|
Brakes & Suspension Forum sponsored by The Tire Rack |
|
|
|
Thread Tools | Display Modes |
11-04-2012, 05:12 PM | #26 | |
Scooby Specialist
Member#: 20952
Join Date: Jul 2002
Chapter/Region:
MAIC
Location: Baltimore, MD
Vehicle:BRZ & Datsun White & White |
Quote:
- Andrew
* Registered users of the site do not see these ads.
|
|
11-04-2012, 06:35 PM | #27 | |
Scooby Specialist
Member#: 90215
Join Date: Jun 2005
Chapter/Region:
MWSOC
Location: Where am I not?
|
Quote:
Thanks for the tips on the loading conditions. As for ANSYS, I'm very limited with its use in Workbench. I'm in a 500 level FEA class, but we've mostly been setting up everything manually through the product launcher as the focus is on the theory behind FEA. So a lot of matrix algebra All my previous FEA experience has been with SolidWorks Simulation with very limited use on assembly analysis. Since it's out of my scope to do account for bushing deflections and there's no way for me to get proper K&C data, I'm going to try and get as much camber gain/reduce camber loss as possible, and try get a linear ride-steer curve. It's really limited to what WinGeo tells me by moving the pick-up points within the packaging constraints. I hate WinGeo... if any of you have used it before, you understand the pains caused by the stupid "loop of death." I can't wait to start my masters and start using ADAMS. I was going to model heat transfer and account for the stresses on the upright (both convection and conduction) from braking, but I hadn't thought of brake torque. Do you know how I could best model that for analysis? jdblock - I am not worried, because there isn't going to be that much adjustment. Maybe an inch or two. I can also modify the pick up points for the strut to correct the KPI. And this is just for research, I'm not actually going to machine this... probably |
|
11-04-2012, 06:41 PM | #28 |
Scooby Specialist
Member#: 90215
Join Date: Jun 2005
Chapter/Region:
MWSOC
Location: Where am I not?
|
|
11-04-2012, 08:41 PM | #29 |
Scooby Specialist
Member#: 20952
Join Date: Jul 2002
Chapter/Region:
MAIC
Location: Baltimore, MD
Vehicle:BRZ & Datsun White & White |
|
11-05-2012, 09:19 AM | #30 |
Former Vendor
Member#: 95419
Join Date: Sep 2005
Chapter/Region:
E. Canada
Location: montreal
|
I will chime in here before things go crazy out of proportion.
We have about 120 hours of FEA done on our model, it is significantly stronger than the OE part to the point that we have no worries about using it on a gravel car. Our current revision will take a lot higher loads than the cast IRON oe part, between 45 - 150% more depending what load and angle, etc. We do have a gravel version that does not have an adjustable steering arm for increased strength. The steering arm for this model is mounted on a plate were the pin for the bearing goes on the lower pick-up point. We do have a version that uses the OE ball joints as well as a steering arm that is mounted in double sheer. As far as the OE all joint, why would you want to use this when you are going so far as to use a billet upright? It is not worth it. Put a proper bearing in there and change the front arms. I will also say, the less adjustment in the part the better. Our non-adjustable models are MUCH stronger than the adjustable ones. -mark |
11-05-2012, 09:30 AM | #31 | |
Scooby Newbie
Member#: 207569
Join Date: Apr 2009
Chapter/Region:
International
Location: Adelaide, South Australia
Vehicle:MY04 Forester XT Red |
I'm following with interest, thanks for the input Mark
Quote:
Just a thought. I agree with ditching the OEM ball joint, surely it's a weak link, why leave it in place. |
|
11-05-2012, 09:35 AM | #32 |
Scooby Guru
Member#: 158864
Join Date: Sep 2007
Location: Aussieland
Vehicle:2003 SOHC destroker |
i know i'm repeating myself here but again, what makes yours better than LIC's or MSI's or whoever else?
no better/different = no motive to buy |
11-05-2012, 09:39 AM | #33 |
Scooby Newbie
Member#: 207569
Join Date: Apr 2009
Chapter/Region:
International
Location: Adelaide, South Australia
Vehicle:MY04 Forester XT Red |
|
11-05-2012, 09:43 AM | #34 |
Scooby Specialist
Member#: 90215
Join Date: Jun 2005
Chapter/Region:
MWSOC
Location: Where am I not?
|
|
11-05-2012, 09:51 AM | #35 | |
Scooby Specialist
Member#: 90215
Join Date: Jun 2005
Chapter/Region:
MWSOC
Location: Where am I not?
|
Quote:
Two things: 1. would you mind explaining how you set up your loading conditions for FEA? 2. could I get a picture of a fully-assembled tarmac version of your upright in the specification you mentioned? I haven't been able to find one yet. Last edited by EarlQHan; 11-05-2012 at 09:59 AM. |
|
11-05-2012, 01:12 PM | #36 |
Scooby Guru
Member#: 306478
Join Date: Jan 2012
Chapter/Region:
South East
Location: chattanooga, TN
Vehicle:2016 Focus RS stock |
Based off the title I thought it was gonna be something silly like "bc it lets me pick up chicks!"
|
11-05-2012, 03:39 PM | #37 |
Scooby Guru
Member#: 71875
Join Date: Oct 2004
Location: Montana
|
I just haven't felt like getting into this big internet argument.
Just because the roll center is an imaginary point and doesn't do as accurate of a job as forced based analysis doesn't mean it's a worthless term and should be ignored. If put a big ball joint extender on the car and keep everything else the same (ride height, alignment, spring rates etc), the car is going to handle differently and you will change the amount of roll. |
11-05-2012, 08:21 PM | #38 | |
Scooby Newbie
Member#: 312567
Join Date: Mar 2012
Vehicle:2012 STi |
Quote:
I feel your WinGeo pain; I've used it in DOS As far as mass-market suspension analysis programs go, ADAMS is my favorite. Optimum K is also pretty good, but is limited to only kinematics. I would say that ADAMS is the Catia or Unigraphics of kinematics/basic simulation programs, and Optimum K is more like Solidworks....if that comparison makes any sense. To model brake torque, just constrain your upright at the UBJ (well, where it bolts to the damper in this case), LBJ, and at the steering arm and apply loads at the brake caliper mounts. You just need to work backwards from the contact patch into the upright to figure out the loads - think about what would happen if the pads were welded to the rotor and you applied a longitudinal force at the contact patch - how would that force be reacted into the chassis? To figure out how much longitudinal force you need at the contact patch for a given deceleration, you will need rotor/pad/caliper geometry to figure out the brake bias. Then just move the force (net moment=0, and make sure you get your forces applied in the correct direction; it's easy to get something backwards here) from the contact patch to the pad/rotor interface, then to the brake caliper mounts on the upright. Here's a hint for calculating the effective radius of the pad friction force - the pad face centroid is a not-perfect, but really close approximation. The actual radius is effected by many things, including pad compound, clamp load, rotor rpm, temperature, and pad wear, etc; but that's a different thread. |
|
11-05-2012, 09:09 PM | #39 | |
Scooby Newbie
Member#: 312567
Join Date: Mar 2012
Vehicle:2012 STi |
Quote:
I haven't done any kinematic analysis on a Subaru, so this is just hunch, but I'm guessing that most of the "improvement" that comes from a roll center adjuster kit comes from fixing the bump steer curve. |
|
11-05-2012, 09:35 PM | #40 | |
Scooby Specialist
Member#: 48377
Join Date: Nov 2003
Chapter/Region:
MWSOC
Location: "They eat fish soaked in lye"
Vehicle:1996 Gutted, built XP class Impreza L |
Quote:
The only reason I would not want the roll center below ground is that it contributes to weight jacking. I don't know how many people here have an Impreza that has been on the verge of grip rolling, but mine was before I redesigned the subframes and suspension geometry to improve the dynamic response both in roll center and weight jacking. I highly recommend starting with "Race Car Vehicle Dynamics" by Milliken and designing from there. Designing anything based off internet forum advice is not the best approach. That said, I would plan on a few FEA case runs to account for different loading conditions such as- 3G bump with 2G cornering, 10G bump by itself, 3G cornering, 3G braking with 5G bump, and so on and so forth. Without knowing the final application of the car it's hard to tell what the limiting condition is. This is the method I used for the design of my suspension pieces. Good luck with your project. |
|
11-05-2012, 10:51 PM | #41 |
Scooby Specialist
Member#: 90215
Join Date: Jun 2005
Chapter/Region:
MWSOC
Location: Where am I not?
|
Thanks guys. Have RCVD, as well as a few of Dixon's books and use them frequently. There's just not enough time left in the semester to get into the nitty gritty of the geometry and I'm not aiming for the "perfect" solution. I'm taking 25 hours, as well as being the FSAE treasurer and compliance engineer, and working on top of all that (not much sleep these days ) I'm aiming for a solution that can be applied to a wide range of ride heights, since there's a business aspect to consider for the project. I was merely using the title to get some eyeballs in the thread to help out.
That being said, with textbooks, there's a conspicuous lack of information on vehicle dynamics, and response from a driver's perspective. Everyone models vehicle dynamics in terms of an open-loop system, but in reality, the driver closes the loop. I'm hoping I'll be able to utilize state-space equations to model a vehicle/driver system in order to create a simulation to get a better idea of how sensitive a driver is and how sensitive the vehicle needs to be because there are human limitations to reaction time and different driver preferences on feel. I think it could mathematically determine a driver's capabilities as well. A bit more out there, but the most important topic in my opinion and it is an underdeveloped topic in my opinion. Maybe it's because engineers don't like to think about the "loose nut." I'm hoping I can do this as the topic of my masters thesis next year. |
11-06-2012, 12:01 AM | #42 | |
Scooby Newbie
Member#: 312567
Join Date: Mar 2012
Vehicle:2012 STi |
Quote:
Agreed 100% on the RCVD recommendation, and since it sounds like you are on an FSAE team, the Optimum G seminars by Claude Rouelle are pretty good and affordable - ~36 hrs of classroom instruction crammed into 3 days for ~$500, just dont take everything he says as gospel, use it like any other resource. It may sound like a lot, but it's probably cheaper per hour than almost all respectably universities. And Claude has some good jokes too. |
|
11-06-2012, 12:25 AM | #43 |
Scooby Specialist
Member#: 5691
Join Date: Apr 2001
Chapter/Region:
BAIC
Location: Redwood City, CA
Vehicle:2001 2.5 RS Silverthorn |
Interesting discussion. Obviously MSI and LIC have spent a lot of time and effort working on the geometry designs. How about thinking about something totally different. Track Subaru's kill wheel bearings really fast. Track Subaru's overheat brakes super fast. What about a liquid cooled knuckle, cooling the knuckle would protect the bearings. Spraying liquid onto the rotors for cooling is illegal in most race series, but instead of taking TIC's approach of putting titanium spacers between the rotor and hub to reduce heat transfer between the rotor and hub, why not cool the hell out of the knuckle and try to use that as an approach to cool the rotor and thereby improve brake life as well?
I'm suspicious that an engineer is about to jump in kick my butt in terms of cooling capacities, but its something I've never seen done, seems really hard to implement on a castknuckle, but wouldn't be difficult on a machined knuckle. |
11-06-2012, 12:38 AM | #44 |
Scooby Newbie
Member#: 207569
Join Date: Apr 2009
Chapter/Region:
International
Location: Adelaide, South Australia
Vehicle:MY04 Forester XT Red |
Good idea, maybe just finning the hell out of it would help too, we used to cool engines that way.
|
11-06-2012, 03:59 AM | #45 | |
Scooby Guru
Member#: 124254
Join Date: Aug 2006
Chapter/Region:
NESIC
Location: Rhode Island
Vehicle:2006 06 SGM Slowbaru "The Scoobinator" |
Quote:
|
|
11-06-2012, 04:26 AM | #46 |
Scooby Newbie
Member#: 207569
Join Date: Apr 2009
Chapter/Region:
International
Location: Adelaide, South Australia
Vehicle:MY04 Forester XT Red |
|
11-13-2012, 09:31 PM | #47 |
Scooby Newbie
Member#: 194923
Join Date: Nov 2008
Location: south florida
Vehicle:2006 gg evo iX |
Here's a link from the dark side that you guys might enjoy..
http://forums.evolutionm.net/evo-tir...uantified.html |
11-29-2012, 01:52 PM | #48 |
Scooby Guru
Member#: 71875
Join Date: Oct 2004
Location: Montana
|
A real chassis engineer seems to think the roll center is still relevant. motoiq just published this:
http://www.motoiq.com/magazine_artic...ll-center.aspx |
11-29-2012, 04:12 PM | #49 |
Scooby Guru
Member#: 158864
Join Date: Sep 2007
Location: Aussieland
Vehicle:2003 SOHC destroker |
too sick actually. obviously missed it
if you were to start producing them, even if they're exactly the same as the LIC's then providing they don't cost the same fortune i think you'd have more than a few keen buyers. |
11-29-2012, 04:22 PM | #50 |
Scooby Newbie
Member#: 207569
Join Date: Apr 2009
Chapter/Region:
International
Location: Adelaide, South Australia
Vehicle:MY04 Forester XT Red |
|
Thread Tools | |
Display Modes | |
|
|