HelpMeHelpU

Just read a good part of this thread and just wanted to say Thanks for all the great info--it's a very impressive collection of cool info.

I have wondered about RAS on a GTS and all of the thread's collective comments are very helpful. Thanks all

mdrums

worf928

Exactly. And I'm not sure what you're envisioning. That's why I asked if you had any insight into RAS's functional requirements.

I'm envisioning one set of requirements and I think you envision/speculate-on/have knowledge of a more complex set of requirements. I'm trying to figure out more-precisely what you are writing about and the provenance of what you write. If the provenance of what I write isn't clear then let me know. You will note that I write things like 'suspect', 'I think', etc.

Based upon what I've read about RAS, I thought I understood what it does and when it does it. Your posts cast doubt on my understanding.

Bottom line: I'm trying to get you to tell me why I'm wrong.

Great! So, you are closer to the 'inside'.

Lot's of folks post information. Sometimes it's right. Sometimes its wrong. Sometimes its hard to understand.

I'm not trying to argue. I'm trying to get a better understanding of what RAS really does and perhaps some insight into 'the guts' of the system.

Of course they are.

I went back and re-read all of our posts. I agree completely with the majority of what you wrote. There were two sentences you wrote that piqued my curiosity. These are the (minority) sentences about which I'm not sure. As I wrote previously: what I think you are writing about is absolutely possible. I'm trying to figure out if what I think you are writing about is the same thing you think you're writing about and if it has actually been implemented or is notional.

I re-post these two sentences below and highlight some specific portions:

The above two bold italic phrases are what interested me. They imply a much-more sophisticated control system than I expected given how Porsche markets RAS. (On the other hand, I can also conjecture reasons why Porsche Marketing might not be allowed to market the more sophisticated RAS operating modes.)

Do you mean to imply that RAS will adjust toe for a small fraction of a second when one tire hits a bump and the corning forces on the other three wheels change? And/or, that RAS will adjust toe one way on braking and corner entry, adjust it again at mid-corner and again at exit under acceleration (by incorporating inputs in addition to speed and steering rate?) This is what I mean when I use the terms 'dynamic' and 'high-speed' as applied to RAS' toe changes.

It would be really cool if RAS did all that. More rear toe-in (up to a point) during heavy braking from high-speed would improve stability. I haven't even thought about how or why specific dynamic toe changes through the rest of the corner would be beneficial. Do you know?

It would be much simpler and safer (due to various failure modes) to simply let RAS command a toe change for a given speed and steering wheel rate and then let the 'passive toe control system' (e.g. bushings, control arms, etc.) deal with all the 'other stuff' (e.g. bumps, friction changes, etc.) This is also 'dynamic' but much-less so that what I think you propose and doesn't require RAS to be a 'fast' control system.

I think that all the benefits you describe would derive from the latter. The former, while possible, seems unnecessarily complex.

Humans are quite slow as far as inputs are concerned in Control System Land. First derivatives of human input are easily calculated by 'slow' control systems that sample and respond to human input at low double-digit cycle rates (e.g. 25-ish hz.) It's the 'rest of the world' that requires higher sample and control rates (i.e. a 'fast' control system that must run at hundreds or thousands of cycles per second.)

I believe that the literature says this. But, that phrase is devoid of information. A passive toe control system (i.e. a well-designed rear suspension without active, commanded, actuators) can be said to 'adapt to driving situations' (by the writers of Marketing Materials.)
Perhaps there's more detailed information if the body of ZF's literature? Is it findable through the posted link?

Of course. This brings of the concept of 'fail-safe'. (As an aside, the "Daytona" link I posted is a case where circumstances led to the control system(s) improperly entering a 'fail-safe' mode.)

Here's what I think:

The toe-out RAS produces at low speed is - I think - unlikely to lead to situations that complicate vehicle stability control (because it's low speed.) Furthermore, attempting dynamic toe changes in mid-emergency (as opposed to simply commanding RAS to a 'fail-safe' static toe) could lead to hard-to-discover failure modes that might compromise stability (heard of Phugoid oscillation?) (If not see this: )

The toe-in change commanded by RAS at high-speed increases stability. At high-speed it is even-more necessary to be vary careful about what the system as a whole does and minimize control complications in exceptional circumstances. The more 'things' you have trying to affect the system's response the higher the likelihood that you will have detrimental coupling effects. And, it's the corner cases (i.e. 'emergency') that have the most 'bugs' that are excruciatingly difficult to find because testing the exceptional conditions is often very difficult, very expensive, or more-or-less impossible. So, designing the simplest control system with minimum necessary inputs and outputs to meet requirements is what you do.

I'm not saying that it isn't possible to field a system that commands dynamic toe changes under non-nominal (exceptional) conditions. It's just a lot harder to get it right. So, again, while possible, I question if VSC does - when stability control 'goes active' - anything other than tell RAS to 'go to neutral' and 'turn off.'

BlueNorther

duh, I posted in old thread instead of an active one on this topic, my bad