Choosing camber curves 1

[BKB001]

I am trying to design a clean sheet front (and rear, but one at a time) SLA suspension without a lot of constraints. I have been through this thread and learned a lot of valuable things, but I am still missing some fundamentals.

The suspension will go into a small light car (think Austin Healey Sprite) that will most likely be slightly overpowered and used mostly on the street with a few trips to local autoX and roadcourses.

My intentions were to provide a KPI of around 14 degrees and a castor angle of around 7 (I have heard that if KPI is twice castor that the camber curve will remain constant while turning). That may be too high, but please let me know.

The main question is about the camber curve. My first draft thought is to make the camber curve exactly follow the body roll with a little extra negative camber (1/2 degree or so) beyond. On a track width of ~45 inches, this means that at 3" compression, the tire should have around 8 degrees of negative camber. This seems excessive. Am I missing something? Is the goal reasonable?

Also, don't think that I forgot about RCs, but given the controversial nature of them on this forum I was going to leave them be for now.

Thanks in advance!!
Brian

 

[GregLocock]

"My intentions were to provide a KPI of around 14 degrees and a castor angle of around 7 (I have heard that if KPI is twice castor that the camber curve will remain constant while turning). That may be too high, but please let me know."

That KPI is very high, I've never heard that rule, even if it is right it seems to me to be a bad recommendation. Also, how do you set your castor? 7 seems OK, but I am interested - there are plenty of cars around with 3.5 , or even less. Also what mechanical trail are you aiming at? and scrub radius?

"The main question is about the camber curve. My first draft thought is to make the camber curve exactly follow the body roll with a little extra negative camber (1/2 degree or so) beyond. On a track width of ~45 inches, this means that at 3" compression, the tire should have around 8 degrees of negative camber. This seems excessive. Am I missing something? Is the goal reasonable?"

Sounds sensible to me, for a track car. It won't be much fun on bumpy roads, and tyre wear is likely to be poor.

"
Also, don't think that I forgot about RCs, but given the controversial nature of them on this forum I was going to leave them be for now."

Bring it on. The reason they are controversial is that they are an endless source of mischief and misunderstanding. Also, you can't really talk about camber curves unless you are going to think about RCH.










Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[BioMax]

I built a full tube chassis Chevelle with a-arms on all four corners.
The camber gain I ran was 1 degree for every inch of travel and a lower front roll center than the rear.
Total travel on all four is 8" and ride height is right at half travel.
I ran a small amount of anti-dive and a very small amount of anti-squat.
The scrub radius is 2"
And 5 deg of castor

The car is primarily a street car and is very well manored. It has not yet seen the track, but so far everyone is pleased.

BTW the car is slightly overpowered and yet remains well manored.

 

[GregLocock]

40 degrees per metre of travel is a pretty typical number for a street car's rear suspension.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[BKB001]

OK, sorry for the delay in response, but I was under the assumption I had already (It has been crazy here)

At any rate - thank you all for your inputs. I of course have more questions that stem from them, but that is what this is all about, right?

Greg, you mention that 40 degrees/m is common for a street car rear suspension. What about the front?

What about for a elise/Z06/S2000 type car?

Now I imagine that my first declaration about KPI and castor was too simple. I would imagine that adding castor would tend to make the camber curve more aggressive under cornering. How much? It would seem to me that if camber were gained in this respect (although maybe having the camber being variable based on steering wheel rotation too heavily is bad?) it would eliminate at least some of the troubles encountered through bumps on the street (unless you were hitting bumps and turning, but that's OK, as long as I don't end up darting about on the highway).

How KPI adds to all of this doesn't sound right to me anymore. To make sure I have this right (I may not) KPI is the angle between the vertical and the upright member of the suspension looking in the direction of travel. It seems to me that this wouldn't affect camber as I had previously thought (from a post on these forums, I will withold URL as I am at work and have to get going soon).

BioMax, your car sounds interesting and similair (although mine will be ~1800 lbs). I have used those same rough numbers for suspension travel and ride height. I have not yet decided on anti's just yet. How is the steering feel with only 5 degrees of castor? Do you run P/S?

The other thing I may be struggling with is that the track of the car I am working with is very short, and that will perhaps mean I do not need the suspension travel under cornering, and my camber curve may have to be steeper than normal.

Now for RC. I had planned on making the RC on the ground or very close to it on the front. The rear will be slightly higher, and the suspension will be very similair. I haven't read anything about RC's that has conviced me to aim for something other than that vague description... please forgive and correct me if I am wrong.

So I have more questions to add to the originals now. My initial approach about camber curves was only valid for steady state cornering, which we don't do that much of on the track. This leads me to the following:
A: Under braking the camber would be too far into the curve to be beneficial and the turn in would be sloppy and bad on tires
B: Under exit, the curve would be not negative enough to help, and would still be sloppy.

Of course neither of those account for the help of castor adding camber while turning.

It seems that if the camber curve was too shallow for the body roll, then the exit would be extremely poor (especially for the inside wheel, which will be at positive camber then), and under turn in, the tire would be at a too positive angle if not for the help of castor again.

Oh - and I was aiming for 0 mechanical trail - do I want any?

I think it is safe to say that I need a good straightening out. It's safer to be ignorant, but more fun to know what you're talking about. I am clearly in the middle - I know just enough to get it wrong.

Any comments would be appreciated.

Thanks again
-Brian

 

[BKB001]

I think I just made some sense in my own head. I can now see how castor will add negative camber under turning (on the laden wheel). I can also now see how KPI will reduce negative camber on the laden wheel. Am I right in saying that they offset the camber curve towards the negative but do not change the actual curvature?

If I have this right, I am going to go straight home and model some things in solidworks to see this for real.

Let me know.

 

[BioMax]

"BioMax, your car sounds interesting and similair (although mine will be ~1800 lbs). I have used those same rough numbers for suspension travel and ride height. I have not yet decided on anti's just yet. How is the steering feel with only 5 degrees of castor? Do you run P/S?"

Brian-
I am by no stretch a racecar driver or even just for sport. I'm a fabricator, nothing more. But I can tell when a car doesn't feel normal. If something is doing it's job right you should not notice it.

I am not running P/S, I had a semi-custom rack and pinion built for the vehicle with a semi-agresive steering ratio. I sucks to "3-point turn" but it feels great on the road. The only thing I wish I could change now would be the scrub radius. There is more bump input then I would like.

As far as KPI, I have heard several theories on what works where. On vehicles that I am forced to use a high KPI, (meaning the top ball joint is closer to the center of the vehicle then a low KPI upper ball joint) I will run more castor to help compensate for the excessive positive camber on the outside tire when turning. The KPI/castor relationship certainly effects camber, but I couldn't tell you how much or how to set it up.

I am looking foward to hearing what every one else has to say about this subject.

 

[BKB001]

Well it seems as part of what I have asked is a bit complicated. If I simplify a bit and guess and check it may be more relevant. Since I now know that the camber produced by castor and KPI will be an offset, I can offset my camber curve due to suspension travel by this amount. I also know that tha castor and KPI have signifiant 'feel' affects on the chassis.

So, for a 1800 lb car with a 44" track, I would think that more castor would be OK because the car is light (I don't intend to run PS) and would help with feel of this light car. It also would help make the camber curve more aggressive as the car approaches a turn.

So I have a Solidworks model made up where I can check the curve, but I have to iterate manually. How does 9 degrees of castor sound for a starting point? I will sweep across 3-9 degrees of KPI to check for what happens to the curve.

Any suggestions?

 

[spdracer22]

I don't mean to thread-jack, but could someone set me straight on the difference between KPI and caster?

 

[BKB001]

They are both inclinations of the upright, just different axes. The KPI is the angle the upright is tilted towards the center of car, and caster is the angle it is tilted towards the rear of the car.

Don't worry about thread jacking - this thread seems to have died.

 

[GregLocock]

Yes, it wastes a lot of electrons to start a new thread.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[spdracer22]

"The KPI is the angle the upright is tilted towards the center of car, and caster is the angle it is tilted towards the rear of the car."

Then how does KPI compare to camber?

My assumption, and please correct me if I'm wrong, camber is the inclination of the tire, and the KPI is the angle of the line connecting the upper and lower ball joints...the KPI and the camber angle don't necessarily have to be the same, but they'd be fixed to one another...

 

[GregLocock]

BKB has it right. The castor and KPI refer to the steer axis (upper and lower ball joint in some suspensions), camber refers to wheel inclination in front view. For a given spindle/carrot geometry increasing KPI increases negative camber by the same amount.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Supercar1]

> 40 degrees per metre of travel is a pretty typical number for a street car's rear suspension. Cheers Greg Locock

I just read about some Baja desert racers. Their suspension travel is measured in *feet*. I did not know that Greg's company uses meters!

Even though I am metric-thinking, I could never turn my brain around to think in deg/mm. It has always been 1 deg/inch +/- something. Good to know that it is ~40 deg/m, thanks, that helped.

 

[Supercar1]

BKB001,

You are thinking the right thoughts. You just need to add realistic numbers to them. Unless you are driving off-road, you will not be using +/-3" of travel on your small track car (44" track). Also, the steeering angles will be quite small, so that caster/KPI thing may have a smaller effect than you think.

Castor and KPI is not what affects your feel. It is the castor and KPI *trails* that do. The KPI trail will mostly affect you in braking and over bumps. Steering jacking too. The castor trail will make the steering heavier. You can offset the knuckle forward of the steering axis to reduce it.

Consider your center of gravity and weight disgtribution. Some car have their inside front wheels up in the air when exiting corners, so the unladen front camber there is not an issue.

Nevertheless, if your car uses its both front tires when cornering and exiting corners then it is a challenge, and usually a compromise between the steady-state cornering and transient handling. That's the best that I was able to figure out myself.

 

[NormPeterson]

If it really is a Sprite/Midget/Spitfire/most other little sports roadster, there isn't enough room for a boxer engine. Footwells start barely aft of the front suspension/crossmember, with enough room in between for the crankcase and maybe one bank of barrels (never mind either head, clearance, plug access, etc.), and the floor is essentially at the chassis low point already.

This sounds suspiciously like an SCCA Solo X-Prepared project with enough displacement such that the required minimum weight would allow for the retention of basic street-legal equipment (the intent of Prepared Category limitations is to retain recognizable exterior features while providing for performance and necessary safety features. Body shape restrictions in this category are aimed at preventing attempts at ground effects or streamlining). A serious effort aimed a little higher than regional competition is probably looking at being somewhat more than only "slightly" overpowered.


Norm

 

[BKB001]

I am glad this thread is getting some more attention, thank you all for the help. I can get a bit more specific now as to my application.

You guys have overestimated me, this will be a track day (and auto-X) car that will receieve regular street duty. It would be great to get it entered in some events, but I am by noi means preparing for a class. I am hoping that I can prepare it for the track with a few minor things for street use. I want to build the chassis from the ground up and set the body on it. This makes choosing the body only important in external dimensions. I was originally looking into Austin Healeys, but the track width is very small and they are very hard to find and $$$ when you do. I have sourced a couple of Datsun 2000 roadsters that will most likely be finding their way into my garage for a song very soon. Going with a Datsun leads me a bit into the engine selection, which will naturally be a VQ35DE nissan engine out of a G35 or 350Z. It is a very lightweight engine (315 lbs dressed but dry) that we can find lots of around here. So we are looking at about 270 hp in 1800 lbs. (the subaru engine is too wide, and I'd really much rather be naturally aspirated). Now I just have to figure out how to cut the darn high windshield down.


Enough of background.

So the width of the Datsuns is a bit more forgiving, it is around 54 inches. The wheelbase is roughly the same, and the extra width should make fitting control arms a bit easier.

That's an excellent point regarding the width, perhaps 3 inches up and down is a bit too much. I wanted it to be driveable on the street, and I figured that was about the minimum... what would you suggest?


The idea of trail hasn't quit distanced itself in my little brain from the castor/kpi idea. If I got it right, castor is the angle from balljoint to balljoint and trail is now far off this line the spindle is... does that sound right?

So, bowing to higher powers, are my latest estimates of 9 degrees KPI and 9 degrees castor off the mark? I was aiming high at these due to the low weight of the car, but I have no idea what I am doing in reality. I want it to have excellent feel to the steering, but I do not want to cause the need for PS.

What values of trail would you suggest? And since I am new to the idea of trails, which direction are the spindles offset for each?

I was trying to set up the camber curve for steady state. One thing would help greatly... what roll angle can I expect from a 1G corner?

Greg, thanks for confirming my KPI/Caster definitions, I probably learned that from on of your posts on this forum anyway.



Thanks all for your help and support. You will all be the reason this car will handle so well.

-Brian

 

[Supercar1]

> ... width ... is around 54 inches. The wheelbase is roughly the same

??? 54" wheelbase?

> ... perhaps 3 inches up and down is a bit too much. I wanted it to be driveable on the street, and I figured that was about the minimum... what would you suggest?

For a small car +/- 1.5-2.0" is plenty.

> ... castor is the angle from balljoint to balljoint and trail is now far off this line the spindle is... does that sound right?

No, that is a caster offset. The caster trail is the distance on the ground between the steering axis and the center of the contact patch, if the side view. In the front/rear view that distance is a KPI offset, a.k.a. "scrub radius".

You can have many different casters, cambers, caster offsets and KPI offsets, but the same caster and KPI trails, and approximately the same steering efforts. When thinking about steering efforts, these trails is what you should target in your design, not the angles. But the angles are important for handling.

It is hard to describe this in one post. Get any vehicle dynamics textbook and it will have everything you need to know in one chapter with lots of pictures.

> So, bowing to higher powers, are my latest estimates of 9 degrees KPI and 9 degrees castor off the mark? I was aiming high at these due to the low weight of the car, but I have no idea what I am doing in reality. I want it to have excellent feel to the steering, but I do not want to cause the need for PS. What values of trail would you suggest? And since I am new to the idea of trails, which direction are the spindles offset for each?

Or you can have, say, 18 degrees of caster and have the wheel center offset a little forward, like on a bicycle fork. That will result in the same caster trail and about the same effort. Got the concept?

To be technically accurate, you need to think about the distance between the center of the contact patch and the steering axis. So you need to multiply these trails by the cosine of the caster or the KPI.

> I was trying to set up the camber curve for steady state. One thing would help greatly... what roll angle can I expect from a 1G corner?

About 2-4 deg/g for your car, with 2 being a racecar and 4 being a daily driver.

 

[NormPeterson]

I think the track on those little Datsun roadsters was still somewhat less than 50". They were more MGB-class cars than Sprite/Midget, but that's still pretty small relative to just about anything current. I've got a small collection of car dimensions that I'm pretty sure I can locate for more accurate numbers.

Those VQ engines are rather nice powerplants, 75 bhp/L NA even in 3.0-liter Maxima trim and completely un-fussy about doing so.

The separate chassis (tube frame?) is probably going to land you in E-Mod, as that construction technique isn't XP or Street Mod legal.


Norm

 

[BKB001]

The wheelbase is roughly the same as the Sprite, not the track - that's what I was trying to say. Norm, as far as exact measurements, I don't know, I was going by internet references... please fill me in if you find better numbers.

Width 58.901 in | 1496.1 mm.
Height 52.201 in | 1325.9 mm.
Wheelbase 89.801 in | 2280.9 mm.
Front Track 50.201 in | 1275.1 mm.
Rear Track 47.201 in | 1198.9 mm.

I was going by width because that's how I am doing most of the calculations. I never knew if track was from the center, the inside, or the outside of the tires, and I know that I can always flare fenders (although bodywork is not fun) if necessary to get a little more room for the 225s I am planning.

I agree, the Nissan engine is a honey - it'll fit in well, even better than a V8, plus it will fit physically better too! I think I am going to go with the newer (but not newest) G35 variant making 270 or so. That should be plenty. I think I may be Megasquirting too, if anyone wants to help with that on another thread or E-mail. That engine will be living in dreamland hauling a 1800 lb Datsun around instead of a 3500 lb Maxima.

I am still confused on trails vs angles, not on what they are physically, but on how they affect the car. I'll have to read up in Milliken tonight.

I understand that building the car this way will class me out of anything I could actually compete in, but that's OK, this one is more for fun than for weekly racing.

As far as wheel travel, if the car leans 3 degrees during a 1g turn, the suspension will have to compress at least ~2.8" for the outside tire to be on the ground. I think I should still aim for that, right?

All previous and future thoughts/comments/slams appreciated.

Thanks guys
-Brian

P.S. Seriously, does anyone know how to cut a windshield on purpose? Those 2000's had a higher windshield to make some governing body happy and it looks wrong by about 3".