Welcome to the North American Subaru Impreza Owners Club Saturday December 27, 2014
Home Forums WikiNASIOC Products Store Modifications Upgrade Garage
Click here to visit TireRack
Brakes & Suspension Forum sponsored by The Tire Rack

Losing traction? Need new tires?
Click here to visit the NASIOC Upgrade Garage...
Here you can view your subscribed threads, work with private messages and edit your profile and preferences Home Registration is free! Visit the NASIOC Store NASIOC Rules Search Find other members Frequently Asked Questions Calendar Archive NASIOC Upgrade Garage Logout
Go Back   NASIOC > NASIOC Technical > Brakes, Steering & Suspension

Welcome to NASIOC - The world's largest online community for Subaru enthusiasts!
Welcome to the NASIOC.com Subaru forum.

You are currently viewing our forum as a guest, which gives you limited access to view most discussions and access our other features. By joining our community, free of charge, you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content and access many other special features. Registration is free, fast and simple, so please join our community today!

If you have any problems with the registration process or your account login, please contact us.
* Registered users of the site do not see these ads.
Reply
 
Thread Tools Display Modes
Old 10-30-2012, 12:29 AM   #1
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default Why Slamming Your Car is a Good Thing

Now that I have your attention...

I will talk about ride height and a below ground roll center in a bit, but first, I'm currently working on my senior design project which is a front upright design for the 5x114 GD cars. I am looking to correct various aspects of the suspension geometry by having an adjustable ball joint location (while being able to use the OEM ball joint) and tie rod location. This will allow for camber (independent of KPI), bump steer, and roll center to be tuned according to the vehicle's ride height, rather than designing it around a specific ride height. Also, if you were to design an upright, what features would you like to see?

However, I need a little help finding some information. I have rudimentary suspension nodes and elements which I've modeled in WinGeo, but I don't trust my measurements and it was also taken with my car which already has the Whiteline RCA kit installed. I've been searching, but could not find the camber curve and ride steer plots which I thought used to be here somewhere.

If anyone has them and can help me out, I'm looking for the OEM curve plots as well as the suspension nodes if someone has them. Or if anyone has already modeled the suspension in WinGeo and has the .gem file, I could use that instead.

If you have this data available, please PM me.

For those who have automotive engineering/FSAE backgrounds I would also like some help. The more academic research I've been doing, the more I'm inclined to believe that the car should be lowered as much as possible, no matter what it does to the kinematic roll center. In fact, I believe the fact the kinematic roll center goes below ground can be ignored and that having it below ground and stay there is actually beneficial. I am ready to be flamed by those who regurgitate the misinformation they read on the internet

Here is my logic:

It doesn't matter where a single point in space is. We are trying to maximize cornering force by increasing contact patch load. While the roll center height does affect this, I have yet to see anyone on NASIOC/IWSTI to use the force-based roll center, EVER. However, the force-based roll center is not easily found. But using some simple logic, in the case of race cars, it can be assumed the majority of the cornering force is done by the outside wheel. In order to increase the total amount of cornering force available, you want to reduce lateral load transfer, which comes from having a lower CG, i.e. lowering the car. Also, by having a roll center below ground, there will be an anti-jacking force which will offset some of the effect of the larger roll moment. Also, an anti-roll bar can be used to increase roll stiffness.

However, in the case of the Impreza, I do believe the camber curve should be corrected after lowering a car to increase the contact patch, but nearly everyone says having a roll center (and the person is almost always talking about the KRC) below ground is bad. Those people also believe having the roll center migrate laterally is a real thing and it's also a bad thing. Does anyone have legitimate, intelligent arguments as to why the roll center should absolutely not be below ground?
* Registered users of the site do not see these ads.

Last edited by EarlQHan; 10-31-2012 at 08:08 PM.
EarlQHan is offline   Reply With Quote
Old 10-30-2012, 12:34 AM   #2
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default

Oh and for those who want to help me by critiquing my upright design, this is my initial design without optimizing the geometry yet.


Last edited by EarlQHan; 10-31-2012 at 08:08 PM.
EarlQHan is offline   Reply With Quote
Old 10-31-2012, 08:10 PM   #3
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default

Hopefully the new title will get somebody in this discussion... sorry for the deceptive title.
EarlQHan is offline   Reply With Quote
Old 10-31-2012, 09:55 PM   #4
rbaldi
Scooby Newbie
 
Member#: 312567
Join Date: Mar 2012
Vehicle:
2012 STi

Default

I can't help you on the geometry part, but I work with race car suspension on a daily basis, so I think I can be of some help there.


To put it bluntly: no. There is no reason that the "roll center" should not be below ground, in fact, there are situations where it is advantageous. A below ground "roll center" (more accurately, n-lines that point at a downward angle (see below for an explanation)) will create a jacking force that tries to lower the sprung mass under cornering load and create less geometric anti-roll, which must be reacted with with bigger springs/anti-roll bars, for the same roll gradient.

Now, a rant about roll centers:

The so called "kinematic roll center" does not exist. There, I said it. Yes, there is a point on the sprung mass where lateral loads can be applied without creating any rolling moment - but this breaks down as soon as you account for deflection - something that, by nature, happens in a suspension system. The vehicle does not actually roll about this point, or in 3D, the "roll axis" (a line between kinematic roll centers of the front and rear suspensions), and it does not capture the effects of jacking forces (the vertical component of the reaction of contact patch loads into the sprung mass).

The force based approach is simpler, can be applied to any suspension type, and is independent of other vehicle parameters. A simple free body diagram will show that slope of the line between the contact patch and the instant center of the suspension (commonly referred to a n-lines, because the line is always normal to the motion of the contact patch) is what determines the geometric anti-roll and jacking properties. Nothing else.

Here's an example - 2 cars both have a kinematic roll center of 10" above ground. One car has a track width of 50", the other, 100". Do these 2 vehicles have the same geometric anti-roll properties?

In addition, the kinematic roll center model breaks down rapidly when vehicles are not symmetric. The force based approach considers each wheel independently, so does not have this issue. </rant>

Here is my postulation: Roll centers have the highest ratio of words typed on the internet to influence on the vehicle's lap time. Your thoughts?


My advice to you is to let the "roll center" be where ever it needs to be in order to achieve the other vehicle parameters that you think you want (I say "think" because, I assume, you don't have any tire data so you're really just guessing). Without moving inboard suspension points, you will be somewhat limited in what you can do, at least in terms of camber gain, but you already know this. I guarantee that a lower vehicle CG, minimizing KPI, having some camber gain in roll, getting enough caster to override the effects of KPI and have some camber gain with steering, and decent packaging so that components can be light and stiff is way more important than the "roll center" location, at least on a vehicle that is not very aero platform sensitive.

Last edited by rbaldi; 10-31-2012 at 10:02 PM.
rbaldi is offline   Reply With Quote
Old 10-31-2012, 09:59 PM   #5
codesoccer
Scooby Specialist
 
Member#: 179573
Join Date: May 2008
Chapter/Region: South East
Location: South Florida
Vehicle:
2002 MBP Bugeye
Flip Flop Tuned on corn

Default

This is interesting. Subscribed.
codesoccer is offline   Reply With Quote
Old 10-31-2012, 10:02 PM   #6
Spenk
Scooby Guru
 
Member#: 7164
Join Date: Jun 2001
Location: Arizona Bay
Vehicle:
2000 Miata LS
Evo Orange Mica (1/644)

Default

Quote:
Originally Posted by rbaldi View Post

Here is my postulation: Roll centers have the highest ratio of words typed on the internet to importance of the vehicle lap time. Your thoughts?


My advice to you is to let the "roll center" be where ever it needs to be in order to achieve the other vehicle parameters that you think you want (I say "think" because, I assume, you don't have any tire data so you're really just guessing). Without moving inboard suspension points, you will be somewhat limited in what you can do, at least in terms of camber gain, but you already know this. I guarantee that a lower vehicle CG, minimizing KPI, having some camber gain in roll, getting enough caster to override the effects of KPI and have some camber gain with steering, and decent packaging so that components can be light and stiff is way more important than the "roll center" location, at least on a vehicle that is not very aero platform sensitive.
Beautifully stated.

Bless you, kind and intelligent sir!
Spenk is offline   Reply With Quote
Old 11-01-2012, 04:05 PM   #7
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default

Quote:
Originally Posted by rbaldi View Post
My advice to you is to let the "roll center" be where ever it needs to be in order to achieve the other vehicle parameters that you think you want (I say "think" because, I assume, you don't have any tire data so you're really just guessing). Without moving inboard suspension points, you will be somewhat limited in what you can do, at least in terms of camber gain, but you already know this. I guarantee that a lower vehicle CG, minimizing KPI, having some camber gain in roll, getting enough caster to override the effects of KPI and have some camber gain with steering, and decent packaging so that components can be light and stiff is way more important than the "roll center" location, at least on a vehicle that is not very aero platform sensitive.
Thank you. Finally someone who knows what they are talking about. I'm not sure why everyone places so much importance to the KRC and where it is located. I think it is the most overused and least understood term in vehicle dynamics among the general population. I think people get too hung up on the kinematics and forget the whole point of the suspension is to get the maximum out of the tire.

I had the same thoughts on the design outcomes as well. I don't have tire data, but all tire curves exhibit the same general characteristics: you want as much normal force as possible, and load transfer reduces the total cornering force available. This is also why I built in camber adjustment via lateral ball joint movement, then you have an effective increase in track width, further reducing LLT, and it allows you to reduce the KPI by using camber plates. I do need to FEA it to see if it can handle the loading conditions, but there's always the possibility of other materials.
EarlQHan is offline   Reply With Quote
Old 11-01-2012, 05:22 PM   #8
rbaldi
Scooby Newbie
 
Member#: 312567
Join Date: Mar 2012
Vehicle:
2012 STi

Default

Agreed on all points. With the size of tires you can fit, and the overall mass of the vehicle, it's pretty safe to assume that the tire are load sensitive within all realistic loading conditions, so the best bang for your buck is to do everything you can to minimize load transfer.

It's hard to tell from your picture how you plan to adjust camber , but I'm assuming it's by moving the lower ball joint laterally? If so, keep an eye on how much the scrub radius and spindle offset increase when you increase negative camber (Maybe not a bad thing on this car, but can make for some really heavy steering on a car without power assist).
rbaldi is offline   Reply With Quote
Old 11-02-2012, 06:04 PM   #9
kos
Scooby Specialist
 
Member#: 118334
Join Date: Jun 2006
Chapter/Region: SWIC
Location: Bouncing off Kerbs
Vehicle:
2005 AW RS + 2006
Time Attack WRX

Default

Earl, you're alive!
kos is offline   Reply With Quote
Old 11-02-2012, 07:41 PM   #10
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default

Quote:
Originally Posted by kos View Post
Earl, you're alive!
lol. Yup, as you can see, I've been nerding it up for the last few years. Last semester of undergrad, then it's off to England next year for my master's.
EarlQHan is offline   Reply With Quote
Old 11-02-2012, 07:45 PM   #11
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default

Quote:
Originally Posted by rbaldi View Post
Agreed on all points. With the size of tires you can fit, and the overall mass of the vehicle, it's pretty safe to assume that the tire are load sensitive within all realistic loading conditions, so the best bang for your buck is to do everything you can to minimize load transfer.

It's hard to tell from your picture how you plan to adjust camber , but I'm assuming it's by moving the lower ball joint laterally? If so, keep an eye on how much the scrub radius and spindle offset increase when you increase negative camber (Maybe not a bad thing on this car, but can make for some really heavy steering on a car without power assist).
Correct, the camber would be adjusted in that fashion. The ball joint can also be located vertically using shim stacks. Everything for the geometry will be accounted for in the final design, but I wanted to create an upright that would allow for the geometry to be corrected for virtually any lowered ride height, rather than setting a ride height window for the users. It will also be able to use OEM ball joints/tie rods or use a solid rod instead of a ball joint if the user wishes to put a monoball in the control arm and use an aftermarket rod end so the tie rod can be mounted in double shear.
EarlQHan is offline   Reply With Quote
Old 11-02-2012, 11:24 PM   #12
rbaldi
Scooby Newbie
 
Member#: 312567
Join Date: Mar 2012
Vehicle:
2012 STi

Default

Sounds like you have everything under control. Keep us informed of your progress, I'm interested to see what your final product looks like and what compromises you make, because after all, engineering is all about compromises.
rbaldi is offline   Reply With Quote
Old 11-03-2012, 10:33 AM   #13
Scooby921
Merci Buckets
Moderator
 
Member#: 88606
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Clarkston
Vehicle:
2011 GMC Sierra
'13 JCW

Default

Quote:
Originally Posted by EarlQHan View Post
Oh and for those who want to help me by critiquing my upright design, this is my initial design without optimizing the geometry yet.

From an engineering side...

You're going to want a lot of those small radii/fillets to be much larger. In a component like this the loads/stresses can be quite large. Small radii corners will concentrate stresses. Is FEA any part of your design project?


<---has designed quite a few automotive engine mount brackets and did a senior thesis on magnesium brackets
Scooby921 is offline   Reply With Quote
Old 11-03-2012, 10:55 AM   #14
Counterfit
Scooby Guru
 
Member#: 124254
Join Date: Aug 2006
Chapter/Region: NESIC
Location: Rhode Island
Vehicle:
2006 06 SGM Slowbaru
"The Scoobinator"

Default

Just sayin'.
Quote:
Originally Posted by EarlQHan
I do need to FEA it to see if it can handle the loading conditions, but there's always the possibility of other materials.
Counterfit is offline   Reply With Quote
Old 11-03-2012, 03:04 PM   #15
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default

Quote:
Originally Posted by Counterfit View Post
Just sayin'.
lol.

Scooby921 - I will do FEA later on down the line. I am aware of the stress concentrations on re-entrant corners. I was just modeling the upright as a concept for now. I don't have the loading conditions, so if anyone has data logs they can supply, I can calculate the loads. Otherwise I will have to make an educated guess. I assume worst case scenario is 2G cornering with 1G bump and use a safety factor of 1.5.

I am doing the modeling in SolidWorks, but I will be doing the FEA with the ANSYS 13.0 plug-in. Although it's not mechanically complex, since it is an assembly and I need to FEA it as such, I'm worried I won't be able to get the contact sets set up properly to accurately simulate the fasteners in shear. Any tips?
EarlQHan is offline   Reply With Quote
Old 11-03-2012, 03:10 PM   #16
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default

Oh and to anyone who still thinks the kinematic roll center needs to be above ground to shorten the moment arm, thereby reducing the roll moment, because the chassis rolls about the KRC, please read:

http://www.neohio-scca.org/comp_clin...namics2007.pdf

Long story short: the KRC is irrelevant
EarlQHan is offline   Reply With Quote
Old 11-03-2012, 05:11 PM   #17
kos
Scooby Specialist
 
Member#: 118334
Join Date: Jun 2006
Chapter/Region: SWIC
Location: Bouncing off Kerbs
Vehicle:
2005 AW RS + 2006
Time Attack WRX

Default

Quote:
Originally Posted by EarlQHan View Post
lol. Yup, as you can see, I've been nerding it up for the last few years. Last semester of undergrad, then it's off to England next year for my master's.
Nothing wrong with that man. Awesome to see you doing good things. Finish your damn car while you're at it!

Have you seen the LIC modified knuckles? Those are a pretty impressive setup.

http://www.licmotorsports.com/produc...ti_lic-0731091
kos is offline   Reply With Quote
Old 11-03-2012, 08:10 PM   #18
EarlQHan
Scooby Specialist
 
Member#: 90215
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Where am I not?
Default

Quote:
Originally Posted by kos View Post
Nothing wrong with that man. Awesome to see you doing good things. Finish your damn car while you're at it!

Have you seen the LIC modified knuckles? Those are a pretty impressive setup.

http://www.licmotorsports.com/produc...ti_lic-0731091
I wish I could! It's been further disassembled so I could do this project. Plus, the cost of living in England is expensive so I need to save money. I did see the LIC uprights as well as the MSI when I was doing research. Both impressive and I did get good ideas from seeing both.

These are just my opinions, but the LIC puts the tie rod in double shear which is great. However, with the LIC upright adds weight, rather than removing it. As for the MSI design, I like how it can adapted for different brakes and it allows for ball joint and tie rod height to be adjusted. However, it doesn't allow for use of OEM ball joints/tie rods (as far as I can tell). Also, I can't be 100% from looking at the pictures, but it looks like the MSI would be weaker and softer than the OEM upright. Both have a limited ride height window and neither have lateral adjustment for increased static camber without affecting the KPI, so I'm just refining a design that's been come up by the hard work of these guys.

(BTW: If either LIC or MSI see this please feel free to critique my design idea. I hope you don't feel the need to defend your uprights. They're both great solutions in my opinion, as I know there are limits of practicality for business purposes, whereas I'm doing academic research)
EarlQHan is offline   Reply With Quote
Old 11-04-2012, 01:08 AM   #19
rbaldi
Scooby Newbie
 
Member#: 312567
Join Date: Mar 2012
Vehicle:
2012 STi

Default

Quote:
Originally Posted by EarlQHan View Post
lol.

Scooby921 - I will do FEA later on down the line. I am aware of the stress concentrations on re-entrant corners. I was just modeling the upright as a concept for now. I don't have the loading conditions, so if anyone has data logs they can supply, I can calculate the loads. Otherwise I will have to make an educated guess. I assume worst case scenario is 2G cornering with 1G bump and use a safety factor of 1.5.

I am doing the modeling in SolidWorks, but I will be doing the FEA with the ANSYS 13.0 plug-in. Although it's not mechanically complex, since it is an assembly and I need to FEA it as such, I'm worried I won't be able to get the contact sets set up properly to accurately simulate the fasteners in shear. Any tips?

Large high speed pothole impacts can be upwards of 10G, especially with a short sidewall tire. Also, don't forget about braking loads - both longitudinal and torque reactions. I would recommend doing most of your stress work with a worse case scenario or 2; you are highly unlikely to hit a 10G bump while cornering at 2G and braking at 1G, so that may be unrealistic. If it were me, I would make sure the upright wasn't going to catastrophically fail at 10G bump (a little localized yielding may be okay here, but you need to decide what you want as your "fuse"-I would think you want a suspension link to buckle before you fail the upright), and use 3G bump, 1.5-1.8G cornering, .6-.8G cornering as a worst case, with whatever safety factor you deem reasonable given your faith in your analysis techniques. I would use another set of loads for looking at deflections, probably 1.5G cornering. It also may be prudent to make sure you don't get excessive toe compliance from brake torque ( I have seen this on some crappy FSAE cars). What are your camber and toe compliance goals? I would think there would be a lot of hand waving here because most of your compliance will come from the rubber bushings.

I have never used the ANSYS plugin for Solidworks, but in ANSYS Workbench, it is really easy to model part to part contact, you can rigidly connect nodes in the contact areas, or specify a spring rate on the connection (a little more accurate, but requires test data or a lot of experience to model correctly). It's a good idea to try to analyze the assembly, but just feeding loads into the bearing areas will be good enough for all but your final iterations, just ignore the stresses at the load sites and constraints and check the bearing stresses via hand calcs. This method works really well on double shear joints (I have predicted within 5% to actual test to failure on a test bench data), but gets a little dodgy on single shear joints because of the addition of the bending moment. But hey, that's what factors of safety are for, right?
rbaldi is offline   Reply With Quote
Old 11-04-2012, 01:21 AM   #20
rbaldi
Scooby Newbie
 
Member#: 312567
Join Date: Mar 2012
Vehicle:
2012 STi

Default

I'm a little surprised that the kinematic-roll-center-doesn't-exist/matter/below-ground-is-okay notion hasn't caused an uproar yet; after all, NASIOC is not the place for facts
rbaldi is offline   Reply With Quote
Old 11-04-2012, 01:07 AM   #21
Purpose_Built
Scooby Specialist
 
Member#: 222879
Join Date: Sep 2009
Chapter/Region: Tri-State
Location: NJ
Vehicle:
2003 WRX 60-1/td06h
Hybrid

Default

Durp, Durp, Durp. Thats how i feel as i read this. Can't wait to finish school!!!
Purpose_Built is online now   Reply With Quote
Old 11-04-2012, 01:23 AM   #22
Counterfit
Scooby Guru
 
Member#: 124254
Join Date: Aug 2006
Chapter/Region: NESIC
Location: Rhode Island
Vehicle:
2006 06 SGM Slowbaru
"The Scoobinator"

Default

Quote:
Originally Posted by rbaldi
I'm a little surprised that the kinematic-roll-center-doesn't-exist/matter/below-ground-is-okay notion hasn't caused an uproar yet; after all, NASIOC is not the place for facts
Massive blocks of detailed engineering talk usually act as idiot repellent around here.
Counterfit is offline   Reply With Quote
Old 11-04-2012, 01:35 AM   #23
jdblock
Scooby Specialist
 
Member#: 235211
Join Date: Jan 2010
Vehicle:
2004 STi
WRB

Default

Any worries about the axle popping out of the transmission when adding negative camber via the ball joint?
jdblock is offline   Reply With Quote
Old 11-04-2012, 04:54 AM   #24
vicious_fishes
Scooby Specialist
 
Member#: 158864
Join Date: Sep 2007
Location: Aussieland
Vehicle:
2003 GD w/EZ30D swap

Default

LIC already do one of these don't they?

http://www.licmotorsports.com/produc...i_lic-08207412
vicious_fishes is offline   Reply With Quote
Old 11-04-2012, 05:18 AM   #25
KillerRabbit
Scooby Specialist
 
Member#: 184772
Join Date: Jul 2008
Chapter/Region: MWSOC
Location: Chicago, IL
Vehicle:
2004 Crazy Slow WRX
World Rally Blue

Default

Very interested, subscribed for further info
KillerRabbit is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -4. The time now is 10:41 PM.


Powered by vBulletin® Version 3.7.0
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Powered by Searchlight © 2014 Axivo Inc.
Copyright ©1999 - 2014, North American Subaru Impreza Owners Club, Inc.