Is Your Brake System Simulation Tool Accurate and User-Friendly?

[Kozy]

Hello

For the past 6 years I have been working (on and off) on a vehicle brake system simulation tool. It takes a lot of inputs about the car and the brake components, along with a rate of maximum lateral acceleration which is used to determine the amount of grip available.

The model then solves for the point of front tyre lock up, a point which of course varies depending on all the factors involved, so for example if you add front bias to the brake force distribution, the overall brake force drops due to earlier front wheel lockup, the rate of acceleration therefore changes, as does the load transfer, stopping distances etc etc.

It provides a detailed analysis of the brake system, some the results include pedal force (accounting for booster assistance calculated from inlet manifold pressure and diaphragm area), load transfer, rate of acceleration, stopping distance, 'dynamic bias' (where 0% is always all four wheels locking simultaneously and 100% is no rear brake force at all) and energy dissipation.

There is also an input for engine power, which when combined with the weight and grip inputs, is used to model a rate of positive acceleration. From this information, it was also possible to work out the relative heating and cooling of the brake discs, based on the amount of energy absorbed by each one, the mass, material, cooling surfaces and also the rate of acceleration between corners.

An oval test circuit is then modeled, with the car accelerating, braking and cornering at rates congruent with the inputs, so for instance a car with more grip will corner faster needing to slow down less for the corners, but also brake harder in a shorter distance. One with more power will accelerate to a higher speed, and need more distance to slow down for the corner.

I am fairly happy with this now, but I am in need of some knowledgeable people to test it out and point out any obvious flaws. Also, I would like to hear any advise on anything that is not clear, as I would like this to be an easy to use tool for anyone looking to upgrade their brakes, not just for engineers with a lot of background knowledge.

The tool can be found here. Any feedback welcome, especially if you find problems / glitches as I need to iron those out!

 

[Simon Bridge]

What are you offering in return for this testing?
i.e. is the resulting simulation software going to be free software or open source?
Will you be publishing your algorithms? Will it add insight to the fields in question?

You are asking for the benifit of years of expensive study and experience after all.

Your best approach is to test the simulation experimentally, see how it's results compare with the kind of real-life situation where it is supposed to be useful. This sort of testing tends to be expensive, but thems the breaks: everything else is just someone's opinion.

 

[Kozy]

Ive tested it myself, I set this up for my car, a 1995 MX5 Miata, for which accurate stock prop valve data is available. These cars are renowned for having a lot of front bias which limits the braking performance in the interests of stability, and I was interested in the best way of improving the system by adding rear bias.

Apparently if I had no prop valve and fed the rear brakes full line pressure, I would achieve near 0% front bias, I.e all four wheels would lock up simultaneously on a hot dry day. It calculated acceleration would jump from 0.8g to around 0.9g. I had previously logged brake accelerations at around 0.8g with corresponding lateral acceleration of 0.9g.

I removed the valve and went out for a drive with the ABS disabled, and found that, exactly as predicted, all four wheels locked up simultaneously.

Data logs, crude as they were, showed acceleration under braking had indeed increased to around 0.9g, showing that the brakes were now making full use of all the available traction.

I have been running that for a year now and have had no problems with it, so consider the bias algorithm (by far the more complicated aspect) to be correct.

The brake temperatures thing is more experimental, I've no way to measure those and it's highly likely I've not got the right equations for modelling the rate of cooling. I know that is actually a very complicated thing to model accurately as it depends on airflow around and through the disc, however I was simply looking to have something that took into account things like the temperature delta between the disc and ambient, the heat transfer coefficient and the surface area of the disc, to provide some insight as to how changes to the discs will affect the relative temperature.

This is only intended as a free tool for use by amateur racers and tuners, however the threshold braking algorithm is quite a piece of work and combined with aspects like prop valve and vacuum servo effects makes this quite unique compared to all other brake tools I have come across, so I will not be sharing that.

 

[Simon Bridge]

OK - anyone sharing their expertise in testing this tool does so on the understanding that an important and unique part of the system will be kept from them.
Good luck.

 

[Kozy]

I am not sure what you think I am asking for? I am inviting people to use a free tool, and tell me if any of the results seems spurious and whether there are any glitches. That does not, in my opinion, warrant publishing all the equations and programming.

 

[Nidum]

There is quite a library of braking system modelling software available :

https://www.google.co.uk/search?q=v...-8&oe=utf-8&gws_rd=cr&ei=sg4qVqGGGsPzUI6QqdgB

There is additional software for modelling complete vehicles .

This is one of the best ones for braking systems :

http://www.plm.automation.siemens.c...lab/automotive/vehicle-dynamics/braking.shtml

Try to obtain use of one of the professional systems and see how yours compares .