[SS-Traveller]

Originally Posted by Sutripta

@SS Finally the experts (faculty) are chipping in.

Originally Posted by Samurai

The trouble here is calling for experts to comment.
Therefore, as long as you use titles like faculty, experts, etc., most people here will be uncomfortable in sharing what little they know.

On a scale of 1 to 10, I think an expert on this thread is someone who's at least achieved 4 or 5 in terms of basic knowledge. We set out to talk about a fairly simple but extremely misunderstood topic, i.e. how all the 4 wheels of a car are driven together by multiple methods. At the heart of such methods is one of the oldest designs of machinery (perhaps second only to the wheel) that runs a car. And it is the arrangement of differentials in multiple patterns that distinguishes the type of drive - RWD, FWD, 4WD (part-time, full-time, on-demand) - and understanding the functioning of a differential is so essential to understanding these acronyms that keeps flying around to everyone's confusion.

Originally Posted by dadu

Great job SS, quite good and technical thread.

You've got to thank Sutripta for thinking this up!

Originally Posted by SS-Traveller

Right, basic engineering class #1...: ..why was the differential (which we talk about like we know all about it but actually don't) invented at all?

So we get this answer...

Originally Posted by headers

One does not need a differential if the speeds are low.i.e speed < say 20 kph. The vehicle & tyres will manage the frictional issues.

Our primary test mule would be the cycle rickshaw - a vehicle with ONE-WHEEL DRIVE! Only one of the two rear wheels is a driven wheel, while the other is free. If it were not, and both wheels were fixed to the axle, then...

Originally Posted by Sutripta

I don't think the puffing and sweating 0.1 HP engine [i.e. the human engine driving the wheels of the cycle rickshaw] is going to like this solution!

So we come back to our old question...

Originally Posted by SS-Traveller

Why do we need a differential at all if the cycle rickshaw can do without one?

And our experts say...

Originally Posted by headers

The differential as the name suggests was invented to accommodate the difference in the wheel speeds that arise due to different radii each wheel takes during a turn.

Originally Posted by sudev

The old horse drawn carriages achieved direction control by turning the front complete axle and wheels on a central pivot. This way the need for differential was all but eliminated.
Now take this forward to why differential comes in....

Originally Posted by Samurai

Differential is required when torque has to be transmitted to an axle whose wheels could turn in different speeds. An open differential always delivers equal amount of torque to each wheel, whether driving straight or turning (ideally).
Open Diff: Equal torque to both wheels, but allows different speeds.

I am putting forward two more explanations of why automobiles need differentials:
1. The final mover for the car is those little palm-sized patches of tyre on the road. And two of them are better than one. If the car was a one-wheel drive like the cycle-rickshaw, the possibility of wheelspin and wasted effort in moving the car would have been much higher. In a straight line, when the car is set in motion, BOTH tyres get equal turning force - and because 2 tyres grip better than one, we get better acceleration.
2. If there was one-wheel drive, say for the left wheel - and if one was driving through multiple S-bends - the left wheel would have to suddenly slow down on a left-hand curve, and speed up rapidly on a right-hand curve, to maintain a steady speed through these S-bends. Obviously, this would have unnecessarily strained the engine and transmission. The differential helps to average out this speeding up and slowing down of the wheels.

[Sutripta]

On a scale of 1 to 10, I think an expert on this thread is someone who's at least achieved 4 or 5 in terms of basic knowledge.

Let us close this by saying 'amongst the most knowledgeable regular participant on this forum.'

We set out to talk about a fairly simple but extremely misunderstood topic, i.e. how all the 4 wheels of a car are driven together by multiple methods. At the heart of such methods is one of the oldest designs of machinery (perhaps second only to the wheel) that runs a car. And it is the arrangement of differentials in multiple patterns that distinguishes the type of drive - RWD, FWD, 4WD (part-time, full-time, on-demand) - and understanding the functioning of a differential is so essential to understanding these acronyms that keeps flying around to everyone's confusion.

You've got to thank Sutripta for thinking this up!

Thanks SS, but suggestions/ suggesters are a dime a dozen. Doing matters, and you are the doer. Remember when you said "then why don't you put up a thread on this", My response was "Me! No. Too much hard work".
(To others wondering whats going on, an explanation. Call started with exchange of Puja greetings, then veered round to the hot topic of the day - the Aria. And then to its drivetrain (of which we still don't know much), and its perceived positioning. It was in that context that I said that there was too much of misconception about 4WD, which needs to be addressed, and also going back a step, about RWD, FWD, etc. )

Since I started the cycle rickshaw question, let me close it. The basic question was - if the cycle rickshaw does not have a differential, why does everything else?

Let me start by saying that our cycle rickshaw does not have a differential not because it does not need one, but because it makes do without one. Just like it makes do without proper brakes, or proper gearing. How does it make do without these. Because at the slightest sign of adverse conditions, the 'engine' jumps off the rickshaw, and manhandles the vehicle.
Not having a differential does not mean it will not benefit from one. It is a cost benefit analysis done by both buyer and seller. This same analysis says that for a powered vehicle, a diff is a must.


How about the other option for the cycle rickshaw (leaving aside lateral thinking alternatives like front wheel drive, as in a kids tricycle) is an axle with wheels firmly attached at both ends (ie a spool). It is really not an option.

When taking a turn, total tyre slippage has to be equal to the difference in the arc lengths transcribed by the two wheels (a slight amount can be accomodated by deformation of the tread area of the tyres). And this slippage will be against the traction afforded by the tyre. You have to give enough torque to break traction. Anything less, and you will stall. And this force (and energy) is totally wasted. Components like chain, freewheel, and axle flanges/ keys are going to be stressed, but before any significant damage, the 'engine' is going to give up. As an experiment to gauge the effort necessary, (and going to waste) may I suggest getting on a pedalcycle, locking the front wheel by whatever means, and trying to pedal forward a few metres.

So the only thing left is the 1WD system used. Let us see where this leads us.

I am putting forward two more explanations of why automobiles need differentials:
1. The final mover for the car is those little palm-sized patches of tyre on the road. And two of them are better than one. If the car was a one-wheel drive like the cycle-rickshaw, the possibility of wheelspin and wasted effort in moving the car would have been much higher. In a straight line, when the car is set in motion, BOTH tyres get equal turning force - and because 2 tyres grip better than one, we get better acceleration.
2. If there was one-wheel drive, say for the left wheel - and if one was driving through multiple S-bends - the left wheel would have to suddenly slow down on a left-hand curve, and speed up rapidly on a right-hand curve, to maintain a steady speed through these S-bends. Obviously, this would have unnecessarily strained the engine and transmission. The differential helps to average out this speeding up and slowing down of the wheels.

I would put more emphasis on 2 rather than 1. Motorcycles transmit quite a bit of power through one wheel.
The other way of looking at 2 would be for uniform rpm of the power source, speed of the vehicle (as seen from its geometric centre, or CoM) would be a function of its direction of travel.
Also we would expect (even demand) that the vehicle behave identically for a left turn as a right turn. For a 1WD, that is certainly not going to be the case.
Powering 1 wheel of a pair is not an acceptable solution. (Proof:- It is not used on the Tata Nano)

Regards
Sutripta

[sudev]

Going OT:
Vis cycle rickshaws - Would it not be more efficient to have power delivered at the front wheel with a hinged driver seat rather than they way it is? Something like rickshaws in Indochina where the driver is at the back and swivels the passenger seat to steer? But deliver the power in direction you want to go rather than straight line and then waste the effort by steering it at an angle?
Oh god does my question make sense?

[SS-Traveller]

Originally Posted by Sutripta

I would put more emphasis on 2 rather than 1. Motorcycles transmit quite a bit of power through one wheel.

Right. Even a T-Rex does that.

Originally Posted by Sutripta

The other way of looking at 2 would be for uniform rpm of the power source, speed of the vehicle (as seen from its geometric centre, or CoM) would be a function of its direction of travel.
Also we would expect (even demand) that the vehicle behave identically for a left turn as a right turn. For a 1WD, that is certainly not going to be the case.

So we've needed to live with the differential all these years, and as far into the foreseeable future as we can see.

Originally Posted by sudev

...power delivered at the front wheel with a hinged driver seat rather than they way it is? Something like rickshaws in Indochina...
Oh god does my question make sense?

Sorry, Sudev, I give up - what does such a rickshaw look like? Any images? Tried Googling, but nothing came out.

To continue...
Now that we are stuck with the differential driving 2 wheels on one axle till (at least) our lifetimes, the problem we constantly face is of losing traction on one wheel, when it lifts off the ground or gets onto a slippery patch on the road. Taking a fast turn around a corner is a major problem. As the car goes round the turn (let's say a left turn), the right wheel is travelling faster than the left (arc with greater radius), and keeps moving the car. But there's this stupid interference called centrifugal force, which wants to tip the car over on to its right - it can't do that all the time, but does manage to lift up the left side to an extent.

Lift up the left side, and the left driving wheel's lost contact with tarmac - and it starts to spin faster than the right. But the right wheel was pushing the car, no? It's got no power to turn anymore - the differential has transferred drive to the faster-spinning wheel (the left one). The vehicle loses speed around a turn as a result. And the engine is howling away, over-revving and overheating, but the car's not going any faster. Damn! How're we going to win that race?

Thus came the need to have differentials where a wheel can spin faster than the other - but only upto a limit. The first use of LSDs was in racing. Off-road vehicles found the LSD to be a nice thing to have too, and adapted it for some more movement potential when one wheel is stuck in mud. There are now a variety of designs of LSDs, which was perhaps described a little too prematurely on this thread - so I'm reposting it here for your reference:

Originally Posted by SS-Traveller

Now LSDs are not standard fitment in all cars, because they're expensive, and in 99.99% of day-to-day driving, a car doesn't need them. But for vehicles venturing onto slippery surfaces frequently, or in very high-performance cars (especially when going round a corner), one of the driven wheels can lose grip. LSD then helps to transfer some of the power from the engine onto the other wheel, and keep the vehicle moving.

So how do LSDs work? There's the fun of it all! They're more complicated than your average FNG (or even A.S.S. technician) can figure out, and as a result they stay miles away from it. But broadly speaking, LSDs are of two types:

1. Torque sensitive - using different methods to work:
____A. An arrangement of gears which does not allow one wheel to spin too fast, before the other side gets its share of turning force;
____B. Clutch plates which are pressed against each other by the freely rotating wheel;
____C. Cones - a solid cone fitting inside another hollow one.

2. Speed sensitive - using these methods:
____A. A viscous coupling with a special "oil" in it, which becomes thicker and stickier when heated up by friction;
____B. A pump and clutch pack - electronics govern a pump, and once sensors tell the system that one wheel is spinning free, the pump is made to engage the clutch - this transfers power to the other side of the diff.

[SS-Traveller]

The most common type of engine and driven axle combination in the first 70-80 years of automobile development, was the FR layout (Front-engine, rear-wheel drive layout - Wikipedia, the free encyclopedia).

A few issues cropped up with the front-engine rear-wheel-drive set-up.

A. The front was heavy with the engine's weight, but the lighter rear had to push the car. Tyres were not getting enough grip to push the car as technology improved and engines pumped out more power. There was a limit to how wide the tyres could be made and fitted, to improve grip.

B. A bugbear called "oversteer" plagued (and still plagues) all RWD vehicles. The rear wheels wanted to go straight, while the front ones demanded that the rears follow them.


Many solutions have been tried throughout the history of automobiles, including

- adjusting tyre pressure,
- changing wheel alignment,
- different suspension and damper settings for front and rear,
- four wheel steering
- front wheel drive (with its own bugbear of understeer),
- four wheel drive
- ESP (electronic stability program) aka ESC and a variety of other acronyms.

[YaeJay]

Originally Posted by headers

One does not need a differential if the speeds are low.i.e speed < say 20 kph. The vehicle & tyres will manage the frictional issues.



The differential as the name suggests was invented to accommodate the difference in the wheel speeds that arise due to different radii each wheel takes during a turn.

Guess this was illustrated in the initial posts [page 1]

Very true, This was explained to me and my friend by his jeep mechanic at Pollachi. Was a true 'bulb' moment for us at 10th std. .


Excellent information, I din't even think of a cycle rick till date!


5*!

[Sutripta]

Originally Posted by sudev

Going OT:
Vis cycle rickshaws - Would it not be more efficient to have power delivered at the front wheel with a hinged driver seat rather than they way it is? Something like rickshaws in Indochina where the driver is at the back and swivels the passenger seat to steer? But deliver the power in direction you want to go rather than straight line and then waste the effort by steering it at an angle?
Oh god does my question make sense?

Sorry, Sudev, I give up - what does such a rickshaw look like? Any images? Tried Googling, but nothing came out.

I think he means something like the 'cycle vans' used in Calcutta to deliver bread and other grocery items early in the morning. Tadpole with a hinged tail.
No, I don't think mechanical efficiency is the issue. Stability and dynamics is. As the tail hinges around, the base supporting the vehicle changes shape. Drastically. On the other hand, the tadpole with properly steered wheels has a few things going for it. Lets close the cycle rickshaw chapter.

So we've needed to live with the differential all these years, and as far into the foreseeable future as we can see.

Till individual electric motors at each wheel take over.

Now that we are stuck with the differential driving 2 wheels on one axle till (at least) our lifetimes, the problem we constantly face is of losing traction on one wheel, when it lifts off the ground or gets onto a slippery patch on the road. Taking a fast turn around a corner is a major problem. As the car goes round the turn (let's say a left turn), the right wheel is travelling faster than the left (arc with greater radius), and keeps moving the car. But there's this stupid interference called centrifugal force, which wants to tip the car over on to its right - it can't do that all the time, but does manage to lift up the left side to an extent.

Lift up the left side, and the left driving wheel's lost contact with tarmac - and it starts to spin faster than the right. But the right wheel was pushing the car, no? It's got no power to turn anymore - the differential has transferred drive to the faster-spinning wheel (the left one). The vehicle loses speed around a turn as a result. And the engine is howling away, over-revving and overheating, but the car's not going any faster. Damn! How're we going to win that race?

Well before overheating, you will have valve bounce/ dropped valves/ con rod through block/ piston through head.

Actually, because differentials is the subject of this post, I think we should be precise. The open differential will supply the same torque to both wheels. The value will be determined by the wheel which requires the least torque. And a wheel spinning in air requires very little torque. Which is also what will be supplied to the heavily loaded right wheel.

Also, today I think with ESP, we are thinking in terms of appropiate traction, rather than traction/ no traction.

Originally Posted by SS-Traveller

The most common type of engine and driven axle combination in the first 70-80 years of automobile development, was the FR layout
Why was it the most common?

B. A bugbear called "oversteer" plagued (and still plagues) all RWD vehicles.
Your bugbear, my fun.

The rear wheels wanted to go straight,
Once again, why?
Both subject for another thread. Not this one

Many solutions have been tried throughout the history of automobiles, including
.....
- ESP (electronic stability program) aka ESC and a variety of other acronyms.

ESP is what interests me. Because I know absolutely nothing about it.

Regards
Sutripta

[SS-Traveller]

Originally Posted by Sutripta

...something like the 'cycle vans' used in Calcutta... Tadpole with a hinged tail.

Originally Posted by sudev

...power delivered at the front wheel with a hinged driver seat rather than they way it is? Something like rickshaws in Indochina where the driver is at the back and swivels the passenger seat to steer?

The cycle van has a 1WD rear drive - Sudev mentions a front drive.

Originally Posted by Sutripta

Why was it the most common?

And the quiz goes on! Rack your brains, gentlemen (uhh... edit: and ladies too!), and if you have an answer, let's have it.

[sudev]

Without google or any lookup in reference I'd say ... uuhhhh .... because .... ahheem ... it was easier to make the non steering wheels the driven ones? And it sort of distributed the weight balance nicely?

The engine could have been at the back mounted to drive the rear wheel but then the front wheels would have had much less weight (and resulting traction) to be able to steer effectively?

Sir, do I pass?

[SS-Traveller]

Originally Posted by sudev

...it was easier to make the non steering wheels the driven ones? And it sort of distributed the weight balance nicely?

The engine could have been at the back mounted to drive the rear wheel but then the front wheels would have had much less weight (and resulting traction) to be able to steer effectively?

Sir, do I pass?

Sudev, so you pass indeed!

Historically, there were two inventions that had not come into being (or the engineering / mass production capabilities of those times could not support it) when automobiles were rolling off assembly lines. These were the constant velocity joint and the transaxle.

What we understand as front-wheel drive today is the FF layout (Front-engine, front-wheel drive layout - Wikipedia, the free encyclopedia), with the engine mounted transversely over the front axle, driving the front wheels. To achieve this, the two essential components were the CV joint (else one couldn't drive and steer at the same time) and transaxle. The original Mini used this layout for the first time.

One might argue that the Citroen Traction Avant was historically the first front-wheel drive mass-produced car, but it did not follow the FF layout - rather, it had a MF layout (Mid-engine, front-wheel drive layout - Wikipedia, the free encyclopedia). Therefore, it could do with a conventional differential, instead of a transaxle. It also had conventional U-joints instead of CV joints, but I'll keave the technical explanation of that to Sutripta.

[sudev]

Thank you professori !!

Before you go on further on trans axles and CV joints. Should we not delve in to connection between gear box and differential? And differing differentials and what mods are possible on standard diffs?

[Sutripta]

Originally Posted by SS-Traveller

Historically, there were two inventions that had not come into being (or the engineering / mass production capabilities of those times could not support it) when automobiles were rolling off assembly lines. These were the constant velocity joint and the transaxle.

Hi SS,
I would not really agree with the assessment on transaxles. The transaxle (transmission and differential in one unit) is an evolution designed to fill a need. Any competent manufacturer could design and build one:- provided he saw a need to. The need to is a result of optimising a different set of criteria. (In the case of the Mini, it was internal volume to external volume.) Other constraints generally saw that the need did not arise.

We shall also not get into subtleties of calling the Traction Avant mid engined. For our purposes of understanding drivelines, it is a front engined, front wheel drive vehicle. Weight distribution should be the subject of another thread.

Joints (non smoking variety). If there are two shafts, driving and driven, at angles to each other, we need a joint in between them. Ideally, both the driver and the driven shaft should turn through equal incremental angles, for all values of rotation, and deflection angles, be capable of transmitting fair amounts of power relative to its size, have good service life, and be economical to produce.

The most common joint is the universal joint. It meets most of the conditions of the ideal joint, except the one of equal rotation.
If for one revolution of the input shaft, the output also rotates once (has to!), where is the problem? It is just that in the course of a complete rotation, sometimes the output shaft will rotate slower than the input, sometimes faster. Thus if the input shaft rotates at constant (angular) velocity, the output, over each single cycle, will show a variation of (angular) velocity. The universal joint is not a CV joint.

If it is not a CV joint, how did we use it even in our front engined, rear drive vehicles. As we are pottering along slowly, we don't see our cars hopping along, do we?

The trick is in this. If the input and output shafts are in the same plane, then if we introduce a shaft in between these, and if we maintain the deflection angle at each end to be the same, then with proper phasing of the UJs at each end, we can exactly compensate for and eliminate the variations. The output shaft will now accurately follow the input. Viola! we have our propeller (cardan) shaft. (An interesting point to be noted is that though the input and output shafts are following each other properly, the propeller shaft itself is not. It is speeding up and slowing down over each cycle. Luxury car makers nowadays object to this.)

However, for front wheels, this solution does not work. If we have to have any steering at all, we cannot maintain the condition of 'in the same plane'. If a UJ is used in a FWD, esp when cornering, it would cause the wheel to speed up and slow down cyclically. For various reasons, this is not acceptable.

The CV joint (by its very name!) solves this problem. Its way beyond my powers to explain the intricacies of a CV joint. Wikipedia, (and other sources on the net) give a good description, with animation, on how it works (as long as you do not ask awkward questions!)

Regards
Sutripta

[SS-Traveller]

Originally Posted by SS-Traveller

Citroen Traction Avant... had conventional U-joints instead of CV joints, but I'll leave the technical explanation of that to Sutripta.

Originally Posted by Sutripta

Traction Avant... is a front engined, front wheel drive vehicle.
...with proper phasing of the UJs at each end, we can exactly compensate for and eliminate the variations.
However, for front wheels, this solution does not work. If we have to have any steering at all, we cannot maintain the condition of 'in the same plane'. If a UJ is used in a FWD, esp when cornering, it would cause the wheel to speed up and slow down cyclically. For various reasons, this is not acceptable.

Sutripta-da (and everybody) - first, Shubho Bijoya to you!

Now I've got to ask for my own understanding - how did the Citroen Traction Avant manage with U-joints?

[Sutripta]

Originally Posted by SS-Traveller

Sutripta-da (and everybody) - first, Shubho Bijoya to you!
Same to you, and everybody else.

Now I've got to ask for my own understanding - how did the Citroen Traction Avant manage with U-joints?

Manage is the right word. It had all the inherent problems one associates with UJs. You accepted the Gallicness of it. For sharp steering angles, it is quite pronounced. The late Mr. Latif (Calcutta camera expert) had one in which it was very noticeable. Mr Sircar last year restored one. I'll ask him to take a tight turn in front of me! I think (not at all certain) somewhere in its production run, the UJs were replaced by CV joints.

Another thing I forgot to mention was that in a front engine rear drive config, the deflection angles are small.

What about the CV joints? Beyond my competence.

Regards
Sutripta

PS- Ive asked another forum member to check out something. Will revert on his reply.

[SS-Traveller]

Originally Posted by Sutripta

Manage is the right word. You accepted the Gallicness of it. For sharp steering angles, it is quite pronounced. Mr Sircar last year restored one. I'll ask him to take a tight turn in front of me! I think (not at all certain) somewhere in its production run, the UJs were replaced by CV joints.
What about the CV joints? Beyond my competence.

Hmm... so the Traction Avant was a leap forward, literally, in it's drive system, but not quite a smooth leap. Never knew it had such problems.

Could someone explain the intricacies of the CV joint in simple lingo please?

Originally Posted by sudev

Should we not delve in to connection between gear box and differential? And differing differentials...

Topic for basic engineering class #2?

In this, could we talk about the different types of shaft that join a gearbox to the differential, as well as types and designs of shafts (axles) joining differentials to wheels?

Originally Posted by sudev

...what mods are possible on standard diffs?

Shouldn't this be a topic for an advanced engineering class?