Hi. Still can't figure out how to accurately model suspensions in the geom/CFG file. :'(
Okay. So I setup a model and added in all the points that I would need and made sure it was accurate. Setup my front and rear suspension using the points from this model. It was modeled at ride height with the wheel centers equidistant from the center of the world, and at 0 height.
In game this setup needed about 5" right height added in via the CFG file. But it worked as expected. Upon further inspection all of the body mounting points were far lower than expected and modeled. So I set about finding a solution. Remeasured my shock figures (I think I finally figured out how PiBoSo is measuring them) and moved all of my suspension mounting points down by 0.2.
The car now sits at the correct height without major ride height adjustments and the body mounting points are in the correct position. But The car doesn't drive as expected. My COG is correct so are all my wheel weights and spring rates. The car just under steers like a dump truck, and the rear suspension doesn't move correctly.
I though oh well it will come around at least I got the suspension to be in the correct position. I'll adjust the spinner shock length so it looks nice on the spinner. That will be a win. I found the correct length, plugged it into the INI file and my RF tire was in the correct position. The LF was completely broken, laying on its side.
The suspension is built symmetrically, how can this happen? And why is it correct in the garage and on the track, but not on the spinner?
Ideas?
Unfortunately it's very difficult to guess where the problem is from a description.
Have you enabled debug mode to see the suspension geometry and data on track?
What suspension type are you using?
Could you please post screenshots or videos?
He is designing the suspension based off of my dirt modified. It is a solid 3-link rear end style.
Here is a video I made of my rear suspension in a heat race.
https://www.youtube.com/watch?v=rXQuuGH89sM
Here are some photos of measurements off my car and what mark modeled.
The first photo is all measured center to center off of mounting points. The tri-angles are Upper control arms, the lower control arms are on the right hand side.
The pin style things are the ball joints, measured from the center of the ball to the tip of the joint.
I measured the shocks at ride height from center of bolt to center of bolt.
The bottom spring and shocks are the rear of the car.
Picture #2 is what we call a "Pullbar" It controls the pinion angle under accel and decel, that is the top drawing.
The lower drawing is a "J-bar" that is what we used to control the rear roll center and rear end from moving left to right.
Picture #3 Is how my rear springs and shocks are mounted to the frame and to the rear end.
Picture #4 is how I measure my ride heights in real life.
The trailing arms that connect the rear end to the chassis are: Left rear, 18* 15" long and Right rear, 10* 17" long. Typically in a dirt car the more Left rear bar angle you have at ride height the quicker and easier the chassis will Lift on the Left rear. The Left rear then creates "rear steer" while hiking the chassis upward. This creates downward pressure through the rod to the rear axle to create what we call as bite or drive. The flatter the Right rear bar angle is, the tighter and harder the car rolls over on to the right rear on corner entry - mid - exit. Typically we lower the rods down on the frame side to tighten the car entry and exit and raise the rods to loosen entry -exit
I can't help but think the car sitting so funky on the spinner is causing this issue. I just double checked everything, and everything is setup symmetrically in the GEOM and CFG and INI file. (except rear lower links) I can find no reason that the LF insists on sitting like it is. I'll work on a video of how the car acts in game with the wrong geom file that works right, and the right geom file that works wrong. I was hoping to catch you today yet if by chance these images help.
Spinner/background
(https://i.gyazo.com/7e9e5fd122ae06fd9d7982e0ff1d41e9.png)
Front suspension - Dual wishbone
(https://i.gyazo.com/861c9d3960a14e0d7beef464091cffb1.jpg)
Rear suspension - 3-link (Green - Panhard bar, blue - pull bar)
(https://i.gyazo.com/4932c972a2466fcc92b9e52445587ca7.jpg)
Could you please post the "showroom" section of the INI file?
And also an on-track screenshot with all the debug numbers?
(https://i.gyazo.com/e835c14ea1c3a9399fe779fdc46578e8.jpg)
[INFO]
name = Modified_BETA
short_name = MOD_BETA
garage_pic=
engine_type = V8
engine_displacement = 350 CI
max_power = 310
gearbox =
drivetrain = Rear Wheel Drive
Weight = 1500
author = Phathry25
[SHOWROOM]
flsusp_length = 0.115
frsusp_length = 0.115
rlsusp_length = 0.135
rrsusp_length = 0.135
[data]
code = 1
cat = Modified_BETA
[cockpit]
ff_scale = 1.15
[race]
length = 20
rolling_start = 1
Please note that the rear suspensions are heavily overdamped.
The critical damping ratios SB ( slow bump ), FB ( fast bump ), SR ( slow rebound ) and FR ( fast rebound ) should be below 1.
The front suspensions geometry seems correct. What's the value of "shock_minlength" for front left and right?
Is there a way to take shock dyno information and implement it into your physics?
Yes Kyle.
Shock_minlength is 0.358
rest_length is 0.175
Both side are identical.
Quote from: Phathry25 on September 16, 2016, 11:27:59 PM
Shock_minlength is 0.358
rest_length is 0.175
Both side are identical.
The total distance between the front shocks upper and lower links looks definitely longer than 0.358 + 0.115 = 473mm. Could you please try a value closer to the rest length in the INI file?
For example:
flsusp_length = 0.15
frsusp_length = 0.15
Increasing the length in the INI file did not help the LF tire.
Total length is 0.485 I'm pretty impressed you can tell 12mm based on those pictures. Should my total for the two values be <, =, or > that total length?
(https://i.gyazo.com/1b3c9c97af8c2482c8d307a52a1d57b1.png)
Could you please upload the car mod somewhere and send the DL link to debug the LF suspension problem?
Thank you.
The problem is in the camber setting... More info tomorrow...
(https://media.giphy.com/media/Q42cpzgcUeZkA/giphy.gif)
Thanks for all your help!!!!!!!! No further explanation needed. I never even thought the car could blow through adjustment ranges on the spinner. Sorry to waste your time.
In the GEOM file, if "camber_setup" is set to 1, then the camber adjustment is done using the spindle link on the upper arm.
Then "camber_steps" sets the number of possible settings.
If "cambersetting" in the CFG file is greater than "camber_steps", then you get a suspension geometry that is out of range.
In the future, a check will be added to avoid out of bounds settings.