Roll Center analysis with steering angle? 1

[ksw100]

Do you have any tips how Roll Center analysis should be thought about when considering varying steering angle? When analyzing Macpherson strut and Double Wishbone suspensions (with caster and / or king pin inclination) I imagine that the systems' roll center will be different if the wheel are pointed straight versus steered at an angle.
To narrow the scope of this discussion, let's focus on road sports cars (i.e. the likes of McLaren's, 911, Supra, Corvette, etc.)
What would be the top goals regarding roll center and steering when designing such suspension systems?

 

[BrianPetersen]

Since no one seems to be touching this ... Thought experiment.

What steer angle do you anticipate the front wheels being turned to during cornering at speeds high enough for it to matter? Hint: do some calculations. For example at the dizzying (moose-avoidance test) speed of 20 metres per second (72 km/h) in the formula A = V squared / R where A is close enough to 10 m/s2 (1 g) then R (The radius that the car is cornering at) is 40 metres. If the cornering radius is 40 metres and the wheelbase is (let's say) 2.5 metres, what does trigonometry tell you the nominal steer angle is (we're not considering tire slip-angle here, but let's just get a ballpark to see if it matters). Does that steer angle matter in the grand scheme of things?

Perhaps imperfections in the design are recoverable at 72 km/h during your double-lane-change moose test, because that's more autocross speed than Nurburgring speed. What about at 200 km/h? What's the steer angle now? (Hint, it's so small that this is going to be more dominated by tire behaviour than geometry ... and you had better have more slip angle up front than out back i.e. understeer if you want to stay out of trouble ...)

Do you care what happens at parking-lot speeds on pavement, when high steering angles could be involved, but the driver can bring the car to a complete stop within 1 second with a quick stab of the brakes no matter what misbehaviour might be encountered?

Does it matter / do you care what the roll-center does at parking-lot speeds when driving on ice or packed snow? High steering angles could be involved ... but little lateral acceleration due to low grip. Easy for the driver to lose control.

If a degree or three of steer angle, in circumstances that matter, results in a drastic change of what the roll-center does with any sensible steering-axis inclination and caster geometry, I'll be a monkey's uncle. Or your suspension design is fubar. I can't tell.

 

[GregLocock]

I'd have thought roll center migration in roll was of much greater magnitude.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[cibachrome]

For example:

https://imgur.com/AzwaX6Q

 

[ksw100]

What software generated those curves? Thank you!

 

[cibachrome]

These results come from a GM developed suspension nonlinear analysis code (named SNAC). It has since been farmed out as SuspensionSim proxied by Altair. Obviously, I have an ancient version kept alive with a Matlab interface.
Here's a question back to you: Now that you can consider all of these departures from textbook fantasy, what are you able to do with this information? Do you have a parametric sim which can use these inputs as constants, functions, or large values of zero? Can you evaluate what the fraction of load transfer error results from ignoring 'RC Migration'? Are you utilizing the other important aspects of the load transfer mechanism (such as tire Mx), and if so does your tire model reproduce tire test data results ? These considerations are pretty important at peak performance, as are Ackermann, diff, power, steering wheel torque gradient, and driver skill-set. If not, don't follow the rabbit down this hole. Save it for track test days.

 

[cibachrome]

more..

https://imgur.com/B9htvkP

Now what ?