In moving from LM P1 to Hypercar, Toyota switched from a circa-2.4 litre V6 twin turbo to a 3.5 litre V6 twin turbo. HPD has gone in the opposite direction, moving from a 3.5 litre V6 twin turbo DPi engine to a 2.4 litre V6 twin turbo GTP/LMDh engine. That is what it has selected to power the ARX-06 that will challenge Toyota for outright victory at Le Mans from either next year or (more likely) the year after.
The Acura ARX-06 brings an IndyCar configuration and displacement to GTP
The IMSA LMDh competitor is permitted to fit almost anything by way of a four-stroke IC engine between a new-generation LM P2 chassis ‘spine’ and the mandated spec transmission, which incorporates a 50 kW/67 bhp MGU. The Bosch-supplied MGU is housed within an Xtrac-provided spec transmission case which has Xtrac gearbox internals. Fitted within the chassis survival cell, the associated hybrid system battery is from Williams Advanced Engineering.
The entire hybrid system is spec, albeit with the VCU controlling it interfacing with the competitor’s own engine control unit. How the MGU is deployed on track is decided by the individual competitor.
IMSA has stated that assuming regeneration targets are met, its full output will be available whenever the driver demands power at the rear wheels. In turn, the only major constraint on ICE engine choice is that, with or without with the assistance of the MGU, it must provide sufficient performance to exploit and respect an IMSA-mandated power curve.
IMSA’s ‘standard’ baseline power curve has a maximum output of 500 kW at 9500 rpm. However, there is actually a window such that the mandated curve can be lifted from a lowest curve peaking at 480 kW up to a highest curve peaking at 520 kW.
In all instances, peak power is at 9500 rpm while maximum rpm is set at 10,000. The scope for IMSA to shift any given competitor’s power curve up or down is one factor within a Balance of Performance (BoP) intended to equalise the performance of all competing cars, including rival FIA Hypercars.
It follows that the engine might only have to supply 430 kW if the lowest curve is applied and the MGU always assists it. On the other hand, it might have to supply 520 kW if the highest curve is applied and the MGU output is not always available.
The AR24e V6 has a 90º bank angle
Note that those figures are measurements taken by mandatory IMSA-monitored driveshaft torque sensors. If we consider the Dossier on the Glickenhaus by Pipo Moteurs P21 in RET 128 (December/ January 2021), a 6.2% loss between flywheel and driveshafts was quoted as the figure Pipo Moteurs works to for the SCG 007 Hypercar. On that basis, 520 kW at the driveshafts implies (in old money) about 740 bhp at the flywheel, so that is the target figure for an LMDh IC engine at 9500 rpm.
HPD’s existing production-based 3.5 litre V6 twin turbo DPi engine has a maximum engine speed of 7050 rpm, and given the current IMSA-mandated boost curve its maximum power is in the region of 600 bhp at the flywheel. On paper, if it were to be run to 9500 rpm it could conceivably output 740 bhp. However, its production base might not make that crankshaft speed feasible, and in any case that provenance must have been a concern when HPD was considering its LMDh engine.
Engines confirmed
So HPD has produced a bespoke LMDh engine, dubbed the AR24e. This is the final confirmation of the four LMDh engines that will debut at the 2023 Daytona 24 Hour race. The others are a naturally aspirated 5.5 litre V8 from Cadillac, a 4.6 litre V8 twin turbo from Porsche and a 4.0 litre V8 twin turbo from BMW. That poses the question as to why HPD has gone for such a small displacement engine.
In going down this route, HPD has been able to draw on Honda’s Formula One as well as its own IndyCar experience of small-displacement V6 turbos. Also, it is surely no coincidence that it currently has the same configuration engine under development for IndyCar from 2024. Those rules call for a 2.4 litre V6 twin turbo with a 90º bank angle, characteristics the LMDh unit echoes.
Both the LMDh and IndyCar IC engines have not only been designed and are being developed by HPD but are also manufactured in the organisation’s Californian facility. Logically, HPD is sharing development costs between them.
The 2024 IndyCar engine will run to 12,000 rpm, and from a plenum pressure of 1.6 bar absolute will produce in maximum boost road course trim around 800 bhp at the flywheel using the latest ‘renewable ethanol’ fuel. It isn’t inconceivable to imagine a version retuned to a peak power speed of 9500 rpm and, given a little more boost, producing 740 bhp at the flywheel using the bioethanol-based ‘sustainable’ fuel now supplied at Le Mans.
Arguably, the biggest difference is the 2500 miles between rebuilds requirement for a 2024 IndyCar engine – far short of the mileage needed to contest 24 hours at Le Mans or Daytona. For that reason alone it is more likely that the new LMDh unit is a specific development in the IndyCar architecture rather than simply a retuning of the same.
Of course, it could instead be a brand new development with its own architecture by coincidence echoing IndyCar 2024. On paper then, what would be the advantage of a such a costly clean-sheet design?
The 2024 IndyCar IC engine’s bore is a mandated 96.5 mm and needs a bore spacing such that it can squeeze within a 460 mm span between monocoque and transmission. A smaller bore could make for a more compact and lighter engine; however, the LMDh regulations mandate a much higher minimum weight for the engine. At the same time, a smaller bore implies a higher level of mechanical stress for a given engine speed – owing to the correspondingly longer stroke – which is not in the interest of 24-hour reliability.
On the other hand, LMDh actually offers a 640 mm-long space for the engine. A larger engine base could be a structural advantage but again the vast implicit investment would seem to make that route unlikely, particularly as any engine-related gain will be neutralised under the BoP concept. At the end of the day it seems such an investment would only be justified if the architecture shared with the IndyCar V6 was not up to 24-hour race duty, which seems highly unlikely.
If we turn back the clock we find that HPD’s very first in-house engine design was the naturally aspirated 3.4/4.0 litre V8 Prototype unit of 2007-13. That in turn was derived from Honda’s existing IndyCar unit, which had been designed by Ilmor to 3.5 litre IRL regulations. The Prototype engine shared the same basic architecture but was otherwise a bespoke design for LM P1/P2. HPD won’t admit it but it would appear the AR24e echoes that approach.
Since producing its pure race V8 Prototype engines, HPD has developed a production-based V6 twin turbo for both LM P2 (at 2.8 litres) and DPi (at 3.5 litres). All it will say in moving from the 60º bank angle of that unit to the 90º mandated for IndyCar is to observe that this is to the benefit of centre-of-gravity height. But if that height is the driver then why not go the whole hog to 120º, which like 60º is the basis of an endurance-friendly smooth-running even-firing V6?
And if 120º is considered to result in a unit that is too wide, it doesn’t follow that the IndyCar-mandated 90º is a logical compromise. A compact 90º V6 with a three-pin crankshaft will naturally have those pins disposed at 120º, yet it will be burdened with uneven firing intervals and an unbalanced secondary rocking couple. However, this endurance-threatening drawback can be overcome by the use of split crankpins.
Once that approach is accepted, one can investigate different bank angles and the consequent split pin overlap within the context of acceptable vibrational characteristics. In developing its current Formula Two V6 turbo, Mecachrome reasoned that opting for a 95º rather than a 90º bank angle beneficially increased overlap, strengthening the split pins and hence the crankshaft overall without vibrational compromise.
The ARX-06 will challenge for victory in next year’s Daytona 24-hour race
In any case, HPD has admitted to a 90º V6 and to only 2.4 litres for the twin turbo AR24e. Considering the implications of this choice gives us an interesting insight into LMDh engine requirements, even if HPD itself will not shed further light. For a start, what are the pros and cons of thus trading displacement for higher boost?
On the one hand, a smaller cylinder means that the loading imparted to the big end by the weight and motion of the con rod/piston assembly is inherently less.
Comparing the AR24e with the rival Porsche engine in this respect, unfortunately we don’t have the reciprocating weight or con rod length for either engine. However, if we assume the AR24e has the IndyCar-mandated bore of 96.5 mm and consequently a stroke of 54.6 mm (for 2396 cc) then its mean piston speed at 9500 rpm is 17.22 m/s. The Porsche has a 95 mm bore and stroke of 81 mm (for 4593 cc) so its mean piston speed at 9500 rpm is almost 50% higher, at 25.55 m/s.
On the other hand, there is the higher cylinder loading attributable to the AR24e’s higher boost requirement. In this respect let us consider as a baseline a naturally aspirated LMDh engine for which we can assume a BMEP somewhere between an unambitious 14.0 bar and a more ambitious 16.0 bar. To obtain 740 bhp at 9500 rpm given a (24-hour friendly) 14.0 bar BMEP requires a displacement of 5.0 litres. Clearly Cadillac is playing safe with 5.5 litres.
For its part, Porsche will require next to no boost to obtain 740 bhp at 9500 rpm from 4.6 litres. By contrast, the AR24e will need sufficient boost to obtain a BMEP of 29 bar. That suggests up to 2.0 bar plenum pressure. While that is rule of thumb, it does indicate the AR24e is likely to exploit a higher plenum pressure than a 2024 IndyCar engine, which as noted will be restricted to 1.6 bar absolute.
However, it should also be noted that the IndyCar boost limit is exploited as far as possible throughout the rev range. By contrast, the AR24e will tailor plenum pressure to the prevailing engine speed so as to match the mandated LMDh power curve. For example, that curve starts at 5500 rpm, at which the ‘standard’ figure is 246 kW, meaning with 50 kW coming from the MGU an engine contribution of 196 kW. That translates to about 280 bhp at the flywheel, hence a BMEP requirement of 19 bar, which implies a substantial reduction in plenum pressure from what is needed at peak power speed.
Pressure points
Also, if the AR24e is given the ‘standard’ 500 kW curve and if it does have the MGU constantly supporting it then at 9500 rpm it will be required to supply 450 kW to the driveshafts, and that in turn implies about 640 bhp at the flywheel. A 2024 IndyCar output is expected to be about 650 bhp at 12,000 rpm, given the 1.3 bar plenum pressure mandated for the Indianapolis 500. Scaling that back to 9500 rpm and increasing the plenum pressure to compensate implies that the AR24e will need about 1.6 bar to obtain 640 bhp.
Also, because the IndyCar engine is governed by a mandatory limit on plenum pressure, the trick is to run it as hard as possible within that context, pushing it to the verge of detonation. The IndyCar engine’s detonation events are constantly monitored during its life between rebuilds. How much detonation to which it is found to be subject dictates how hard it can be pushed on an ongoing basis. By contrast, the AR24e can be run to the mandated power curve using engine settings that in theory keep it clear of detonation.
It is also worth noting that these days the stress caused by boost can be countered by, for example, the use of steel pistons. The Achilles heel of any such turbo engine is the head gasket, and there are even suggestions that the AR24e has a combined head and block for each bank, which HPD declines to deny.
On top of that, continual real-time in-chamber combustion monitoring can be used, even on track given current technology. The use of direct injection and sophisticated engine control can further help keep detonation at bay. HPD notes that the AR24e’s ECU is Formula One specification and uses in-house developed software including drawing on Honda’s Formula One energy management expertise.
Mark Crawford, HPD large project leader for the ARX-06 ,remarked, “Critical to the project was a clean-sheet hybrid powertrain control system, brake-by-wire and vehicle dynamics control system – all written in-house at HPD. This control system architecture was implemented on a Formula One-spec ECU hardware platform. HPD also uses its custom, in-house developed ultra-high speed data logging system.”
The ARX-06 has the smallest displacement of next year’s LMDh contenders
Interestingly, for the AR24e there is the possibility of using pre-chamber combustion, which Ilmor has admitted its 2024 IndyCar engine will exploit. We can assume that HPD’s rival engine will have the same technology.
To recap, pre-chamber combustion means the charge is spark ignited within a secondary chamber, which discharges fast-moving heated jets through a series of nozzles into the primary chamber. That provides multiple ignition sites in the primary chamber, which leads to rapid and highly stable combustion. In the context of the AR24e, this approach is known to significantly mitigate the tendency to detonation.
Interestingly, the LMDh regulations, unlike the IndyCar ones, permit the use of a pneumatic valve return system, although given the 10,000 rpm cap that is unlikely to be an advantage – and is something else to go wrong over 24 hours.
As an aside, the 2.4 litre IndyCar engines will be married to a spec hybrid system in the manner of the new LMDh engines. The IndyCar system will offer more than half as much power again. As we reported in RET 140, the IndyCar system was scheduled to have started its testing on track in June but that has been postponed to the end of the year, apparently as both manufacturers have experienced a need for crankshaft modification when the MGU is in operation.
The driveability factor
When it comes to LMDh engine performance factors, given a mandated power output, arguably top of the list is driveability. Other considerations are fuel consumption, cooling requirement, durability and performance degradation between service intervals.
Driveability inherently suffers as boost levels increase. However, it is worth noting that contemporary management of turbo engines has done much to overcome the inherent response drawback. Moreover, here the MGU can be used to provide instant torque to supplement the IC engine’s output. Incidentally, LMDh permits variable valve timing technology, but with the MGU augmenting the IC engine it is again most probably an unnecessary complication.
When it comes to the question of torque control to match the permitted driveshaft measurement in real time, in theory the more cylinders the finer the control. On the face of it, the rival V8s score in that respect. Again though, the MGU’s assistance arguably overrides this consideration.
Interestingly, Crawford is quoted as saying, “Simulation work was carried out for a novel intercooler packaging and anti-tune induction concept to reach the performance targets while allowing the downsized engine to meet the 500 kW rules target without damaging combustion ‘events’.”
What is meant by ‘anti-tune induction’ is anyone’s guess. RET contributor John Coxon says, “This description smacks to me of some kind of intercooler in the intake system that has a function to alter the performance of the engine, one that can alter the tune of the engine, up or down, to tailor it to the regulation-demanded power curve. Part of that could be achieved using the intercooler, but to tailor it more precisely might need some kind of variable geometry [VG] intake manifold or even some kind of variable Helmholtz resonator.
“I have some experience of using these methods in the past: the resonator on a four-cylinder production engine in the 1980s to smooth out the torque curve, and in the 1990s a VG manifold on a naturally aspirated V8. In my mind, one idea could be to ‘retune’ the engine using a VG manifold and then 'top up' the torque curve to the target as necessary using the MGU. In using the MGU to top up the IC engine’s power it is possible to control the overall output torque much more accurately.
“With the alternative of varying the boost pressure, it takes time to empty or fill the plenum chamber to change the pressure, and to my mind that might introduce control issues. Altering the inlet tract length however would be much quicker and easier to control, and the intercooler could be integral with the VG system.”
Fuel efficiency
When it comes to fuel efficiency, on paper fewer cylinders is better if only thanks to inherently less friction. Alas, given the BoP, that isn’t as significant as it seems to be at face value. Under the BoP, both maximum energy consumed per stint and refuelling time can be adjusted on an individual car basis.
Indeed, in the 24 Hour Prototype race report elsewhere in this issue, Toyota’s Hypercar powertrain development chief Masakiyo Kojima makes the point that fuel efficiency strategies such as the use of pre-chamber combustion are not necessarily worth the complication. Mind you, as noted above, the AR24e might exploit pre-chamber combustion to help counter the threat of detonation.
Then there is consideration of engine size and weight. The 2024 IndyCar engine has to come in at 118 kg, whereas fully dressed the LMDh engine must tip the scales to 180 kg. Clearly, the AR24e will be very heavily ballasted (and by regulation it needs 6 mm additional crankshaft height and the sump is the ideal place for ballast, to keep the centre-of-gravity height low).
Assuming the AR24e has the IndyCar architecture, there can be a 180 mm spacer between it and the transmission, should HPD want to push weight forwards. The LMDh racecar has to weigh in at a minimum of 1030 kg, while the wheelbase is mandated at 3148 mm. Unlike in Formula One, the weight distribution is not regulated into a narrow window. However, the homologation process does lock in whatever figure is selected (within a tolerance of ±0.5%).
Could the AR24e score in terms of car weight distribution potential? Interestingly, the current grandfathered Oreca Rebellion R13 LM P1 car is in effect an upgraded LM P2 car, and it has always been a challenge to get its aero and weight distribution biased far enough forward to exploit the LM P1 tyres provided by Michelin. In effect, LM P1 tyres are the same size front and rear, whereas LM P2 cars were designed around smaller fronts.
The LMDh cars will follow Hypercar in having 18 in rims front and rear. Could a heavily ballasted, forward-shifted AR24e help put more weight onto the front tyres, to the advantage of the Oreca LM P2-based Acura ARX-06 racecar? If so, doubtless its BoP will be adjusted accordingly.
That is the way of things in LMDh. Engine-wise almost anything goes – so long as it doesn’t provide a clear advantage over the opposition. If it does then the BoP will inevitably be tweaked to cancel out that benefit.
In the final analysis, given a specified power curve, HPD is really complicating things in having to use significant boost to reach the targets. While there are workarounds for the disadvantages, boost inevitably creates lag, and even if the MGU can fill in the deficits, that is energy that cannot be used elsewhere.
Cadillac, Porsche and BMW by contrast can pretty easily always be on power target and use energy more optimally. Any ballasting benefit of a small boosted engine will doubtless be balanced out with the BoP. Ultimately the only significant thing BoP won't balance out is reliability, and in that respect arguably a highly stressed, complex system is not what is prescribed. But at the end of the day, there are some hugely talented engineers at HPD who will make the best of what they have been given to work with.
It is such a shame though that HPD engineering talent isn’t going into making a machine that can be more competitive than the opposition, rather than the BoP having everyone averaged out. Some folks say the new Hypercar/GTP era will rival the majesty of Group C, but in the Group C days engineers were allowed to make a better mousetrap.