65 replies to this topic

[Dusky]

Moke spider, I find your suggestion for the pressure sensing quite disingenuous.

 

Both you and Nick have failed to grasp that the specific braking situation for which this system is being proposed is NOT THE most effective stop. Go and RTFQ.

 

Nick, go out to your Mini or Mini based variant, sit in the driver’s seat and push the brake pedal as hard as you can. Then do the same at night, when it’s raining, when it’s dry, park it on some cobble stones, in a forest, do the same again when it’s icy, jack it up in the air, take the wheels off try any circumstance you care for. Now if you could measure the pressure in the braking system each time you performed this test do you think it would have a uniform value?

 

Then take yourself for a drive, anywhere you like, any surface, any weather, try a field. But get over 30mph to make it more exiting for yourself and then repeat what you did when parked, push with the same force and don’t let go. Now ignore any noise the tyres might make, the direction you’re going in and the angle you’re travelling. Would the pressure in the braking system be similar to the value when you were parked? Is this type of braking normal?

If you put a pressure P on the pedal, the pressur ein the system will always be P*X, with X being the leverage factor of the brake pedal .
Force on the brakes themselves depends on the master cylinder bore ,brake line bores,...

Its simple physics. F= p x A.  But it WILL remain constant unless you pop a hose.
If you do not understand this simple physics, please don't consider 'inventing' a brake system.

Edited by Dusky, 26 October 2015 - 01:52 PM.

[Dylan8660]

I was getting round to the simple physics but it looks like I have to take this slowly. Pressure = Force/Area. Your right the area is constant. But what am I saying about the force in those questions?

[Dusky]

I was getting round to the simple physics but it looks like I have to take this slowly. Pressure = Force/Area. Your right the area is constant. But what am I saying about the force in those questions?

The force to press the pedal as hard as you can will be the same too.

[nicklouse]

 

 

Nick, go out to your Mini or Mini based variant, sit in the driver’s seat and push the brake pedal as hard as you can. Then do the same at night, when it’s raining, when it’s dry, park it on some cobble stones, in a forest, do the same again when it’s icy, jack it up in the air, take the wheels off try any circumstance you care for. Now if you could measure the pressure in the braking system each time you performed this test do you think it would have a uniform value?

 

 

what do you mean by Uniform?

 

and no the pressure in the system will not be constant. and that is the issue.

[Dylan8660]

Thanks, so with that in mind there is now a pressure value that is predictable and repeatable and therefore detectable. It is associated with an abnormal braking scenario that unless your brake system is in need of serious overhaul will have your brakes locked up, all that without a wheel sensor in sight.

 

If you don’t feel that this is a realistic scenario and will never occur then I envy your good fortune and long may it continue.

 

If you did find yourself in this situation do you think that an automated form of cadence braking might be beneficial? Applying that to the existing system will be much harder than simply detecting when it might be needed.

[Spider]

Moke spider, I find your suggestion for the pressure sensing quite disingenuous.

 

"disingenuous

 

adjective

adjective: disingenuous

not candid or sincere, typically by pretending that one knows less about something than one really does."

 

 

Dylan8660, I take exception to your remark here.

[fenghuang]

and rightly so, I think.

[Ethel]

You can't use pressure as a trigger point. As it is not a constant figure. It all depends on the friction of the road surface and tyres. This is not constant or even equal. And that is the point at where what you are currently suggesting fails. Some thing that acts on the whole braking system without any consideration to the situation is down right dangerous .

See if you can get hold of "Vehicle and Engine Technology" by Heinz Heisler there is some good reading in there on braking systems and how they work and what is needed.

 This is true, and where I think we're arguing over confusion rather different views. ABS detects differences in traction (tyre grip) and then compensates by reducing the line pressure to each brake accordingly. It does it many times a second, faster and more accurately than any human could, giving a cadence braking effect.

 

Theoretically, you could fit any braking system to a car and avoid having to IVA it as long as you didn't collect enough points on the DVLA scale. I'm really not clued up on MoT's with ABS, I suppose if it's there it'd have to meet the standards in the same way retro fitted seat belts do. On the legality front, ringing up your insurers and asking for cover on your home-made braking system is likely to be the biggest hurdle

 

Bike ABS sounds interesting, for starters it'll be subject to much bigger variables that should make it more adaptable and less vehicle specific.

[fenghuang]

FWIW I've only ridden ABS equipped bikes a few times but didn't really like them. A lot of bikers don't. To be honest I don't know if they work on both wheels or just the front, but I can't imagine they are a linked system - both wheels would have to work entirely separately, or they'd be lethal. And on those I've ridden there's a toothed rottor in the front wheel (as you'd expect).

[Icey]

Caveat here - I've not ridden a bike with ABS but ABS on bikes has a significant effect on stability (which is the priority on two wheels); there are a few YouTube videos showing some great examples. Honda and BMW (most now probably) developed linked braking systems for the same reason (stability).

 

Anyway - this ABS/cadence thing. My tuppence worth.

 

Firstly, forget using a Raspberry Pi. Accurate real-time I/O with a full Linux kernel is not all that easy to achieve, yes you can tinker with servos and what-not but that is nothing like sampling sensors and monitoring sub-second timings. In most factory systems you'll usually find ASICs doing this kind of work, why? Because they will have run through stringent code quality checking and will be running the lightest code base possible, a few hundred lines at most. Compare this to the Linux kernel - 15 million lines (depending on how you compile it).

 

Secondly - Arduinos. Getting closer but again, unless you can write code that's MISRA (or the like) compliant and you can harden the hardware so that it can handle the harsh and noisy automotive environment, I'd forget it.

 

Electrical engineering and embedded software development are highly paid careers - they are hard to do! I work with developers who write code for military and space vehicles, to make this stuff work properly (i.e not fail when you need it most) is largely beyond amateur efforts.

 

 

And, if I'm understanding this thread correctly, the real heart of it - can you use line pressure to detect a loss of traction?

 

I don't see how the answer can be 'yes'. Simplify the problem - you want an input agnostic system that can adjust brake effort based on available traction. It has to be input agnostic because the pressure you apply to the brakes has no direct mechanical relationship with the available traction. ABS works because weather you stamp on the breaks or use them 'normally' the system doesn't car, all the system is doing is linking braking effort to available traction.

[nicklouse]

Now I would never have used agnostic in that way but it makes sense and a Google shows where it's use comes from.

And unlike say a launch control system a braking system has to monitor all the wheels and act on each wheel separately.

Remember the dual circuit braking system is only there to try and provide some braking if one brake fails. And the results are variable and depend on the split system you have. The results are either scary or very scary.

[monkey]

I feel I need to weigh in with my opinion.

Firstly it takes an awful lot LESS pressure to lock brakes on ice, with poor tyres than it does in the dry with good tyres. This is one basic example of how the pressure required to lock brakes is NOT constant.

Secondly, anyone who thinks that pressure required to lock wheels is a constant should not be playing around with any braking system designed not to lock them

Thirdly, I applaud any engineering attempts, BUT this is public safety and you would need to be very careful about how safe the system is, including fail safes to ensure safety. Also, as has already been said, good luck getting insured with home brew anti lock devices on your brakes - I know I wouldn't insure you without seeing intensive and conclusive testing results.

[Dusky]

The pressure to lock brakes is obviously not constant. A pressure limiter would be the only thing needed then.

But how is the pressure inside the system different for a given pedal depression X?

Physics don't allow it to be variable. ( yes, it'll vary a % or 2 because of heath etc, but nothing significant).

[Ethel]

Plenty of sense spoken there. Though the OP did recognise the difficulties of ABS in the original post.

 

The underlying phenomenon that all this tackles is that static friction is greater than sliding friction, which hits us with a double whammy - if the tyre skids your ability to decelerate is degraded,as well as losing your ability to steer as it becomes pretty much as slippy for the tyre contact patch in all directions. Plus, the lost reaction force that was turning the  disc between the pads means the that locks too, with static friction, and you'd need even more grip on the tyre to get the wheel turning and give you your control back.

 

Our problem revolves around the condition of things moving or not, so a motion detector must be the most appropriate trigger. What we do with it has a few options.

 

We could attempt to find the optimum braking condition by decreasing the line pressure until the brakes stop locking - while releasing and reapplying the brakes rapidly so we can sample the sliding friction we need.

 

We could just use cadence, equal on-off "pulses" of braking, and rely on the fact that because the regained static friction under the tyre is greater we gain more than we lose than a half on half off application appears to offer at first glance.

 

Couple both so your "half 'n half" tactics aren't applied for more than a brief moment so your motion detector can see if you're out of tyre traction trouble and you might have something workable.

 

I fully agree you wouldn't want to degrade your underlying plain old braking system, hence my suggestion to use a vacuum servo. In effect you'd have upgraded the system before you start. Yes, you could still get locked brakes in the sans-servo half of the operating cycle, but that would give you wider hysteresis for we mere humans to analyse and react to with our left foot.

 

(pun specific font)

[Dylan8660]

Moke spider, Are you saying that you were actually being serious? I though you were better than that.

 

Icey, No your not, but you aren't alone.