Suspension Design Analysis

[Murtaza]

I'm uploading the analysis of our suspension design in XLSX format. The tow files are archived.

Our design sports the following specs

1400mm track width at front
1350mm track width at rear
1600mm wheelbase
8deg Caster at Front
0deg Caster at rear
5deg Kpi Angle at front and Rear
3.8mm scrub at front and rear

The CG of the car will be approx. 350mm above ground, with the rear heavily loaded apprrox. 65R/35F weight distribution.

I'm hoping that you guys can take a look at it and give your opinions/suggestions.

Analysis1 (http://dl.dropbox.com/u/17445821/Analysis1.csv)
Analysis2 (http://dl.dropbox.com/u/17445821/Analysis2.xlsx)

Thank You.

[Murtaza]

I'm uploading the analysis of our suspension design in XLSX format. The tow files are archived.

Our design sports the following specs

1400mm track width at front
1350mm track width at rear
1600mm wheelbase
8deg Caster at Front
0deg Caster at rear
5deg Kpi Angle at front and Rear
3.8mm scrub at front and rear

The CG of the car will be approx. 350mm above ground, with the rear heavily loaded apprrox. 65R/35F weight distribution.

I'm hoping that you guys can take a look at it and give your opinions/suggestions.

Analysis1 (http://dl.dropbox.com/u/17445821/Analysis1.csv)
Analysis2 (http://dl.dropbox.com/u/17445821/Analysis2.xlsx)

Thank You.

[Chris Craig]

Which tyres are you planning on using????

I'd be very watchful of the weight distribution aswell,i think that may cause you a few problems!!!

[Murtaza]

R12 145/70. The tires are sponsored and the same are compulsory for all teams. We are participating in the Indian event and as the engine (800cc M800) has also been sponsored which weighs at a massive 130+ kg's we're stuck with such a weight distribution.

Also I have a doubt with the Pitch angle. Is it measured from the vehicle wrt the ground or to the Pitch centre? Because in our analysis we have taken with respect to the Pitch Centre. Is it correct?

Any help would be appreciated.

[SNasello]

First off, it looks like you are using optimumK software to analyze your suspension, in which case there are no tire deflections included. When you don't include tire deflection, the pitch angle wrt the pitch centre(or wheel centres) is the same as the pitch angle wrt the ground.

It looks like you are combining pitch and roll/pitch and heave in the same simulation. What is your reasoning for this? Where did you get the inputs for these motions?

When we look at suspension kinematics, we look at heave first, then roll, then roll with steering, which we base off understeer gradient and total roll stiffness.

Developing true combined situations is difficult without track data.


Seeing as your main concerns are with the weight distribution, I would go a bit more extreme on the track width ratio, ie. make the rear a lot smaller than the front to promote more load transfer at the rear. You can fix understeer with a really stiff rear suspension also but you will probably lose overall grip.

[Murtaza]

Yes we're using OptimumK for the analysis.
We were trying to reproduce a turning situation for one of the analysis.

Analysis for the Heave, Roll and Roll-Steer gave us satisfactory results. Will upload them tomorrow.

We're trying to straighten out the weight distribution problem.

Regarding your suggestion : we cannot go much smaller than this track at the rear due to the size of the engine.

This may be too direct a question and a silly one at that too...As we have a very heavy car we were thinking of making the car more stiff, and designing for ride freq. of 1.9Hz rear and 1.7Hz front. Is it ok?

And one last thing. How much is the normal unsprung mass of FSAE cars in terms of % total mass and average absolute mass?

[SNasello]

It is good that you are trying to represent real turning situations, but you should be cautios that the time histories of the angles you are using for input are representative.

For the ride frequencies, 1.9 and 1.7hz aren't really stiff for a formula student car. The wheel rates for your car might end up being fairly stiff however. Due to the relative mass/tire grip for your car, you will be relatively stiff in ride in comparison to what you need in roll.

[Sormaz]

"stuck" with the weight dist.???

sounds like your chassis/integration team isn't of much help!
you get to place your engine and driver don't you?

[Ken_D]

Out of curiosity, is it the V-twin out of the suzuki bike, or the 3-cyl from the maruti m800? 290lbs (130kg) is absolutely massive.

[Murtaza]

It's the 3cyl Maruti 800 engine. And the weight includes the whole powertrain. Exhaust manifold, cylinders, Transmission, differential and all..

When we incorporated anti dive at the front, it changed the intended design by some amount. How do I ensure that doesn't happen, or atleast stays almost the same?

[Thomas MuWe]

First of all:
Thumbs up for posting your work here.

If you include anti - dive to some amount, you have to change your pickup point - normally at the chassis! :-) This often changes your other kinematics - especially your toe curves. So there is not a simple magic fix to that.

Think about that the change of the anti dive will also cause a shift in your pitch axis.

The thing is though that without having sufficient tire data you can just think about have a reasonable amount of camber / toe - change in heave and roll.
Also I think that your ride frequencies are not very high. So maybe an increase in your spring stiffness or a different ratio of your rocker / deflect lever can help.

Unfortunately you did not include your caster moment arm in the front. With such a heavy vehicle it can be pretty hard for the driver to drive the car because he / she needs a lot of power to turn the steering wheel.

Best of luck with your design.

[Murtaza]

These are the analysis for the Roll, Roll+Steer, Heave. Analysis (http://dl.dropbox.com/u/17445821/Analysis.rar)

@Thomas: We haven't yet finalized the Steering system.

ARB's haven't been incorporated yet in the initial design. We may add it later to tweak the suspension.

As we don't have tire data, we think we may settle for a figure of 2.0Hz at the front and 2.15Hz at the rear. Any suggestions on these rates? Our car will weigh around 350kg's with a weight distr. of 60R/40F at best.

[BrandenC]

It's nice to see some actual design work being critiqued here rather than people just asking for answers.

A couple of suggestions:

First, if you don’t plan on including ARB’s in your design it would be a very good idea to look at where your lateral load transfer distribution is compared to your weight distribution. If you end up using arb’s you’ll probably do this anyway.

Second, with respect to the impact of sprung vs unsprung weight it really depends on the surface you’ll be running on. I’ll guess that most teams on 13 inch tires will have nearly 100lbs that are unsprung. Considering most of these cars are between 500-600lbs total (with driver), it’s quite a large percentage of the total mass.

Luckily you don’t have to get everything right to have a fast car or learn a lot about vehicle dynamics. Plan to spend a significant amount of time testing, it’s pretty easy to change springs and see how different wheel rates affect the car.

[Murtaza]

Thank you all for your inputs.

We realized a few anomalies in our design and really stupid and major ones at that. SNasello was kind enough to point them out to us.

[Bobby Doyle]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by SNasello:
Seeing as your main concerns are with the weight distribution, I would go a bit more extreme on the track width ratio, ie. make the rear a lot smaller than the front to promote more load transfer at the rear. You can fix understeer with a really stiff rear suspension also but you will probably lose overall grip. </div></BLOCKQUOTE>

Are you suggesting an extremely rearward distribution of lateral load transfer because of a rearward weight distribution?

[BillCobb]

What's the reasoning behind the 8 deg front caster? Any understeer effect is linear range and that sure isn't going to help the foundation weight distribution oversteer. (it will be "tighty-loose". Besides, the tierod loads and gradient will produce unacceptably high steering efforts. If that's your weight distribution, use two different tire sizes to balance the car. Otherwise you'll need a sewer pipe for a front bar.

[Thomas MuWe]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by BillCobb:
What's the reasoning behind the 8 deg front caster? Any understeer effect is linear range and that sure isn't going to help the foundation weight distribution oversteer. (it will be "tighty-loose". Besides, the tierod loads and gradient will produce unacceptably high steering efforts. </div></BLOCKQUOTE>

It is not the angle which causes the steering effort!

[The_Man]

Firstly I logged on to the forum today after a few months and I can't be more happy to see this thread.

Seems you have put a lot of effort into the restricting roll center movement and done a good job of it. About the frequencies, the problem you might face with such low frequencies (low stiffness) is that your car may tend roll too much, your analysis is for 3 deg of roll and I assume you want to restrict your roll to the value. I personally have found this difficult to do this without having too much jacking. hence I have ended up with 3.0-3.5Hz frequency, which is on the higher side but not entirely unacceptable. Then again, I have usually worked with a lower Roll Center height compared to you. What I am trying to suggest is get an idea of the spring stiffnesses you require to restrict your vehicle roll and see if the frequency value you get is close to the one's you have been targeting, do not be scared of going slightly stiffer.

Best of Luck!

[Adambomb]

First off I'd like to say this looks like a great effort so far. Very refreshing compared to the "tell me how to do this" threads!

Personally I'm a fan of rearward weight distributions and low ride frequencies. Although to do this with the same tires all around, BillCobb is right, you'll need a lot of front bar. Actually just listen to him in general, he is way smarter and more experienced than me anyway!

Also, with lower ride frequencies you'll definitely want to have sway bars, or else you'll need a lot of suspension travel (and therefore end up with a large roll angle) to keep it from bottoming out in cornering. Unfortunately I can't look at your kinematics because the website is blocked from where I'm getting on the internet.

As for sprung/unsprung distributions, I'd our 2010 car is on the extreme end of the spectrum with about 45 kg (100 lbs) unsprung on a 170 kg (373 lbs) car, and we could really only stand to lose maybe a kg out of each corner by switching to centerlocks and CF rims, everything else is pretty much ragged edge with regards to our constraints. Unless of course we go to 10s; I'm a bit "out of the loop" on the design this year being away from school. As for what effect that has on the car, honestly I don't have a really good answer for that, other than the generic thought to imagine what impact a larger unsprung/sprung ratio would have on transients, specifically bump, and how the sprung mass would react to a given input from a larger unsprung mass.

[brettd]

Who says you will need a lot of front bar just because it's rear heavy? We used to run cars with 57:43 that didn't even have front ARBs...

[exFSAE]

Believe front roll stiffness was the concept he was getting at = where yes, all things being equal, as you move CG aft you'll probably need more front load transfer to balance it out.

[BillCobb]

If the car has rear weight bias, designers and developers usualy also have a ride frequency and ride frequency ratio in mind in order to have "flat ride" at some designated speed. This minimizes dynamic body (hence tire) load disturbances during encounters with ride events (i.e. bumps, tar strips, and cross seams). It clearly makes damper tuning more effective, too. That's what what 4 and 7 post testing is all about.

If the chassis is balanced for steady state handling using springs alone and limit handling performance is acceptable, the ride quality (comfort as well as dynamic tire loading) will be disturbing to the car and to the driver. Most rear weight bias cars demonstrate this problem (called "hitching"). The effects are most noticable during heavy braking or accelerating, too. If your butt doesn't care or can't pick it up, fine. If your driver sits up in the seat to imagine his azz dragging during launch, its an issue. It certainly looks funny from the Peanut Gallery, too, and a Judge would see it.

My caster comment still applies. You may think you need that to get some lareral force compliance understeer from the steering system, but steering parts are stiffening springs. It's not that effective at high force levels. Perhaps you would use it to have the driver key into where tire force gradient is maximum, but that's a poor way to do it. Driver training and testing with instruments does that better. Otherwise, steering effort gradients will be too high. This is caster angle I'm discussing, not caster offset.

[Kevin Hayward]

With the usually high steering angles of FSAE cars with quite a lot of Ackermann to improve initial turn in there is usually a general lack of front end grip. This means that the front roll moment distribution is often lower than what you would normally expect in order to get the car working. This means that a rear weight bias car can usually run the same size tyres without running a stiffer front spring or front anti roll bar. There is more than one rear weight bias car with same size tyres with rear anti roll bars and no front anti roll bars.

Another issue tends to be tyre temperature. Most often you are dealing with temperatures on the low end of the operating ideals, even with the sticky rubber. This problem is worse with the front end of the car than the rear. The end result is that within the ranges of rear bias we see in the competition that a FSAE car can be both balanced well for steady state handling and ride on springs alone, when using the same tyre size.

...

However when looking at the tuning of these cars the biggest gains are found focusing on turn in. This is where the different tracks make a big difference. In Oz the track is very twisty and there is nothing that resembles a steady state corner. This rewards a car that changes direction quite quickly. However I remember the Silverdome, where there were some corners of reasonable length where the steady state tuning came into play a lot more. I think that one of the reasons we have not seen real decreases in the non-aero times of the skidpan (even with much better rubber) is that teams are running more Ackermann than they used to, and not adjusting it (or being able to adjust it) for the skidpan.

For the original poster I recommend having a car built early, a big box of springs, and the ability to tune with roll bars. Above all things do not underestimeate the effect of dampers on the vehicle (including on the skidpan). You can dramatically alter the behaviour of the car through intelligent damper tuning even in steady state situations. Also do not disregard damping effects on heave, roll and pitch frequencies.

Kev

[buggaero]

A 1600 on that weight bias approaches the classic P911 scenario. Do consider your yaw and is it still going to be a "race" ??

The numbers you ran were good, but they can be equally fatal considering the corner speeds on that track where you wouldn't see slaloms or hairpins but straight stretches apart from skid pad.