With the chemical symbol MoS₂, this black crystalline compound occurs as the mineral molybdenite, the principal ore from which molybdenum metal is extracted. It is commonly used as a solid lubricant, thanks to its low-friction properties, which are similar to those of graphite, as well as its high load-bearing capabilities and the fact that it is relatively unreactive, being unaffected by dilute acids and oxygen. Most usefully when it comes to engine applications, it has good thermal stability, up to 350-400 C in an oxidising environment.

Molybdenum disulphide was first discovered more than 250 years ago, when the lubricating properties of an unknown ore were noted in 1744 by Johann Alexander Cramer. The ore was similar to lead, galena and graphite, and these substances were labelled with the Greek word ‘molybdos’, meaning lead-like. In 1778, a Swedish scientist named Carl Wilhelm Scheele identified molybdenite as the sulphide of a distinct metallic element by heating it to yield a white oxide powder. At his suggestion, Peter Jacob Hjelm, another Swedish scientist, successfully isolated the metal in 1782 and named it molybdenum.

The first use for the material was as a strengthening agent in steel production, a use to which it is still put, but it was not until 1935 that it was used for its lubricating properties. A German engineer, Alfred Sonntag, had designed a huge machine to simulate aircraft vibrations, but it failed due to friction between the moving parts. He tried many lubricants to solve the problem, but none had sufficient load-bearing capability to be effective. However, he came across an 18th century text that mentioned the lubricating properties of molybdenite, and on using it as a lubricant found it to be highly effective. After this discovery, Sonntag developed a method of purifying molybdenite, which contains traces of quartz, into the powdered lubricant that is in use now.

Molybdenum disulphide takes the form of microscopic hexagonal platelets, with several molecules making up each platelet. These platelets are attracted to metal surfaces which, when combined with sliding force between metal parts, results in a thermo-chemical reaction, creating a protective coating of MoS₂ on the parts in question. This coating can withstand pressures of about 500,000 psi, and as such makes MoS₂ an attractive option for use on components where boundary lubrication is an issue, such as the interface between camshafts and tappets.

While the application of MoS₂ as a dry-film lubricant was established in the mid-20th century, using it effectively in oil took longer to perfect. The problem was that the particles would not stay in suspension in oil, leading to the particles forming a sludge that could block oil passageways (while also negating the material’s lubricating benefits). However, once methods were found to prevent this, MoS₂ has proved to be a highly effective anti-wear additive. For example, tests undertaken at the Argonne National Laboratory, in Illinois in 2012* showed that adding MoS₂ nano-particles, 50 nm in size, to a polyalphaolefin base oil showed significant reductions in friction between the piston skirt and cylinder liner on heavy-duty industrial engines. These same benefits can also be realised in transmission oils, where boundary lubrication is far more common.

As with any technology though, the use of molybdenum disulphide as an oil additive is not a silver bullet. For example, it only acts a friction reducer under boundary lubrication conditions; in hydrodynamic and full-film regimes, the particles do not come into play and some studies have even shown that they can actually marginally increase friction. However, provided its limitations are recognised, it can have considerable benefits when used correctly.

* Nicholaos, G., Demas, Elena, V., Timofeeva, Jules L., Routbort, George R. Fenske, “Tribological effects of BN and MOS 2 nanoparticles added to polyalphaolefin oil in piston skirt/cylinder liner tests”, Argonne National Laboratory, 2012

Written by Lawrence Butcher

As its name suggests, the primary role of the scavenge system is to scavenge oil from the engine and pump it to the oil tank, ready to be pumped back into the engine by the pressure pump. In most dry-sump systems, particularly those that are additions to previously wet-sump engines, one pump body will take care of both pressure and scavenge operations. From a packaging perspective, this can be less than ideal, particularly within the tight confines of a Formula One engine. These engines, whether they be the previous generation of V8s or the new V6s, therefore tend to have separate pumps for scavenge and feed duties.

In the first iterations of Honda’s V10s the system relied on a single-stage scavenge pump. This drew oil from a sump pan that was split into chambers to reduce pumping losses associated with the movement of air and blow-by gases from cylinder to cylinder, caused by the rise and fall of the pistons. In later iterations of the V10, and then the V8, Honda moved to a multiple scavenge pump set-up as this was shown to provide a more efficient solution, scavenging more oil for less power consumption.

In effect this meant that each sump chamber had its own dedicated scavenge pump, which then fed into a common centrifugal oil-air separator before feeding into the oil tank. The layout Honda settled on until the end of its involvement with Formula One featured five scavenge pumps located on the right-hand side of the crankcase and a single oil feed pump on the left-hand side of the front of the motor.

On particular problem the team encountered during the development of the scavenge system was ensuring the consistent presence of oil at the pump inlets. Using the analogy of a straw, if the end of the straw is not completely immersed in liquid then it is almost impossible to suck the liquid up. The same applies to the inlets of the scavenge pumps, which have to be completely immersed in oil to operate.

Initially, Honda’s scavenge pump relied on the volume of oil displaced by the crank and rod rotating to fill the pump inlet. Unfortunately this approach could not guarantee a consistent supply, particularly under cornering where the g-forces were such that they pushed the oil away from the inlet. To combat this, Honda incorporated an oil trap into the inlet of the pump. This took the form of a small cavity just next to the pump inlet that acted as a reservoir that was filled by the oil thrown from the rotating components, ensuring a constant supply regardless of the g-forces.

Interestingly, owing to high operating pressures of 20-40 kPa, Honda also opted for a compression-type pump design rather than a positive-displacement pump for the scavenge system, the reason being the increase in pumping efficiency such a design offered. However, it did mean having to include a pressure relief system in each scavenge. This was because the pumps were specified to transport the light oil-gas mix that the crankcase contained; if they pumped no aerated oil, such as when the engine was initially started, the pressure increase caused by the higher-viscosity fluid could have caused damage.

As with the feed pump system examined last month, the level of development that went into optimising this relatively mundane engine subsystem is impressive, and highlights the level of refinement necessary in engines such as those in Formula One to ensure optimum performance and reliable operation.

Written by Lawrence Butcher

Taking as an example the dry-sump system found on Honda’s V8-engined cars in use until the end of 2008, there are in fact two pumps on the engine, one feeding the oil and the other scavenging it from the sump. These pumps had torocoid rotors, with four inner and five outer teeth, which provided excellent volumetric efficiency.

A key design requirement for Honda was that each section of the engine should get the correct volume of oil, at the correct pressure, all the time. Ensuring that this happened was no easy matter, for example the valvetrain needed a high volume of oil at low engine speeds, but the demand did not rise with engine speed. Conversely, the oil supply to the main and big-end bearings, supplied though a centre feed in the crank, needed to rise in pressure as engine speed rose.

One major problem the team faced was a loss of feed pump performance as engine speeds increased (to more than 18,000 rpm eventually), due to suction cavitation in the pump. One potential solution would have been to reduce the pump speed and increase its size, but given the ever-present pressure to reduce mass and packaging size, this was not considered to be a viable option. Instead, Honda identified the root causes of the cavitation.

There were found to be two clear contributing factors. First, an insufficient volume of oil was being drawn into the rotor suction chamber during rotational transfer at high speeds. Second, the oil that was being drawn into the rotor chambers was leaking out because of centrifugal force. To cure the first issue, Honda replaced the thick, single rotors with thinner twin rotors, while the second was addressed by redesigning the suction port, so that as the suction chamber filled with oil, the port closed, preventing leakage.

The modifications were a resounding success. In 2005, Honda’s feed pump was rotating at 12,800 rpm but suffered a considerable decline in pressure at these speeds. With the modifications, which allowed for the basic pump design to be retained, flow rate increased by 30% for no reduction in speed.

Another interesting area of development for Honda was the pump rotor material itself. Initially, these rotors were constructed from sintered aluminium; however, in a bid to reduce rotating weight, and hence overall engine losses, Honda experimented with different materials.

The last generation of engines, used during the 2008 season, featured an inner rotor made from sintered magnesium, with an outer rotor of plastic. Due to the soft nature of these materials – particularly the plastic – the team had to develop a new profile for the rotors, which retained pumping efficiency but reduced surface pressure between the teeth. If the tooth geometry had been left the same as that used for the harder materials, the level of friction between the teeth would have increased, negating any gains from the rotors’ reduced mass.   

This is only a small insight into the level of optimisation that goes into just one component in a Formula One powertrain, but it shows the lengths engineers will go to in order to find performance gains in such series and the interesting paths down which such investigations may lead them. 

Written by Lawrence Butcher

In the early days of racing it was perfectly acceptable to simply vent crankcase pressure directly to the atmosphere. However, given that the gas being vented contains a mixture of unburnt fuel, combustion debris and oil mist from the crankcase, this approach became unacceptable in the post-war era as environmental concerns started to grow.

The simplest solution to containing these by-products is a closed circuit breather system, where the case pressure is vented directly into the engine inlet. Over time, these have evolved into quite complex systems to manage case venting at varying engine loads and speeds. For example, the systems used on modern roadcars will use valves that prevent oil mist being sucked directly from the engine at low speeds when case pressure is low and inlet vacuum high.

While some race engines feature such systems their complexity is not appealing, and neither is the ingestion of oil mist into the inlet charge. Thus, in racing applications, by far the most popular method of controlling oil expelled through any vents is to use a catch tank with either a separate or integral oil separator.

The most basic type of oil separator consists of a simple volume through which the blow-by gases flow. As the gas enters the volume, its velocity slows, allowing oil to drop out of suspension and pool in the bottom of the volume. The oil can then be either fed back into the engine or into a catch tank. Taking the void approach one step further is the use of what is known as a ‘labyrinth’ system of baffles. These force the blow-by gas to slow down by directing it around tight corners, and again the oil drops out of suspension.

More complex are centrifugal-type separators. These cause the gas to spin through a chamber, with the oil droplets separating out and running down the chamber walls. ‘Driven’ centrifuges, where the cylinder is actively rotated, are one option but they are rarely used because of their complexity. Far more common are non-driven centrifuges, where the gas enters a circular chamber at a tangent and is encouraged to flow in a circular motion, creating enough centrifugal force to throw the suspended oil against the chamber walls.

Most catch tanks with integral separators will use either this type of design or a labyrinth-type baffle. Note though that most dry-sump oil tanks also rely on the centrifugal approach to help de-aerate scavenged engine oil. 

The most important design consideration for an oil separator or oil catch is its cross-sectional area. This needs to be large enough to allow for the speed of the blow-by gases entering the device to drop below 1 m/s as they transit the separator/tank to the vent. It is also worth having an inlet angle that is tangential to the wall of the tank, in order to promote a circular motion regardless of whether centrifugal force is destined to be the sole method of oil separation.  

Written by Lawrence Butcher

While not yet common in racing machinery, significant savings in packaging and weight can be made by sharing lubricating/cooling oil between the electric motors and any transmission systems used. (There is a possibility that such an approach is taken on Audi’s current R18 LM P1, where the front-mounted MGU features a dedicated gearbox/differential unit.) For example, General Electric unveiled a motor in 2013 that uses regular automotive transmission fluid that can be shared with a vehicle’s transmission.

Such sharing considerably reduces the complexity of cooling and lubrication systems on hybrid vehicles, and in a racing context that means less pipework and thus less weight and a smaller packaging envelope. If the same fluids can also be used for cooling energy stores, then evidently further savings can be made. Simplifying these cooling circuits could also have potential benefits in relation to aerodynamic performance. Beyond simply reducing the number of pipes under a car’s body, allowing for tighter body packaging, individual coolers for hybrid and other systems can also be eliminated. Although the same levels of heat rejection will still need to be accommodated, the use of a single cooler allows for more efficient packaging and less ductwork, potential reducing overall vehicle drag.

While not directly related to lubrication, another factor that needs to be considered is the impact of electricity on the life of bearings. Most electric motors used in motorsport are ac, three-phase traction motor/generators. Since batteries provide dc, inverters (also known as variable frequency drives, or VFDs) are needed to convert the dc to ac. These have an unfortunate side effect – they induce unwanted voltages on motor shafts. Without effective, long-term electrical grounding, this shaft voltage can erode and eventually destroy motor bearings. The bearings can be insulated, but this can simply move the problem elsewhere.

The effect is exactly the same as that found in an electrical discharge machine, with the current moving from the motor’s shaft to the bearing, removing material. The need to find ways of mitigating this effect is not a new one, and electric motor manufacturers for sectors such as mass transport and heavy industry have developed various solutions. For example, one system being investigated by some automotive manufacturers is the use of a conductive fibre ring around the motor shaft, which provides an earth path that bypasses the bearings.

The point illustrated here only scratches the surface of what is a new and emerging area of vehicle lubrication. As mixed and pure EV powertrains appear in increasing numbers, efficiency gains relating to lubrication improvements will no doubt be an area of great interest to those involved in racing them.  

Written by Lawrence Butcher

With this in mind, it should come as no surprise that some engine developers have harnessed the power of CFD to create fully dynamic simulations of engine flows, encompassing both the properties of the fluids running in and around the engine, and accounting for factors such as heat rejection. Such advanced levels of simulation are exceptionally complex though, requiring large amounts of computing power to obtain high-resolution solutions. Luckily, with the emergence of open source CFD packages such as OpenFoam, and cloud-based computing, the ability to analyse specific areas of interest have been brought within the reach of a far wider range of engineers.

One particular area of the lubrication system into which CFD can provide useful insight is the design of oil pumps. The pump is arguably the most important component in the lubrication system, so in a highly optimised race engine its operation needs to be as efficient as possible.

The most common pump used in race engines, which invariably feature dry-sump lubrication, is the gerotor type. This consists of an external and an internal rotor, with the outer rotor driven by the rotation of the inner one, which is connected to a driveshaft. The pump body will have one or more inlets and outlets, and as the rotor rotates, the teeth of the inner and outer rotors disengage on the inlet side, creating a partial vacuum that draws oil in. The oil fills the volume between the teeth, and as the teeth begin to re-engage, it is forced out of the outlets.

Theoretically, the flow rate of the oil is a function of the gears’ rotational speed, but in practice leakage through the gaps between the gears reduces efficiency. It is therefore beneficial for the pump designer to be able to model exactly what is going on inside a pump. While it is possible to physically test pumps on a rig, comparing different rotor geometries and body designs requires the production of multiple components. However, by using CFD to perform such analysis, changes need only be made in the CAD models, with the only limitation on the number of iterations that can be tested being the computing power available.

Provided the CFD model is properly validated, it can also be used to investigate other important phenomena, for example the formation of bubbles and cavitation in the pump body, or the impact that variations in features such as the shape and angle of the inlet and outlet ports has on oil pressure and flow. Being able to assess the benefits of changes in these areas can have a small but significant impact on overall engine efficiency, leading to potential gains in both lubrication system effectiveness and reductions in parasitic power losses.

It would be fair to say that, given the current state of development of CFD and its now widespread availability, such detailed investigations are no longer the preserve of the big engine manufacturers and are within reach of smaller engine developers.

Written by Lawrence Butcher

Teams struggled with broken axles, flanges and other rear suspension parts, and with prematurely worn-out bearings and seals leaking under the increased stresses. As Eric Warren, director of competition at Richard Childress Racing, explains, “It was a huge learning experience. The reason that the rule had been 2º for years was to prevent failures. NASCAR figured that as people were running housings with toe-in (a practice outlawed for 2013), you should be able to run with a degree more camber.

“Everybody was in a mad dash to get the rear axle grease and the materials for the half-shafts and drive plates to survive under the new loadings. The major issue was heat generation, and that was breaking down the greases we were using at the time. So we had to move to greases that were rated at much higher temperatures.”

So how are different types of grease formulated, and what sorts of additives are used to tailor their performance to particular applications?

Much as with oils, greases are formulated around base oil stocks and additive packages (see RET-Monitor November 2012 and December 2013). However, where grease differs is in the addition of thickeners, which give it its consistency, with the concentration and type of thickener dictating how ‘thick’ the final grease is. It should be noted that a grease can have a very thick consistency but in fact be based around a low-viscosity base oil.

Numerous types of grease thickeners are currently in use, each with its own advantages and disadvantages, depending on the particular application. The most common thickener types are simple lithium soaps, lithium complex and polyurea.

Simple lithium soaps are often used in low-cost general-purpose greases, and perform relatively well in most performance categories at moderate temperatures. Lithium complex greases provide improved performance, particularly at higher operating temperatures. A common upper operating temperature limit for a simple lithium grease might be 250 F (120 C), while that for a lithium complex grease might be 350 F (176 C). Polyurea is growing in popularity, and like lithium it has good high-temperature performance as well as high oxidation stability and bleed resistance.

The type of thickener in a grease is key to its performance, so in an application such as NASCAR Cup, where high temperature stability is key, only a lithium complex or polyurea grease should be considered. Fortunately for the motorsport customer, there are a lot of high-performance greases on the market, many of them originating from the aerospace and power generation industries, that provide the performance needed in even the most demanding environments.

Written by Lawrence Butcher

Traditionally, the excess pressure is dealt with by an oil pressure release valve that bleeds oil back to the sump if the optimum pressure is exceeded. Ultimately, this means that when the pump is flowing more oil volume than required, it is contributing excessively to parasitic power losses.

One development that could provide improvements and reduce these losses is the advent of variable-displacement oil pumps. Rather than rely on a release valve to control oil pressure, the volume of the oil pump itself is actively controlled in relation to oil pressure. In the past two years, a number of production cars have started to feature this type of pump as manufacturers strive to improve engine efficiency in order to meet increasingly stringent economy standards.

Both VW/Porsche (VW Audi Group) and GM use variable-displacement designs in their roadcars, although those from VAG differ in operation from those from GM.

The VAG group pump is of the gerotor type, where inner and outer rotors are used to pump the oil. The inner rotor sits on a driveshaft and drives the outer rotor. Because the inner and outer rotors have different rotating axes, more space is created on the suction side due to the rotating motion. The oil is drawn in and transported to the pressure side. On the pressure side, the space between the teeth becomes smaller again, and oil is forced into the oil circuit.

In a variable-displacement pump, however, the outer rotor is enclosed by a control ring that can move the position of the outer ring relative to the inner ring. The control ring is kept in place by a regulator spring, and only begins to move once the oil pressure pushing against it exceeds the spring pressure. When this happens, the control ring turns, decreasing the space between the inner and outer rotors, resulting in less oil being transported from the suction side to the pressure side and forced into the oil circuit. With less oil, the oil pressure is lower, and the nature of the system means that the volume of oil pumped is constantly varied in order to maintain the correct oil pressure.

The unit from GM, currently found in its 1400 cc turbocharged I4 Ecotec engine, works on a similar basis but relies on a vane-type pump rather than a gerotor unit; its operation is shown in the video below. Vane-type pumps are typically used in situations where low-viscosity fluids need to be pumped, and because of this – plus their extra complexity and therefore higher cost compared with gear pumps – they have not tended to be used in automotive oil systems. However, they are beginning to be seen in an increasing number of automotive applications thanks to their more efficient operation compared to geared pumps. 

A vane-type pump generates flow using a set of vanes that are free to move radially within a slotted rotor that rotates in an elliptical chamber. A typical configuration uses an elliptical cam ring, with the rotor spinning in a cylindrical housing and a pair of side plates to form the pumping chambers. The changing volume of the cavity between adjacent vanes creates the pumping action as the rotor rotates. In the case of GM’s pump, the variable output is achieved in much the same way as with the VAG unit, with an oil pressure-controlled ring varying the pump’s displacement by changing the volume of the pump’s cavity.

It is early days for variable-displacement oil pumps, but with manufacturers claiming reductions of about 60% in their power consumption, it is only a matter of time before they see wider adoption. 

Video: GM variable-displacement vane pump

Written by Lawrence Butcher

These additives allow the lubricant to work at a wider temperature range and at higher loads as well as prolonging the life of the lubricant and the mechanical parts involved. Oil producers will keep their exact additive packages a closely guarded secret, but in general terms these additives can be put in to the following categories. 

Friction modifiers (FM)

These additives are introduced to lubricants to reduce friction. Typically they consist of chemical compounds that have a high affinity for metal surfaces and possess long alkyl chains that are adsorbed on the contact surfaces in either single or multiple layers, resulting in a thin film forming. The nature of this attachment means FM molecules only function in the boundary or mixed lubrication regimes, for example at the camshaft-lifter interface. 

Anti-wear (AW) additives

Compared to friction modifiers, anti-wear additives are adsorbed more strongly onto metal surfaces, and are able to reduce wear of the surfaces involved thanks to a thicker film that serves to separate the surfaces even under high loads. The most common AW additive is a compound of zinc, sulphur and phosphorus, called ZDDP (zincdialkyldithiophosphate), the benefits of which were covered extensively in a previous RET-Monitor article.

As noted in that article, ZDDP has fallen victim to more stringent environmental regulation, so oil manufacturers have undertaken extensive research into similarly effective alternatives, including non-metal based compounds featuring a high phosphorous content. Alternatively, particularly when dealing with engines that use hydraulic tappets, the use of low-friction, wear-resistant coatings such as DLC has removed the need for additives such as ZDDP. However, in engines such as the flat-tappet units used in NASCAR’s Cup series, ZDDP is still the only viable solution to ensuring component longevity.  

Extreme pressure (EP) additives

EP additives are used primarily in applications that see exceptionally high contact pressures, and are more often found in transmission rather than engine oils. Both AW and EP additives are activated by pressure and temperature as well as shear forces, although EP additives tend to activate at higher temperatures than AWs.

At elevated temperatures, they ‘chemisorb’ – chemically bind – at surfaces, the compound being adsorbed onto the metal surface by a chemical reaction. However, at lower temperatures they are more likely to physically adsorb. Therefore, under moderate loads, AW additives will provide the protection needed, with the EP additives coming into play in more extreme conditions. 

Viscosity modifiers

The purpose of viscosity modifiers is to allow the lubricant to work effectively over a wide temperature range. They are particularly important in race engines that tend to run at higher temperatures than general automotive units. For example, one oil supplier to NASCAR teams had to develop a new viscosity-modifying additive package to ensure oil film stability at the elevated temperatures encountered when cars were drafting, as a result of the practice’s impact on airflow through the car’s oil coolers. 

Viscosity modifiers tend to consist of polymers of high molecular weight that unfold at high temperatures, leading to an increase in viscosity. At low temperatures they curl up, leading to a less pronounced increase in viscosity. By carefully selecting the viscosity modifiers, an oil can be produced that has a low viscosity at normal operating temperatures, thus reducing frictional losses, yet still retains its lubrication characteristics, preventing excessive wear, if temperatures rise.

Written by Lawrence Butcher

Ultimately, debris in the oil – be it from external contamination or the breakdown of components such as bearings – is seriously detrimental to an engine’s longevity. The most common filter type, found on everything from a sedate shopping sedan to an 850 bhp NASCAR Cup car, is the ‘cartridge’ filter. First developed in the 1950s, these filters may look simple externally but the differences between the ones found at a local part factor and those developed for racing use are considerable. While the basics of their operation are similar, the needs of a race engine compared to a road engine are very different.

The most important difference relates to the level of filtration, and the impact this has on oil pressure and flow. Logic dictates that one would want the best level of filtration possible, in order to trap the greatest quantity of debris. As an example of the level of filtration available, a high-quality automotive filter will trap about 80% of particles between 8-10 µm in size. However, research by one manufacturer of racing oil systems has found that, in the vast majority of cases, this degree of filtering is not necessary and could in fact be detrimental to the overall performance of the oil system.

The research found that some very fine filters produced an unacceptable level of pressure drop across the filter because of the restriction of the filter element, which in some cases was sufficient to cause the oil bypass valve to open and feed unfiltered oil to the engine. Such occurrences would be almost undetectable on most engines, as the impact on overall system pressure (gauge pressure) would be minimal, and the first indication that something was amiss would be an increase in wear rates.

The company’s study indicated that particles smaller than 20 µm are not a particular concern for engines with short service intervals, which encompasses most race engines. With this in mind it was able to create a filter that provided complete filtration of particles down to 27 µm, resulting in a maximum pressure drop across the filter of only 2.5 psi.

Another factor to be considered in filters destined for racing use is the suitability of the filtering medium itself. Since the 1940s, cellulose-based materials have been the norm, with the oldest filters using simple cotton fibres. Most fibres used in mainstream oil filters are derived from wood pulp, with the fibres being made rigid by adding resin. For most applications, cellulose is still more than sufficient, providing filtration down to around 20 µm; for racing though, a synthetic filter material is the only real choice.

There are two key reasons why this is the case – flow and temperature/pressure resistance. Cellulose fibres are relatively thick, while synthetic ones are much thinner, so for a set mesh size the synthetic filter will provide a greater area for oil to flow through. Regarding the resistance of the fibres to temperature and pressure, synthetics maintain their structural integrity under extreme conditions much better than cellulose. Thus in race engines subject to elevated temperatures and running higher oil pressures, they are an ideal solution.

Written by Lawrence Butcher

There are many similarities between race engine and aero engine development, with designers of both striving to find the most efficient and lightest possible engineering solutions. In the case of oil coolers, the increasing heat rejection characteristics of the Merlin over its predecessors meant a new cooler had to be designed that provided sufficient heat transfer in a compact and lightweight package. It was therefore logical that post-war racecar engineers would, for the very same reasons, adopt the design.

The design has been optimised over the years, but the same basic features have remained. The intention is to provide maximum contact area for the oil and air, while minimising any reduction in oil pressure. This is achieved by using what is known as a bar and plate construction. The fluid tubes are created by sandwiching two square bars between two flat plates, with the assembly then brazed together. These tubes are fixed between two end tanks through which the oil enters and exits, with the voids between the tubes taken up with closely packed fins which provide most of the cooling area.

This type of construction provides far better flow and heat rejection characteristics than simple round tube and fin coolers, which used to be the standard in automotive applications. These simply consisted of a round tube with fins attached, with the oil flowing from one end of the tube to another. Although cheap to produce, the large number of tube bends creates a significant pressure drop, and a much larger cooler is needed to obtain the same heat rejection as a bar and plate type cooler due to the smaller tube and fin surface area. There are a number of variations on the basic bar and plate cooler, including types that use seamless C-section tubes to increase burst pressure by eliminating seams.

Beyond the benefits of the basic construction, most high-quality bar and plate oil coolers have their efficiency further increased thanks to the inclusion of ‘turbulators’ within the fluid pipes. When fluid flows through a plain tube the fluid nearest the wall is subject to frictional drag, which has the effect of slowing down the fluid at the wall. This laminar boundary layer can significantly reduce the tube-side heat transfer coefficient and consequently the performance of the heat exchanger.

In an effort to promote turbulent flow, many cooler manufacturers fit shaped elements within the oil pipes. The shape and size of these elements can have a considerable impact on the pressure drop across a cooler, so their design is a trade-off between increased cooling and ensuring that an acceptable pressure drop is maintained.

For the same reasons, some oil cooler designs also use turbulators on the air side, with turbulent flow increasing the volume of air coming into contact with the cooling fins. In many racing applications though, where the smallest possible coolers are often specified to help keep frontal area (and thus drag) low, turbulent flow within the cooler is to be avoided at all costs.

Written by Lawrence Butcher

Before studying how aeration occurs, it is pertinent to take a quick look at the three forms of air in oil – dissolved air, air bubbles and foam. When present in oil in these forms the air is referred to as ‘bound’ air because it is bound, or connected, to the oil. Dissolved air is invisible and is harmless to oil function as a lubricant, but it can be released as bubbles and/or foam. Bubbles are small air pockets entrained and dispersed throughout the oil, while foam consists of pockets of air on the surface of the oil separated by thin liquid films. Air that is separated from the oil is referred to as ‘unbound’, or free air, and can become bound air through various mechanisms.

The level of aeration in the oil is determined by a range of factors, including engine speed, oil type and oil condition among others. It can be a particular problem in high-revving race engines – not only does the high rpm of components such as the crank churn the oil to a greater extent, the oil also flows through the lubrication system faster. This means it has less time to rest in the sump or the oil tank, where air bubbles would naturally separate out.

Another component that can significantly increase aeration in the oil is the oil pump itself, particularly in dry-sump lubrication systems. The aeration occurs on the scavenge stage of the pump, and the oil being sucked into the pump is more of an air-oil foam; as the oil is pulled between the pump’s gears it is further aerated. That is why oil tank design is of utmost importance for maintaining oil condition. As described in a number of previous RET-Monitor articles under the Oil System keyword, tanks are constructed so as to maximise the amount of air oil separation as the oil is fed down through the tank.

The problem of removing air from scavenged oil has also led to a number of devices designed to condition the oil before it reaches the tank. The most recent, and interesting, of these is an inline air-oil separator that uses centrifugal force to separate oil and air. Through the use of carefully designed channels, the air is removed through one outlet, with the de-aerated oil flowing through another.

While some oil is carried out with the air (which is fed back into the oil tank through a secondary inlet), most is returned through the main tank inlet in a far less aerated state than if it simply flowed from the pump. In terms of performance, the manufacturer claims a 30-70% reduction in aeration over running a direct scavenge feed to the tank, with the level of de-aeration dependent on rpm and other engine operating conditions.

On a final note, it is worth considering that oil pressure also has a direct relationship to its susceptibility to aeration. The greater the oil pressure, the more air it is able to absorb and, if the pressure drops, a portion of this dissolved air will form into bubbles. Therefore, if a high-pressure oil system suffers a pressure drop, more air bubbles will be introduced into the lubricant, potentially contributing to problems such as pump cavitation. Conversely, if a system is running at a lower pressure then there will be less air to release and, thanks to Boyle’s law, the size of these bubbles will also be larger, meaning they will separate from the oil more quickly.

Obviously the susceptibility of any particular lubrication system to aeration depends on a multitude of dynamic factors, but hopefully this provides an overview of the core issues.

Written by Lawrence Butcher

This presents some unique challenges from a lubrication perspective, not least in the area of oil selection. Modern emissions legislation has seen the make-up of engine oils change considerably over recent years, with an ensuing unforeseen impact on the reliability of engines running solid lifter. This month I want to investigate these issues and the solutions devised to counter them.

Thanks to regulation changes by various legislating bodies, particularly the US Environmental Protection Agency, the quantities of ZDDP (zinc dialkyl dithiophosphate) used in most engine oils has been restricted. The intention behind such legislation is to protect emissions-reducing components such as catalytic converters, which can be damaged by the inclusion of such additives. However, the reduction in ZDDP has had an unforeseen impact on racers running so-called ‘flat tappet’ race engines, ranging from NASCAR Cup racers to Californian drag racers running air-cooled VW engines.

ZDDPs are a family of coordination compounds, originally developed by Castrol, that feature zinc bound to the anion of dithiophosphoric acid. These uncharged compounds are not salts; they are soluble in non-polar solvents, and the longer-chain derivatives dissolve easily in the mineral and synthetic oils used as lubricants. It is important to remember that in boundary lubrication, surface asperities make contact with each other even though the lubricant supports much of the load. The level of friction is determined by the shearing forces necessary to cleave these adhering asperities, and wear and friction can be reduced through the use of additives that reduce this contact.

In most modern engines, whether for road or race use, the loss of ZDDP is not a major problem. These engines feature overhead camshaft designs that do not experience the same boundary lubrication issues at the cam lobe-lifter interface as old-fashioned pushrod V8s, which often feature solid flat tappets [Fig. 1]. Compared to engines that use roller or hydraulic lifters, the loadings at the point of contact between a flat tappet and the camshaft lobe are exceptionally high, so there is a greater need for additives in the oil to help protect them. Over the years, ZDDP has proved to be the perfect solution here, and thus its removal from most motor oils caught some racers by surprise. Suddenly they were starting to experience previously unseen problems, particularly excessive wear of cam lobes.

To counter these problems, several oil manufacturers started producing specialist oils for motorsport use, with substantial ZDDP content. However, it is worth noting that not all zinc (ZDDP) additives are the same. ZDDP does not begin to be effective until it is subject to heat and loading and, depending on the exact formulation used, different additive packages react at varying levels of heat and load.

ZDDP also has different ‘burn’ rates. Some zinc additives have slower burn rates that require more heat and load to activate than others. With this in mind, several manufacturers of engine break-in oils make a point of using ZDDP formulations that activate at relatively low heat and load levels, providing the maximum possible protection for components when they are at their most vulnerable.

Fig. 1 - A solid valve lifter, often referred to as a flat tappet, in this case from a VW air-cooled engine used for drag racing

Written by Lawrence Butcher

The use of turbo-supercharging though brings some considerations regarding lubrication systems, above and beyond those required with naturally aspirated engines.

Turbo-superchargers have various types of internal bearing, with either rolling element or journal bearings being commonplace. Regardless of their type, however, they all are subject to very high load and temperature variations. A turbo compressor can spin at anything up to 100,000 rpm under race conditions, so consistent lubrication is vital. Invariably, a turbo-supercharger will receive its oil supply from the engine, although there is at least one unit on the market with a self-contained lubrication system.

In this unit the main bearings are located in the cold-side compressor housing rather than between the compressor and turbine housings. A cavity is built into the front portion of the compressor housing in which specially formulated high-speed bearing oil resides. This oil is transferred to the ceramic ball bearings via a pair of wicks through which a mist of oil is drawn; as the bearings are not flooded in lubricant, drag is reduced, and due to the bearings’ location, lubricant temperatures are also kept low.

However, in most conventional turbos the bearings are located between the hot and cold sides of the turbo, and are subject to much greater heat transfer from the exhaust gases, so a constant flow of oil is needed. This flow lubricates the bearing and contributes to cooling. Many modern turbos often feature additional water cooling, but some still rely solely on the oil. Regarding the plumbing of turbos, it is widely known that the bore of the exit pipework for the turbo oil supply must be far larger than that of the inlet, to accommodate any aerated oil. Similarly, unless a scavenge pump is fitted, the oil exit must be above the level of the main engine sump to allow the lubricant to drain.

One key factor that is often overlooked is filtration of the turbo’s oil supply. The three main causes of turbo failure are excessive heat, ingestion of foreign objects into the compressor wheel and oil contamination. Most engine oil filters will trap particles larger than 100 microns, which is not ideal for the longevity of a turbo’s bearings, so most turbo manufacturers recommend that the lubricating oil is filtered down to 10 microns, particularly if a turbo uses rolling element bearings.

Given the high shaft speeds, even the slightest contamination can lead to degradation of the bearing surfaces – even carbon particles deposited in the oil from the combustion process can cause problems. The solution is to run a secondary filter element in the turbo oil feed lines, providing the necessary supply of clean lubricant.

However, it is not a straightforward case of ensuring that the filter element can trap particles of the required size; it must also still allow sufficient oil flow to prevent starvation of the bearing. With packaging space often at a premium, filter manufacturers have had to develop specific high-flow filters to ensure that filtration as well as flow requirements are met.

On a final point, it should be noted that because of the generally small size of turbo oil supply lines and filters, regular inspection is imperative. Any build-up of deposits in these lines can restrict flow, so the measures put in place to protect the turbo bearings can actually become the reason for their failure.

Written by Lawrence Butcher

For many years, the route to reducing frictional losses using lubricating oils has been to reduce oil viscosity, thus lowering the sliding friction between components. Unfortunately this invariably has the knock-on effect of reducing the film strength of the oil, a less than ideal solution in highly loaded contact areas such as those between gears. Here the weak film strength can lead to a boundary lubrication condition, to the detriment of component reliability as well as friction.

However, a new avenue of developments in engine and transmission oils looks set to reduce power losses, without reducing lubricity. These developments stem from advances in nanotechnology, which refers to the engineering of structures at an atomic level to produce materials that would not be available through traditional chemical engineering means. By re-engineering conventional materials at an atomic scale, their fundamental properties can be altered considerably, and in the realm of lubricants it has been found that compounds engineered at the nano level can prove highly beneficial in reducing friction.

Many oils use additive packages containing elements that can also act as dry lubricants, and new nano-additives can be used in the same way but with improved characteristics. In contrast to the flat plates of conventional solid lubricants that lubricate by sliding, inorganic fullerene-like (a fullerene being a molecule composed entirely of carbon), nanotechnology-based solid lubricants have structures of progressively smaller concentric spheres, sometimes as many as 20 or more, nested within each other like a Russian doll. These act as miniature ball bearings that roll across surfaces, providing lubrication.

This multi-wall structure offers many advantages. As the outer shell wears away during use, it exposes an identical shell underneath. This automatically renewing, self-healing feature maintains lubricity.

In comparison tests with conventional boundary lubricants such as molybdenum disulphide, nanotechnology lubricants are reported to have reduced friction by up to 25% while increasing load capacity by up to 80%.

A typical test to compare the sliding friction performance of different lubricants involves a high-frequency reciprocating rig in which a steel ball is loaded against a reciprocating plate. In one comparison test, the friction coefficient of a standard automotive oil was 0.17, while the average film strength (measured by electrical resistivity) was 84%. A representative race oil showed a friction coefficient of 0.11 and film strength of around 75%. Another showed friction below 0.1 at higher operating temperatures but highly variable film strength, averaging 34%. However, a nano-additive oil of the same viscosity recorded a friction of 0.06 at 75 C while retaining a film strength of 98% – halving the friction without losing any film strength.

Obviously gains of this magnitude, coupled with potential increases in reliability, could be highly beneficial to teams looking to increase performance without expending vast resources on component development.

Fig. 1 - Nanotechnology-based additives take the form of microscopic ball bearings

Written by Lawrence Butcher

The biggest differences in lubrication needs between an engine fitted in a motorcycle and one fitted in a car are the levels and direction of forces the oil is subjected to. Notably motorcycles lean, in the case of racing bikes very heavily, during cornering. This means the oil is subject to the same forces as oil in the sump of a car, but the orientation of the bike means it remains in the bottom of the sump, rather than being pushed up the sides as it is in a car. Because of this, the oil pick-up system will be optimised to operate under these conditions, so when placed in a racecar (which does not roll) oil starvation and surge can become a problem. The other issue is from a packaging perspective, as standard motorcycle sumps tend to be very deep, which is less than ideal for attaining a low centre-of-gravity height. A lower-profile sump is therefore often desirable.

There are a number of ways to ensure that a bike-engined-car does not suffer from oil starvation. The simplest, though least effective, is simply to overfill the sump with oil. This though can lead to high oil usage and increased windage. A second solution is to use a sump baffle system, as in car engines. These can be very effective, but this effectiveness depends on the engine type and the orientation of its installation.

The Yamaha R1 (1999-2003) engine is a good example of where a baffle system can be a suitable solution. Taking the experiences of various sidecar racing teams, it appears the R1 can get away with a simple baffle plate, a relief shroud and a small overfill of oil. This provides enough performance in a sidecar outfit, capable of cornering at over 2 g, so it should be sufficient for cars producing medium amounts of aerodynamic grip. It is also negates the weight, complexity and expense of a dry-sump system.

A further tool that can assist in ensuring regular lubricant flow is a moveable oil pick-up pipe. Several manufacturers of sump systems specifically for bike-to-car transplants offer sump plates with a flat bottom and swinging pick-up to cater for this.

For some applications, even the most highly optimised wet sump will be insufficient to deal with the forces generated. Many bike-engined racers, especially those used for hillclimb or sprint events, have extremely powerful aero packages and are capable of producing prodigious amounts of downforce, resulting in cornering g-forces that would overwhelm the best wet sump. It is therefore necessary to fit a dry sump lubrication system to prevent engine damage.

There are plenty of engine-specific systems on the market to cater for this need. Many use the original engine’s oil pump, with extra pick-ups in the sump, while others rely on an external pump. In some applications, particularly those using Suzuki’s 1300 cc Hayabusa engine, the scavenge pump for the system replaces the engine-driven water pump (while retaining the original pressure pump). Water pumping duties are then completed by a separate electric pump.

From a packaging perspective, this type of system is ideal, as it adds very little to the size of the engine envelope. It should be noted though that most bike engines do not have useful accessory drive pulleys to drive external pumps in the way car engines do, so the drive to any external pump must use existing gear drives from the engine.

Some of the most in-depth research regarding dry-sumped motorcycle engines has been undertaken by Formula Student/FSAE teams, which invariably run such engines. For those looking for more information, a search of Google Scholar will flag up many student papers on the subject.

Fig. 1 - Lubrication systems in bike-engined cars such as the Radical SR3 need careful consideration to ensure sufficient oiling (Courtesy of Radical Sportscars)

Written by Lawrence Butcher

All oils are made from a 'base stock', but while traditional mineral oils are derived from crude oil, the same is not true of all synthetic oils; the 'synthetic' refers to the fact that the oil has been manipulated at the molecular level to change its properties. This can either be manipulation of highly refined mineral stocks (originating from crude oil) or from man-made base stock derived from sources such as natural gas or esters.

The list below shows the different types of base oils. Many synthetic oils on the market are produced from hydro-cracked Group III stocks; however, for motorsport-specific oils only groups 4 and 5 will normally be used.

There are five families of base oil:

• Mineral oil Group I - conventional base oil derived directly by refining crude oil
• Mineral oil Group II - as above but more highly refined
• Hydrocracked Group III- extra highly refined base stocks
• PAO Group IV - pure synthetic hydrocarbon (SHC)
• Esters Group V - fully synthesised from a reaction between acids and alcohols.

These base stocks will have a very uniform molecular structure, as shown in Fig. 1, which allows them to perform far more consistently as a lubricant. This is important, because the base stock, even with the inclusion of additives, is responsible for the majority of lubrication properties in an oil.

The key advantages of synthetic base oils from a race engine designer's perspective is their improved temperature resistance, heat rejection and film strength characteristics.

Synthetics are simply more tolerant than petroleum oils to extreme heat. When heat builds up in an engine, petroleum oils quickly begin to burn off. They are more volatile; the lighter molecules within petroleum oils turn to gas, and what's left are the large molecules that are harder to pump. Synthetics have far more resistance as they are more thermally stable to begin with, and can withstand higher temperatures for longer periods without losing viscosity, a characteristic that can be further enhanced by including viscosity-modifying additives.

The uniformly smooth molecular structure of synthetic oils also gives them a much lower coefficient of friction than mineral oils - and less friction means less heat. Also, since each molecule in a synthetic oil is of uniform size, each is equally likely to touch a component surface at any given time, thus moving a certain amount of heat into the oil stream and away from the component. This makes synthetic oils far better heat transfer agents than conventional petroleum oils.

Mineral oils have very low film strength in comparison to synthetics. The film strength of a lubricant refers to its ability to maintain a film of lubricant between two objects when extreme pressure and heat are applied. This is especially important where lubrication is marginal, such as around main bearing shells or at the camshaft- lifter interface. Synthetic oils will typically have a film strength five to ten times higher than that of mineral oils of comparable viscosity. Even though heavier weight oils usually have higher film strength than lighter weight oils, an SAE 30 or 40 synthetic will typically have a higher film strength than an SAE 50 or 60 petroleum oil.

Different oil suppliers will have their own synthetic base stocks, and these will be used as the foundation of motorsport-specific synthetic blends. It is from this starting point that chemical engineers will begin to modify the oil to provide specific characteristics. The level of this 'customisation' will vary depending on the final user: a general motorsport oil for modern multi-valve engines will have an additive pack that can be used across a range of engines, whereas oil destined for a Formula One team will have an additive pack tailored exactly to their engines' specific needs. Next time we look at oils, we will investigate these additives further.

Fig. 1 - A representation of the molecular make-up of mineral-based and synthetic oils

Written by Lawrence Butcher

Oil quality

There are a number of sensors on the market that allow engineers to assess oil quality in real time, potentially saving an engine from expiring prematurely. These sensors fall into two brackets - those that monitor the quality of the oil itself, and those that measure the quantity and type of metallic particles in the oil.

Oil quality sensors work by assessing the dielectric constant of the oil - a dielectric is a type of electrical insulator, most commonly used in capacitors - and comparing it to a reference figure from a known sample. This information can then be transmitted over the car's telemetry link, allowing trackside engineers to tell if the oil structure is degrading.

Sensors to measure the quantity and type of metallic particles in the oil work by collecting metallic particles and analysing their composition via a remote diagnostics unit that can discern between the tiny metal particles that are always present in the oil and larger particles that could indicate the imminent demise of the powertrain.

One such sensor on the market is designed to replace a conventional magnetic oil plug or sump plug in either engine or transmission applications. It is a dual-channel device that can detect not only the quantity but the type of metallic debris build-up in the oil. It attracts metallic debris to the face of the sensor (much like a conventional metallic plug) and measures the particle build-up via remote electronics. One output channel provides data on very fine particle build-up, the other on the build-up of larger metallic objects (which could result in the aforementioned demise of the engine).

Oil-level sensors

A second but equally important aspect of any lubrication system is accurate monitoring of oil levels. Oil tanks in racing applications tend to be complex items, having to control oil levels during extreme cornering and acceleration while taking up the minimum space possible. In the bad old days, oil levels were gauged through the medium of a simple sight glass, but in modern racing things are rather more advanced.

The oil tank is a fairly hostile environment for sensors so they need to be suitable for use in high operating temperatures, and are often hard-mounted to the engine, so they need exceptional vibration resistance as well. Most of these sensors are solid state, the lack of any of moving parts reducing the chances of failure, and they tend to be made from carbon fibre or titanium to fit a team's precise packaging requirements.

The oil-level data will be another key parameter monitored from the pit lane, and temperature sensors are often incorporated into the level sensor. If a sudden drop in oil level is noticed, it gives a team the opportunity to take preventative measures and possibly prevent a catastrophic failure.

Written by Lawrence Butcher

This means that any coolers used need to be fully optimised, but CFD and wind tunnel resources are an expensive and limited resource so inevitably cooling system testing takes second place to overall aero development. However, there are some specialist facilities that will test for both oil and water cooling systems using purpose-built test rigs and wind tunnels.

For initial evaluations of new cores or cross-correlation studies, simple benchmarking exercises can be conducted quickly and cost-effectively in flat-face templates, onto which the cooler cores are mounted. These are then placed in the test section of a wind tunnel, and the inlet and outlet oil temperatures measured, as well as the air pressure either side of the core. However, where more detailed studies are needed or specific cooling problems have been identified, engineers will look to replicate the on-car installation as closely as possible using the actual car's ducting or even complete sidepods.

The key tools for assessing the cooling efficiency of a core are a number of temperature sensors placed at the inlet and outlet of the core and pitot tubes to measure the inlet and outlet airflow. As temperature-controlled flow air is passed over the outer 'finned' surface of the radiator, a temperature-controlled flow of fluid is passed through the internal tube passages. The system is then run through a predetermined test matrix of differing load and airspeed conditions, with the level of heat dissipation being derived for each condition. More complex air pressure measurements can also be taken using arrays of pitots located behind the cooler core to ascertain pressure distribution across the cooler face and to help identify problems such as stagnant flow areas.

Beyond the basic measurements of inlet and outlet coolant temperatures and airflow velocities at the cooler face, Laser Doppler Anemometry (LDA) - a technique for measuring the direction and speed of fluids, using lasers and 'tracer' particles in the test fluid - and thermal imagery to provide visualisation of core performance can also be used. The use of thermal imaging cameras gives a very clear idea of surface temperatures on the core, allowing for hot and cold spots or blockages to be identified quickly. LDA also allows for detailed analysis of air velocity over the cooler.

In its simplest form, LDA crosses two beams of collimated, monochromatic and coherent laser light through the fluid under test. The beams are normally obtained by splitting a single beam, to ensure that the beams remain coherent. These are then intersected at their focal point, where they interfere with each other, creating a set of straight fringes. The lasers are aligned to be perpendicular to the fluid flow, and as the tracer particles pass through they reflect the light into a photo detector. By measuring the Doppler frequency shift of the light it is then possible to calculate the velocity of the tracer particle and thus the velocity of the fluid.

The end-product of this process produces an image very similar to that created during a CFD simulation, but derived from actual test data. As with the thermal imaging, the process allows for a very rapid assessment of the flow characteristics of a particular installation; changes can then be made and re-assessed to determine the most effective method of installation.

Fig. 1 - Oil cooler requirements need to be accommodated into the overall vehicle aerodynamic package (Photo: Lawrence Butcher)

Written by Lawrence Butcher

For many years the standard has been reinforced nitrile rubber hosing with a stainless steel overbraid. Pioneered for use in the aerospace industry, these provided both excellent strength and protection to vital lines and, when combined with screw-type end fittings, excellent reliability. While this was the benchmark for many years, the never-ending quest for reduced weight and more compact packaging in racecars has seen hose manufactures develop new types of hosing to help this cause.

The one big problem with standard reinforced rubber hosing is that there is a finite bend radius that can be achieved before the hosing begins to collapse in on itself, compromising flow. To overcome this, several manufacturers now offer hoses with a convoluted inner core, or alternatively a spiral 'spring' inner reinforcement. The convoluted core allows for the hoses to follow a much tighter radius without collapsing, an additional benefit being that the hose is better suited for use in vacuum applications thanks to the additional rigidity.

The only downside of the convoluted hose is the potential for reduced flow rates and the creation of unwanted turbulence. To counter this, several hosing manufacturers have created hybrid hoses that feature an inner hose with a convoluted outer surface and a smooth bore inside; giving excellent flow characteristics while still maintaining flexibility.

Another considerable improvement in hose construction has been the introduction of PTFE (polytetrafluoroethylene) to replace rubber in hose construction. PTFE has a number of advantages over rubber - notably, it is far more resistant to aggressive fluids such as biofuels with a high ethanol content, and some hydraulic fluids. More important though, especially in the case of oil systems, PTFE has a much higher temperature resistance than rubber, making it particularly useful in application where high temperatures are to be expected. For example, on some superspeedways, NASCAR teams will run oil temperatures in excess of 280 F (140 C), which is beyond the reliable temperature range of most rubbers; PTFE hoses are reliable beyond 400 F (205 C), making them an ideal choice.

The only disadvantage with PTFE is that it is not as strong as rubber, and thus needs additional reinforcement, so the construction of the hose differs. Generally a rubber hose will have fibres incorporated into its structure, added before curing, which increase its strength. In the case of PTFE hoses, most manufacturers will house the inner tubing in an outer sleeve made of a more resilient material, which can be either another hose, made of a material such as silicone, or a metallic of fibrous braided sleeve.

The final key area of development in hosing has been the materials used to protect the inner core. As mentioned, the traditional way to protect rubber hoses is by using a stainless steel overbraid, which provides a very abrasion-resistant covering. It is heavy, however, and where weight is an issue manufacturers have looked to new materials to provide protection. The most popular solution is to use Kevlar or Aramid fibres woven together to form a sheath, which provides a hard-wearing finish (though still not as substantial as stainless steel) for a fraction of the weight. For high-pressure applications, steel is still the favoured medium as its strength prevents expansion of the inner hose.

A quick glance through a hose supplier's catalogue will present a bewildering array of different hose types, with varied performance and price levels. It is therefore imperative that careful consideration be given to the type used and the potential benefits or disadvantages it can provide.

Fig. 1 - An example of a reinforced PTFE hose; note the use of a silicone outer sleeve around the PTFE inner

Written by Lawrence Butcher

The matter of fluid movement within enclosed spaces has been the subject of a great deal of study in a number of areas outside motorsport, notably in the marine industry and among mainstream automobile manufacturers. The first efforts at reliably simulating the behaviour of oil in tanks was in fact undertaken to understand its movement within bulk crude carriers in rough seas - in essence giant oil tanks!

Advances in modern CFD/FEA capability and high-performance computing systems mean it is practical now for racecar designers to incorporate slosh simulation (encompassing any fluid tanks, not simply oil) into the overall design process. This reduces the need for physical testing and reduces the chances of component failure on track due to oil starvation, while also allowing the behaviour of fluids to be taken into account in relation to a car's dynamic behaviour.

The interaction between a fluid in a tank, the tank and any air within the tank is a very complex one, and it is only recently that commercially available CFD/FEA packages have been capable of simulating such interactions. The two key methods used to simulate conditions are VOF (volume of flow) and arbitrary Lagrangian-Eulerian (ALE) formulations, both of which are able to account for the interactions between material and fluid as well as the boundary flow conditions on the surface of the fluid.

The VOF method is a numerical technique for tracking and locating the fluid-fluid interface. It belongs to the class of Eulerian methods that are characterised by a mesh that is either stationary or is moving in a certain prescribed manner to accommodate the evolving shape of the interface. As such, VOF is an advection scheme - a numerical recipe that allows the programmer to track the shape and position of the interface, but it is not a standalone flow-solving algorithm; the Navier-Stokes equations describing the motion of the flow have to be solved separately.

Several commercial CFD packages integrate VOF capability, although it is only in the past few years that the ability to conduct true VOF simulation has been incorporated into them. VOF methods are very effective for calculating the behaviour of fluids but are limited to this role, so if other factors such as heat transfer from the fluid to container material need to be studied, a different set of codes is needed. This is of particular note in racecar design, where items such as the oil tank can be packaged very closely to other components, where heat transfer could be an issue.

It is here that newly developed codes based around an ALE method can prove very beneficial. The ALE is a finite element formulation in which the computational system is not fixed in space as with the Eulerian-based VOF, or attached to material as in Lagrangian-based finite element formulations. When using the ALE technique in engineering simulations, the computational mesh inside the domains can move arbitrarily to optimise the shapes of elements, while the mesh on the boundaries and interfaces of the domains can move along with materials to precisely track the boundaries and interfaces of a multi-material system.

While very effective for fluid movement simulation, the big advantage of ALE-based finite element formulations is their ability to be reduced to either Lagrangian-based finite element formulations by equating mesh motion to material motion, or Eulerian-based finite element formulations by fixing the mesh in space. This means that only one FEA code is needed to perform the range of simulations from fluid flow and fluid-structure interactions through to heat transfer.

Such capability is not widely available in commercially produced packages, but this is likely to change soon. Whichever approach is chosen, it is now well within the abilities of teams with a good computing capability to accurately assess the behaviour of fluids throughout the vehicle and optimise designs to ensure that the performance or reliability of systems is not compromised.

Fig. 1 - Fluid slosh simulated in OpenFoam, a feely available open source simulation package

Written by Lawrence Butcher

Some drag racing classes require the use of wet sump oiling systems, while some competitors prefer the simplicity of not running a dry sump. Interestingly, the problems that often arise in relation to ensuring sufficient oil supply are not always where one would initially expect to find them. Clearly the forces exerted on the oil under hard launches are considerable, but ensuring sufficient pick-up is usually simply a case of locating the oil pick-up at the rear of the sump, eliminating the possibility of starvation.

It is in fact under braking that problems can arise, with users of wet sumps often suffering a considerable drop in oil pressure during braking at the end of a run or after the burnout. One solution was for drivers to kill the motor as they crossed the line, before braking or deploying the parachute, in addition to gradually slowing down after burnouts. However, neither approach is ideal and racers have been looking at better ways of controlling the oil within the sump.

There are number of approaches here. The starting point for improving the oil supply is by increasing the capacity of the sump. Due to issues with ground clearance and engine height, this is often achieve by designing extra horizontal capacity, often called 'kick-outs'. Increasing the capacity is a good start, but better control of the oil's movement is the real solution.

One of the oldest methods is to use a horizontal baffle across the sump, with a series of one-way trapdoors, the purpose being to allow oil to flow to the pick-up under launch but restrict its flow forwards during braking. The key disadvantage with this type of design is the time it takes for the trapdoors to close as the oil tries to move back along the sump, with significant quantities of fluid transferring before the traps are shut.

One solution by manufacturers of drag racing sumps is to replace the trapdoor system with a series of ball valves. These respond much faster than trapdoors, and extensive track testing has shown that they considerably reduce the volume of oil slosh. While baffles and check valves can aid oil control, careful design of the sump shape is also important to ensure that their function is optimised. Sumps will often have a stepped profile, with a deeper section at the rear to ensure there is always enough oil at the pick-up.

As mentioned earlier, ensuring a reliable oil supply is not the only important factor. In many drag racing classes, competitors are separated by hundredths of a second at the finish line, so any horsepower gain at the wheels - no matter how small - can prove decisive.

Power loss due to windage can be significant, especially in large-capacity V8s, which have very big reciprocating components. Windage plates or screens have long been a popular solution here. These take many forms but all try to reduce the potential for oil to splash from the sump onto the crankshaft. There is, however, some debate as to how effective windage plates are, with some manufacturers feeling that they actually increase parasitic losses if they're not properly designed, due to the restriction of oil return to the sump.

Another approach to reducing the effects of windage is the use of oil scrapers, positioned close to the crankshaft, which remove excess oil attached to the crank webs.

Although the wet sump may be a rarity in modern racing, competitors who use them are still pushing their development to try to circumvent the inherent disadvantages of this type of oil set-up.

Fig. 1 - A custom-built sump for a V8 drag engine. Note the substantial 'kick-out' at the rear of the sump and the integral baffle plate in the shallower section

Written by Lawrence Butcher

It was clear that the 375 hp produced by the 255 cu in V8 was not going to be competitive in 1964, regardless of the quality of chassis, so work began to create a new, twin cam unit based on the Fairlane block. The overall development of this engine is interesting in itself, but of greater interest are the problems and subsequent solutions the Ford engineers found regarding the lubrication system. Although the engine in question was developed over 40 years ago, the lessons learnt are still applicable today.

Initially, the design team carried over the existing dry sump system from the pushrod engine, consisting of single scavenge and pressure stage pumps, with the oil stored in a tank towards the front of the car. The first problem the designers ran into was providing sufficient oil to the cam lobes, which necessitated an increase in size for the pressure side of the pump. This in turn placed a greater demand on the scavenge stage, not only in terms of the volume of oil needed, but also as a consequence of the increased tendency for the oil to foam as it drained back from the camshafts.

To counter this, the scavenge stage was doubled in size while the pick-up pipe diameter was increased by 50%. The result was a sufficient supply of oil to the bearings, in addition to adequate scavenging of the crank case.

Another problem the engine suffered from was excess oil being ejected from the breather system. The cause was identified as a combination of two issues: windage from the action of the crankshaft, combined with the large amount of throw-off from the reciprocating components caused by the large bearing clearances. It must be remembered that bearing and machining technology was not as advanced as it is today, and to ensure sufficient cooling and lubrication, race engines tended to run very generous clearances. The breather also needed to cope with the volume of air and gasses being drawn from the sump by the scavenge pump. The problem of breathing was solved by simply increasing the diameter of the tank breather system to reduce the internal tank pressure.

To increase the effectiveness of the system further, a second scavenge pump was subsequently added to the system, which also insured against failure of either pump. Overall, the changes made to the system sound very familiar to those one would undertake today in converting a road engine for race use. In summary they were:

• Redesigned oil pan baffles to control case windage
• Installation of baffles to prevent excess oil build-up near the timing gears
• Better control of oiling to the cams, and redirecting the sump return to prevent oil draining over rotating parts
• Increasing the scavenge pump capacity, with the addition of extra stages to improve effectiveness
• Venting of the oil tank to reduce blow-by

It is encouraging to note that even the mighty Ford had to deal with many of the problems that still beset competitors working on road engine development today - although most clubmen have to go without the benefits of a motor corporation's r&d resources!

Fig. 1 - This is the final iteration of the 1964 oil system; note the twin oil scavenge pumps with independent pick-ups (Courtesy of Ford)

Written by Lawrence Butcher

Running wet sumps has caused problems for many years, as one engineer from a notable engine manufacturer explains, "We've said for a long time that a dry-sump system would be cheaper, more reliable and easier for everyone. Big parts of the budgets are routinely spent trying to make a wet-sump system with the standard production sump pan work in a racing environment. It's been a nightmare for every project we've been involved with, right back to the Sierra Cosworth days."

So what can engineers do to optimise what is essentially a flawed system? Here are some of the approaches used to mitigate the disadvantages of running a wet sump.

Baffles:
Baffles or windage trays are one of the most commonly used systems to control oil in a wet sump. They often consist of horizontal and vertical plates within the sump, and often feature one-way gates to prevent oil surging away from the pick-up. In the case of the S2000 regulations, the cars have to retain the standard sump pan, so controlling the slosh of oil within is vitally important.

However, baffles need to be carefully designed to ensure they do not in fact restrict oil flow towards the pick-up, or slow the return of oil from the upper half of the engine. With powerful CFD simulation packages now becoming more accessible outside the upper echelons of motorsport, engineers have a powerful tool to assess the effectiveness of baffling. Whereas previously a system could only be tested by running the car and logging the oil pressure traces, designers can now get far closer to an optimum solution without making large numbers of prototypes.

Movable oil pick-ups:
One other solution to preventing oil starvation is a pivoting oil pick-up pipe. Instead of trying to keep the oil near the pick-up, the pick-up is able to move around in the sump, using the same force vectors that affect the oil.

However, to be effective, a pivoting pick-up will often require the use of a dedicated sump pan, in order to allow the pick-up to move freely. This means that if the standard manufacturer's sump needs to be retained, the effectiveness of any pick-up design is severely limited.

External oil reservoirs:
If it still proves impossible to maintain reliable oil pressure using either of the above methods, another option is the use of a pressurised, external oil reservoir. In recent years this has been the route adopted by a number of teams competing in the British Touring Car Championship.

The reservoirs are pressurised up to normal oil pressure and then the release of oil is controlled by a pressure-sensitive switch. If the pressure in the engine oil system drops below a preset level, the valve opens, allowing extra oil into the engine to maintain pressure.

Although it's effective, there are a number of disadvantages with the system, notably additional weight and complexity. It also does not give complete insurance against pressure loss - as the author discovered, at the cost of a crankshaft and set of rods, after a pressurised system could not handle the oil surge in a boxer engine!

Fig. 1 - A wet-sump system with built-in pivoting oil pick-up, designed specifically for off-road racing applications

Written by Lawrence Butcher

One example of this is the development work undertaken by the now defunct Minardi team in the production of a new gearbox for the 2005 Formula One season. The team had been working on its own gearbox design in 2003, which was then raced in 2004. During the ongoing development programme, the team's engineers identified the gearbox oiling system as an area for potential performance gain.

The original iteration of the gearbox incorporated a built-in lubricant pump, with the oil stored in a sump section. The volume of oil in the gearbox was singled out as an area where gains could be made, both in terms of reducing power loss due to friction and reducing weight in terms of capacity. This potential was narrowed down to the following areas:

• Reduce the quantity of oil used
• Improve gearbox lubrication
• Reduce parasitic power loss
• Increase the rigidity of the casting

It is in the approach taken to achieve these goals that things begin to get really interesting. Unhappy with the capability of its simulation software, the team decided that being able to visualise fluid flow within the transmission case would be beneficial. To this end, it contracted an outside company to produce a rapid-prototyped centre gearbox section, in clear material. Initially the team had tried to produce the component using its own in-house SLA (Stereo Lithography) facilities, but the properties of the SLA material made it unsuitable for such a structural application. It was for this reason that an outside contractor was called in to produce the section using a much stronger SLS (Selective Laser Sintered) material.

The aim of this approach was to visually asses the movement of lubricant inside the casing, with the gearbox mounted on a low-power dynamometer. The first key change to the system was to relocate the oil pump from inside the casing to an external position, and eliminate the sump, replacing it with an oil tank. This made the system 'dry' and immediately allowed a reduction in the casing's internal size. The effectiveness of this approach was verified by testing a rapid prototype of the original sumped system, which showed that the new dry sump provided ample lubrication.

During these initial tests it also became apparent that the volume of oil being supplied to the main gearbox bearings was excessive, which entailed a reduction in the size of the oiling holes. This proved to be effective, and the lubricant pressure increased with no subsequent increase in pump flow rate. The original oiling holes were 2 mm in diameter, which was later reduced to 1 mm. Under test conditions this equated to an increase in oil pressure of 0.18 bar at 6000 rpm, while still maintaining a pump flow rate of 3.08 litres per minute.

During this testing it was also noticed that, as a consequence of the casing's internal architecture, oil was being distributed to areas of the gearbox where it wasn't needed. Several iterations of the casing design using alternative geometries were tried, until an optimal solution was found. The end result was that the volume of oil in the casing, and hence overall weight, was reduced, with the added benefit of a stiffer casing structure.

With the case design and dry-sump oil system optimised, the design was subsequently produced using a rapid casting process, and using rapid-prototyped cores to produce a titanium casing. Although this is only one strand of a complex full-car development story, the project provides a valuable insight into what can be achieved with a little engineering ingenuity.

Fig. 1 - The rapid-prototyped gearbox containing red hydraulic fluid to aid visibility of flow patterns

Written by Lawrence Butcher

Obviously correct sizing of the 'plumbing' is important - there is little point in having 11/16 in pipes if the oil inlets are only 3/8 in, for example. But what is of greater importance is the effect the lines and fittings can have on system pressure and, more important, the likelihood of cavitation in the oil pump.

Cavitation is a subject in its own right, but in essence turbulent flow, or a reduction in NPSH (net positive suction head) at the inlet to the oil pump can cause considerable damage and wear to the pump components. The most common cause of cavitation in a pump is the presence of tight bends close to the pump inlet. A bend causes pressure losses due to both friction and a change in momentum.

Fig. 1

These losses are dictated by the bend angle, the curvature ratio and the Reynolds number (Re) of the pipe. The Reynolds number is fairly easy to deduce, although some generalisations may need to be made. It is important to note that the internal finish of the pipe will have a bearing on its flow characteristics - a corrugated pipe will have a different Re value from a smooth bore item. Generally, ribbed inner pipes are used due to their ability to bend without deforming, but several manufacturers now have products that feature a smooth inner, yet still give acceptable bend radius.

The Reynolds number for a pipe is calculated using the following equation:

where Dh is the hydraulic diameter of the pipe, calculated using the formula, Dh = 4A/P, where A is the cross-sectional area and P is the wetted perimeter (in the case of a pipe, this is the circumference); v is the mean velocity of the object relative to the fluid (in the case of a pipe, the object is stationary), which can be calculated from the pump flow rate and the pipe area; µ is the dynamic viscosity of the fluid; and p is the density of the fluid (kg/m³).

With the Re defined, pressure drop due to pipe bends can be ascertained using the equation:

where fs is the Moody friction factor in a straight pipe (a dimensionless number that relates pipe roughness, Re and friction factor, and is usually represented as a graph); p is the fluid density; u is the mean flow velocity; Rb is the bend radius; D is the tube diameter; Ø, is the bend angle; and kb is the bend loss coefficient. Values for fluid density (of most common oils and other fluids), dynamic viscosity, Moody friction factor and bend loss coefficients are all available at www.engineeringtoolbox.com along with a host of other pertinent information.

While a drop in oil pressure can be a problem, turbulent flow can be more of an issue. This is where selection of pipe unions can be of utmost importance to pump life and efficiency.

Fig. 2

Fig. 2.1

Fig. 3

Fig. 3.1

Figs. 2 and 2.1 show a cheap and pretty nasty 90º union supplied with a dry-sump system. Admittedly it can be packaged into a very small space, but the bend is practically a right-angle turn, consisting of two intersecting drillings with a minimal radius. This is going to introduce a lot of turbulence into the flow, whereas Figs. 3 and 3.1 show a union that has a nicely radiused bend; while it will still introduce turbulence, it will not be to the same degree as the first fitting. The compact nature of the first fitting does appeal in terms of packaging, and several manufacturers now supply forged fittings, which are compact yet still retain a reasonable bend radius.

While this appears to be a minor detail, and one that could easily be overlooked, it could have a considerable impact on both the efficiency and longevity of an oil system.

Fig. 1 - Oil lines and fittings can play a far greater role in the longevity of an engine than some may think

Figs. 2 and 2.1 - A cheap oil line fitting. Note the drilled inlet and outlet, and lack of radius on the bend

Figs. 3 and 3.1 - A high-quality fitting. While it takes up more space than the previous item, flow characteristics will be far superior

Written by Lawrence Butcher

The design of the oil storage tank is of vital importance to the efficiency of the system, with the tank required to perform a number of tasks beyond simply holding oil. There are several key factors that dictate the design of an oil tank, relating to the specific requirement of the overall vehicle package.

As oil is pumped around the engine it is subject to an extreme degree of agitation, both when running through bearings and as it travels back to the bottom of the crankcase. The crankshaft, con rods and other reciprocating parts churn the oil into a foam-like consistency, of greater volume than the original fluid. It is for this reason that the scavenge stages of an oil pump are generally of a greater size or number than the pressure stages, in order to account for this greater volume.

With the oil removed from the engine, it is then the task of the storage tank to separate the oil-gas mixture back to a substance that can be used as a reliable lubricant. This is achieved by controlling the flow of oil into the tank through the orientation of the inlet and, usually, a series of baffles. The oil is generally fed into the tank in a fashion that will encourage it to 'spiral' into the main oil pool, with a number of flat plates used to slow the oil down to allow more time for it to de-aerate. The plates also act as baffles to restrict vertical movement of the oil under high g-loads, and can be complemented with vertical plates to control lateral slosh.

In high-end applications, such as Formula One and Sports Prototypes, extensive CFD simulation is undertaken to establish how the oil will behave in the tank. This is essential to allow for the minimum quantity of oil to be run without compromising flow to the engine, with every extra litre contributing to the overall weight of the vehicle.

Constructing a tank in the way described above is simple when there is space to package a regular cylindrical unit, but matters become more complex when installation space is at a premium. This is the case with many top-level single-seat racecars, and current Formula One technology is a case in point.

In a current-generation Formula One car, the tank is generally housed on the front of the engine in order to optimise both space and weight distribution. This does not allow for a simple cylinder, however, and the tank's shape is dictated by the space between the engine and the safety cell. The result is a tank that requires many more compartments and baffles to control the oil flow effectively.

Most high-performance race engines run at a negative crankcase pressure, with the dry-sump pump capable of creating a considerable level of vacuum. The oil tank therefore needs sufficient venting to allow the gaseous mixture evacuated from the engine to escape, otherwise the efficiency of the pump is compromised.

From this brief overview it can be seen that oil tank design is one of the key components in an effective oil system, with the overall vehicle package and engine requirements dictating the end product. As with any race machine, every component needs to be optimised, and the humble oil tank is no exception.

Fig. 1 - This fabricated tank on an Arrows Formula One car shows the level of complexity in a high-performance oil tank (Courtesy of Concept Racing)

Written by Lawrence Butcher

For many years, and where natural flow of air around the sump was insufficient, the most common intermediary was an oil-air heat exchanger. Made from a series of pressed or stacked plates, or a system of flattened tubes and fins, these were highly efficient but suffered from packaging problems, often requiring long hoses or special ducting to a position on the body of the vehicle. Mounted near the front or sides, these coolers are also vulnerable to accident damage.

A more elegant solution to cooling the oil is to use an oil-water method, one of which is the Laminova laminar flow oil cooler. It consists of two concentric aluminium extrusions, one of which forms the outer housing into which the engine oil is introduced. The second, a far more complex affair, can best be described as a thick-walled tube, on the inside of which are a number of fully enclosed channels running through the tube wall parallel to the axis of the tube. On the external face of this tube is a series of very fine, 3 mm-high radial fins, 0.2 mm thick and 0.3 mm apart. A quick look at the diagram might give a clearer idea.

As the oil flows across and around the outside of this tube, the manufacturers claim that normally the flow would become turbulent and significantly increase the pressure drop in this part of the oil circuit. However, by introducing channels into the finning running the length of the extrusion, this is somehow avoided and the flow around it remains laminar. Compared with other designs, the pressure drop throughout the oil circuit is very low, while the heat transfer across the boundary between the oil on the outside and the engine coolant on the inside is very high.

As a general rule, turbulent flow in a cooler produces excellent heat transfer but high flow losses, while if the flow remains laminar heat transfer isn't generally quite so good but the flow losses are at a minimum. In many traditional designs, these parameters have to be balanced to give the optimum solution, but in the case of the Laminova, the huge surface area more than makes up for the loss in performance over turbulent flow designs. To optimise the heat transfer as well as selecting the correct size of unit, optional internal cores are available to give the best coolant flow characteristics at minimum pressure loss.

So is this a better design than the traditional plate or tube-and-fin design? Typical of an engineer, the response is: it depends. To use an oil-into-water cooler, the main engine cooling circuit needs to have sufficient extra capacity and therefore a slightly larger cooling area than is otherwise necessary. On the plus side, however ,the heat exchange between the oil and water is likely to be much better, and the design a much more robust affair. Positioned much closer to the engine, accident damage is less likely, and even in the case of engine failure, modular construction allows the unit to be dismantled and easily cleaned before re-use. Also from an engineering viewpoint, since hose runs will be much shorter, the pressure drop in the oil circuit should be at a minimum; and since the engine coolant warms up quicker, at start-up the oil will warm up much quicker, giving better engine protection during this phase.

In refreshing the parts that other fluids cannot reach, the Laminova oil cooler may not be the answer to all applications but it must surely be one worth looking at.

Fig. 1 - Unique core of the cooler
Fig. 2 - Internal core options

Written by John Coxon

Consisting of a pump, a filter, some bearings and one or two other components - not forgetting, of course, the oil - when assembled under the cleanest of conditions and serviced regularly to the same standards, we can expect the engine to last for the full design life (generally at least 150,000 miles, or 240,000 km) and possibly even much longer. The oil pump will pick up the oil from the sump and deliver the volume required, with the surplus returned to the sump via the oil pressure relief valve (PRV). As the engine gets warmer or the speed increases, the oil volume demanded will increase, and to maintain the overall system pressure, less oil will be discharged back to the sump.

The problem arises when - or should I say, if - the demand for oil is greater than that supplied by the pump and when the PRV remains closed. At such a time the pressure in the system will fall. Designed correctly and compensating for the small amount of wear that will inevitably take place, this will normally only happen at very low engine speeds, when high oil pressure is seldom required.

But if, say, the owner or garage puts in a slightly thicker oil than that recommended - a 10W/60 instead of a 10W/40 or 0W/40 for example - then at a given temperature and relief valve pressure setting, the increase in back pressure caused by the thicker oil will cause a larger amount of oil to be returned through the PRV back to the sump. In most instances, so long as the lubricant gets to all the critical parts and readily drains back to the sump then little real harm may take place, but if say the relief valve couldn't cope with the extra volume of oil returned or if, once through the valve the flow effectively 'choked' in the return pipe, then the pressure to the engine oil system could continue to increase. Not so worrying if this increase is only slight but potentially catastrophic if it blew out an O-ring or an oil filter seal somewhere in the system.

The use of higher viscosity oils, however, can introduce other issues as well. In one engine I can remember, approved only for use with 10W/40 oils, filling with a 10W/60 caused the relief valve - which rarely saw such an opening with the correct grade of oil - to jam wide open at high speed. Naturally, when the engine returned to idle and since the PRT was still wide open, the gallery oil pressure simply disappeared. This particular issue was eventually solved at the design stage when the casting length supporting the relief valve stem was lengthened, such that jamming in its locating bore became impossible thereafter. It's a reminder that occasionally designers can get things wrong, but it sometimes takes the actions of the uninformed to highlight it.

Thankfully, most OE engines no longer give access to the PRV since it remains hidden somewhere deep inside the pump. But when these valves are fully accessible, particularly in those systems using an externally mounted dry sump, serious problems can arise. In one incident, the customer of a well-respected pump supplier installed his pump as per the maker's instructions and adjusted the relief valve pressure to 80 or 90 psi, when the engine was hot and at idle speed. Switching off the engine and retiring for the evening, he came back the following morning when the engine had cooled down and the ambient temperatures had fallen to single figures. Upon firing the engine again, the gallery oil pressure simply went off the end of the gauge and oil proceeded to spurt out through the seals and all across the engine bay.

To err, they say, is human but to really make a mess takes a gallon or so of engine oil.

Fig. 1 - The working parts of the oil pressure relief valve

Written by John Coxon

When circulating around an engine, the lubricating oil is subject to temperature and pressure fluctuations, which alongside the shearing action of the rotating and reciprocating components can introduce a gaseous element into the body of the oil, in the form of small bubbles. These can consist of products of combustion, air or even components of the lube oil that have 'boiled off' in the bearings and come out of solution.

Whatever they might be, the net effect will, surprisingly, be to increase the viscosity of the oil. While this is theoretically good for the valvetrain the presence of this gas can have serious consequences for the rest of the engine - cavitation in the oil pump and bearings, for instance, or inconsistent operation of any hydraulic elements such as cam phasers, switching elements or hydraulic tappets. Consequently, as an engineer I prefer to have my lubricant as 100% 'solid' oil or as close to it as possible.

In a dry-sump engine much of this can be centrifuged out at some stage, but in a wet sump the opportunities to separate gas and oil are far fewer. Often it is simply the case of 'out of sight, out of mind' and since in a wet sump you can't see the oil level you would naturally assume that everything is OK. In a dry-sump system you will immediately know if all is not well when the oil level rises in the tank. But in a wet-sump engine such luxuries are not available, so the only way is to sample the oil itself. Taking a 100 ml sample of the oil-air mixture from a tapping in the gallery at the back of the cylinder head, diverting it into a measuring cylinder and observing the final oil level as the air gradually rises to the surface, you might be very surprised!

To alleviate the problem you have one of two choices. First, you can increase the time it takes for the oil to drain back to sump, allowing time for the air bubbles to coalesce and come out of solution. A gently sloping shelf underneath the crankcase leading to a deep 'well' where the oil can be picked up by the pump again will help here. Or second, you can increase the volume of oil in the sump by making the sump bigger. In a way, the latter amounts to much the same thing and allows time for the air to come out of solution. But whichever way you choose, with the wet sump you still have the problem of oil surge and a slightly higher engine, neither of which is desirable for racing.

Air in the oil of a wet sump engine may be something you don't ordinarily think about but just because you don't see it doesn't mean it isn't there.

Fig. 1 - Measuring the air content of engine oil

Written by John Coxon

To those assigned with cooling the engine and students of vehicle design, these decisions are no doubt all too familiar. But when it comes to the oil system, however, especially dry-sump systems, I doubt if many have even thought about it.

With hoses running from the engine to the scavenge pump, and from the scavenge pump to the oil tank as well as from the oil tank through the filter and back to the engine again, there would seem to be plenty of opportunity to stick in the odd cooler here or there. But when it comes to deciding the optimum place, what should we be thinking about and where should it go?

Let me reassure you right at the start, as far as I am concerned there is no right or wrong place. The important feature, however, is that the fluid should enter the engine at the required temperature and the pressure drop across the whole system should be wholly acceptable at the flow rate required.

Some ways may be slightly less efficient in terms of the size of the cooler required while saving on the hose lengths needed. Similarly, the positioning of the cooling matrix may be more critical from an aerodynamic viewpoint and, although important, size and weight may not be the overriding characteristics.

In the end, there are basically two main approaches - running the scavenge oil through the cooler or to run the pressurised oil through the cooler. In the case of the former the oil, as a mixture of oil and air/combustion gas, will have a considerably greater volume; in the case of the latter the cooler will experience full gallery oil pressure, and will consequently need to be stronger and therefore heavier. Alongside this, with scavenge cooling the radiating matrix may need to be larger to cope with the reduced thermal efficiency, while that of the pressurised version will be smaller and more compact.

However, for the convenience of plumbing, it is often better to cool the scavenged oil. But depending on the temperature increase across the engine a hot oil is much easier to de-aerate than a cold oil, and so as well as all the above, the size of the oil tank may also be reduced if the oil is easily de-aerated. On balance, I therefore much prefer to mount the cooler in the pressurised line, for not only is the oil in a better condition and heat dissipation more efficient, the flow through the scavenge system will be less restricted.

In the end, the best method is the one that works, and while opinion is divided it is hard to get it totally wrong. The important thing, however, is to place the filter just prior to the oil going into the engine main oil gallery.

Fig. 1 - Oil cooler on the Gold Leaf Team Lotus, Lotus 49, when the oil was cooled using a separate system

Written by John Coxon

For a vee engine, therefore, we are likely to see at least four scavenge stages to the oil pump. The problems, however, start to multiply when this installation includes one, possibly even two exhaust gas-driven turbochargers.

Unlike mechanical, crankshaft-driven superchargers, turbochargers need a high-pressure oil feed not only to lubricate the fully floating bearings but to cool them as well. Engine oil at engine main gallery pressure is consequently taken from a convenient tapping point and directed into the top of the turbo bearing housing.

Here, when the turbo is running at high speed and the fully floating bearing system is working as it should, the oil is chopped up and mixed with the air and any exhaust gas leakage to form a dirty whitish-brown mixture that can have the texture of whipped cream at times. This falls down the much larger oil drain and discharges into the sump above the level of any oil in it. And so long as the turbo unit is mounted high enough for this mixture to drain into the top of the crankcase - which is what turbo manufacturers specify - everything else being equal, all should be OK.

A problem arises, however, when the engine packaging boys get their way and the delicate balance of pressures inside the turbocharger bearing housing become disturbed, with potentially catastrophic results. Insisting on low-mounted turbo units, one either side and hidden snugly from view in the car sidepod, this can present major problems to the engine designer.

You see, oil simply does not flow uphill. And when the turbo bearing housing is lower than the engine sump, the tendency would be for the oil to flow from the sump into the bearing housing, rather not the other way around as nature intends. Creating a depression in the sump might help, but the tendency for the oil to drain back towards the turbo will always prevail.

If I had my way, this would be the first thing I would change. If liquids could flow uphill naturally, without any coercion, I am sure many common engineering problems could be solved overnight! It might create some more but we wouldn't need to provide additional scavenge underneath each turbo to direct the oil mixture back to the oil tank. We probably also wouldn't need a small-volume/catch tank underneath the turbo bearing housing to collect the oil and allow enough time for at least some of the air to diffuse out.

If oil could flow uphill we wouldn't therefore need a balance pipe to vent this air away from the bearing housing to a catch tank or the engine breather system, and life would be much simpler. At the same time though, it wouldn't be quite so much fun for us engineers!

So the next time you see turbochargers mounted low down in the engine sidepod, just think about the poor engineer who had to reverse the laws of nature to get the oil to drain uphill.

Fig. 1 - Sidepod-mounted turbocharger installation

Written by John Coxon

In the case of the transmission, the object is to transfer the tangential force from one rotating shaft to that of another, while for the pump it is not so much the transfer of force but that of the motion of the trapped fluid. Looked at in this way, it must be clear that when asked to perform different tasks, the design of the gears must be subtly different.

In transmitting motion between parallel shafts, external spur gears are the ones most often used for gear pumps and transmissions alike. The side or flank of the tooth, that which does all the pushing work, is usually shaped in the form of an involute and is essentially unchanged between the two.

At the base of the profile, where this flank meets the body of the gear, is the fillet radius or root while at the top, where the flank meets the outside radius, is the tip. For transmissions where shaft misalignment under load can easily occur, the tip can often be crowned to provide relief. Also, since the teeth can be heavily loaded, provision has to be made at the base of the tooth to ensure that the profile blends into the root diameter. This has to be big enough to minimise stress concentrations but not so big as to weaken the tooth overall.

When it comes to gear pumps, however, the loads experienced are nowhere near those in a gearbox, so 'crowning' at the tip of the tooth is not as necessary and - dare I say it - even undesirable. Indeed, its provision may actually reduce pumping efficiency as a result of excessive leakage. Since the tooth bending loads are also smaller, the root radius can be also be reduced.

In essence, and in terms that gear people might recognise, the best gears are those that have been cut to the whole depth. It is essential therefore that the root radius is kept as small as possible to keep clearances to a minimum.

In the real world though, the temptation is for pressure and scavenge pump manufacturers to use standard 'off-the-peg' components. Made in thousands under closely controlled conditions, these will be much more accurate and with fewer defects, making them very much cheaper than any bespoke units likely to be produced. Designed for power transmission, they will most likely have relieved tips and larger root radii than perhaps is ideal, but the quality of manufacture and the accuracy of the profile will more than make up for any loss in pumping efficiency.

Although the requirements between the gears of a transmission and that of a pump may differ in a number of significant details, in reality - though it hurts me to say so - the actual differences are likely to be very little, if any at all.

Fig. 1 - Shaft or pump gears?

Written by John Coxon

Once we have removed the oil from the engine, however, one disadvantage is getting back into a state whereby we can introduce back to the engine. Filtering and cooling the oil are relatively simple, but removing the air and gaseous products of combustion takes a little more consideration.

When flowing through the bearings, the engine lubrication oil is subjected to a high degree of shear. As it leaks out of the bearings, the constant thrashing it receives from the pistons, con rods and crankshaft increases its overall surface area and mixes it up with the gaseous mixture that resides in the crankcase with it.

Depending on the age of the oil and the amount of agitation received, the resulting foamy mixture can take the form of whipped dirty light-brown cream. With a volume greater than that of the oil that was pumped into the engine, and which appears to flow not quite so readily as the base oil it replaces, the mixture can be difficult to handle.

In a wet-sump engine this mixture is left to settle out as best it can. The gas contained within it is held in the form of small bubbles by the action of the surface tension of the oil. These bubbles have to coalesce into larger bubbles and then into yet even larger ones before finally separating out into gas and oil.

Under the conditions prevalent in the average engine sump, this is a time-dependent exercise according to parameters such as the temperature and the viscosity (and hence surface tension) of the oil. Manufacturers of wet-sump engines assist this process by providing a gently sloping shelf underneath a significant area of the crankcase to slow down the movement of the mixture and allow time for the components to separate before it falls into the well of the sump. By picking up the oil towards the bottom of this well, enough time has elapsed for most of the gas to separate, and the oil is ready to be pumped through into the bearings again.

In dry-sump engines, the churning of the scavenge pump(s) is likely to increase the amount of gas entrainment into the oil, so it is even more important to separate the two. Once scavenged, the oil is pumped into a cylindrical tank such that the angle of incidence imparts a circular swirling action of the oil within it.

The size of the entry pipe and the volume of flow rate through it (the tank) and its diameter are such as to act like a centrifuge, with the air-gas mixture moving towards the centre while the oil keeps to the outside. Picked up again at the base in its de-aerated form and pumped through a cooler and a filter, the oil can be directed back towards the engine again. Since the total volume of oil in a dry-sump system will almost always be greater than that in a wet sump, and assuming the bearing flow rates between a wet- and dry-sump engine are the same, the greater volume of oil in the dry-sump system will allow even more time for the oil and gas to separate.

A dry-sump oil system containing more oil than a wet-sump equivalent? That seems a bit of a paradox to me.

Fig. 1 - Dry-sump tank showing the position and angle of incidence of the oil intake

Written by John Coxon

The geometry of the gerotor oil pump is developed from such a simple approach. Rolling one circle around another, but in the special case when the radius of the smaller circle is exactly half that of the larger, a shape similar to that used in a Roots-type supercharger is produced.

The Roots-type supercharger is simply an air pump inserted into the intake system of an engine to force air into it. But when scaled down and fitted to a dry-sump oil system, the idea takes on a greater significance as a scavenge pump. And if the Roots supercharger is still a popular way of pressure-charging an engine, the same concept when used as a scavenge pump is perhaps more appropriate.

For its size, the Roots-type pump is the most efficient method of extracting an air-oil mix from the engine. With its large intake and output ports, and relatively cavernous chambers - varying only slightly in section throughout compared to other types of pump - the resulting flow is smooth with only a low level of pulsation.

The downside is the challenge of manufacture, so to keep the pumping efficiency high the clearances between the rotors have to be kept to a minimum without actually touching either each other or the pump wall. At the same time though, if we increase the clearance and veer away from the geometrically correct shape, the pumping efficiency will inevitably fall.

Because of this required accuracy of manufacture, the lobed rotors have to be geared together using a system of spur gears with minimal backlash. But extra gears and the problems associated with machining, together with all that extra material in the rotor, means only one thing - extra cost. With the development of cheap gerotor parts sintered to a net shape and which don't require an additional outlay of gears, it is easy to see therefore why in recent times, despite the benefits, Roots-type scavenge pumps are no longer popular.

This problem of the clearances required has a direct influence on the choice of material. Using conventional machining techniques, rotors tend to be solid and therefore, if made in scuff-resistance steel, comparatively heavy. Since the aluminium housing with a higher coefficient of thermal expansion would expand away from the steel rotor, these clearances would increase with temperature and the efficiency of the pump fall.

Ideally, therefore, rotors should be made of an alloy aluminium with a thermal expansion similar to that of the housing. This will also minimise the weight issue. Unfortunately aluminium may not be the best rotor material because, unless covered with a much harder wear-resistant coating, the debris which always passes through a scavenge pump at some time or other will very easily damage the surface.

Coatings of this type create even more expense, which could be the reason why we don't see Roots-type scavenge pumps more often.

Fig. 1 - The Roots-type scavenge pump

Written by John Coxon

So every 2500 or 3000 miles, I forget which, I would dutifully warm the engine of my ten-year-old Triumph TR4 and drain the oil, allowing at least 30 minutes for as much of it as possible to drip into the battered container underneath. Then I'd clean the sump plug and carefully replace it, ensuring that the copper seal was still up to the task, then replace the oil filter with the same fastidiousness and fill up the engine with the required amount of fresh oil. At eight or nine pints if I remember correctly, this was no cheap task even back then!

Subsequently, carefully removing the spark plugs and with the ignition coil disconnected, I would carefully crank the engine until full engine oil pressure was reached. Replacing the plugs and reconnecting the coil, the engine would be fired up and after a few minutes the filter housing and sump plug examined for leaks.

Later on in life, when the TR4 was replaced by the tow wagon, whenever the oil in the racer was changed or the engine fired up after a period of inactivity (usually 24 hours or more) a similar procedure was adopted. Whether I was changing the oil or intending to start the engine after a period of inactivity, the plugs would be removed, electrics disconnected and the engine spun over until at least 40 psi showed on the gauge.

I mention this because rarely do I see this procedure enacted inside the paddock early in the morning when firing engines for the first time. Possibly because roadcar oil-drain periods are much longer these days (up to 20,000 miles in some cases) or possibly because of the difficulties of removing spark plugs and isolating electronic fuelling and ignition systems, the practice of priming the oil pump before firing the engine for the first time or after oil changes seems to me to have been long since forgotten.

On wet-sump engines when the pump is inaccessible, the practice is perhaps excusable. Even so, I have seen production engines with wet sumps that if left for a few weeks without running would quite easily lose their oil pump prime, with catastrophic results. But on dry-sump engines, especially if the oil tank is located some distance away from the engine, it is inexcusable and courts disaster. Of course, there will always be some residual oil left in the feed oil gallery and bearings, but with remote mounted dry-sump oil pumps some or even most of this oil may have drained away once the engine has been stopped.

To fire the engine and introduce combustion loads into the bearings therefore seems to me to be something akin to a lottery. In some cases as well, if the oil tank is mounted higher in the chassis, oil has been known to drain down into the oil pan and cause hydraulic locks when the engine is cranked for the first time after a long lay off. Cranking without the plugs will indicate a problem without the possibility of serious engine damage.

Even cranking without firing or removing the spark plugs is surely better than nothing, but you'll have to disable the fuel injectors or else you might introduce other issues such as fuel entering the oil system and the problems that brings, not to mention the bore wash.

No, I may be old-fashioned or fastidious (or both) but the only way to ensure your oil pump is fully primed without risking any damage is to spin it over without any plugs. At the very least you might discover any issues before any serious damage can occur.

Oh yes, and just listen to your old dad now and again! He's probably been there and learned the hard way - from his mistakes.

Fig. 1 - Simple dry-sump system

Written by John Coxon

In the 1960s, a dry sump was totally unheard of outside the top echelons of the sport. At club level it probably meant a stone through the base of the engine and a trail of oil. To counteract oil surge in corners, competitors used to baffle the internal portions of the sump to delay, if not totally prevent, its onset. Horizontal and vertical plates would be welded into the sump in the hope they would prevent the oil from sloshing around yet still enable it to drain back to the pick-up pipe. Incidentally it was always the horizontal plates that seemed to be more effective than the vertical ones, presumably because they prevented the oil from disappearing up into the crankcase.

Sometimes volume had to be introduced by welding a monstrous selection of steel plates on the side, rear or front of the sump. The increased volume of oil thus created would ensure that under surge conditions at least some of it might reach the pick-up, and in so doing prevent total disaster. Ingenious swinging trapdoors could also be used, as well as what I think is the neatest idea of all - a swinging pick-up pipe, which floats around in the base of the sump and, hopefully, goes where the oil goes.

If restraining the oil in the sump or chasing it around could be considered as one choice, the other must surely be the use of some form of oil accumulator. Consisting principally of a giant piston and a spring/pressurised volume in a tube, the idea is to pressurise an amount of oil that can be released back into the main oil gallery, should the pressure in there fall momentarily. Fiendishly clever, the system can also be contrived to retain the pressure of the oil in its cylinder while the engine is stopped and then by opening some form of tap, releasing it into the engine and effectively 'pre-lubricating' the bearings before cranking.

Since the vast majority of wear takes place in the first few seconds of firing this is highly desirable. When regulations require competition engines to run with 'wet' sumps, the advantages of accumulators such as these can be very attractive.

Ingenious this might be, however, the big advantage of a dry sump is that it enables the engine to be positioned much closer to the ground. Removing the sump and positioning the oil in a more advantageous position not only enables the engine to be lowered but gives an opportunity to reposition the oil tank to benefit vehicle handling. Better cornering and reduced weight transfer under braking therefore gives the competitor so much of an advantage, and when systems such as this are so relatively inexpensive and work well, why make life difficult?

Fig. 1 - A baffled sump

Written by John Coxon

At first and fairly self-explanatory really is whatever is pumped into the engine, should come out. To avoid the engine gradually filling up with oil, the scavenge pump at a very minimum, has to return all this to the external oil tank. In addition there will inevitably be a small amount of combustion blow-by. As power unit engineers we try to keep this to a minimum because exhaust gas blow-by leaking into the crankcase is effectively lost combustion pressure and hence power. High blow-by can sometimes equate to low oil consumption but in general we tend to regard anything over certain limits as some kind of failure in the ring pack.

As anyone who has ever played with engines before will tell you, an engine whose crankcase is vented to atmosphere will always deliver more power than one that is fully sealed. The build up of pressure in the crankcase acting against the piston and rotating masses therefore has a positive effect in reducing the power and hence as a converse, any reduction should have the opposite effect. Race engine designers therefore are a canny lot and realise that if you size the scavenge pump(s) to pump much more out than you are pumping in, then the overall reduction in crankcase pressure should equate to more power. Crankshaft seals therefore not only have to prevent oil leaking out, but air leaking in.

But there's more. We mustn't forget that the primary reason for any pump is to not necessarily just to remove the oil from the engine but to manage that flow of oil through the engine as well. Oil will inevitably flow from a place of high pressure to low pressure but when mixed with air this draining process may take somewhat longer. To encourage the oil to flow faster not only should the pressure difference be increased (by increasing the vacuum in the sump) but the introduction of an air bleed strategically placed will also help. The exact position and size of this air bleed is the province of trial and error and in the end, a full understanding of what is actually going on inside the engine.

While various rules of thumb exist about the size of the air bleed, say in proportion to the oil supply pump are helpful, it still leaves us with the initial question of how big should our scavenge pump(s) be? Well, unfortunately, to my knowledge no definitive guide exists but a wise old pump man told me if your total scavenge pumping rate is around twice that of the pressure pump then that's a good enough place to start. Introducing small holes into the rocker covers obviously increases this for a given crankcase depression while introducing narrower internal rotors with side (cheek) rings reduces the amount pumped. In the end you will be limited either by the depression in the crankcase and the quality of your crank seals or the power used to turn the pump against the extra produced by the engine. Either way you will be generating more hot air than I.

Fig. 1 - Dry sump pump with one pressure stage and two similar sized scavenge stages

Written by John Coxon

Now before you start to think that I'm some sort of crackpot with a grudge against the filter manufacturers, I've done my homework. Once many years ago I had occasion to run a series of tests on an engine with the oil filter removed. The tests, part of a research project into engine wear using radioisotopes, were conducted in an engine dynamometer test cell, monitoring the build up of radioactivity in the oil due to the minute quantities of wear that was taking place in the top ring reversal point on one of the cylinders. At first we ran a traditional set up monitoring the build up of radioactivity in both the oil and filter, but for some reason, the exact details of which have long been forgotten, it was necessary to remove the filter and continue testing purely monitoring the condition of the oil.

Testing continued apace with a series of tests each lasting up to ten hours of wide open throttle running on a proprietary 1600cc gasoline engine. After each test the oil was carefully drained, flushed and replenished with a new, slightly different test lube. Altogether we must have completed some 15 or more tests with very strong and repeatable radioactive wear signals coming from the oils. Altogether with flushing cycles and warm up cycles, the engine must have covered 200 or more hours of quite arduous testing, but it was at the cessation of testing when we received our greatest shock. Once the engine had cooled down and had been carefully stripped adopting all the usual techniques and safety precautions necessary for handling (only very slightly) radioactive engine components, there was no visible evidence of wear of any sort on any of the critical parts. Wear there most obviously had been since we had monitored it during the tests and confirmed it using more sensitive off-line radioactive measurements. And the wear we had monitored in the oil signal was greater than we had measured in the oil during earlier tests when the filter was still in the line.

The conclusion to all this was while the filter was obviously trapping some wear debris, when removed, that debris was retained in the oil no doubt held in suspension by the detergents but it didn't do any harm to the engine overall. The tests were short (less than 10 hours) and the oil changes were made under ideal conditions, so why can't we do the same with our sprint car engines and dispense with the oil filter? Discuss.

Fig. 1 - A typical oil filter.

Written by John Coxon

Low or no oil pressure at all is a common enough problem and while the obvious (to some) culprit might be the oil pump, in 999 cases out of 1000 this does not turn out to be the case. Their logic is simple. No oil pressure - it must therefore be the pump. But in reality there are 1001 other, more likely causes. At this juncture, it is well worth remembering that an oil pump is designed to produce flow. The resistance to flow as a result of restrictions in the oil galleries, bearings and journals introduces the pressure element in the equation and therefore should anything change in these there will be a corresponding change in the oil pressure. Replacing the oil pump on such occasions is therefore most unlikely to resolve the issue.

Funnily enough, the most common reason for low oil pressure is simply lack of oil. Although the theory on oil volatility with modern 'synthetic' oils would tend to suggest otherwise, in practice many of these 'thinner', less viscous oils do seem to require more frequent top up than older products. Whether this really is the case or not is difficult to ascertain, but the stark fact is that vehicle manufacturers tend to have more warranty claims put down to low oil levels (in other words high oil consumptions) than ever before. This may of course have something to do with owners being less likely to check their oil levels on a frequent basis but whatever the cause, the result is still the same.

Another cause of no oil pressure, as silly as it may sound, is actually a faulty gauge or sender unit. In competition cars, gauges, whether they are mechanical or electrical have a hard time of it. High underbonnet temperatures and rigidly or stiffly mounted engines create the most difficult of regimes in which excessive vibration transmitted through to the chassis or the shock loading experienced by an off-road event will almost certainly cause any electrical or mechanical gauge to fail in time. I once measured the 'G' loadings on an electrical component mounted off the side of the cylinder block of a 4 cylinder engine mounted somewhat rigidly in an engine dynamometer cell. At resonance, the accelerations measured were in the region of 25-30 times that due to gravity, which proves if proof were ever needed, the hostile conditions under which many of these components are expected to live.

Blocked filters and loose or partially restricted oil pick up pipes are another less obvious cause but the funniest if not a slightly scary one, is that of the wrong dipstick. As an engine manufacturer you would never believe the trouble that this seemingly simple yet highly practical approach to measuring the oil level can cause. Most engine manufacturers will have different engine derivatives according to the intended vehicle in which it is to be installed. In many cases with different levels of 'dress', that is positioning of things like power steering pumps, generators and air conditioning pumps, the position of the dipstick will need to change for reasons of access. Different positions of the dipstick require different lengths and while this can be readily controlled during manufacture, in the service workshop practices may not always be that robust. You can therefore appreciate that after changing the oil and replacing the correct volume of oil for that car, having temporarily misplaced the dipstick another was found to 'plug' the hole. Slightly longer and calibrated for a different dress of engine, the red oil warning light came on long before oil level reached the minimum point - on the dipstick.

If ever there was a moral to these stories it is this. If you should ever be in the position of experiencing low or heaven forbid, no oil pressure, check the simple things first before pointing the finger at the oil pump.

But better still, check your oil regularly whether race or road.

Fig. 1 - Perhaps the most critical part of many oil systems - the dipstick.

Written by John Coxon

A complex process having one of many causes, for an adequate explanation we need to understand and explore some of the more basic principles of fluid mechanics. In particular if we consider the case of steady flow along a streamline as the velocity increases at a point then it is an established fact that the pressure will decrease. In the case of a liquid however, this pressure cannot drop below that of the vapour pressure at the temperature of concern. If for any reason the pressure falls to the vapour pressure then the liquid will boil instantaneously and tiny bubbles will appear at that point in very large numbers. As these bubbles are carried along in the flow there will come at time when the pressure will rise again to that above the vapour pressure and the bubbles will instantly collapse again as the liquid condenses. This process of condensing will cause a void or cavity to be created and the surrounding liquid will rush in to fill the space available. Moving in from all sides the liquid will collide at the centre causing an extremely high-pressure zone, which together with the associated shock waves generated may be enough to severely damage the surface of any metal component in the vicinity. The surface doesn’t have to be exactly at the point of the implosion since another process called ‘water hammer’ can transfer this energy rapidly through the liquid. Either way the surface may be subject to considerable fatigue damage, rapidly eroding the surface of the component. Gear oil pumps, as a result of careless design or inadequate servicing, are particularly prone to this phenomenon and high-speed racing engines especially those with dry sump systems, are likely to suffer most.

Unlike most other engines, by nature of their high speed and the fact that air is often encouraged to pass through the crankcase, racing engines tend to dissolve more air into their oil. The presence of the swirl tank at the end of the oil return line will separate out and vent much of this, but nevertheless the feed to the oil pump will almost certainly contain a small but not insignificant amount of dissolved air which will be kept in solution by the pressure of the pump. If the pressure of the oil in the system drops to below that required to keep this air in solution, the air will quickly vaporise and then collapse again as soon as the pressure increases. Depending upon the pump design and local flow conditions, cavitation wear can therefore take place at any point in the system but because of the rapid changes in flow velocity, mainly within the pump rotor. Gear pumps running at very high speeds, much higher than those originally intended, are particularly susceptible to this form of wear.

Poor tank design, air leaks on the intake side to the pump as well as restrictions in the oil pick up arrangement (including inadequate hose sizing) can be other causes of this problem.

Written by John Coxon.

Years ago, and I can’t even remember exactly when, I had occasion to place a small window in the side of a prototype engine. It was the sort of thing we did in those days if we had a problem or there was something we wanted to understand. The same technique has been used with turbochargers but that’s a story for another time. On this occasion however, getting the oil out of the engine was proving to be troublesome and so we wanted to see precisely what was going on in the crankcase. Since this particular episode was well before the advent of video and no one even thought about cine, crouched down beside the engine we could clearly see the connecting rod and the inside of the sump. But when the engine fired all this disappeared into a mass of dirty brown - gloop. Obviously a mixture of oil and air and filling the crankcase void in seconds, it oozed out of just about every port or orifice you could see (and some you couldn’t) within seconds. As a perfect example of an understatement, clearly something was amiss! Stepping out of the test cell at the end of the day with more than just an earful of syrupy mess, the lesson was well learned that in order to get the oil out of the engine a pressure gradient had to be created between the sump and the outside world just to encourage things along.

In fact an independent electric oil pump was installed and the oil scavenged out at the base of the sump and into a separate tank and before long, still crouching beside the side of the engine the mist literally and slowly began to disappear and I could see once again the blur of the rotating crankshaft. But would you know? At a fixed engine speed, as we gradually increased the speed of the pump and the crankcase depression began to increase, at the same time the power to the engine also increased. And not by just a small amount either. After many years the memory is a little vague but something in the region of 3-5% seems to ring some kind of bell.

In later years when dry-sump systems became more readily available, as a rule of thumb, I was always told, the scavenge pump should be at least of twice the size of the pressure pump but as rules of thumb go this is only a guide. Mounted on the same shaft as the pressure pump the eventual size and number of scavenge pumps will be a compromise based around the depression created in the crankcase, the increase in engine power at the crankshaft and the sealing technology used, not forgetting, of course, the efficiency with which the oil-air mixture is extracted.

What goes in does eventually come out - but just consider that it needs a little bit of help to get it back into the tank.

Written by John Coxon.

In earlier articles we have established that oil pumps are invariably of the positive displacement type and it therefore follows that the power needed to drive them will be proportional to the volume flow rate produced and hence the engine speed. This flow rate will also vary with the temperature (and hence viscosity) of the oil and the clearances in the bearings but essentially for a given temperature the power consumed by the pump increases more or less linearly with speed. Traditionally the size of the oil pump is based around the flow requirement of the bearings at or near the engine idling speed. At this condition, classical bearing analysis assumes that the void between the journal and bearing will run full of oil but that any leakage from the sides will be made up by the flow from the oil pump.

As the speed increases most oil pumps therefore will be sized and have sufficient capacity to more than make up for the increasing leakage losses until the pressure reaches a certain pre-set value. At this point any further oil pumped will be directed back to the sump via the pressure relief valve. And while the output of the pump rises in proportion to the engine speed the demand from the engine bearing system does not. This additional pressure over and above that needed at the bearings therefore causes unnecessary extra pumping work. If we could tailor this output to that closer to the bearing requirement then pumping efficiency could increase and energy savings could be made.

The lube system in a modern engine can consume somewhere around 3-5% of the overall engine power output. Since these powers can be relatively small (a matter of a few bhp) and the complexity of producing a reliable variable flow pump onerous, manufacturers have often kept to the traditional approach of a simple pump and pressure relief valve. However as continuing improvements in fuel economy or performance becomes more difficult, the need to match the flow output of the pump with the demands of the engine assumes a higher level of importance.

One way of achieving this could be by using an electrically operated independent oil pump. While these have been suggested, I am not aware of anyone save on some kind of research engine who has as yet taken up this approach. Another way more recently introduced into the world of heavy-duty diesels is the split gerotor principle. Flow variation is achieved by dividing the rotor set in two along it’s axis and altering the radial position of the offset (the difference between the centres of the inner and outer rotors) of each of them but in the opposite direction. Much more complicated than the traditional approach, in this way however, the actual displacement can be varied to suit the demands of the engine.

Written by John Coxon.

The ‘gerotor’ pump is a more recent introduction to the automotive world. Although its basic design concept goes back to the beginning of the 19th century, the ‘GeRotor’ principal (derived from combining the words GEnerating and ROTOR) consists primarily of two elements; an inner and an outer rotor. Designed using a trochoidal inner rotor from which the outer rotor is developed, the inner has one less tooth than the corresponding outer while at the same time the centrelines of each run at a fixed eccentricity. Rotating about their respective axes the chamber volume between the two rotors increases and then decreases which when linked to crescent shaped ports, can be used to pump fluids. Having been around quite some time, the automotive business didn’t fully appreciated its advantages until the early 1980s such that today it is perhaps the most commonly found oil pump type available.

Unlike the internal gear pumps mentioned at the start, the gerotor pump has no crescent shaped divider. The shape of the lobes is such that the surfaces of each are always at a tangent to and almost touching that of the lobe opposing it and in sliding contact with it. This, it is claimed, keeps an oil seal between the two elements and prevents backwards slippage as the oil progresses through the pump. A major advantage to this kind of pump over previous oil pump designs is one of packaging and the reduced pressure fluctuations in the delivery port. As engines are moving towards crankshaft triggered ignition and no longer requiring a distributor, oil pumps running directly from the nose of the crankshaft were more desirable. The flexibility of the design and the reduced costs of manufacturing (using netshaping techniques) also proved very attractive. Furthermore, the opening and closing of the intake and delivery ports over much longer periods than gear pump designs, is less likely to introduce turbulence and cavitation at higher engine speeds. For this reason gerotor pumps are often more popular in high speed engines.

Where space is limited and large amounts of air could be entrained in the oil, lobe pumps may be considered. A rather special case of a trochoidal rotor and based on a design similar to that of the Roots-type supercharger, two lobes when rotating together offer a much increased displacement for the same installed volume. Unlike any other trochoidal system the lobes have to be geared together separately and do not touch, but despite this large amounts of air/oil mixture can be easily handled at small pressure differences. This makes them excellent for use in oil scavenge systems. Difficult to make and highly sensitive to the clearances between lobes, these are not a popular solution.

Written by John Coxon.

In any high performance unit the oil has two primary functions: to lubricate (obviously) and to cool. For more mundane applications we can also include things like ‘to protect against corrosion’ or ‘ to minimise the build up in deposits’ but since I am assuming that the oil won’t be left in the sump for long let’s just stick to these two.

In lubricating the engine, conventional theory suggests that there has to be oil flow at a sufficient pressure to fill the void between the journals and the bearing shells and take care of the amount leaking away at the edges. To this must also be added a small amount bled off to the cylinder head to feed the cam bearings, tappets etc. While this also takes some of the heat away with it, there will also most likely be a series of oil spray jets feeding directly off the main oil gallery directed at the underside of each piston to keep these cool. The exact details of how much and the distribution of the flow will depend upon many parameters, including things like, the bearing clearances, the relative temperatures and, of course, the viscosity of the oil.

Thus the demand is for a supply of oil at more or less a constant flow rate and delivery pressure for a given engine speed. As engine speeds increase so will the flow losses, but this will be more than made up by the output of our positive displacement pump, assuming it is geared directly to the speed of the crankshaft. Any oil over and above that required will be deposited back in the sump or oil tank by the pressure relief valve. Oil pumps best suited to this kind of duty have traditionally been gear pumps, the gerotor pump or occasionally in certain circumstances, the lobe pump.

Gear pumps, or perhaps more exactly external gear pumps, consist of a pair of meshing gears, one driven with the other one idling and they work on the principal of oil being trapped in the void between successive teeth and the outside of the casing, as the gears rotate. Since there is line contact between the gear teeth and the clearance at the sides is very small, there is little or no room for the oil to leak away and so the oil entering the cavity is that delivered to the outlet port. Thus for every rotation of the gear a fixed volume of oil is transferred from the intake port to the outlet. A common fitment on all engines at one time, these would be installed in the sump of an engine and driven from a skew gear off the camshaft, which at the other end of the shaft, could often be found the distributor.

Still widely available, cheap to make and very reliable, these pumps can also be readily found in external dry sump pumps. However in recent years partly as a result of cavitation at higher flow rates this type of pump is giving way to the gerotor pump.

We’ll look into this and the other pump sometimes used, the lobe pump, next month.

Written by John Coxon.

Some oil pressure will obviously be needed to create a flow through the various other passageways and obstacles irrespective of the crankshaft rotational speed but the flow of this oil must be sufficient to make up for the oil lost out of the bearings and ensure that the journal and bearing surfaces are not in direct contact at any time. To create this pressure and generate the required flow the designer can select a pump from one of two main categories: the positive displacement type (PDP) or that from the rotodynamic group.

There are many different sub divisions within these categories but positive displacement pumps as the name suggests, displace a constant volume of liquid for each revolution. Rotodynamic pumps however, do not rely on simple displacement but convert the input power of the pump into kinetic energy of the fluid. A typical example is that of the centrifugal pump, which accelerates the liquid using a rotating device known as the impeller. Similar to that of the compressor in a turbocharger, here fluid entering the centre of the impeller is thrown out radially at high velocity, and is subsequently converted into pressure energy through the resistance downstream. But in selecting the most appropriate design it is important to review their characteristics in the light of the intended application and to understand that these two categories, independent of their precise design, behave so very differently.

The most important characteristic for any pump must be its flow versus the final delivery pressure and, in the case of any positive displacement pump, this is almost totally insensitive to the back pressure it is pumping against. In any internal combustion engine at a given speed this back pressure can vary significantly depending mainly upon the radial clearance in the bearings and the viscosity of the lubricating fluid and since our centrifugal pump has a variable flow rate depending upon the back pressure this doesn’t appear to be the best application. Furthermore positive displacement pumps are also fairly insensitive to changes in the viscosity of the fluid inside the pump, while centrifugal pumps are generally highly sensitive. A very important characteristic if you consider that the viscosity of say a 10W-60 lube oil could be over 140 cSt at 40 deg C but down to 12 cSt at 150 deg C. Even at this limited level of investigation, the positive displacement would therefore seem to be the most suitable of the two types and when geared to the crankshaft in some way, the flow produced would be roughly in line with that required in an engine.

We’ll look at the range of suitable positive displacement pumps next month.

Written by John Coxon.

In part I guess this attitude stems from our road-going transport habits. Buried away deep inside the power unit and out of sight, the only time we might ever think

about it is when that little red light on the dashboard starts flashing or, in some more modern cars (where this light is omitted), strange grinding–type noises start emanating from under the bonnet. For our racer, to be fair, we do think about these things but so long as the oil tank is topped up and the oil pressure gauge shows ample oil pressure, do we really care or even want to know more?

When used for serious competition, most engine oil systems used would now be classed as ‘dry sump’. This means that the oil pump not only has to supply the oil to the engine bearings, pistons and cams but also has to scavenge that same oil out of the engine afterwards. In this way, or so the theory goes, we can reduce the depth of the oil pan (and hence lower the position of the engine in the car) and totally eliminate the oil pressure surge issues associated with high ‘G’ cornering and braking. But getting the oil out of the engine isn’t always the easiest of chores. To start off with, it is far from 100% oil. Sprayed all round the engine crankcase and camshaft chest and mixed with copious amounts of air by the endless churning of the reciprocating and rotating parts, the resulting cocktail can look more like a dirty brown blancmange than the best quality synthetic you poured in. Getting this out of the engine at the bottom therefore requires multiple scavenge pumps which will then need to direct the mixture into some form of ‘swirl’ pot to de-aerate it prior to it going back into the pressure pump and hence onto the engine again. As a very minimum most designers will recommend that the scavenge pump(s) should have at least twice the capacity of the pressure pump and in some cases sometimes as much as four times, depending upon the engine and its oiling characteristics. Since dry sump pump systems are modular there are therefore no excuses for having insufficient scavenge performance available. The reduction in crankcase pressure and improved windage losses could also give other, more tangible benefits.

In a ‘wet sump’ engine, all this is removed from view and tucked well away when this air-oil mixture in the crankcase will slide down the gently sloping shelf forming the base of the sump, and from thence into the well. This process delays the oil sufficiently as it drains back and enables it to release much of the entrained air, and without us even having to think about it. As a young and inexperienced engineer I remember once, many years ago I hasten to add, trying to get the oil out of a prototype engine. This particular research engine although having an integral sump was intended to have only external oil and water pumps and without the budget for a proper scavenge system it wasn’t long before we had all manner of oil/air mixtures coming out of each and every orifice on the engine as well as some we didn’t even know about! In the end we managed to remove the oil from the engine and de-aerate it using other means but needless to say, if ever a lesson was well learned, that was it.

At least one engine builder of repute has been known to run an engine without a sump just to find out what is happening in the crankcase. It may sound bizarre but the lessons he learned will be most surely worth it.

So, design your oil system carefully. Look after it as you would your own heart and may you have a long and fruitful life together.

Written by John Coxon.

Communising parts between the three series should eventually reduce costs by eliminating near-redundant development programs for parts that serve equivalent functions, according to Dr Andrew Randolph, engine technical director of Earnhardt Childress Racing.At the present time, entrants in the Sprint Cup Series are permitted five oil pump scavenge sections where the other two series use four. In any case, the dry sump oiling system needs to use a single oil pump of no more than 9-1/2 inches length and 3-1/2 inches cross-section.The primary differences, aside from scavenge sections, is overall length of the system including seals, bearings, adjusters, bolt-on end plates and covers – but not including the front end of the shaft – which is 10-3/4 inches for the Cup cars and 10 inches for the Nationwide cars and Camping World trucks. There are also differences in the oil pan, where the Cup series allows segmented and radiused oil pans, while the other two require open box-style pans, he explained.

The idea of a common oil pump has bounced around for a few years now but gained impetus in 2009 as sponsorship dollars began to dry up across the board. “Throwing away slightly used parts that could otherwise go to the Nationwide or truck teams just doesn’t work in this economy,” Dr Randolph said.Finally, he acknowledged, the rule makers in all three series are talking with one another and even asking teams and engine builders for input as to what solution would make the collective series more affordable for everyone – in particular the Nationwide and truck groups.The big question, of course, is which configuration NASCAR will adopt – if it follows through on this idea – for all classes and how the teams will work within the decided strictures. There is bound to be whining from all sides.“The lowest cost option would be to spread the Nationwide and Camping World pump and pan to Cup,” said Dr Randolph. “With one less section, it limits what you can do with the pan design and deceases high speed power by about five horsepower in ‘open’ configuration.”

Initial cost will be a factor: replacing pumps, pans and oil lines – which can be very expensive when you are looking at anywhere from 8-15 engines per one-car/truck team. Additionally, teams must re-optimise their complete oil systems, including plumbing adjustments and revisions to oil pans. Certain engines might need adjustments to the cylinder head and to the valley drain system, but that will not be known until the project is undertaken.Oil pumps and ancillary components certainly aren’t the most glamorous of parts in the NASCAR bin but they could serve as many first-line soldiers do: clearing the path for more progressive changes that allow NASCAR to keep costs under control without impacting action on the track. Adopting a single oil pump is a change that would be totally transparent to the fans, according to Dr Randolph, and make the long-term cost of racing less expensive to the teams.It is just up to Sprint Cup, Nationwide and Camping World rule makers to find a middle ground. “I think, down the road, other engine parts could be communised and that would serve to make ‘trickle down’ almost cost free,” said Dr Randolph.