fully replaced with additives to prevent all wear because the viscosity is the main factor that keeps moving parts separated and from contacting each other in the first place. Film thickness (generated by viscosity) can't be totally replaced with film strength (provided by AW/AF additives).

Are you or were you an oil formulator? ... sounds like it.

I did 31 years in the oil industry with the last 12 playing around with the slippery stuff (mainly global PCMO development with the odd foray into HDDO). Was doing the equivalent of a billion litres a year of finished oil. Best job I ever had but towards the end, surrounded by the worst people I'd ever worked with. I could cope with wear, oxidation & shear but not envy, malice & spite. Threw in the towel at 53 for a quieter life. Best decision I ever made.
 
PS when I met my wife, I was a smoker. Perhaps 'strongarmed' is too strong a word to describe what she did but she 'powerfully encouraged' me to give up the cigs & I stopped after about 18 months. 26 years on & importantly still alive at 65, I'm rather glad she did. Sometimes a bit of external influence isn't such a bad thing.
Probably a good result of some "strong arming" ... good for her and you. (y)

Here's how the whole strong arming of the US auto industry got started to go thin - see attached PDF.
 

Attachments

  • EPA 5W-20 Letter.pdf
    332.4 KB · Views: 89
Yes the cutoff seems to be around 2.6 cP from what I've seen aswell. Trouble is fuel dilution starts at the piston rings, the dilution will be at it's worst there. How much fuel does it take to drop 0.2 cP? The rest of the engine can likely cope with less viscosity just fine.

I often find cars (mostly diesels, but some petrols aswell) with the oil level 20% over maximum. The owners didn't add anything, or lift the bonnet, and the last oil change was done where I work so in absence of missing coolant it must be water (but doesn't look like it) or fuel. That measn the additive concentration went down 20% at least, and fuel content could actually be higher, if the pcv blew out oily fuel vapours into the intake... Do we really want to start these cars out with 2.6 or even 2.9 cP oil knowing they won't stay there?
 

"Due to their greatly reduced volatility and good low-temperature performance, new base oils of API Group II-IV allow the formulation of lighter automotive viscosity grade oils, such as SAE 5W-40, 0W-30 and even 0W-20, to achieve better fuel economy. However, as explained in Figure 4, the use of thinner base oils increases the risk of engine wear unless appropriate friction modifiers are simultaneously deployed in the formulations. By shifting the Stribeck curve to the left in Figure 2, friction modifiers cause an equivalent shift of the wear and the frictional losses curves in Figure 4. The result is that the“optimal viscosity” point corresponding to the greatest fuel economy also is shifted to the left, towards lower viscosities. In practice, however, it is wise to prefer a somewhat heavier oil to a somewhat lighter one to further minimize wear."
 
I did 31 years in the oil industry with the last 12 playing around with the slippery stuff (mainly global PCMO development with the odd foray into HDDO). Was doing the equivalent of a billion litres a year of finished oil. Best job I ever had but towards the end, surrounded by the worst people I'd ever worked with. I could cope with wear, oxidation & shear but not envy, malice & spite. Threw in the towel at 53 for a quieter life. Best decision I ever made.

Welcome back Joe :)
 
I'd have to go search my computer for the link to the wear study paper this figure came out of, but in basic terms it looks like around HTHS of 2.6 (about what most xW-20 is) is where accelerated piston ring wear starts kicking in. There are other wear studies out there showing that lower viscosity oil can cause more wear. I think there was one study that used different viscosity oils that were blended with the same additive package so they could see just the viscosity effect.

View attachment 45644


I agree that with more modern oils, the AW/AF additives can take to place of viscosity to some degree to help mitigate wear, but I don't think viscosity alone will ever be fully replaced with additives to prevent all wear because the viscosity is the main factor that keeps moving parts separated and from contacting each other in the first place. Film thickness (generated by viscosity) can't be totally replaced with film strength (provided by AW/AF additives).

Are you or were you an oil formulator? ... sounds like it.


It appears that wear also correlated with engine RPMs. Most modern engines cruise at legal highway speeds at 2000rpm or less.

Another point, how does this accelerated wear equate to miles ? The question has been asked before. Does a 300k mile engine only go 290k on 20 grade oil?

We may never know the answer.
 
Probably a good result of some "strong arming" ... good for her and you. (y)

Here's how the whole strong arming of the US auto industry got started to go thin - see attached PDF.

A couple of things...

As regards the EPA letter, if I read it correctly, the request to use 5W20 came initially from Ford & Honda. The EPA letter in effect lays out the conditions required to stop these OEMs pulling a fast one, using 5W20 to validate their vehicles & then using a heavier weight oil thereafter for factory & service fill oils. To me, that's not unreasonable. I personally wouldn't hold this up as evidence of 'strongarming'.

As regards the graph of piston ring wear rocketing up as you go beyond 2.6 cP HTHS, I can see why this might be concerning... and also why it actually isn't. This graph reflects a philosophical fault line in the way oils are viewed & tested. That's to say oils are tested to fit the extreme & end up not honestly reflecting reality. On the strength of this chart, (which presumably like the EPA letter, got issued a couple of decades ago) I should today be able to go out with my pressure gauge & find tens of millions of cars that have run on xxW-20 oil & which badly fail a basic compression check, on account of their top rings being worn away to nothing.

I strongly suspect that is very much NOT the case because the graph is misleading. It's misleading because the tests have been carried out with a bulk oil temperature of 130°C. Had the tests been carried out at 100°C (which is the more typical maximum temperature for engine oil in normal service) things would have looked far more benign. Had the tests reflected true driving reality where the oil cycles between cold & warmed up (10°C to 100°C) it would be even more benign still.

I referred to two industry tests earlier in this thread, the Sequence IIIG & the Peugeot TU5. Both of these tests run flat out with the bulk oil temperature controlled to 150°C; the first for 100 hours, the second for 72 hours. At this temperature mad things can happen to oil. It can oxidise like crazy. I've had oils where the kinematic viscosity at 100°C has risen from 14 cst to over 5,000 cst! I've also had tests where the the test stops itself because the sump has emptied itself & oil pressure drops to zero. After a while, you twig that these tests, fun as they are to investigate, have absolutely no bearing on reality whatsoever & you should never try to extrapolate from them.

Anyway, that's my two pennethworth...
 
One last anecdote...

I bought my wizzy, little Daihatsu Sirion in 2006. The factory fill oil was 0W20. After 9,000 miles, the engine got its first oil change & because I could, I put the used oil in for full analysis (what can I say? It was free back then!).

The results were interesting. First off the KV100 of the oil was a mere 5.0 cst! That is mega-thin by engine oil standards. The reason? 8% fuel dilution (think I did this by TGA so very accurate). Never dug into why it was that low but it never got that high ever again so I suspect it was simply the rings still bedding in.

The most remarkable results though were the wear metal numbers which came off the ICP. Basically there were none! It took me a while to get my head around this because such a thin oil should have caused bad wear right? Then I realised that whilst the oil was relatively thin, if the oil spent most of those 9,000 miles below 100°C, then it would in absolute terms be relatively thick, thick enough to maintain a decent film strength & prevent wear.

I loved that car for seven years. In the end I gave it away to my son-in-law who promptly drove it into a herd of deer!
 
Back in the day, we used to have this thing called the Harrier jump-jet. It could take off vertically, hover...and get this...'bow' to the crowds at airshows! Of course we had other aircraft too... Tornadoes, Jaguars, venerable Vulcans, etc but the Harrier was 'what made us great'.

Which is odd because objectively, as a fighter aircraft, it really wasn't that great. Too expensive, too complicated, too slow, too thirsty, inadequate range, too little load carrying capacity plus a marked propensity for killing its pilots in accidents but hey, those bows. Oh we clapped & cheered!

As the decades have all too quickly ticked by, I have grown to regard the HTHS test as something not unlike the Harrier; an interesting & clever technical oddity which sounds important but actually isn't.

It's technically clever because it measures fresh oil viscosity at conditions of relatively high temperature (150°C), high speed & high rotating shear (like what you get in bearings). Under these conditions, the long chain polymeric molecules that make up the oils VII can 'align' & cause a partial but TEMPORARY drop in the oil's viscosity.

To which I say, so what? First off, unless you're hammering it at Daytona, your bulk oil temperature should never get anywhere near 150°C. Typically, bulk oil temperature tracks your coolant temperature so should top out somewhere around 100°C. I'd hazard a guess that over the last five years, my car's oil hasn't even hit 100°C because moderate speed/light load driving doesn't overstress the cooling system.

Let's also talk about the relative importance of HTHS as a test. I never kept track but as a calibrated stab in the dark, over the course of a typical crankcase oil development, I'd estimate I'd run 5,000 simple, drop tube kinematic viscosities for every ONE HTHS test. Say you were doing a Euro 5W30 program. Typically, at the start of the program, you'd check if you could make a commercial finished oil with 3.5 min HTHS without falling foul of the upper KV100 limit of 12.5 max. If yes, you might then ignore HTHS until you write up the Candidate Data Package at the end of the program. That's at the development stage. When it comes to the commercial blending of oils, I doubt if HTHS ever gets checked as part of anyone's regular QC regime.

And let's also acknowledge that the importance of HTHS is regularly & systematically undermined by the same people who insist that it's important & essential! Historically, the European OEMs were fixated on the importance of having oils which have an HTHS of 3.5 min. This figured prominently in the early ACEA specs & the CCMC specs which preceded ACEA. But then A1/B1 appeared allowing 2.9 min HTHS followed by A5/B5 which demanded you meet all the wear requirements of A3/B4 but yet allowed 2.9 HTHS. And while all this is going on, the Yanks have moved en masse to 5W20 with 2.6 min HTHS, demonstrating that it is realistically do-able!

My advice is try not to get too overly fixated on HTHS. I used to be 'that person' but I'm better now...

My F150 Ecoboost routinely sees 240°F/115°C coolant temps. Pretty much any time I tow and it is over 25°C ambient I can get there without trying.

Plenty of trucks here in the US can run very high oil temps under load.
 
Last edited:
Back
Top