Micropitting in Gears

This podcast episode dives deep into the persistent and often misunderstood issue of micropitting in gears, guided by expert insights from Ben Wainwright. The conversation begins by defining micropitting and examining how it manifests in real-world applications, asking whether it remains a practical concern in modern gear systems and what fundamental factors—such as lambda ratio, surface roughness, and contact conditions—drive its formation. We explore how micropitting should be properly characterized, which surface finishes are most effective at reducing risk, and the complex role lubricants play, including whether antiwear and EP additives mitigate or inadvertently promote damage. The discussion then turns to testing: how micropitting is currently evaluated, whether today’s tests truly reflect field failures, and what an ideal test might look like if designed from first principles. The episode closes with practical guidance for gearbox designers, common misconceptions still seen in industry, trusted rules of thumb, and a forward-looking question on whether micropitting is something we will ultimately engineer out—or simply continue to manage. 

Our Guest:

Benjamin Wainwright completed his studies in mechanical engineering at Imperial college London in 2017. During his undergraduate studies he was heavily involved in the maintenance of the 1902 Car mascot of the City and Guilds. This along with a keen interest in the fundamentals of engineering inspired Ben to take on a PhD in the Imperial College Tribology Group focusing on surface fatigue of gear and bearing applications. Ben was awarded his PhD in 2022 and shortly after continued his career within the R&D department of PCS Instruments. Here, together with his fellow R&D engineers, he has contributed to the development and release of a system to allow research of electrified contacts as well as high speed twin disc machine focusing on high speed EHL.

Marc Ingram has over 10 years experience working in Tribology.  His masters degree in Chemistry and PhD in Engineering (Tribology) helps him easily bridge the gap that so often exists between lubricant formulator and field/test engineer.  He has worked with all major oil and additive companies after a successful career at Imperial College, PCS Instruments and Afton Chemical.  Marc has extensive knowledge of all tribology test methods inc

For more information on STLE, please visit https://www.stle.org/ If you have an idea for our podcast, or interested in being a guest, please Email STLE Director of Professional Development Robert Morowczynski at rmorowczynski@stle.org .   Also, we love your feedback, please take a minute to provide us with your thoughts at Perfecting Motion Podcast Feedback or Email us at rmorowczynski@stle.org.  

Transcript

Welcome to Perfecting Motion, an STLE podcast series that talks with members and
industry professionals about current issues and trends impacting the global tribology
and lubricants community. Hello, and welcome to our STLE podcast, Perfecting Motion.
I am Bob Morsensky, STLEs Director of Professional Development. And today we have
with us Benjamin Wainwright from PCS Instruments. And our host today, and back for
another episode, is Mark Ingram from Ingram Tribology. Today, Mark and Benjamin will
be talking about micropitting and some of the innovative ways we are seeing
advancements in the field. Mark, I'll head over to you. Thanks for being here. Hey,
thank very much, Bob. Yeah, today our subject will be micropitting, which is a
failure mode normally seen in is. We have one of the experts of micro -pitting with
us today, Benjamin Wainwright. We are going to discuss the background of micropitting,
the mechanisms, and how people can prevent it. So, Ben, how would you define micro
-pitting? So for me, micropitting is a form of rolling contact fatigue,
and it's due to the imposed stresses by asperity interactions of contacting surfaces,
mainly in rolling sliding lubricated contacts of hardened steels. So these asperity
interactions, they create pressure spikes and stress concentrations. So it's often kind
of described as rolling contact fatigue on the asperity level. And it predominantly
kind of occurs when the roughness is relatively large compared to the predicted
lubricating fill. And of course, it differs from macro pitting, which is influenced
by the stresses of the entire herzian contact. It's it's more predictable than
macropitting because you're not looking at like a single point of failure which can
have a very large scatter but you're actually looking at thousands of like parallel
similar fatigue events all happening at the same time so it's not a case of first
time to failure but actually material loss for either you know area or volume and
you essentially kind of it's obeying the law of large numbers from this point
I like the way you describe it as been like a parallel, lots of parallel fatigue
mechanisms happen together. That that's like a really nice way of describing it,
actually. So then in the field, Ben, how is it manifested like in gears and
bearings? It can kind of have a few kind of manifestations. I think, yeah, you're
right. You know, gears and bearings are some of the most prone machine elements to
micropitting. It can appear in essentially micropitting, as its name describes,
is pitting on the order of tens of microns. And what can happen is you develop
this large crack network with these micropits within. And it can cover quite a large
appreciable area, so it can have other kind of names like grey staining, peeling,
depending on which application you're focusing on. And This kind of large crack
network area is responsible for the erosion of significant amount of material, leading
to Bayes through micropitting. Though another way it can manifest itself is in a bit
more of a sporadic way, so you can have singular isolated pits on the surface.
This is still micropitting and still governed by the same physical laws that
micropitting, but it can look quite different because you haven't got this dense
crack network. And this is often not really responsible for larger amounts of
material loss through macropayin, but maybe more of initiation points for larger
propagating cracks, which can create macro pitting instead. You mentioned the kind of
scale of micropitting. How different is it to like a normal way? You know,
are we 10 times the loss in volume, or where are we on that kind of scale? It's
really difficult to answer something like that, I suppose, because micropitting wear
are, they're kind of competing planar modes. If you have a lot of wear, I don't
think you'd expect to see a lot of micropitting because you'll either wear away the
roughness, disparities, or you'll wear away your microcracks. So conditions that
promotes a lot of wear, you know, appreciable enough to, you know, a cause of
failure are not going to promote a large amount of micropitting. But I suppose in
terms of tolerance of machine elements, the erosion of material, the effect is the
same. You're going to end up with noisier gears, things failing through by vibration
and perhaps even generating larger pitting as well. Is micropitting still an issue in
the field? Yeah, I would say so. And maybe even in some cases, just with certain
trends, because we all like to try and make our gearboxes as small as possible, as
power dense as possible. And we want to use lubricants that are as thin as possible
to avoid other losses like churning losses. So we're kind of compromising surface
integrity for efficiency. And in that way, we can end up developing more micropitting
that we might actually expect because our lubricating films are not good enough to
prevent completely surface interactions. So in that way, we're going to see
potentially more incidents of micropitting as we try and increase power density. Maybe
the question isn't, is it still an issue in the field? As in the premises, it's
decreasing. It might actually be increasing. Is that maybe? Yeah, well, the thing is
you try and improve in one way and you might end up promoting, you know,
micropitting or another failure mode in another way. The same thing is if you kind
of mitigate one failure mode in one area, you end up failing by something else in
another area. And that might be back in the example I gave, that might be
micropitting. Of course, there are probably plenty of examples in other situations and
lighting modes. Okay, so if we kind of summarize that then, so micro -pitting is
still an issue, happens in sliding, rolling, contacts, is a competing mechanism to
normal way, likely to be lots of, described really well by you as like a parallel
fatigue type mechanism with lots of fatigue events happening on the same surface.
Okay, so In terms of the mechanisms then, what is driving this failure?
You know, why are these little pits happening and losing so much volume of the
parts and failing? Yeah, I mean, so micropitting is influenced by a huge number of
contact parameters and effects. I think fundamentally, because micropitting is a
fatigue process, You can think about it as you require for fatigue a cyclic stress
and a history of that stress. And so migraitting on the asperity level is driven by
the cyclic stresses of asperities interacting. So the first thing you've got to talk
about is, well, what drives migraping? Well, of course, it's the surface roughness of
those contacting disparities. How sharp that roughness is, the wavelengths involved in
that roughness. And then, well, if you talk about roughness, then you've got to talk
about film thickness. And that brings us on to that quite famous parameter that
everyone likes to use when describing micropatting. And that's the lambda ratio, which
is simply the ratio of the film thickness to the roughness. So you have some kind
of metric for the lubrication quality through lambda ratio. And you have your surface
roughness. And Both these parameters, they are very important in driving microping
because they give you the level of the stress that is being opposed on the
surfaces, but also the number of contacting points of the surfaces too is driven by
both those parameters. And of course, well, we've mentioned where already,
but then, yeah, the level of wear is very important to drive micropitting. So you
can imagine if you're wearing your surfaces a lot more, you're actually reducing the
sharpness, taking away the peaks of the asperities. They're becoming, well, rounder,
more conformal, and they're not going to impose sharp stresses, and you're also going
to reduce the number of contact experiences there. Contact pressure, of course, will
have a huge effect. You actually end up, if you increase contact pressure, you
increase the stresses to the point where asperities are all plasterized, and you also
increase the size of the contact as well. So you bring more asperities into play
through two mechanisms through contact pressure. Slider ratio as well is very
important. By increasing the level of sliding, you're actually increasing the number
of disparities coming into contact with the opposing surface per cycle. So if you
imagine if you're an asperity in a pure rolling contact, You will see just one
Asperity contact as you traverse one cycle. But if there's a significant amount of
sliding, you might actually see two, three or four contact cycles within a single
macro contact. And then there's the direction of slider ratio too. So quite often we
see in the field, especially in gearing, where the direction of sliding changes at
the pitch point. Microputting preferentially forms on the didendum of the gear and
the direction of sliding relative to con. So the way I like to imagine a gearing
slider ratio kind of a setup is if I'm on the dedendum of a gear and I'm watching
the tip bit of the opposing gear come into contact and it's rolling towards me.
The direction of sliding from where I'm stood on that gear is in the opposite
direction to that rolling contact coming towards me. And that means that direction of
sliding, the tractor of forces act to kind of open up the contact of the crack
mouth. So cracks propagate in the opposite direction to traction. And that's, to me,
that happens because the traction of force acts to, in one direction,
open up the crack mouth as it rolls over, well, actually, as it is just over the
contact mouth, it is going to open up the crack mouth. However, in the other
direction, it will act to close it. And there's also that other mechanism of fluid
entrapment into the crack can act to propagate the crack even further because you've
got to get that pressurized fluid. There's been some great models on that, which
have shown that if that is the case, it really does properly propagate the crack
much quicker in a certain direction. Well, so there's quite a few driving mechanisms
there, isn't it? I think I've caught them all. So we've got roughness, which can
drive the stresses and just the number of contacts. Obviously, if those are higher,
you're likely to drive micropitting further. Film thickness, you know, if that's lower
and the lambda ratio is lower, you're likely to have bigger stresses and more
micropitting. got contact pressure, just increasing the size of the contact,
you know, putting more contact patches in play, where,
which kind of can prevent it, funnily enough, by reducing the roughness.
Yeah, I think that's it. Yeah, that's all correct. Yeah, yeah. The cycle of the
disparities, more, more disparities driving through if the slider ratio is higher
means more contacts per cycle, I guess, or per rotation or however you define it.
Directionality of the side of ratio, so the detendom of the gears being negative to
the direction of rolling, so that would promote crack growth mostly. And then you
have the fluid entrainment into the cracks causing like a hydrostatic pressure.
So, yeah, I bet this is the subject to your PhD. So There's quite a complex
interplay of all these mechanisms going on, is it? Which ones are the main drivers,
if you had to choose a couple, I guess. Which ones are the main ones driving it?
Well, roughness will have to be up there. Very important. And one of the things I
found in my PhD was that the roughness itself is much more influential than the
lander ratio. In fact, using the lander ratio on its own doesn't really give you a
great indication of the level of micropit you might expect. For example,
you can have a consular -lander ratio, lots of different levels of roughness. I would
expect the higher roughness to micropit more because it actually creates higher
asperity contact pressures. Whereas if you reduce those peaks,
though it's very slopes, you actually reduce those contact pressures and therefore
reduce the level of cyclic fatigue in that respect. So if we stay with like,
I suppose we'll start talking about solutions now, but going directly on from that
answer, from the materials point of view, what kind of roughness parameters can we
change to reduce micropitting? You know, how can we reduce it,
you know, and a cost of, I guess there's a cost effective, there's a cost effect
here as well, but, you know, bearing that in mind, how would you reduce it? I
mean, to be honest, if you could go as smooth as possible, I think the way we
should first start looking at this problem is how do we define the surface or the
roughness of the surface? At the moment, the most common approach is to use
parameters such like RA and RQ. These parameters, they're not particularly good at
capturing the important aspects of the surface when it comes to micro -pitting. For
example, R &RQ are both what we call height parameters. They only look at the
heights within the roughness profile. They contain no information at all about
wavelengths in the profile. So you could have some very tall kind of peak, but if
they have a very large radius of curvature, that would have the same IQ as a very
kind of spiky ruffer's profile with very low rate of curvature. You could get the
same IQ, but you're going to get very different levels of micropayting with both
those surfaces. So, firstly, there is that, and there are other options we can use
to define our surfaces. So parameters which contain more information on the
wavelengths are something that we call hybrid parameters would be something we called
RPK, which is derived from the bearing ratio curve. RPK looks at just peaks of the
roughness as well. So unlike RA and RQ in which you're defining your parameter based
on both the peaks and the valleys, the valleys are something that doesn't make
contact with the other surface at all. So it's not really that important in the
micropicking problem. RPK focuses on just the peaks. So that's another kind of a
reason why RPK might be a strong contender as something to define our surface in
terms of how it might promote micropitting and then of course you have other aspects
that look just at the wavelength something i spent a bit of time looking at was a
parameter called you order correlation length which gives you a single value based on
what really is the kind of the most dominant wavelengths involved in the profile.
So you can imagine, like, if the most dominant wavelengths is very, very long,
you'll have quite shallow slopes, normally quite good at preventing micropitting, your
autocorrelation length will be quite large. Then the opposite is true. So if your
autocoridating is very low, you'll have quite peaky, quite high frequency within your
roughness, which gives you those sharp asperities and sharp disparities lead to higher
contact pressures. So a week or so ago, we interviewed the chat from MPL.
I forget his name now. One of the things he was saying we should be doing was
when we published roughness of surfaces was to include something like a bearing curve
to show it fully. Yeah. That was one of the. Was that Timothy Cramps? Tim, yeah.
Yeah, sorry. Yeah, he's great. Yeah, no, he's right. I think although that is entire
chart rather than a single value. So it becomes a little bit more, a little harder
to distinguish which surface is better than which. As it,
you know, it requires more interpretation. Yeah. Whereas a single value is very easy
to kind of interpret, especially if you just want to pass over a quality control
check to say, you know, what surface roughness should I accept this gap bearing?
Yes. It's nice to have a single parameter which does that. Yeah, I like see. I
like your RPK idea as well. If we look then on the practical side, you know, when
you, when you hover gear and if it's ground and if it's polished, how can people
like practically reduce it? You know, is there a polishing stage they should get to
or like, you know, a grit size of the polishing paper? What's the, I'm less
familiar in like of gears you know how what would be the way of reducing it that
way i suppose yes adding additional finishing techniques to the end of a machining
process is is going to cost more money but you could finish off the teeth with an
additional grinding process by choosing grits of higher grit values and i'll get you
that'll get you there perhaps there's some kind of polishing slurry technique you
could use as well to polish the surfaces too. But I'm not going to just say that
everyone shouldn't now polish their gears. A, that's expensive, and B, just because
you've removed micro -pitting, you might end up promoting another failure mode instead,
like pitting. So it's really important that way you do the research first to make
sure that what you're doing doesn't exacerbate another fading mode that you weren't
expecting. Okay, very, very sensible. In terms, if we think in the lubricants term
now, how can the lubricants prevent micro -how can they slow it down or prevent it?
So lubricants are really, really important. We wouldn't get rolling contact fatigue
without a lubricant. So lubricants can control that initial wearing -in process so
that they can allow some kind of control over how much of the roughness can be
removed, perhaps as a way of delaying the effect of the anti -ware additive,
so that you can initially remove a small amount of material, kind of reduce the
sharpness of disparities before that tribo film, and additive package kind of kicks
in and protects our surfaces. I suppose we can start talking about the additives
involved, because you mentioned the first question, like, what are the driving
factors? I mean, lubricant chemistry is obviously extremely important in driving
micropitting, anti -ware additives, especially, because they protect that roughness,
ensure that those asperities are preserved and continue to stimulate fatigue cycles
onto the opposing surface. So anti -ware additives are kind of a driving force for
micropitting and something that we need to control how they act. We don't want a
particularly aggressive anti -ware additive acting on a very rough surface, which would
exacerbate the whole micropitting issue. We would like that anti -ware additive to
either be delayed or that surface's initial surface to be smoother, which is two
kind of ways that we might look at trying to control the level of micropitin that
we're getting. Friction modifiers as well. So we mentioned a bit about traction, open
up that crack mouth, increasing the speed of crack propagation. So if you have a
friction modifier, you actually reduce that level of traction on the surface. So you
slow down those, that crack propagation too. So friction modifiers are actually quite
a nice way of reducing the level of micropitting. And some nice work that Mao did
at Imperial College, he used a friction modifier, showed that it reduced micropitting,
showed how the anti -ware additive exacerbated micro -pitting increased it. And then
when he had both the friction modifier and the anti -ware additive together, he
showed that the friction modifier, because it's a competing tribo film additive was
actually delayed the effect of the anti -wear additive as well, which was Calvinized
little conclusion. You mentioned roughness and how the different process of machining
can control it. So the argument that there are kind of slow acting chemistries that
essentially polish the surface, and I write in saying that would then reduce the
RPK, like reduce the peak severity on the surface? Yes,
yeah, I expect that. So is it primarily, are the additives acting mainly as like a
secondary effect in a way that they're not reacting fast enough, you know,
allowing a little bit of polishing way to happen, and then they're reducing the
roughness, so you're less likely to get micro -pitting down the line? I think that's
the argument you make. I think that's the desirable effect, yeah. At the moment, if
an anti -wear additive really acts too quickly and protects a rough surface, I would
expect that would increase the level of micropotining scene in the system. So if we
can control the chemistry, delay that effect, allow a little bit of control wearing
in. We could probably prolong our micropicking design life there.
And then your friction modifiers act in a similar way by preventing the anti -wear
chemistries to take hold and form the tribe of film and then you're more likely to
get wear and reduce micropitting that way by reducing roughness.
Yeah, so to friction modifiers, reducing friction does reduce the level of
micropitting because you reduce the tractor forces and reduce the forces that are
affecting crack propagation so that has one effect the secondary effect i've only
seen in in one paper about it delaying the effect of the antiware additive and i
thought that was really interesting and i'm not a chemist by trade so i don't know
the full chemical reaction behind that but it looked like a very kind of a
promising promising like take on the whole kind of how do we design our package to
prevent micropitting. I suppose another way of looking at maybe countered the argument
would be once the films are formed, it would help us, you know, a lot of these
tribal films are actually designed to be very, very smooth. You know, I know you
probably preserve the roughness beneath the tribofilm or the roughness is kind of
preserved. Yeah, how does that, what do you think about that? If the tribe of films
designed to be relatively smooth then. If it's designed to be smooth,
I suppose that that is great. I think like you said, the tribal film forms onto
the surface, so it would conform to that initial roughness that's already there.
Especially if that roughness hadn't been worn away, the tribal film would just add a
secondary layer onto that roughness and preserve that from wearing away. I suppose in
a very polished surface, the tribal film can actually add roughness of 100
nanometers, especially if it's quite patchy. But to be honest, I've never seen 100
nanometers worth of roughness being from a tribofilm significant in terms of
accelerating micro -pitting. That kind of roughness wasn't the driving force, and
that's normally happening on the softer opposing surface in a classic micropitting
test, for example. Okay. Do you think the chemistry kind of alters the failure
mechanism or just kind of delays it? So if, is micropatin always going to kind of
happen eventually, even if you do things right? Or is there a way of, you know,
if you do what you think is required? We delay the tribofilm formation enough to
allow a little bit of polishing on the surface. We have a nice, smooth tribofilm,
which is thin and low friction. But we still have a roughness. You know, we still
have a measurable roughness. Yeah. Does micropayton still occur just further down the
line, or what's that effect? I think with something like additive delay in the
effect, so let's say we design a great additive package, and it reduces our
roughness a bit, and reduces our friction, and it delays micropitting. Micropitting is
still a fatigue process, and at the moment, they still suffer from rolling contact
fatigue, and we still have some asperity interactions. I would expect that to be
ongoing, but because we've managed to reduce our asperity stresses enough, that
micropitting is not happening so quick that is now the limiting factor in our
system. So now that micropatting is not limiting our system, something else will
undoubtedly be the limiting factor in our system. It could be another machine element
in our transmission, or it could be another failure mode, like pitting.
So if you remove micropitting so much that your service is smooth, then macro
pitting is probably going to be the next cause of And in some ways, micro -pitting
and macro -pitting are both a competing failure mode, and micropitting can act as the
initiating of macro -pitting as well. During my PhD, I was able to show that if you
have significant levels of micro -pitting, so much so that you remove the material
from the surface so quickly that large cracks cannot form, you can't really produce
macro -pitting, Even if your counter pressures are high enough that actually under
these conditions with a smoother surface, you would normally produce macro pitting.
And then on the other side, if micropitting is so slow, where you produce just one
or two sporadic pits, and this is kind of normally happening on smoother surface
like bearings, those micropits themselves could act as initiation points.
And because for larger cracks, and because we're not removing material very quickly
through lots of amount of micropitting, those cracks can actually propagate deeper
under the surface, producing the macro pitting failure. Okay. In your kind of opinion
then, what makes like a good micropitting test? How would we kind of test for this
mechanism, bearing a mind of things we've already said about, you know, controlling
the roughness, how to get chemistries, you know, are we going to look for,
you know, we're keeping the load the same and increasing the roughness, or are we
going to increase the load, like an increase the contact pressure, or, you know,
how have you designed your micropitting testing? Yeah, that's a great question. I
think that it's one of those questions which is really open -ended and, you know,
what is the right answer for this? I think there are two approaches to designing a
micro -picking test because, first of you know, there's a really, micro -neged effect
has a very complex nature, loads of compounding effects. So just because your micro
-picking test, your lubricant performs very well in this test doesn't mean it's going
to perform very well in every application. The first thing I'd say is if you have
the time and you have the resources running, testing under conditions as close as
possible to the application might prove very, very useful. However,
if you just want a screening test to just check for loads and loads of different
lubricants, then an accelerated micropagint test is what you want. And that is
something that you can run, which would accumulate a large number of fatigue cycles
very quickly, creating a hardness difference between your test specimen and the
counterfaces that it runs against. That's really useful because you want to promote
micropitting on the test specimen and you want your counterfaces to be essentially a
kind of constant in the test. Hopefully the roughness of those doesn't change too
much and that means that you can control that aspect because roughness, if that
changes without the test, it's going to really affect the end of course we've
already mentioned that lubricic lubricant chemistry has a strong effect on how that
roughness evolves in the test so it's really difficult to kind of control that
parameter in such a way that you're doing it realistically but also it's not the
dominating factor because you want to bring out the effect of other parameters that
you might be more interested in rather than the roughness so i I think that's
really important to consider. And then because it's an accelerated micro -pitting test,
we want to use what we define as a negative slatter ratio where the contact motion
is in the oscillarition traction, like we see on the didendum of gears. We want to
apply a significant amount of contact pressure. However, not too high because we
don't want to promote macro -pitting to be the pha -moded. We want micropitting to be
are phaeton mode. If it's too high, the hurts in stresses might dominate and perform
the macro pit before micropitting has really had a chance to establish itself. Can
that happen? Like, you know, if your contact pressure is just too high, you would
form a macropit much before micropitting happens. I've never seen that. Have you seen
that happen? Yeah, definitely, yeah, yeah. Evening, so if I control everything else
and I set up my experiment in a way that other than contact pressure. I'm confident
that I'm exacerbating or I have a contact set up that really promotes micro -pitting
and then I increase the level of contact pressure. And if you do the numerical
solutions as well, it's quite interesting. So what you find is if you increase above
a certain point, you end up having every single asperity in conduct will be placidly
deformed so that it has reached as limiting contact pressure. And because of that,
all those kind of disparities end up being their curves is curvature increases and
when the depth of which those asperities stresses reach the subsurface also increases
because it's proportional to the contact area of each asperity the contact radius of
each disparity and that means those asperity stresses are pushed deeper into the
context into the surface which which promotes macro pitting. It seems to promote the
growth of larger cracks, which at some point then interact with the, with the
Hertzian stresses a lot more, promoting it even further. Okay. It's like a really
accelerated kind of pitting test.
It's interesting, you know, the way you're describing if you could set up a test
is, it's always interplay between accelerating a test and keeping it realistic is the
challenge, isn't it? Absolutely, yeah. Reducing the hardness of your, of your kind of
test specimens who it induces failures and, and you're keeping the hardness harder
on, on the counter surfaces. The argument about then the control and the roughness,
you know, there'll be probably a mismatch there as well, as opposed to saying using
the same hardness. Yeah, so if you, yeah, you're right. And because the test
specimen is after its roughness is often just worn away immediately. So it's not
really playing a huge effect. It's just the counterfeaces roughness, which matters.
Having a hardness difference doesn't mean a lot of applications do have a hardness
difference between the compassing surface. Gearing is a classic one, especially when
you have two gears of different sizes. You want the gear, because it sees many more
cycles than the larger gear. So in order to reduce contact fatigue on the smaller
gear, you harden it, a bit harder than the wheel, I suppose, the pinion version of
the wheel. The wheel is tend to be softer. Okay. In the current test we have,
you know, I suppose the SIG micro -pitin test, you know, do you think that's like
representative field failures? You know, one with the increase in load and the
duration endurance phase at the end, you know, that particular test? I suppose,
in a way, no, we don't often see in applications steady steps in load. But I can
appreciate what that test is trying to achieve. It's almost trying to achieve a
limiting conditions in which we expect to see micro -pitting being in the failure
modes. And increasing pressure really does massively increase the level of micro
-pitting. Of course, if you do it too much, I expect you would get pitting, but
then that is also material dependent as well. So I can see where we're aiming
towards it. It's the SIG test, a representative test. It's, the S &G test is an
accelerated micro -picking test. And to its credit, it does use real gearing,
unlike disk testing. So you are getting a variable slider ratio across the contact,
variable contact pressures. It is an expensive test. It is a long test, and gears
aren't normally designed like that, but then it is an accelerated micro -picking test.
So in a way, it has both benefits and drawbacks. And also, it's been a test been
around for a long time, so there's a lot of historical data on the F -ZG. So if
you've been doing a lot of F -ZG tests, and you have a lot of data on lots of
different lubricants, that's going to be very useful to design future lubricants on
that kind of testing, but does it represent what happens in real applications? Not
exactly. You know, similar to lots of these tests we work with in the lubricants
industry, isn't it? They've been around a long time. They may not be perfect, but
it may be the best thing we've currently got, you know, and you're right, there's a
lot of data on them, and they are, to be fair, they they are challenging to the
chemistries. You know, not everything passes. You know, you have to work hard to get
oils to look good in those tests. Personally, I prefer it if the load was constant.
I think that would make a big difference. You know, there's a few effects on there,
I think, aren't quite true. But anyway, the development's been done, and it was done
many, many years ago, and we're stuck with it. If you could design one from
scratch, you know, How would you do it? You know, so I put you in a tough
position, right? Your job is to qualify lubricants for use, I don't know,
in a wind turbine, something really expensive. And if you get it wrong, and you put
some oils up there and they break, it's your fault, okay? And someone else comes
along to you and says, right, I've got this new oil. It's brilliant at everything.
You've got to qualify that to make sure it's not going to cause micropitting in
your wind turbine or gearbox or whatever. What test would you like to run on it
and what? Yeah. So if I have an application involved, like you said, I would first
investigate, well, what are the contact parameters inside that application? So what
have I got? What was my surface roughness? I would calculate my lambda ratios, my
film thicknesses. I would look at the hardest of my steals, if possible try and
make specimens out of the same steals, I would be running at what I would consider
that application's harshest conditions. To try and make the damage happen an
appreciable amount of time, I don't want to go too harsh because I don't want, I
would still like my test to be as representative of that application as possible. I
would be using the same lubricant as what's currently failing and try and replicate
the damage, be checking my specimens are producing the correct damage that is similar
to what's being seen in the application. I think that's really important actually to
just to confirm that what you're producing in your micropitting test is actually the
same type of micro -pitting as what's being produced in your wind turbine. You need
to have that check. Okay. If I have time, I would section my specimens as well to
ensure that I am generating those microcracks. And then once I am confident that
that I'm producing that, then I would go on to change. If I'm constricted to only
be allowed to change the lubricant in this application, because the OEM has said,
you know, the gearbox is fine. We just want the lubricant change. You're the
lubricant manufacturer, for example. We just, you go and sort this problem out for
us. Then I would be testing my different lubricant formulations using that test
method that I devised. However, From what I've seen in industry, there seems to be
a little bit of a kind of knowledge gap between the OEMs and the lubricant
manufacturers, where lubricants don't know exactly the kind of conditions that the OEM
gearbox's transmissions are really under. And of course, that means that when it
comes to design of their lubricants, that they are in a way sometimes stuck with
these standard tests, unless they have a very good relationship or they manage to
acquire the transmission that they could test on themselves. You bring up a very
good point that, yeah, the ideal situation, I agree, is probably to design a test
with full knowledge of the mechanics of your system, your gearbox. And I think that
was a really nice thing to do. It got me thinking as well is, you know, I put
you in that position just on a thought there for fun, but it begs a question,
isn't it? You have these two people sat on either side of this test, which might
not be perfect communicating through these particular numbers, which might not suit
either of them, you know, because the OEM will want protection on their gearbox and
the lubricant manufacturer want to show the quality of their oil preventing it in
the field, but they still are communicating through maybe a test which doesn't match
to show that performance and to protect that asset. So like you said,
designing something perfect, exactly what you need is would benefit both. Obviously,
it'll take time and require clever people and things to do it. But it's probably
achievable, isn't it? I think there's another way of doing it. I think we could all
spend a lot of time is actually understanding the fundamental physics behind
micropitting. So we know it's a cyclic stress history -driven process,
it's a fatigue process. And if you know what your additives do to the surfaces
under a lot of different applications, you can say, well, actually, I know my
additive package will protect the roughness in this situation. Therefore, we shouldn't
be using this additive pack when the surfaces are so rough because that will lead
to a lot of micropitting. We should only be using an added pack in a smoother
environment, or I know that my friction modifiers will only work under this
condition, these conditions, et cetera. I think a lot of it comes down to what is
the underlying process driving the micro -pitting? And if we understand how to reduce
those factors, then we can design our lubricants in a certain way to benefit. And
if we know that under certain conditions, our additives behave in a certain way. We
don't need to go in, just because our additive package behaves, let's say, for
example, our additive package behaves very well under high contact loads, but poorly
under low contact loads. It might not perform very well. You know, micropate might
be exacerbated a lot under the lower contact conditions because our added package,
which is designed for higher. Just isn't working very well. So we don't, yeah. Okay.
We're coming to the end now. I'll finish with the thought from you. Is micropitting
a problem? We'll eventually engineer it out, or do we just need to continue to
manage it? What's your thoughts there? I don't think it'll be something that we ever
really truly prevent, but it's something that we, in the end, we'll just be able to
control and manage to a certain level that it won't be the main failure mode to
our machine elements and so that our machine elements can survive for their design
life. So we'll design machine elements, we'll understand how much micropitting is
acceptable in that design life. And then hopefully that's all we get during the
design life because we've designed it properly, a machine element lubricant and all.
I feel like most current steels are required for operation. They do fatigue over
time. So So if they don't fail by micropitting, they will fail by something else or
something else in the system will fail.
So we're always going to be able to tolerate some level of micropitting. It will be
there, but we'll control it in a certain way that it just won't be the limiting
factor anymore. Yeah, it won't cause like significant loss of material and then
pitting and then eventual failure of the tooth or something. And if you keep
improving, you know, we're back at the point where things are lasting an incredibly
long time. But if things last forever, then they're probably over -engineered. You're
probably doing it, just like the Victorians used to do. We need some failures to
keep us in doing something useful, don't we? So let's, let's, hopefully it continues
for a bit longer, Ben, for me and you to have something to do. I'm sure there'll
be still failures within our lifetimes mark. Exactly. Oh, well, thank you very much,
but anything you'd like to say before you kind of sign off. Yeah, thanks for having
me on here. It's been a pleasure to talk about micro -pitting. I've spent a lot of
time thinking about this over the past few years, so thank you. That's fantastic. I
appreciate you coming on and talking in depth to everyone. I really do, and you've
obviously prepped really well. Thank you. And giving some brilliant insights and being
an expert, it's been a great insight into the actual, I didn't realize how complex
It was, all these interplaying mechanisms for it, and it's a lot of food for
thought. So hopefully we'll continue the conversation at some point, and maybe we'll
test some of your theories. But you'll see you, Ben. You take care and see you
soon. Thank you. Cheers, Mark and Benjamin, we would like to thank you for taking
the time to talk with us and share your story that I'm sure others can relate to.
And thank you for your support of STLE. We appreciate you sharing your expertise
with us. We would love to hear your ideas for future episode and would love some
feedback in the comments with the link survey in the description. For those
interested in learning more about STLE, please visit our website at www .s .org. Thank
you for joining us, and remember to keep your viewers turning, keep your creativity
flowing, and keep your world in motion. Bye. Thank you for joining us on this
episode of Perfecting Motion, brought to you by the Society of Tribologists and
Lubrication Engineers, the Premier Technical Society, in the Tribology and Lubrication
Industry. STLE's mission is to advance the science of tribology and the practice of
lubrication engineering in order to foster innovation, improve performance of equipment
and products, conserve resources, and protect the environment. STLE supports its
members with a variety of technical, educational, and professional development resources
and programs. To learn more about STLE, please visit our website, www .stle .org.