Miller 3 Wheeler: Suspension Design

Miller 3 Wheeler: Suspension Design Part 1: 5-12-2014

Suspension Development


Suspension design can get really deep, really quick. I will attempt to document this in a clear understandable way. I’m not an expert by any means on this, but I know enough to be dangerous! I know how not to have a suspension set up, I learned that the hard way back in college. I was designing the front suspension for the Formula SAE race car project for our senior design class. Knowing basically nothing about suspension design, and not understanding much on the reading I had done on it, I went to town and designed the suspension, both front and back. Then we did the fabrication, and eventually got the car driveable. It was only later during testing at an autocross event that we understood how poor the suspension was. I had designed the wheels to stay perfectly parallel when they moved up and down, which kept the same camber all the time. Ok for bumps, but super bad when it comes time to corner. The outside wheel tipped out when hard into a corner, and the car just pushed in a corner instead of turning properly.

What you want instead is some camber already in the wheel, and then some increasing camber as you go into a corner and the weight transfers. Camber is how much the tires tip in/out when you look at the car from the front or the rear. Typical is that the top of the tires tip in towards the center.

Several years later, I was back in the Formula SAE realm, only this time as an adviser for the team.  After 3 years of advising, and several very good handling cars, I had a much better understanding of suspension development.  Continuous learning is so important!!!! Its not so hard to eat humble pie when you can learn from it!

This project gives me the opportunity to learn further, and apply what I’ve already learned as I get a deeper understanding of this.  Now, this vehicle will not be an autocross star, but there’s no reason that it can’t be a good-handling, comfortable vehicle, and implement as many good handling characteristics as I can.

As I’m starting on this, I did some more research to refresh my knowledge, and found some nice links:

Link to a Formula SAE powerpoint by a design judge which has some really good diagrams and explanations.  Worth the time to read through.  I’m going to be referring to that many times.

Another good one is a long forum post on suspension design: Race Car Suspension Class is the post title.

I looked a buying a suspension analyzer program, which I may still do, but decided to try to determine as much as I can in the CAD model.  A little more time consuming, but doable.

Based on the research information, the first step is to start with the front view of the suspension, and work on determining the locations of the ball links, which determines King Pin Inclination (KPI), Scrub Radius and Spindle Length.  From there you can determine the Instantaneous Roll Center, which then helps you pick the lengths of the A-arms. I ended up making two sets of models, one that allows me to rotate the body, and another that allows me to move it up and down.  Trying to make those together as one model can make it way too complicated and too hard to decipher the results.

The Rotating Model

By rotating the body, it simulates body roll during a corner, so if you are making a right hand turn, the body will roll to the left, and transfer most of the weight to the left front tire.  Camber during this is important, and what I messed up many years ago.  If that tire tips out too much, it looses traction and the car will push.

fv-susp-skel-rot leftfv-susp-skel-rot flatfv-susp-skel-rot right

Inside corner wheel               Neutral position                       Outside corner wheel

I don’t have these quite where I want them yet, but it’s pretty close. Here’s the up/down model:

fv-susp-skel-vrt middle fv-susp-skel-vrt down fv-susp-skel-vrt up

Neutral Ride Height                 Full Down                                   Full Up

After the front view suspension is determined, then you can move onto the side view and work on caster. Caster can sometimes be harder to understand, but the easiest way is to picture a grocery cart caster wheel.  The steering point (pivot point) is in front of the contact point of the wheel, producing mechanical trail.  This allows the wheel to follow nicely and go in a straight line.  This is Positive Caster.

Negative Caster is not desired, and will produce a vehicle that is not stable!!!!!  Race cars have very small amounts of positive caster so they are close to the unstable boundary, but make them much easier to turn!  Dragsters have lots of positive caster so they naturally go straight, but can be difficult to steer into a corner.  We are looking for a nice balance of cornering and going straight. My research shows that the caster steering angle is typically inclined back from vertical between 2 and 7 degrees.  I chose 5 degrees of caster, and for the size of my wheels/tires, this produces ~1.2″ of mechanical trail. Caster also influences how much camber is added or subtracted during a turn, so we need to combine the Front View camber model, and the caster model to see how the two affect each other. So here are the views with the body in a neutral position, and then inside corner and outside corner, you can compare the camber in each by looking at the top view. fv-sv-tv neutral fv-sv-tv neutral 2

So the camber looks pretty decent there, let’s see how it looks if the body is rolling 5 degrees: fv-sv-tv rot2 fv-sv-tv rot Almost perfectly straight up and down, so still pretty close to what we want. So now let’s see what happens if we’re under braking in a corner, no rotation though: fv-sv-tv braking2 fv-sv-tv braking Still have favorable camber on the tires. Once last scenario, acceleration, not nearly as critical as weight has transferred, but still worth looking at:

fv-sv-tv accel 2 fv-sv-tv accel Now that those areas are looking good, we can start to determine the steering details, like Ackerman vs Anti-Ackerman.  Can also start to think about shock & spring placement, I’m leaning towards a remote placement with a bellcrank and pushrods (like an Indy car!!). More to come later….