Approximating neutral steady state - without tire data [AWD]
Approximating neutral steady state - without tire data [AWD]
(OP)
Hello everyone,
[first post, so here goes!] We are currently designing the suspension for a new AWD vehicle, used for gravel-only [rallycross]. Think 550bhp 1000kg, very much like FIA WRX supercar.
I've spent quite some time reading up on websites like this, as well as RCVD etc. but currently find myself lost in details and need some "higher-level pointers" that help me understand what steps to take to arrive at some sensible goals.
My question is actually twofold:
A - is it possible for us to ballpark some spring/ARB values that approach a neutral handling car in steady state cornering, without having tire data available?
We will be using the same tire/rim combo on all 4 corners. We fully appreciate the need and use of testing to arrive at a desired setup, but the thing I'd like to avoid is starting out with a car that needs to be patched up because of some fundamental issues.
MAIN concern at this point is questioning integrating ARB's at either front or rear, or none at all
B - perhaps people with some experience with related cars would like to comment on some of our design parameters below? Would be greatly appreciated!
A - stuck in design process
The entire car is modeled out in 3D CAD, including weights.
AWD car, double wishbone on all 4 corners. Direct actuated spring/damper unit on lower control arm.
The suspension kinematics have been engineered iteratively using base Solidworks models combined with excel / VSUSP to arrive at what I think we want. Rates/frequencies/roll stiffness/load transfer have been evaluated using excel sheets (RCVD formula). So basicaly input spring/ARB choices and evaluate the outputs.
The steps I took so far:
-select suspension kinematics using models, to arrive at desired values
-extract sprung/unsprung weights and CoG locations from model
-Total weight distribution [56% front]
-select/trial spring rates to arrive at supension frequencies [1.74hz F / 1.79hz R]
-select/trial ARB and include tire vertical stiffness estimate to arrive at ride rate and total roll rate/stiffness
-roll stiffness distribution [60% front]
-use CoG vs neutral roll axis to arrive at load transfer F/R and TLLTD [58% front]
-calculate roll sensitivity [2.3 deg/G]
-input estimated max latG, verify:
--max steady state roll
--wheel travel used, travel reserve to bump stops for track irregularities
--actual wheel loads
This is where I'm stuck. Is it possible without tire data to say something usefull about steady state cornering balance?
B - car data, feel free to comment!
-AWD gravel, rallycross
-1020kg inc driver/fluids
-double wishbone all corners
-direct acting spring/shock on lower control arm, linear [Motion Ratio 1.49 to 1.51]
-uniball joints on every pickup, pickup points triangulated to feed forces into tube spaceframe (for compliance red.)
-total weight distribution [56% fr]
-design max latG: 1.3G
-wheelbase 2350mm
-tracks 1720/1720mm
-tire dia 640mm
-RCH 98mm fr / 118mm rr (1mm vertical change in roll, no gradient change in roll)
-roll sensitivity [2.3 deg/G]
(-tire vertical stiffness assumption: 300N/mm)
-susp frequencies [1.74hz F / 1.79hz R]
-currently small front ARB included
-roll stiffness distribution [60% fr]
-TLLTD [58% fr]
-jounce travel 140mm
-rebound travel 100mm
-static camber -1* fr / -1.5rr
-approx 50% camber recovery fr+rr
-caster -7* fr / 0* rr
-KPI 9*
-anti's: design includes multiple inboard pickups, one setting for 0% anti's all round, will be including pickups to accomodate approx 40% antidive and 20% antisquat.
-roll angle @ 1.3G: 3.0deg
-chassis stiffness not evaluated yet, but read somewhere proper design goal is >10x total roll stiffness
Many thanks for anyone commenting!!
regards,
Tom
[ps non native english speaker, hope everything makes sense!]
[first post, so here goes!] We are currently designing the suspension for a new AWD vehicle, used for gravel-only [rallycross]. Think 550bhp 1000kg, very much like FIA WRX supercar.
I've spent quite some time reading up on websites like this, as well as RCVD etc. but currently find myself lost in details and need some "higher-level pointers" that help me understand what steps to take to arrive at some sensible goals.
My question is actually twofold:
A - is it possible for us to ballpark some spring/ARB values that approach a neutral handling car in steady state cornering, without having tire data available?
We will be using the same tire/rim combo on all 4 corners. We fully appreciate the need and use of testing to arrive at a desired setup, but the thing I'd like to avoid is starting out with a car that needs to be patched up because of some fundamental issues.
MAIN concern at this point is questioning integrating ARB's at either front or rear, or none at all
B - perhaps people with some experience with related cars would like to comment on some of our design parameters below? Would be greatly appreciated!
A - stuck in design process
The entire car is modeled out in 3D CAD, including weights.
AWD car, double wishbone on all 4 corners. Direct actuated spring/damper unit on lower control arm.
The suspension kinematics have been engineered iteratively using base Solidworks models combined with excel / VSUSP to arrive at what I think we want. Rates/frequencies/roll stiffness/load transfer have been evaluated using excel sheets (RCVD formula). So basicaly input spring/ARB choices and evaluate the outputs.
The steps I took so far:
-select suspension kinematics using models, to arrive at desired values
-extract sprung/unsprung weights and CoG locations from model
-Total weight distribution [56% front]
-select/trial spring rates to arrive at supension frequencies [1.74hz F / 1.79hz R]
-select/trial ARB and include tire vertical stiffness estimate to arrive at ride rate and total roll rate/stiffness
-roll stiffness distribution [60% front]
-use CoG vs neutral roll axis to arrive at load transfer F/R and TLLTD [58% front]
-calculate roll sensitivity [2.3 deg/G]
-input estimated max latG, verify:
--max steady state roll
--wheel travel used, travel reserve to bump stops for track irregularities
--actual wheel loads
This is where I'm stuck. Is it possible without tire data to say something usefull about steady state cornering balance?
B - car data, feel free to comment!
-AWD gravel, rallycross
-1020kg inc driver/fluids
-double wishbone all corners
-direct acting spring/shock on lower control arm, linear [Motion Ratio 1.49 to 1.51]
-uniball joints on every pickup, pickup points triangulated to feed forces into tube spaceframe (for compliance red.)
-total weight distribution [56% fr]
-design max latG: 1.3G
-wheelbase 2350mm
-tracks 1720/1720mm
-tire dia 640mm
-RCH 98mm fr / 118mm rr (1mm vertical change in roll, no gradient change in roll)
-roll sensitivity [2.3 deg/G]
(-tire vertical stiffness assumption: 300N/mm)
-susp frequencies [1.74hz F / 1.79hz R]
-currently small front ARB included
-roll stiffness distribution [60% fr]
-TLLTD [58% fr]
-jounce travel 140mm
-rebound travel 100mm
-static camber -1* fr / -1.5rr
-approx 50% camber recovery fr+rr
-caster -7* fr / 0* rr
-KPI 9*
-anti's: design includes multiple inboard pickups, one setting for 0% anti's all round, will be including pickups to accomodate approx 40% antidive and 20% antisquat.
-roll angle @ 1.3G: 3.0deg
-chassis stiffness not evaluated yet, but read somewhere proper design goal is >10x total roll stiffness
Many thanks for anyone commenting!!
regards,
Tom
[ps non native english speaker, hope everything makes sense!]
Talk To Other Members
RE: Approximating neutral steady state - without tire data [AWD]
From what you have already defined and described, I'm guessing that you currently have a vehicle with a substantial amount of 'understeer' based on wgt distr. and tlltd (assuming some sort of standard tire section/mass ratio) and the understeer function climbs as Ay increases. That's 'neutral' if you mean the understeer is increasing linearly, or if you could manage to keep a linear range understeer to be constant as a function of Ay, that's another 'neutral' meaning, and then we have a 'Neutral steer' category to describe a case for which the front and rear axle slip angle gradients with Ay are equal so that the linear range Understeer Gradient is Zero and stays there until something dramatically happens (as in your 550 hp wakes up and/or your drifter driver drifts off to drifting driving. If you catch my drift...). I use the cornering compliance (sideslip gradients) terminology because THEY fully describe the dynamic properties as well as the steady states AND we can tell which end of the car is not up for the challenge. 'Understeer' only tells us that one end of the car may be out of spec but doesn't identify which end.
So you now need to add the Big Science picture to your Controls synthesis framework ! You can't ignore the dynamics, either, because your 'neutral' car could tax the driver's courage as well as ability(s).
PS: What's the tire size ?
RE: Approximating neutral steady state - without tire data [AWD]
It's pretty linear because them there tars are bodaciously stout on the light loads. Response times for such very low understeer (and make that CONSTANT understeering) are acceptable. And you can get to a desired max lat value with you fingers crossed. I have no powertrain shenanigans in here so how you dial that in will make a fine difference. This in not gravel, so you get a 'tighty-loose' car. If you are game, I can post a simplified version of this tire for you to play with in Excel or Simulink that does about 90+% of what my tire model entails. It include the aligning moment and overturning moment and camber elements, too. (which you can NOT ignore in you present state of drifting.
RE: Approximating neutral steady state - without tire data [AWD]
All the (excellent) tire modelling discussion aside...
Is this a real car you're actually going to build?
If your question really is 'should I plan for anti-roll bars' the answer is almost undoubtedly yes. Even if you obtain the 'neutral' behavior you're after in whatever context you mean, how will you tune without them? Are you going to carry around a bunch of springs and take the whole car apart when the driver requests 5% less roll stiffness in the front? Are you going to rely on damping and tire pressure changes alone, and just deal with the giant limitations that will provide you?
Even if, during your analysis, you determine that according to simulation one end doesn't need (or want) an ARB, you will do yourself a massive favor by planning for one (chassis/upright hardpoints, packaging, etc) in the event that the real world doesn't match what you determined on paper. Which is likely to happen if you're racing on gravel.
RE: Approximating neutral steady state - without tire data [AWD]
Sure, make it to need a front bar (easily adjustable blade type) would be advantageous until you figure out the balance that works.
RE: Approximating neutral steady state - without tire data [AWD]
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?
RE: Approximating neutral steady state - without tire data [AWD]
Are you driving on gravel or tarmac? If gravel, your stated max lateral acceleration (1.3 g) is not going to happen unless you have jet propulsion with the jet engine applying thrust downwards ... or sideways.
We don't pay attention to ride frequencies. We do pay attention to keeping the suspension off the bump stops. Gravel implies bumpy unimproved roads. That implies needing lots of suspension travel if you want to keep the tires in contact with the ground and not skip off the tops of all the bumps. That implies soft spring rates and compliant dampers and as little antiroll bar as you can. Let it roll. 2.3 degrees per G looks way too stiff. I'd guess three or four times as much for a gravel setup.
The counterpoint to soft springing is how you deal with landings after "yumps". Do you have to deal with that?
Look at the desert racing "trophy trucks" for this situation taken to the extreme. Lots of suspension travel, very soft springing, lots of roll in corners.