Suspension analysis poll

[John Block]

That time of year again when suspension design starts popping up.

If you had a computer application for suspension analysis, what would you want to get in the way of output? Now more fun, what do you think your advisor or Professor would want you to get from this magic application?

[John Block]

That time of year again when suspension design starts popping up.

If you had a computer application for suspension analysis, what would you want to get in the way of output? Now more fun, what do you think your advisor or Professor would want you to get from this magic application?

[exFSAE]

I want a button that says "optimize" and you click it and it makes the fastest car.

[Rear Admiral]

agreed ^

adviser? What's that?

[Anvit Garg]

A professor or faculty adviser would probably want some sort of sensitivity analysis.

Possibly a report of the effect that changing spring rates does, ride height, and other tangible settings.

I believe theoretical changes such as roll center locations, instant center lengths, scrub, etc are less appealing to an adviser who is helping out of sympathy, and has no interest in your project.

:_(

[js10coastr]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by John Block:
That time of year again when suspension design starts popping up.

If you had a computer application for suspension analysis, what would you want to get in the way of output? Now more fun, what do you think your advisor or Professor would want you to get from this magic application? </div></BLOCKQUOTE>

I'm a bit biased but, see OptimumK + (nested for loop * simulation software). The for loop would allow you to learn a bit more than an optimization routine... DOE would work just as well.

I'm not sure advisor wise, but as an alum working in motorsports I'd want students to understand the parameters and what effects they have, and why one would change these parameters. For example, why different cars prefer different roll center heights... tradeoffs in IC lengths, spring rates etc etc.

[scott_rfr]

From a non biased approach I agree with js10coastr. OptimumK--&gt;Matlab and you can do a lot. I was once lost trying to create a full transient sim with all kinds of stuff. Stick to the basics, MMM or steady-state, remember steady-state sim does not mean yaw moment has to equal zero. Some look up functions, loops, and you can run lots of setup variations fast and see what is going to make the big difference. This really has helped setup on our current car in testing. Plus you have the added benefit of really learning whats going on. You might think you understand this stuff then try writing a sim and you really start to see where the blanks are and fill them in.

Scott
Rutgers SAE

[Jersey Tom]

Eh. Weird question to ask.

As for the faculty (or alumni) adviser part.. I'd say I wouldn't want the software available at all. OptimumK is enough.

It's a huge benefit in industry (race-related or otherwise) to have skills in practical programming, and modeling real world systems in varying degrees of complexity. They're must-have skills really. You don't get that as much with an off-the-shelf solution. Plus, you generally can't modify the off-the-shelf solutions.

From an "adviser" standpoint I'd want students to learn how to:
<LI>Write good clean code, quickly. Preferably including object-oriented methods, error catching, GUIs, etc
<LI>Learn how to take a complicated physical system (racecar), identify what's most important, and how to model it on a computer. IMO it's a lot better sometimes to have a few small programs that do specific things (just look at tire data, or aero data, or kinematic data, or steady state cornering) as opposed to trying to make the all-encompassing simulation.
<LI>Validate their model with real world testing

Let each team figure out what's most valuable and meaningful to them. In '02 or '03 the CU guys had a license of ADAMS.. which was worthless since we didn't know nearly as much about the vehicle system as we needed to, in order to make it worth while. On the other hand, a simple trimmed cornering sim may be of limited use for a team or organization at the highest level of competition. There's lots of in-between. Maybe some teams want to use look-up data for K&C, maybe others want to solve it with a kinematics sim. Maybe some teams don't even know what K&C testing is. Etc etc.

What might be more helpful would be brief primers on specialized stuff or small things you might miss. Stuff like...

<LI>How empirical tire models work, overview of a few different types, strengths and weaknesses.. maybe F&M overview in general. How many guys here understand the effect of plysteer?
<LI>Why aligning torque has a big impact on understeer and not just steering effort (knew a guy on a Cup team who forgot this point..)
<LI>How tire loaded radius data can either have almost no impact, or massive impact on your vehicle handling (even on non-aero cars).
<LI>True explanation of why frame rigidity is important, and how much you need with numerical examples (most explanations are very hand-wavy)
<LI>How various differential types work and the different ways one might model them
<LI>What K&C testing is, why it's useful even at this level


Stuff like that. Get the students primed on the bits and pieces, and leave it to them to assemble it into something meaningful. Kinda like a Claude seminar... but instead of just getting the slides with some hand-written notes that may or may not make sense after you scribble them... a handful of available "white papers" describing things to a fair level of detail, but also in a digestible format. Almost what I attempt to do with some of my SCCA FB devo blog stuff, but more in-depth. And less profanity. Similar to the OptG "Tech Tips" that Matt and/or whoever had been putting together back in the day.

Just my $0.02.