Andy's Axioms: "Everything That Glitters Ain't Always Gold": Shear-Thinning of Hydraulic Oil
In a previous article, https://www.dhirubhai.net/pulse/hydraulic-oil-your-problem-its-pump-andy-hessler/, I promised a following article to add some greater depth and clarification.?
Let’s talk about hydraulic oil.?It’s one of my favorite things to talk about, after all.?Specifically, let’s talk about viscosity, because viscosity is the most important property of a hydraulic fluid.?Viscosity is that curious quality of a fluid that makes it try to be just a little bit like a solid.?It is the fluid’s internal resistance to flow.?It is the quality most strongly associated with lubrication as it relates to film strength.?And it has a personality all of its own.?Unlike many other properties of matter that seem to vary with just one or two other parameters, viscosity seems to be affected by several parameters at the same time.?In my years of testing fluids I’ve become fairly convinced that even the phase of the moon seems to affect viscosity.?It’s just plain hard to nail down what is actually going on sometimes.
For this discussion…we’ll only look at viscosity from a very simplistic point of view, because there can be long discussions on its role in tribology.?There are plenty of terrific technical books written on that subject.?And, in most cases it seems that even in those tremendous technical volumes all of the math, theory and brilliance applied to solving a problem assume the reader knows what viscosity is present in their system.?For us, our system will just be a simple gear pump.?I say simple.?But, it’s only simple in that it’s easy to convey the necessary pictures to the reader.?Anything described here about gear pumps applies equally to piston, vane and screw-type pumps as well.?All of these thoughts apply to any tribological system in fact.?Any system that has surfaces in close proximity to other surfaces with some fluid lubricant in between experiences many levels of physics.?And here, we’ll talk about shear-thinning.
In my search for an all-weather hydraulic fluid for a gear pump system, I set off to conquer the problem by mastering conventional wisdom.?The problem seemed simple.?In order to have an all-weather hydraulic fluid, that fluid would need to satisfy the needs of the system from the coldest temps (-40°C) to the highest temperatures generated in the system (75°C or so). Keep in mind, these particular systems didn’t have any active cooling mechanism.?No oil coolers.?No heaters.?Whatever heat could be dissipated by the generic structure of the hydraulic components would limit the maximum temperature the system would reach.?But, there was no way to actively drive that temperature down.?What one quickly discovers in a system like this is that as the oil gets hot…the system slows down.?If you try to run the system in a cold environment…you’re at the mercy of the oil thickness.?If the oil is too thick, the pumps will cavitate.?And, cavitation will lead to failures before too long.?So, in this particular situation, the answer had always been to run arctic fluid in cold environments and “ambient” oil in regular warm environments.?That’s a pain, of course.?But, it solves the general problem of being able to operate in two different environments when there are no heaters or coolers available.
Conventional wisdom seem to suggest that if I could find an oil that had the properties of the arctic fluid and the properties of the ambient fluid, that one fluid should be sufficient to span the entire temperature spectrum.?This lead to the discovery of the magic of the Walther equation and how it can be used to plot viscosity as function of temperature as a straight line on a graph. (REF https://www.astm.org/Standards/D341.htm) All you have to do is pick your viscosity at 40°C and 100°C and connect the two dots and that’s the answer to all of your problems!?The flatter the line, the higher the viscosity index.?So, this young engineer quickly figured out that all he needed to do was find an oil that ran through the same 40°C point and had an appreciably high viscosity index.?With that knowledge, it was a simple task to hunt down fluids that fit that category.?Now, back in the late 90’s there only a few fluids that met this criteria.?Today, it’s a much different story.?But, it was at this point the painful lessons were handed down to me, one after the other.
Gear pumps, being a fixed displacement type of pump, take a given volume of oil and sweep it around the housing and push it out the other side.?The geometry sets up a theoretical swept volume.?What comes out the other side is the actual volume or actual flow.?The ratio of these two values are of course the volumetric efficiency of the pump.?Volumetric efficiency multiplied by the mechanical efficiency totals the overall efficiency of the pump.?And, for our purposes here, the mechanical efficiency is not quite as interesting. So, that’s the last it’ll be mentioned here.?Oh, but the volumetric efficiency is such a wondrous beast in itself.?For most gear pumps the volumetric efficiency changes depending on the pump’s speed, pressure and viscosity.?There is a sweet spot where the pump is efficient and quiet(ish).?And, those systems, being driven by an electric motor, were optimized for sound quality as much as volumetric efficiency.?Pressure and speed were in the control of the operator depending on what they were doing on the vehicle…and what they were lifting.?So, the actual performance of the system became a very strong function of the oil viscosity.?Speed was fairly fixed.?Pressure was in a narrow range.?But, viscosity could vary quite a bit…and during hard work cycles, performance would noticeably drift off as the oil got hotter and hotter.?This was, as one could guess, quite a bit worse if the unit was using arctic fluid…which started out thinner…and got so thin that at high temps it would no longer be able to perform its job as a lubricant for the pump.?Oh…those poor chewed up gear pumps.??
I quickly offered up solutions that I was able to plot out on the ASTM D341 graph paper.?Nice flat lines with oil at 46 cSt at 40°C (or so) and at the high end the oil would stay thick.?In fact, these oils promised (on paper…gotta love paper promises) that they’d actually be thicker than my standard oil at the highest operating temperatures.?Barrels of oil were ordered and vehicles were drained, refilled and tested.?And I failed…again…and again…and again.
This is where I was fortunate to come across Mr. Ted Selby at the Savant Group.?And, he kindly explained to me what I was missing. (Check out his site and resources here )?I’d probably still be beating my head against the wall if it wasn’t for his help.?Simply put, I made the mistake of neglecting physics and believing marketing material.?That isn’t to say anything bad about marketing people.?If an engineer like myself can completely miss the physics, it’s hard to get too mad at marketing folks that have less access to the physics.?And that was the glitter that kept tripping me up.?It wasn’t anything insidious.?But, I took what was presented to me as being sufficient information for solving my problem.?And, anyone familiar with selecting greases or oils know this very well.?You’re presented many physical parameters.?What each lubricant manufacturer presents will vary a bit.?But, they’ll almost always provide those two magical points for viscosity at 40°C and 100°C.?With those points you can fill in the Walther equation?and plot out the viscosity for the oil at any other temperature.?And, on the ASTM D341 graph, you can plot out beautiful straight lines that provide that same prediction.?I became single-minded in my pursuit of finding values of viscosity at those two points that met my criteria.?Along with that you’re often given the even more magical number for a fluid’s temperature viscosity relationship called its viscosity index (VI).?The higher the VI, the flatter the viscosity temperature relationship. Now, one can paw though reams of tech specs and just scan for high VIs and find good candidate fluids.?(Oh…the dream of using silicone oils!?Viscosity Indexes in the 400’s!)
The humbling part that Mr. Selby explained to me was that hydraulic oils are non-Newtonian.?(Ikr…of course…everyone knows that ??)?Well.?More specially, hydraulic fluids are (or more accurately “were” as this is increasingly less and less the case) made with these crazy long chain molecules called viscosity index improvers.?That’s how they get those glittery eye-catchingly high viscosity indexes.?Oh, and those fancy molecules,?polymethacrylates, polyisobutylene, and other things that no one can/will disclose that are in the oil to do the thickening at high temperatures…they have this weird tendency to experience this thing called shear-thinning.?He also explained to me that the standard values in the catalogs are non-sheared values.?The ASTM D445 viscosity test is done with basically a zero shear rate.?The shiny, glittery gold-like numbers that I was chasing weren’t gold after all.?They weren’t devoid of value.?But…they had terribly little value in relation to what I was trying to do.
So, what is shear-thinning??As best as I can explain it, this is when a complex fluid like a hydraulic oil has sufficient freedom within its structure to allow its component parts to select their own orientation along the direction of flow.?These long chain molecules used to improve the viscosity of a fluid at high temperatures don’t have the structural rigidity to keep their shape as they are forced through a gap.?They flatten out.?And, when they flatten out they no longer provide the thickening effect they were meant to have.?They thin out in proportion to the shear rate they experience.?And, in gear pumps, they experience a ridiculously high shear rate.?It was hard for us to even simulate these shear rates back then to accurately measure a fluid’s shear-thinning properties. (It’s somehow easier in the year 2020 than it was in 2000)?A fair estimate of the shear rate is around 6 million reciprocal seconds (1/s).?Without having perfect dimensions to work with, I estimate it’s roughly the same for piston pumps.?(Though, I have not done a thorough study of this).
领英推荐
So, the oil isn’t experiencing the viscosity I was predicting it would.?The very unfortunate part was that the lubricant companies themselves were not really paying much attention to this characteristic.?They would occasionally talk about a permanent loss in viscosity.?That is, the fluid can experience some amount of temporary viscosity loss due to the shear-thinning effect.?And, it can experience a permanent loss in viscosity at those same VI improvers rupture completely.?They’ll mechanically fail and get chopped up.?When a lubricant manufacturer talks about “shear stability”, they are talking only about this permanent loss in viscosity.?This is important for the overall life of the oil.?But, it doesn’t (strictly) have a bearing on the temporary reduction in viscosity that this flow alignment causes. Being Newtonian simply means a fluid’s viscosity is not affected by the shear rate.?In that sense, a fluid can be completely Newtonian (water, mercury) or along a spectrum of “Newtonian-ness”.?A straight anti-wear (AW) hydraulic fluid with a VI of 100 is mostly Newtonian.?So too are unmodified polyalphaolefins (PAO), polyalkylene glycols (PAG), and polyolesters (POE).?One will note that these other base oils have a naturally higher viscosity index (I’ll write another article on why VI is a terrible number…but for now I’m stuck referring to it).?Phosphate esters have notably terrible viscosity indexes, but they’re still Newtonian.
All of my testing was resulting in failure simply due to the fact that the oil was still pretty thin.?While the equation predicted one viscosity, that equation didn’t account for the shear effects in play.?And, so, the pump’s volumetric efficiency still suffered even though the catalog suggested the oil should have solved the problem.?Remember…catalog numbers are just catalog numbers.?Don’t build your rocket based on catalog numbers.
Now, to make a long story short…the trick is to be able to identify fluids that have a high viscosity index and also have very Newtonian behavior.?This is still rather challenging…as there are just a handful of places that can test fluids like this for the public.?I still have to go through a long explanation of what I want to test and why whenever I talk about shear-thinning. In fact, this article is doing double duty in that it written to help some folks I’m working with gain the necessary perspective on my challenges. There are quite few lessons I need to constantly keep reminding myself of that come up in this whole story. ---?
There are many levels of physics at play all at the same time.?Catalog numbers are just that…catalog numbers.
There is probably an expert out there that knows exactly what you’re up against.?Seek their help.?
As soon as you think you’ve figured something out…be prepared to be humbled ??.?
Always keep your chin up and always believe in the impossible.
And, of course not everything that glitters is gold. ;)
(Enjoy the jams)