All Wheel Drive

[DannytheRadomski]

I know it has been attempted, but no team has found a good AWD design. Would it really be worth it? There are a lot of negatives: cost, time, effort, I'm sure lots of maintenace, worse fuel economy, etc... I can't find any positives (except for the BA factor, which doesn't help at comp.).
This isn't slamming teams that have had a fairly good AWD system. None of the systems I've heard of are what I like. I'm just trying to figure out why it could be a good idea. There may or may not be any logical reasons for having AWD, I just want to see.

[Will M]

Would it really be worth it?

Maybe; set up a lapsim and see.
It is a trade off as you know so I guess those teams felt it was net positive.
Some of the electric set ups look promising.

-William

[Warpspeed]

If you are racing on ice, mud, or gravel, it's a good idea.

On dry bitumen it isn't.

[PatClarke]

Originally posted by DannytheRadomski:
I know it has been attempted, but no team has found a good AWD design.


Delft ??

[Matthew Newman]

I know the UWA REV (Renewable Energy Vehicle) team is planning to have drive motors in each wheel for their electric vehicle. I'm not quite sure on the progress of it though.

[mech5496]

Combustion or EV? On EVs take a look at Zwickau, Delft and ETH. On combustion, might wanna take a look at Yamahas 2Trac system....

[Francis Gagné]

For a combustion car, look at the Palatov D1 (formerly the DP1). There is a blog about the initial development of the car http://dpcars.net/ I admit, 1000hp/ton is a lot of power.

At an FSAE Level, it did have and advantage on our basic LapSim. Think at how the tires gets loaded in the skidpad, awd helps there also. In the tire limited section of accel too. Does the weight, complexity and additional design time is worth the gain? No idea. The flexibility of EV systems seems to make the answer yes. Also the EV offers on the most part the possibility of torque vectoring, traction control, regen and ABS on all wheels.

Another drawback is car feeling. An awd feels very differently. You have to start designing with torque steer in mind.

[TMichaels]

As it has been said by the posters before:
The EV teams will probably disagree with AWD making no sense. For IC it is of course a totally different story, but if you watch this video of Delft's AWD Acceleration at FSG2012, you will probably have less doubt:
http://www.youtube.com/watch?v=u9e6I5D7Z3c

It does not prove that it must be faster on track, but fun to watch http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Max Trenkle]

If you do a free body diagram of all 4 tires in AWD, FWD, and RWD in a corner, you'll begin to understand many things... http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Warpspeed]

4WD with an electric vehicle certainly makes much more sense, particularly with dynamic braking.

And it also makes sense where essentially unlimited power is available, under very low traction conditions.

But 4WD on a good dry grippy surface will be much more difficult to drive to the limit, and all the extra rotating drive train widgets will add to total vehicle mass and take up space.

Being able to throttle steer around a corner with a bit of opposite steering lock is fast and extreme fun.
And with a bit of practice can be done very consistently lap after lap.

Four wheel drive is not at all like that.
It sticks like glue, right up to the point where it suddenly lets go, essentially without much warning.
In the hands of a champion driver it will do everything a Lapsim program says it can do.
In the hands of a rookie driver it can bite pretty hard if you make just one little mistake.

Carroll Smith once said something to the effect that the true speed of a race car is determined by how fast a competent driver can consistently take it without losing control.
I think that is the key.

It's not what Lapsim says it can do, so much as how far a driver feels he can push it without totally losing control.

[DannytheRadomski]

I'm not satying AWD as a concept is bad, but applying to to a Formula car built by college students. Is it worth it then? I'm thinking IC, ansd that makes it uch harder to justify.
Also, how exactly does the 2trac system work? I have never been able to figure out how hydraulic and pneumatic motors work.

[Freddie]

Danny, I don't know if you have found this, but Öhlins (they were deeply involved in the development of the 2Trac-system as they were a part of Yamaha back then, if I remember correctly) has published some info on their system:
http://www.ohlins.com/Checkpoi...-The-Complete-Story/ (http://www.ohlins.com/Checkpoint-Ohlins/2WD---The-Complete-Story/)

Also, if I got it right Monash tried out a combustion 4WD car in something like 2003/-04, but I haven't found any info about it. Can't say I've worked hard searching for it, though ... If you find something on it, I'd love to look into it as well.

[Warpspeed]

I cannot think of a more useless application of all wheel drive than a motorcycle.

High CG and short wheelbase make rearward weight transfer under acceleration extreme.
If you have enough power to lift the front wheel clear off the ground, and any decent bike can !! What purpose is there driving the front wheel ?

Now bikes run a LOT of positive caster, great straight line stability, even better under heavy braking.
But if you apply tractive force BEHIND the king pin axis with a lot of positive caster, it is likely going to rip your arms out and turn you straight into the weeds.

So what is the actual purpose and benefit of all this wonderful technology on a two wheeler ?

[Z]

Originally posted by Warpspeed:
I cannot think of a more useless application of all wheel drive than a motorcycle.
...
So what is the actual purpose and benefit of all this wonderful technology on a two wheeler ?
Tony,

Answer = The "fesh-fesh" (http://www.asphaltandrubber.com/racing/fech-fech/) of Dakar.

Z

[Warpspeed]

Z....


Originally posted by Warpspeed:
If you are racing on ice, mud, or gravel, it's a good idea.

On dry bitumen it isn't.

I obviously forgot to include racing on fesh-fesh and also racing on a rotting banana skin surface, would also be another example.

[Z]

Tony,

The discussion had moved on to 2wd motorbikes, specifically off-road ones. Also, you might rethink your comments about excessive Castor being a problem (wrong on several levels).

Z

[DannytheRadomski]

I'm sorry, but I'm more interested in finding out if AWD IC FSAE cars are a good idea.
But its extremely hard to control the direction a thread goes in, so here's my input. Yamaha put something similar to the 2trac system on a superbike, and they claim that they shaved five seconds off each lap with the two wheel drive. I think it would be cool to see it on some production street bikes. Each would have their merits: purists would choose RWD, but some people will go with the new 2WD.
http://www.gizmag.com/go/2351/ "Though the first motorcycle to be offered with Yamaha's 2-Trac system will be an off-road competition machine based on the WR450F enduro machine, tests on an R1 1000cc supersport road bike have shown an incredible speed differential on wet tarmac - tests at the tight Swedish Karlskoga roadracing circuit showed the 2-trac-fitted R1 to offer a whopping five second per lap advantage over a standard machine when the circuit was wet."

[Warpspeed]

Originally posted by DannytheRadomski:
tests at the tight Swedish Karlskoga roadracing circuit showed the 2-trac-fitted R1 to offer a whopping five second per lap advantage over a standard machine when the circuit was wet."

Ah yes, WHEN THE CIRCUIT WAS WET...

[Freddie]

Could the AWD maybe be used as a form of traction-control for an FSAE car? The traction controls we normally implement restricts engine power which restricts power transmitted to the rear wheels (I simplify a lot here, as I don't really know all details of different solutions), if there is a way of getting power towards the front wheels instead of restricting power could that be an alternative solution?

If I remember correctly, Nissan did use some sort of system like this for the Skyline GT-R in maybe the late 80s / early 90s, having an option where the front wheels would get maximum 10-15% of the power, or something in that magnitude. This would (supposedly) make the car feel like a RWD car but reduced track times significantly. Does anyone else have any more info on this?

[Warpspeed]

The Nissan GTR has basically the output from the gearbox solidly 100% connected to the rear diff in the conventional RWD way.

A multiplate wet clutch can be progressively engaged to transfer (some) torque to the front diff. This clutch is modulated by hydraulic pressure from an electric pump located in the boot, which in turn is controlled from a microprocessor.

The microprocessor has inputs from four wheel speed sensors, and two X and Y accelerometers, and throttle position. The software to operate this wet clutch being fairly complex, and not well understood.

The system reputedly works very well during straight line acceleration and braking, and in the wet. However, many people do not like some of the handling characteristics this car has in the dry when driven really hard on a grippy race track with slick tires.

It seems the software has a mind of it's own, and it can sometimes fight what a racing driver is trying to do.

If you read the road and track tests, and join in some of the GTR Forums, the 4WD control system is always a lively topic of debate, along with many ways to "fix" the perceived problems.

It's a bit like diffs in FSAE cars.
You get to choose between a spool, locker, various torque and speed sensitive LSDs or a fully open diff.
They all have various advantages and disadvantages, and the more radical you go trying to get traction, the more the handling can suffer as a result.

With 4WD, all the same problems exist, but now you have THREE diffs to screw things up with.
There is no perfect best 4WD system.
The various approaches to the problem have their own advantages and disadvantages in various situations.

[Max Trenkle]

The only real benefit I see in AWD is in the straights, as we have so much trouble getting power down. My idea for a AWD system is have it active only within certain angles of steer, or have a nice little function of %power to the front wheels vs. steering angle. This way you still get the proper RWD functionality in the corners, and when you're coming out of a corner, you can put the pedal to the metal.

[NickFavazzo]

AWD is beneficial on more places than just the straights, if the rears are operating at their max and the fronts have long performance to spare, why not use it?

[mech5496]

As I see it, the only possible solution in a IC car would be by hydraulic motors in the front. That is because there is no way any CV joint would tolerate the steering angles we design for. If it was light enough I might go for something like 2Trac (if the laspims give an advantage). Furtehrmore, one could use some sort of electric actuated valve to do some torque vectoring on the front wheels. If the system is not terribly complicated or heavy it might worth a try.

[rjwoods77]

Guelph did a AWD car (2004?) that used an front diff and front axle set off an ATV so it is possible to have the "necessary" steering angle.

[Max Trenkle]

Nick, you're gonna put out vectors from the front tires NOT (edit: not all of the force heads in the direction you want to go) in the direction you're trying to go (in a corner) if you're putting power through them. So how does this help you steer (get more acceleration in the direction you want it)?

[Owen Thomas]

Max, I believe Nick was referring to corner exit, where the rear wheels are at or past their limit while the fronts are still gripping (or counter steering). Correct me if I am wrong here.

Putting drive to the front wheels does indeed induce force in the opposite direction of the corner (understeer), but in this case it would balance the car. Definitely would require a different driving style.

[NickFavazzo]

Max, if you can use braking longs (4 wheels) to increase yaw (positive yaw moment) why can you not use forward acceleration to decrease yaw (create a negative yaw moment)? Provided you have available traction.

Owen, spot on with the corner exit statement. Though I think every car (and every tune) has its own style though, nothing odd there.

There are bikes with hydraulic drive on the front wheels, this sort of system seems like it would be suitable (I haven't actually looked into it much though).

[Markus]

From assembly point of view IC+AWD would be possible, I'd consider hydraulic system the lightest and simplest to implement without looking deeper into it.

But the rules... You would probably need some 3-5kg of shielding and in the end they would still figure out something to not let it run... And likely for a reason, pressured fluids can get nasty.

[Max Trenkle]

Well that's the idea of having the steering angle controlled AWD: as you exit the corner, you have the power you want going to the front wheels. As you turn in, you gain the RWD feel you want. In the straights you can make maximum acceleration from your engine.

Hopefully that's more clear. I was exaggerating about "only" being useful in straights, but I just wanted to be able to compensate for understeer upon entering the corner. Once you're in the turn, it's definitely possible to get unwanted oversteer, but coming out of the corner you start to get your AWD back. I think that you could develop a couple of relationships to test on the track initially, and tune them to decide how much AWD you want, when and where you want it.

Edit: And about not having U joints that can handle the steering angles... You can totally just weld like two or three U joints together and those can be you axles! You don't actually need the "axle" part. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Bemo]

The electric car from Delft from last year showed pretty well how much benefit you can gain from AWD. Of course it is much easier to distribute drive power with four independant electric motors than with an ICE. It would depend on the reaction time of the hydraulic motors if you can use the full potential of the concept.
Something to take into account would be the efficiency of the hydraulic motors. I don't have any ideas of the numbers, but you would have the efficiency of your hydraulic pump multiplied with the efficiency of the hydraulic motor. That could possibly result in quite a big point penalty in fuel. Also in cost you will lose some points.
The same question as always in engineering - is the advantage worth the disadvanteges http://fsae.com/groupee_common/emoticons/icon_wink.gif

[rjwoods77]

Just noticed that Guelph AWD car is in these pictures in reference to my previous post...

http://fsae.com/eve/forums/a/t...20528151#24720528151 (http://fsae.com/eve/forums/a/tpc/f/8356059423/m/52110721411?r=24720528151#24720528151)

[Warpspeed]

One fact is inescapable.
Vectoring torque around all four wheels with some kind of automatic control system will have a very profound effect on the handling characteristics and vehicle dynamics.

The simpler the 4WD system is, the more predictable it will be, and the easier it will be for the driver to learn to drive and come to terms with.

At the other extreme, some software program hooked up to multiple sensors independently controlling the torque fed to each wheel will have two fundamental problems.

It can never be set up to be optimum for absolutely every possible situation.
And it may sometimes work against the driver inputs, which may be fun or frightening, but never fast.

If you are seriously considering doing this, I might humbly suggest first you try building an electronic controlled rear limited slip diff, and write some software to control it.
That would be a pretty good project in itself and a great start to something a lot more complex.

[HenningO]

Originally posted by Warpspeed:
At the other extreme, some software program hooked up to multiple sensors independently controlling the torque fed to each wheel will have two fundamental problems.

It can never be set up to be optimum for absolutely every possible situation.
And it may sometimes work against the driver inputs, which may be fun or frightening, but never fast.



What about Traction Control? There are numerous racing series where automatic control of the drive torque helps the driver to go faster (even if the system is working against the drivers inputs).

A well designed torque vectoring system should not be more difficult to drive compared to a "a simple 4WD system". The challenge of course is to understand what is difficult for the driver to learn and come to terms with. A well designed/setup torque vectoring system will however get you closer to the optimum in every possible situation compared to a passive system. No race car operates at the optimum, the challenge is being as close as possible to it for the longest period of time.

Given the development resources required for such a project, a FSAE team must strongly evaluate the return on such an investment (as been discussed in many threads on this forum).

[Warpspeed]

Originally posted by HenningO:

What about Traction Control?
Traction control, like ABS is something to limit excessive driver input.

But consider how traction control might reduce initial standing start acceleration.
You give it a bit of a rev to start off, dump the clutch, and the throttle slams almost completely shut to totally prevent any wheelspin above some absolutely minimal amount.

Now in theory that should be faster, but in practice any serious drag racer will be extremely unhappy with that.

The all knowing computer has totally foiled what the driver was attempting to do.

[Bemo]

Of course implementing vehicle dynamic control systems needs some work to prevent undesired effects. But there is no need to be too scared of it. There are several examples of FSAE cars equipped with fully automatic launch controls which work extremely well.

The electric cars in FSE show pretty well the potential of such control systems. Of course it is more difficult to realise such a system in an ICE car, but it is definitely not impossible and there is no question that you can gain a lot.
To be able to react to different track conditions you just need some control parameters which can be adjusted from the cockpit. This way you can even change setup between two autocross runs without having to leave the lane or you can adjust stuff during driver change in endurance, for example if your drivers have different driving styles, thus requiring different setups.

The potential of this increases massively if you have AWD with the possibility of controlling drive torque at each whell independantly.

[DannytheRadomski]

Going back to the Yamaha 2Trac system (I just like it most because it is almost inherent torque vectoring), would it be feasible or even allowed in FSAE?

[Warpspeed]

I still think a gun driver with a bit of practice under his (her) belt will beat some computer nerd trying to adjust fifteen different variables on his keyboard while cornering at the max G.

[Bemo]

The problem is that this driver doesn't have four throttle and four brake pedals. It would be possible to build, but driving this would need quite some practice http://fsae.com/groupee_common/emoticons/icon_wink.gif

By controlling each wheel independantly you increase the theoretical maximum performance of your car.

I'm also quite a fan of conservative solutions, but you must accept that a computer is often much better in controlling stuff than a human (if you have the necessary input data and proper algorithms).