## Automotive Engineering Q

I've got an idea for a predictive type of braking system. The general theory is great, however I am not a Mechanical Engineer at all ... just an idea incubator

Anyway ... my idea requires the use of deceleration of the vehicle. Are there typically sensors already on the car that measure the actual rate of deceleration? My system depends greatly on knowing exactly how quickly the car is slowing down - either through applying the brake pedal, emergency brake or simple downshift/deceleration.

It's my understanding this information should be able to be gathered from any car with ABS as the ABS system has many different monitors related to braking and deceleration. Can someone tell me if this is true and if so, what these monitors are, what they measure and what unit they use to measure? Useful info would also be how one may tap into this. If you can't provide specifics, I'd be happy to just have better direction.

I would outline my whole idea here, but I'd really like to get a prototype and patent it before I go blabbing about it on the internet. I'm sure you understand.

Sorry if I posted this in the wrong forum too ... I was trolling around looking for the best place to post and this seemed best.

 PhysOrg.com science news on PhysOrg.com >> 'Whodunnit' of Irish potato famine solved>> The mammoth's lament: Study shows how cosmic impact sparked devastating climate change>> Curiosity Mars rover drills second rock target
 Recognitions: Gold Member Science Advisor Staff Emeritus ABS systems have one wheel speed sensor per wheel, that's all. The point is that the ABS computer can detect rapid decelerations of a wheel, which would be caused by lock-up, and can then pulse the brake on that wheel. The sensors just produce a signal whose frequency is the same as the wheel's rotation. If the wheel is rotating once per second, the signal has a frequency of 1 hertz. - Warren

 Quote by chroot ABS systems have one wheel speed sensor per wheel, that's all. The point is that the ABS computer can detect rapid decelerations of a wheel, which would be caused by lock-up, and can then pulse the brake on that wheel. The sensors just produce a signal whose frequency is the same as the wheel's rotation. If the wheel is rotating once per second, the signal has a frequency of 1 hertz. - Warren
That's a great start, knowing how the measurement is made. I guess at that point i would have to determine thresholds as to what is slower decelleration and what is rapid decelleration.

I was really hoping to find data after it is processed by the on board computer that may already have thresholds and/or switches set for different types of decelleration. Although, I guess since it's an ABS system, it probably only has one threshold/switch to either pulse the braking system or not ... just detecting lockup more than decelleration.

Warren - thank you very much for your contribution. If anyone else has some more specific data, I'd appreciate it.

Thanks.

-Joel

## Automotive Engineering Q

 Quote by chroot ABS systems have one wheel speed sensor per wheel, that's all. The point is that the ABS computer can detect rapid decelerations of a wheel, which would be caused by lock-up, and can then pulse the brake on that wheel. The sensors just produce a signal whose frequency is the same as the wheel's rotation. If the wheel is rotating once per second, the signal has a frequency of 1 hertz.
It seems that if the sensors relied on full revolutions of the wheel to produce any signals at all, that slamming on the brakes on an ABS car would cause much tire smoking and sideways sliding of the car. Perhaps ABS sensors are able to detect partial rotations.

 Recognitions: Gold Member Science Advisor Staff Emeritus http://www.asashop.org/autoinc/aug2001/mech.cfm Indeed, hitsquad, most "tone wheels" on cars have on the order of 50 teeth, allowing the ABS computer to detect 1/50 of a wheel rotation. - Warren

 Quote by chroot http://www.asashop.org/autoinc/aug2001/mech.cfm Indeed, hitsquad, most "tone wheels" on cars have on the order of 50 teeth, allowing the ABS computer to detect 1/50 of a wheel rotation. - Warren
Keep those links coming. That post really helped me by identifying what sensor to search for ... I was THAT far behind the 8-ball that I didn't even know where to start searching. If anyone has any other informational articles on the CAB (controller antilock break) referenced in that link, that would also be helpful as it seems to be the device that decyphers the signals sent out by the WWS.

-Joel

 Additionally ... Can I use an existing WSS for a second purpose? IOW, can I simply tap into it and read the voltage without affecting that voltage or would I have to install my own WSS? I assume I could use an existing sensor, but I figured I would ask.
 Recognitions: Science Advisor If you used a sufficiently high input impedance on your tap of the existing signal you should have no issues. Before you get too far into experimenting you may want to read up on the Mercedes braking system in their new SL where the push of the brake pedal is only one input into its brake application. It compensates for lots of factors, and even automatically dries the rotors when they're wet. Its the first of many braking systems coming out that take ABS to a new level, its effectively going to be a brake-by-wire world for the applications that need it. Ford has a development mule that detects an oncoming collision and applies the brakes fully and tightens the seatbelts faster than a typical human reaction time. Getting things sorted out for DOT and liability concerns is sure to make this a technology unseen for a while though.... Cliff
 Recognitions: Gold Member Science Advisor Perhaps looking at Airbag sensors, rather than ABS sensors would be more fruitful?
 For quick experiments, you can rig up a simple pendulum weight operating a wirewound "pot". This will give you a zero to 12 v signal that varies with deceleration. The pendulum is non linear so you might prefer a small weight connected to a spring, sliding on Nylon or some such, and operating a linear "pot". By playing with resistances in series with the pot you can tailor your signal to the decelerations you are interested in.
 Recognitions: Gold Member Science Advisor Staff Emeritus Freddie, That's probably one of the most convoluted ways to measure acceleration I've ever heard. - Warren
 I'm currently working on something like this. The mass spring system is not a bad idea actually. However I would opt out for an accelerometer. You can get a 2g accelerometer for < 15\$ (however oem units require electronic knowledge to hook up) An airbag unit will not be accurate enough, its meant to measure impacts, 100g's or so and you wont get the resolution you want (rather, that I assume you want) Another way to do it is just to measure the frequency change in the WSS (wheel speed sensor). Log time and find the difference between the frequencies, that is proportional to acceleration, assuming no sliding, of course. If you lock the brakes or ABS comes on you're hosed. A piezoelectric accel sensor would be more robust, if not more accurate, but only in a straight line. HTH
 "Collision Warning with Auto Brake - a refined warning system that makes the car brake by itself if the driver doesn't act when a rear-end collision with a moving or stationary vehicle is imminent." "The new system will be available in the Volvo S80, V70 and XC70 at the end of 2007." Source: http://www.technologynewsdaily.com/node/7894