Providing a low-friction bearing capable of operating at the very high speeds a turbocharger sees – more than 150,000 rpm – and which can also survive the high temperatures experienced within a turbine housing is no easy task. For many years, the most common solution was a simple bronze-bushed affair, with an oil supply and in some cases a water cooling jacket. In high-performance applications, this approach has given way to more efficient ball bearing systems, which generate less friction and therefore help to increase turbine efficiency, allowing for faster spooling and thus reduced turbo lag.

Ball bearing designs also reduce the amount of oil required to provide adequate lubrication compared to journal bearings, and this lower oil volume reduces the risk of seal leakage. Ball bearings are also more tolerant of marginal lubrication conditions, reducing the possibility of failure.

While such systems are undoubtedly more efficient than a simple bush, there is still a lot of scope for improvement. The answer to achieving greater durability and lower friction has been the introduction of ceramic-based bearing elements, using silicon nitride instead of steel to provide a number of advantages over regular steel bearings and races. Ceramic bearings are generally available in one of two types – pure ceramic, where all the elements of the bearing are made from ceramic material; and hybrid, where ceramic balls or rollers run in steel rings.

The advantages of ceramic bearings are considerable. Ceramic bearing balls can weigh up to 40% less than steel ones, depending on their size and the materials used, and this can reduce centrifugal loading and skidding, allowing hybrid ceramic bearings to operate 20-40% faster than conventional steel bearings. As a result, ceramic and hybrid ceramic bearings have less inertia, and thus in turbo applications can increase response rate over steel bearings. Ceramics also need far less lubrication than steel, further reducing the oil supply requirements over both ball bearing and journal bearing systems.

The thermal characteristics of ceramic balls are also very attractive, as they experience minimal expansion even at very high temperatures. To give an example that quantifies the performance possible with such bearings, one turbo manufacturer worked with its bearings supplier on producing a hybrid ceramic ball bearing turbo unit that saw peak speeds of more than 220,000 rpm, with a shaft temperature in excess of 300 C – conditions far beyond those in which a traditional steel bearing could survive.

From this it is clear that ceramic technology provides the best current solution for high-efficiency turbocharger bearings. There are other potential development avenues when it comes to further reducing friction in turbo applications, but currently they exist only in theory or in very large machinery applications. For example, the use of foil-air or magnetic bearings offers the prospect of almost completely eliminating friction, but a practical application of these technologies in an automotive turbocharger has yet to be realised.

Written by Lawrence Butcher

Although in the past Formula One engines have used rolling element bearings for applications such as the main and rod bearings, in recent times plain journal-type bearings have been the favoured solution. One area that Honda focused on with its bearing developments was the con rod big-end bearings. In physical terms, these were similar in size to those one would expect to find on a mass-production engine of 600 cc or so; however, the loads they experience are substantially greater. According to data from the bearing development programme, Honda showed that a standard production bearing can expect to experience loads in the region of 20 kN vertically and 5 kN horizontally. By contrast, the Formula One rod bearing saw loads of 50 kN and 20 kN respectively.

As engine speeds rose throughout the first decade of the 21st century, and regulations demanded longer service lives from power units, Honda began to find that its existing bearing solutions were not up to the task, and occurrences of bearing damage or failure started to become common. This led the team to investigate new bearing materials and designs. The solution was a move to a silicon-bronze bearing material, instead of the copper-steel bi-metal bearing type it had been using.

The existing bearings were of quite traditional construction, with a steel backing supporting a copper-lead intermediate layer that was topped with a thin coating of an alloy containing lead, copper and silicon. This bearing make-up provided low friction, but could not withstand the high contact pressures of the Formula One application and could not effectively dissipate the high levels of heat built up under running conditions.

The silicon-bronze bearing, however, was constructed without a steel backing, consisting instead simply of a silicon-bronze lining coated with a thin layer of lead-indium. While having marginally lower ‘slideability’ and a lower level of heat dissipation than the pure copper material, it provided a bearing shell that was stronger in all respects than the steel-copper item. Also, the heat transfer rate of the whole bearing assembly was higher than the copper-steel item, as the steel – which has relatively poor thermal conductivity – was removed from the equation. 

The results of using the new material were impressive: its greater strength and heat dissipation characteristics allowed lower-viscosity oil to be used, reducing windage losses. More important, bearing failures became rare, allowing the engines to exceed the regulation-imposed working life of 1500 km. 

Written by Lawrence Butcher

Increased bearing load results in higher bearing surface temperatures which, after all the analysis has been undertaken and the tests completed, results in thinner, less viscous oils having even thinner oil films at the higher temperatures involved. The inevitable conclusion to all this for a fixed flow rate of oil is even smaller journal-to-bearing diametrical clearances.

As these clearances diminish, however, the shape and topological nature of the bearing becomes more important, making them more susceptible to fatigue or failure due to seizure, particularly in the initial stages of running where conformability of the bearing is at its most critical. The trick therefore is to make a bearing with reduced running clearance, which will bed-in quickly to the dynamic requirements of the rotating journal.

Under a (very powerful) magnifying glass the surface of a bearing looks like a series of peaks and valleys. These peaks are often referred to as asperities, and they are the key when it comes to reducing friction. A plain bearing works most efficiently in what engineers refer to as the beginning of the ‘hydrodynamic’ part of the lubrication regime, which is when the relative movement of two surfaces separated by the lubricant generates a pressure to force those surfaces apart. In doing so, the opposing surface asperities are forced to disengage, allowing the surfaces to glide across each other with the minimum of drag.

‘Mixed’ or ‘boundary’ regimes – when these asperities are in partial contact – may occur when the oil flow is reduced, at high oil temperatures or excessive loading, and at these times the bearing is particularly vulnerable to failure. It is therefore at these times during its early life, before the assembly has had time to fully conform, that the risk of failure is at its greatest.

Clearly any plain bearing design therefore needs to have a surface finish that holds the oil and retains it as long as possible. At the same time, the surface must have no barriers to the movement of the oil in the direction of rotation, while effectively discouraging it from escaping out of the sides of the bearing.

Such a surface can be obtained using a method of boring that produces a surface roughness parallel to the movement of the journal with a cross-section such as that shown in Fig. 1. This retains the oil in the grooves, minimising the flow escaping to the side and with the reduced supporting area quickly conforming to the dynamic movement of the system, creating a uniform polished bearing and giving maximum support in the minimum of time. Bearings of this nature may appear to wear quickly after initial start-up, but taken over the life of the power unit they have a lower risk of seizure and greater durability – and, ultimately, a higher loading capacity.

Bearing wear was rarely a problem with older engines, but the use of lower-viscosity oils in more modern units is creating a new set of challenges.

Fig. 1 - Microprofiled bearing

Written by John Coxon

In Al-Sn-Si bearings the tin content, which can be as much as 40%, acts as the soft component distributed within the aluminium matrix. As a separate phase this produces a diffuse network along the edges of the grain boundaries of the aluminium to give the bearing surface its low frictional properties under boundary lubrication conditions. On the other hand, the high-hardness silicon is distributed across the aluminium matrix and serves to polish the mating journal while retaining the lubricant on the bearing surface to support hydrodynamic lubrication. 

Aluminium-tin-silicon bi-metal bearings consist of two layers bonded onto a steel backing plate. The intermediate layer, which consists of pure aluminium or an aluminium-manganese alloy, is bonded to the steel backing strip, after which the Al-Sn-Si is roll-bonded on top. The actual process might vary from one manufacturer to another, but the intermediate layer is there to help adhesion, stabilising the bonding between the outer layer and the steel shell, and also helps to serve as a cushion should edge loading be applied to the bearing in service.

Once bonded to the steel backing strip, bearings are formed into a semicircular shape and processed to the required bearing size In this state, the bearings are soft compared to copper-lead high-performance derivatives and so some method of hardening has to be introduced in order to improve their resistance to fatigue. The most common method is solid solution treatment, whereby the bearings are heated to 400-550 C and then rapidly cooled. Unfortunately, while this method increases the strength of the outer bearing layer it decreases the bonding strength of the intermediate layer, creating a brittle intermetallic layer next to the steel backing. The formation of this brittle aluminium-iron (Al-Fe) compound, although dependent on temperature, is also a function of the chemical composition of the aluminium alloy next to the steel.

This was an issue in early stages of development but was ultimately resolved by replacing the aluminium or aluminium-manganese intermediate layer by one consisting of aluminium and around 7-8% silicon. Instead of forming FeAl₃, this aluminium-silicon material forms an Al-Fe-Si compound, at a much higher temperature than that used during heat treatment of the outer bearing material. The presence of the silicon would appear to suppress the formation of the brittle FeAl₃ compound, producing a bearing with a much higher resistance to fatigue, but perhaps even more crucially the higher temperatures allowed would prove beneficial in optimising the composition of the outer bearing material to create even greater strength.

Just like the human body, once the problem of fatigue is overcome then the person can go from strength to strength.

Fig. 1 - Aluminium bearings after testing

Written by John Coxon

However, 70 or so years later the link was made between the toxicity of airborne lead and the adverse health effects on children, so within a decade the presence of four-star leaded fuels in the West was largely eradicated. Once removed from the urban air, as part of ‘end of vehicle life’ legislation, the authorities in Europe have now removed it from virtually all areas of road-based automotive applications. So while lead-free bi-metal bearings are now common and working satisfactorily, for high-performance tri-metal units – especially in highly loaded diesel engines – another approach was needed. Cue the development of a new type of bearing overlayer.

Unfortunately lead has some very desirable properties when it comes to bearing materials. It is soft and has a comparatively low melting point, and these characteristics are essential in high-performance engines where peak bearing loading can be very high. But as well as these properties, lead has what is referred to as ‘lipophilic’ properties, in that it has an affinity for some types of oils, lowering the surface tension and making the bearing surface more ‘wettable’. All this – and of course its ability to absorb heat during the melting process – makes adhesive wear at the boundary with the crankshaft journal highly unlikely.

Combining lead with indium or copper (to add strength), however, made it possible to accept high peak loads; the loads would be reacted by the presence of the oil film, but if this broke down then any friction generated would melt the lead, taking heat away from the area, and the molten lead would be smeared to a position elsewhere in the bearing, giving up its heat back to the lubricant flowing past and re-solidifying. The result would be a redistribution of the load in the bearing, a situation often referred to as ‘bedding-in’. Under such conditions the presence of lead was ideal, for not only did it make the bearing shape eventually conform to that required, it also made seizure of the journal most unlikely, even under the most adverse conditions.

As a conclusion to a study for this lead replacement therefore, materials were evaluated for strength, melting point, thermal conductivity and adhesion to crank journal surfaces, so the tri-metal application using bismuth and silver was suggested. With a low melting point (271 C), pure bismuth was favoured as the top surface in contact with the journal, while beneath it and next to the base layer (a copper-tin based product) was the silver, chosen primarily for its high thermal conductivity.

Both layers were around 5 microns thick, giving a maximum thickness of the overlayer of 10 microns. At this thickness, optimum fatigue resistance was achieved with, at the same time, adequate ‘embeddability’ against the risk of dirt or foreign matter in the oil. Furthermore, by altering the crystal orientation of the bismuth and electroplating it into pyramid-type shapes, bearing surface wettability was improved, minimising the chances of seizure.

Including silver as a separate layer was in my opinion a masterstroke. As a distinct layer, and being easily picked up by ICP (inductively coupled plasma) oil analysis equipment, the presence of silver in the oil sample can indicate a certain degree of bearing wear long before the situation becomes critical.

It’s not often you get such warning signs of impending disaster in an engine. 

Fig. 1 - The bismuth-silver overlayer

Written by John Coxon

However, the report went on to explain that second in the ranking of bearing failure, jointly with faulty or incorrect bearing assembly (another manufacturing issue), came “insufficient or unsatisfactory lubrication”. Described as the point at which the continuous oil film between the surface of the bearing and the journal is interrupted, the result is some kind of surface damage – polishing, wiping or smearing and scuffing, leading eventually to the engine’s demise. However, unlike the ingress of foreign matter, the lack of a continuous oil film can mean many things.

In the engine build process for instance, bearings have to be pre-lubricated with oils to ensure the existence of the oil film right from the moment the engine is cranked for the first time. There may be engine oil in the sump, but in the time taken to pump this oil through the system to the bearings for the first time the engine crankshaft may have revolved many times. Without any lubricant at this critical time, serious bearing damage will take place. During engine build therefore it is common to use special ‘build’ oils with a viscosity generally far greater than that usually specified for normal engine running. This extra viscosity ensures that the oil doesn’t drain away in the time between assembly and first fire, but also during the build process it can help centralise the bearing cap upon assembly – the other joint second issue highlighted in the report.

In the special case of the bearings, this oil should introduce no new additive technology to the engine other than that destined to be used in the engine sump. Since once the engine has started to rotate, the lubrication regime will be totally hydrodynamic, no additional oil additive technology is strictly necessary – save perhaps for anti-oxidants and possibly a corrosion inhibitor. Detergents, dispersants, friction modifiers and anti-wear additives are not really necessary, unless of course you want to use the same oil to lubricate the camshaft. In this case, certain aggressive camshaft designs may benefit from additional anti-wear technology to prevent scuffing during this initial cranking, so the best advice I can give is to use a product recommended for engine build by your engine sump oil supplier. This will most probably do the job and be compatible with the engine oil you eventually intend to run.

Once the engine is running, you should have no further issues with the engine bearings, provided of course you use the correct grade of oil specified by your engine designer and change it regularly. The engine manufacturer will spend a lot of time selecting the appropriate grade of oil to make sure you get that little bit of extra performance and yet protect the bearings for all eventuality, but that is a story for another time.

Fig. 1 - Big-end bearing showing signs, among other things, of foreign body contamination

Written by John Coxon

Typically a cylinder block machining line may machine, say, 100 blocks in an eight-hour shift. During this time the block would be line-bored to, say, 80.00 mm diameter and then fine-honed to the final dimensions of 80.014-80.032 mm. The limits to manufacture therefore would be stated as being 80.014 mm at the maximum metal condition and 80.032 mm at its largest or minimum metal condition, a difference of 18 microns.  The crankshaft line, however, making roughly the same number of crankshafts in the same time period, may be grinding the main bearing journals to a tolerance close to ±0.010 mm (10 microns). Here the journal size, once ground and polished, would be, say, 75.000 mm at its maximum size and 74.980 mm at its smallest. Given that we can machine these components to such accuracies, it still requires a system of selective assembly to build engines to the correct running clearances. In these days of mass production, where parts are purposely designed to tolerate large variations in finished dimensions, this is an expensive but nevertheless the only realistic option.

As part of the design method and the limitations placed on them by the chosen supplier and/or legislation, engine designers will have selected the type of bearing for the application. Durability trials will establish the limits of the bearing-to-journal clearance, which has to allow sufficient oil to flow to cool the surfaces but at the same time generate enough pressure to keep the surfaces apart throughout the full duration of the sign-off tests and hence proposed engine life. Using low-viscosity engine lubricants, and with clearances strictly controlled to a matter of plus or minus a few microns, the bearing selection procedure still has to be simple.

The process starts out at the manufacturing stage. For the cylinder block this means that after the line boring and honing operation, each main bearing will be gauged to determine its size. Since our total machining tolerance is of the order of ±0.018 mm, the possible bearing shell outside diameters will be split into two or three different bands, each with a band width of around 0.0040 mm and each one represented by a code. This data will then be transferred to side of the block to be read either manually or by machine at the assembly stage.

Likewise, the crankshaft journal diameter will have been also measured and the sizes coded, and again this data will have been etched in some way on the crankshaft, usually on the machined face at one end.

To select the precise main bearings it must be remembered that although visually they are the same, the upper and lower main bearings will have different part numbers, with each bearing colour-coded with up to four or five different colours – typically blue, green, yellow, brown and so on – reflecting their exact size. Selecting the correct bearing using the appropriate colour(s) from a matrix that lists main bearing bore size against the exact journal diameter is therefore simple, remembering of course that where differing colours are demanded, these can be fitted in either upper or lower bearing  positions.

It may sound complex but believe me, it is much easier to do than explain.Fig. 1 - Typical bearing selection matrix 

Written by John Coxon

But with bearing shapes nowhere near circular and clearances between shell and journal measured in microns these days, how do we determine whether a bearing is up to the task throughout its lifetime?

One way during development is to repeatedly strip, examine and rebuild the engine bottom end, looking for markings on the bearings that would indicate wear and perhaps something going wrong. This is obviously very time-consuming, not to say expensive in terms of man-hours, but visual inspection is probably still the best method. Dismantling the engine also serves as a check on the assembly process and the 1001 other little things that can go wrong at this stage.

Another way to check for bearing wear is by using a technique that exploits radioactivity. It is very good at understanding the engine conditions under which wear is generated at a specific place or location in the bearing, and has been developed in the past to understand the limits to viscosity in a particular application. At the time, and before the general ban on lead in automotive applications, the lead overlay in a bearing would be irradiated to produce the radioactive isotope bismuth²⁰⁶, and bearing wear would be assumed by detecting the build-up of radioactivity in the oil filter or engine oil. The method doesn’t take into account the shape of the bearing surface during use or the smearing effect of the soft overlay across the bearing, but with a measurement resolution of nanometres per hour, the technique was certainly very sensitive (if a bit expensive). 

The easiest way to screen engines for bearing wear, however, is to analyse the elements of the lubricating oil. For little more than $15-20 (£10-15) a small 100 ml sample of oil can be analysed using a spectrometer of one form or another. In the laboratory this could be of the inductively coupled plasma type or, at the track, well-heeled teams might use rotating disc electrode systems, which will typically screen any oil sample for the 20 or so elements commonly used in engines or engine oils to levels of parts per million.

Understanding the type of bearing design, and after reviewing the other likely metals that will be sources of wear, one can pick out those materials attributed to bearing wear. The presence of copper, for instance, could mean con rod little-end bearing wear; it could also mean big-end or main bearing wear when using copper-based products. Aluminium, however, is a difficult one. It could be attributed to piston ring top land erosion of contact between skirt and cylinder bore, or it could be down to the aluminium-tin bearing material.

Either way, it might be prudent to strip the engine for further examination or to look for the presence of tin in the oil sample. One engine I know uses bismuth at 99% purity as an overlay that was flashed onto an aluminium-tin bearing using a very thin layer of pure silver to assist adhesion. To screen for bearing wear I looked for small amounts of bismuth in the oil, but when silver started to appear during the test I knew that the bismuth overlay was being breached. Not necessarily a bad thing though. 

But for only a few dollars (or pounds) it often amazes me why race teams or enthusiasts don’t analyse oil more regularly.

Written by John Coxon

But I guess times have changed, and although some time ago many things may have not been seen relevant to mass production, today the unending effort to extract even more fuel economy out of future road-going vehicles is going into areas no-one considered possible even a few years ago. The potential benefits of ceramic ball bearings over traditional steel rolling element versions therefore, however slight, would seem to be one of these, and is hence perhaps the reason Formula One is no longer shying away from their use.

Before we go further though, let me outline a little of the terminology. Full ceramic bearings consist of an outer race, an inner race and intermediary balls all made from a ceramic material. On the other hand, hybrid ceramic bearings are those where only the balls are ceramics – a sort of halfway house incorporating most of the benefits of ceramic technology but limiting the cost.

All told, there are three basic types of ceramic available – silicon nitride (Si₃N₄) and silicon carbide (SiC), which are both black, and zirconium dioxide (ZrO₂), often called zirconia, which is a sort of dull white. All of these are lighter than steel, with a higher modulus of elasticity and lower coefficients of friction and expansion. One interesting property is that even though they have low coefficients of thermal expansion, at high temperatures they don’t appear to continue to expand at all. If all that wasn’t enough, components made from them can have a far superior surface finish, which (obviously) will never rust. In many cases the surface finish is such that any friction generated may be a little as 10% of that from a traditional steel version, such that lubrication with oil or grease is not strictly required at all.

Improved service life, high temperatures in a lubrication environment which is hostile, and new (again) to Formula One? This change of heart by the FIA seems to me to prepare the way for a new range of turbocharger applications, where the traditional fully floating bearing arrangement can be readily replaced by the latest in ball bearing canister designs.

Forgive me though – is this just another example of modern automotive r&d supporting Formula One, and not the other way round as we are frequently told?

Fig. 1 - Hybrid ball and roller bearings (Courtesy of SKF bearings)

Written by John Coxon

When it comes to the fundamental design there are two types of small-end bearing – the fully floating design, where the piston pins are free to rotate in both the piston and the con rod, and the fixed variety, sometimes referred to as pressed pins, where the pin is an interference fit in the rod. From time immemorial the fully floating variety would seem to have been the favoured one, but the method of retaining the pin and prevent it from ‘floating’ into the cylinder wall using little circlips or bent pieces of wire can cause many problems if not properly developed for high-speed running.

Favoured when engines are repeatedly dismantled and re-assembled, as in the case of race engines, floating piston pin designs are now considered to be superior to fixed pins in terms of lubrication, resistance to scuffing and noise generation. The pressed pin version, however, which historically has more often been favoured by volume engine manufacturers – and despite its name – is rarely pressed into place. Instead, manufacturers prefer to heat the rods in an oven, and the expanded pin hole contracts upon cooling, gripping the pin tightly. Assembled in this way the interference fit can be as much as 0.001 in, with 0.0015 in recommended for race engines. Problems often crop up start during disassembly though, when pressing the pin out when cold can broach either the surfaces of the pin or the little-end bore.

For the fully floating designs, early bearing materials are most likely to have been babbit (a mixture of lead and tin) but for the higher loads and after the phasing out of lead in manufacturing, modern materials tend to be bronze-based. In some cases these might be steel bushes faced with the bronze and pressed into the rod small end, but often this might just be the internal bore of the small end plated with the bronze itself. In some cases, engines may run with steel on steel when the piston pin runs directly in the unplated surface of the steel rod. More important though is to ensure that, whether pressed or fully floating, the little-end bearing gets plenty of lubrication – not only to ensure a minimum oil film thickness but also to cool one of the most hostile regions in the engine

Since the small-end bearing is not pressure-fed with oil, the lubrication of this bearing arrangement is a complex phenomenon. It’s a combination of two actions, the first of these being that of the squeeze effect due to the pin moving up and down relative to the rod and piston bores during the reciprocating cycle. Here the downward motion of the piston relative to the bearing and the oil supports the major part of the load. At the same time, the expanding gap underneath ‘sucks’ in the oil. When the relative motion reverses, the lower oil film comes into play, sucking in the oil in the top gap and so the oil flow to cool the area is maintained.

The other action is the rotation of the piston pin due to the inevitable friction in the bearing. This creates a second ‘wedge’ effect, reinforcing the squeezing effect of the up-and-down motion of the piston. Depending on the clearance in the bearings, the type of engine, the masses of the components and the speeds involved – and a host of other parameters – analysis of the small-end bearing is therefore highly complex.

It may be small and, being buried deep inside the piston, often difficult to get at, but in the search for higher engine outputs, understanding what goes on here is becoming more and more important.  

Fig. 1 - The ‘squeeze’ effect

Fig. 2 - Small-end bearing

Written by John Coxon

A plain white metal, bi- or tri-metal engine main or big-end bearing has to cope with a range of conflicting requirements. First, it has to be strong enough to support the loads. As engine speeds increase along with combustion loads, this is fast becoming a critical issue and where once simple bi-metal designs may have coped, stronger, more expensive, tri-metal units, often based around copper, are increasingly being used. As well as being stronger, the copper performs another function in that it conducts heat so much better than other materials.

A third characteristic of a plain bearing – and one which is ignored at your peril – is the idea of ‘embedability’, the ability to trap small pieces of dirt in its surface. This quality is essential in the early life of the unit, given that the surface of the bearing where it makes contact with the crankshaft journal has to be soft enough to ‘soak’ up those tiny bits of dust or metal swarf inevitably found in a new engine just after it’s been built. These are the result of machining operations or airborne dirt which, unless medical standards of cleanliness are pursued, can never be completely removed during manufacture.

At the same time the bearing also has to be capable of deforming to the shape of the crankshaft as it rotates during the initial period of minutes or hours of running. This would also appear to be essential if we are to avoid early-life bearing failure, and seems to be particularly important as lubricating oils get thinner. The strength of the tri-metal bearing therefore has to be compromised by adding a conformable surface layer or overlayer, which is comparatively soft and can melt and simply disappear if too much heat is generated.

In a tri-metal bearing design the steel-backed high-strength copper-based lining is initially electroplated by a micro-thin nickel layer, onto which the much softer tin-based overlayer is deposited. The nickel acts as a form of barrier, preventing the diffusion of the tin migrating away from the overlay and reducing its strength.

Wouldn’t it be nice though if we could add yet another layer which after a period of time would diffuse into the overlayer and gradually make it harder. Thus after the period when the risk of swarf and running-in debris had passed or been absorbed by the soft overlayer, components in this fourth layer would slowly diffuse outwards to the overlayer and, once in situ, work towards strengthening it and helping it support greater loading. Well I am reliably informed that this now is the case, and the addition of this ‘diffusion’ layer, once fully absorbed into the initially soft outer layer, increases its strength by an extra 15%!

‘Having your cake and eating it’ seems to be the term that springs to mind.

Fig. 1 - The four-layer bearing 

Written by John Coxon

Designed to control the end-to-end movement of the crankshaft to somewhere in the region of 0.004-0.006 in (0.1-0.15 mm) depending on the design, in small reciprocating engines thrust bearings take the form of either one-piece half-bearings incorporating integral flanges or loose separate semicircular, flat washers arranged either side of a central bearing. Prevented from turning in some way, and whereas the bearing journal relies on the difference between the journal diameter and radius of the bearing to create a ‘wedge’ of oil, the thrust bearing is limited by the fact that it is flat. That makes it much harder to create and maintain an oil film, so the only way oil can be introduced between it and the mating face of the crankshaft is by adding some kind of slight taper or including an oil groove directing oil spilling from the radial bearing.

Whichever way you choose, the loads able for support are far lower, making the manufacturing of both thrust bearing and crank that much more critical. In fact, because the two surfaces are flat – just like two metrology gauges when ‘wiped’ together – the tendency may be to stick. Thus if the axial load is so great as to displace the lubricating oil, causing the oil film to collapse, the surfaces could be inclined to stick together, overheat and then fail all the more readily.

While this may be an increasing trend given the increased axial loads, the only other way for the thrust bearing to fail is down to manufacturing or the (accidental, I assume) inclusion of dirt. Setting aside the most obvious concern of dirt, the difficulty in machining crankshaft thrust surfaces must not be underestimated either. Completed using the side of the grinding wheel during the same operation as the grinding of the journal, the surface not only has to be flat but also perpendicular to the axis of rotation of the crank.

The grinding wheel can also sometimes leave a swirl pattern of scratches, leaving a criss-cross pattern reminiscent of the honing patter in a cylinder bore. If not completely polished out, these can be highly effective in wiping away the oil film, and short-term failure will inevitably follow. So while the bearing design may be difficult, if the crankshaft thrust faces aren’t polished to perfection, all your efforts in engine build may come to nothing.

Fig. 1 - Semicircular crankshaft thrust bearing

Written by John Coxon

Inserted from the front or rear of the engine, the bearing diameter has to be more than twice the cam centre line to nose height. However, in the case of overhead cams, where the engine height may be critical, separate bearing caps gives the opportunity to remove this restriction, and the limitation is more likely down to loading and the materials used. In this case, from the handful of cams I’ve surveyed recently, for overhead cams using separate cam bearing caps, journal sizes can be anywhere of the order of 18-25 mm for lightweight contemporary designs.

Going back in history, camshaft bearings were effectively just smaller versions of main or big-end designs. Babbitt materials were used on vintage units and, when shell bearing designs became available, some form of aluminium-tin mixture (usually around 20% tin) would be deposited directly onto a steel backing. These would be dropped into place during assembly and, if memory serves, they rarely failed in any significant way so long as the flow of lube oil was maintained.

In more modern times, and to save cost without impairing durability, cam journals are more likely to run directly in the aluminium cylinder head for production engines. Presumably the spare graphite in the structure of a typical cast-iron cam, and the silicon in the cylinder head aluminium mix being compatible, gives no need to introduce yet another unnecessary component, and this of course will no doubt also keep those in charge of the finances happy as well.

Lately though, what with the relentless push to reduce fuel consumption, engine technology – particularly with respect to valve gear – is taking on an interesting twist. Although the cam journals in the valvetrain contribute little to the overall friction, efforts are being made to reduce this friction wherever possible. Thus on camshafts, plain journals are now being replaced by small roller bearing technology, and despite the obvious increase in complexity and commensurate cost, the difficulties of designing small split-roller cages are being overcome.

For those not already in the know, the big advantage of a roller bearing over that of the traditional plain bearing is that it eliminates the boundary layer effect when relative movement between shaft and journal is low. Theoretically, a roller bearing has point contact only, and with no sliding (just rotation) friction is low.

But when it comes to camshafts some ingenuity is needed. A roller bearing must have three elements – an inner race against which the rollers move, the rollers themselves held apart by a cage and the outer race. In the case of this latest cam roller bearing design the inner race is that of the camshaft journals, while the 18 or so 1 mm diameter rollers are held in a rigid plastic cage split so that it can be assembled around the cam journal. The outer surface, rather than relying on the aluminium of the cylinder head, is a pair of 0.7 mm thick steel shells that locate in the cylinder head using the hole drilled for the oil feed.

This ensures that the outer shell is correctly aligned and the oil feed is not restricted through poor build assembly. More interestingly, however, the split line of the bearing doesn’t follow that of the bearing cap. To ensure that the rollers are disturbed as little as possible when crossing this split line, the outer bearing is split in the shape of a sine curve, as shown in Fig. 1.

Perhaps this is the shape of cam bearings of the future?

Fig. 1 - Early aluminium-tin camshaft bearing in its upper housing alongside a more modern equivalent

Written by John Coxon

With oil film thickness measured in microns (0.001 mm) it is therefore critical to the performance of any such bearing to ensure that the dimensions, overall shape, surface quality and cleanliness are held within tight limits. Thus in designing a bearing, the critical parameter is that of the clearance between journal and shell, which is sometimes referred to as the ‘eccentricity’ of the bearing.

The function of the oil in any bearing design, however, is not purely to lubricate; another major aspect of all engine lubricants as applied to bearings is to remove any heat generated. So if a bearing appears to be adequately lubricated but the heat generated through the friction within the oil film is not dissipated, the bearing will eventually overheat and fail. While it is imperative to have sufficient viscosity in the fluid to support the loads generated, too much viscosity will reduce the flow of the oil and possibly also the rate of heat removal. If this rate of removal is insufficient, the bearing may become damaged and eventually fail. 

Thus, if a particular engine was designed to run on an SAE 30 grade oil, for example, then using an SAE 60 oil for competition purposes might not produce the results intended. The increased viscosity might be able to support the increased loading, but the reduced flow rate through the bearings could introduce other issues. In reality, competition engines, through their widespread use of wide-open throttle settings and heavy rates of fuel enrichment to develop maximum power, will load often excessive quantities of unburned fuel into the lubricant, reducing its viscosity. At the same time, competition bearings may have greater clearances to allow for higher oil flows, so oil with a higher viscosity when diluted with fuel may be within the design limits of the bearings.

If you like, the oil clearance in the bearing will have an optimum value, taking into account the combination of design parameters associated with the actual lubricant in the sump rather than what it was originally filled with. Greater bearing clearances will cause an increase in the oil flow and give lower oil and bearing temperatures. But at the same time these clearances will give a less uniform pressure distribution within the bearing as well as higher peak pressures, and higher peak pressures can lead to increased loads and consequently reduced fatigue life. At higher pressures, the minimum oil film thickness resulting could also be lower, producing asperity (metal-to-metal) contact between journal and bearing shell, creating increased wear.

While outwardly straightforward, designing and developing ‘simple’ crankshaft and big-end bearings is a lot more complex than you might think.

Fig. 1 - Bearing eccentricity

Written by John Coxon

For this pressure to be maintained, the flow from the oil pump needs to make up for the losses through bearing leakage around the system. But if either the oil flow or its pressure should fall below that required then the bearing will be irretrievably damaged.

That is not the full story though. In this context we have assumed that the oil is a uniform and single entity that is incompressible and behaves in a consistent way, more or less as a Newtonian fluid. This is far from the case.

Simply looking at the products of combustion and the amount of gases circulating within the system, the oil in any crankcase will be contaminated with unburned or partially burned fuel, water and, often in large amounts, air. Some of this air will be dissolved in the oil while some of it will be only entrained in it. If the temperature inside the sump is greater than around 100 C then the water will evaporate and take no further part in this discussion, as will any entrained air that is centrifuged out of the system via some means of separation (swirl pot or other separator). What remains in most practical situations is unburned fuel, dissolved air and, of course, the oil.

Regular readers will know that the pumping action of the rotating journal in a plain bearing builds up a ‘wedge’ of high-pressure oil in the converging zone between journal and bearing to support the load. In the diverging zone [Fig. 1] the oil pressure will fall away, and if this sudden change in pressure causes the solubility of the fuel or oil to exceed the saturation pressure at the prevailing temperature, a process known as cavitation can occur.

It has been suggested that there are two types of cavitation. The first is associated with air or its constituent gases dissolved in the oil. When the pressure falls to that of the lowest around the bearing, these gases are purported to come out of solution and will be carried around the oil circuit to be dissolved back into it as the pressure increases once again. This type of cavitation is not particularly severe and is thought to have minimal detrimental effects.

The second type is that of vapour cavitation, and because it is changes from a liquid to a gas and back again during the process, it is potentially the most damaging. As a big-end journal rotates in the bearing shell, the load and its direction will change significantly throughout the engine cycle. At the high speeds of a race engine, the frequency at which this happens is so high and the time taken for it to happen so short that the oil pressure in the oil rapidly falls and the fuel dissolved boils instantaneously. As the pressure rises again, the bubbles caused by the evaporation will collapse and go back into solution, the repeated action of which may ultimately fatigue the bearing surface.

Often discounted or even unknown by some engineers, bearing cavitation can be highly destructive and difficult to diagnose.

Fig. 1 - Distribution of oil pressure in a plain bearing

Written by John Coxon

As a journal rotates within a bearing the lubricant between them creates the well-known ‘wedge’ effect effectively separating the two. Maintaining this separation at all times at the main bearings is relatively simple, but ensuring that the big-ends do so as well is slightly more difficult.

There are two ways we can oil crankshaft bearings - feeding the fluid down the middle of the shaft and thence out to the mains and big ends in succession or, more commonly, through drillings in each main bearing housing down (or up) from the main oil gallery. The former creates minimum splash but compromises the mechanical design of the crankshaft, while the latter, although producing more splash (increasing resistance to the rotating crank), is more practical and ensures that each bearing (mains or big ends) receives, in theory, more or less the same amount of fluid at all engine speeds.

In order to transfer the oil to the big-end bearings, however, grooved main bearings are required. These carry the oil from the oil gallery feed, around the main bearings and into a drilling, carefully placed in the crankshaft and up to the con rod. A grooved bearing introduces the lubricant to the centre of the bearing from which it flows outwards, but in doing so the effective area in contact with the crank journal is reduced, along with the load-bearing capacity of the bearing. On the other hand, a plain main bearing shell has the maximum load-bearing capacity but cannot get the oil to the con rod. Clearly a compromise has to be reached.

The most practical solution is to limit the grooving in the (main) bearing, not only in its width (which controls the flow rate) but also in the extent of the groove around its circumference. Replacing the grooved lower main bearing with a plain one (to take the higher loads) while using the grooved bearing at the upper end seems to satisfy this dilemma. The big end may not receive a continuous flow of oil but so long as it is introduced during the period when the bearing load is low, the oil thus entrained should be able to maintain a minimum film thickness and prevent surface-to-surface contact.

Minimising the oil flow and reducing engine friction but maintaining the minimum oil film thickness at acceptable wear rates, and at all engine operating conditions? Welcome to the wonderful world of power unit engineering!

Fig. 1 - Grooved main bearing shell

Written by John Coxon

Copper-lead or perhaps aluminium-tin, with or without an overlayer, a plain journal bearing will never be able to function as desired without lubrication. Forced between two surfaces (the bearing and its journal) by the action of the oil pump, this oil together with the rotating action of the journal creates a 'wedge' of the stuff that keeps the surfaces apart, hence reducing the friction between them. If the pressure in this wedge is sufficient and the asperities of the journal and those of the bearing surface no longer interlock, the bearing is said to operate in the hydrodynamic regime. If these asperities interlock though, even if ever so slightly, friction (and hence wear) will be generated and the lubrication regime will become mixed or in the worst case boundary.

Of the many components comprising a modern engine lubricant, only the viscosity and the friction modifiers, if included, can directly affect the friction generated. Altering the shear rate within the bearing assembly by increasing the clearance can reduce friction, but the extra clearance will need higher oil 'make up' or flow, and so any performance released by reducing the friction may be consumed again by the extra lubricant to be pumped. On the other hand, too fine a clearance will never allow the oil to flow, and the heat generated - which is otherwise taken away with the oil as it leaks away - will inevitably lead to bearing failure.

Once the engine has fired and the journal is rotating, the main effect of the lubricant is attributed to the viscosity of the fluid at the prevailing temperature and the induced shear rate of the fluid. As this rate of shearing increases, the viscosity modifiers in the oil (which are designed to control the oil's viscosity with temperature) will lose their effectiveness and the oil will shear down to the viscosity of the base fluid. In a bearing therefore, one of the critical parameters is that of its viscosity at high speed and high shear rate, and is often referred to as HTHS. However, a lower HTHS viscosity lubricant would most likely produce a lower minimum oil thickness, which is so critical to the durability of the bearing. Under these conditions the presence of the friction modifier would come into play, for not having a major influence under hydrodynamic (or EHD) conditions, as the film thickness reduces, its presence could help reduce the friction by another way.

The selection of the lubricating oil for high-performance engines is a little more complex than most think.

Fig. 1 - A typical engine bearing

Written by John Coxon

And so for many years the tri-metal copper-lead substrate applied to a steel backing shell with a flashing of tin-lead was the bearing system of choice for high-performance gasoline engines. They gave the mechanical strength required but at the expense of other properties. They were comparatively expensive to make though, so to simplify manufacture a reticular (or honeycombed) tin-aluminium alternative was developed. A high-strength homogenous material with excellent bearing surface properties, these 20% tin-aluminium ingots were rolled down to a suitable thickness and roll-bonded onto a continuous steel sheet and formed to shape.

But as we have said, copper-lead has disadvantages. Compared to babbitt, copper-lead, containing 20-30% lead, was much harder. At normal running temperatures this was 60 Brinell compared to the 20 of the babbitt. The hardness of copper-lead reduced the embeddability as well as the conformability of the bearing, and consequently, since shafts always distort to some degree or other, the outright strength of the bearing could never be fully used. Furthermore, the copper-lead materials at the time had poor resistance to corrosion, with the lead phase being attacked by the acids in the lubricating oils of the day. Once this lead was removed the qualities of wear resistance and good boundary layer lubrication were gone.

The addition of a very thin (0.0005-0.0015 in) overlay in the form of either lead-tin or lead-indium at the very best delayed things, but because of the low strength it was easily removed by any dirt present, no matter how little. The fatigue resistance of the overlay was also very poor, giving rise to the dilemma of restricting the performance of the bearing due to this overlay or using the full performance of the copper-lead but being compromised on durability. Such compromises are often common in engineering.

The way around this issue of course was to develop a new range of alloys using soft(ish) aluminium with added tin to reduce the so-called solidus temperature (the temperature at which the solid begins to melt). With increasing levels of tin, at about 7% content the tensile and yield strength, as well as ductility, fall away quite sharply. This is because of the tendency of the tin to surround the aluminium grains.

The eventual solution was to be found not perhaps so much in the addition of other elements but by the breaking up of the tin surrounding the grain by a process of cold working followed by annealing. After this treatment the tin remains continuous along the grain edges and not across the faces, producing a continuous and therefore strong aluminium base matrix. The tin effectively forms a network similar to that of a honeycomb, and is hence referred to as 'reticular'.

Aluminium-tin bearings at around 6-7% of tin may not be the ultimate in bearing performance but they still currently account for the bearing technology in many of the world's high-performance road vehicles.

Fig. 1 - Plot of bearing load at seizure against percentage of tin in an aluminium bearing

Written by John Coxon

But bearing failures do occur, and sometimes the designers do not get it right, but more likely these days, some other event in the engine build process goes awry. Thus when a bearing exhibits, say, fretting, fatigue, or failure due to cavitation then the fault generally lies in the engine design process. In other cases, however, when for instance the failure is wear-related, things are not always so clear-cut.

In theory, all bearings will be subject to some degree of wear. At initial cranking, when the surface asperities of journal and bearing are interlocked, any relative movement is bound to generate friction and a bit of wear. Once the crank is moving, however, and the 'wedge' of oil is generated by the relative movement of the two surfaces, so long as the loads can be supported by the oil film then - again, in theory - no further wear can take place. I say 'in theory' simply because engineering components are not perfect things, and bearings journals are never perfectly round, or in some cases the bearing housing is not totally rigid so that bearing needs to adjust itself to accommodate these slight imperfections. Once 'accommodated' though, any wear will gradually reduce to a stable geometry. In the past we would have called this 'running-in' or 'bedding-in' - a process very familiar to those over a certain age.

But other forms of wear can be more serious, and they will all have much to do with either the cleanliness of the build environment or that of the lubricant used.

The most common form of bearing damage is often referred to as 'embedded debris' damage. Much as the term suggests, foreign particles become embedded in the soft overlayer, often surrounded by a ridge or 'halo' where the displaced bearing material has been forced up and then smoothed over by the passing of the crankshaft surface.

A result of poor cleanliness during build or poor filtration, most of the material found originates from the drillings in the crankshaft, oil galleries or casting sand when the quality of the end-of-line wash leaves much to be desired. Dust or dirt in the atmosphere settling on the bearings during build can also cause similar but less critical debris damage. Modern bearings, because of their lack of soft lead overlays, are perhaps less tolerant towards this kind of dirt.

Other forms of damage can occur, for instance, debris is caught between the bearing and its housing. Pressed into the inside diameter of the housing, the resulting distortion will feed through onto the running surface, resulting not only in accelerated wear at the point where it touches the rotating journal, but a localised increase in the temperature caused not only by friction but the reduced capability of transmitting the heat away and into the structure of the crankcase or cylinder block.

It may be obvious, and it may be like telling your granny to suck eggs, but complete cleanliness during the engine build process - while essential - is in practice quite difficult to achieve.

Fig. 1 - Main bearing after only a few hours' running. Notice the extreme wear on the right-hand side and the scoring as the engine finally failed

Written by John Coxon

At its most basic level, this running clearance can be described as the difference between the journal diameter and that of the traditional shell bearing at 90º to the parting line. For passenger cars with a nominal journal size of, say, 50 mm (2 in) this will be somewhere between 0.001 and 0.002 in, but as performance increases this is usually increased by up to 50% to take account of the extra flow and cooling needed. But with oil film thickness measured in microns, the importance of dimensions and shape in bearing design cannot be underestimated. And once machined to high levels of accuracy to produce the geometry desired, the objective is to maintain it under all conditions, both dynamic and thermal.

Take the case of a con rod, for instance. While the forces due to combustion will push down on the crankpin, those of an inertial nature will try to tear off the bearing cap as the piston disappears up the bore again. This cyclic load produces an extra-special challenge to the designer since the clamp load between cap and upper rod has to be sufficient to prevent movement of the cap. The traditional method of manufacturing is to machine the clamping faces of both cap and upper rod and then, bolted back together, machine the bearing housing diameter to the size required. This is time-consuming and consequently expensive, and to ensure accurate alignment during engine build it may require dowels concentric or offset, to the con rod assembly bolts. This will reduce the possibility of the cap moving - however slightly -under dynamic loading as well, and ensure the bearing stays accurately aligned throughout.

Nowadays, however, to ensure accurate alignment at less cost the method more frequently used in engine manufacture is that of fracture cracking. With this process the con rod is manufactured in one piece, and the intended big-end split line weakened in some way. Assembled on a jig fixture, the rod is split across this weakened point by a sudden impact, producing a clean break that will assemble back together accurately. Furthermore, the shape of the surfaces produced will give a considerable amount of lateral location, which will minimise any tendency of the cap to move or the bearing to 'walk' sideways during use.

Of course, not all con rod materials can be fractured in this way. Steels with any level of ductility are clearly unsuitable, and much development is ongoing to produce higher-strength steels with this brittle characteristic, but since production savings of the order of 25% are claimed don't be surprised if the traditional con rod and bearing cap arrangement becomes a thing of the past in OE engines in the years to come.

Fig. 1 - Fracture-split bearing cap

Written by John Coxon

When assembling half-shell bearings into, say, a cylinder block main bearing housing, even the most visually challenged can't help noticing that at the first sight the bearing does not appear to fit. Laying the half-shell across the half-moon of the cut-out leaves the bearing standing several millimetres proud, which only after slight pressure clips into place. Referred in engineering parlance as the 'free spread', this is designed to hold the shell in place during the engine build.

The more observant, however, might notice that even when seating as it should, the shell is still slightly proud, and no matter how hard you try the situation persists. Believe it or not, I have heard of some engine builders actually filing this flush to the level of the clamping face. This is totally wrong.

Bearing crush, as it is called, is incorporated for a number of reasons, for when clamped together the compression in the bearing shells provides a high resistance to the shell turning or progressively creeping in the direction of the crankshaft, as well as supporting the thin and otherwise flexible structure during use. Furthermore, under compression the shell will take up the shape or the aperture prepared for it, and sitting firmly and securely in place will permit efficient heat transfer away from its source - that is, the bearing. When assembled in its housing, this bearing crush will generally vary between 0.001 in for road transport applications, increasing to around 0.002 in for competition units depending on the bearing diameter.

If bearings are assembled with too little crush, the bearing will be loose and free to move backwards and forwards in its housing, and after disassembly highly polished zones will be seen on their reverse side and or possibly on the parting line face. Known as bearing shuffle, the loss of radial pressure will also result in inadequate contact with the bearing housing, leading to heat build-up within the bearing itself and possible damage to the bearing surface.

Although the incorrect practice of filing the shell flush can result in this, more probable causes are insufficient clamp load in the bolt (the bolt binding or bottoming in its thread) or burrs or foreign bodies holding the bearings caps apart. When bearings are assembled together, giving too much crush, the pressure in the shells causes them to buckle inwards around their weakest part. Evidenced by polished areas on the bearing surfaces, assuming the shells are to the correct size and specification then the culprit would otherwise seem to be either over-tightening of the clamping bolts or incorrect machining of the bearing housing.

So next time you are trapped on the Underground or the Subway, crush in the wrong place may be uncomfortable but in the right place it has its benefits too.

Fig. 1 - Bearing crush or 'nip'

Written by John Coxon

Ceramic bearings fall into two distinct types - full ceramic units when outer race, inner race and balls are all made from a ceramic, and 'hybrid' units when for reasons of cost, only the inner ball/rolling elements are ceramic. With the latter, although the inner and outer races will be superfinished in steel, the full assembly still retains many of the advantages of the full ceramic. Furthermore, bearing ceramic materials can be classified into three common materials used: silicon nitride (Si3N4) - which seems to be the most favoured currently -silicon carbide (SiC) and zirconium dioxide (ZrO2), sometimes referred to as zirconia. The first two are black in colour while the third is white, but despite these differences, all have significant advantages over steel.

It has always bemused me that a ball or roller bearing had any significant friction because, at a theoretical level at least, only point contact with no relative sliding motion exists between the inner race and the ball or roller, or the ball/roller and the outer race. Think of it as an inner gear of an epicyclic gearbox surrounded by star wheels and an outer annulus, and you will understand that in theory, since there is no sliding motion, no friction should be present.

However, as well as the inevitable small amount of slip, energy will be needed to rotate the roller around first its own axis and then around the axis of rotation of the whole assembly. Ball or roller inertia therefore plays an important part in the 'friction' within a ball or rolling element bearing. So when the density of the ball or roller is around 40% that of steel, this constitutes a major reduction in the energy required to move them. Also, the forces in the bearing and its 50% greater stiffness over steel produce less distortion and generate less heat than its steel equivalent.

This reduced amount of heat, and the lack of steel-on-steel contact generating microwelding between the components, also means that adhesive-type wear will be substantially reduced. Ceramic bearings therefore run cooler and require much less in the way of lubrication than traditional materials.

Lighter weight, reduced friction, less wear and lower operating temperatures all add up to a longer operating life (sometimes by as much as ten times) or greatly improved performance; a sort of win-win-win-win situation, which is difficult to ignore. In the words of a recent UK TV advertisement, "Now you don't often see that do you?"

Fig. 1 - Ceramic bearings

Written by John Coxon

Running for most of its time under benign, hydrodynamic lubrication, such a bearing nevertheless has to cope with a conflict in characteristics. A bearing has to be strong, being able to take high cyclic loads for long periods of time with little or no wear. It needs to resist seizure and physically prevent itself from sticking to the crankshaft journal. Embeddability - the ability to retain small foreign particles from the oil - needs to be good, along with the ability to accommodate small misalignments of the crankshaft. If that wasn't enough, the material also needs to resist corrosion, cavitation and have a high enough melting point to cope with any heat that is inevitably generated. In short, a bearing needs to be strong and soft at the same time, and how this is achieved depends on the intended application.

The simplest approach for most light and medium duty applications is the bi-metal bearing. An aluminium-tin (up to 20% tin) alloy attached to the steel backing using a thin layer of pure aluminium will provide much of the above but without the strength. Adding small amounts of silicon, copper or other transition metals (chromium, manganese, nickel) will improve the strength but not to the point now demanded by most current high-performance engines. For applications such as these, a more complex approach is required - that of the tri-metal bearing.

Here the aluminium alloy is replaced by a much stronger copper-based alternative, which includes a high proportion of lead (up to 25%, where regulation allows) to act as a solid lubricant. Somewhere between 0.25 and 0.4 mm (0.01-0.015 in) thick, small amounts of tin can be added to strengthen things further. However, copper-lead-tin alloys may be strong but they lack the softness required for embeddability and that other necessity, conformability.

For these we have to introduce yet another layer - the overlayer. At around 0.01-0.02 mm (0.0005-0.0008 in) thick, this overlay will usually consist of a lead and 10% tin mixture on top of a nickel barrier. While the lead-tin mixture will absorb any dirt or fine debris in the oil, the nickel is present not necessarily to bond it to the lower layer but to prevent the tin from progressively migrating towards the copper in it.

The problem with this overlayer, however, is its lack of strength, and although small amounts of copper can be introduced, strangely the thicker the layer the less its load-carrying capacity, and so overlays are always very thin. But as in many things the thickness is very much a compromise. Too thin and any slight wear or misalignment will uncover the copper layer; too thick and the surface may start crazing, causing partial flaking as the layer parts company with the rest. Either way, bearing failure quickly follows and is why in a modern tri-metal bearing the overlay is most critical.

These days lead is now banned in most OE automotive components, so bearing manufacturers for production vehicles - probably the worst affected - have had to come up with lead-free substitutes, particularly for the overlay. Tin-based alloys with around 6% copper deposited on a nickel barrier layer is one approach. Other proprietary methods include mixtures of solid lubricants (molybdenum disulphide, graphite or PTFE) and hard, wear-resistant particles (silica, alumina or silicon carbide/nitride) all bound up in polymer resins. Sprayed onto the base bearing material using suitable solvents, these are cured at temperature. Although race engines are currently unaffected by these directives, the increasing concern over the use of lead in any manufacturing process may force race bearing manufacturers down the same route.

So do you still think bearings are simple?

Fig. 1 - A 'simple' tri-metal bearing with lead-tin overlay

Written by John Coxon

Now before we start reading too much into it, this was a survey conducted by an aftermarket bearing manufacturer, and presumably included a range of internal combustion engines of varying age and condition requiring some form of repair, and probably excluded most race engines. Nevertheless, even in a clinically clean build environment, upon stripping an engine it is surely a dishonest person who has never seen evidence of any dirt in the bearings. Fine dust from unfiltered air, small metallic particles from the piston ring or cylinder bore - even the odd grain of sand from the casting process - all may be regularly present, and all eventually lead to scuffing in the bearings, depending of course on the size of the particle and ultimately whether or not they are trapped by the oil filter. In the selection of any bearing material, therefore, it won't surprise you to hear that this almost inevitable presence of dirt features very highly.

Scuffing in a bearing is simply the cause of localised welding or 'pick-up' between the asperities on the surfaces of the journal and the bearing. It's caused when a foreign particle, larger than the minimum oil thickness, becomes lodged between the surfaces and is embedded into the bearing surface; the bearing alloy moves to form a raised lip around the particle, retaining the particle in the crater thus produced. If metal-to-metal contact between this lip and the shaft surface occurs then damage to the rest of the bearing surface could result, with eventual seizure a possibility. The characteristic to resist this scuffing is referred to as the 'embeddability' of a bearing, and depends on two factors - compatibility and conformability of the materials used.

Compatibility is the ability of the bearing material to resist localised, solid-phase welding. For this, the adhesive component of the friction during the contact with the journal needs to be low. Conformability, on the other hand, is the ability of the bearing material to move and conform to the shape of the shaft and all the minor imperfections including any dirt that may present.

Conformability is therefore a function of the softness of the material and the thickness of the layer on the backing shell. But soft materials as a rule do not support high loads, so the trick of the bearing designer is therefore to seek a compromise, balancing the trade-off between load-carrying capacity (which requires strength, particularly that in fatigue) and embeddability of the material or its softness.

Lead, perhaps the softest of all engineering materials along with its neighbour in the Periodic Table, tin, therefore make excellent bearing surfaces when the loads are low. Babbitt metals containing lead and tin are a perfect example, and are still preferred in many pre-war race engine applications. For higher dynamic loads, and when bonded to a steel backing shell for strength, copper (in the form of a lead bronze, with its ability to disperse heat) is still favoured for more recent competition engines, but is mixed with a sizeable portion of lead to give it a certain amount of embeddability. Increasing the amount of lead reduces the load-carrying capacity but improves the embeddability - essential for a good bearing. So while a commonly used competition lead-bronze bearing at 14-20% lead has a 30% higher load-carrying capacity, its sister material at 20-26% is considered to be the better bearing.

So even in the cleanest of environments you can't afford to ignore the dirt.

Fig. 1 - Evidence of scuffing (and much more)

Written by John Coxon

If you go back in history, ball or roller technology was a common sight. Before World War I, the leader in four-valve overhead cam engine design, the Peugeot L3, used ball races on the three main bearings of this four-cylinder unit. The big-end bearings were reserved for plain white-metal bearings but by this action the design intention, I think, was made clear.

Later on, in the mid-1950s and even when 'modern' thinwall steel-backed shell bearings were readily available, Mercedes-Benz still committed to a system of roller bearings on all main and big-ends. Here the multi-piece crankshafts necessary to take the roller system were assembled together using the Hirth process - a system of radial serrated journals clamped together axially by large screws. Even in 1970, legendary engine designer Mauro Forghieri is reported to have preferred roller bearings on the four main bearings of the 312B but eventually settled for just two, one at either end of the one-piece crankshaft.

Clearly the issue is the crankshaft and the strength of multi-piece units to enable assembly with the traditional roller bearing. In theory at least, the rolling point contact of the roller bearing over the sliding technology of a more traditional approach should give a significant advantage, and is perhaps why bearing manufacturers are now beginning to look at the concept again. The major difference over earlier designs is the development of split roller technology, which can be readily assembled around traditional one-piece forged or machined-from-solid crankshafts.

Analysis of friction within the traditional internal combustion suggests that around 20% comes from the main and a further 10% from the remaining big-end bearings. This comes from the drag induced in the oil film within the bearings and the high oil pressure needed to generate it. Rolling contact bearings, by their very nature, have considerably less friction, and since they do not require anywhere near as much lubricating oil they should therefore have a significant advantage.

It has been calculated that the friction of a roller bearing is around 10% of that of a traditional bearing at about 3000 rpm. If this holds true on the test bed then the FMEP (Friction Mean Effective Pressure) due to the bearings of an engine of 0.4 bar on an engine at 6000 rpm could be reduced to much nearer 0.1 bar or less - a small but nevertheless useful increase of around 2% in net power output.

In practice, however, no such advantages have been observed yet. Even with significantly reduced oil flow, the action friction measured is equivalent only to that of the best designed shell systems, which is particularly surprising. Clearly, modern shell bearings are much more efficient than many of us are aware and, after factoring in the extremely high levels of durability now expected, these apparently simple devices are highly underrated.

Fig. 1 - Typical Split bearing cage

Written by John Coxon

Dangerous enough as it may appear and rather than people stepping out in front of pit lane traffic, my primary concern is for the engine bearings - an issue that, perhaps rather predictably, has escaped many correspondents.

When an engine is stationary, the crankshaft settles onto the main bearings, and with no oil film separating them metal-to-metal contact occurs. At the point of initial cranking, there will be no oil pressure since the oil pump itself is stationary; there will therefore be no lubricant present to separate bearing and journal. Students of tribology refer to this as boundary lubrication conditions, and for a couple of revolutions or more during cranking and initial fire, bearing wear can take place.

With lubrication conditions so very different from those that exist during the normal high-speed, high-load running, bearings intended to cope with frequent stop-start regimes will need to be redesigned. A Formula One engine is designed to start only a limited number of times, and increasing this by a factor of, say, three, four or maybe ten will give bearing designers much to think about.

In the world of road-going transport, where hybrid internal combustion/electric motors are becoming more common, the issue is a very real one. Before the introduction of this so-called stop-start technology an engine might be expected to start somewhere between 5000 and 10,000 times during its life. That is 5000-10,000 times with little lubrication present and the accumulated wear that will inevitably occur. The introduction of stop-start technology when the engine switches off automatically as the vehicle comes to rest, and restarts as soon as the throttle is pressed again, has increased this by perhaps ten-fold, so bearings are now being designed to cope with this extra level of distress when the bearing is so vulnerable.

The ban on the use of lead - perhaps one of the best bearing materials available - in road transport engines under a European Directive for 2011 has encouraged a whole new range of alloys based on aluminium with various overlays, one of which is copper-tin. However, the introduction of stop-start technology has necessitated the development of an alternative overlay consisting of PolyAmidelmide (Pal) resin binder, into which undisclosed ceramic additives are finely dispersed.

These provide the desired properties of wear resistance, mechanical strength, thermal conductivity and embeddability so essential for a bearing at low to medium speeds. At high speeds, however, the thermal conductivity of the substrate is limited, so if the product is to be used in higher-speed engines then it requires more development. While not directly applicable to Formula One, since lead is not strictly banned under the terms of the Directive, nevertheless there is a clear challenge to produce more durable bearings should the ban be adopted.

So while everyone else is arguing about the safety aspects of electric pit lane running, spare a thought for the bearing designer who, unbeknown to most, will have to make it all work.

Fig. 1 - Polymer-coated bearing (Courtesy of Federal Mogul)

Written by John Coxon

The issue here is so-called 'turbo lag'. Primarily a function of the lack of exhaust gas energy upon opening the throttle, and therefore an inherent issue with the technology, spool-up times (the cause of turbo lag) can still be reduced by minimising the rotating inertia of the wheels and attacking that perennial favourite of engineers - friction in the bearing assembly.

The traditional turbocharger bearing arrangement was described here in RET-Monitor a couple of months ago. A fully floating concept consisting of two journal bearings at either end of the shaft supported by an additional thrust bearing next to the compressor, the most dominant effect on the friction inside the rotating assembly is this thrust bearing.

Described by some turbo engineers as the 'Achilles heel' of the design, the thrust assembly generally comprises a fixed thrust bearing and a rotating ring. When oil is fed into the wide end of the tapered pad on the thrust bearing, the action of the rotation will drag it back into the narrow end, building up oil pressure to withstand the axial load. For most of the time, however, the direction of this axial load will vary; it very much depends on the pressures in and around the compressor and turbine wheels at any particular instant. The net axial load is then taken by one of the two opposing faces depending on its direction.

In hybrid turbos where bigger compressors are used, the compressor dominates and pulls the shaft towards it. If air restrictors are required (as in the case of some engine formulae) the situation is exacerbated and turbos need a lot of extra thrust capacity. For these units, 360º-type thrust bearings as opposed to the normal 270º versions are commonly found.

But since friction is the enemy of good low-speed transient response, the major push of late has been towards the use of ball bearings in the form of a cartridge replacing both the journal and thrust bearings. Consisting of a single sleeve containing a set of angular contact ball bearings at both ends, for a given radial load the friction difference between these and the standard plain journal is negligible. However, under axial loadings irrespective of direction, the traditional advantage of the ball bearing begins to show, allowing the central core of the turbo to spool up much quicker from low speed.

Although improved engine response results, the system has to be designed so that ideally it keeps the ball races in contact with both outer and inner races at all times and at both ends, irrespective of the direction of the axial thrust. Failure to ensure this could easily result in rapid failure at the speeds experienced, since rolling element bearings are notoriously poor in such circumstances.

And since turbochargers will become increasingly common in the coming years, reliability could be a major issue.

Fig. 1 - Comparison of spool-up times

Written by John Coxon

In a properly designed bearing, the oil forms a thin film between the journal and the bearing, and separates the two such that no contact should take place so friction is therefore minimal. Only during stopping and starting, when boundary lubrication is present, is friction high and the bearing ever at risk, or if highly flexible cranks are used when the bearing is loaded at its edge.

So the 'slipperiness' of the bearing material is not the main issue. More particularly, it is the strength of the material, its ability to absorb the loads placed upon it, and the ability to cope with the heat that may result. Ideally, and often forgotten, a good bearing material should also be able to fail without damaging the journal that runs within it. In this way, costly crankshafts can be saved at the expense of much cheaper bearing components. As load factors increase with engine performance, this last factor becomes harder to achieve.

In the case of white metal, the options are generally either lead-based or tin-based. The stronger of these are the tin-based bearings, which will not only accept much higher loads but much higher surface speeds as well. Known as Babbitt metals - not to be confused with lead-based alloys which are often erroneously referred to also as Babbitt metals - these are typically 80-90% tin (Sn) mixed with 8-12% antimony (Sb) and about 3-6% copper (Cu). When dissolved in the tin, the antimony forms a eutectic of very hard crystals of SnSb which, being lighter than the melted mixture, will ordinarily float to the surface to give a hard, upper layer on top of a relatively soft, lower one.

The material in this state would be too hard at the top and too soft underneath for satisfactory bearing applications. The addition of copper to the melt upon cooling produces dendrites of Cu6Sn5, which interlock, preventing the free movement of the hard SnSb constituent and producing a more consistent material, more easily used by the bearing casting industry. A bearing produced in this way gives a more uniform distribution of hard load-bearing particles within a soft malleable matrix with a much lower melting point. So, as the bearing wears, the softer metal erodes to create a refuge for the lubricant, while the harder interstitial particles provide the actual bearing surface.

When considering the elemental composition it is important to remember the effect of lead. In times past the manufacturing method for making tin invariably included an unavoidable impurity of up to 0.35% lead (Pb). When introduced into a tin-antimony mixture, this impurity surrounds the SnSb crystals and weakens the microstructure. Forming yet another eutectic with tin and antimony, the lead reduces the melting point of one component to 143 C, which is only slightly above that of the bearing's normal operating temperature and is clearly unacceptable for any form of competition. When supplying tin-antimony bearings for competition use, only the purest form of tin, free of lead, should therefore be used.

To many, white-metal, Babbitt bearings may be considered old-fashioned or even slightly downmarket. However, in their more modern, much purer form they can be superior and more accommodating to more modern bearing materials when used against older type and more flexible cranks.

Fig. 1 - Fully machined white-metal bearing ready for installation

Written by John Coxon

Manufactured from aluminium, or more likely some form of leaded bronze, these bearings are loosely restrained axially using a pair of 'snap' rings, one either side, and will therefore rotate along with the shaft according to the friction balance within the system. Simple in concept it may be, but the design of such a bearing is a lot more complex than you might think.

In the early days, when simple bushes were pressed into a housing, the compressor-turbine assembly became unstable at certain critical speeds. This would lead to excessive out-of-round movement of the shaft causing metal-to-metal contact between shaft and bearing, and the possibility of the compressor blade tips making contact with its housing. Increasing the shaft size and hence stiffness would move these critical speeds higher up the rev range, but at the same time increase rotating inertia and reduce spool-up times. Likewise, increasing the tip-to-compressor housing clearance prevented contact in this critical area but bearing durability was still compromised.

The not-so-obvious solution to all this was to allow the bush to rotate in its own housing and optimise the clearances, first between the shaft and bearing and then between bearing and its housing.

The original concept of the floating bearing was to minimise friction, for although frictional forces are low, at 100,000-150,000 rpm when friction is proportional to the square of the speed, this amounts to a significant amount of lost power. However, it was quickly found out that oil flowing between the bearing and its housing introduced an element of damping, preventing this metal-to-metal contact and improving durability. If the inner clearance between shaft and bearing is less than the outer clearance between bearing and housing then the sleeve will rotate at a fraction of that of the shaft speed, and the overall power consumed is reduced.

However, perhaps more important, an element of damping is introduced into the system, which will reduce the instability in the shaft. This enables the shaft to run freely through its critical speeds without the possibility of compressor wheel blade tips coming into contact with its housing. In practice, while it may be assumed that the bearing runs at about half that of the shaft speed, this apparently is rarely the case - a figure like 40% is more likely, but this depends on many factors. In one common design the figure is nearer 25%, while that of a slightly larger version is about 30%.

At typically between 0.0010 in and 0.0016 in (0.025-0.04mm), the inner clearance controls the load-carrying capacity of the shaft while the outer clearance, typically somewhere between 0.0026 in and 0.0034 in (0.066-0.086mm), controls the damping in the system. In production designs, actual clearances are generally selected based on running the bearing at the lowest possible speed when the durability of the bearing is at its highest.

Turbocharger bearings may to the layman appear to be simple but a considerable amount of engineering goes into their design and development.

Fig. 1 - Traditional turbocharger bearing layout

Fig. 2 - Typical turbocharger bearing

Written by John Coxon

In fact, the thin-wall shell-type crankshaft bearing had originated in the US in the 1920s as a replacement for the earlier thick-wall/poured-in-place type, which had consisted entirely of lead-based white metal. Although white metal offers first-rate bearing properties, its fatigue strength is poor. The thin-wall alternative consisted of a thin steel backing with a thin white-metal lining. This composite offered superior fatigue strength with another advantage - superior precision in manufacture.

Vandervell bought a licence to manufacture the patented thin-wall bearings for the UK and European markets. He was soon improving them, and Ferrari was an early convert in the racing world. These days the company established by Vandervell in the UK still supplies bearings to race engine constructors, although it's now under the ownership of the MAHLE Group. Its high-end race bearings are known as V-series in recognition of where, within that global organisation, they are still made.

The tradition started by Vandervell in 1949 extends these days not only to Formula One but also to the likes of Le Mans, NASCAR and the NHRA. V-series bearings are widely used in these series, and these days they are 'tri-metal' productions, indicating that they are a composite of three rather than two layers. But what hasn't changed is the use of lead alloys in the inner two layers.

This is perhaps surprising, given the trend in the automotive industry to eliminate lead from all engine components. This trend is being widely encouraged by governments - indeed, the EU has now decreed that automotive bearings for production vehicles must be lead-free. In time, as with the removal of lead from gasoline, those who govern racing will most likely come to ban lead in bearings.

In the meantime, MAHLE and the former Vandervell and Clevite operations owned by it produce unleaded bearings for production cars and bearings containing lead alloys for high-end race engines. Through recent RET-Monitor articles, author Dieter van der Put has investigated the use of lead in crankshaft bearings for race engines. He eventually gave the opinion that there is no clear-cut advantage from the use of lead.

However, whereas unleaded bearings have been proven in roadcar applications, high-end race engine running conditions are very different. The overwhelming empirical evidence today is that the vast majority of high-end race engine engineers are prepared to pay a premium for the protection afforded by tri-metal bearings containing lead alloys. Given the number of highly talented and experienced engineers involved across the various disciplines - and current concerns to save costs - this can be no coincidence, so we have withdrawn van der Put's opinion on this matter from the RET-Monitor.

One wonders what that visionary Tony Vandervell would have made of all of this. Sadly he and his green-painted Thinwall Special Grand Prix Ferrari, which won the 1951 International Trophy at Silverstone, have long departed.

Fig. 1 - Lead is still used in race engine bearings

Written by Ian Bamsey

A number articles, in both Race Engine Technology magazine and here in RET-Monitor, have been published on con rods, mostly regarding overall con rod design and material. Bearings have been discussed as well, on which the absence of a small-end bearing bush was the main topic of interest, and which I believe is still not really common practice.

Let us make the step towards the lead-free bearing bushing material. As described in a previous article, lead-free bearing bushes have higher hardness compared to regular leaded bearing materials. To provide some numbers, typical leaded bearings have hardness values in the 80-120 HB range, whereas lead-free bearing materials are typically about 135 HB. In general, this means lead-free bearings have higher loadability than leaded bearings, but are less forgiving as regards conformability. Later in this article I will give an example of where this is of significant difference and where the transfer to lead-free materials could lead to engine failure if a number of design and manufacturing aspects are not properly executed.

But first I would like to discuss further the mechanism of sensitivity on bearing loads in the small-end bearing. When discussing highly loaded engines we should differentiate between two situations - loading due to combustion load, typically in turbocharged diesel engines with high peak firing pressures; and loading due to inertia load, typically in high-revving gasoline engines

Regardless of the type of loading, it will lead to bending/deformation of the piston pin, resulting in an oval shape of the pin. In severe cases, the amount of deformation can be such that the pin can no longer move freely in the small-end bearing bush. This will almost always result in damage to the small-end bearing, and hence total engine failure.
To overcome this a typical solution is to give the small end a so-called trumpet shape, which can be done by precision machining after installing the bearing bush. During this operation on the small-end bore, a somewhat larger diameter (about 5-6 microns) will be machined towards the outside of the bore. The exact final shape needs to be tailored to the specific situation, by means of determining an initial geometry by Finite Element Analysis, after which engine testing and small-end bore measurements will be used to make final adaptations at a detailed level.

What can be observed during engine inspection after an engine run is that the edges of the small-end bore have become dark. As a rule of thumb, the trumpet shape is correct if the outer couple of millimetres show an even and supported area of contact, not a line contact.

When transferring to lead-free bearings the detailed geometry and machining quality of the small-end bore shape is of major importance. Due to the fact that the hardness of the material is higher, it takes more time to run-in properly. Running-in protocols are known to have been adapted to longer duration in combination with lower loads when using lead-free bearing bushes. All this is done in order to give the components more time to adapt to their final shape.

There are two other areas of attention when looking at the trumpet shape that are more critical using lead-free bearing materials - the 'waviness' of the bore shape and the radial machining grooves in the bearing bore.

Both of these will lead to local contact between piston pin and bearing. A difficulty with these parameters is that the con rod might be within its drawing specifications and tolerances, where these local deficits, not separately specified, might lead to engine failures. Typical failures can include rotation of the bearing in the bore, loss of bearing material, deformation of the bearing, excessive heat generation and damage to the piston pin and piston bearing bores.

It is in these circumstances that the lead in leaded bearing materials still has the advantage due to its self-contained lubrication and conformability properties.

Written by Dieter van der Put

A number articles, in both Race Engine Technology magazine and here in RET-Monitor, have been published on con rods, mostly regarding overall con rod design and material. Bearings have been discussed as well, on which the absence of a small-end bearing bush was the main topic of interest, and which I believe is still not really common practice.

Let us make the step towards the lead-free bearing bushing material. As described in a previous article, lead-free bearing bushes have higher hardness compared to regular leaded bearing materials. To provide some numbers, typical leaded bearings have hardness values in the 80-120 HB range, whereas lead-free bearing materials are typically about 135 HB. In general, this means lead-free bearings have higher loadability than leaded bearings, but are less forgiving as regards conformability. Later in this article I will give an example of where this is of significant difference and where the transfer to lead-free materials could lead to engine failure if a number of design and manufacturing aspects are not properly executed.

But first I would like to discuss further the mechanism of sensitivity on bearing loads in the small-end bearing. When discussing highly loaded engines we should differentiate between two situations - loading due to combustion load, typically in turbocharged diesel engines with high peak firing pressures; and loading due to inertia load, typically in high-revving gasoline engines.

Regardless of the type of loading, it will lead to bending/deformation of the piston pin, resulting in an oval shape of the pin. In severe cases, the amount of deformation can be such that the pin can no longer move freely in the small-end bearing bush. This will almost always result in damage to the small-end bearing, and hence total engine failure.

To overcome this a typical solution is to give the small end a so-called trumpet shape, which can be done by precision machining after installing the bearing bush. During this operation on the small-end bore, a somewhat larger diameter (about 5-6 microns) will be machined towards the outside of the bore. The exact final shape needs to be tailored to the specific situation, by means of determining an initial geometry by Finite Element Analysis, after which engine testing and small-end bore measurements will be used to make final adaptations at a detailed level.

What can be observed during engine inspection after an engine run is that the edges of the small-end bore have become dark. As a rule of thumb, the trumpet shape is correct if the outer couple of millimetres show an even and supported area of contact, not a line contact.
When transferring to lead-free bearings the detailed geometry and machining quality of the small-end bore shape is of major importance. Due to the fact that the hardness of the material is higher, it takes more time to run-in properly. Running-in protocols are known to have been adapted to longer duration in combination with lower loads when using lead-free bearing bushes. All this is done in order to give the components more time to adapt to their final shape.

There are two other areas of attention when looking at the trumpet shape that are more critical using lead-free bearing materials - the 'waviness' of the bore shape and the radial machining grooves in the bearing bore.

Both of these will lead to local contact between piston pin and bearing. A difficulty with these parameters is that the con rod might be within its drawing specifications and tolerances, where these local deficits, not separately specified, might lead to engine failures. Typical failures can include rotation of the bearing in the bore, loss of bearing material, deformation of the bearing, excessive heat generation and damage to the piston pin and piston bearing bores.

It is in these circumstances that the lead in leaded bearing materials still has the advantage due to its self-contained lubrication and conformability properties.

Fig. 1 - Piston pin damage due to local bearing deficits

Written by Dieter van der Put

First though I would like to pass on a remark made to me when talking to a well-known bearing shell supplier. When discussing the possible reasons why leaded bearing materials are still favoured over lead-free materials, I was told, "Leaded bearings have been around forever, so people are in their comfort zone with them."

At first I was surprised by this but, on consideration later, I could understand it. It's most probably the same reason why we are still so familiar with the traditional internal combustion engine - not forgetting the fact that the introduction date for the End of Life Directive has been postponed to the middle of 2011. This of course has also reduced the focus on lead-free alternatives.

Now to the engineering side of this matter. Until now, there have been two factors pointing to uncertainty over using lead-free materials - conformability/embeddability and temperature - with temperature perhaps not as strong a factor because oil degradation seems to be more sensitive to higher oil temperatures than the bearing material itself.

As far as the current status regarding bearing shell materials is concerned, in general there are two approaches to their use.

High specific bearing load applications will typically use a lower hardness running layer on a higher hardness base material. An example here is 'Sputter' bearings, which were developed to the limits of galvanic surface coating technology. These bearings are often used in the highest loaded areas of the bearing system (the lower main bearing shell and upper con rod shell) in combination with lower hardness bearing shells for sufficient embeddability.

With Sputter bearings, the base material will need to provide sufficient conformability against deformations. The running layer is too thin to accommodate this function.

Medium and low specific bearing load applications use low-hardness running layers in combination with a higher hardness base material. High-revving race engines - as in Formula One, especially those with large bore sizes - tend to have a certain design freedom for increasing bearing shell widths. As can be read in the latest RET magazine (issue 49, September/October 2010), the main bearings of the Toyota RVX-09 engine have a diameter of 44 mm and a width of 18 mm, resulting in a ratio of 41%. The bearings of the con rods are even more squared, at a ratio of 49% (D: 36mm/ W: 17.5mm). In combination with modest peak firing pressures, this leads to medium specific bearing loads, enabling the use of running layers with very low hardness (10-15 Vickers). An advantage here is very good conformability, combined with an exceptionally high embeddability.

As can be read in one of my earlier articles on lead-free bearings, deformations were looked at from the viewpoint of an initial deformation of the bearing bore. Apart from the initial quality of the machined bearing bore, deformations will also occur during running of the engine, based on the crank train loading and engine block stiffness.

To get a feel for how much deformation can occur, we can make the following comparison. As a rule of thumb, typically the bearing clearance (diameter) is often taken as 0.001 of the bearing diameter. This leads to approx 44 µm bearing clearance. Without having real and objective data of Formula One engines at hand, finite analysis simulations of similar designs show deformations during engine running, which are in the magnitude of half this bearing clearance. This leads to the conclusion that these levels of deformation are of significant influence on the bore shape.

According to a well-known bearing shell supplier, the most common reason for failure in bearing shells in high-revving race engines is wear of the bearing surfaces due to contamination. Lead is able to embed particles of contaminants into the leaded areas, so they don't damage the running surface of the crankshaft. It remains a challenge to develop lead-free alternatives with similar properties in terms of conformability and embeddability as their leaded predecessor.

Having talked to a number of insiders in the race engineering industry on this topic, one can start to understand why the industry is reluctant to develop an alternative to this sensitive bearing system - particularly when there is no urgent legislative reason to do so.

Fig. 1 - Mahle Clevite bearings (Courtesy of Mahle Clevite)

Written by Dieter van der Put

The data here is to be read as relative between the two materials: due to the differences between the several racing classes, absolute figures cannot be given, although the information provided can be taken as representative. I am sure though that a fair number of race engineering colleagues would have more detailed information on their specific applications, which hopefully they are willing to share.

The temperature in the bearings is very difficult to measure directly. The most direct approach is to measure the temperature of the bearing shell in its main bearing bore using thermocouples. It is assumed that a temperature delta of about 5-10 C between the bearing back and the bearing surface exists.

The long-term stability of regular oils is secured until about 140 C. As the last link after the main bearing in the oil feed chain, the con rod bearings will see oil temperatures about 10 C higher, and based on this the maximum oil temperature limits in the main oil gallery can be determined.

The next step is to declare the variables on which the bearing temperatures need to be mapped. In this particular case, oil pressure and rpm are taken as variables, against which the temperature data is investigated. Some factors need to be taken as constant, to be able to predict the isolated differences on the parameters under investigation. In this case, the main gallery oil temperatures were kept constant (controlled by coolant temperature). Clearly, oil specification is another important factor, so this is kept constant by using one oil specification throughout the measurements.

Initially, leaded bearing shells were built into the engine, and bearing shell contact temperatures recorded during an rpm sweep. A significant temperature rise with increased rpm was observed: over a 50% rpm rise, a temperature rise of 10 C was measured. Rpm range was in the higher regions of the total rpm band.

Interestingly, not all the bearing shells showed the same temperature rise coefficient. This confirms that bearing loading can be considered as the major factor in temperature rise. The rpm rise is the same for all bearings, which assumes the same temperature rise coefficient over all of them.

In this particular engine, main bearing no 4 has the highest load, hence the coefficient was higher at this bearing than the others. In this case the difference between coefficients was a factor of two between the main bearing with the highest and lowest loading. Also, the measurements showed that the temperature within a single bearing could differ significantly between the front and rear of the bearing shell (looking in driving direction of the engine) - up to 6 C difference could be measured at one engine load.

Subsequently, the trade-off between oil pressure and bearing temperature was measured, where the oil pressure was reduced using an external oil pump. The goal was to validate the assumption that oil pressure is not a factor in oil temperature, while oil volume is. The pressure build-up is a characteristic of the hydrodynamic bearing itself, so long as enough volume is provided to the bearing. What was learnt from this test is that even a reduction in the oil pressure of about 50% led to a temperature rise of only 3 C on the main bearing temperature, which is negligible.

After these first tests on leaded bearing shells, the engine bottom end was dismantled and lead-free bearings installed, and the same test protocols carried out to determine the difference between leaded and lead-free under standardised conditions.

From these measurements it emerged that only a very small difference in bearing shell temperatures between leaded and lead-free bearing shells is to be expected. Temperature differences were not consistent between the different main bearings over the total rpm band range, but in summary a difference of about 2-3 C was measured.

In conclusion, these results lead back to the initial statement that no significant difference can be observed on bearing temperature between leaded and lead-free bearing materials. I acknowledge that these results might not be representative for all engines in all race applications, but I do feel they show that leaded bearings are not necessarily a must for racing, and that lead-free alternatives are up to the job.

As mentioned earlier, I would be very interested in any data-supported feedback from professional race engineers working in this area.

Fig. 1 - MAHLE Clevite bearings (Courtesy of MAHLE Clevite)

Written by Dieter van der Put

In addition to possible sensitivity issues concerning roundness (shape and steps in the circumference of the bearing bore) and the lower conformability and embedability of the unleaded, harder materials, the temperature inside the bearing is considered to be of major importance in the selection of the bearing material.

The temperature in the main bearing needs to remain within certain limits so as not to lead to deterioration of the oil film, in terms of its thickness and the oil quality. Above the temperature limit, the lubrication between crankshaft and bearing will quickly fail, leading to major engine damage. Racing engines will run up to 50% smaller oil film thicknesses than a typical OEM roadcar, which run in the region of 0.5 to 1 µm.

The oil temperature inside the bearing is a consequence of several factors, based on the main oil gallery temperature and mass flow, and depends on oil heat-exchanger capacity and the amount of oil going through it. It is this oil, which is fed to the bearings, where a temperature rise occurs due to friction loading, based on rotational speed of the crankshaft and crank train loading.

Oil temperature in the main oil gallery can be limited by increasing oil mass flow or oil heat-exchanger size, or both. The nature of the racing game, however, is that the heat exchangers are chosen to be as small as possible due to the tight engine and vehicle package in relation to aerodynamic boundary conditions, and the oil mass flow is as low as possible in order to achieve the lowest possible oil pump driving power, and reduce friction losses in the engine. Given these contradictory boundary conditions, compromises need to be made to get to the best overall performance.

Taking these factors into account, it becomes clear that the thermal conductivity of bearing shell materials is of major importance in ensuring that the heat transfer from the oil itself, through the main bearings to the crankcase, is achieved as efficiently as possible. Current developments on the lead-free alternatives have progressed significantly on the roadcar side, but given the more extreme application in performance engines, further steps need to be taken to come to a true and reliable replacement of leaded bearings, which still have the advantage over their lead-free alternatives.

It will remain the task of the bearing suppliers to think of and further develop lead-free bearing material specs, with or without applying coatings to them, to achieve improved thermal conductivity as well as sufficient conformability and embedability to be able to compete. The addition of coatings might support in achieving these targets, especially for running-in behaviour, as long as it is taken into account that coatings are not always known for their good thermal conductivity properties.

As usual, market demand will determine the efforts going into these developments to achieve feasible products. The question remains though of how much 'back-up' the developments get from the legislative directives. As mentioned in the former article on lead-free bearings, the EU End-of-Life Directive applies in principle to the OEM roadcar markets only, but this Directive cannot be neglected by the racing industry in years to come.

And although it will be possible to introduce measures to make a controlled separation of lead-containing materials from worn-out racing engines much easier in comparison to the roadcar industry, the development towards the removal of all lead-containing components will need to be continued in order to achieve the 'green' racing environment this industry needs to remain acceptable in the modern world.

The question will not be whether but when the racing industry makes the step towards removing all lead from its race vehicles, and whether this will be at its own initiative. Until then though, it is my expectation that developments here will remain a low priority.

Fig. 1 - MAHLE Clevite bearings (Courtesy of MAHLE Clevite)

Written by Dieter van der Put

Since the crankshaft system does not directly rotate in the bearing bore, we will look at the bearing shells in between. As John Coxon wrote in the May 2010 issue of Race Engine Technology

magazine (issue 46), in the past few years the 'End of Life' EU Directive has required that all lead be removed from the vehicle, no matter how small the amount. Where the Directive has 'supported' the automotive industry to develop lead-free type bearing materials, race engine designers have not yet transferred their focus to the lead-free alternatives. So let us look at what the areas of concern around lead-free bearing materials could be.

Bearing materials have their own specific properties regarding fatigue strength, rate of wear, corrosion resistance, conformability and embedability. In general, materials with harder, stronger properties are required for fatigue strength; softer properties are required for embedability and conformability.

The advantages of bearing shells containing lead are conformability and embedability. Particles can be 'bonded' into the softer material, and therefore do not lead to damage on the bearing shell and/or crankshaft. Conformability to local deformations with lead-containing bearings is less critical due to its soft mechanical properties. This article will therefore focus on the conformability aspect of the main bearing bore in relation to the lead-free alternative materials.

Conformability is typically shown on technical drawings with the cilindricity symbol and can be divided into two areas - shape and step.

The initial shape of the main bearing bore depends mainly on machining capability. In service, combustion and inertia loads will lead to deformations of the bore as well. Stiffness differences between cap and housing are significant parameters to achieve a homogeneous deformation, in order to be non-critical to the bearings.

Local steps can be characterised as sudden geometry changes over the bore diameter, and are mainly a consequence of damage or steps in the area of the split line. Local deformation in the area of the split line can consist of positioning tolerances during assembly of the bearing cap, where a positioning feature is typically used to align the cap with the crankcase.

Unfortunately, these positioning features require certain tolerances. Measurements have shown that a so-called cracked main bearing is dimensionally more stable over the split line in comparison to more traditional positioning methods such as pins, bushes or machined positioning surfaces.

After assembly of the bearing shells, the resulting roundness at the level of the split line remains a focus of attention during the running-in phase of the engine (and its bearing shells). During assembly, the slightly proud standing bearing shells are pre-tensioned, which will result in a radial force on the bearing bore geometry (called 'crush').

Depending on the stiffness differences between housing and bearing shells, a certain resulting geometry will be 'set'. To achieve sufficient clearance at the split-line location, the bearing ends are often tapered to some extent, creating a more or less oval resulting geometry.

The intersection between the tapered geometry and the nominal diameter of the bearing shell is the next critical zone. Contact between crankshaft and bearing shell edge can lead to local contact and initial wear of both components. Lead-free bearings with, for example, a lead-free bronze intermediate layer are significantly harder and therefore have less conformability, so bearing shell suppliers provide running-in layers on the bearing surface, with thicknesses of the order of 0.0001", to give significant improvements in the conformability behaviour of the bearing.

Ultimately, the 'End of Life' Directive will not miss out on motorsports. Given the need for it to become more eco-friendly, the race engine industry will need to work in accordance with the Directive. But questions remain as to which developments in lead-free bearing materials can be carried over from the regular automotive industry.

Although circumstances in race engine technology might be different from the roadcar industry, with additional developments on the structure and design of the main bearing area - as well as on bearing materials - it should be possible to make significant steps towards the widespread use of lead-free bearing materials. Who will be first though, and thus prove a winner?

Fig.1 - Roundness measurement of a main bearing bore

Written by Dieter van der Put

Recalling his days at Tyrrell, Renault's Formula One chief designer Tim Densham says, "Normally you would just open the SKF book and pick out standard bearings for most of the shafts." Things have moved on somewhat from those days, although even then some companies did make custom bearings for the teams.

Rule changes have played an important role in modern bearing materials in Formula One and, although introduced several years ago, the use of ceramic bearings is now more important than ever - where permitted. Christian Klatt, vice-president of sales at German ceramic bearing manufacturer Cerobear, says, "At least one component is always made from ceramics - typically the rolling elements, the balls or the rollers, and the races are made from special steel which is more advanced than conventional bearing steel."

Standard bearing steel, 100Cr6, is good for most bearing applications, and its quality has improved even in recent years through better cleanliness. But depending on the application or requirement, there are improved bearing steel grades available that offer better temperature resistance, corrosion resistance and particularly fatigue life.

In motorsport - perhaps more than in any other application - weight and space are critical and, as a result, bearings are operated at Hertzian stress levels that one would never see in industrial applications, sometimes above their static load capacity. So there is a demand for higher grade bearing steels in motor-sport (such as Cronidur 30 and X 40 CrMoVN 16 2) that enable designers to downsize the bearings, maintaining the same service life or even increasing it.

High Nitrogen Steels (HNS), which have a higher nitrogen content and a lower carbon content, were introduced into motorsport bearing applications in 1999. They have fewer carbides, and carbides have a positive effect, giving strength to the material. On the other hand, if they are too big and their proportion in the structure is too high, they can act as sources of fatigue failure.

Hybrid ceramic/steel bearings have outperformed standard bearings in various tests that show a ten- to 100-fold improvement in service life, depending on the testing conditions. But in Formula One in particular there has always been a demand for improved solutions, so research is being carried out to develop a new bearing steel that is even better than HNS.

With steels, however, 'superiority' does not necessarily mean greater hardness for them to be better. HNS (HRC 58-60) provides the same or an even lower hardness than 100Cr6 (HRC 58-62), but it is the structure and its resistance to Hertzian stresses that is important, so it is this that has been the focus of development to achieve a material with a superior, finer and more homogeneous structure. The result should be a better wear and overrolling resistance; (this is resistance against being rolled over, as opposed to rolling resistance which is friction), which results in a longer fatigue life, particularly under compromised lubrication conditions.

Brittleness could be the downside, although this is not yet proven but still likely, as brittleness in steels often rises with hardness, which is why the development of this new steel is aiming for a finer structure, as this would reduce brittleness. But brittleness in general would not pose any problems for rolling elements in hybrid bearings and plain races in ceramic bearings, although this would not be recommended for some special types of highly loaded flanged bearing races that depend on the load and geometry of the application. Si3N4 balls are used in highly shock-loaded applications such as wheel bearings, gearbox and diff bearings, with very few problems.

It would be wrong therefore to give the impression that operating hybrid bearings is a risk because the materials are brittle. If this were the case, ceramic bearings would not be used in applications from Formula One to the Space Shuttle's main engine.

Current Formula One rules require that a gearbox is run over four races before it can be replaced, so it is important to know that the unit will last the course. As there is no testing allowed, a used 'box may be installed for Friday practice but, according to Densham, bearing failures are not unheard of.

Densham says, "We have had several analyses done on the bearing loads and we know which ones are the heavily loaded ones. If you have a problem, the first thing is to squirt a bit more oil in there, and that is something we can do quickly, whereas to have a new bearing made takes a fair amount of time."

Fig. 1 - Double-row crankshaft bearing
Fig. 2 - Flanged hybrid bearings
Fig. 3 - Selection of needle and roller bearings for engine, differential and gearbox
Fig. 4 - Valvetrain bearings

Written by Glen Smale

The rolling element bearing does consume some power. While mechanical losses are worth a little more than you might think, there are actually two gains made for every reduction of mechanical loss implemented.

Every horsepower retained through eliminating loss is a horsepower of extra power available from the engine. But there is also an additional gain made because every inefficiency comes out of the engine as heat and a good proportion of that waste power has to be removed through the radiators - either water or oil.

Taking a clean sheet approach to engine design, there is a predictable method of confronting the limitations of rolling element bearings and how far they can be pushed. For a given load requirement, a smaller bearing can help tremendously with the detail packaging of the drives or the ancillaries, and can help shrink the size and therefore weight of many parts of the engine.

The degree to which the load is shared out between adjacent balls or rollers is a function of the precision within the bearing. Any variation in size from one ball or roller to the next will result in an uneven sharing of the load and will give rise to larger peak loads than if all parts were exactly the same size.

As a consequence the greater the precision in the manufacture of the bearing, the better the load carrying capacity. Indeed in top line motor sport where works teams might be working directly with bearing manufacturers, they could go one step further and actually measure and grade elements to assemble into hand-built matched component bearings. This then gives as near even load distribution as possible.

The same goes for non-circularity in either race. This is a function of the precision of manufacture of the races, but also the uniformity of the bearing housing and shafts. This is also a big issue because as big loads go through lightweight structures the resulting deflections will tend to make the bearing races deflect.

While we normally find the amounts of deflection microscopic, when you are trying to share the loads between relatively stiff rollers or balls, the loaded shape of the housings and shafts has to be considered.

With a too-tight fit on either race, a builder potentially closes up the running clearance in the bearing and increases the loads too much. Finally, the effects of running temperatures need to be considered on the bearing system, focusing not only on heat and cold, but also on any transient running conditions in the case where one component might heat up faster than others.

With the rolling element bearing, one must also examine the prospects of micro-welding. As the element continues to move on, the surface is torn apart resulting in adhesive wear. Most people believe that micro-welding only occurs between similar materials, but this is not the case. Micro-welding can even take place between steel and ceramic surfaces with some combinations of load, temperature and lubrication.

Fig. 1 - Rolling element bearings from Cerobear

Written by Anne Proffit

Bearings reduce friction by providing smooth/polished metal balls or rollers, and a smooth/polished inner and outer metal surface for the balls to roll against. These balls or rollers carry the load, allowing the device to rotate smoothly.

In general a rolling element bearing is a bearing which carries a load by placing round elements between two surfaces. These surfaces are referred to as the inner race and the outer race. The relative motion of the races causes the bearing elements to roll, with little or no sliding. Bearings are normally selected on the basis of a requirement to carry a given load for a given period of time. Rolling contact bearings are designed to carry pure radial loads, pure axial loads or a combination of the two.

The bearing designer is confronted with the problems of designing a group of elements which make up a rolling element bearing. Parameters such as fatigue, loading, heat, corrosion resistance and lubrication, to name but a few, must be considered. There are many types of rolling element bearings, each designed to carry a specific kind of load. Each of the different types of bearing contain either a ball bearing, a roller bearing or a needle type bearing. A needle bearing is an elongated roller bearing.

Rolling-element bearings may rotate at over 100,000 rpm. Maximum rolling element bearing speeds may be specified in 'DN', which is the product of the diameter (in mm) and the maximum revolutions per minute (rpm).

There are also many material issues for bearings. For example, a harder material may be more durable against abrasion but more likely to suffer fatigue fracture. Therefore, the material varies with the application, and whilst steel is the most common for rolling element bearings, plastics and ceramics are also in use.

A bearing can last indefinitely; longer than the life of the machine, if it is kept clean, lubricated, and operated within its load rating. Also, every effort needs to be made during manufacture to make sure the bearing materials are sufficiently free of microscopic defects. Note, that cooling, lubrication, and sealing are also important parts of bearing design.

The operating environment and servicing needs must also be considered in bearing design. Some bearing assemblies require routine addition of lubricants, while others are factory sealed, requiring no further maintenance for the life of the bearing or assembly. Although seals are appealing, they increase bearing friction, and a permanently sealed bearing may have the lubricant contaminated by hard particles, due to bearing wear, which will abrade the bearing.

Fig. 1 - Schematic diagram of a rolling element bearing.

Written by Eric Smart

The bi-metal bearing is constructed of only two layers, consisting of a steel backing with an inlay of an aluminium alloy. Typical bearing materials used in the construction of bi-metal bearings are a combination of the following materials: Tin, silicon, lead, copper and aluminium. The bearing material may sometimes be described as aluminium-silicon. The silicon is what bearing manufacturers have determined make aluminium a preferred material for modern bearing applications. The small particles of silicon, are finely and evenly dispersed throughout the bearing material, and since it's an alloy, it can maintain its properties throughout the entire bearing thickness resulting in a more consistent, reliable performance.

Debris is a major cause of bearing failure and it can be left over from machining operations or due to wear of the component parts. Debris can get trapped between the journal and the bearing surface and lead to damage of the bearing surface and/or scoring of the journal, and could ultimately lead to failure. Debris is an important point with bi-metal bearings due to issues over embeddability.

Bi-metal bearings are considered harder than tri-metal bearings. The surface of the bearing contains hard and soft areas where the soft areas allow embeddability, whilst the harder areas do not. Embeddability is the ability to embed hard particles and foreign debris within the bearing material. However, due to the hard and soft nature of the bearing surface, it is possible for debris to become trapped between the bearing surface and the journal leading to wear. Therefore, to counteract this effect, the bi-metal bearing is machined with micro-grooves to retain the hard particles, these particles eventually being washed away by the lubricating oil.

Note that the different combinations and quantities of alloying materials, as used by the different manufactures, will provide different bearing characteristics. Whereas the aluminium-silicon alloy may produce a hard bearing with partial embeddability, another aluminium alloy mix may provide a bearing that offers much improved embeddability and higher load carrying capabilities.

Bi-metal bearings do possess conformability, which is the ability to adjust to minor misalignments during assembly.

In general tri-metal bearings are said to have a greater load carrying capacity then aluminium bi-metal bearings. In terms of bearing selection, that is, whether to use tri-metal or bi-metal bearings, it may come down to how much load the bearings have to withstand and the life expectancy of the bearings. For example, bi-metal journal bearings may be used in a high load application for a short period of time; for example a drag car, having one or two runs. If high load carrying characteristics and longevity are required, then tri-metal bearings may be selected. This example of bearing selection is not meant to be exhaustive, but to illustrate that a trade off between the two bearing types has to be made. Note also that cost will also be a factor.

In comparing bi-metal to tri-metal in severe load applications, there will no doubt be some cases where the bi-metal bearings will outperform those of tri-metal. If the roughness of the sliding surfaces is reduced, the load carrying capacity of the bearing is increased. Hence concentrating on bi-metal bearings in this way, could provide load carrying benefits.

The future trend for journal bearings may be headed toward the bi-metal type, but tri-metal bearings are still in demand. In all probability, bi-metal and tri-metal journal bearings may continue to co-exist for quite some time.

Aluminium alloy journal bearings are not new. What has propelled them forward are possibly the latest methods of manufacture and the inclusion of materials such as silicon. Bi-metal bearings have the benefit of higher melting points and greater fatigue surface strengths, than tri-metal bearings.

Fig. 1 - Typical bi-metal journal bearing

Written by Eric Smart

Journal bearings operate between the extremes of no lubrication on the one hand, to thick-film lubrication on the other. Therefore, bearing material must possess compressive strength in order to withstand the loading due to the gas pressure and also fatigue strength to withstand the cyclic loading of the rotating journal. It must however be soft enough to allow foreign particles to be embedded and yet resistance to wear, resistance to corrosion and a low coefficient of friction are also important considerations.

All bearing materials have a pressure-velocity operating envelope, within which they function safely. The maximum pressure that the bearing can accept is determined by the hardness of the surface while the maximum velocity (of the journal within the bearing) is determined by heating within the bearing. This heating may be due to friction heating caused by metal-to-metal contact, as well as due to shearing of the oil film. The velocity is therefore restricted by the thermal conductivity of the bearing material.

Crankshaft journals are case hardened to the extent where they can abrade the foreign particles. The bearing surfaces are kept soft as this allows the foreign particles to become embedded in the soft bearing material and reduces the probability of wear elsewhere in the engine. The soft material should have a low modulus of elasticity, to enable it to deform within its elastic limit and therefore to absorb the hard particles. This property of the bearing material to absorb hard materials, is known as embeddability. Another property of the soft bearing material is known as conformability which allows for slight misalignments of the journal and the soft bearing material.

Due to the mechanical properties of the soft bearing material, it should be squeezed out due to the forces acting upon it. However, to prevent the soft bearing material from being squeezed out, it is made thin and is then supported by a much harder material. The rationale behind the thin layer of soft material is as follows: Consider a thick layer of plasticine between two blocks of wood under pressure – this leads to the plasticine being easily deformed and squeezed out of the sides. As the plasticine gets thinner, more pressure is required to cause it to flow laterally, and subsequently more pressure is required to deform it. The thin layer of plasticine that remains between the blocks cannot be squeezed out altogether. That would require an infinite pressure.

This effect of not being able to squeeze out all of the soft material is known as the principle of plastic constraint. It is used in bearing design and enables a very thin layer of soft metal to be deposited onto the hard bearing material. This layer is thick enough to embed dirt particles yet thin enough to support the journal forces without being squeezed out.

The soft bearing material acts like a lubricant when there is little or no lubrication, and due to its low melting point, it may also melt locally to prevent seizures occurring. If the layer of soft metal did not exist, frictional heating due to metal-to-metal contact would lead to a much higher bearing temperature. This would lead to gross atomic bonding between the journal asperities and bearing asperities followed by seizure. All things being equal, this would lead to bearing damage with little or no damage to the crankshaft journal.

Because the thin overlay of bearing metal can be worn away before the end of the normal operating life of the bearing, it is normal to have a second thicker, harder layer of bearing material beneath the soft thin layer which is bonded to a steel backing strip. In the event of loss of the soft, thin layer, the harder layer can still protect the journal.

If the roughness of the sliding surfaces is reduced, the load carrying capacity of the bearing is increased. Hence the requirement for a very good quality finish of the bearing surfaces is an important means of increasing load capacity, reliability and service life. Surface treatments can be used to improve the quality of the journal surfaces.

Written by Eric Smart.

When two surfaces roll or ‘roll and slide’ in contact with one another with sufficient force, maximum stress is developed beneath the contacting surface. A sub-surface crack appears which is initiated by this maximum stress. The crack then propagates to the surface during operation. The lubricant under pressure, enters this crack and then assists in the formation of surface pitting. This pitting constitutes a fatigue (fatigue is defined as slow crack growth) type failure and would be expected to occur in the journal bearing.

Hence, knowing the surface fatigue strength of contacting materials is quite important in order to resist this pitting type fatigue failure and is obtained by testing. The test results are made available in the form of a load stress factor or a wear factor. Hence the surface fatigue strength may be defined as the number of cycles at which the first tangible evidence of fatigue (pitting of the surface) is observed. Such failures are often called wear because they occur over a long period of time. This type of wear is not to be confused with abrasive type wear. Wear, which is normally divided into two types, namely adhesive wear and abrasive wear, is generally undesirable because it could lead to increased tolerances and possibly catastrophic failure.

Adhesive wear is associated with the adhesion between atoms of the asperities of the two contacting materials where, generally, wear fragments will be removed from the softer material asperities. Asperities (surface roughness) are like peaks of a mountain range which are present on the journal and journal bearing surfaces. If work hardening extends into the asperity, the size of the fragment sheared off will be larger. In order to minimise the size of the wear fragment either decrease the loading between the journal and the bearing or increase the hardness of the journal.

Abrasive wear occurs when, for example, a hard particle digs into the softer material. These wear particles can come from the wear particles due to adhesive wear. These dislodged softer particles become oxidised (due to oxygen in the lubricant) converting them into hard oxidised particles which then rub against the surfaces and cause wear. The journal bearing wears the most and is then replaced. Note, these former asperities are not the only source of wear in an engine as the products of combustion, dirt and dust particles also contribute to journal bearing wear but such abrasive wear can be reduced through good filtration.

Frictional forces overcome friction and are undesirable in bearings because of the power they absorb. When two surfaces are in contact with each other, the asperities of one surface sit on top of the asperities of the other and the resultant friction is due to the sliding over (shearing) or breaking of the contacting asperities. Surfaces even when polished and ground have asperities. The load pressing the surfaces together is supported solely by the contacting asperities and as the contact area of each asperity is very small, the stress on each asperity is therefore very large. At low loads the asperities deform elastically but at high loads the asperities deform plastically and require more energy to shear them.

The work done in overcoming friction is mainly converted to heat at the sliding surfaces. In order to minimise these frictional forces, it is necessary to coat the asperities with polymers or soft metals. Many journal bearings have a thin film of soft metal on the surface of the bearing, for example, white metal. The softer metal forms a lubricating film and has a relatively low coefficient of friction, which should be low enough to prevent seizure in the case of boundary lubrication, or no lubrication at all.

Written by Eric Smart.

Clearances should be small enough to obtain the maximum load carrying capability of the bearing. If the clearance is too small the bearing temperature will be too high and the minimum film thickness will be too low. As the bearing wears, the effect on the bearing performance must be considered, as this leads to a decrease in the bearing temperature and an increase in the flow of oil through the bearing with a knock on effect on hydrodynamic and boundary lubrication.

Lubrication

The success of modern journal bearings is due to the understanding that lubrication, the function of which is to separate the surfaces, is an integral part of journal bearing design. A journal bearing may be subject to four types of lubrication, namely: Hydrodynamic, hydrostatic, boundary and solid film. However, only hydrodynamic and boundary lubrication will be considered here.

Hydrodynamic lubrication does not depend upon the introduction of lubricant under pressure, but it does require an adequate supply of oil at all times. The oil film pressure is created by the moving surface dragging the lubricant into a wedge shaped zone at a sufficiently high velocity to create the pressure to separate the surfaces. Heat is generated within the bearing due to the work performed by shearing of the oil film as the journal rotates in the bearing. A pressure fed system is used to force greater oil flow through the bearing to improve cooling. Hydrodynamic lubrication is sometimes referred to as full film, fluid, or thick film lubrication.

Boundary lubrication occurs when the build up of full film lubrication is not possible. This may occur with a change in operating conditions leading to very high bearing temperatures, or during periods of start up and shut down. Bearing wear which results from thin-film lubrication may also be improved by the inclusion of additives to the lubricant or surface treating the bearing. When a bearing operates under hydrodynamic and thin film conditions together, mixed film lubrication is said to exist.

Instability of the oil film known as Oil-whirl or Oil-slip has not been considered here, while solid film lubrication has also been ignored.

Bearing Materials

A bearing material must possess compressive strength to withstand the gas pressure loading and fatigue strength to withstand the cyclic loading of the rotating journal and temperature variations. It must be soft enough to allow for wear and the embedding of foreign particles, but it must also have a low modulus of elasticity to allow for deformation, within its elastic limit. The rate of wear of the bearing material and a low coefficient of friction are also important considerations while other considerations of bearing materials include reliability and corrosion resistance.

Typical bearing materials for high performance engines would be copper-lead and lead-bronze. Some copper-lead bearings may be given a thin coating of pure lead followed by an even thinner film of indium. The lead and indium diffuse into one another, providing a coating to the copper-lead bearing. This coating is intended to last the life of the bearing, assists in the running-in process and provides surface protection.

Written by John Smith.

On a mile-and-a-half oval like Chicagoland Speedway, where the CWTS raced the final weekend of August, PME owner Mark Smith acknowledged that he is “always fighting temperatures because they never lift here. They’re running wide open throttle all the time, so the temperatures are always up and you always have to control your oil temperatures. We used the tapered spacers here and they take about 200 horsepower away, giving us 8000 rpm maximum for about 700 horsepower.”

Of course with lower speeds he can play with his clearances a little bit. “It’s like all stuff; the key is making sure we keep everything cool and look at our oil flow, making sure we’re in the areas we want because you don’t want the oil to take power away,” Smith told me at Chicagoland.

According to Bill McKnight of MAHLE Clevite, PME Engines uses the MB3229V, a 2.017-inch main set with the Vandervell lead-indium overlay. “It is very popular in NASCAR because this set allows teams to transition from the 2.017-inch shaft diameter to the 2.00-inch shaft and utilize the same block.” The mains are available in X, standard, 0.017”, 0.018”, 0.019” and 0.010” sizes and McKnight said PME usually takes trays of 25 pieces per order.

PME has experimented with bearing coatings in the past but Smith doesn’t feel it’s terribly critical now because these particular bearings he’s been using – after careful research – are “almost bullet-proof. We coat every once in a while because you always go back and try something again. But for the most part, we leave it alone and let it do its job.”

For rod end bearings, PME specifies the CB1798H, which has a 1.850-inch shaft diameter and measures 0.709-inch wide. “This H series bearing has a thin Babbitt overlay that is slightly harder than the lead-indium overlay of the V material,” McKnight said. “It also has a medium amount of eccentricity and a high level of crush, suitable to the NASCAR application for which it is intended.”

Smith says he runs his bearings for about 350-400 miles. “There’s been times when we run them a secondary time,” he revealed. “You measure the spread and you measure to make sure you still got the clearances and that the departing lines are correct. For the most part, 1000 miles is about as far as you run them.”

He appreciates the fact that his supplier is proactive with the size bearings he needs. “They have minimum diameters on the crank so they’ve come up with some new bearings that have helped us out. The housing for the con rod is a standard size now,” Smith said.

Written by Anne Proffit.

In all the nitromethane-burning NHRA classes, it’s “okay to have a high and long cylinder pressure curve, but when we experience detonation, it goes right off the map,” Medlen explained. He likes to see about 12,000 PSI, “but when it detonates the cylinder pressure spikes to 26,000!” Every bearing manufacturer has methods of dealing with this cylinder pressure. “It depends on the material.”

Medlen uses a three-layer ‘V’ bearing rather than an ‘M’ series unit. “It’s all about balance. A bearing that is too hard ends up hurting the crankshaft and one that is too soft won’t withstand the loads we see in our engines. We have to get the maximum out of the bearing without killing the crank; it has to be durable enough not to overheat the crankshaft and hard enough to last under these severe conditions” of a 1000-foot trip down the dragstrip in just over four seconds.

The ‘V’ bearing is stronger and has more density to give support to the crankshaft. The lead-iridium surface layer provides both embeadibility to hold trash and flexibility to handle crankshaft and housing bore deflection seen in these conditions. Most nitro runners are using the V-series of engine bearings these days and most, like Medlen, swap out their uppers every run. “We swap out the lower bearings every 6-8 runs,” he said.

The latest development from Mahle Clevite is in coatings. “They have a proprietary coating that keeps the surface temperature down and doesn’t erode the bearing,” Medlen said. He used bearings with this coating for the first time at Seattle on Neff’s car and also in the A Fuel dragster category, where Courtney Force won her first Wally. “We are pretty excited about what we see,” Medlen said.

While the coating is not necessarily new, this is the first time it has been used in the nitro-burning categories. This manufacturer experimented with dry-film bearing coatings on nitro engines but found the standard bearing coatings used so successfully in NASCAR and other racing venues were unable to withstanding the extreme wiping action of the crank. “The standard coating was often completely gone in one pass,” said company spokesman Bill McKnight, “making its effectiveness questionable at best.”

Working with coating supplier H.M. Elliott Inc., the well-known bearing manufacturer has developed a “new coating” that is substantially more durable and shows real promise to bearing and crankshaft life extension in nitro-powered machines. “Our initial feedback has been extremely positive. The new coating is far more durable and we’re pretty excited about the possibility of offering a better bearing to this select group of customers.”

In addition to Medlen, Jim Oberhofer of Kalitta Motorsports, perennial Top Alcohol Funny Car champion Frank Manzo and Terry Haddock, owner/crew chief/driver of his own Top Fuel NHRA dragster have been making test runs with these coated bearings.

McKnight anticipated testing and development continuing into the autumn racing season with the possibility of having the new coated bearings available to all customers before the Auto Club NHRA Finals at Pomona in November.

Written by Anne Proffit.

Still, the Plymouth, Michigan manufacturer has interests beyond the scope of recognized motor sport. SKF is one of the major backers of the local University of Michigan solar-powered race car. The University of Michigan solar team’s Continuum competed in an 1800-mile race across the Australian outback in 2008

and placed first in the 2500-mile North American Championship – its fifth victory in nine races.

Currently the University of Michigan team is preparing for October’s World Championship race across Australia, hoping to become the first American squad to win the World Championship since General Motors’ entry won the inaugural, 1987 contest.

According to Kenneth C. Fegely, business engineer for SKF’s Racing and Engineering Services, this competition was an easy fit for his company. “We have been in the business of reducing friction and increasing efficiency for over 100 years,” he reminded me.

“We see the practical application of solar technology as one area of the future. It is not just the solar portion that interests us but the extremely efficient and clever use of the limited energy to get as much practical work as possible from it.”

The solar car has an in-hub electric motor with an efficiency of over 95 percent, according to Steven Hechtman, team leader for the Ann Arbor-based University of Michigan solar racing team. “Our team places a large focus on reducing any friction within the motor that would decrease this efficiency,” he said.

Working together since 2007, “SKF provides us with low friction deep groove ball bearings (DGBB) for our motor and wheels. These ball bearings exhibit extremely low amounts of friction, ensuring that the solar car is able to cruise with minimal resistance,” Hechtman explained.

The new Infinium race car has undergone more than 1000 miles of testing, using the hybrid DGBB with low drag seals. Because the drive wheels are the electric motors, “Our bearings are not only supporting the rotor; they also support the weight of the car and driver, and they handle the manoeuvring forces as well,” Fegely stated.

“When we first constructed our motor with SKF bearings, we were immediately impressed by the superior performance,” Hechtman told me. His previous edition motor without the SKF product was “only able to spin freely around its axle for a matter of seconds, while the new motor with SKF bearings spins for a number of minutes before coming to a rest.”

The Infinium solar racer is a three-wheeled vehicle with one drive wheel and two non-driven wheels, each with two DGBB’s. The bearings used on the [rear] driven rotor/wheel are a basic size 6206, which computes to 30x62x16 mm, Fegely told me. “The non-driven wheels use something smaller: a 6004 20x42x12 mm bearing.

“These are a Hybrid deep groove ball bearing with low drag seals, low drag cage, special internal clearance and wide temp lubricant. We chose the bearing specifically for its load carrying capacity and excellent rolling characteristics.” The hybrid term designates a steel ring with ceramic rolling elements. “The silicone nitride balls are harder, lighter and require less lubricant to operate than a standard conventional all steel bearing.”

Despite being comprised of dissimilar materials (steel and ceramic), the bearings have no tendency to micro-weld to each other under loading, as occurs with steel on steel. “This reduces the requirements of lubricant and allows a lower viscosity lubricant to be chosen, helping reduce parasitic losses due to the grease,” Fegely explained. “The low drag seal strikes the right balance of keeping contaminants out, yet reducing friction due to seal lip contact.” It is not possible to run the bearing without a seal, as the vehicle operates in all environments during competitions.

During the Australian race, vehicles travel through the outback where they are subject to dust storms and temperatures of 110°F. Although the race route itself is fairly straight, competitors do need good cornering capability during qualifiers for the race, which occur at the Hidden Valley Speedway, a road course where Australian V8 Supercars compete. They also test their car at NASCAR tracks like Michigan International Speedway, though the American Race is also on stretches of long, flat roads.

Will this technology morph into mainstream motorsport? While many aspects of the DGBB could transfer to another racing application, Fegely believes it is the “process that you use exactly what you need for the requirements,” that should be paramount in choice. “What works great in a solar racer may well fair miserably in a rally car – although some of the features like ceramic balls, low drag seals – may transfer over nicely.”

With service life of 5000 miles and most solar races being conducted in half (or less) that distance, the DGBB as used in the University of Michigan solar Infinium race car performs the three jobs any bearing is subjected to: support, locate and reduce friction. We will have to wait until October to see how it fares in this competition.

Written by Anne Proffit.

In a good example of a reverse technology flow - series production to racing - bearings originally developed to cope with extreme cylinder pressures found in the latest generation of high speed passenger car and commercial diesel engines have found their way into Formula One and sport car engines.

In the mid 90’s, vehicle manufacturers found that the bearing materials available to them that used bi and tri metal construction processes were unable to provide the durability required for full engine life when the cylinder pressures exceeded 160bar. Enter ‘sputter’ bearing technology.

Sputter materials and processes for engine bearings were first developed by Glyco, part of the Federal-Mogul engine components group. The sputtering process produces a material that combines the high wear-resistance properties of an aluminum-tin sliding layer with the extremely high-load withstanding capacity of a cast copper-lead-bearing metal layer.

Sputter overlays are deposited by Physical Vapor Deposition (PVD) method, utilising argon ions for bombarding a cathodically connected target, made of the final bearing coating material, normally Al20Sn or Al40Sn.

The sputter process starts with the base bearing shells being loaded into a vacuum chamber which is then evacuated before the introduction of argon gas. The gas is then heated until it enters a plasma state causing positively charged argon ions to form.

Then follows an etching and activation stage during which stage the bearings become the cathodes and the argon ions bombard the bearing substrate to both clean it and provide good adhesion for bearing surface material.

In the next step the current is reversed and the bearing becomes the anode and a nickel or nickel chrome (Ni or NiCr) diffusion layer is deposited on to the cathode target. The argon ions bombard the target knocking the atoms of nickel (nickel and chromium) out from the target. The atoms moving off the target meet the bearing – now the anode - substrate surface and stick to it, creating a diffusion layer of 0.00004”-0.00008” (1-2 µm) in thickness.

Finally, the working surface overlay is deposited. At this stage the cathode is a target made of an aluminum-tin alloy, Al20Sn or Al40Sn. Atoms of the target are knocked out by the high energy ions and are deposited on the substrate surface forming the finished AlSn overlay.

The sputtering process provides an extremely homogeneous distribution of tin within an aluminum matrix. Hardness of aluminum-tin sputter material is about 90 HV, which is three times higher than hardness of aluminum-tin alloy prepared by conventional methods (casting). Cast copper based bearings or high strength aluminum based bearings are commonly plated by sputter overlays.

Load carrying capacity of sputter bearings is highest of all bearing materials, being in the region of 100-120 MPa (14500-17400 psi).

There are however some disadvantages with sputter bearings. High production costs caused by the slow deposition process and low soft anti-friction properties due to poor material compatibility, conformability, embeddability – the ability of a bearing to allow small particles to embed themselves into the soft bearing surface, thereby cause no further bearing damage – have typically confined the use of sputtered bearings to premium and high performance niche markets.

However, a significant cost reduction can be achieved by a combination of a sputter bearing shell in the high load conrod position, with a common tri-metal bearing shell being used in the less loaded cap position.

The physical vapor deposition process produces an exceptionally uniform alloy matrix with superior wear resistance, capable of withstanding the very high loads found in very high speed gasoline engines and the current generation of racing diesels such as the Audi V10 and Peugeot Le Mans series engines.

Written by David Wood.

“Clevite Nitro V series bearings maintain that fine line of being hard enough to withstand cylinder pressures of 10,000 psi, yet soft enough to tune the engine.”One key to successfully running a nitro-powered drag racing engine is the ability to “Use the upper rod bearing as a tuning tool. After every run we measure the thickness to see how hard that cylinder was working,” down the 1000-foot drag strip.“For example, the stock shell measures 0.0625-inch thick. If after the run the shell measures 0.0535-inch thick (a difference of 0.009-inch), that would indicate that the cylinder is working harder than we would like so we would make the jet sizes larger in that cylinder.” The opposite would hold true if the shell was squeezed only 0.0005-inch, meaning Antonelli was likely not working the cylinder quite hard enough.Antonelli said that the same style main and rod end engine bearings have been the JFR team’s choice for at least the last three to four years. There have been recent developments to increase bearing life. “We work closely with Clevite to ensure we know as much as possible about the dynamics of our engine bearings, the oil film, the firing load and deformation that occurs to the entire rod-crankshaft-bearing assembly when the cylinder fires,” Antonelli said.Extreme surface damage to a bearing can inflict permanent damage to the crank; materials advancement in Clevite rod journal bearings often allow Antonelli’s crew to “hit the crank with some 400 grit emery cloth, leave it in the car and do battle again.”“If we suffered severe damage to the surface of the bearing, the JFR crew would need to replace the crankshaft. And that [crankshaft] is a $3500 item these days!”The JFR teams carry sufficient main and rod end bearing inventory for five to six events on its trailer. “We change the lower shell on main bearings two through five every run and the upper shell on two and four, then three and five every third run alternating,” Antonelli told me. “The upper rod bearings are changed out every run and the lower (cap) rod bearing every other run.”

The MAHLE Clevite bearings used by JFR are widely used in alcohol and nitro-burning NHRA classes. The CB1512V rod bearing is the mainstay for Hemi-powered engines, according to MAHLE Clevite’s Bill McKnight.This bearing has an overlay of Vandervell’s lead indium design with an intermediate layer of cast-copper-lead applied to a steel back. It is specially made for the Hemi engine that is the basis for most nitro-class engines.Two different main journal sizes are available and legal for NHRA nitro classes. About 80 percent of engine builders use the traditional 2.750” size while the balance opt for a 3.00” main. “It is a matter of personal preference,” McKnight said.All the JFR Funny Car teams run the 3.00” mains. The JFR BOSS 500 Ford Top Fuel engine is campaigned by John Force and Mike Neff; Force Hood and Robert Hight currently run the Hemi style engine.McKnight points out that, while the BOSS Ford uses the same 3.00” main bearing journal size as the Hemi, a different main shell is required for the Ford housing bore.“Both the BOSS and the Hemi three-inch shafts utilize five straight-shell bearings and thrust washers instead of a flanged thrust bearing,” McKnight noted. The thrust washer (TW120S) is the same for both the Ford and Hemi three-inch mains.