Double Wishbone suspension - Magic Triangle????

[michaelplogue]

Up front, I want to say that I'm a total noob with this stuff, and that I'm not an engineer, mechanic, or have any experience in this subject. That being said: I'm in the process of designing a trike which I've started discussing in this thread:

http://www.exocars.net/showthread.php?t=4211

As mentioned there, my "Plan-A" is to use a standard old-style, VW Beetle front axle beam. "Plan-B" would be to go with possibly a Mustang II front end - depending on local availability. "Plan-C" would be to use a double wishbone setup. Having done some research on the web (from forums like this one and others), I have come to the realization that these systems are much more complex than they appear. So I've been doing some reading up on the subject, and have been playing around with some computer models to see what I could come up with - with my limited knowledge and experience.

This is where hopefully all you guys come in. I seem to have stumbled across a specific geometry that - at first glance - would seem to simplify things quite a bit. It's all based on an isosceles triangle, with the base being the ball center-to-center height of your vertical uprights, and the apex having an angle of 22 degrees. By superimposing this triangle on the frame of your vehicle, any point along the long 'legs' of the isosceles triangle are eligible pivot points for the base of the a-arms. It appears to work for both equal and un-equal length arm setups. Obviously, the shorter the arms, the less body roll it would be able to handle and still keep both tires flat on the road.

This first illustration shows a setup with equal length arms, and vertical uprights.



The black boxes are obviously the tires, the red line is the vertical upright and the base of the isosceles, the green lines are the a-arms, the thin blue lines are the long legs of the isosceles triangle (meeting at the apex at a 22 degree angle), and the corners of the blue box representing the pivot points for the A-arm base.

As you can see in the lower picture - which illustrates the 'neutral position' - the a-arms follow along the path of the triangle legs. The upper image shows the maximum body roll that can be achieved while keeping both tires flat on the ground.

The following shots show the same setup using shorter and longer, equal length arms with the maximum allowable roll.






As mentioned before, this also works with unequal length arms. The simple rule is that the a-arm pivot point must be along the legs of the isosceles triangle.

This one illustrates a shorter, upper arm.



.... And this one with longer upper arms.




(An interesting observation regarding un-equal arms: By having the top arm shorter, it appears that the center of gravity will rise upwards, whereas, shorter lower arms actually brings the center of gravity downwards).

This will even work with canted, vertical uprights. Same rules apply: Base of the triangle in-line with the vertical uprights, a-arm pivot points anywhere along the long legs of the triangle.



I've tried this with all kinds of arm lengths and configurations, and it seems to pan out every time - with varying degrees of maximum roll.

Now, my further research showed that this 'flat tire' configuration is not optimal, as it does not allow for the lateral 'roll' of the tire when making a turn. To take this into account, most setups have the wheels sitting with a negative camber while in the neutral position. This can be easily achieved with this system by setting it up as described above, and then shortening your coil/shocks to bring the car body down, resulting in the tops of both wheels tilting inwards for the negative camber. This way, the tire on the outside of the turn will have maximum grip with the road - taking advantage of the lateral tire roll.




Now, this doesn't necessarily take everything into account, such as tilting the assembly 'forward' to obtain an anti-dive configuration. However, it's seems to me that it would still work - in theory at least.


So.... Somebody please tell me I'm a whack-job, and that I shouldn't quit my day job. That way I can quit wasting my time on this, and go back to Plan-A or Plan-B.


........ Or you can tell me I'm a super-genius and pump up my self esteem....

.

[golftdibrad]

I'm having difficulty understanding your models, i'm used to animations or curves in excel :p

Remember to consider you case in bump as well....

Some general 'guidelines' of thumb when designing your unequal length double wishbone suspension:

A little negative camber gain is good
Long links are good
bottom link longer than the top is typical
roll center should be above the ground but below the CG for typical designs. As the car rolls it shouldn't stray from its lateral position (parallel to the ground) more than a few inches.

Get Carrol smiths books on racing cars. They are well worth the money.

[michaelplogue]

Here's a quick and dirty animation that might illustrate it better.

http://www.youtube.com/watch?v=V37DSq6Xa2k

[golftdibrad]

I dont think your roll center is in a happy place.

[cordycord]

Listen to Brad--and Carrol Smith. Shorter upper a-arms--when placed horizontally or when the frame link is below the (upper) ball joint--will induce negative camber during suspension compression. A favorite lower/upper ratio seems to be 3:2, not counting other variables.

Also Google the term "slip angle", and prepare to wad up all of your suspension diagrams.

[cheapracer]

Originally Posted by golftdibrad

I
Some general 'guidelines' of thumb when designing your unequal length double wishbone suspension:

1/ A little negative camber gain is good
2/ Long links are good
3/ bottom link longer than the top is typical
4/ roll center should be above the ground but below the CG for typical designs. 5/ As the car rolls it shouldn't stray from its lateral position (parallel to the ground) more than a few inches.

Get Carrol smiths books on racing cars. They are well worth the money.

No electricity at the factory today so I have some time.

1/ Good for what? It's no good for braking and you can only use as much or need as much gain as the body rolls in unison and thats just for the rear, at the front we have KPI and caster also to add into the equation. The only foundation your statement has is to oppose saying that positive camber gain is good which for most examples of course it's not. Interesting to note that Formula Vee is one of the fastest cornering non aero racing cars out there yet have no front camber gain at all other than what they can produce by altering caster.

2/ Nope, wrong. Not that I blame you it's posted around the net so much who wouldn't believe it? Want examples? easy - for our lot lets take the Ariel Atom, lovely long arms, probably helps to sell the car - now go and speak to the people who actually drive them hard and find out they have to add around 3+ degrees of static camber to avoid positive camber and present a flat tyre surface while cornering. Long arms upper and lower arms do not allow sufficient camber change and are only suitable for ground effect F1 cars that need to stay flat at all times but look back in history when F1 cars relied on mechanical grip and what do you find? Yup, short arms.
See Atom picture below generating positive camber through the top arms not arcing inwards enough (because they are too long) to draw the top of the wheel inwards to achieve ideal neg camber. Note that Atoms generate around 0.9 cornering g's, quite low for a 600kg car. You can read this and similar as well - http://forum.atomclub.com/index.php?topic=9555.0

3/ I'll get back to that with Cord's post.

4/ The theoretical RC (TRC) is a most misunderstood concept in many forums and something that may surprise, most car suspension designers for the major car companies pay little regard to it because it's not real. It is a useful start tool to visulise some initial basics but once the car starts to drive down any normal road the theory RC goes out the window and the practical RC (PRC) comes into play but that is also only one (at each end) of the factors.

Simple explanations - For example, where is the TRC for a motorcycle? Of course it's at the point that the tyres touch the ground, let go of the bike and the bike will fall, but as it falls it rolls around the ever moving PRC, the TRC is still at the bottom of the tyres but you can see as the bike falls the PRC moves to the sidewalls.

Example 2, FR reverse trike 3 wheeler, the TRC is at the contact patch of the rear tyre but as it leans in a corner the tyre leans up on it's edge - the PRC is right at that edge.

Now for 4 wheels we have too many effects to consider for this short day but I hope you can begin to imagine some of the effects - roll bars for example put pressure on the outside tyre therefore the PRC is migrating in that direction (the car is starting to roll more around the outside tyre than the inside) to the point where the car goes up onto 2 wheels, not hard to see that the PRC is then at the outside tyre contact patch, there is no other point for the car to roll around even though the TRC is still somewhere close to where the paper calculations put it.
Note again the Atom below, it's TRC is pretty close to the ground but you can see that lateral force on the higher CG has lifted the PRC higher and towards the outside tyre which has compressed and the inside tyre is looking very light.

I suggest strongly to people who want to play with theoretical RC's is to forget them and play with and understand the effects of the roll centerline, thats the line you draw between the front RC and the rear RC, that is far more useful to understand. Think about where and how the CG will fall over that line and it's effects on corner to corner weights etc.

[cheapracer]

Originally Posted by cordycord

Shorter upper a-arms--when placed horizontally or when the frame link is below the (upper) ball joint--will induce negative camber during suspension compression. A favorite lower/upper ratio seems to be 3:2, not counting other variables.

Firstly there are a number of race and road cars that use an inclined top arm of near equal length but the inclination, usually as severe as 30 degrees, arcs inwards far faster than the level lower arm. A surprising number of F1 cars used this before ground effects. This 60's Ferrari uses a combination of both inclination and length difference and is a good shot for people to understand how the top of the wheel is being pulled in to gain camber, notice the body roll but the tyre is near perfect to the road...... http://www.conceptcarz.com/view/phot...-F1_photo.aspx

I guess every new builder has to start somewhere and I guess if you averaged it out many arms parrallel probably end up somewhere near your 3:2 ratio and it's not the worst advice one could give but it is a fallacy and more of an effect because thats the way it is in the Locost book which has so much effect on so many cars built. I guess most of us have seen "I'm using the same measurments as a Locost for my arms"...

Firstly understand 'Scrub' is the sideways/lateral movement of a tyre during suspension travel, a little scrub really isn't a big drama.

Now lets imagine a static lower balljoint and as the wheel goes up and you get camber gain as the top of the wheel goes in BUT the wheel is pivoting on the lower joint so the tyre goes outwards at the bottom at the same time - scrub.

This change depends on just where the lower ball joint is located and fbecause of road car ground clearance many OEM uprights in use for kit cars are quite high therefore if you use a very long arm that has very little lateral movement in it's stroke you can end up with large lateral scrub at the contact patch. Racing cars with custom uprights try to get the lower joint as low as possible to help eliminate this. Some racing cars even have the joint inside of the tyre and mere mm's away from the road surface to have camber gain with low lateral scrub. Example - http://www.conceptcarz.com/view/phot...12F_photo.aspx

Yes these are old racing cars but those cars are totally reliant on mechanical grip, no aero.

But for our street/track cars with OEM uprights a shorter lower arm is a better bet so that the scrub can be controlled from becoming excessive after initial camber gain.

note; That is a shorter lower arm (and top arm) than something like on the Atom, not a lower arm shorter than the top arm.

If you don't understand I will draw some qikpics to help.

[golftdibrad]

cheapracer:

I suggest you spend more time addressing the topic of the op vs tearing down my arguments and contradicting yourself.

I am not an expert on suspension kinematics, I did a school project on a couple of cars so I know more than most. I also know that suspension design is all about COMPROMISE in what you want it to do and how you want it to preform. I'm not looking to re-write the Carol Smith books on the internet.

My suggestion to the op remains, get this: http://www.pegasusautoracing.com/pro...oduct=TO%20WIN

or at the very least just Engineer to Win by Carroll Smith.