hths vs kinematic viscosity

[GMorg]

I want to preface my question by stating that I would like to learn something as opposed to starting an argument.

Lately, I have read a lot of posts that suggest that HTHSV is the primary measurement that should be considered when deciding on a viscosity. I recognize that some will suggest that OEM recommendations should be the primary factor, but I am trying to keep the discussion to those of us that consider additional factors beyond the OEM recommendations due to rarer applications or motivations. Again, I would like to avoid an argument about this point.

From the point of view of a tight-clearance lubricated surface, HTHSV seems to be an appropriate consideration of how a fluid behaves. However, it seems that the lubrication system is a system and that HTHSV can influence the overall pattern of fluid flow through the entire system.

Given two fluids with the same HTHSV but with different KV, I would think that more of the low KV oil would be diverted through paths of lower resistance to flow as compared to the high KV oil. The low KV oil would therefore have higher residence time in tight clearance areas and would not cool those surfaces as well. Although the HTHSV is the same, the lower KV fluid has greater flow rates in wide-clearance paths. The low KV fluid would cool some areas better due to this behavior, but would cool other areas poorer. Therefore, I would think that both KV and HTHSV have be balanced for a particular lubrication circuit.

On the other hand, for two oils of equal KV, but different HTHSV, the overall pattern of oil flow is more likely to be very similar between the oils. However, the tightest bearings could have dramatically higher flow rates. The assumption is that most of the oil flow does not pass through the tightest bearings.

So, it seems to me that HTHSV is very important, but that KV should not be ignored. It seems inaccurate to suggest that an oil should be considered a grade different than the KV suggests because the HTHSV is outside of the typical range for that grade.

Am I missing something here? Is it impossible for the ratio of HTHSV to KV to get to the point where too much oil is diverted through paths of less resistance to flow?


I have two examples:
1) I have build a few drip systems for doing extractions across columns. For each solvent, I have had to dramatically alter the resistance at each outlet based on the KV of the solvent. A high KV fluid will flow to every outlet while a low KV fluid will take the path of least resistance resulting in insufficient flow at the more distant outlets. It take much, much more flow from the pump for low KV fluids to get sufficient flow to the distant outlets if the resistance at individual outlets is not adjusted for the difference in viscosity.

2) You can see a similar phenomenon when adding lubricant to loose bearing (like some farm equipment). If you pump a grease through the zerk at a fixed rate, you can eventually fill the entire bearing space with grease. On the other hand, if you try the same thing with and oil, it will run through a portion of the bearing and then run out having never covered the entire surface.

 

[RamFan]

Awaiting Caterham...

 

[TFB1]

Originally Posted By: GMorg
You can see a similar phenomenon when adding lubricant to loose bearing (like some farm equipment). If you pump a grease through the zerk at a fixed rate, you can eventually fill the entire bearing space with grease. On the other hand, if you try the same thing with and oil, it will run through a portion of the bearing and then run out having never covered the entire surface.

But a engine bearing is spinning and will distribute the oil in just one rotation...

 

[jaj]

GMorg - I'm with you on this issue. Flow of a fluid under relatively low stress (like being pumped around the oil galleries of an engine and dropping back down into the sump) is entirely determined by the fluid's kinematic viscosity. For many of the roles that oil plays in an engine, HTHS is an irrelevant measure. For a few places, like sleeve bearings and piston rings, it's very important.

The relationship between HTHS and KV is not deterministic in a modern engine oil (it is in a strict Newtonian fluid, but that's not engine oil). The only reason that HTHSV is measured at all is because there is no formula that can take the KV40 and KV100 (the two measures that uniquely define the VI) and calculate the HTHSV. If such a formula existed, the industry wouldn't spend a considerable amount of money measuring it.

So, in response to your question, kinematic viscosities are the crucial measures of viscosity for any kind of low-stress flow. HTHS is the crucial measure for cases where you have high enough temperature and stress to render viscosity index improvers ineffective.

The reason that the kinematic and HTHS viscosities are often grouped across commercial oil grades is that they are all targets that blenders have to meet when designing an oil. HTHS has to be high enough to meet the standard, but low enough to deliver the best fuel economy. KV100 and KV40 have to be in the right range for the numerical oil grade, and lastly the low temperature performance has to be in range for the "W" rating.

One oil company that produces oils that don't fit the "usual" commercial product model is Redline. Redline 5w-20 has an HTHSV of 3.3 cPs, a level more typical of a commercial 10w-30.

 

[dparm]

The relationship between HTHSV and KV is actually more indicative of the shear resistance of an oil, among other things.

The way I understand it, we use HTHSV as a representation of "operational viscosity" because it is at a very high temperature that is more representative of the highest temperature it will see in areas like the bearings.

 

[CATERHAM]

http://www.bobistheoilguy.com/forums/ubbthreads.php?ubb=showflat&Number=2276634&page=1

 

[GMorg]

Thanks for posting the link to the other thread. I should have included it from the beginning. My experience with other fluids is similar to the data in the thread above, in that, at very high flow rates my extraction manifolds would flow from all outlets. At lower flow rates, the problem is more pronounced.

CATERHAM: Do your observations (or those of others) hold at low RPMs and thus lower flow rates? In other words, is the correlation between oil pressure and HTHSV nearly 1 at 1500 or 2000 rpm like it is at 6500 RPM?

I recognize that a simple manifold is not an engine and that ether, acetone, and toluene are not engine oil. However, I would like to be able resolve my observations with yours. I worry that there are special police that get involved when the laws of physics seem to be getting broken. With the current data, is seems that KV is not relevant in an engine at high RPM and at the same time, KV affect how fluids escape a circuit.

 

[LoneRanger]

Originally Posted By: CATERHAM
http://www.bobistheoilguy.com/forums/ubbthreads.php?ubb=showflat&Number=2276634&page=1


I learned a lot in the time spent reading post #1 in that thread. Thank you, Peter.

 

[JAG]

I tried but don't understand how the examples relate to both KV and HTHS viscosity. The latter is at one high temp of 150 C and a shear rate if 1 million per second. The former is at any temp. And a very low shear rate, driven by gravity only. Polymeric thickeners are what make the relationship between viscosity and shear rate a dependent one. The higher the shear rate, the lower the viscosity, generally. That's because the stress causes them to align in a way to reduce resistance to flow (reduce viscosity). When the stress is removed, the polymers that were not torn apart in the process go back to their original shape. That's called temporary shear. Permanent shear refers to when polymers are torn apart in the process and when the stress is removed, the KV is lower than what it was prior to the shear stress. Studies have found that when KV drops X percent from permanent shear, the HTHS viscosity drops around X/2 percent. This is because the polymers affect the KV much more than they do the HTHS viscosity.

The shear rate and temp. in the HTHS test were arrived at by analyzing a lot of data to find a good correlation between it and minimum oil film thicknesses in journal bearings and between piston rings and cylinder walls. I do not recall what engines were tested or what the oil temps. or load or RPMs were. To find this correlation with the HTHS test, they tried various temperatures and various shear rates. Each part of an engine causes different shear rates and temperatures, which varies by the engine type, load, oil properties, RPMs, and many other things. Yet, they ended up using one shear rate and temp. for the HTHS test. It does NOT completely replicate what oil experiences in every part of a hot engine; for example, oil on cylinder walls near TDC and BDC sees low shear rates because the piston speeds are low there. Oil consumption was found to decrease when polymer-containg oils with the same HTHS viscosity as monograde oils were used and it was attributed to higher actual viscosity of the multigrade oils on the cylinder walls. The HTHS test was necessitated by oils with polymers in them becoming prevalent. Monograde oils without polymers are Newtonian fluids so the viscosity is independent of the shear rate.

Those solvents are probably Newtonian fluids and the fluid shear rates were probably always much lower than 1 million per second. Sorry if my post does not address your comments or questions.

 

[GMorg]

No need to apologize JAG.

I agree that HTHSV sheds light on the effective viscosity under temporary shear conditions. The measurment can help insure that KV does not imply a level of performance that is not available in high stress areas. However, some oils, in the absence of VIIs have relatively high HTHSV. I actually think that this phenomenon may be related to surface interactions with more polar oils, but that is a different topic.

My curiosity is related to the idea that the effective viscosity under high stress is different than during bulk flow. So, in my mind these areas of the engine act as restrictions to flow at an outlet. Since different areas of engine would have different shear stress, then each "outlet" could have an independent restriction very similar to having different valves at each outlet from the oil circuit. So, it seems to me that as the relationship between KV and HTHSV is changed between formulations, then the actual flow through the engine would be different. Perhaps the effect is minor, but it seems rational that the effect would be present.

In addition, based on my observations with other fluids, I would predict that the phenomenon would more pronounced as KV is reduced. IN my example, adjustments to manifold outlets is required to balance flow depending on the KV. For motor oil, each HTHSV would create unique "outlets" at the edge of bearings or the end of squirters, or the ends of pushrods, ect. The balance of flow through the engine would likely be different depending on the ratio of KV to HTHSV and the threshhold of the effect at certain lubricated surfaces.

Perhaps with one formulation, only the main and rod bearings serve as restricted flow outlets but with a different formulation the cam bearing also become a restricted outlet.

I hope this is making some sense...

 

[CATERHAM]

Originally Posted By: GMorg

CATERHAM: Do your observations (or those of others) hold at low RPMs and thus lower flow rates? In other words, is the correlation between oil pressure and HTHSV nearly 1 at 1500 or 2000 rpm like it is at 6500 RPM?

Yes the engine rpm doesn't play a role that I've noticed in the relationship between the KV100 vs HTHSV. I could have chosen a lower rpm for comparison purposes but I'm more interested in the maximum OP readings at close to maximum rev's.

From a practical point of view HTHVS is sometimes referred to as "bearing viscosity" for obvious reasons. The Red Line Oil company often refers to it as such since most of their ester based oils have particularly high HTHSVs relative to their KV100 values when compared to petroleum and PAO based oils.

Another point I didn't emphasize enough early in that thread is the effect of an oil's Viscosity Index in the comparison.
While viscosity change with temperature is virtually linear at high temperatures, an oil's VI although kinematically derived will affect an oil's operational viscosity at 100C. In other words, two oils with the same HTHSV rating but markedly different VIs will have disparate viscosities at 100C. Small differences of 10 points or so won't be noticeable but 50 points or more would be very noticeable. For example a 5W-20 dino with a 150 VI and the ultra high 216 VI Toyota 0W-20 both have a nominal 2.6cP HTHSV but the 0W-20 will be close to 9% lighter at 100C in terms of operational viscosity. Of course that viscosity difference grows dramatically at lower temp's and declines at higher temp's.
That's an extreme example to emphasize that when comparing the HTHSvs of different oils, while you can ignore their KV100 spec's, you still must take into consideration their VIs.

 

[JAG]

GMorg, I think I agree with what you are saying. Your second main paragraph above is supported by he fact that oil pressure in a hot engine is more dependent on HTHS than KV viscosity. Your observations with the solvents is a new thing to me so I can't explain it in terms of what I know about motor oils in engines. It's interesting though. Fluids act in strange ways at times.

 

[JAG]

Originally Posted By: CATERHAM
snip... For example a 5W-20 dino with a 150 VI and the ultra high 216 VI Toyota 0W-20 both have a nominal 2.6cP HTHSV but the 0W-20 will be close to 9% lighter at 100C in terms of operational viscosity. Of course that viscosity difference grows dramatically at lower temp's and declines at higher temp's.

How did you arrive at 9%?

 

[jaj]

I don't follow CATERHAM's logic through any of this discussion, so I'll stay out of that part of it.

A more direct way to respond to GMorg's original question is to consider that it takes a certain level of pressure in the manifold to push the fluid the full length of the manifold header. Regardless of viscosity, this pressure head is probably pretty constant - it's determined by the pressure needed to overcome friction traveling down the manifold tube.

A thick fluid flows more slowly than a thin fluid, and it builds up a pressure head (basically it piles up) at the entry that's sufficient to push fluid past the friction and out to the full reach of the manifold. Because it's thick, it flows slowly and it doesn't take a lot of volume to maintain the head pressure and keep flowing the full length.

Thinner fluid flows faster, and so it takes a higher volume input to maintain sufficient pressure to reach the full length of the manifold. Reduce the volume, the pressure drops and the fluid stops making it all the way to the end.

The only viscosity that matters in any of this is the kinematic viscosity of the fluid. It's low pressure (the "head" is measured in millimeters) and low shear.

 

[Shannow]

Originally Posted By: CATERHAM
From a practical point of view HTHVS is sometimes referred to as "bearing viscosity" for obvious reasons. The Red Line Oil company often refers to it as such since most of their ester based oils have particularly high HTHSVs relative to their KV100 values when compared to petroleum and PAO based oils.

Another point I didn't emphasize enough early in that thread is the effect of an oil's Viscosity Index in the comparison.
While viscosity change with temperature is virtually linear at high temperatures, an oil's VI although kinematically derived will affect an oil's operational viscosity at 100C.


This...

I've designed and troubleshot bearings inn some very big, very high power spinny stuff....and I always look at the KV at the operating temperature for the Reynolds, Sommerfeld etc. numbers.

When looking at my vehicle engines, fuel economy, and protection, I always look at HTHS if it's available....you can see some of my past posts wanting engine oils to be rated on some cold flow characteristic and HTHS, rather than the barn door kinematic ratings that we currently have.

The difference is that most of the oils I play with in industry are straight base-stocks, and behave predictably in manifolds, bearings, and drain systems (and most importantly in straight cut gears).

Most of the stuff that I use in my engines have VI Improvers, and a lot of them perform in KV tests morethan fine, but don't do well in HTHS, even if there is no permanent shearing from the event.

My belief is that bearing operating clearance, wear, and friction is most accurately reflected by HTHS...BUT...the reason that it became important was that "Viscosity Improved" oils changed the traditional measures of KV, without improving the performance of the oil where it actually matters, while traditional KV measurements on straight oils were enough to predict performance.

One of the reasons that I think Amsoil's "straight" 30 10W-30 is a stellar concept.

BTW, I've seen a very common "ISO 32" HV oil rapidly become a dirty ISO "24" oil in high speed, high shear herringbone gear systems.

 

[Shannow]

Originally Posted By: jaj
I don't follow CATERHAM's logic through any of this discussion, so I'll stay out of that part of it.

A more direct way to respond to GMorg's original question is to consider that it takes a certain level of pressure in the manifold to push the fluid the full length of the manifold header. Regardless of viscosity, this pressure head is probably pretty constant - it's determined by the pressure needed to overcome friction traveling down the manifold tube.

A thick fluid flows more slowly than a thin fluid, and it builds up a pressure head (basically it piles up) at the entry that's sufficient to push fluid past the friction and out to the full reach of the manifold. Because it's thick, it flows slowly and it doesn't take a lot of volume to maintain the head pressure and keep flowing the full length.

Thinner fluid flows faster, and so it takes a higher volume input to maintain sufficient pressure to reach the full length of the manifold. Reduce the volume, the pressure drops and the fluid stops making it all the way to the end.

The only viscosity that matters in any of this is the kinematic viscosity of the fluid. It's low pressure (the "head" is measured in millimeters) and low shear.


"Caterham's method" pretty much agrees with you.

The pressure drops along the oil paths after the pump, are pretty kinematic, and pretty low.

The pressure drops from the point at which the oil enters a bearing, circulates, and leaves, are probably in the high 90s of the percentile pressure drop from the pump, and most of that pressure drop is in the areas in which the oil is resisting being ripped apart, the HTHS area (not necessarily exactly that temperature and shear rate, but "high".

So measuring the oil pressure in a fairly quiescent stream well upstream of feeding the bearings is pretty close to taking a measurement just prior to them.

You've also got to realise that in an engine, the pumps are positive displacement, so one revolution is one unit of oil that has to be delivered (in an ideal system), so oil pressure (backpressure) is a measure of how little the oil in question wants to pass through the couple of "thou of bearing clearances versus the relatively massive oil passages.

 

[GMorg]

From the discussion here, I think that the difference between my manifold and the average engine is emphasized in Shannow's and jaj's comments. If the resistance to flow in the majority of the circuit is low relative to the resistance to flow at the outlet points, then the KV becomes irrelevant since the bottleneck to flow is at the outlets where HTHSV is likely the major parameter of concern. In my example, with thin fluid, the outlet must be adjusted such that the majority of the fluid flows beyond any given outlet so that it can travel to the next outlet.

I now conclude that the phenomena that concerned me are at work in an engine, but they are irrelevant in magnitude as compared to the effects of HTHSV. So, I would guess that the relationship between KV and HTHSV does have an effect on how fluid flows through the entire oil circuit, but that the effects are so minor that it would be very rare to observe a real-world impact.

So, at operating temps and beyond, it seems that KV can be ignored. At temps much below operating temps, KV should be an important parameter. For the undefined range of temps between those two situations, I am guessing that it just doesn't matter.

Several posters have suggested HTHSV and KV at very cold temps should be the major parameters for grading oil. For engine oil, I think I understand that point of view and can support that position. Thanks for the input guys...

 

[jaj]

Originally Posted By: GMorg
...since the bottleneck to flow is at the outlets where HTHSV is likely the major parameter of concern...


You're almost there. HTHSV is simply not a factor in this. HTHSV is measured at a shear rate of 10^-6 seconds. To put that rate of shear in physical terms, it's two surfaces a millimeter apart (0.040") moving at 1000 meters per second relative to each other. That's 2,237 miles per hour, or Mach 2.9! I doubt that speeds even 1/100th of that arise in your testing manifold.

That's why I have trouble with the notion that HTHSV and oil pressure are related. HTHSV only manifests itself for a few milliseconds as the oil passes through the narrowest part of the "hydrodynamic wedge" that keeps sleeve bearing surfaces apart. As that happens, oil pressure spikes to over 1000 psi, shear is around 10^-6 seconds, and the VII's briefly lose effectiveness, dropping the viscosity to the HTHSV. As it exits the wedge, the VII's recover almost instantly and the oil starts displaying its KV rating again. Similar brief events - sliding piston rings, camshaft followers, and so on - occur in other places in the engine, but most of the time the oil flows and behaves the way its kinematic viscosity says it will.

 

[CATERHAM]

Originally Posted By: jaj
Originally Posted By: GMorg


I have trouble with the notion that HTHSV and oil pressure are related...but most of the time the oil flows and behaves the way its kinematic viscosity says it will.

I think you're over thinking this.
As I mentioned previously HTHSV is not referred to as bearing viscosity for nothing but to differentiate it from the unreliable KV100 measure that we usually rely upon when comparing oil viscosities at operating temp's. An OP gauge directly measures bearing viscosity in a running engine.
About the only time the KV100 comes into play is when the oil is draining back into the sump, not in any lubrication role.

 

[GMorg]

I agree that HTHSV is measured under the conditions that are described above. However, I do not agree that loss of VIIs performance occurs only under those conditions.

In fact, any time that laminar flow exists, linear molecules begin to orient with the flow. I work with some of nature's longest water soluble polymers. The solutions act like a gel when not in motion. As the fluid begins to flow, the effective viscosity can drop dramatically at even low flow rates and low pressures. In fact, for a while after the flow is stopped, it can be restarted without much effort because the molecules are so large that are not significantly displaced by brownian motion or diffusion- they are still oriented with the historical flow.

In other words, the phenomena observed during an HTHSV measurement can occur at other pressures and temperatures too. I would argue that in spinning bearings the effective viscosity is always lower then the KV whenever long, linear molecules are in solution - even at room temperature. The exception would be whenever the fluid also contains molecules that have strong interactions with the sliding surface. In this case, there is a counter force that orients the surface-interacting molecules across the flow.