Do rear brakes activate always, despite regenerative braking?

So (if I have understood correctly) in terms of the OP’s opening question:

The short answer is no, the rear friction brakes in a FWD HV with regenerative braking (like a gen4 prius) will not necessarily operate each time the brake pedal is pushed.

The long answer is something like: the outcome from the Electronically Controlled Brake System is affected by lots of variables; things like load distribution in the car, whether the road is rough or slippery, whether you are driving around a corner, your speed, the slope, the traction battery SOC, the drive mode selected, and how hard you press the brake pedal. The car computers are intended to integrate all of this through input of many sensors into interconnected TRC, VSC, EBD, ABS, and Hybrid Control systems, then control the hydraulic (friction) brakes as appropriate through a combination of hydraulic pressure from several pumps and the solenoids that send or relieve hydraulic pressure in each wheel cylinder, controlling each brake caliper. If you don’t mind Russian websites you can go here to get an idea of the interactions in a rav4 (or see specific models, probably copied from TIS).

In a gen4 prius, the control system no longer includes sensors of pressure in each wheel cylinder or position of each caliper. Some models have no obvious PID for brake pedal position or pressure (sometimes Toyota calls it ‘regulator pressure sensor’). It may be possible to ‘see’ some of what is going on with the friction brakes by logging a combination of hydraulic pressure and solenoid PIDs. The system details, and therefore the best methods to monitor them, vary between models. It may be necessary to use Techstream as not all OBD programs handle ‘status’ PIDs.

At the level of the ‘brake actuator assembly’ the following table is helpful (but in normal braking all of these are off, other solenoids control the amount of friction braking):



At some level of brake pedal pressure, depending on the other variables mentioned above, friction braking will be activated on one or (more often) several wheels. As observed above, this happens, at last to a slight extent, at quite low brake pedal pressure under some circumstances. The fail-safe mode for braking typically uses friction brakes (but it may need a lot of pedal pressure, and not provide much feedback through the pedal).

This may elaborate the many helpful contributions early in this thread.

 

Here is part of a (1 sec interval) log from today in my rav4 HV that shows (among other things):
(a) regen without friction while pedal is pressed.
(b) engine braking without regen or friction at high SOC (no pedal press)
(c) friction followed by regen, with gentle engine braking and without pedal press (who would have guessed?)
(c) various ratios of regen to friction depending on SOC, pitch etc.

All un-noticed while driving; the car computers take care of it all without fuss.

 

so lets just say battery is full and im driving on a strong steep downhill ( 6% for 12miles) , engine will brake as if im in a B mode, but gasoline is burnt , so only way to avoid the ICE run is by speeding up? i find this situation many times when it happens i just accelerate to 90 mph cause thats the only way to prevent the ICE from kicking in and to drain the battery a little, or else the enine braking will annoy me fo bout 10 minutes non stop.o for the next 10 miles of this downhill after battery is max charged

 

Fuel isn't used(*) for engine braking (B mode, D mode, or in any reasonably modern ICE vehicle, hybrid or otherwise).

The engine is moving in a Prius when engine braking is occuring, and it does indeed sound like it's revving up, but that's just everything working in reverse - the wheels are what's moving the engine, not the fuel combustion.

The work required to compress the air inside the cylinders converts this energy to heat. This is called adiabatic heating and is central to engine braking. There is also friction inside the engine. Both of these result in energy being transferred from the turning wheels and eventually converted to heat inside the engine, thus having a braking effect. But, like I said, no fuel is required for this.

Confusingly, the EV light will turn off when the above happens. I guess it's because the ICE is engaged to the drive train, even if it's not using any fuel.

You don't need to speed up or accelerate to reduce your fuel use. If anything, this will increase it by either making the ICE use fuel now, or possibly later to replace the charge in the battery that you just used.

Sure - the engine braking noise is annoying, especially when present for prolonged periods. I 100% agree. But that's better than wearing your brakes out sooner, overheating them, or overheating/overcharging the traction battery.

Remember that traditional ICE vehicles have been doing exactly this for decades, and hence why we're taught to change to a low gear when going down hills (and the engine will rev up and make noise accordingly). The difference is that in a traditional ICE vehicle you can reduce how much noise your engine is making by selecting a higher gear and compensating the the reduced engine braking by using more friction brake pressure. The Prius doesn't have gears though and it's all controlled via computers so we don't have this luxury.

(*) The above being said, the ICE will still use fuel if it's not up to temperature, if more heat is required by the climate control than the current temperature of the ICE coolant can provide, if the traction battery is low and there's a greater 12V electrical drain than can be safely supplied via the traction battery alone, and some other stuff situations.

 

Trivia for the day: the gasoline engine's braking effect is produced by acting as a vacuum pump, sucking air across the closed throttle.

In any braking effect, the work done by the car's momentum on the wheels has to be turned into some other kind of work, and in this case it's being turned into the work of pumping across the closed throttle.

It also takes work to compress air above the piston, yes, but that doesn't help, 'cause you get that work right back once the piston crosses top dead center; the compressed air just acts like a spring. But you don't get back the work of pumping across the throttle (because the intake valve closes before that would happen), and so that's how this braking effect can get rid of the energy.

It's different with jake braking in diesels, because diesels don't have a throttle, so they can't do what our engine does. Instead, they have to use compression, and they need a way to not get the compression energy right back again pushing the piston down. So the Jacobs brake changes the exhaust valve timing to pop off the compressed air into the exhaust right at the top of the compression stroke, after all the work has been done to compress it, and before it gets to give that work back to the piston. That's what makes jake braking so loud.

 

I think Prof is right, because heating means the energy recovery will be a lot less than 100%. But I learned a lot from ChapmanF, especially about the difference in diesel truck engine braking.

Who says the Prius drives like a truck?

 

I've done a few static regen tests in the Prime in both EV and ( HV mode staying within HV modes EV range to keep the the Engine from starting ), My observations are that it takes somewhere between 4x and 5x the amount of battery capacity to climb a grade as is regained from regen descending the same grade at the same speed both up and down. DrPrius helped to confirm by showing Amps used both up and down.
I'm aware that short static tests don't reflect the dynamic nature of regen while driving normally, especially with other traffic around, though I do think that 4 to 5 times the amount of energy is needed to speed the Prius up on a flat smooth road as is regained from regen while slowing down from that same speed on the same road, is pretty close.

 

The adiabatic heating Prof referred to as the piston goes up becomes just as easily adiabatic cooling as the compressed air 'spring' returns that energy pushing the piston down again. So the only heat energy that gets away is what net amount happens to escape into the cylinder walls and coolant during the portion of the piston travel where the air temperature is above the wall temperature. (Which is another way of saying, the amount by which the process isn't quite adiabatic.) The rest, you get back.

Diesels need that Jacobs mechanism to pop off the compressed air at the top there, so you don't get the energy back.

We, happily, have a throttle, so we can do the same kind of thing on the vacuum side. If the intake valves stayed open, the vacuum could also act springy and return energy by pulling the piston back up after the intake stroke. But the valves close before it can do that, and we don't even need a separate mechanism for making them close then, because that's what they normally do. (And it wouldn't be returning all the energy anyway, because unlike the compression 'spring', which is normally in a completely sealed combustion chamber, the vacuum is constantly being diminished by more air that has come across the throttle plate and needed energy to do that.

 

Interesting. I did not know about the vacuum effect until ChapmanF raised it. I would love to see an energy budget for the various contributions to engine braking torque, but I have not found one. I guess it depends somewhat on vehicle speed, engine speed and design tricks.

Here is a nice part of an explanation from Abishek:

"In gasoline engines during engine braking the butterfly valve in throttle body would be almost closed. So when piston moves down during suction stroke it's similar to what happens when you close the end of a syringe partially and try to suck in air, it will be very hard and absorbs a lot of energy to do it. Then you compress whatever air is drawn in and and some heat will be produced. A portion of heat generated is removed to surrounding while the remaining energy of compressed air pushes piston back during power stroke (except there is no power being generated). Then at exhaust stroke you push the air out to atmosphere. Here you lose energy during suction mainly along with some energy being lost during compression and exhaust stroke in addition to friction."

Maybe it is not so much from sucking air (or compressing or blowing air) as from failure to suck air (the work to move the piston against a vacuum)?

To get a benefit, there must be deceleration fuel shut-off (DFSO), which modern fuel-injected cars mostly have.

 

It is the work to pull the piston against a vacuum. Then the following exhaust stroke, the exhaust valve is opened immediately and air rushes in to fill the vacuum with regular atmospheric pressure, which must then be pushed out.

The compression and expansion strokes are mostly a wash, compressing and relaxing a spring.

If the exhaust valve would stay closed until almost the very end, then the intake and exhaust strokes would also balance out with the same spring effect, stretching then relaxing a spring. It is the immediate opening of the exhaust valve (necessary during normal fuel burning operation) that ruins the spring effect, so the work used to "stretch the spring" of the vacuum during intake cannot be recovered.

 

It would be interesting to get a handle on just how much work we're talking about.

If there is a long steady downgrade in the area (with light enough traffic / high enough speed limit), and it's steep enough, you can find a critical speed for the car, at which engine braking cannot slow you down. At lower speeds, you can use engine braking to slow, and above that speed, you'll continue picking up speed despite maximum engine braking, and have to tap the brakes occasionally to hold a speed. Right at the critical speed, the engine will rev right up to the maximum RPM the car uses and hold your road speed close to constant.

That speed, the car's weight, and the grade can be used to calculate the rate of work being done by the engine.

I've been in that situation any number of times, I've just never happened to think about working the numbers. (Or, ok, I've thought about it, but not buckled down and done it.)

A shortcut way could be if the descent is long enough to fill the battery. Start the descent with the battery not near full, and cruise control set to hold the speed, and watch the charging kW on a scan tool—a little more empirical than grabbing some figures for the car weight and the grade and doing the math yourself. As the battery charges up to 70% SoC the regen will be replaced by engine braking, and (assuming the grade's roughly constant) we can assume the engine braking work is roughly the same in kW as what was going into the battery just before.

It'll be a respectable amount ... I'll say some tens of kilowatts (or some thirteens of horsepower), anyway.

Certainly it's a lot less than what the brakes can do in short bursts. But the brakes get super hot in a short burst of hard braking, and the engine can do its maximum braking for as long as you like, with its operating temperature holding steady. So the work being done has been shed from the car, in some combination of the air out the exhaust and the air through the cooling system.

 

On two particular grades I've traveled many times, well beyond filling the battery, and wondered about similar things: 2012 Prius, about half loaded, posted as 7% grade, critical speed 52-54 mph. At 50 mph or less, it slows down, while 55 or more, it creeps up.

I compute (and hope someone will double-check) that as 25.5 kW of gravitational power, or about 34 HP. But remember that it also includes more than just 'compression' (or 'vacuum') engine braking: air drag, tire rolling resistance, and drivetrain mechanical friction.


Alternate formula, for vacuum pumping component only, for the simple case of 4-stroke Otto cycle engine with fixed valve timing: 0.5 * (engine displacement) * (crankshaft spin rate) * (intake manifold vacuum, i.e difference between ambient and manifold pressures), with appropriate unit conversions.

I should to try to resurrect some old calculations on my former 2.5L Subaru, but it was a much smaller figure than the Prius-on-a-downgrade calculation above. All the other drag losses well exceeded the raw engine 'compression' braking.

 

I wonder if the e-VVT is programmed to change timing to get better results while using regen.

from 1995 toyota VVT-i
Toyota Develops New VVT-i Engine Technology | Toyota Motor Corporation Official Global Website

 

Another reason it might be simpler to do the same run right across the 70–80% SoC regen-to-engine-braking transition, and just look at the reported electrical power being recovered just before the engine braking phased in. That's already net of the air drag, rolling resistance, friction bits.

Do we mean the same by 'regen'? I use it for when energy is recovered into the battery ... then engine braking is used in place of regen, if the battery is full.

There's a certain road speed where MG1's rpm limit is hit if the engine's not turning. Above that road speed, the engine always has to be turning, even if it's not used for power.

Below that speed, the best results for regen would be with the engine not turning. Above that speed, for best regen, you'd want the engine resistance as low as you can get. Maybe you'd open the throttle some to reduce the vacuum. (But then, the more air you let in, the bigger compression 'spring' you get. I wonder where the optimum is.)

Once you've filled the battery and can't do any more regen, then you'd want the engine resistance as high as you can get it.

As for the intake timing, Toyota has published what the limits of adjustment are (they're summarized in this post for three generations anyway). Maybe somebody can work out theoretically what points within that range would be best for low resistance (during regen) and for high resistance (during engine braking)?

 

Primes B works a lot differently than the non Primes do.
It might be difficult to explain what I call the Prime enhanced golf cart mode which for lack of a more precise description is (staying within the EV range of HV mode). The Prime will run EV as long as the Prime programming allows it, (until one of the thresholds is reached which turns on the engine). There is a complete set of rules / thresholds the govern when and for how long the engine turns on and the length of time it runs. And if a warmup cycle is needed in HV mode to bring the coolant temp up to around 130 F at slow speeds, higher coolant temp at higher speeds, the temp thresholds drop along side mpg temp related thresholds.
Plus if there weren't already enough of these kinds of odd behaviors, some have claimed the the 2020 Prime refresh has changes some of the thresholds for engine on even lower than the 2017 - 19 models.

There is an engine on speed of 20 mph, for some instances of engine on in HV, temperature dependent engine on (around 60 F coolant temp ( not sure which coolant loop yet )) and ambients around freezing, and probably whenever coolant temp is below 60 F. But even in freezing ambient temp if coolant level is above around 60 F the engine will not turn on until the 20 mph threshold is reached.

if the car is left in EV and B and speed goes over 20 mph, after letting off the go pedal the engine starts consistently.

Probably others I haven't seen yet too, especially at higher speeds with lower ambient temps..

In EV the only time I'm aware engine on (are from others posting about it) is downhill when traction pack is at 100% SOC. Overwise I've never had the engine start while in EV mode and engaging or disengaging B, even on the coldest nights or winter, at least that I can recall. Not that I would run the Prime again like that if I could have winters 1 and 2 todo over again, even with a heated cabin while charging and preconditioning at least once per drive cycle.

I don't even pay all that much attention to regen while I'm driving. Yes, I see the graphics, and I use B a lot, but there isn't really all that much difference in using B as having a feather touch on the brake pedal, other than timing the slowdown or stopping distance. If when approaching a stop in B and it's timed wrong, a feather touch on the brake pedal will enhance the regen and adjust the stopping distance much the same as timing the B mode stop down to about the 3 mph limit on the flat. Pretty much the same either way.
I've also noticed that regen seems to get used faster than EVSE charge. Kind of like the difference between a non sine wave and pure sine wave inverter connected to a sensitive electronic device. Or another way to describe it would be a dirty charge of fluctuating amp / watt input compared to the relatively steady computer controlled charge program while hooked to the grid or whatever the grid like source that the car is connected to.

 

There seem to be different opinions about the relative importance of various factors in engine braking, based on (a) thermodynamic theory of ideal gasses - like the Wikipedia article (b) insights of experts -mechanical engineers / engine system designers and (c) observations from those who perform real-world experiments - alas with too few measurements. See the input from Jag and BrendanM in the second half of this (often hijacked) thread - Engine Braking - On the road

This may be a sign that we need to refine the theory to fit the observations.

There is no doubt that Toyota hybrids can perform engine braking, though it is far short of the ‘hill descent control’ in some off-road AWD cars. Indeed, it is probably even more clever in a hybrid, where excess electricity from regenerative braking (excess meaning the battery is deemed “full”) can be consumed to spin MG1, thus revving ICE. Atkinson-cycle engines and e-VVT seem well-suited to engine braking under some theories. But even things like design of the exhaust system can make a substantial difference to engine braking (and help to explain why some small diesels without a throttle plate or Jake valves do show substantial engine braking).

Perhaps someone with a genX Prius can do the experiments to determine the contributions of various elements to engine braking in that model under particular conditions. But it would not be easy because the car computers may intervene without notice to change things in the car. fuzzy1 and vvillovv have clearly given it some thought already.

As it stands, I give credit to both Prof and ChapmanF for stating elements that probably make a substantial contribution to overall engine braking. It has been a very interesting thread hijack.

 

Every so often it is good to be reminded that there are other online communities far less collegial than this one.

Boy, was that a long, noisy thread. It pretty much took until Jag's first appearance, four days in and two thirds down the thread, before there was anybody more interested in working out what was going on than in calling other people idiots. I liked Jag's creativity in coming back (over three different days) having developed new examples to help explain things other participants were handwaving over.

Other guy seemed to specialize more in misrepresenting others ' posts to dunk on what they weren't saying. "That last post was intended for [...] who thinks it takes no energy to compress air." Seriously? Also kind of a Johnny One-Note on stuff like "Get it now? An ICE is not designed to be an efficient air spring" when nobody else's point depended on it being designed to be an efficient air spring. Others had just been pointing out that the compression and power strokes happen to be a spring—designed or not, efficient or not—and the degree of inefficiency can be quantified (essentially, how much of the heat departs into the cylinder walls and how much of the pressure gets past the rings), and that's your limit on how much of the braking work you can give it credit for. Jag responded constructively, with a rough number:

Jag also went an asked an ME and got an interesting response, giving more credit than expected to basic mechanical losses, and less to throttled pumping.

One thing the other guy harped on a lot was having tried flooring the throttle during engine braking (in a non drive-by-wire car, so it actually happened ), and finding there was still a respectable braking effect that way too.

There are clearly a couple of effects being traded off there. As you floor the throttle, you greatly decrease the manifold vacuum, and that contribution to the braking effect. At the same time, you greatly increase the amount of air going into the cylinders, the amount of air being compressed, the pressure it's being compressed to (though still you get most of that back), and the amount of air pumped through the exhaust plumbing. So it's possible the increasing effects are able to account for about as much braking work as the decreased one.

That could mean that when a source (like the Wikipedia article cited in the first day of the thread) says SI gasoline engines do it with closed throttle and vacuum producing the majority of the effect, it isn't necessarily saying that they couldn't do it with wide throttle and exhaust pumping instead, but only that closed throttle and vacuum is the way they always have.

That was the simplest approach in older cars because closed throttle was naturally what you got when the foot's off the accelerator. In modern drive-by-wire cars, both choices could be theoretically available, so one could wonder why closed throttle and vacuum is still the choice.

The other guy responded to the post with the Wikipedia link with the usual "Wikipedia is not exactly a peer review journal" mantra for discrediting a Wikipedia article without the hassle of showing anything mistaken in it. What makes that such a waste is that, even though Wikipedia articles are written and edited by random dogs on the internet, the things they say have cited sources, and sometimes those turn out pretty interesting.

In that Wikipedia article, the source for vacuum being the majority of the braking effect turns out to be an 18-year-old Ford patent about how an engine with electromechanical valve actuation (EVA) could have a bunch of additional choices (compared to the closed throttle and vacuum in prior art) for how to do engine braking. So it partly answers an earlier post here about VVT-i, too.

One thing it covers in the Introduction and Summary is probably the reason that closed throttle and vacuum is still being used instead of open throttle and exhaust pumping: they're trying to minimize how much engine braking cools down the catalytic converter and takes it out of its efficient operating range. The open throttle pumping approach would be sending so much cool uncombusted air out through the exhaust it would cool the catalyst way faster, and also load it up with so much stored oxygen that it would be lazy about breaking down NOx when the engine fired up again.

The scheme in the patent wants to go even further, and do engine braking with no flow out the exhaust at all. With EVA, it could just close the exhaust valves for the duration, and accomplish either expansion (vacuum) or compression braking, just by different choices of when in the cycle the intake valves are opened and closed. The air being pumped could largely stay in the intake manifold being shuttled around from one cylinder to another.

It includes a bunch of graphs showing interesting stuff, like engine braking torques achievable in the different modes.

Matter of taste, maybe, but I still prefer to reserve "regen" or "regenerative braking" for when the energy is being recovered to go into the battery. Once the battery's full and we're not doing that, it's just kinetic energy of the car, passing through the transmission, spun off in the engine: engine braking as in any other car. Sure, in our transmission, the path it follows is partly from ring gear to planet carrier mechanically and partly from MG2 to MG1 electrically to the sun gear and the planet carrier from there, but that's nothing but inside baseball about how our transmission works. It isn't, at that point, behaving differently than anybody else's transmission during engine braking.

 

Nice summary ChapmanF. I am not sure whether the Ford patent graphs are based on measurement, theory or imagination. Patents do not require any proof that an invention works, just an ‘enabling disclosure’ of how to practise it. Some people (like the car care nut in his video on the topic) believe that MG2 does generate electricity during engine braking, and that electricity is used to power MG1. If true, this is a neat addition to the non-hybrid method. I certainly see negative torque on MG2 (and RM in the rav4 AWD) at least sometimes while engine braking.

As I understand it, during engine braking in a petrol OTTO or Atkinson (with the fuel off):

During the intake stroke, the piston moves down to TBC. Normally air would be drawn in past the throttle valve(s), through the intake manifold, and past the inlet valve(s). But the throttle is closed, so the movement of the piston creates a partial vacuum. Work must be done on the crankshaft to move the piston against this partial vacuum. That work comes (ultimately) from the wheels: hence engine braking.

The seal of the throttle plate (or butterfly valve) is not perfect, so some air is sucked past: hence a partial vacuum, even after many piston cycles would otherwise deplete air from the intake manifold. During the compression stroke, the partial vacuum does some of the work to move the piston back up, but the energy recovery is incomplete, maybe up to 85%. The remainder is lost as heat through the engine block and coolant (and noise): hence engine braking.

In an Atkinson engine, the inlet valve(s) may stay open for part of the compression stroke, so some of the air that has been sucked in is pushed back into the intake manifold. The remainder is compressed, and compression causes heating. Normally there would be an explosion of compressed fuel and hot air somewhere near TDC, but this is engine braking, so there is no fuel to explode.

Therefore in what is normally the power stroke (powered by the combustion of fuel) the compressed air does some of the work to move the piston back down but the energy recovery is incomplete, maybe up to 85%. The remainder is lost as heat through the engine block and coolant (and noise). Also in the Atkinson cycle the power stroke is longer than the effective compression stroke, which creates greater vacuum. So even more work must be done on the crankshaft to move the piston against this partial vacuum. That work comes ultimately from the wheels: hence engine braking.

In the exhaust stroke, the exhaust valve(s) opens (a little earlier in some designs). Air rushes in from the exhaust manifold to negate the partial vacuum in the cylinder, then this air must be pushed out again through the exhaust valve, against whatever back-pressure is inherent in the exhaust (and anti-pollution) system design. Moving (pumping) this air requires work on the crankshaft to move the piston up. That work comes (ultimately) from the wheels: hence engine braking.

Any friction in the mechanical system generates more heat (and noise) that escapes from the closed system. And every conversion between energy forms (eg potential to kinetic) is less than 100% efficient. Remember that the most efficient ICE is below 40% efficiency in combustion mode. You can’t defeat entropy. So work must be done to run the engine in braking mode. That work comes (ultimately) from the wheels: hence engine braking.

The braking effect increases at higher engine speeds, so hybrid systems gain an advantage by powering MG1 (ultimately from the wheels) to rev the engine for engine braking. It may be less annoying if we think how clever this is. The relative contributions of all these elements depend on the details of system design and driving conditions.

 

I've been driving a hybrid for going on 2 decades now, like a lot of others in this forum and hybrids really do behave a lot differently when slowing down, than most / all normal gas powered cars do..

what makes it hard to follow the discussion at times, is that there are the typical variables missing in the descriptions, ie manual, automatic, cvt, which leaves the reader wondering exactly what is being described in which description as the writer might think it's not really as important to define those details for the reader.

Generally, any driver that owns a manual experiences engine braking whenever they down shift, where another driver with the same model having an automatic will experience a lot less engine braking, so much so that engine braking seems non existent to that driver, even though the engineer knows that there is engine braking happening even with the automatic.

It can get even weirder with a cvt ie: lean burn as in some of the honda hybrids and other such tricks programmed into the ECUs, which in my experience, when the driver can get the car into that mode and hold it in that mode for more than a few seconds, or the car just dumps into that mode and stays there by itself, behaves much like the engine braking a driver gets while driving an automatic.

It's already a confusing topic and the lack of exactly which instance of engine braking we talk / write about can make the discussion extremely difficult for those wanting to learn more about it, to follow.
And even more difficult at times for the writer, wondering how much detail is needed so the reader can follow along and understand better.