Roll couple, Roll axis??

[pav8427]

Now that I have got the bugs out of the front end, it's time to move on to the rear.
I've got a pretty good grasp on SVSA lengths,Anti-sqaut,Under/over steer, etc.
What gets me baffled yet is roll axis,roll couple and some of the other stuff in regards letting the front and rear play nice together.

Anyone care to run an explanation by me?
Any input pertaining to info available from Performance Trends would be good too.

BackYard Engineering terms would help.

Thanks, Doug

[chicane67]

In simple terms... hopefully this wont be too vague.

Roll axis = The roll axis is the imaginary line through the vehicle from front to rear that the vehicle rotates about during cornering. The tire contact patches apply force vectors through this axis.

Roll couple = The roll axis is a line that joins the front and rear roll centers. If the vertical distance between the roll axis and the center of gravity is not zero, a torque (roll moment) equal to the centrifugal force times the distance between the center of gravity and the roll axis will be exerted on the sprung mass, causing the body to lean towards the outside of the turn. This force is called the roll couple.

[David Pozzi]

In Perf Trends, the Front Roll Couple is the percent of the "Total Roll Load" that is handled by the front suspension. Usually a front engined Camaro would have aprox 75% Front Roll Couple. In that case, the front suspension is receiving 75% of the cornering load in roll only. Front weight can vary from car to car, so the ideal number will vary and even be reverse on a rear engined car.

Click on the text on the far left and an explanation box will appear with more info. Our Camaro is currently understeering at 86% but Perf Trends predicts "Neutral" Go figure!

If you downloaded the manual, near the back of it are general recommendations. The roll axis is usually parallel to the "Mass centroid axis" . Slice a car like a loaf of bread, each slice has a center of gravity height, put a dot there. Assemble the car and connect the dots, that's the Mass Centroid Axis line. The rear is usually higher than the front so the rear RCH is also higher to match it.

FLLD, "Front Lateral Load Distribution" is the percent of total roll stiffness relative to the weight on that end of the car. So if you have 75% of the load, and 75% of the stiffness, the percent of total is 50%. A good starting point is 55% rear FLLD, according to the program. This is like dividing the car into two halves, then computing the weight to roll stiffness and comparing them front to rear to achieve a relative F/R balance percentage. This is a number that can be easily compared between different cars with different front to rear weight percentages and is a better indicator of oversteer or understeer. I need to go back and read the manual on this, I might have it wrong.

If a spring change is made it will show up in the FLLD and a bar change or adjustment can be made to get the balance and FLLD back where it was.
David

[pav8427]

Chicane, You say 'the distance between the CG and roll axis'.
Is there a sweet spot for that or a reasonable range to keep that?
Currently I have 14.2"

Dave , My initial mock up is telling me that I have 81/19 with 'some' understeer. How much is some? What do you figure caused the difference between your Camaro's actual performance and the PT reports?

Also with regards to a converging 4 link, I have read that in a triangulated set up that its roll center is at the IC of SVSA.
And w/parallel set up using a PHB,the roll center is at the C/L of PHB.

How well do them to points get along when dealing with a Conv. 4 link(uppers parellel, lowers kicked in 10* at front) and PHB?
Should I shoot for a specific range to keep it?Up hill,Down hill?
I know under/over steer comes into the equation at some point.

Where does 'level ground roll rate' come into play?

To give you some more insight my initial set up looks like this:
SVSA length = 92.3"
Anti-sqaut=101%
IC height is at 16.2"
RRC=9"

This is going under a 65 Chevelle w/BB 4sp.
Hopefully 3400# and decent front/rear ratio.
At the moment I am not looking at anything that has to do with springs/roll bars/shocks yet These specs. are with a decent ball park guess on sizes.

Thanks, Doug

[David Pozzi]

The anti roll bar rates are calculated from measurements taken which is not the same as actually measuring them. The anti-roll bars are mounted in poly but the computer doesn't "know" that, and probably assumes they are solid mounted. The tires are unknowns to the computer other than width, chassis torsional stiffness is another unknown factor. Perhaps the program is more accurate with a car that has less front weight percentage, ours is something like 56%.

I'm not up on four link design, but if you have an issue of "Chevrolet Power" the GM parts book, there is some info on finding rear roll centers there. The roll center axis should parallel the "Mass Centroid Axls" wherever that is. I always have read the rear RCH should be higher than the front on many cars it's near the height of the rear axle tubes.

Are you expecting the 10* upper links to provide lateral location? Sounds like an iffy deal.

Level Ground Roll Rate shows how much the car will lean, I believe it shows it for 1G lateral.
David

[pav8427]

The uppers are parallel and the lowers are kicked in. No,I'm not, I am running a 30" pan hard bar that is adjustable from 9-15" off ground level.

[chicane67]

pav- Depending on the system, something between 12 and 16 inches. Looks like your right there.


Speaking of calculations for Sta-bar rates... I figured out the equations for those who like to "Bring on the tech".

For those of you that do not... (and you know who you are)... then dont. Lord knows giving you the right tools may eventually lead to your own death.

David, PM your e-mail again... Ill send you a printable version of the attached PDF.

[pav8427]

Holy crap batman. Looks like a conversation between sober and one to many Millers.

[boodlefoof]

Little Falls, MN! My grandparents recently moved after living there for a looong time (the Holmquists).

[Norm Peterson]

Also with regards to a converging 4 link, I have read that in a triangulated set up that its roll center is at the IC of SVSA.
And w/parallel set up using a PHB,the roll center is at the C/L of PHB.

How well do them to points get along when dealing with a Conv. 4 link(uppers parellel, lowers kicked in 10* at front) and PHB?

Should I shoot for a specific range to keep it?Up hill,Down hill?
I know under/over steer comes into the equation at some point.

There's a construction that you have to go through to find the geometric roll center. Please correct me if I'm wrong, but I am assuming that your "parallel" description for the uppers refers to the plan view.

In plan view, extend the LCA axis lines until they intersect. That gives the longitudinal coordinate of the location at which the LCAs effectively provide lateral restraint. At 10°, it won't be a lot, but it's necessary to consider it in order to solve the problem. Anyway, in side view, again extend the LCA axis line and mark that at the longitudinal point determined above. That's one point, and you're looking to find a second point in order to draw the axle's own roll axis.

The other point is typically a similar convergence point determined from the uppers, or it may be defined by a PHB or Watts link or other mechanism that defines lateral restraint. With your arrangement that I currently understand to be something of a Satchell link with the lowers at a much smaller convergence angle than usual, you'll need a PHB or something similar. Note that the "convergence" of parallel in plan view uppers "occurs" at "infinity", which can be freely translated as "no help at all for developing lateral restraint".

Your roll steer will be the slope of the line drawn between the two defined points. In your case, it can be taken as a line between the LCA convergence point and the midpoint of the PHB, as seen in side view. This is the quantity that you may have seen with suggestions for 3% or so (slope = 0.03), or for which 7% is slightly too much for serious performance intent. I think staying closer to 3% will solve a separate issue involving the fact that only four links are required to properly locate the axle while this arrangement has five (doing so should minimize the amount by which the "extra" link gets in the way of what the other four would do otherwise during roll).

Where that roll steer line crosses a vertical line drawn through the rear axle (still side view) is your geometric roll center. As a practical matter, it will be only a little lower than the midpoint height of your PHB (assuming that the PHB is higher than the LCA convergence in order that the roll steer be in the direction of vehicle understeer).


Norm

[David Pozzi]

Tom,
Your Private message mail box is full!

[chicane67]

Ok. That helped.

[pav8427]

Norm, I'm catchin pretty much everthing your pitchin except for the 3% figure. How is that figured exactly? 3% of what?

John, I have been here for 12 years and still don't know a lot of people.
Where abouts did they live. Does this mean you have some Yankee blood in ya?

Thanks, Doug

[pav8427]

Rise/run? As in highway grades?

Doug

[Norm Peterson]

Rise/run? As in highway grades?

Doug

Yup. Expressing it as a 'percentage' makes it easy to relate how much the roll steer is relative to the amount of roll. 3% would be 0.03° roll steer per degree of roll (these numbers all look pretty small, but you would feel the difference).

Off-topic: sounds like there's some civil engineering in your background.


Norm

[pav8427]

Nope,Just remember some of the grades we pulled when I visited my dad in central Alaska being referred to in %.
I'm more of a Backyard Engineer.

Doug

[pav8427]

I had to do a little rework. Looks like if I drop ride height at rear to level out lower arms it takes care of the roll steer dilema.
This actually levels the car and now I'm looking at something like;
SVSA=56.59
Anti-squat=97.8%
IC=9.625"
RC=13"
Roll axis to CG moment arm=12.24"
This will give me 3% roll understeer.
Interestingly, with my initial plot, Perf. Trends was saying I had 'some' understeer. When really on paper it figured out to be right around 7.5% oversteer.
Do you suppose the '5-link' issue is confusing it?
I'm curious to know also. If I were to build a three link using these same dimensions, What kind of results do you think I would end up?

Doug