51 responses to “#F1 Forensics: ERS and Software – A new Pandora’s Box brought to you by KwikFit Insurance”

1. Thanks! Nice to see this all so clearly now.

   • Glad you like it mate 😉

2. I thought there was a minimum weight of 20 kg for the battery pack? And a maximum of 25 kg? Of have Mercedes found a way around this?

   • Myrdinn 20/25Kg for a 4MJ battery pack, according to rumors Mercedes could use an only 2MJ battery pack..maybe the trick lies here..

     • Hmm, I see your point, but in the technical regulations I haven’t been able to find any reference to the size of the ES. The only limitations I found are on energy flow from the MGU-K into the ES, which is limited to 2 MJ per lap, and the energy flow from the ES into the MGU-K, which is 4 MJ per lap.

       The weight restrictions for the ES don’t mention anything about storage capacity and the minimum weight being a function thereof. Of course it could still very well be that Mercedes have found that they have no need for an energy buffer (because as I understand it that’s basically what the ES is) larger than 2 MJ, but that wouldn’t give them a weight advantage in my view.

       It would be very interesting to know whether Mercedes has found a clever solution for their energy flow management that would allow them to run with smaller capacity and therefore lighter batteries, as this would maybe allow thicker (and thus heavier connectors etc) which have less resistance, increasing energy transfer efficiency and reducing the cooling requirements of the ES.

     • Oh, I just noticed that I overlooked the requirement that the difference between maximum and minimum state of charge of the ES cannot be more than 4 MJ at any time the car is on track. This would of course make it pointless to have an ES with a capacity larger than 4 MJ.

   • You are absolutely correct; I thought the same thing when I read the article. From the technical regulations:

     5.4.3 The total weight of the part of the ES that stores energy, i.e. the cells (including any clamping plates) and electrical connections between cells, must be no less than 20kg and must not exceed 25kg.

     This sort of puts a hole in the Mercedes weight advantage argument; sorry Lorenzo, maybe time for a rethink. Good article otherwise; please keep them coming.

     • Maybe the ES itself could not be “lighter”, but the weight saved by the less cooling requirements should be evident 🙂

       Obviously the rumors needs to verified first.

       • Interesting theory, but it would only really help if their car was overweight to start with as the only other positive is that they may have slightly more blast to put as low as possible and in optimum positons to help balance as the cars and driver still have to conform to the minimum overall weight limit

         • Also, front/rear weight balance is by regulation, so there is no chance to improve handling by biasing the front or rear. I believe the c.g. of the battery pack is also defined (but haven’t had time to look it up, so could be wrong).

         • I believe there can be a 3% difference in the weight distribution from to rear, which equates to 20.73kg. Hence why I feel this is a case of misinformation, as I have said before if the ERS boost is worth 2-3 sec a lap then there is no way an engineer in hos right mind will give up around half of that just to get 12kg of the 20kg spare needed for ballast. (as no-one given the choice would run over weight) I don’t buy it Lorenzo, I’m very sorry mate as I like your writing. They will know for surw that once the other PU suppliers get sorted they will definitely have a 0.5-1.5sec s lap deficit which can’t be made up by having lighter battering as the cars have a designated overall minimum weight.
           Good theory but it doesn’t stack up in my humble opinion.

         • Dear Clear View, the size of the ES does not necessarily equate to the amount of energy delivered to the MGU-K (as I tried to explain in another comment on this page). If Mercedes have found a way to run the PU with a smaller ES, that most likely means they have found a smarter way to manage the energy flows between the MGU-K and MGU-H. Bypassing the ES saves two conversions (AC -> DC and back), which means higher efficiency and lower cooling demands.

         • Would it not cause am issue in qually? I can see how it wouldn’t matter souch in the race as they can only harvest enough energy to refill the ES per lap, BUT in qually they would need to harvest 2mj in the 1st half of the lap to be able to use full ERS during the second half of the lap.

           I would love to see some telemetry showing how much energy you can potentially harvest in the lap, against what the rules state you can harvest. If you could potentially harvest 8mj per lap then having a small ES wouldn’t matter as you can charge it with 4mj lap as you don’t need 4Mj in 1 lump and can just keep topping up the 2Mj ES as the lap progresses. So, it’s , use a bit, charge a bit, use a bit, charge a bit. I like it, very clever way of thinking.

         • @ Clear View:

           4.2 Weight distribution :
           For 2014 only, the weight applied on the front and rear wheels must not be less than 314kg and 370kg respectively at all times during the qualifying practice session.
           If, when required for checking, a car is not already fitted with dry-weather tyres, it will be weighed on a set of dry-weather tyres selected by the FIA technical delegate.

3. Fantastic piece (particularly liked the closing statement).

   First question that jumped to my mind was what happens when the radio fails and the engineers can’t tell them what ‘mode’ they should be in? I assume they have a standard mode to switch to, but how much slower will that be?

   • Most likely they already have a plan B if radio fails, or they can communicate with the driver trough the pit wall (the good old table), obviously it will not be as good as the one suggested by the track engineer

     • Do you know if the team cam flash up messages on the smart screen on the steering wheel? I know the drivers have a confirmation buttons with yes and no on them and they can also communicate what the tyres are doing (graining or blistering etc) and which tyres they want for next stop. It would make good sense to send a text message to the steering wheel as the driver could then check it when not in mid corner or fighting the back end so he isn’t distracted by a sudden radio message and can read it on a straight and confirm yes or no, or do as directed and then confirm yes or no. I think this is why there wasn’t much talk between Fernando and Kimi to their race engineers about the issues they were having.
       Just a thought?

       • Having driven FF, F2000, and Atlantic (a long time ago!) I’m not sure how feasible reading text during a race is. The cars are very stiff, with lots of vibration and buffeting and little time to actually read something. I lined up all my gauges so the needles straight up meant everything was okay (oil pressure, temperature, rpm maximum) and even included an idiot light for oil pressure; this meant a simple glance read all the gauges, with no necessity to actually read. That’s why digital gauges are pretty useless in a race car. Reading text? Probably not, but who knows.

         • It was just a question really, I was looking for alternative ways of commutation. I don’t mean a full text message as you and I would write but say something along lines of “M6 T4” which could translate to Multifunction switch to position 6 and toggle to setting number 4, just as an example, they can read their lap times so a code of a handful of characters in my be feasible.

       • i would think it would be covered by;

         8.5.2 Pit to car telemetry is prohibited.

         • That covers changing any parameters of the ECU it doesn’t say about instruction to the driver, it was only a thought. It won’t be long and the drivers will be using Google Glass and they will navigate menus as normal but the info will apear on the eye piece.

4. Lorenzo, I always enjoy your writing and the level of insight is outstanding.

   So if I’ve got it right, as the season progresses and the ERS is tweeked by respective teams, I guess the more effective the ERS is, then the more overall ‘power’ will be available for racing.
   Example: if there was 1000mj of energy available for the race (this is not the correct amount but just for eg purposes) at the 1st race in Oz, then the chances are by next Oz GP there could 1100mj due to the optimization of the energy recovery. Allowing to the drivers to push harder as the race lenght is still the same but they have more energy to expend. This is where the ‘unlimited’ transfer of power from the H to the K is going to get exploited and also any changes to the IC by way of ‘reliability’ updates.
   So the more ‘power’ that can be extracted from the 100kg of fuel by the IC, the H and the K is the deciding factor on how fast a driver can go in the race. So say 305km race with 1000mj = 1hr 30min race time, but if you had 305km with 1500mj = 1hr 15min race time (just an example)

   I think I’ve got it……..

   • Thanks Clear 😉

     Yeah, the power that could theorically be recovered from the H and K is bigger than those that can be recovered from the ICE, as the more you push it harder, the more fuel you burn ( and also by the flat power curve ). Meanwhile the energy recovered, and exchanged between H and K in unlimited.

     It would be interesting if ERS power would have no time per lap limit, then we should see a dramatic increase of BHP from ERS, maybe something about 1000BHP

     • 1000bhp with 1000n/m of torque, now we’re talking Lorenzo. Maybe you and I could sit on the TWG and really get thing done LOL.

       • I guess with unlimited time ERS use we could easily break the 1000BHp limit

     • I think the time restricting of the ERS is a bit of a red herring (miss information, it’s a English saying) as yes it works out as 33sec at 160bhp but it’s more like 10sec at 200bhp then next corener exit 4 sec at 100bhp and another at 6sec at 130bhp, hence why in the regs it’s in Mj (megajules) not BHP. This thing of 33sec a lap is to draw a comparison to last years Kers system. As I say, if we look at it in Mj they will deploy it as needed, there is a max limit of power deliver (120kw I believe) and again in can b deplyed in a number if different ways to suit that precise section of track or race situation. So the 33 sec of boost in a lap is really a very very poor way to explain it if you want to know more than just the basics, this is why people often get confused of thing in F1, as by putting it in a basic format so the majority get the idea you maybe loose some of the cleverness and ingenuity that is the real story behind it.

       • Eg- full boost 33sec or half boost for 66sec so if you look at how long the driver is requesting throttle (1%-100%) and look at the level of demand he is asking for (I’m pretty certain that the boost is connected to throttle request) then the boost can be dread a lot further than the idea of 33sec boost a lap actually paints.

       • Just look at these calculations and you understand why they say there is 33.3 seconds a lap of full ERS usage (and 6.6 seconds of KERS). These calculations also make it clear why it is said that if you use less you power you can use ERS longer per lap (At least that’s what Sauber has said in the comments on one of their beginner’s guide to 3024 rules series).

         60W = 60 Joule/second
         400 Joule / 60 Joule ≈ 6.6 seconds

         120W = 120 Joule/second
         4000 Joule / 120 Joule ≈ 33.3 seconds

         And I don’t know if it is true but I wouldn’t be surprised if one of the reasons why this year a transformer is in the car to transform the normal power to higher voltages is because it would allow the MGU-K to work at different power outputs.

     • In my opinion, there is no limit in the regulations on the amount of time the MGH-K can deliver power to the crankshaft (or if you wish, the total amount of energy that can be delivered to the crankshaft by the MGU-K). However, the MGU-K power is limited to 120 kW (which equates to roughly 160 BHP), so if you want to get to 1000 BHP, it requires you to get 840 from the ICE, which seems impossible to me within the fuel flow limit of 100 kg/h.

       The only limits on the MGU-K besides its power are the 2 MJ output into the ES and the 4 MJ input from the ES; the energy exchange between MGU-K and MGH-H is unlimited. So the 33 seconds at 160 BHP from the MGU-K amount to the ‘boost’ that can be obtained by draining the full amount of energy allowed from the ES into the MGU-K. This says nothing about extracting energy from the MGU-H to power the MGU-K or vice versa, which can be done at any point in time for as long as one desires.

       Note that if Mercedes were to run an ES with a capacity of 2 MJ, this doesn’t mean that the energy flow from ES into MGU-K is limited to 2 MJ per lap. A simple way to understand this is the following thought experiment: during the first 16 seconds of the lap, they transfer 2 MJ from the ES to the MGU-K. In the second part of the lap they charge the ES again and during the last 16 seconds of the lap they boost again with the MGU-K with energy from the ES.

       • So I am pretty much on the money then, this stupid thing of saying about boost/time is a pile of hippo stool. It’s max energy allowed to be deployed from the ES and harvested by the K and how each team and driver use it to suit them best and get the best laptime is down to software adjustments on how much and when it is used.

         • I think you’re right, except that the total amount of energy that can be harvested by the MGH-K is unlimited, only the part of the harvested energy that can be sent to the ES is limited to 2 MJ per lap.

       • Yes but, the boost is only needed on acceleration so they boost then top us ES into a corner, then use boost again put of corner and harvest again into the next. My point is that putting time comparisons against it is rather missleading and not that helpful for getting your head around things.

         • Out not put sorry

         • I concur on the time comparison, that is just confusing people, including me until I looked up the energy flow chart in the technical regulations.

           I’m not sure about them not boosting on the straight when looking at the top speeds that are higher than last year with a less powerful ICE. But most likely the electrical energy for that comes from the MGU-H (since at full revs the turbo would otherwise build up too much pressure anyway, so you might as well harvest that excess turbo pressure).

     • There is a finite amount of energy available in 100kg of fuel, at present, according to the PU manufacturers, the PU’s have gone from 35% efficiency to around 42% efficiency (including ERS) so that still leaves 58% of the energy in the fuel being waisted (either heat or sound) so for the sake example and argument there is 1000kw of energy in the 100kg of fuel, if your engine is 42% efficient you are harnessing 420kw for 305km which equates to a certain maximum speed for the race. Now if you can male your engine 45% efficient then you have 450kw for the same 305km which will allow you to cover the race in a quicker time.
       Sorrt for all the posts, I’m thinking out loud and I’m looking for corroboration on my thoughts or to be told I’m way off the mark.

       • I think the efficiency that you refer to here is for the ICE. The overall efficiency of the PU has to have increased more than that, otherwise they would be much slower with a 33% reduction in available energy for the race (150kg vs 100kg of fuel).

         But the basic argument I believe is correct: they ICE only has a certain efficiency, the rest of the potential energy in the fuel is wasted in the exhaust and cooling water (equal proportions in first order approximation). The heat in the cooling water is lost, but the energy in the exhaust (both heat and kinetic energy) can be partially recovered. This is what the turbo does. Part of the energy harvested by the turbo is then used to compress the air going into the engine, the excess energy is harvested by the MGU-H.

         • Thanks Myrdinn, I’m quite glad that I’m grasping the finer details of these new PU’s. I guess if they are using 30% less fuel the efficiency must rise by 30% that would mean a step from 35% efficiency last year to to 45.5% efficiency (30% of 35 is 10.5), so I wasn’t far off.

         • I’ve really enjoyed our little exchange of comments, it’s what separates this site from others, where idiots argue half truths to the death without taling anything from the responses they are receiving.
           Thanks again.

         • @Clear View The pleasure is all mine. A big part of my fascination with F1 lies in trying to understand the unbelievable innovation shown by the teams and an open exchange of ideas with other fans only increases that.

           My motivation for thinking it is the efficiency of the ICE was the following: the cars are about 2% slower over a race distance compared to last year. They are heavier, the tyres are harder and the downforce has been curbed. This all results in lower cornering speeds. To loose so little compared to last year requires them to go faster along the straights (and accelerate faster out of the corners).

           However, lap time gained in corners is cheap in terms of fuel consumption, whereas lap time gained on the straights is expensive. This is because air resistance goes with velocity squared.

           So an overall efficiency gain of 20% (42/35-1) for the PU would be too little in my eyes to explain the comparatively small difference in lap time that we see. I suspect the 20% efficiency gain of the ICE comes from the downsizing and turbocharging (lower number of cylinders and lower revs both mean less friction losses in the engine).

           I’d be very interested to see some overall efficiency numbers from these PUs, but I’d also be very surprised if the manufacturers did provide them. ICE and turbo technology is fairly mature and there are no spectacular gains to be made anymore. This is not true however on the electrical side (also see comment from The Engineer below)

5. Pedants corner.

   The motor is polyphase and not DC. http://dropcanvas.com/nkdih

   The conversion from AC-DC-AC is via an inverter. http://dropcanvas.com/g5yb8

   Those pictures are from last years kers system. This year things are much beefier, and possibly a newer technology. I have a feeling that Lorenzo is on the right path, when he talks about Mercedes and their electrical drive system. In the last few years there have been a number of developments in new designs of electric motor. Much of this has been spurred on by the requirements of EVs. But the other issue is the shortage of material for powerful rare earth magnets. China, the main supplier, put a restriction on their exports. The links below, will give a small taste of the current R&D, for those interested in this area.

   http://www.qmpower.com/content.aspx?page=technology

   http://www.yasamotors.com/

   • Hi Iain – thanks for that

     Just for your info – if you post more than 2 links in your comment – the wordpress software puts it into moderation.

     Apologies for the delay in you seeing it up on the site

   • Interesting stuff. Although I suspect that the export limits from China don’t really affect F1 PUs since the production volumes are so low. They will go for the best performance no matter what.

     • Certainly F1 wouldn’t be affected by the export restriction. I read today that the WTO? have passed judgement against China in this matter.

6. Good overview, bit shaky on the facts.

   I believe that the advantage that mercedes is currently showing is down to the fact that they sat down and analysed all aspects of the PU prior to development.

   One of the first aspects of the analysis would have been where energy could be generated, recovered, lost and used. Subsequently they would have looked at the efficiency of each element.

   The analysis would have identified an area where energy loss is quite significant – charging and discharging an energy store based on Li-ion technology.

   In their documentation Tesla state that the best efficiency they are achieving is 86% using optimum charging cycles. So even if F1 technology managed to achieve 90%, this would still give a 10% loss per cycle – approx 200KJ! This lost energy is typically converted into heat.

   So, IMHO, the first design aim would have been to minimise the use of the ES.

   How would this be achieved?

   As mentioned elsewhere passing generated energy between the MGU’s, which is unlimited!

   Any energy generated by the H is passed to the K, and a commensurate reduction is made in the fuel being passed to the ICE if total output is too high.

   When K is generating (under braking), the energy is passed to the H keeping it spinning and consequently charging the cylinders with air, which will also retard the ICE by being compressed. Additionally as that compressed air passes through the exhaust, energy from it will assist in keeping the turbo spinning! Also the cylinder will be charged in readiness for when it is required, just needing the fuel to be injected.

   Should there be any spare capacity in the ES, it will also be replenished. Conversely if there is an excess of energy the brake by wire will apply a greater braking force to the rear wheels.

   Because the turbo is being kept spinning at the appropriate speed, there is no requirement to have a turbo that has a quick time response and so mercedes have used a turbo that is super efficient at high flow rates and consequently causes very low exhaust back pressure (look how dramatically different the mercedes turbo looks in comparison to the Ferrari/Renault).

   So while reducing the usage of the ES, the performance of the ICE has also been improved.

   The big question is how much energy is the H generating?

   The K is only allowed to provide 120KW. If the H is generating an average of say 50KW, the max that can be drawn from the ES in full boost mode is 70KW.

   2MJ of ES would provide 28 secs of full power boost – the more efficient the H the more time that would be available.

   Of course for this to work reliably you will need a rock solid K and H.

   How did I come up with such a convoluted theory?

   Well firstly, I graduated in the 70’s with a degree in Mech Eng and Systems and Control – and have worked in both motor and digital fields. Otherwise –

   Listen to the mercedes PU – it has a low frequency exhaust roar, without a typical turbo whistle (compare the Ferrari/Renault). Under braking there is a very pronounced whine – this implies to me that the K is working hard and also is using straight cut gears to ensure reliability.

   As mentioned above, the look of the turbo.

   And of course the PU performance.

   • In other words, Merc has thought long and hard about the concept of *usable* energy vs. maximum storable energy? Clever indeed.

     BTW, that turbo application has a real world potential to significantly improve mileage figures.

   • Thanks Engineer for your interesting thoughts. I have to say that while I also believe that Mercedes has done the best job so far, I can’t quite imagine the Renault and Ferrari engineers not going through the same process of analysing the system from all angles before they embark on a 250 million dollar project.

     Could you elaborate a bit on your argument that the compressed air exhausted from the cylinders under braking when keeping the turbo spinning using the MGU-H is beneficial. Wouldn’t you rather keep the turbo spinning but avoid it building pressure (for instance by opening the waste gate)? Compressing air requires quite a bit of energy of which a large part is then lost from the system in the intercooler. But maybe I’m missing something.

     • Instead of using energy from the ES to alleviate turbo lag, the turbo is kept at the optimum speed by energy from the braking process. That energy would otherwise be disseminated by the rear brakes. The energy that isn’t used from the ES can be used later on for boost. The amount of energy that needs to be recovered by the K into the ES is much less.

   • could the use of super capacitors be useful when it came to passing energy directly between the K and the H sad I can’t imagine the turbo with produce electricity at the correct volts/watts/Hz/Amps, what ever that use so it needs to be converted into ‘useable’ energy so chargeing a super capacitor or 2 would mean it’s gone nowhere near the ES and would obviously drained before the ES energy was harnessed. Perhaps this is why RedBull had soooooooooo much of a problem with dissipating heat? They were said to be using them last year and RedBull have designed their own ES and aren’t using the Renault supplied one as the other teams are using, hence why Renault said some of their issues where still with their PU but RedBull specific.
     What are your thoughts @Engineer?

     • could the use of super capacitors be useful

       No – Tech Spec only allows 5KJ of energy to be stored outside the ES.

       State of the art switch mode rectifiers and inverters have efficiencies in the high ninety percent – the Tech Spec specifies that it assumes 95% efficiency in conversions when checking energy transfers, so the actual values must be higher or RBR would be shouting from the roof tops!

       Incidentally the MGU’s work in the 300-400 volt range – that is why there are warning lights for the martials to show that the ES is isolated.

       Last year RBR put their super-capacitors in the gearbox, presumably using the oil to cool them!

7. Great article! Cheers Lorenzo, it has cleared many questions for me and even given me more of a reason to get excited about this very technical form of racing. My only gripe in terms of the actual race series would be that – as many have stated before – drivers are by nature racers, meaning that in the long run it (asside from the megabucks) would be less than satisfying to have to conserve energy, rather than deploy it in an out-and-out attempt to finish first.

   I can see that as a part of the big picture and being a team sport (particularly now, given the complexity of the “power units” (uggh, hate that term…)) the engineers are essential to the effort of getting optimal performance from the car at any point around the track. My admiration for the drivers has grown enormously with the level of technical manipulation on a corner-by-corner basis that these new cars require, with the engineers in your ear the whole time. I wonder if Kimi likes it…

   Cheers!

8. Wow so many comments ! I’m getting a little bit lost reading all of them LOL ! I give you a pic of the Renault V6 unmounted from Maldonado’s car

   Take a look : http://tmblr.co/ZAF-Ys1BVILJ_

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