3-d Suspension modeling
#3
EvoM Guru
iTrader: (1)
I haven't seen any of the systems that properly take into account slider axis so Ive done all my modeling in solidworks and mostly 2D. Slider axis vs steering axis is very different on the evo and I'd rather have it right in 2D than wrong in 3D
I need to refine my rear model though I'll probably just stick to a 3D point map in solidworks.
I need to refine my rear model though I'll probably just stick to a 3D point map in solidworks.
#4
Evolved Member
I did it in lotus... will measure the final effect on the car to confirm if it really accurate..
#5
Evolving Member
Modeling in general is tough, a tiny mis-measurement will cause a huge effect in measured bump steer etc in a model.
What process was used to come up with the point in these models? I have done 2d before with the car perfectly level and using a plumb-bob to measure the mounting points on the floor....but for 3d i feel you would need scanning or something to be accurate enough to make setup changes on it.
Zack
What process was used to come up with the point in these models? I have done 2d before with the car perfectly level and using a plumb-bob to measure the mounting points on the floor....but for 3d i feel you would need scanning or something to be accurate enough to make setup changes on it.
Zack
#6
Evolved Member
I measured it first with the car level on a big lift used for alignment... then I removed the front subframe and knuckles and measured them individually... as some things are hard to measure on the car.. like steering rack height..
Ideally you need something like faro arm..
Ideally you need something like faro arm..
Trending Topics
#10
Evolved Member
iTrader: (8)
I pulled all the frame pivot points and then measured all the components individually. I don't think I ever got to measuring out the rear hub carriers though.
Similar to Dallas, I just did it in Solidworks. You can define sketches, planes, and reference points to pick up all the geometry aspects and use the 3D motion simulation to generate plots of everything.
FWIW, I think you kind of end up in this fantasy land though with it. You can generate all kinds of data with a chassis based reference pretty easily. But trying to mimic chassis motion (front and rear simultaneously) and then turn it into ground relative data to truly under stand what the tire sees...that starts to get way more time consuming unless you come up with a way to automate the whole process.
That's likely where the mentality of just making it adjustable is probably best. You can iterate for days and probably get no where...or setup some decent test courses in car and start adjusting until it works the best.
Similar to Dallas, I just did it in Solidworks. You can define sketches, planes, and reference points to pick up all the geometry aspects and use the 3D motion simulation to generate plots of everything.
FWIW, I think you kind of end up in this fantasy land though with it. You can generate all kinds of data with a chassis based reference pretty easily. But trying to mimic chassis motion (front and rear simultaneously) and then turn it into ground relative data to truly under stand what the tire sees...that starts to get way more time consuming unless you come up with a way to automate the whole process.
That's likely where the mentality of just making it adjustable is probably best. You can iterate for days and probably get no where...or setup some decent test courses in car and start adjusting until it works the best.
#11
Evolved Member
after doing the simulation in lotus susprog I think people are way overthinking this one.. Evo needs more castor, preferably around 8 deg, maybe a bit more track up front if you can squeeze it inside the fenders, and keep the steering axis inclination to a minimum, i.e. not increasing camber by moving top mounts but modifying the hub/damper connection instead.
Roll centre relocation, by dropping the outer ball joint, is a nice mod also. The ammount will depend on your running height and the wheel/tire combo you are using. I would say that on a stock upright 25 to 35 mm drop is ok, depending on how you do it. Any more and it is tricky without a custom upright, such as what Dallas does.
What a 3d sim will do nicely is tell you how much you have to drop the outer tie rod to get minimal bump steer. You have to remember that with a drop in the outer A arm ball joint you need to drop the tie rod too... and the more you drop less stiff it will be.. (again, unless custom upright)
Since evo uses a mcpherson, there is no real magic that can happen from manipulating the suspension in 3d sim.. compared to a proper double A arm...
Roll centre relocation, by dropping the outer ball joint, is a nice mod also. The ammount will depend on your running height and the wheel/tire combo you are using. I would say that on a stock upright 25 to 35 mm drop is ok, depending on how you do it. Any more and it is tricky without a custom upright, such as what Dallas does.
What a 3d sim will do nicely is tell you how much you have to drop the outer tie rod to get minimal bump steer. You have to remember that with a drop in the outer A arm ball joint you need to drop the tie rod too... and the more you drop less stiff it will be.. (again, unless custom upright)
Since evo uses a mcpherson, there is no real magic that can happen from manipulating the suspension in 3d sim.. compared to a proper double A arm...
#12
EvoM Guru
iTrader: (1)
What Kiki said is pretty much my take on suspension. My goal is usually try to get back close to OEM geometry but with more caster and camber. I did make another thread talking about caster, ball joint, and bumpsteer. There I showed the a simple table for what tie rod to run with which ball joint based on how much caster you have.
Kiki, one thing you lose with going the low steering axis route is tire clearance and slider axis. Slider axis is what determines your camber gain on a macstrut car along with your instant roll center (goes without saying, shorter slider axis = more camber gain but more RC migration). I typically run quite a bit SAI because I need to fit 295s under OEM pulled fenders.
Its always a balance of the variables, but as I've made parts and played with lots of odd and/or mathematically correct setups, its all kinda out the window on what exactly is best. Two extremely different setups can often produce nearly identical lap times just in different ways. Even next year I'm going to try a "no front swaybar" setup because reasons
Kiki, one thing you lose with going the low steering axis route is tire clearance and slider axis. Slider axis is what determines your camber gain on a macstrut car along with your instant roll center (goes without saying, shorter slider axis = more camber gain but more RC migration). I typically run quite a bit SAI because I need to fit 295s under OEM pulled fenders.
Its always a balance of the variables, but as I've made parts and played with lots of odd and/or mathematically correct setups, its all kinda out the window on what exactly is best. Two extremely different setups can often produce nearly identical lap times just in different ways. Even next year I'm going to try a "no front swaybar" setup because reasons
#13
The SusProg3D program is set up to do full car suspension simulation of the front and the rear together I believe. I believe you can pitch, roll, etc and all 4 corners react together. It also can solve geometry problems for you I think. For example, in the pics I posted, my friend used it on his 190E to solve OEM bump steer issue. It basically told him where to move one of the rear links to make bumpsteer nearly zero for normal rear wheel movement. That would be impossible to do manually on a multilink setup like the rear of our cars. There are infinite possibilities of where to move each pivot.
So sounds like no one has really went that in depth on the rear of the EVO 7/8/9 platform? Besides probably the big name aftermarkets like Wisefab or Brypar.
So sounds like no one has really went that in depth on the rear of the EVO 7/8/9 platform? Besides probably the big name aftermarkets like Wisefab or Brypar.
#15
EvoM Guru
iTrader: (1)
The SusProg3D program is set up to do full car suspension simulation of the front and the rear together I believe. I believe you can pitch, roll, etc and all 4 corners react together. It also can solve geometry problems for you I think. For example, in the pics I posted, my friend used it on his 190E to solve OEM bump steer issue. It basically told him where to move one of the rear links to make bumpsteer nearly zero for normal rear wheel movement. That would be impossible to do manually on a multilink setup like the rear of our cars. There are infinite possibilities of where to move each pivot.
So sounds like no one has really went that in depth on the rear of the EVO 7/8/9 platform? Besides probably the big name aftermarkets like Wisefab or Brypar.
So sounds like no one has really went that in depth on the rear of the EVO 7/8/9 platform? Besides probably the big name aftermarkets like Wisefab or Brypar.