question on chassis stiffness

[Frank]

im after some advise about chassis stiffness

perhaps the answers are in a book i haven't read or understood properly

could someone please give me some guidance?

just thinking about a chassis stiffness for a minute

if the value was very very low...

what happens?

i think it means rolling moment resistance will match the F:R weight distribution of the car

a front heavy car will oversteer
a rear heavy car will understeer

and it will be impossible to "adjust" your rolling moment distribution

your stuck with the one behaviour

i suspect its time i mathematically analyse the chassis torsional stiffness that is required (preferable/essential?)

im thinking of writing a simulink sheet

i suspect quite a few of you have done this before?

cheers in advance

Frank

[This message was edited by Frank on March 13, 2004 at 08:01 AM.]

[Frank]

im after some advise about chassis stiffness

perhaps the answers are in a book i haven't read or understood properly

could someone please give me some guidance?

just thinking about a chassis stiffness for a minute

if the value was very very low...

what happens?

i think it means rolling moment resistance will match the F:R weight distribution of the car

a front heavy car will oversteer
a rear heavy car will understeer

and it will be impossible to "adjust" your rolling moment distribution

your stuck with the one behaviour

i suspect its time i mathematically analyse the chassis torsional stiffness that is required (preferable/essential?)

im thinking of writing a simulink sheet

i suspect quite a few of you have done this before?

cheers in advance

Frank

[This message was edited by Frank on March 13, 2004 at 08:01 AM.]

[NJM.]

Frank,

I came across a paper on this subject for a NASCAR chassis. The paper is out of Clemson and shows how chassis stiffness changes roll stiffness.

http://www.ces.clemson.edu/~lonny/pubs/journal/sae983051.pdf

Some basic ideas on minimum stiffness I have seen are 10 times the roll stiffness of the car. This is however not as important as I once thought. Having a stiff car is not as important as having a car that is homogenous. You dont want stiffness to fall off at one section of your chassis and be stiff everywhere else. If say you have a car that has a very stiff front and soft rear you will have the front respond to an input and the driver will be waiting on the rear to catch up. It will create a very uncertain feeling about what mode (understeer/oversteer) the car is in. Also NASCAR builders purposely do not conform to the 10X roll stiffness rule. For a car that is so dependant on tires (low aero, big weight) and susceptible to changes in track conditions, having a stiff chassis makes the car harder to keep in its balance range. A softer chassis, although it can never achieve the same level of handling, will be much more drivable over a larger range of conditions.

Hope this helps

[Frank]

thx man

http://fsae.com/groupee_common/emoticons/icon_wink.gif

[RagingGrandpa]

Cornell wrote an SAE technical paper (2000?) on torsional stiffness , targets, and testing. I know our university has all the recent SAE papers on file too so it was free for me. Check and see if you can't get ahold of it.

[Frank]

Answering my own question here...

Yes Frank,
it is true that

if: the chassis has low torsional stiffness

then: the Elastic Roll Moment Ratio approaches the Weight Distribution Ratio

http://www.uq.edu.au/fsae/graph1.jpg http://www.uq.edu.au/fsae/graph2.jpg

this was calculated using "roll axis on the ground" method.. an assumption that IMO isn't too terrible

http://www.uq.edu.au/fsae/model.jpg

[This message was edited by Frank on March 15, 2004 at 06:12 PM.]

[Kevin Hayward]

Frank,

I think that the spring analysis that you have done is a pretty good start to determining ideal chassis stiffness. A couple of papers back it up.

My question concerning the analysis is how do the transients work. Sure this works steady state but the dampers must temprorarily increase roll stiffness. This would mean that in the critical corner entry phase the roll moment distribution may be linked to your weight distribution only approaching what it should be when achieving steady state.

I only put this idea forward having tuned with what is a very stiff FSAE chassis (7000 Nm/deg). I was very confused when small changes (especially damper changes) had quite an effect on the car where they did not have the same sort of effect on our 3200 Nm/deg chassis of 2002. As of yet I have not modelled how the dampers affect this analysis but I reckon it could be something worth looking at.

Maybe I'm just way off base on this one.

Regardless I think that there is a lot of justification for sloppy chassis' on FSAE cars that goes around. Yet going sloppy is in direct contrast to the accepted practice of race car design.

Are our models complete?

Cheers,

Kev

UWA Motorsport

[Frank]

I hear you Kev, and yeah, I reccon a floppy chassis will FUBAR the transients

i really only did that analysis to try and test my prediction (which was correct)

our current car is floppy as hell, and changing springs and ARB rates seems to do little..

there's a joke going around here that im about to throw "dumbshocks" at the front to get it to understeer http://fsae.com/groupee_common/emoticons/icon_smile.gif

either that.. or add 5kg over the front axles

this year i ended up taking an interest in chassis (last year suspension) (( i pick on the weakest point each year ))

Personally i don't have the academic interest/patience to dive right into transient behaviours right now .. im just trying to lay a foundation for "brainiacs" in the future

another theory i have is that transient behaviours are far less critical in a low speed car with little inertia.. i actually said this to a design judge, and got caned on it... but seriously, at these speeds and (very low) levels of driver skill (and im more concerned with driver repetitiveness than actual lap times) i find it very hard to imagine huge gains from in depth transient analysis (I barely have the patience to watch shock speed histogram DA)

there's a guy joining our team ATM, and a lecturer (who wouldn't know a rod end from a capscrew.. although he does know a bit about the practicalities of DA in the general sense) has set this guy the task of DA'ing the bejesus out of our car

ok
point 1: who's gonna drive it
point 2: the chassis is floppy
point 3: what tyres are you planning to use

in my mind there seems very few teams who get to the level of requiring the knowledge base of transients

im not knocking FSAE, or teams, but this is how it seems to me

but as usual, ill probably back flip on these comments in a month or so

there's one thing im sure about though, a floppy chassis is unnecessarily heavy, simply because the load paths are crap

another thing, after doing viewing a couple of hundred FEA analysis' , im VERY dubious of the stiffness figures I see bandied about concerning FSAE

have a look at the extreme things we're going to do around the driver, or have a look at GT's front roll hoop supports

without these, or VERY effective skins, you're going to be hard pressed to get stiffness (without making it heavy)

we all reckon there should be MANDATORY weight in, rolling road dyno , AND chassis stiffness test at the comp. http://fsae.com/groupee_common/emoticons/icon_smile.gif

and hey, we're as guilty as anyone for BS'ing the judges

[Kevin Hayward]

Frank,

I agree with you on the knowledge base on transients. We are starting to really get into the dampers and we keep finding out how little we know.

Also agree that a lot of chassis stiffness figures around seem a bit high. The only real way to get the high figures is through a high walled monocoque http://fsae.com/groupee_common/emoticons/icon_smile.gif Although I am well aware it is not a solution for every team. It just seems a bit ridiculous to me that so many people seem to think that a spaceframe chassis can match a half decently designed CF mono.

With our '03 car the transients take on new meaning. We are able to pretty much have whatever corner entry / exit behaviour we want within the tuning range of the dampers.

And while these cars are light, low polar MOI, short wheelbase they still do not instantly go steady state and the dampers can make a pretty noticeable difference to behaviour. I'm still unsure as to how much of a difference it makes to the stopwatch but it certainly makes a difference to driveability. Well the rebound adjustments do anyway ... still looking for a lightweight well priced damper with decent compression adjustment ... know of any?

But as I said in the last post we weren't able to notice such changes with our last spaceframe car, which was quite stiff for a tubeframe.

Maybe we can have a chat about this over a beer in December.

Kev

UWA Motorsport

[Frank]

yeah, im starting to wonder just how long it is til we get over mono-phobia

the team has just decided to take a new car to england

(not sure if anyone has built a car in 6 months before. but we're about to have a crack at it)

so its new chassis building time, the welding is being done off campus ATM

are you guys building a new mono this year?

[Kevin Hayward]

Frank,

Do you seriously mean a different car than the one you had in December for a competition that is only a few months away?

If you do ... good luck to you. That would be a huge effort.

We'll have a new mono with a few improvements. Not as many innovations this year ... we are more focused on development.

Kev

UWA Motorsport

[Frank]

yes kev serious

http://www.uq.edu.au/fsae/bwoken.jpg
http://www.uq.edu.au/fsae/serious.jpg