godlameroso wrote: ↑
Mon Jan 01, 2018 6:29 pm
Imagine enough electrical energy to both remove the compressor as a load on the engine, with the MGU-K going at full tilt. There is no regulation as far as the capacity of the ERS, just how much you can harvest or deploy per lap. It's a no brainer to take the 5kg weight penalty and stuff as many exotic battery cells as you can in the ES casing. Particularly since power from the ES to the MGU-H Is unlimited.
I wonder what the maximum power the MGU-H is capable of in motor mode?
If it takes ~100kw for ~5 bar MAP. Does that mean the MGU-H needs to be that powerful? Or does it only need to be powerful enough to create a lag free environment? ie maintain a minimum and maximum turbine speed.
If so it allows you to size your turbo for extreme efficiency as the turbine doesn't have to be efficient over as wide a range of efficiency islands. Provided of course you have enough electrical energy.
Again if your combustion process is good it's easier to make that compromise. As you can just harvest K->H->ES, and you can take that engine penalty, still have more power and need less fuel to do it than everyone else.
Personally, I tend to think of this from the opposite direction - just another way to do it and both ways are equally valid. Let me explain how I see this.
The energy flow from MGU-H to ES is unlimited in both per-lap limit (amount) and in charge/discharge speed (rate), and given that the MGU-H can be used as a flywheel to allow some sort of alternative route from the K to the battery, then you'd have to work pretty hard on the sizing of both the MGU-H and MGU-K than if you ignored this route and optimised them against more traditional, simpler sizings. Both of those components are going to have higher limits than they would otherwise have, if you didn't see this route and optimise your PU around using it.
The first thing I'd say is that the 2MJ and 4MJ limits for K-direct-to-ES need to be maximised, so the ES is going to have to be able to store and discharge 4MJ per lap for the working life of the control electronics and energy store, purely from this workload alone.
All rechargeable batteries lose capacity with each charge-discharge-recharge cycle, so an ES battery at end-of-life ( which is what, 10 weekends next year?) needs a lot more than 4MJ full-to-empty at the start of its life in order to have that at the end of life. But it's not a 4MJ lap limit at all anyway. It's far more than that because of the flywheel route.
Consider the sizing of the MGU-H if you're using it as a flywheel a lot of the time during a lap.
If the flywheel is being used to charge the ES, then you're going to want it to be able to accept power from the MGU-K at maximum rate for the MGU-K (120kW) during the flywheel spin-up, and it's going to need to be able to dump that kinetic energy into the ES at some rate you decide.
You might guess/start off with at the same 120kW that was being used to generate electricity for the ES, so a 50/50 split of time spent charging and discharging.
I think that would be really ineffective, since there is an *unlimited* path from the MGU-H to the ES. You'd actually want that to discharge into the battery as quickly and for as short a time as possible - just for example, you might opt to have the MGU-H dumping power back into the ES at 240kW or 360kW, or more.
If weight and volume weren't factors, you'd want the rate to be as high as was possible with any known generator and battery technology, because the faster you can discharge energy from the flywheel to the ES, the more time you can spend receiving energy at the regulated 120kW limit from the MGU-K. That 50/50 time slicing of the charge/discharge of inertia from the flywheel is a bad assumption. You want it to be 99% charging, 1% discharging or better.
That leads to a sizing discussion about how fast the MGU-H could feasibly dump power into the CE and ES, how it would switch between charge/discharge and how much time the MGU-K would actually spend charging the ES using this route. It is unlimited in both rate and duration in the regulations. Why not a 240kW MGU-H, capable of time slicing 1/3d of the time putting its power into the ES and the other 2/3rds of the time charging up from the MGU-K?
This opens up possibilities. The whole system is also reversible, ES powering up MGU-H at some huge rate, then that power being used to discharge into the MGU-K at its limit. The time when the ES is spinning up the flywheel of the MGU-H, it would also be powering the MGU-K directly at 120kW.
It does mean that there is a >4MJ route from the ES to the MGU-K. In fact, it's essentially unlimited. You would want an ES capable of generating 120kW for the whole time the car needs to accelerate around the whole lap at the longest duration on the throttle (I suppose this will be somewhere like Spa, Monza or Baku). It would dump the 4kW permitted directly to the MGU-K only during the timeslices where the MGU-H wasn't able to provide the indirect route because it was being spun up itself by the ES. Total energy out from the ES through the K would be a hell of a lot more than 4kW by adding together the flywheel and direct routes. The ES would need to be capable of outputting 120kW as well as whatever limit the MGU-H was running at (eg. another 240kW), meaning its real limits would need to be potentially 360kW and upwards of 10MJ, for the sake of example.
Essentially, the 25KG limit for the ES becomes a hard limit. The sizing of the MGU-K would be such that it could run 120kW for a lot longer than the 33s the 4MJ direct ES to MGU-K would suggest - maybe double this. The sizing of the MGU-H would not be the amount of energy produced by the turbine and otherwise lost to the wastegate, it would instead be capable of large rates, in order to permit effective flywheel operations. The CE would need to be capable of switching 120kW from MGU-H to ES to MGU-K very quickly and smoothly for several races.
The ES thermals would be quite hard to control, with the amount of work it would be doing and the charge and discharge rates. The MGU-K is going to be physically bigger and will need more cooling as it's going to be active a lot of the time. The MGU-H is going to be doing a hell of a lot more work than it would have been doing if it was just a power takeoff for unwanted boost pressure. The ES and CE are both going to be throwing electricity around at much higher rates than the MGU-K basic 120kW, and the CE in particular is going to be switching quite a lot.
That's how I see things anyway. Forget 33s of deployment per lap, followed by whatever can be dripfed from the turbine in sustaining mode. What you're interested in is, how much power can you get out of the ES using the unlimited route. It's feasible that with an ES with enough capacity, the direct ES to MGU-K route would *only* be used to fill in the times when the MGU-H is being topped up with rotational inertia (at an unlimited rate!).
It'd be interesting to see what the capacity and discharge rates would look like for 25KG of lithium ion batteries, at a maximum, using modern cells. Also, it'd be interesting to know how many seconds per lap the hardest duty cycles are for tracks on the 2018 calendar, because it could provide a maximum required specification for quali mode.
Edited to add: Just looked at the FIA's notes for the ES and the regs state a maximum of 4MJ SoC difference from top to bottom in a lap. This means that the route out of the ES and into the MGU-K is likely direct for most of the 4MJ, but the charging idea still stands as a way to bypass the 2MJ ERS-K to ES limit.