Open top menu
14 Sept 2015

Unless you've been hiding under a rock, or arrived at this juncture in the DeLorean that McLaren teased us with at the beginning of the season, the McLaren and Honda partnership has thus far been rather disappointing.  Earlier in the season I looked at the RA615H's architecture:, whilst this piece I did earlier in the season serves as a reference to the architecture of the other powerunits:

Whilst writing this article I stumbled across a couple of videos on Youtube from a Japanese source, as such the audio is of no use to me (although, if you can translate the videos to English, please get in touch as certain aspects of the video would be quite interesting).  However, I can use some of the images for reference and as such thought I'd go through some of the stills as an expansion of the article.  However, please bear in mind the images are simply used as a reference source for me to help explain certain areas, I'm not suggesting everything they depict is wholly accurate (this is the covert world of F1 afterall).

Honda, like the rest of the powerunit manufacturers were given the chance to upgrade their units throughout 2015, utilising the tokens they hadn't spent before the season started, or in the case of Honda an average of the other suppliers, given their baseline began in 2015.  As such Honda had 9 tokens at their disposal, allowing them to add performance.  Furthermore, parts that retain the same design can be ungraded without a token spend if being done in order to improve reliability and/or reduce costs (performance may be achieved as a by-product).

So, what have they done this season?

Of the 9 tokens that Honda could deploy during 2015 they have thus far spent 5, two were used in an update for the Canadian GP, whilst a further three were used in Belgium.  The five tokens spent so far appear to have centred around improving combustion, bringing the I.C.E closer to the performance of Mercedes, Ferrari and Renault.  Compartmentalising performance in that way can be counter-productive though, with the Turbocharger intrinsically linked to ICE performance, which in turn requires the MGU-H to be operating at the optimum level.  In terms of peak power / bhp it would be easy to draw comparisons between the different powerunits, the problem is that the RA615H doesn't provide the power consistently.  A shortfall in the Energy Recovery System (ERS) means that they don't have enough energy to deploy around the course of a lap.

So, where's the shortfall?

In these screen grabs from the Japanese video (1) we can see that the intentions of Honda are clear, package the Turbine, MGU-H and Compressor within the V of the I.C.E.
They go on to show a comparison of the RA615H against the PU106A, highlighting the difference in size and arcitecture between the two.
Energy should be being both recovered and deployed almost all the time, in order that the maximum amount of power be dispensed, something that can be difficult to achieve if you have the wrong architecture.  As such the Honda turbo is the polar opposite in size when compared to the leading one of Mercedes.  Split on either end of the ICE, the Mercedes turbo and compressor are probably twice the physical size of their Honda counterparts.  

This was seen as a coup by Honda who firmly believed the turbo sizing would be perfect, as it would spool up quickly providing the requisite boost to the ICE and meant the footprint of the powerunit was reduced as the Turbo sat between the ICE's vee.  The problem with this is that the MGU-H isn't doing any meaningful recovery, be it lower in the rev range or higher up. The MGU-H acts like a wastegate, only instead of reducing pressure it slows the shaft whilst recovering energy.  In the case of the larger Turbo used by Mercedes they have to operate in a bit a of a bubble, but in doing so can recover much more energy.

So why's that important, can't they just use the energy from the battery (ES)?

Yes, they could use energy stored in the ES, however, the energy matrix only allows storage of 2mj's of energy from the MGU-K whilst the MGU-K can spend 4mj's per lap, making for a 2mj deficit, not even accounting for the MGU-H's requirements.

ES limitations aside energy transfer between the MGU-H and K (in either direction) via the controller is unlimited, meaning recovery of energy via the MGU-H is imperative.  Taking this route also means there is a negligible power loss too, as the AC/DC transfer to the batteries (ES) is less efficient, wasting energy and creating heat as a by-product.  Sound complicated? Well it can get even more complicated when you know that the energy flows can all be working at the same time.

For example: The MGU-H is recovering energy as the Turbo is producing too much power, it is sending some of this directly to the MGU-K (via the controller) without conversion losses, meanwhile the MGU-K is also receiving power from the ES (with conversion losses) in order that it can deploy the full 120kw (approx 160bhp).

But if you can only deploy 4mj's of energy, at 120kw isn't it all spent after 33.33 seconds?

No, this is one of those misnomers that cropped up early in 2014 with the broadcasters referring to the amount of energy at the drivers disposal.  By-and-large it seems to have been glossed over now, however for most that appears to be the narrative they understand, as the broadcasters haven't rectified it, instead ignoring their mistake and hoping no-one noticed.

It isn't 33.33 seconds for several reasons, one of which I've explained above, with the energy transfer between the MGU-H & K being unlimited.  The other reason is that unlike KERS, that was deployed with a button press and so can therefore be given a linear time reference (again only based on the maximum 60kw-80bhp being deployed) the MGU-K is deployed according to the current energy map being used and the throttles position relative to the revs.

As you can see this makes energy management an absolutely crucial part of the entire energy deployment scheme.  McLaren have this season often suffered at the behest of high fuel consumption and that is due to the extra demand put on the petrochemical resources when the ES is depleted and the MGU-H cannot supply the requisite energy to the MGU-K.  Mercedes are streets ahead of the competition in this respect, with an ability to cope with the demands of any given situation.

So, did someone get their sums wrong? The turbo seems woefully undersized and as such A) the MGU-H isn't needed to keep it spooled and B) cannot harvest enough energy from it to help fill the ES and/or pass energy to the MGU-K.


I have said repeatedly since the start of the new regulations that fuel is one of the key performance differentiators.  This was something that both Ferrari/Shell and Renault/Total were seemingly slow to realise but did make progress with throughout 2014 and seem to have narrowed the gap further in 2015.  Remember the current generation powerunits have less cubic capacity, have to be a third more fuel efficient and without the ERS produce 2/3rd's the power.  This means they have to be supremely more thermally efficient than their predecessors and lubricants play a significant role in this especially fuel due to the interaction of the petrochemical and electrical systems.  This puts McLaren and Honda at a disadvantage, something they must have realised during 2014 as they struggled against Mercedes with the same powerunit, that's because their lubricants are supplied by Mobil1.  Now I'm not laying any blame at their door, I'm simply pointing out they're another chink in some seriously damaged armor, especially given they only supply McLaren, limiting further the data that is being supplied to the lubricant partner.  Have they worked with Honda as tirelessly as Shell have with Ferrari or Total have with Renault? I don't have the answer to that, all I can say is they would need to throw some serious weight behind the challenge.

I've not even mentioned Petronas yet and with good reason, their partnership with Mercedes has quite literally fuelled a revolution, providing a clear advantage that took years to perfect and with whom Mercedes worked very closely with during the initial phases of the new regulations.  The bad news, not only for McLaren is that gap is about to widen further, as Mercedes development plan initially meant they'd happily dispose of 7 tokens for 2015.  With the homologation regulations all but broken for 2015, thanks Ferrari and Red Bull, in-season powerunit development has ensued and those 7 tokens that would have effectively been locked out to Mercedes have just been spent.

Mercedes HPP spent their remaining 7 tokens in a powerunit upgrade that only the 'works' team ran at Monza, albeit only Lewis Hamilton raced owing to a coolant leak on Nico Rosberg's car.  The 7 tokens were used inline with a development curve that Mercedes have followed with Petronas and exploit an even more lethal combustion of their fuel, providing yet more performance from the ICE, which in-turn will increase the potential electrical energy yield.

Let alone 1988 I bet Honda wishes it was 2008...

Regaling in their past triumphs McLaren used some sickly advertising tools to welcome back Honda, a love-in that has lasted about as long as an Elizabeth Taylor relationship.  Whilst they continue to try and show a unified front the pressure is building, afterall McLaren is a stalwart, a team that most could never foresee being in such a predicament.  Lest we forget though, that Honda's last F1 engagement went out with a whimper at the end of 2008, when the sport was on the precipice of huge changes.  The F1 they knew and left behind is distinctly different to the one we have today and could be another problem for the Japanese marque.

Honda's last foray in F1 was at a time when excess was at its peak, wind tunnel and track testing were almost unlimited and as such data could be validated with ease.  Did Honda overlook these facts when committing to the new programme?  Dyno's are a wonderful tool and Honda have many, both at their headquarters in Sakura and at the Mugen facility in Milton Keynes, however, the results they portray can often be skewed by many conditions (environmental factors, driving style etc).

These two images from the video (1) show an iteration of the RA615H in one of the dyno rooms.
In these three images from the Mugen facility in Milton Keynes (video 2) we can see that the dyno room features so additional equipment in order to simulate how power is produced at the wheels but doesn't go as far as replicating other chassis conditions.

Graphical representation of the simulation that is being run on the dyno
We know that the other manufacturers either started their research with (Mercedes) or now have access to (Ferrari and Renault) full chassis dynomometers, which are able to more closely represent the conditions a powerunit undergoes (See AVL's setup below).

Images copyright AVL who provide this type of equipment to leading motorsport teams.  Marrying the powerunit to a chassis which is also mounted on a 7 post rig adds many more layers to the test environment, increasing the relevant data that can be garnered under simulated conditions.
Honda clearly have access to a chassis dynomometer, shown here with what looks like a Super Formula car mounted on it. It appears to use a rolling road configuration rather than the hub dyno's used in the AVL setup.
This isn't to say that Honda haven't got access to a similar setup to the one provided by AVL, I just don't have the proof they do.

For those that aren't aware of how the teams/engine suppliers work during a race weekend we can see in the following images the operations room in Sakura.

Engineers in Sakura study the data transmitted from both cars as they aim to improve performance and/or identify any problems during the race weekend.  A similar room also forms part of McLaren's facilities in Woking.
At the European races McLaren have additional engineers onsite to study the data, utilising the 'motorhome'.
A sample of some of the data screens the engineers work from, in images 1&2 we can see vital data regarding the powerunit, such as MGU-H speed, turbo pressure, temperatures etc.  These figures change to a red status if they go beyond their intended thresholds.  In the image above the teams use this dial style display in order that they can see gaps between each car, using GPS data this visualisation is key in deciding strategy, ideally keeping the driver in clear air.
McLaren aren't helping matters

Pointing fingers at Honda is easy, it is clear to see they've made errors in their powerunit design but what of McLaren?  Ron Dennis has repeatedly alluded to the fact that they made the right decision in switching from Mercedes powerunits and seeking an alliance with Honda and he's not wrong.  Red Bull will find this out to their peril when their relationship with Renault finally comes to an end and they're stuck with Ferrari PU's.  The 'works' teams will always get a slight edge over the customer teams until such point the FIA do something about it.  This has been exacerbated this season by the destruction of the homologation procedure, meaning different specifications of powerunits are available, with the 'works' teams getting first stab at using them.  This aside, ECU lockouts protect certain parameters being accessed by the supplied teams, requiring a request to the supplier for its use (You might recall such an instance with Williams in the past).  These lockouts are supposed to be about protection of the powerunit, increasing their lifespan, however, intrinsically that also holds back performance.

On the chassis side of things McLaren have been eager to portray the MP4-30 as the best thing since slice bread but it too has its problems.  Far better than the MP4-29 the '30' is an improvement but still not the 2nd best chassis on the grid I have seen claimed.   The 'size zero' rear end is impressive but comes with its own complications that won't help the Honda PU.  As I have already discussed thermal efficiency is an important part of the performance envelope, meaning cooling all of the components is critical.  McLaren have opted to use an intercooler to cool the charge air, which we find mounted in the right hand sidepod, whilst the ICE radiator resides in the left.  Considering the other 'works' teams of both Mercedes and Ferrari opted to use chargecooler arrangements (air-liquid-air) it seems strange that McLaren-Honda have opted to tread this path, the advantages of the chargecooler system seem apparent to me in terms of architecture and thermal efficiency but like McLaren many of the other teams still use standard Intercooler setups.  The space within the sidepod is therefore at a premium, given both the need to house both large coolers and the demands of a shrunken coke bottle region.  This led to the team placing coolers in the saddle position above the ICE, something that has been rectified with the Belgian GP upgrade.

The older 'saddle' position for the ERS/Turbo oil coolers is highlighted in yellow on the upper image, whilst the new orientation used since Belgium is highlighted in Green in the lower one.  You'll note that in the new configuration two coolers are present, however, this was the case in the saddle position, it just wasn't as obvious.  It's an interesting change that coincided with a 3 token PU spend, 1 of which goes toward these changes and the other two are thought to be focused on the ICE's combustion elements.
The changes made to the oil coolers have also led to some bodywork changes, as we can see in the upper image the cooling outlet (highlighted in green) has been reduced significantly when compared to the Belgian GP spec in the lower image.  Now it's not to say the team wouldn't have reduced the size of the outlet for the demands of Spa and Monza in any case, but to me it is clear the reduction is at least in part due to the cooler changes.  It will also come with some aerodynamic improvements as a by-product, with the exhaust plumes direction and energy less effected by the slow energy dispensed from the previous cooling outlet.  After all the teams are using the exhaust plume to assist the airflow structures generated by the diffuser and rear wing.

Rumours have surfaced over the last few days in regards to Honda using a Super Formula car to validate their concerns that the McLaren chassis is as much to blame for their poor performance as the powerunit itself.  I scoffed at the thought of this initially but having looked at it laterally it's not actually that far fetched.  If we take a look at the Sporting Regulations:

22.1 Testing of Current Cars (TCC) shall be defined as any track running time, not part of an Event, in which a competitor entered in the Championship participates (or in which a third party participates on behalf of a competitor), using cars which were designed and built in order to comply with the 2014, 2015 or 2016 Formula One Technical Regulations. No competitor may sell or make available a car of the current year to any third party without the full knowledge of the FIA.

The problem with the wording of this regulation is that it opens somewhat of a grey area up for the engine manufacturers, as it only infers cars designed to comply with the 14,15 or 16 F1 technical regulations.  A Super Formula car is not designed to comply, however, using it as a mule for development does mean making it representative of a current car in order that the powerunit and ancillaries fit.  A very grey area and I'd suggest something that the manufacturers have an agreement with the FIA not to do, considering they aren't directly responsible for following the sporting regulations.  IF this kind of test were to take place I can see the regulations being swiftly amended for 2016 as Mercedes, Ferrari and Renault will swiftly follow suit with their own 'mules'.

Although I think the idea of actual track testing using a Super Formula 'mule' is a little outrageous perhaps what isn't is the same application of the 'mule' only this time mounting it on the full chassis dynomometer.  It would certainly give Honda some real world validation data to show McLaren how and where they believe the MP4-30 is weak.

Where do Honda go from here?

2015 is a write off, frankly if it were not for the data they are gathering at each GP you'd almost forgive them for not racing at the remainder of the races.  It appears to me that Honda already clearly understand where their shortfall is coming from and have spent their 5 tokens thus far on preparatory areas for the 2016 powerunit, making changes to the I.C.E that will prove more fruitful once they are free to spend more tokens in 2016 and/or items that will be locked out of development for 2016.  Honda still have 4 tokens at their disposal and will likely homologate a powerunit using those tokens before the season draws to a close, afterall they can't be carried over to 2016.

The Turbo and MGU-H are likely facing a complete redesign, which would eat 12 tokens at worse case, giving plenty of scope for other modifications around the powerunit. Some items would have to be changed to accommodate their fitment, such as the plenum but that is a sacrifice worth paying.  There is also a question mark surrounding the MGU-K and whether it was sized correctly for the demands, with recent images showing it is roughly twice as small as the Mercedes unit. 
Image courtesy Ferran Figuerola - The MGU-K is highlighted here in green, mounted under the left cylinder bank in a position shared by the other engine manufacturers
Images: Left courtesy Ferran Figuerola whilst right is from Auto Motor und Sport more specifically Tobias Gruner

As we can see in the comparison above the MGU-K Honda is using is roughly 1/2 the size of the Mercedes unit.  Size is not an indicator of performance, Honda may well have found a way to miniaturise their unit compared to Mercedes.  However, with the mistakes made on the Turbo and MGU-H it is plausible that the 'K' is able to adequately supply the 120kw assistance just not over a sustained period.  The other take away from these side by side images is heat control, Mercedes have clearly made efforts to shield the MGU-K from the exhaust that lies directly above, whilst Honda's log style manifold although shielded is in much closer proximity.
The Mercedes PU106A (2014 powerunit) you'll note the shielding placed around the exhaust manifold protecting the surrounding area from heat soak.
Honda engineers examining their MGU-K (albeit blurred out for the video)

The engine manufacturers are permitted to 'spend' 25 tokens for 2016, although it is unclear at this stage if we will continue to see the abuse of the homologation period for 2016 with in-season development allowed once more.  Of these 25 tokens the ERS if it were to be redesigned in its entirety (which is highly unlikely) would account for 24 tokens, leaving only 1 token for the development of other areas.

One of the problems for Honda is the fact that although they have 25 tokens to spend, so too do their rivals.  McLaren fans should take some solace in what Ferrari have been able to achieve, as they too made mistakes in the design of their Turbo, MGU-H and exhaust layout for 2014 that they rectified, marginalizing the gap to Mercedes.  Their fight isn't over yet though as the gap widened once more with Mercedes last token spend showing the continued development of the powerunits is crucial.

However, 25 tokens may not be enough for Honda with none of the manufacturers yet at the point of diminishing returns and so it may be prudent for them to lobby the other manufacturers and FIA for some concessions.  I think this goes beyond what is equitable and moreover what is in the best interest of the sport as one of the great teams stares down the barrel of a further fall from grace.  The concessions would be dependant on how the homologation process works for next season and if an agreement is found to continue in-season development.

If the manufacturers, FIA and WMSC decide to defer the homologation period for all manufacturers I'd suggest that Honda be allowed to follow the 2015 matrix with 32 tokens from 61 modifiable items, whilst Mercedes, Ferrari and Renault have 25 from 51 modifiable items.  However, if the homologation date is to be retained I'd suggest that Honda still be allowed the 32 tokens from 61 items, whilst be allowed in-season development.  I think the largest issue facing a protracted use of the matrix is Renault, should their 12 token spend in Austin prove fruitless they too may be looking for concessions, which in-turn muddies the waters for Honda.

Whatever happens in terms of their token spend, the homologation period or any concessions that may be given make no mistake Honda face an uphill battle, they're some way off the mark and catching the likes of Mercedes and Ferrari may be a bridge too far for 2016.


  1. You allude to the fact the chassis may not be as good as they say - what do you base this on ? . Alonso says of the 2.8 seconds missing at Monza - 2.4 is accounted for in the PU while .4 is in teh chassis ? You have done a good job on the PU side - please follow up on teh chassis side too .

    1. Poor McLaren fans, their whole defense for the entire season has been that their chassis is "ok". How sad

    2. lol exactly Anon. I think only McLaren fans are expecting anything impressive at Singapore

    3. Its difficult to see (visually) where McLaren is with its suspension. As for the aero, one can quickly conclude that the team is following an archaic philosophy. You can see it most clearly on the FW, but also on the way it tries to condition the air towards the back of the car, and on the diffuser.

      In my personal opinion, the lead on philosophy started to shift from Redbull to Mercedes during the 2013 season. In Mercedes started to adopt a unique approach in 2012/2013. I wasn't entirely convinced by it then, but now we can clearly see that that is the philosophy to follow.
      Redbull and Mercedes have completely different approaches to VG's (especially on the FW), however the results don't seem to be too different. They are different approaches, but they do seem to work efficiently, each in its own way.
      McLaren seems to lag behind. It seems that there is a lack of understanding in regards to aero. One can speculate in many different ways, but poor correlation between track and CFD/windtunnel could be part of the problem.

  2. i think merc, RBR and STR and ferrari are still way ahead of mp4-30 chassis

  3. The real question is, why would anyone cut slack at Honda, while noone , for example, cut slack on Ferrari this year?

  4. I cannot wait for this stupid Power Unit age to end, and it appears that I'm not alone.

  5. Great stuff Matt!

    I've published a piece (and portuguese and not so detailed) 2 weeks ago and got great feedback.

    I think Honda will overcame this all and fight back next year.

    And for McLaren, Peter Prodromou is just starting a new philosophy, we need to be patient.

  6. Given Honda's performance to date McLaren must be worried. McLaren knows how good or bad their chassis is: most teams do: a 20KPH deficit at Monza is either a testament to having a missing 120KW or a very draggy body. Peter Prod is nobody's fool whatever readers and fans think.

    I suspect the Size Zero concept is at least partially responsible for a number of the issues; particularly the H and cooling.

    The key issue is the "H" as you say Matt, and this is why McLaren are worried. Boulier is on record saying that McLaren has seen the Honda solution: next year! It is this very slow development rate that is concerning, and if Honda makes another mistake it may be the end of a once great British team.

  7. Really the car is totally crafted so patiently which gives the full comfort and ease. Honda Jazz is available in three variants and all are powered by the.2L, 1198cc, 4-cylinder, i-VTEC petrol engine with 5-speed manual gearbox. automoves


Total Pageviews