Practical guidelines for suspension design for beginners 2

[roller_delawer]

Hi all,

I am a mechanical engineering currently working on Interior Trim parts for a living.

However I've always wanted to learn about vehicle dynamics and suspension design with the idea of making my own kit car (or at least try to do so). So I've read a few books about the topic (the famous Milliken and Milliken and others) but I've always found that there's a gap between understanding camber/caster/scrub radius etc and the proper method to correctly approach the design of a suspension with some specific handling goal in mind.

Reading this forum, most of the question look beyond that gap for me: even if I understand the technical terms, I struggle to understand the questions and the way you are designing the suspension in order to put those questions and answers in context.

Is there any book, online training, or something like that which would help me to learn that practical approach of this science? How did you started in this world? What else do you recommend me?


Thank you very much!!


EDIT: I know that there are several titles of practical suspension design aimed at hobbyist making a "locost" kit car in their garages. But due to my engineering background and the idea of this being a possible realistic career path for me, I am looking for something more technical than that.

 

[cibachrome]

Suspensions connect the forces and moments from the tires to the chassis/body based on cammanded moments from the driver. The entire vehicle operates at the whim and comfort of the operator/driver on a roadway. So, given a set of desired ride and handling response performance targets or goals, a combination of tire and suspension properties are selected as subsystem parameters. Given what's available and affordable as a front and rear set of tires, suspensions (with steering on one end or both) are formulated based on cost, mass, durability, manufacturability, repairability, crash-worthiness and cool-factor (lore of what sells well).

Often we see this flow down process seriously mis-played by all the hype and hoopla from sandbox 'professionals' who build a car around a suspension type and bandaid the tires and the masses to try to make it work. Some more successful than others.

What you need to start with is a selection of what performance targets/goals you think you need or can live with. This sometimes takes 'competitive assessment'. You measure the steady state gains, response times, bandwidth, linearity, max values etc of vehicles you like and don't like. In my case, I have done this so much that I can write down the goals for just about any car or truck you wish because of all the testing and characterizing ("finger printing") done thru my career. Then select the performance you think you want and try to put it together with a parametric simulation. If a solution can be found (close or exact), you start scrounging around for hardware that could deliver the goods. Hardware that's cost effective, geometrically friendly, and utilizes the features of the tires, roadway and driver in a comfortable manner. The parametric simulation tells you 'why' you meet the goals (or why you don't). As opposed to a multi-body simulation which tells you that what you build could do the job (or fails). Much of this depends on the qualifications of the driver(s) too. My mother though good handling meant "easy to park". For my father it was 0=80 mph time with good launch.

Imagine listening to radio or music from a source (CD, cassette tapes, vinyl or shellac records) thru an amplifier (1, 10,100, 1000 watts) 1 channel, stereo, 4 channels, 16 channels) with PA speakers, Bose, AR, EPI or earphones) and BTW: your listener is stone deaf. All they REALLY want to listen to is a news broadcast but are sold on Pink Floyd. What items would you choose? Pick the speaker wires first ? It's the same for vehicle dynamics: the physics is easy. Some NEVER get it right. Some can get it awesome on the first try.

 

[BrianE22]

Did you look in the FAQs for this forum?

 

[SwinnyGG]

If you want a very basic intro to vehicle dynamics, the Carroll Smith books- Engineer to Win, Tune to Win, etc are pretty decent at explaining things at a 30,000 foot level.

RCVD would give you the deep dive into the math.

 

[BUGGAR]

Learn and master the secrets and myths of the "Roll Center".

 

[GregLocock]

How I got started was designing the suspension for a solar car. Rather excitingly this was a rigid frame connecting the three wheels, the body was suspended off the frame (Lotus twin chassis, in retrospect). The reason this was selected was that wheel alignment was paramount since rolling resistance was one of the causes of drag. The front wheel was on a bicycle fork (in effect). On its first outing the drivers complained that it was too twitchy, so I reluctantly modified the jig the forks were built on to give more trail, and dropped the original forks in. I was about to attack them with hacksaw and welder, when I thought, hmm, I wonder if they'd bend? To my astonishment, and anger, the answer was yes, I could bend the forks with my bare hands. So, I explained this to the team, and said we could test the new geometry (which worked) but we'd have to check for bending after each significant event. Then I took on the job of designing a spaceframe style front fork that would fit inside the wheelspat and was stronger. That worked. Then we got on with tuning the coilover springs (easy we didn't really need anything other than a soft ride) and the shocks (I had to soften them off and in the absence of any guidance just made the holes bigger in the pistons) - my first and last attempt at building a shock. Finally we had to set the ride heights so that the aero body was at zero pitch, that was worth about 20% on power required at cruise speed.

Meanwhile Bob bought every single tire he could find and rolled them along a carpark with a drum to give them the right weight. From this he worked out which tire had the lowest RR (a cheap BMX with natural rubber inner tubes, as it happened).

Solar car is a simple project because the goals are easy to define, whereas a production car has to do many things right.



Cheers

Greg Locock


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

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

 

[GregLocock]

Oh, and another vote for Carroll Smith. Except for his theory about roll axes.

Cheers

Greg Locock


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

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

 

[SwinnyGG]

Except for his theory about roll axes.

Is this a reference to his book stating that the roll axis can be accurately produced by connecting front and rear roll centers with a straight line?

I learned this one the hard way, and lost one of those "young engineer who thinks he's right but isn't" arguments we were all part of when young and educated in name only.

My understanding is that he later acknowledged that this concept was incorrect, but I'm not sure if it was known at the time he first wrote it in a book.

 

[BUGGAR]

The "roll center" grasshopper.

 

[GregLocock]

jgKRI -yes. It's wrong from basic physics, it didn't have to be 'found out'.

here's the solar car

Cheers

Greg Locock


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

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

 

[BUGGAR]

Find how to determine the torque in a four bar linkage and you can "snatch the stone from my hands". For now, forget about the "roll center" concept. It's misleading and poorly named and won't get you what you want.
"Combining" the front and rear roll stiffness' will be covered later. My bill will follow, too.

 

[roller_delawer]

Hi all,

Thank you for all your answers, some of them are really interesting.

Regarding the FAQ, I didn't know there was a specific FAQ for this forum, but I've taken a look, and it looks really interesting. I will read it more in depth.

Regarding my initial question, some people recommended the RCVD, as I said initially, I've read this book as well as Allan Staniforth's "competition car suspension", and others. And, while being completely mandatory for this topic, I am looking for something that goes one step further and really helps me to start with the hands-on work with some real world figures, graphs or advices of how to apply all that theoretical knowledge into a real design.

I could plot all the different parameters for a suspension, and even if I understand more or less what they mean, I would be lost when trying to set a specific value as target, or when making compromises, I'd be completely lost trying to understand which one I should prioritize.

Is there anything out there that helps me with these things?

I haven't read any of Carrol Smith's books, might that be more similar to what I'm looking for?

 

[roller_delawer]

What you need to start with is a selection of what performance targets/goals you think you need or can live with. This sometimes takes 'competitive assessment'. You measure the steady state gains, response times, bandwidth, linearity, max values etc of vehicles you like and don't like. In my case, I have done this so much that I can write down the goals for just about any car or truck you wish because of all the testing and characterizing ("finger printing") done thru my career. Then select the performance you think you want and try to put it together with a parametric simulation. If a solution can be found (close or exact), you start scrounging around for hardware that could deliver the goods.

Hi Cibachrome,

You've just nailed. But that's precisely my problem, theoretical books don't help much when you're trying to set up real-world goals, or making real-world compromises.

Because of my experience as an interior trim engineer, I can easily understand what I need to deliver to fulfill my customer needs. And when things get complex and I have to give up something, I know how that's gonna affect the final result. However, with suspension design, I'm totally lost.

And this is kind of a vicious circle, I don't start creating my first suspension cause I'm not sure what to do, and I won't be sure what to do until I have designed a couple of them...

Thank you again!

 

[BUGGAR]

Suggestion?:
Go out in the garage and build a half scale model. Make the parts adjustable (I used all thread rod and carburetor rod ends). Buy some springs (use tension springs reversed to simulate compression springs). Set the spring mounts so you can get progressive spring action as well. Set it up on a vertical board with a track along the bottom for your "wheels". Hang a weight on a string from your adjustable center of gravity and yank the string sideways, and up and down, and play with it until your significant other yells, "When are you going to stop dancing with that thing and come in here and help me?"

 

[BrianPetersen]

You haven't told us what sort of vehicle you are attempting to build.

Upper and lower wishbones with very high spring and damping rates work well for a Formula 1 car or other roadracing car designed for use on smooth pavement. Doesn't work well for rock crawling.

Beam axle located by linkages with links as long as practical and very soft long-travel springs and dampers work well for rock-crawling. Doesn't work well for roadracing.

When you start wanting high grip and decent ride quality and good steering behaviour and decent stability at high speed is when it starts getting tough.

 

[roller_delawer]

You haven't told us what sort of vehicle you are attempting to build.

Hi Brian,

My idea is to design a track-oriented double wishbone suspension for an existing car shell, for a restomod project.

But, as I said, being an engineer and working for a car manufacturer already, I can see some kind of career path possibilities here. So I would like to learn as much as possible.

 

[BUGGAR]

Something like Heidts, et. al. is doing?

 

[BrianPetersen]

OK. Pavement. Roadracing. You need good grip and high-speed stability, decent grip feedback through the steering (this one is tough), you will be using modern high performance tires, you need enough space around the spindles for good brakes, and ride quality doesn't matter. Because you can use relatively stiff spring and damper settings, antidive doesn't matter at the front (use the springs), roll centers aren't as meaningful as some people make them out to be (DON'T let them be high, use low single-digits inches above ground level, use the springs and antiroll bars for the rest). Because you can use relatively stiff spring and damper settings, you don't need much camber gain with bump travel ... use the springs and antiroll bars to keep body roll under control. Get the idea? Another thing ... bump steer is bad.

Now, how much space do you have to work with? The availability and location of structural attachment points to the vehicle will be a BIG governing factor in what type of general layout you use.