UKC

Why do EVs need normal brakes as well as the regen? I understand that the current regen is more of a faster slowing than braking to stop, but could they be engineered to make the brake pedal for proper braking actually do regen braking so no need for brake pads? 

I am aware that a not insignificant amount of the particulate pollution comes from brakes and tyres, so was musing about ways of that being solved and it occurred to me that if there was a way of using the regen for it, we could have both better range and fewer particulates. 

Please, in your replies, assume very limited mechanical knowledge! My knowledge stretches to (pedal) bike mechanics!

You can’t regenerate much when the battery is nearly full.

Hydraulic breaks work with a total electrical systems failure - although if the boost servo dies you’ll need to push hard.

Most cars can break faster than they can accelerate - so a drive system and battery specced for acceleration will never be up to doing an emergency stop level of breaking.

Breaking is 4 wheels, many EVs are 2WD - again a limiting factor for hard breaking. 

Regeneration can’t do the final stopping - although active motor power can.

You still need some thing to stop your car rolling down a hill when you're not in it.

Actually the motor can do that. If you short out the motor it will be very difficult to turn and effectively braked. This is done in cordless drills to stop the chuck turning so that you can change the bit.

In reply to Girlymonkey. It would be possible to make an electric car without brakes. If the battery was fully charged the generated power could be dumped as heat - which is what the brakes are doing after all. The only reasons you need mechanical brakes are to brake the non-driven wheels and as a backup in case the electrics fail.

In one (hyphenated) word..... Emergency Stop.

That one trumps all other reasons

Depends on the motor type surely? 

Locomotives have been doing that for decades, but it’s a challenge for EVs; ours regenerates at half the motive power, and an emergency stop from 70 mph would need to regenerate and thermally radiate at something like 4 times the motive power.  By the time you’ve beefed up all the parts of the electric drive train I suspect it weighs more than brakes, as well as being far more complex; not a good trade off for something rarely used.

Also handy for torque vectoring through breaking on current EVs that use one motor and differential per wheel pair...    Shame you can’t buy the torque vectoring AWD Leaf - https://www.topgear.com/car-news/electric/nissan-has-built-high-performance...

there were experiments done with magnetic braking (as well as England's first experimental "escape lane") on Crawleyside (Stanhope) in the 1970s after the Crawleyside coach crash that happened outside our house in the late 60s.   I can't remember the outcome (and it would be very interesting to know) but it was a very steep hill with numerous deaths to its name.  The Softley bank on the far side of the dale going up towards Bollihope/Teesdale had a similar record with soldiers being killed in the 70s when an armoured vehicle overturned on the way down.

It does depend on the motor type, yes. Doesn't work with an induction motor. But I think most cars use a BLDC motor which would work.

My response was more theoretical though anyway. It could be done and be made to work, but I can't see anyone actually doing it, for the reasons you've listed.

At a fundamental level, not really.  It's physics that says an electric motor/generator can't passively act as a parking brake.  The magnitude of an induced electrical field is related to the rate of change of the magnetic field, and vice versa.

If the wheels aren't turning the motor (acting as a generator) is generating no electrical current.  It can only resist movement if there is movement to resist, and can only apply an infinitesimal torque to resist an infinitesimally slow movement.

Your not kidding, I pulled the Telma retarder out of our crane once to change the clutch, probably weighed 300kg. They are eddy-current brakes and air cooled but in a car? Anyway cars have to have two independent braking systems and fitted to all four wheels so there's no reason to add anything extra.

I agree with you on the infinitesimals; no shorted motor can resist motion unless it’s moving.  But an induction motor isn’t going to do that when moving either...

They could. The problem is mechanical brakes are extremely powerful, many hundreds of horsepower. That means for most cars (not so much electric supercars) that some of the electric drivetrain needs big upgrades. Motors can be builtto take brief but very big overloads. The controller can't, that's sized to suit thr maximum drive power and no more as upgrades are very expensive. The battery has a maximum charge rate too, generally it's lower than the maximum discharge rate so e braking is slower than acceleration. There are ways to store small ish amounts of energy very rapidly in capacitors, hybrid batteries, flywheels and otger mad options. Also you can dump the surplus you can't store as heat if the controller can handle it. Basically this is all veryexpensive for little gain so we see weak e brakes.

Finally there's safety, your mechanical brakes are redundant and reliable, they work (sometimes sub-optimally) whatever fuses pop, whichever computer craps out. Powerful e brakes require a lot of working circuitry and software.

As high power electronics (capacitors and IGBTs) fall in price and as batteries get more sophisticated we will get better e brakes but the humble disk brake is going nowhere, it'll just get a bit lighter.

Jk

So basically, it sounds like there are some ways of making it work but it's totally not practical. 

Thanks folks. Another great chapter of the UKC encyclopedia 🙂

I agree - the car is going to roll down the hill very slowly.  It's enough resistance to engage a chuck on a drill but it doesn't hold it dead still and nor can it.  I was more making the point that e.g. induction motors won't offer breaking force when shorted as there're no fields without external power to excite one of the fields.

I have no idea how fast it would roll down the hill - slow enough and its effectively a break.  I suspect it's not slow enough!

It seems to me that the power electronics are not dropping in cost as fast as either solar panels or batteries, and are on their way to becoming a limiting factor... ?

dissipate the energy boiling water to put in your heating system when you get home

The aim is no hydraulic brakes, according to a changing room conversation with a squash opponent a couple of years ago, but not yet there. JLR development employee 

I was going start saving for a Corvette C8 - the first ever 'vette to come in RHD, and now mid-engined.  I figured start saving now, buy a used one in 10 years time.

Then I read more about it. It's still hydraulic brakes, but the connection from the pedal to the booster is electronics.  Nooooo!  To no great surprise, they had to halt sales and do a recall due to a problem with the brake-by-wire system.  

I don't think I'm gong to like the future of cars.

Given Boeing cant get it right with planes and a $xxxxm budget, doing it in cars does sound dubious. 

Do you mean put a current through the motor coils or simple short it out, because those are very different things. Shorting it out, costs nothing, electrically , but won't stop the motor.

Although Boeing has screwed up big time with the 737 Max, practically all their other planes have been amongst the most successful aircraft of all time.

Shorting out the phases of brushless AC or DC motors will stall them (assuming their torque rating is large enough relative to the angular momentum of the system being braked). I’ve had tyre shredding braking applied during a thermal breakdown of phase to phase insulation in an electric formula 3000 car. The problem is dealing with this amount of energy in a short period of time and/or when the battery is fully charged. A bank of supercapacitors or a flywheel works well, but then you’re humping that mass around unused for 95%+ of the time. 
I ran a project for EADS Airbus on electric braking for aircraft, but the mass of the regen braking always “outweighs” the benefits

Edit

it can be done in aircraft, but it costs a lot of seats, so a no go with the operators

I get you.  I wasn't disputing your point about induction motors (or others that don't use permanent magnets), just pointing out a more fundamental reason why it wouldn't work with any motor.  Paradoxically, I think a frictionless passive electrical parking brake would basically be the same as a perpetual motion machine.

I'm pretty sure the brake that locks the spindle of a cordless drill while you tighten/loosen a keyless chuck is mechanical btw.  A really posh one might also use the motor as a brake to bring it to a rapid stop, I'm not sure.  Some circular saws, routers and other tools certainly do.  (But they're generally spinning with a bit more inertia than a drill or screwdriver bit and have no need of a 'parking' brake, they usually have a manual spindle lock if necessary to change the blade/bit/whatever.)

There is an entire class of electric vehicles that don't have mechanical brakes btw.  They don't have much of anything else either - pretty much just batteries, some electronics, a motor and a single wheel.
youtube.com/watch?v=i2OwOEHQ4vA&

Hard to tell from the consumer market but there's 1200V kit available now at about the price of 600V devices last time i looked a few years ago. Seems like the trend is to higher voltage.

As for lock up motor-brakes, you could do it with a switched reluctance motor and a permenant external field (like a magnetic door lock). Pointless but possible, parking brakes don't need huge holding torque. 

Jk

I do wonder though - if you had a super-conducting motor with serious magnets, just how close to stationary could it get on a steep hill? Enough to move imperceptibly slowly? 

Those wheels are insane.  Not my kettle of fish...  

How much fuel is left in the tanks for a typical landing?  Some stealth stuff uses the fuel as a heat sink for critical parts of a flight...  I see no down sides to this.

I suppose the missing part is a conductor that can go to very high temperatures without melting, and whose resistance doesn’t increase much with temperature, to give you a high specific power electric -to-air radiator.  Dopant based quantum tunnelling conduction in a ceramic matrix?  

The brake on my cordless drill is purely by shorting the motor. I've taken it apart and had a look.

As others have pointed out though, you couldn't make a parking brake like that because it would creep very slowly. But you still don't need discs though. My Leaf has a mechanical lock when you put it in park, this is independent of the parking brake that works on the discs.

Esp as, say, liquid hydrogen would enable power converters and electrical machines to operate at superconducting temperatures. Again, was on another EU consortium project with Airbus on their cryoplane concept. Everything comes together when there are fully all- electric aircraft.

I think his point was the reliability of advanced software safety systems cannot always be trusted as much as you might like to think, rather than that air transport is dangerous.

Three points certainly hold true in my experience:

- applying rigorous quality management processes to software development adds a lot of cost

- most software engineers don't like doing it

- (to clarify, not my area of experience but...) most car companies will be complete infants when it comes to understanding the processes and governance required

It's certainly a valid thing to be concerned about.

If you read up on the issues in their Starliner program, very similar concerns might arise there.

A very big part of the problem was the pilots didn't know the control response issues (that the company did). The trained pilot response made a potential failure situation worse. The entire design approach was unsafe for a passenger jet then, but in some weird scenario where the dumb design was approved, fully trained pilots might have been able to avert tragedy.

That works as a brake to decelerate a spinning spindle but does not, indeed can not, work as a spindle-lock to immobilise it while it's not moving.  Did you try turning the shorted motor by hand while you had your drill in bits?  You can move it quite easily.

As others have pointed out to you, because you alone have claimed otherwise in this thread. 

"You still need some thing to stop your car rolling down a hill when you're not in it."
"Actually the motor can do that."

But by all means please do feel free to lecture me about it now as if you were right all along. 

As a parking brake, of course that's true.  Electric vehicles would not have disc brakes if that was the only reason for needing them.  There's nothing new about immobilising the transmission instead, or as well.  (Direct drive would tend to limit the options a bit obviously, you can't immobilise the gearbox or the diff if you don't have one.) 

Automatics have had the option to put it in 'park' and effectively jam the gearbox since Noah was a lad.  Old Landrovers have a single separate drum parking brake in the middle of the transmission at the transfer box.  (Which is absolutely rubbish, nobody in their right mind relies on that alone to stop their 'series' Landy rolling down a hill.)

This makes quite a nice GCSE physics question, to show that the electric motors on an EV do not have enough power to brake within the highway code distances.  The figures claim that a car at 70mph (=31m/s) should be able to brake to zero in 75m.

s=(v+u)/2 x t gives a braking time of 4.8sec.  KE = 1/2mv^2 so (for my 1.8 ton e-Niro) = 865kJ.

P = E/t = 179kW.  But the motors are only 201bhp = 150kW.  Therefore they can't transfer energy quickly enough.  Real world, of course, braking distances are quite bit shorter than the highway code's 40-year-old figures and the motors are more underpowered.

But to address your (legitimate) question, as long as you don't drive like a 17-year-old, you basically never use the mechanical brakes.  Ease off the accelerator early and regen will do almost all of the work, and the mechanical brakes only kick in at very low speeds when the regen braking becomes inefficient.  

When I returned my previous car - an Outlander PHEV - at 45,000 miles, it still had the original pads and they showed basically no wear.  

I just went and dug out the old drill gearbox, and guess what? It does have a mechanical lock too, as well as the motor brake. I'll shut up now

Yes.  No.  Er.. haven't a clue. 
If you've got superconductors you can use for that though, it shows a bit of a paucity of imagination using them to build a slightly better version of the kind of electric car we already have doesn't it?

Ha ha.  I deliberately posted the maddest video I could find. 
They're not strictly speaking legal in the UK currently, like electric scooters, but they're not entirely bonkers for commuting.  The smaller, lighter ones are as easy to take with you on a bus or train as a chunky briefcase or a bag of shopping.

Phew.  And you posted just in time that I don't have to take mine to bits too.

The motor has to be turning, relatively fast, for this to happen, shorting a stopped motor and then trying to turn the shaft will be easy.

Back to the drill and they are mechanically locked, not simply shorted out. The same with a lawn mower, there's a mechanical brake which is held "off", electrically, while cutting. Dropping the power drops the brake.

Larger machines often have DC injection to stop them more quickly, indeed there are devices to do this retrospectively, for dangerous equipment, table saws come to mind. There was a change in the law to stop these machines from slowly spinning down after use. They need to be actively stopped.

I agree. It is my experience that rigorous testing and debugging is the sine qua non of software development.

Somewhere out of mind I have a copy of the MISRA C book.  It's all stuff about writing code that's not dangerously broken in the C language, towards the embedded systems side.  Really I see it as a damning indictment of the C language.  It came free with a Keil compiler I used commercially, I assume someone at Keil wanted to get rid of their copy.  To quote wikipedia:

MISRA C is a set of software development guidelines for the C programming language developed by MISRA(Motor Industry Software Reliability Association). Its aims are to facilitate code safety, security,  portability and reliability in the context of embedded systems, specifically those systems programmed in ISO C / C90 / C99.[1]

But this is all about low level safety in the code produced, not bigger picture stuff like making sure you've thought about the role of the physical layer in your cryptographic system or your use of a fundamental physical law (inverse square) as a form of authentication outwith the software.   

I was born and bred on Fortran in the mid '60s, but by the mid '80s, I had grown to loathe that language in spite of it being the standard in the company I worked for - they regarded it with almost an religious fervour. That's when I started to use C and other languages (and Kernigan and Ritchie's book on C was one that I bought). I never really got to like pointers. I used VB.net, later. 

I am completely unfamiliar with the use of fundamental physics in cryptography: I always left the security aspects of our software to professional programmers.

Incidentally, you have inspired me to look into Python. At first glance that looks pretty much like any other language, with its own slightly different syntax, but I suppose its popularity is based on its libraries and powerful functions.

I’m not sure how many programmers would get the big picture needed to understand the ensemble system with one time passwords as a contactless car key.  

I think that’s a pretty apt take on it.  Python as a language is a mixed bag IMO; 99% of the time I don’t need its dynamic typing, and static typing would help spot daft mistakes earlier on.  Likewise the object model is way more dynamic than most people need.  But the libraries available for it are phenomenal in quality and breadth.  It’s also trivial to plug C code in to it which is helpful when you have a critical section here or there, and when you’re working with a lot of custom hardware and associated IO.  

I think a diverse interest in “data science” has driven a lot of its recent increase in popularity.

Mostly I use it as a “glue language” to stick stuff together - custom tight loops on C, IO/networking and 3rd party libraries.  It’s a very good glue.