Why are Hondas engine so bad? Explained.

f1_engine_honda.jpgSo to explain the complexities Honda are facing in getting things right, first we must all get get a basic understanding what the things are.  This article is in no way complete and (using F1 language,) “for sure” is a very simplified scenario of why Honda are struggling to produce power, any kind of fuel economy, and not least, reliability.

What we do know is that the Honda PU is heavy on fuel, down on power, and has a tendency to commit suicide (you would wouldn’t you).

Firstly we should point out that the latest incarnation of a Honda power unit features an internal combustion system unlike most systems you would relate to.  Oddly it bears more resemblance to an archaic diesel engine than perhaps the latest highly efficient petrol offering in the showroom.  So please take what you know about ignition systems and put it to one side.

So the question.. Why is the Honda engine so bad?

Well lets again start with perhaps why it isn’t, we also know that the Honda engine in the paddock is classed as a GOOD engine (that’s right) and when it is fully developed it is likely to be a match for Merc and Ferrari.  Sauber might not be so daft..

We also know that new combustion technologies have been used by other teams that usual refer to it as a Thermally Efficient engine.  It is a tricky system to get right, ask Renault.  But when its right, its right.  Ask Mercedes.

I will assume a general understanding of how an Internal Combustion Engine (ICE) works so I am not going to include that in the comparison.

It is unclear as to exactly which new technology Honda are using, there have been reports across the internet in the past but TJ13 have some reservations about the reliability of the original sources.  It could be one of two types.  HCCI or TJI.

The easiest way I can explain the difference between Homogenous Charge Compression Injection (HCCI) and Turbulent Jet Injection (TJI) technology is to understand the fundamental difference between the two and then use a timetable of events.

The fundamental differences are depicted as follows :-

Ignition No Artificial Ignition Source Artificial Ignition Source (i.e. Spark Plug)
Fuel Mixing Location Outside Cylinder Bore Within Cylinder Bore
Ignition Control Compression Temperature Electronic Spark

HCCI Comparison – no spark plug, 


The basics of TJI are to pre-ignite an air/fuel mix to cause a pressurised flame that is forced into the combustion chamber through tiny holes in the pre ignition chamber at precisely the right time to produce the explosion.


This is far more efficient than say a single spark plug which will ignite in a much  less uniform manner out from the electrode gap. The “spark” from a spark plug will have what is known as a flame front at the outer edge of the fire which is what causes heat and emissions. The best way to visualise how a spark ignites is to drop a marble into a flat sided tray of water and watch the first ripple – that exactly how a flame travels from a spark plug. The resulting explosion of the fuel/air mix can be seen as an arc cone down from the electrode into the bore igniting the fuel/air as it goes.

In the TJI, there are multiple, highly pressurised and controlled flames that are injected through multiple nozzles in the housing into the bore like a ray gun, causing what I can only describe as a chaos of exploding fuel/air throughout the bore. In other words, one big bang in one go igniting the fuel/air throughout the bore at the same time.

So, TJI can produce the same (and more) exploding force as a spark ignited engine, using less fuel as it all ignites together. The weaker the mix in the chamber the more power. Seems marginal in real time but this is the difference that makes one engine slightly better than the other.


The idea behind HCCI is that there is no artificially produced ignition source. No spark plug.  The fuel/air mix is ignited in the same way as diesel, i.e. compression. However, unlike a diesel engine where the air is compressed then injected with fuel, the fuel and air are pre mixed prior to entering from multiple angles into the combustion chamber.

This means that when the resultant explosion occurs, its throughout the bore in one go (as per TJI) and not in an arc out from the injector (like a spark ignited engine)

This again will produce the same exploding force, but using less fuel.

Timetable of Events

To make it easier to understand, this is roughly what happens and when

1 Piston begins its downward stroke Piston begins its downward stroke
2 Inlet Vale Opens Inlet Vale Opens
3 Pre Mixed Fuel Air Enters Air Enters
4 Piston Begins Its Upward Stroke Piston Begins Its Upward Stroke
5 Fuel Is Injected In
6 Fuel is drawn to the pre combustion chamber
7 Correct compression is reached
8 Spark ignited the fuel in the pre combustion chamber
9 Correct compression is reached Flame is expelled back into the bore
10 Fuel/Air In Bore Explodes Fuel/Air In Bore Explodes
11 Resultant Energy Forces Piston Down Resultant Energy Forces Piston Down


In a TJI engine, the mapping is finely tuned to depict the amount of fuel and air that enters the bore and then precisely when spark ignites. In theory pretty standard stuff, similar to a conventional engine but hard to get things just right

With HCCI, things are even harder to control. As the sole form of igniting the fuel is from the compression (which generates heat). The way to control that is more complex. If the compression is too high too soon, the mixture will ignite too early, not high soon enough and the mixture will ignite too late to make it efficient.

The obvious way to control it is through variable timing of the valve, but that would probably mean it would be less efficient…and that would mean that less of the allocated fuel could be used purposefully and hence reduce the power of the engine.


We have to conceded to a minute understanding in comparison to dedicated engineers, and despite years in engine mapping, these guys have forgotten more than we will ever know about these engines, the problem Honda are having is in controlling the compression within the bore in order to obtain efficiency. The optimum band in which the detonation of the HCCI engine is at peak is minute and getting it to occur at that time is proving difficult.

Engine dyno testing is one thing, putting it into a car where many other variables are added is another. To give but one example; air pressure that can change enough in one lap let alone a race. The mapping then becomes so complex that it would almost have to re-write itself constantly to maintain itself.

Ineffective combustion = miss fire, poor power and economy through lack of efficiency

Detonation at the wrong time is a big problem.  If the piston is at the wrong part of the stroke cycle you get a  “knock” (an audible miss fire), and low power whilst it all gets adjusted.  The unburnt fuel is effectively passed through the exhaust valves and out the exhaust.  In short if the engine does not give up (and it probably will), it will be slow and terrible on fuel. Pistons are not meant to push when they should be sucking.

Running high boost adds to the issue (and I think the Honda does this, a lot).  Hot countries or altitudes will effect the issue.  This makes air intake temperature high.  This means the cylinder fills with poor quality and less dense air, so the fuel that gets mixed to match the ratio (road cars are generally 15:1) is also going to create low power, so running in dirty air adds to the issue.  These are issues that cannot be replicated on a dyno OR rolling road.  In reality it needs real life experience and tiny adjustments, it needs track time and Honda have had the least.

Fuel Mix

An experienced F1 engineer told TJ13:

The mapping includes software tables that define certain fuelling conditions.  E.G when airflow is X, fuel is Y and and ignition is Z.  These engines need to be so fuel efficient that it is difficult to keep out of the detonation zone inside the cylinders and if you get much of that, its game over.

This can be especially difficult at certain rev zones and can cause big torsional oscillations through the power train…and I dunno – kill gearboxes.

So why is Honda struggling?

Well time is an issue, its had the least time to get this right and they have had no experience yet.  There were rumours that Mercedes were going to assist Honda in certain aspects of development, my money is on the mapping.  Mercedes have now backed that truck up.

It has been suggested that Honda should strap the PU in a GT car and drive it round its own F1 track, but again – the mapping is such a fine art that unless the airflow is identical to the F1 car, and it’s in traffic, there will probably be little to gain.

Then there is another idea, and it is just a theory BUT..

The fuel has many additives, one of which is OIL (see previous stories on extra tanks and bans for next year).  Engines, tolerances and especially mapping are designed around fuel, oil and other additives that keep temperatures inside the cylinders down and results in a more reliable and predictable combustion.  Understanding fuel, additives and they tiny details on how they effect combustion is ESSENTIAL to any mapping engineer.

Honda had developed all of its previous power units to specifically use Exxon-Mobil fuel but ahead of the 2017 season a very late switch to BP-Castrol fuel was needed after McLaren signed a new deal with the the British supplier.  This would have been a massive blow.  Engine tokens are ditched this year but this change needs real miles. Track time is non existent, and dynos are almost useless past producing a base map.

So I am going to say it… Have McLaren created a large chunk of the problems through a last minute trade deal?

Baku saw a new generation engine used, and SKY reported that they had bought it along basically as a dyno test.  They retired and returned to the old engine because they feared the gearbox failure (power train oscillations in specific rev ranges anyone?) may have damaged the engine. It may have but I don’t think they ever intended on using it for the race.  Just some track time.


18 responses to “Why are Hondas engine so bad? Explained.

  1. Truly interesting writing. However some questions from a layman. How is HCCI being ‘controlled’ in terms of ignition timing? The only parameter seems to be fuel mix since the valve timings are fixed (and variable is verboten). How is HCCI being kept consistent across different temperatures? Indeed you say this is the trick, I’m just wondering what exactly it is that allows HCCI to work.

  2. Staying out of F1 for that extra year allowed the Honda engineers to take far too much notice of what the others were doing and confusing themselves with too much information. That stopped them from utilising a season’s worth of precious track time to R&D their own basic designs.
    It’s one thing to stand back for 12 months in a bid to watch and understand where Hybrid F1 PUs were going, but it’s an entirely different thing to try to copy concepts which are so incredibly complex without actually running it yourself and learning from scratch.
    Most of the time it’s the intricate troubleshooting of the initial gremlins that teach us the most when we start a new project. The rest is just fine tuning and honing things to a perfect cutting edge.
    The Chinese copy technology all the time but seldom succeed in making the replicated item work as well as, let alone better than, the item they are copying. Simply because they try to do it cheaper and quicker than the original.

  3. Great effort Steve!
    As for the road test being the answer to find a solution to their problems, surely there must be a way to come close to a f1 simulation, without breaking the rules.
    Even if it means that they have to heavily modify a super formula or an indy car. I know this demands a lot of hours and a lot of preparation. But waiting till the engine gets better by itself surely isn’t the way to do it.

  4. Thank you, I found this an incredibly interesting article and very well written (I understand thermodynamics pretty well, but cutting edge engine technology’s a proper egghead’s eggheads field 🙂 ).

    I was thinking to ask if you can explain to me what exactly is mapped to what when engineers talk about engine mapping (something I hadn’t found a satisfying answer to) but voila, you already did – thanks!

    Taking note of the fact that it’s of course to some extent purely speculative, your line of reasoning makes a lot of sense to me viz. Honda – it explains a lot of the uncomfortable location between rock and hard place that they currently find themselves in. I hope they find some kind of breakthrough forward, bc this really is fun for no one.

    Thank again, I really enjoyed reading this article! Cheers from Berlin.

  5. thanks for that Steve, a great explanation of the concepts for a non-technical armchair racer like me 🙂

    • Even disregarding the weight of tanks to carry the oxygen this will not give any more efficiency. The Nitrogen portion the air is essential to expand with heat and drive the piston. The Oxygen/Fuel component of the cylinder charge actually decreases in volume as it is converted to heat. It is important to remember that although the text says explosion there is no explosion in an engine, this would destroy it – just a rapid burn with heat and expansion of the charge

  6. Interesting matter ! Thanks !

    The technology surrounding HCCI this still is very secretive. Not only because of F1 competition, but also because the advantages it can (and will) bring to production cars.
    Controlling HCCI ignition under different loads and different rev’s seems to be quite difficult. I think I read somewhere that HCCI is therefore only used in some rev. or load ranges.

    HCCI has been tested by some (most of) the big car manufacturers for many years now. GM were running HCCI prototypes a couple of years ago. Mazda is on the brink of introducing an HCCI engine in their 2018 Mazda3. What we know about that engine is that it will have a 18:1 compression ratio, which is 4 points up from Mazda’s current SkyActiv engines (14:1).

    What we also know is that Mahle, who offer a TJI system, claim that the raised efficiency through their TJI system, allows for an increase of the compression ratio by up to 4 point (it’s on their website).
    So, could Mazda be using HCCI at certain load levels and TJI at other load levels? And, could that also be the case with the current F1 manufacturers?

  7. I have heard rumours of similar claims, and the farting backfire noise where the ignitions switch over, but honestly, I don’t know. That would be a flipping mapping nightmare.

  8. Matt Somer’s 2015 article was very interesting with lots of photos. Which still suggests Honda isn’t short of dyno simulation equipment.

    I’m still guessing the problem was not knowing where the performance gaps were. Perhaps Gilles Simon was telling Honda the problems were in the ERS, when the real problem was the ICE efficiency.

    HCCI sounds interesting. Presume that heat is used to create the combustion, which could help efficiency because charge cooling might not be needed much if at all.

    If Honda did have HCCI which nobody else has, then it would pay not to let on the secret. If it is true that Mercedes had TJI from Mahle, it was not widely publicised. The public smokescreen was that the split MGU-H/MGU-K was the secret to their success rather than the huge efficiency gain of TJI.

  9. All very nice Steve, but seems to have a major flaw. In your introduction you state the HCCI fuel mixing is “outside the cylinder bore” (first table). You further state, in your explanatory table 2, step 3, that for the HCCI “Pre Mixed Fuel Enters” and then correctly note in step 5 that for the TJI ignition “Fuel Is Injected”. What you are saying is that the TJI ignition injects fuel into the cylinder but in the HCCI engine the fuel is mixed prior to entering the cylinder.

    FIA Technical Specifications, Article 5: Power Unit states in section 5.10.2 (Fuel Systems), and I quote:

    5.10.2 There may only be one direct injector per cylinder and no injectors are permitted upstream of the intake valves or downstream of the exhaust valves. Only approved parts may be used and the list of parts approved by the FIA, and the approval procedure, may be found in the Appendix to the Technical Regulations.

    5.10.2 clearly states that “no injectors are permitted upstream of the intake valves….” I think your article needs a serious rethink. An earlier comment was also right to the point: if the HCCI ignition source is simply compression and fuel is, according to you, mixed prior to entering the cylinder how is ignition timing controlled at all?

  10. After some thought I can see that timing might be controlled by the mixture ratio (rich ignites before lean maybe; seems like a nightmare at high revs though) but my primary comment still stands: FIA Technical Regulations clearly prohibit introduction of fuel anywhere but by direct injection into the cylinder by approved (single) injectors.

  11. If you use HCCI can you keep a normal spark plug to ignite just before the point where you want the compression ignition to happen just to help it along? I think Mercedes were testing road car HCCI a few years back and they used compression ignition only for part of the time according to magazine reports at the time.

  12. I totally agree with the fuel supplier change affecting development however Honda have been working on HCCi engines for over 10 years and have there own HLSI variant. Outside help was probably refused as the engine is possibly far more advanced than even the illmor/Merc engine. McLaren might as well stay with them now .

    • Seriously? You think the Honda engine is more advanced than the Merc or the Ferrari? Just because a manufacturer has been studying a concept doesn’t mean they can make it work. Look at the Wankel, just about every car manufacturer got the patent rights to try it, heck Chevy even had a rear engine 4-rotor ‘Corvette’ prototype that probably would have been badass, except that only Mazda figured out how to make the engines last.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.