Is Subaru AWD more reliable and efficient than transverse AWD?

Is Subaru AWD more reliable and efficient than transverse AWD?

Other urls found in this thread:

youtube.com/watch?v=NZLYse15cj4
torque-vectoring.belisso.com/
youtube.com/watch?v=ooQRxlChvMw
twitter.com/SFWRedditImages

Audi's isn't, and Subaru's is very similar

no, that image is wrong

Do you think anybody on this board is qualified to answer your question? Look at the answers you've got so far, OP.

>FWD-based Haldex """"""""""AWD""""""""""
>in ANY WAY similar to symmetrical AWD

The main advantage of Subaru/Quattro is consistent power delivery because it can be all mechanical, in this day and age of electric or clutched rear drive.

Subarus is the best reasonably priced AWD available. Audi's Quattro was originally symmetrical and I believe some vehicles still have the symmetrical although all of their reasonably priced cars have shitty Haldex AWD.
Haldex is the best AWD for oblivious consumers, but it's the worst for performance driving.

As another user said, mechanical AWD is currently still the best option for high reliability and performance applications. It will soon be outpaced by electronically controlled applications, especially as electric motors become more and more used in AWD applications.

Penis

The A4 and up can still be had with Torsen-based "Real" AWD. Just stay away from "quattro ultra" models.

>Requires power to travel through several 90° turns
>Every turn you loose (sic) power

What the fuck? Don't tell me you actually believe this

no don't fall for that bullshit
it contains the same amount of 90 degree bevel gears

...

...

...

>tldr Subaru fans are plebs

Obviously you have never had a dynamics class or heard of drivetrain loss.
The more surfaces the power needs to transfer through, the more power you lose. This is why FWD cars have a very low percentage of drivetrain loss.
Front engine RWD cars have a slightly higher drivetrain loss percentage. Subaru's AWD system has a much higher drivetrain loss than FWD or RWD cars would. And another AWD system that needs to make more turns to transfer the power will lose the highest percentage of power through the drivetrain.

This has been thoroughly documented and is well known among the transportation industry. This is one of the big tradeoffs for AWD cars. This is why crank horsepower and wheel horsepower are different.

As an example. A RWD car might loose 15% of its power when transferring it to the wheels. An AWD car might lose 25% of its power to the wheels.
In a car that makes 300hp at the crank, that's a 30hp at the wheels difference, which is substantial, and why many high performance vehicles opt for RWD over AWD, not including the added weight of AWD.

...

Dual motor master race

Increased performance and efficiency.

indeed
we are only two or three gen's away from good battery's and proper cooling
the power grid and charging infrastructure has to pick up
already there are too many electric cars to be charged in some places
they have 1970s fuel crisis like queue

There is one reason electric will never surpass mechanical for a driver's car: predictability. Throw a Subaru into a corner in the snow. Do it way too fast, even. You will be able to pull it around the corner with the same inputs every time because it is so rock solidly predictable.

Electric is MORE predictable... Especially when the tech gets better.

OP, Subaru has several different AWD systems. But yes, most of them are the most efficient AWD systems in existence today.

>Do you think anybody on this board is qualified to answer your question?

Yes, actually. Many people here are very smart and experienced. I, for one, have a mechanical engineering degree from an engineering-only university and have worked:
-Quality engineering at a supplier (Bosch) factory.
-Powertrain engineering at a full-vehicle manufacturing plant.
-Test engineering at a full-vehicle test site.

...and I'm sure there are plenty posters here with more knowledge/experience than I.

>Is Subaru AWD more reliable and efficient than transverse AWD?
Yes and yes. They're mechanically simpler, with less drivetrain loss. However, some FWD-based AWD systems like Haldex will be FWD untill slip is detected, which can be more fuel efficient.

>Linear power flow
As opposed to what?
>Lower CoG
Lower, but further forward. A lot further forward, which is what causes terminal understeer.
>Equal drive shafts
You can get that in any longitudinal layout.
>Power loss because 90 degree turns
This is the only sane argument in your picture.

>Do you think anybody on this board is qualified to answer your question?
We probably have several F1 engineers lurking on here. A while ago, one of them even posted, although he ran off during a race (I still think he's part of Honda because of that).

Apart from that, there's plenty of dudes with mech engineering degrees and such.

>There is one reason electric will never surpass mechanical for a driver's car: predictability.
Confirmed for not having driven any electric racecars.

Active AWD torque vectoring is heavenly.

Putting your life in the hands of electrons, instead of steel.

So... how are we communicating to each other right now?

Steel.

Which has kinetic energy.

Which is made by converting chemical energy.

Which is released by...

A computer-controlled spark.

Seriously though, you can program electric motors to be perfectly linear. Probably better than any LSD, even. I know you probably don't trsut what you can't see, but you should get over that.

I'm not dying if my computer fails, I might be crashing into a wall if a ecu fails.

>I might be crashing into a wall if a ecu fails.
How, exactly?

sahara masterrace

The stuff you call torque vectoring, because OEMs market it like this, is most of the times actually single wheel braking. You trust it, speed into the corner, ESP fails, you dead, RWD torsen guy laughts and continues to drive.

A few notes:
>Your ICE (most likely) requires an ECU to run...
>Steel is made up partly of electrons.
>Most engines have very little steel in them.
>Computers generally fail from user error; embedded systems prevent this unless you physically damage them.
>Don't argue about something you know nothing about you filthy normie.

>The stuff you call torque vectoring, because OEMs market it like this, is most of the times actually single wheel braking.
That's passive torque vectoring. I was talking about active torque vectoring.

Now give me a single example of an ESP failure (not a driver failure) that actually caused a torque vectoring mishap and a crash. Either that, or stop trolling.

Actually, the SQ7 is explicitely market for its torque-vectoring as far as i know - works like a charm
youtube.com/watch?v=NZLYse15cj4

Again, passive TV, not active.

Are you sure you know what you are talking about?
Maybe you google your term "active torque vectoring" and then rethink it.

Stop derailing the thread, we are talking about active torque vectoring (by braking or with several electric motors) vs passive torque vectoring

>Are you sure you know what you are talking about?
Yes. We use it in our AWD electric racecar, and our native term literally translates to ''active'' torque vectoring, as opposed to what we call ''passive'' or brake torque vectoring.

>Maybe you google your term "active torque vectoring" and then rethink it.
Well I'll be damned: Subaru trademarked ''Active'' torque vectoring, even though it's clearly a brake vectoring system. We'd call that ''passief'' (passive of course) in my language.

Perhaps something is going wrong in my translation here.

Now, your translation is absolutely correct.
Passive torque vectoring = mechanical differential
Active torque vectoring = active differential OR single wheel braking OR differently driven electric motors
torque-vectoring.belisso.com/

Which makes perfectly sense, because electronics actively adjust.

>single wheel braking
This is actually considered passive torque vectoring in our courses.

And look at the SQ7 with its massive understeer thanks to the active torque-vectoring which made 0 impact on the outcome.
Besides the fact you shouldnt drive this tank there.

>What the fuck? Don't tell me you actually believe this
If you think tbat's wrong, how do you personally explain horsepower figures being different from a crank vs from wheels on the dyno?

>And look at the SQ7 with its massive understeer thanks to the 5000+ lbs kerb weight
Ftfy.

No amount of torque vectoring can fix Newton's second law.

Yeah, yeah, the point was Audi marketing the powerful acrive torque vectoring while crashing their fucking test car like pros

I want to make a Bugatti Type 35 replica with this

>Subaru
>reliable

>Throw a Subaru into a corner in the snow. Do it way too fast, even. You will be able to pull it around the corner with the same inputs every time because it is so rock solidly predictable.
>implying

There's Apex's problem, he did it in the dry.

>Say something retarded
>Get laughed at
>STOP DERAILING

Congrats, you found someone who didn't know what they were doing!

youtube.com/watch?v=ooQRxlChvMw

Why not Lan Evo?

objectivley wrong

>Never took a physics class
How does that even not make sense?

That doesn't make sense to me because brake vectoring/single wheel braking requires an active system to apply the brakes in a specific way based on current conditions.

The only way such a brake system would be truly passive is it the brakes on each wheel were applied automatically without the input of sensors or computer systems. A mechanical torsen LSD or the VLSD center diff like in the WRX are passive in that they simply react to a difference in the distribution of torque (ergo which wheels have the greatest or the least traction). There is no system that says "OK differential, apply 60% of torque to the outside wheel!"

Behold
+All the benefits of AWD
+All the benefits of FWD
+All the benefits of RWD
+Lightweight
+Less drivetrain loss
+Great traction
+More fuel efficient

it can only surpassed by one wheel drive
or perhaps land sailing

I GOT YOU
Steering is trimmed so you're not going off to one side.
Brakes can be used one at a time to steer
Rear engined so you're pretty much driving a top fuel dragster
SUPER LIGHTWEIGHT
Barely any drivetrain loss
Can get hektik (on one wheel) with a quarter of the horsepower you'd need in your old AWD car.
Subaru, I'll take a 22B for my signing bonus.

FUCK
This should be used for econoboxes, and my above design for old AWD designs.

Actually, even better. Delete the old steering system in the front and just have the steering wheel tied to the brakes.

It's true, though. Each turn requires some device to redirect the force (e.g., U-joint, CV joint, bevel gears), and any of those come with unavoidable friction losses.

Any transverse AWD car is a cunt to work on, I can tell you that much. At least with a subaru everything driveline related is basically the same as wrenching on conventional RWD, aside from the driveshafts up the front, which are usually pretty easy to remove when necessary.

Meanwhile, it's actually easier to do a clutch on a RAV4 by dropping the engine and gearbox at the same time. Shit's fucked.

>Test engineering at a full-vehicle test site.
Mechanic here, tell me: who is responsible for arranging components on a car?
I want to round them all up and drown them in acid.

We considered them passive, because it's not actively sending power to wheels like an electric unit. Instead, it's cutting power.

Still "thinking" Haldex is FWD based.

Found the Haldexcuck.

It is FWD-based, and FWD-biased.

Sup TU/E

Sup Delft.