Grease is the word; a complicated word...

General cycling advice ( NOT technical ! )
Brucey
Posts: 33204
Joined: 4 Jan 2012, 6:25pm

Grease is the word; a complicated word...

Postby Brucey » 11 Apr 2018, 1:22am

when I first saw a pot of grease I thought it was like Vaseline; just a substance that was inherently viscous. A while later I got a little proper schooling on the topic and was told that grease was unlike Vaseline, since grease was a mixture of mostly oil (~80-85%) together with a thickener of some kind (~15%) and other additives (0-5%).

My brief introduction to greases also stated that in most cases the oil was the dominant thing when in service. Most ball bearings (in industrial and automotive applications) operate using either 'elastohydrodynamic lubrication' (EHD) or even full hydrodynamic (HD) lubrication at higher speeds whereby a film of high pressure oil cannot escape from the oncoming ball, which runs over the film, and the metal parts actually don't make any contact under normal conditions.

In this story the thickener was said to be relatively unimportant; all it had to do was to stop the oil from escaping too quickly and not to thicken the grease at speed, when the EHD or HD mechanism prevailed.

Even simple thickeners in grease are clever; many fluids are said to be 'Newtonian', i.e. that the viscous drag forces are in direct proportion to the shear speed. But grease thickeners are an example of a non-Newtonian fluid, i.e. the viscous drag forces are not in proportion to the rate of shear. Some fluids are said to be 'dilatant' i.e. they thicken when sheared; a mixture of cornflour and water is like this. If the viscosity increases in time whilst being sheared (eg at constant rate), the fluid is said to be rheopectic.

Grease thickeners have the opposite property; the more they are sheared the less viscous (relatively) they become, hence they are said to be 'pseudoplastic'. Grease thickeners also have a 'memory' in that if they have been recently worked, they stay less viscous than they were initially, for a time. This is known as thixotropy. Other common substances exhibit both these properties; for example tomato ketchup is both pseudoplastic and thixotropic. [In practical terms this means that ketchup often won't come out of the bottle, but when it does it comes rather suddenly (because it is pseudoplastic). If you want it to come out more easily, just shake the bottle; it will remain more easy-flowing for a short time afterwards because it is thixotropic, and will come out of the bottle more easily for a minute or two afterwards.] You can read more about this kind of thing herehttps://en.wikipedia.org/wiki/Non-Newtonian_fluid#Types_of_non-Newtonian_behaviour. Be aware that the actual properties of real substances may not give a nice smooth curve when plotted on a graph; there may be all kinds of humps and bumps.

Greases are commonly characterised by producing what is known as a 'Stribeck curve'.
Image

The vertical axis is frictional drag. Commonly the horizontal axis is portrayed as being a single variable (speed, load, or viscosity) but strictly speaking it is a dimensionless quantity that includes all these variables; (speed x viscosity/load). If two of these variables are held constant (or at least not varied intentionally) the other can be used as the horizontal axis. So the right side of the graph represents high speeds, high viscosities, and low loads, and the left side of the graph is low viscosities, low speeds and high loads. The three zones are (from R to L) full HD lubrication, EHD (or mixed mode) lubrication, and what is known as 'boundary lubrication' in which there is contact between the two surfaces, such that it causes an appreciable amount of drag, and wear is likely to occur. In an ideal case the HD regime very closely resembles that of the oil alone, as if the thickener was not present at all.

Note that the practical effect of this is that the load-bearing capacity of a ball-bearing is very much dependent on the speed of the bearing; at speed the maximum load can be very much higher than at low speeds; a difference between the static load rating and the dynamic load rating (EHD or HD mode) of x3 or even x10 isn't uncommon.

Now, when you whack a grease into a bicycle bearing, it is either a special bicycle grease (that may not be that good in fact) or is more commonly a repurposed grease, i.e. it was really designed for some other application and it has been repackaged (at seemly vast cost... :roll: ) for cyclists to buy. 'Where's the harm in that' you might ask...? Well greases have to withstand all kinds of things other than just keeping the balls happy; they have to resist corrosion (including (road) salt water corrosion for extended periods), not to separate, react with water, it has to keep seal lips happy, not to react with metals and/or plastics or materials other than steel that might be used in the assembly, not to strip coatings (like paints) from other surfaces...the list goes on.

But there is one matter that is like the elephant in the room; repurposed greases are usually designed around a requirement for good EHD and HD performance, because that is the regime where most bearings operate. Now here's the thing; most bicycle bearings are moving so slowly that they don't ever operate in the HD regime! In fact they barely get into the EHD regime. Most of the duty cycle is in the boundary lubrication regime, which is (for most other applications) only important during start up or under exceptional shock loads.

In other words most repurposed greases are not likely to be anywhere near optimised for bicycle bearings; if they work at all it is seemingly by accident or because 'something is better than nothing'.

There are a load of additives that are used to bolster low speed performance including EP (extreme pressure) additives, anti-wear additives, and solid lubricants (like Teflon, hexagonal boron nitride, graphite and MoS2). But these are not added to many other greases in enough quantity to do much good in a low-speed bearing and in fact if the grease is rated for high speed (HD) operation it is virtually certain that the levels of certain solid lubricants are lower than they could be; the balls need to roll not scuff even at high speeds and solid lubricants tend to prevent that.

So for example some greases have MoS2 in and yet they are still rated for use in (high speed) car wheel bearings. These greases might have only ~0.5% MoS2 in them; this does some good but contrast this with the 50% MoS2 that is used in some assembly pastes; ideally a bicycle grease for (say) a headset might be expected to benefit from having 5-10% MoS2 in or something, but you try finding one....

So there are some greases that are designed for things like open gear trains in corrosive environments, which are vastly better than a standard (eg. car wheel bearing) grease when used in bicycle bearings. They usually have higher viscosity base oils, lots of solid lubricants and EP additives, and can have superior water and corrosion resistance. In some cases the difference is enough that you could take the same parts, set up in the same way, and they could last five or ten years without further attention whereas with an inferior lubricant they might be scrap within six months.

Disappointingly most greases that come inside bicycle hubs and other bearings used in bicycles fall into the 'inferior lubricant' category. In fact cartridge bearings are a case in point; these are usually manufactured with a ~50% fill of a grease that works well in the HD regime and doesn't have especially good corrosion resistance. Low speed bearings can have ~75% fill (last longer and less room for water to get in...) and the grease can be much better for the task in hand. Bearing manufacturers will make bearings with special grease fills, but at a cost; most hub manufacturers don't bother to spend the money. There are exceptions; Hope have NSK provide bearings to their specifications and Phil Wood take commoditised bearings and pack them with their own grease specially.

To be continued...more in the next post.....
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

Brucey
Posts: 33204
Joined: 4 Jan 2012, 6:25pm

Re: Grease is the word; a complicated word...

Postby Brucey » 11 Apr 2018, 1:27am

In the previous section I described how thickeners have very unusual properties and thus help the grease do its job, which is usually to work more or less like oil alone at speed, but also to keep the oil in place between times so that the bearing doesn't run dry. Also that (mainly because of the speeds of operation) there is certainly a mismatch if not an outright conflict between the properties required of (say) a car wheel bearing grease and a grease that is ideal for bicycle bearings.

In this section I aim to cover some of the issues surrounding the make-up and use of greases with different thickeners, together with some of the constraints (both practical and scientific) surrounding grease selection.

First a word on thickeners. They are often 'soaps' chemically speaking and they work in a slightly mysterious way. Under a microscope they generally appear to be a network of fibres. These fibres entangle with one another at rest (in a time-dependent way) but when the grease is sheared, they have the ability to (presumably straighen and align with one another then) slide past one another with much lower friction than you might expect.

There are many different thickeners that are used in commercial greases and whilst many of them are chemically similar this is not guaranteed; for example a 'calcium grease' might use one of several different chemistries and whilst there are terms such as 'calcium complex' that can indicate differences these are by no means universally applied in the same way. There are also mixtures of thickeners. The manufacturers tend to play smoke and mirrors to a great extent, and of course begrudge giving information away that may be of benefit to their competitors.

The thickener type has a big influence on many of the properties of interest in bicycle bearings such as water resistance, tackiness/stringiness, mobility, and separation. If you change greases or accidentally mix them, not all thickeners mix with one another in a benign fashion. This is primarily of interest to service engineers responsible for expensive equipment that has a pumped grease system in it; the wrong grease mix might put the equipment out of commission so they are often reluctant to change (even to a 'better' lubricant) unless it can be proven that it won't cause trouble. There are compatibility charts for mixing greases, but should be applied with the caveat that you might not know exactly what is in your grease to start with.

This is one such chart
Image

I cannot vouch for its accuracy and indeed you may find other charts with slightly different indications. One main cause of incompatibility is that the thickeners react with one another and the resultant mixture is too thin, or 'no longer grease-like'.

Even within thickener types there is plenty of scope for variation. For example this article (read only if you are very keen indeed)

http://coastlubricants.com/wp-content/uploads/2017/01/Lubrizol-WhitePaper-2016.pdf

concentrates on the manufacture and varying properties of greases made with just one thickener, calcium sulphonate. Each of the other thickeners has their own story which is likely to be equally convoluted. Note that quite small variations in manufacturing route can make substantial differences to the properties of the thickener even though it might still be 'the same type' as used in another grease. There is also the possibility of an interaction or synergy with the oil that is used in the grease, or indeed the additives. Quite a number of additives also thicken the grease, for good or ill. A number of chemical additives will chemically react with some thickeners but not others. Thus overall, cooking up a grease is at least as complicated as cordon bleu cookery, but with rather less flavoursome results.

When choosing a grease one only has a few sources of information; word of mouth, the manufacturer's blurb, the technical data sheet, and any safety data (MSD) you can get your hands on. None is complete and none is 100% reliable, since any can be guilty of the sin of omission at the very least. But if you can get access to all four then you have a reasonable chance of knowing how a grease will behave without a lot of trial and error.

The safety (MSD) data sheet is pretty much required by law for any lubricant used in industry and without it a risk assessment cannot be carried out, i.e. without it you might poison your staff or customers. It ought to state explicitly if there are any hazardous substances in the grease. The manufacturers tend to cover themselves with vague catch-alls like 'avoid prolonged skin exposure' but if they say 'avoid skin exposure' you can rest assured that there is something pretty horrible in the mix and it may/should say exactly what. There are separate standards for greases used in food processing equipment and (obviously) some of the most powerful (and potentially toxic) additives are not found in such greases.

The manufacturer's blurb is useful in that it often omits certain applications from the list of those suitable; it might be a 'niche grease' that sells for a premium (which doesn't mean that it is special necessarily; often the same stuff sells in differnt niche packages for different markets) otherwise you have to imagine that they usually wish to sell to as many different applications as possible with a single 'general purpose' grease and that in commoditised applications they will try to sell a grease that is versatile but that is only just good enough for any given application, and sells at the cheapest price.

The technical data sheet usually contains information on the base oil viscosity, as well as results from a variety of tests, either equipment specific or (more often) various standardised tests from ASTM and others. Interpretation of these test results is not easy; for example whilst a higher four-ball test score is a good sign, it mightn't relate to your application and might come at the expense of something else. In particular a lot of the load bearing tests are carried out using powerful equipment and this results in not only high local temperatures (at asperities etc) but high global temperatures during the test, which may alter the activity of certain chemical additives. Those tests that involve high temperatures don't usually relate that well to bicycle use; a bearing that is falling apart very rarely gets at all hot on a bicycle, except at asperities.

There is a useful primer on grease technical data sheets here

http://www.skf.com/group/products/lubrication-solutions/lubricants/understanding-grease-technical-data/index.html

One of the aspects of testing that is important but not entirely relevant to bicycle conditions is corrosion. Greases are tested for water-washoff and (in some cases) for resistance to corrosion in artificial sea water. It isn't clear to me that such tests are carried out whilst the bearings are operating in a boundary regime, which means that there may be little or no actual metal-to metal contact during the test and therefore some main driving forces for corrosion (being crevices and metal to metal contact) are likely to be absent in the test. It is also the case that whilst lubricant loss is obviously important, the greases appear not to be tested to see if they absorb moisture much or not. Given that the thickener is a soap, in many cases the soap will absorb water to some extent (which may be contaminated with salt) and this will eventually corrupt the grease in various ways, often by destabilising the soap so that it either thickens or thins. A 'pass' in the standard tests for washoff or salt water resistance is therefore encouraging but doesn't necessarily translate to useful corrosion resistance in bicycle bearings.

Looking at it from the other direction, it is as well to ask how bicycle bearings suffer when they do corrode. A good deal of corrosion damage can occur when the bike is not being ridden; the weird round marks that are often seen on balls, cups and cones are produced by corrosion cells that operate around each ball when the bike is standing. [This pretty much has to be the case, since the corrosion is often in places where the balls do not run on the raceways.].

If there is corrosion during operation, the result is usually bearing parts with fine pitting all over the contact surfaces; remember the bearing is usually operating in a boundary lubrication regime and whenever microscopic asperity contact is made the temperatures are locally very high for a fraction of a second; this makes the chemical activity (reaction rate) very much higher than normal and the result is usually microscopic pitting wherever parts make contact. The increased surface roughness (not to mention debris that is floating around in the system, which forms a 'third body' in tribological parlance) aggravates the situation and a mild case of grease contamination can turn into a failed bearing sooner than you might imagine. Note that when steel parts corrode, the corrosion products turn the conditions very much more acidic than normal and this both further accelerates the reaction rates and further destabilises the properties of the lubricant. It is a real double (if not triple) whammy.

[On a practical note, if you have a cup and cone bearing that has been damaged in this way, what can you do? Well my advice is that you have little to lose by cleaning out the worst of the crud, reassemble (using the same ball bearings, even though they will at best have a matte (i.e. microscopically rough) finish on them) repack with the most corrosion-resisting grease (with EP additives and solid lubricants) that you can find and then to run the bearing for a few weeks. If the bearing repidly goes a long way out of adjustment, it is wearing too fast and it is probably shot. If it goes very slightly out of adjustment this is to be seen as a reason to be optimistic; the chances are that when you inspect the bearing the surfaces will have improved (ie. will have become smoother) and if assembled with fresh ball bearings and more of the same grease, you might well be able to run the bearing for a much longer time without ill effect. BTW the cones can usually be polished (when mounted in an electric drill) using ~400-600 grit SiC abrasive paper, which helps too; the hard surface on the cone usually extends far enough that a little can be lost from the surface (by wear, corrosion, or polishing) without great ill-effect.]

Grease manufacturers don't go into details regarding anti-corrosion additives but any grease that is intended for an exposed service (look for a recommendation for use on fifth wheels and/or exposed gear trains, that sort of thing) will be thus fortified and for bicycle bearings there are (unless you intend to eat the grease) only plusses if such additives are present.

Some EP (extreme pressure) and anti-wear additives work by chemically reacting to the metal surfaces and inhibiting wear, eg via 'plucking' mechanisms (in which there is cold-welding at asperities and then chunks of the raceways/balls are pulled off). These additives often only work (react) at high temperatures; although bicycle bearings do not run hot the local temperatures are high enough at asperity contacts to allow such additives to do good rather than harm. Many wheel bearing greases use thinner base oils (so that they can run at high speeds) and use EP additives to bolster low-speed performance. So do many gear oils; in fact weight for weight, most Hypoid gear oils seem to have more EP additives than many greases.

Solid lubricants are more or less mandatory if you wish to have the lowest wear rates in the boundary regime. These are not used to the full extent possible in typical car wheel-bearing greases because they are expensive to add and they also have a thickening effect that is undesirable at high speed. The latter is of course of no consequence to a bicycle bearing, and for many bicycle applications I'd suggest that a grease with a high content of solid lubricants does no harm at all, and can only do good.

Note that some solid lubricants have such a powerful thickening effect that they may comprise the bulk or even all of the thickener in the grease; my understanding is that some PTFE greases work like this. It also raises the possibility that the thickener (even if it is not usually considered to be a lubricant in its own right) might usefully contribute to the lubrication in the low speed regime, which is explored further in part 3.

Solid lubricants often have interesting side-effects. For example if the solid lubricant is in the form of particles of ~10um in size, it may have the effect of usefully smoothing surface irregularities of about this size. It may also directly inhibit 'third body' action; 5um-sized wear particles don't get a chance to cold-weld (and therefore pluck) bearing surfaces if they are separated by solid lubricant particles which are bigger than this. Graphite is electrically conductive but PTFE is an insulator and MoS2 is more of a semiconductor, (with either a direct or indirect band gap over 1V depending on how it is measured). This means that MoS2 (and PTFE) has a corrosion-inhibiting effect, in that the balls are insulated from the raceways if there is MoS2 present and certain corrosion cells are less likely to develop.

Other additives to lubricants can include various things which can be loosely referred to as 'polymers'; these can change the rate at which the oil viscosity changes with temperature but more usefully some turn the lubricant more tacky or stringy which helps it cling to and coat the bearing parts, or provide an EP characteristic. Similar considerations apply to what happens in contact seals; without a film present at all times the seal cannot work properly. Some greases tend to 'ball up' in narrow gaps and this can let the water in directly. Having a little leakage outwards is no real problem (unless it gets on your brake discs I suppose) and helps prevent corrosion next to seals (which will corrupt the grease, given time). Needless to say if there are seals there is no harm in having a lubricant that is free to circulate within the hub (so a SFG or even an oil might work) and it should keep the seal lips wetted properly; very many modern bike parts seem to work best with a decent SFG inside them. However if there are no seals then a thicker grease (eg #2 ) is required and there should be more emphasis on corrosion resistance.

Now you can't mix additives and thickeners willy-nilly in greases but we are now shaping up to have a good idea of a durable bicycle bearing grease; It should contain

- a soap/thickener that resists water
- a high viscosity base oil (so we have more EHD running and less boundary lubrication)
- corrosion inhibitors (that resist salt water and resist acidification of the grease even if there is some corrosion)
- solid lubricants, more than is usual in a car wheel bearing grease.

Now I've used finish line PTFE grease for a number of years and it is more than OK. It contains a viscous fully synthetic base oil (1600 SUS which is ~350cSt I think) and the PTFE provides a good solid lubricant effect. Typically hubs which have been set up using this grease spin very freely and show very little signs of wear. However if I am to nitpick it doesn't contain enough anti-corrosion additives to resist bearings that are already corroding, and it doesn't wet seal lips as well as an SFG would, but it is OK. A 1lb tub costing about £15 will last most folks (with a few bicycles to maintain) five or ten years. It is clean to use and (AIUI) it is fairly non-toxic too. Similar arguments can be made for other 'bicycle specific greases', I am sure, but life is too short to try them all, and anyway in most cases you have (literally) no idea of what you are actually buying; at least Finish line publish some useful data....

However can we come up with anything better by looking at industrial lubricants?

I've come up with a few suggestions; base oil viscosity (in cSt) is in brackets


Shell Gadus S2 V220 AD grease (220 @ 40C); comment; a bit low on base oil viscosity but readily available and not too expensive. Useful levels of corrosion resistance and MoS2. Better than the average MoS2 wheel bearing grease.

SKF LGEM 2 (500); comment; like the above, and better with improved base oil viscosity but less easy to get hold of and fiercely expensive.

Castrol Molub-Alloy 777-2 ES (950). This is more like it; super base oil viscosity and lots of stinky additives that resist wear, corrosion and water ingress. No SFG variant available, not easy to find but not too expensive (per kg) when you do. [This grease is similar to one of my long-term favourites, Castrol SBX, which is probably even better in fact, but it seems SBX is no longer available in the UK]

Gear oil: Castrol Alphasyn EP (synthetic PAO oil used ) in 220-320-460-680 @ 40C viscosities.

There are of course many other possible candidates; this is just a (fairly random) selection.

Now the mysterious thing is that few of the above greases use synthetic oils, even though they are proven to have many desirable properties. Presumably this is for cost reasons. Castrol SBX does use synthetic oil and appears to be all the better for it. However the same manufacturers make and use synthetic oils (eg the Alphasyn) in other product lines. I have even considered adding thickeners and other additives to synthetic gear oil to make 'the right grease'.

Keeping in mind that bicycles have some unusual requirements, some of the above greases come pretty close to our proposed ideal; you could use any of them and you would get a very worthwhile improvement over (say) a typical wheel bearing grease like castrol LM. However it is very likely that the optimum grease for bicycle uses does not yet exist and will have to be made.

In the next part I will assess some special cases, and show how ongoing research may yet help us to come up with a 'perfect bicycle grease'.

cheers
Last edited by Brucey on 12 Apr 2018, 2:07pm, edited 1 time in total.
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

Brucey
Posts: 33204
Joined: 4 Jan 2012, 6:25pm

Re: Grease is the word; a complicated word...

Postby Brucey » 11 Apr 2018, 1:27am

In the previous part many of the factors surrounding grease make-up and grease selection were outlined. In this part some special cases are examined and there is also a glimpse into what might be possible in the future.

The subject of seals was touched upon previously; essentially most contact seals cannot work and indeed will wear rapidly if there is no lubricant film present at the seal lip. However if the grease inside (say) the hub is a #2 grease it will have limited mobility inside a bicycle hub and the lip film will not easily reform should it be disrupted. You can help avoid this situation by regularly using a little spray lube on the outside of the seals, after each ride (don't spray then ride; the solvent in the spray lube will still be there, and the seal will likely rub and wear). Spray lube also helps displace water from the area. However if you use a more fluid lubricant inside the assembly, the seal lips can be wetted at all times and the service life of all the parts can be greatly extended. If you choose a lubricant that is as fluid as possible, consistent with it not escaping from the assembly too quickly, this is usually best for the seal lips. Some hubs are well enough sealed that they can be entirely lubricated with gear oil, but for most a SFG is a very good choice. Either will work well inside most freehub bodies.

Another special case is cables. The characteristics that make for a good grease elsewhere (e.g. pseudoplasticity and thixotropy) may result in a cable that is 'sticky' when it is first moved, or subject to small forces. Again a much thinner lubricant (consistent with it staying inside the cable) is usually a good solution. However it may be that with certain thickener systems there is less sticking friction; this is not usually measured or decribed in data sheets so it may be a matter of trial and error. Note also that the polymer liners and coatings used in cables can be attacked by certain oils; they may soften or swell, and either can cause a cable to go draggy. Unfortunately there are so many possible combinations of materials that it is not possible to say what greases will definitely work in what cables without trying it. Once you have found a good combination, it isn't a bad idea to stick with it. I commonly use a little FL PTFE grease together with more synthetic oil to make a runnier cable lube inside standard shimano outers and it seems to work OK, but it is not the only way that works by any means.

Going full circle, it has been suggested that Vaseline might be a good lubricant for bicycle bearings. However it will be very different from other lubes; it doesn't shear-thin appreciably so the viscous drag will always be high at ambient temperatures. It is extremely water-repellent and sticks to surfaces well, and its sheer viscosity may allow better running in the EHD regime. It is relatively chemically inert so it may not be easily corrupted even in harsh conditions. So whilst it lacks some of the features in an ideal grease (such as corrosion inhibitors and an ability to re-wet seals) it might work OK in some bicycle bearings.

So, what of the future? Until very recently my understanding of the way most oils and greases work could be neatly captured in this chart;

Image

There are three plots; the uppermost one is essentially the stribeck curve (as per part one). The lowest one is the wear rate, which (as you might expect) follows the stribeck curve shape too. The middle plot is of lubricant film thickness. As you might expect at high loads/low speeds the film thickness becomes smaller and that explains the increase in both friction and wear rate. That most conventional lubricants behaved similarly to this was more or less received wisdom, and you can see the 'ideal' aiming point (lowest drag and wear etc) for a single set of service conditions marked in each plot, thought to be virtually unattainable in any bicycle bearing.

However more recently I have become aware of results that are different to this. For example this research (funded by SKF)

https://ac.els-cdn.com/S0301679X15004612/1-s2.0-S0301679X15004612-main.pdf?_tid=a5a707d8-0d65-4358-913d-992bf614eb54&acdnat=1523358473_5e2f26a5c96ffa215188bcb75777ba06

appears to demonstrate that

a) there are lubricants that develop increased film thicknesses at low speeds (vs the base oil only) presumably as a result of the particular characteristics of the thickener and

b) that the frictional drag of the grease in the same regime is actually lower than it would be in the base oil alone, too.

The second result is particularly surprising. Presumably these things mean a low wear rate too, although I think that has yet to be demonstrated. It opens up the possibility that, at low speeds, you could have grease that lubricated very well, gave low drag, and provided very good protection for the parts, quite unlike the usual 'boundary regime' result. Maybe there are some commercial greases that already work like this and we simply don't know it, but this characteristic is thought to be a result of particular combinations of base oil and thickener only.

I think the thrust of the work (as usual) is towards a grease that gives low drag and good lubrication in the HD range (eg through use of a low viscosity base oil), whilst being bolstered in some way enough to be 'adequate' at low speeds but this line of research shows great promise for development of a really good lubricant for low-speed bearings too.

I think that is about all, folks....

cheers
Last edited by Brucey on 13 Apr 2018, 11:19am, edited 1 time in total.
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

rjb
Posts: 3144
Joined: 11 Jan 2007, 10:25am
Location: Somerset (originally 60/70's Plymouth)

Re: Grease is the word; a complicated word...

Postby rjb » 11 Apr 2018, 8:03am

Nice one Brucey, early breakfast viewing, beats ketchup TV :lol:
I should have paid more attention to my tribology lectures, but you have explained it nicely. :D
At the last count:- Focus Variado, Peugeot 531 pro, Dawes Discovery Tandem, 2 Dawes Kingpins, Raleigh 20, Falcon K2 MTB dropped bar tourer, On One Pompino, Longstaff trike conversion on a Falcon corsa. :D

User avatar
iow
Posts: 388
Joined: 27 Jul 2010, 11:01am
Location: isle of wight

Re: Grease is the word; a complicated word...

Postby iow » 11 Apr 2018, 8:36am

thanks - look forward to the next instalment...
mark

Samuel D
Posts: 2615
Joined: 8 Mar 2015, 11:05pm
Location: Paris

Re: Grease is the word; a complicated word...

Postby Samuel D » 11 Apr 2018, 9:00am

Should not the full HD regime start a little farther to the right than the vertical line on the graph indicates? Otherwise I’d be curious to know why friction initially decreases with increasing viscosity or velocity after the change from EHD to HD.

For that matter, the distinction between the EHD and HD mechanisms is not clear to me. Is it something to do with surface asperities breaking through the lubricant film in the EHD case but not the HD case?

If the horizontal axis is linear and starts at zero and bicycle bearings “barely get into the EHD regime”, it seems to me that increasing the dimensions of the bearing would be tremendously useful. Even a small size increase would have a strong benefit because it would simultaneously reduce load and increase speed, pushing operation to the right of the curve.

It intrigues me therefore that things like bottom brackets appear to have moved from large quarter-inch balls to smaller ones and are now going smaller still. For example, the Shimano Dura-Ace SM-BB9000 and SM-BB9100 have 19 small balls whereas previous Shimano external bottom brackets had 15 slightly bigger ones (if I understand correctly). Maybe this is an admission that there’s no hope of getting out of the boundary lubrication regime in this instance.

I’ve encountered some of these concepts before but never seen the Stribeck curve that so neatly wraps it up. Cool graph.

One thing I wonder about is how viscosity (of the base fluid) is measured and described, and what sort of range we’re looking at across commonly available greases. In other words, to what extent we can hope to push bicycle bearings into the next lubrication regime by seeking out high-viscosity lubricants.

Brucey
Posts: 33204
Joined: 4 Jan 2012, 6:25pm

Re: Grease is the word; a complicated word...

Postby Brucey » 11 Apr 2018, 11:13am

Samuel D wrote:Should not the full HD regime start a little farther to the right than the vertical line on the graph indicates? Otherwise I’d be curious to know why friction initially decreases with increasing viscosity or velocity after the change from EHD to HD.


the breaks between the regimes are not very clearly defined; I agree that they could be moved around a bit. EHD can be regarded as a 'mixed' mode, containing elements approximating to both HD and boundary modes. In the second and third instalments I have some interesting things that show more clearly the (often surprising) effect of the thickener on the properties of the grease; in both boundary and EHD regimes the frictional drag is partly comprised of real friction (arising from metal to metal contact) and partly from shearing in the lubricant; this does not always play out as you might expect. There is also the matter of what happens when the potted curve intercepts the Y axis; this is important in some applications but not in others

For that matter, the distinction between the EHD and HD mechanisms is not clear to me. Is it something to do with surface asperities breaking through the lubricant film in the EHD case but not the HD case?


pretty much. IIRC the idea is that the deformations in the parts are primarily elastic, and therefore wear rates can be low in this regime.

If the horizontal axis is linear and starts at zero and bicycle bearings “barely get into the EHD regime”, it seems to me that increasing the dimensions of the bearing would be tremendously useful. Even a small size increase would have a strong benefit because it would simultaneously reduce load and increase speed, pushing operation to the right of the curve.

It intrigues me therefore that things like bottom brackets appear to have moved from large quarter-inch balls to smaller ones and are now going smaller still. For example, the Shimano Dura-Ace SM-BB9000 and SM-BB9100 have 19 small balls whereas previous Shimano external bottom brackets had 15 slightly bigger ones (if I understand correctly). Maybe this is an admission that there’s no hope of getting out of the boundary lubrication regime in this instance.


traditional cycle parts have evolved by trial and error but you could argue that they might as well have been designed around the idea that the parts will usually still be sufficiently durable even if they are not very accurately made. Thus (to a first approximation) if the front hub as 3/16" balls, the headset and pedals have 5/32" balls, the BB and rear hub have 1/4" balls, the service life will usually be OK provided there is some lubrication and the preload is not too high. Although the parts operate in the boundary regime the wear rates can be low enough that (if anything) the bearing surfaces might get smoother as time goes on. Provided the wear debris is flushed away (or otherwise accommodated) and the contact loads are not so high that you get subsurface fatigue (which is usually caused by excessive preload and is usually what causes bearing surfaces to break up) then there will usually be a good wear life. With traditional parts the balls are sized so that normal service loads can be withstood using a small fraction of the balls that could share the load (in a perfectly manufactured and adjusted bearing) but in designs that employ cartridge bearings (and others with small ball bearings in) there is a greater need for load sharing and a greater need for good tolerances. In bearings where load sharing is required, any small problem with setup or lubrication will result in a very much reduced service life.

I’ve encountered some of these concepts before but never seen the Stribeck curve that so neatly wraps it up. Cool graph.

One thing I wonder about is how viscosity (of the base fluid) is measured and described, and what sort of range we’re looking at across commonly available greases. In other words, to what extent we can hope to push bicycle bearings into the next lubrication regime by seeking out high-viscosity lubricants.


That is a good point. Base oils in a typical wheel bearing grease (eg Castrol LM) are often in the region of 100cSt (or less at running temperatures) but there are greases with base oils that are nearer 1000cSt. This can only help in most bicycle bearings, provided the drag does not become too high. In high speed bearings there is a strong motivation for reducing the base oil viscosity so that bearing drag is reduced in the HD regime. By contrast at low speeds the viscous drag may be affected by the thickener as much as the oil, and not always in the way you might expect.

Because (I think) a lot of deep groove bearings are 'scale models' of one another (i.e. the size of the balls scales to the diameter of the bearing) and the bearings are selected to withstand static and dynamic loads imposed that also scale with bearing size, it surprisingly often works out that the shaft speed that corresponds with the transition into the EHD/HD regime is between about 500 and 1000rpm. If you install a bigger (heavier, more expensive, potentially draggier since there will usually be more rolling elements) bearing and/or employ highly viscous lubricants, you can lower this speed and you might get a bicycle bearing into the EHD regime more of the time.

But.... the peak loads occur at the lowest speeds (starting off, climbing) and almost nothing you can do will save that from being a boundary lubrication regime. In the highly loaded condition might well be where most of the wear occurs and it may be a case of coping with that as best as possible rather than stopping it altogether.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

pliptrot
Posts: 530
Joined: 12 Jan 2007, 2:50am

Re: Grease is the word; a complicated word...

Postby pliptrot » 11 Apr 2018, 6:12pm

As always Brucey, a superb and informative piece. The danger is that we will start making requests...........

philvantwo
Posts: 367
Joined: 8 Dec 2012, 6:08pm

Re: Grease is the word; a complicated word...

Postby philvantwo » 12 Apr 2018, 8:20am

And you never mentioned John Travolta once!! :lol:
[color=#FF0000][/color]

Brucey
Posts: 33204
Joined: 4 Jan 2012, 6:25pm

Re: Grease is the word; a complicated word...

Postby Brucey » 12 Apr 2018, 2:09pm

second part now posted...
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

David9694
Posts: 208
Joined: 10 Feb 2018, 8:42am

Re: Grease is the word; a complicated word...

Postby David9694 » 12 Apr 2018, 9:49pm

We take the pressure, and we throw away conventionality, belongs to yesterday
There is a chance that we can make it so far
We start believin' now that we can be who we are, grease is the word


Which brand of grease is right for me?

Brucey
Posts: 33204
Joined: 4 Jan 2012, 6:25pm

Re: Grease is the word; a complicated word...

Postby Brucey » 13 Apr 2018, 11:18am

er, Brylcreem?

-something different for the bike, obviously.... :wink:

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

Brucey
Posts: 33204
Joined: 4 Jan 2012, 6:25pm

Re: Grease is the word; a complicated word...

Postby Brucey » 13 Apr 2018, 11:23am

part three now posted above.
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

User avatar
Sweep
Posts: 4829
Joined: 20 Oct 2011, 4:57pm
Location: London

Re: Grease is the word; a complicated word...

Postby Sweep » 13 Apr 2018, 12:22pm

Come come brucey :) Simple receommendations please for us non techies.

A bearing grease and a grease for more general bike lubing - ie: greasing bolt threads.
Sweep

MikeF
Posts: 3690
Joined: 11 Nov 2012, 9:24am
Location: On the borders of the four South East Counties

Re: Grease is the word; a complicated word...

Postby MikeF » 13 Apr 2018, 9:54pm

All very interesting.
I frequently use Duckhams Keenol grease. I don't where it would appear on the charts. It has zinc oxide additive and is recommended for chassis lubrication, which I presume (maybe wrongly) is low speed lubrication, and marine applications. It's not now made though. The grease now sold as Keenol does not have the same appearance as Duckhams, so I think it must be different.
"It takes a genius to spot the obvious" - my old physics master