rpmagazine
|
| posted on 29/6/08 at 10:49 AM |
|
|
syd, which single seat class with wings has more than 30-40mm total suspension movement in normal operation?
www.racemagazine.com.au
|
|
|
|
|
Syd Bridge
|
| posted on 29/6/08 at 02:43 PM |
|
|
I really don't know if you're deliberately trying to be awkward or what???
What I was trying to say, is that a go-kart is only a car without suspension(whereas in fact, the go-kart suspension is the stiffness of the kart
combined with axle widths), and that single seaters today are pretty much the same as karts. Did I mention suspension movement?
Nothing changes in the dynamic. The first thing any (race) car does on commencing turning, is to roll about the front outer contact patch, then the
front and rear outer, and then things start to complicate.
The prime 'rolling' couple is the CoM about the outer contact patches. No matter how soft or hard the suspension is. Always was, and
always will be. Newtonian physics and all that. I enjoy watching road saloons and sedans of any sort racing, particularly in the corners, and seeing
all this in action. Those big old Jags at Goodwood are like ballerinas, dancing and rolling all over the place.. Give it a try, you might just learn
a thing or two, by thinking about what you are seeing happening in front of your eyes. Take a digi video along, and play it back as slow as possible,
frame by frame. You'll be surprised by what you see. The video is an amazingly powerful diagnostic, and learning, tool when searching for
answers.
How the chassis behaves relative to the suspension is dictated mostly by spring stiffness, wherever they are placed. You can have a single swing axle
each side with the hubs rigidly attached at the outer ends. Attach (stiff)springs from the outside to the chassis as high as possble, and the
behaviour will be different to that when the springs are terminated at a shallow angle, or soft springs. Swing axle lengths become irrelevant, as does
all the fairyland RC stuff.
Does the suspension move the chassis, or the chassis move the suspension?
You gotta decide if the monkey is swinging the elephant by its tail, or the elephant swings the monkey,...so to speak. 
And therein lies the dilemma.
Cheers,
Syd.
Edit. Having read all that above, it should now be very obvious to all, that the importance of keeping the CoM as low as possible, can never be
overstated. And thus keeping the rolling moments minimal.
[Edited on 29/6/08 by Syd Bridge]
|
|
|
rpmagazine
|
| posted on 29/6/08 at 11:35 PM |
|
|
what I am trying to ascertain is why you are relating this information to this discussion and our previous discussions of road/track cars?
Suspension design on a single seat track car places an overall emphasis on factors that have very little to do with cars that require suspension
movement.
FWIW if I were designing another single seat car I would place far less emphasis on what the RC does or where it is...I think we shall have to
continue to disagree.
The available CoM is most often defined not by what is perhaps perceived as 'perfect' or 'ideal' but what we can reasonably
achieve given the available components, the budget and the intended function and what we can construct. I know I can lower the 190kg engine in my car
50mm from where it is, great for CoM. But in doing so I need a new bell housing, gear change mechanism, change to coolant circuit, clutch release
mechanism, a dry sump system. In doing so I would increase my total expenditure considerably and I do not have the flexibility there. So I would need
to give up the option of quality dampers and the engineering in getting them set up correctly. Frankly the dampers are far more important.
[Edited on 29/6/08 by rpmagazine]
[Edited on 30/6/08 by rpmagazine]
[Edited on 30/6/08 by rpmagazine]
www.racemagazine.com.au
|
|
|
cymtriks
|
| posted on 5/7/08 at 09:21 AM |
|
|
Ok here's my take on this debate we keep hearing between Syd and others.
Point 1
Think of the car as a mass supported by a system of linkages. If this car was supported by very soft (by which I mean a theoretically total softness)
springs then it would be free to move about on its supporting linkages and would move about the geometric roll centres in every direction.
Point 2
Think of the opposite case. Make the springs very stiff or just replace them with rose jointed links. The car's linkages still move but only by
microscopic ammounts. The important thing now is the CoG, the roll centres just drop out of the picture from a practical point of view. The car now
appears to roll about the CoG with the springs just holding it up. The "suspension", in terms of deflection anyway, is now the tire. This
is exactly what I think Syd is describing
Conclusion
In a road car the result will be some way between point 1 and point 2 so the car will tend to roll, in cornering and squat/dive, about a point between
the geometric RC and the CoG. Given the stiffness of a road spring I would expect this point to be closer to the RC for a road car and closer to the
CoG for a race car.
Comments all!
Especially Syd, is this what you've been getting at?
|
|
|
Syd Bridge
|
| posted on 5/7/08 at 09:53 AM |
|
|
Geez, I'm now convinced I don't write in English!!!  Can someone interpret for me??
Did I not say that the car rolls about the outer contact patches, due to the roll couple formed by the CoM trying to go about the contact patch.
Cheers,
Syd.
|
|
|
Doug68
|
| posted on 5/7/08 at 11:28 AM |
|
|
I think this is the simple point Syd is making...
In theory...
And in practice...
Doug. 1TG
Sports Car Builders WA
|
|
|
TheGecko
|
| posted on 5/7/08 at 02:39 PM |
|
|
Nice pictures Doug = I think they do help to illustrate the point.
My take on this:
In pure geometric terms it may be true to say that the roll centre is the point about which the suspension appears to pivot. In practice, as
anyone who's done any roll centre modeling will know, the RC's move all over the place so the car obviously isn't actually rolling
around that point.
In kinematic terms, the fixed point is the contact patch of the outside tyre and everything else moves around that.
The problem is, calculating the geometric roll centre is (relatively) easy whereas calculating the kinematic behaviour is much harder. I'm of
the personal belief that geometric roll centres are still a useful fiction for comparing behaviour of different suspension designs as long as we
don't start believing the car actually rolls around them! I might have a bash at rejigging my RC spreadsheet to work backwards from the outer
contact patch and calculate the position of everything else i.e. chassis roll and both inner and outer camber angles. To do that really requires some
knowledge of roll stiffness, roll couples and the cornering force and their relationship to teh suspension geometry. My 3 semesters of engineering
(from 20 years ago) are now looking pretty shaky! (I eventually completed an IT degree).
Do there exist amateur level tools for doing these calcs?
Dominic
|
|
|
v8kid
|
| posted on 5/7/08 at 05:09 PM |
|
|
What does it matter?
Surely all we are interested in is making sure the tyres get the best grip possible which means keeping them upright or 1 or 2 degrees negative.
In ultimate terms the outside tyre is the most important and in transition terms the inside tyre is only important while there is significant weight
on it.
Apart from that keep it low keep it central. You can't change the basic laws of physics no matter how fancy the linkages or theory!!
So use one of the loads of free spreadsheeets around and check out that the wheels are always 1 or 2 degrees negative for your range of bump and roll
and Bobs your auntie. Also I'd think it would be handy if the changes in camber were smooth with no sudden changes - would this corelate to the
"roll center migration" I wonder?
Trouble is I find it damn difficult to achieve it and after a few hours of plugging away at a computer get bored and go for a pint
Three nights so far and I still havn't found a suitable compromise between the available chassis mounts and ones that give good camber change -
Its a hard life being a Locoster I'm off for a pint
Ah I should have mentioned we want to have both ends changing by the same ammount so that the oversteer/understeer ballance stays the same. Some cars,
including mine, tend to oversteer more in fast corners so adding a bit of of understeer at the limit makes the driving less tiring.
Now I'm definitly going for that pint
[Edited on 5/7/08 by v8kid]
|
|
|
kb58
|
| posted on 5/7/08 at 06:44 PM |
|
|
http://www.2kgt.com/MarkOrtiz/2004_6.pdf
http://www.2kgt.com/MarkOrtiz/2004_8.pdf
Mid-engine Locost - http://www.midlana.com
And the book - http://www.lulu.com/shop/kurt-bilinski/midlana/paperback/product-21330662.html
Kimini - a tube-frame, carbon shell, Honda Prelude VTEC mid-engine Mini: http://www.kimini.com
And its book -
http://www.lulu.com/shop/kurt-bilinski/kimini-how-to-design-and-build-a-mid-engine-sports-car-from-scratch/paperback/product-4858803.html
|
|
|
cymtriks
|
| posted on 6/7/08 at 08:11 AM |
|
|
After Syd's last comment here's my revised take on this debate we keep hearing between Syd and others.
Point 1
Think of the car as a mass supported by a system of linkages. If this car was supported by very soft (by which I mean a theoretically total softness)
springs then it would be free to move about on its supporting linkages and would move about the geometric roll centres in every direction. The
important thing, in terms of how much roll occurs, will be the difference between the geometric RC and the CoG.
Point 2
Think of the opposite case. Make the springs very stiff or just replace them with rose jointed links. The car's linkages still move but only by
microscopic ammounts. The important thing now in roll is the height of the CoG above the road, the roll centres just drop out of the picture from a
practical point of view. The car now appears to roll about the outer tyre contact with the springs just holding it up. The "suspension",
in terms of deflection anyway, is now the tire. This is what I think Syd is describing.
HOWEVER....
A search of the internet reveals a few threads on "force based RCs" which, according to people who claim to have modeled these, represent
the true RC.
These are described as being higher than the geometric RC, part way between the geometric RC and the CoG
NEW Conclusions...
In a road car the result will be some way between point 1 and point 2 so the car will tend to roll, in cornering and squat/dive, about a point between
the geometric RC and the tyre contact patch. Given the stiffness of a road spring I would expect this point to be closer to the RC for a road car and
closer to the tyre contact for a race car.
However, as Syd has said, this is the initial event and after this it becomes "complicated" (his words!) so to continue the conclusions
AFTER the initial turn in event and roll about the tyre contact the car settles on its suspension and adopts a stance partly based on the suspensions
geometry, but...
Is this second event controled by the geometric RC, a force based RC or some other contact patch/RC/CoG relationship?
Comments again guys! We may yet figure this out!
|
|
|
rpmagazine
|
| posted on 6/7/08 at 09:39 AM |
|
|
Geometric RC's and the theories behind them are an old tool to try to understand what happens within a very dynamic event. It is a tool clearly
with flaws, but it has been around a while now as and with any theoretic perspective is has become somewhat reliable and predictive of certain
events...but it is still only a tool.
Force based RC's are used with high end programs that we as mere enthusiasts cannot really access i.e. Adams etc. Agan it is a tool that
manufacturers use and they would not spend millions without some reason, but they do also back it all up with a massive amount of testing and
correlating test results with the programs that they use. This is one area that virtually all enthusiasts fall down on, partly because we don't
have the time/skills/resources but also because we simply want to enjoy what we have created.
What does it matter?...well why do so many go to such trouble to work it all out? I was able through an acquaintance, able to ask some questions of
the designer of the McLaren F1. He answered one of the questions by using the GRC theory and he took it seriously enough.
My concern is that we are treating the contact patches as static things are they simply are not, nor are they fixed points.
The other aspect to this is that we are discussing a single element of whole a cornering event, with the implication that this is the most significant
element in a vehicles handling and IMO it is not really.
WRT keep it low and central, it is all a balancing act (no pun intended), particularly as to what is possible within our respective skills sets and
finances. A good example is my drive-train: I *can* feasibly lower it 50mm but to do so is a lot of work and would absorb a lot of finances that are
required for say good dampers. So I make the decision based on my set of priorities. I have a total CoG of 432mm which is reasonable and I need the
money to buy dampers so there is no need to change. The dampers will have more of a positive effect than a 12mm OA lower CoG.
[Edited on 6/7/08 by rpmagazine]
[Edited on 6/7/08 by rpmagazine]
www.racemagazine.com.au
|
|
|
kb58
|
| posted on 6/7/08 at 03:50 PM |
|
|
quote:
Originally posted by rpmagazineWhat does it matter?...well why do so many go to such trouble to work it all out?
If you spend millions on an F1 team and lose by 0.01 seconds, that's a big deal. For most of us we're happy with getting 80% of the way up
the expense/return curve, before it heads straight up in order to get that last 20% - we just can't afford to go there. However, teaching
ourselves to understand what that last 20% is about is well worth it, though information on force-based roll-centers is not common yet. The articles I
posted are very helpful.
That said, I'm not sure it makes any difference for the majority of us. That is, most of us aren't good enough drivers to even see the lap
time improvement provided by the slightly better suspension. It's like a very small signal improvement in a sea of noise; it's going to be
very difficult to measure.
[Edited on 7/6/08 by kb58]
Mid-engine Locost - http://www.midlana.com
And the book - http://www.lulu.com/shop/kurt-bilinski/midlana/paperback/product-21330662.html
Kimini - a tube-frame, carbon shell, Honda Prelude VTEC mid-engine Mini: http://www.kimini.com
And its book -
http://www.lulu.com/shop/kurt-bilinski/kimini-how-to-design-and-build-a-mid-engine-sports-car-from-scratch/paperback/product-4858803.html
|
|
|
rpmagazine
|
| posted on 6/7/08 at 10:25 PM |
|
|
Actually, having spoken to some F1 designers I might suggest that they do not take RC's very seriously given the other priorities such as aero.
Same for CART.
The engineers I know working with Adams etc are all OEM engineers and they have refined their understanding with empirical work too.
The person assisting me (actually he is doing nearly all the work - a choice I made given the information and process is in the public domain) works
with Adams all the time. He noted that his work in Adams seems to constrain the GRC much as we have discussed in the past (he has checked a few times
to satisfy his intellectual curiosity), but it is not an intention but more of a correlation.
www.racemagazine.com.au
|
|
|
v8kid
|
| posted on 7/7/08 at 09:25 AM |
|
|
Guys,
What’s with these TLA's? Its just too much effort to follow the argument and decipher the latest three letter acronym. Easy to fall into I
know.
I’m beginning to have some considerable sympathy with Syd he offers an alternative explanation, poses that we should think about it, and gets sniped
at from all quarters.
That would be OK if the counter argument was original but when it runs along the lines of – a very important person says so or look at this link to
someone else’s articles - there is not a lot of credibility in that as an original argument now is there?
I freely admit I do not understand the significance of geometric roll centre migration or indeed 50% of Marks articles and since I have a couple of
engineering degrees I am conceited enough to think it isn’t stupidity.
The original questions posed were to very well chosen and have made me think. In fact the unasked question made me think more.
I reckon the answers should be Yes positive would be a good idea, No roll plus, Never ever and the unasked question – what are the other wheel pair
doing at the same time.
Syd is absolutely correct when we are thinking about one axle, we were the ones who jumped to conclusions. He is also absolutely correct about the
basic physics, what the roll centre myth is doing is trying to give us a way of understanding how the front and rear axles communicate.
Remember the chassis is rigid (hopefully) and the suspension is not (also hopefully). If the roll resistances at each end are different the difference
is transmitted via the chassis. I think I’ll do a Syd here and let you fill in the gaps.
Depending on the arrangement of the wishbones the roll is resisted via the springs ( and very expensive dampers) or the wishbones. That has an affect
on timing only when considered in conjunction with the above.
We also have to bear in mind the original question there is no point in getting the blunt end to communicate the way we want with the sharp end if the
camber is all wrong.
Do you remember the great debate over analogue and digital recording systems? I was suckered and bought a very expensive Lynn Sondeck system. The
engineering was awesome. When after a few years I really listened to CD’s I was blown away and sold my analogue system for a much cheaper and better
sounding high street setup. The moral is - the basics source was right in the CD and no amount of gilding the lily on the analogue system could
improve on what was never there in the first place.
Same with the fundamental architecture of the chassis - it’s worth the effort to get it right in the first place.
So there it is the thoughts of Chairman V8Kid. Can I ask you a simple question to be answered simply in your own words? “what is the effect of roll
centre migration” (no TLA’s)
I don’t think anyone knows!
|
|
|
Syd Bridge
|
| posted on 7/7/08 at 10:21 AM |
|
|
Someone agreeing with me!!!! I better check I'm still alive!
Until you lot let go of those cursed 'roll centres', of any flavour, then any hope is lost.
Rollcentres of any type DON'T come into the design process of a road or race car. Anyone telling you that they do, and professing to be a
designer, is pulling your whatever. They may be examined 'after the fact', when the car is up and running, just as an academic exercise,
but they just don't dictate design parameters.
It's all about packaging, but the race car can then be designed for specific behaviour which the road car compromises on, and placement of
critical components can be properly designed in.
Move a spring mount up or down a little, indeed even change spring poundages, and all the GRCs, and even the dynamic, can change appreciably, with
even some very small changes in springs and support placements. It's the big picture, not the fuzzy little one, that counts.
As I put previously, 'does the monkey swing the elephant, or the elephant swing the monkey'?
Cheers,
Syd.
|
|
|
rpmagazine
|
| posted on 7/7/08 at 11:18 AM |
|
|
V8 and syd, the problem we (us doubters have) is that we have exhaustive literature, tools and well respected professionals constantly working in the
field all noting that the way we have chosen is a reasonable way forward.
On the other hand we have Syd, an essentially anonymous forum member telling us they are all wrong and only he has the truth and that so far it comes
down to 'the car rolls about the outer contact patch and then it gets a lot more complicated'
Apart from the mistrust of an obviously messianic allusion I have not yet seen any theory from syd.
So perhaps this is the time: Run us through the whole 'more complicated' bit of the cornering event and also inform us as to your specific
background so that we can judge the extent of your experience. At the moment all I can see is the links you have to what looks like a ladder chassied
911 replica powered by a cast iron Ford Essex V6.
www.racemagazine.com.au
|
|
|
pbura
|
| posted on 7/7/08 at 11:46 AM |
|
|
As I understand it, in a symmetrical suspension (one in which the geometric roll center is constrained), the force-based roll center and the geometric
one will be very close. So, it's a decent tool for the home builder to use to estimate spring requirements, etc., in the absence of data
acquisition sensors and racks full of tires, springs, shocks, and antiroll bars.
For the designer, there's no sacrifice involved in centering the GRC. You don't give up camber or Cg. So, I don't understand what
all the fuss is about. Nor Syd's vehemence.
Pete
|
|
|
v8kid
|
| posted on 7/7/08 at 03:30 PM |
|
|
Blimey we are getting a bit heated aren’t we? You know its not heresy to question or argue beliefs held by the majority nor do we live in a society
that treats those who differ as outcasts. Do we?
I don’t think all well respected engineers do think the same way and there are well respected engineers who question the value of roll centres. Dare I
mention Mitchell? Dixon? Engineers not journalists so their arguments are not as forcefully presented.
So what if we do get the GRC, (now look what you’ve done I’ve slipped into TLA’s myself) so it does not move. What difference does it make?
How does it affect the cars handling? Specifically what effect does roll centre migration have?
Third time I’ve asked with no answer and that’s because no-one knows. – So what difference does it make achieving something if you don’t know what you
are achieving?
Jolly good original question by garage19 what? It certainly got me thinking that the only important thing is maximising the tyre grip and balancing
the grip between the ends of the car at all times.
That means the camber has to be appropriate and the roll stiffness has to change the way we want it to with wishbone movement.
Now we are expected to believe that roll centres are more than just a measure of the roll stiffness and restraining the movement of the roll centre is
some holy grail to be achieved without regard to camber.
Admittedly the roll center is a very rough approximation and the relative heights of the instantaneous roll centres front to rear do indicate which
way the roll balance is likely to go but we can calculate that from charting the roll stiffness.
For all Syd is berated for not giving detailed explanations I don’t see his detractors making any contributions – apart from saying it must be so ‘cos
a clever man says so..
So here’s your big chance answer the question.
P.S. Since you ask I have a HND in EEE ( Electrical and Electronic Engineering) , a degree in Mechanical Engineering, and work as a Design Engineer
but I don’t think that matters as long as you can structure a good argument. If I recall correctly Cymtricks is no slouch either if I seem to remember
reading a very informative dissertation on chassis design correctly backed up with experimental results.
|
|
|
Syd Bridge
|
| posted on 8/7/08 at 09:09 AM |
|
|
Mock as you may, Mr. Magazine, but one of those 'ladder chassised 911's' (which are actually space frames if you got close enough to
look) designed and built by an aussie mate of mine, nigh on 20 years ago, would and did kick ten bells out of genuine 911 turbo's of the day
around a circuit, and powered by a measly 200 hp Cosworth(with a cast iron block!), against 300hp+ of the 911. Those cars will still do it today, and
do at track days.
The essence of the superiority is in the suspension. Double wishbones all round and coilovers. And the 'ladder framed grp 911' also was
about 200 kg's lighter.
Then I was fortunate enough to be invited to work alongside some very well known designers, who laughed when I brought up the subject of 'roll
centres'. To a man, they all asked, 'have you been reading Staniforth or something similar?'
It was a couple of sentences by Gordon Murray, whilst writing about the McLaren F1, that made me ask pertinent questions, and as I worked more within
the confined space that these designers operate in, I gathered the answers and the mindset that these fellas operate with.
Today, this small band still keep things close to themselves, and I don't have anywhere near the knowledge they have on the subject.
But, like them, I'm willing to learn and more importantly, to think for myself, and question things that just don't make any sense to me
whatsoever. Like 'roll centres', no sense whatsoever.
Think for yourself, I've given you enough clues to come to your own conclusions.
Cheers,
Syd.
I've heard it said, that sarcasm is the poor friend of anger, or the lowest form of wit. Both fairly sad, in reality.
[Edited on 8/7/08 by Syd Bridge]
|
|
|
rpmagazine
|
| posted on 8/7/08 at 09:33 AM |
|
|
LOL!
Whilst I think both posts above are quite insightful as I am neither heated or mocking.
I have simply asked for more detail on Syd's theories and asked for some background so I can judge what weight to give to his original theories,
particularly given his habit of openly mocking others as 'armchair experts'. I think it is a perfectly reasonable question to ask,
obviously you do not V8kid and Syd seems slightly outraged!
V8kid you have also mistaken another aspect of my post, I was not asking for formal qualifications as frankly I rate long experience just as highly,
hence the phrasing of my question. Should you wish to make up a defining question I will pass it on and get an answer for you.
As for original thought, I think that most on this forum and I include myself, have enough trouble working accurately with accepted theories that are
known to have a desirable effect.
[Edited on 8/7/08 by rpmagazine]
[Edited on 8/7/08 by rpmagazine]
www.racemagazine.com.au
|
|
|
v8kid
|
| posted on 8/7/08 at 11:39 AM |
|
|
Eh!
I think thats my case QED then!
Pity it was looking to be a jolly good discussion and I certainly clarified my own thinking although it I didn't think my point would be so
easily accepted.
Cheers all
David
|
|
|
rpmagazine
|
| posted on 8/7/08 at 12:49 PM |
|
|
you had a point?
I missed it in all the noise, what was it?
www.racemagazine.com.au
|
|
|
mr henderson
|
| posted on 8/7/08 at 12:52 PM |
|
|
quote:
Originally posted by rpmagazine
you had a point?
I missed it in all the noise, what was it?
I missed it too.
|
|
|
garage19
|
| posted on 8/7/08 at 03:58 PM |
|
|
Sooo..... back to the original question...
In the car/application I have previously described would I be way off the mark if i had camber gain that happened to be pretty equal to body roll?
|
|
|
Alan B
|
| posted on 8/7/08 at 04:54 PM |
|
|
quote:
Originally posted by garage19
Sooo..... back to the original question...
In the car/application I have previously described would I be way off the mark if i had camber gain that happened to be pretty equal to body roll?
That's what I'd be aiming for......
|
|
|
