Some decent reading material

A while back, Neil Roberts (of Swift Engineering) was nice enough to send a copy of Think Fast over here. I told Neil I'd give him a more in depth review once I got a good chunk of spare time - but that hasn't quite happened yet!

From what I did have the chance to read, I'd recommend it. Similar in some regards to what you'd find in the Carroll Smith series - maybe has a bit more "why" than just "what," but still reads well without getting bogged down. Has a lot of stuff I agree with, and a few subtle points I'm not entirely on board with (a couple tire specifics). Definitely some good driver points that I'm becoming more a fan of now that I'm working directly with a team. I also particularly liked some points on tire side-slip drag, as well as the entire section on cheating.

As far as target audience goes, is it probably a good pick up for the FSAE, SCCA, or semi-professional racer? Absolutely. Is it going to bring about some tremendous "ah-ha!" moments for well-established pro stock car or pro open wheel engineer? Probably not. But hey, Neil's still working professionally on this stuff, guy can't give away too much.

I should write a book one of these days - after we've won at least one pro level championship, and after I'm retired from competitive racing. If/when I do, it will probably be as thick as RCVD - or much more so. I want to have something for everyone:

1. "Underclassman level" - the basics .Qualitative over quantitative. An important thing IMO is establishing how to approach problems of this nature (as I've said many times - top down rather than bottom up). 
2. "Upperclassman level" - blending in more quantitative work, putting numbers on things, elaborating on theory, etc.
3. "Graduate level" - Would assume there's already a firm understanding of concepts and focuses on novel applications, methods work, etc. Math heavy.

I feel as though there's a lot of material that blends #1 and #2. I might even break #3 into Masters and PhD. Masters level material would probably be along the lines of RCVD, and Tire and Vehicle Dynamics. If you have a pretty firm understanding of things at that level, you're probably doing alright for yourself. PhD level probably doesn't see the light of day, as it's the good shit that wins championships, and is what keeps me employed while trying to develop it!

Bit of a conundrum

Been a while. Turns out this NASCAR engineering thing keeps me busy all day, and some nights and weekends. However, I still enjoy it. Love it. Learning an absolutely immense amount of stuff seeing the other side of my previous employment.

Hopefully settling in a bit, at which point I can get back to this stuff. Here's the catch: at this point it would be tough not to use all my new knowledge in designing this thing, and blogging about all that shit has the potential to give away some competitive advantage. Hell, Ed Lover knows better than to do that.

"Givin away all the tricks? Get the fuck outta here widdat bullshit"

And speakin' of bullshit, y'all - we haven't won a race yet this season. Going to have to work on that. In any way, you see the source of my motherfuckin quandary .May have to limit future stuff here to just pictures of CAD, mechanical design, or other bits and pieces. Maybe even take a whole new direction on things? I'm open to suggestions. Comment away.

After 2 weeks at a pro race team...

I tell ya what, it's been busy! Figure I'll give an update while I'm not doing design work in this transition period.

Par for the course is about a 7:00-6:00 or 6:30-6:30 (that's 0700-1800 or 0630-1830 for you military and central European types). I've also been doing a 7 hour drive between Charlotte and Akron every weekend to clean and pack stuff at my old apartment, since I had all of two days between ending my first job and starting my second.

It's fun though. Challenging, with an immense backlog of work to do. Pace is wild. Finally starting to do some work with an impact. Gave my report in advance of this weekend's race at Phoenix, where the #22 just qualified 2nd. Really had the pace to be a pole lap by a wide margin. Had the fastest car in first practice by half a second, so the up-front work by the race engineers can't be too far off.

Hopefully on Monday morning the feedback will be, "Tire info looked good," rather than, "Yeah that data was way off, we had to work with something completely different."

In any event, as I said, still in the process of moving shit down there. Hopefully I'll pick things up in another couple weeks.

Some of the best words in motorsport engineering

I started getting into suspension design and vehicle dynamics in late 2005 during an independent study (upright redesign). Going on early 2011, I'm just now getting comfortable with how to understand things at a total vehicle, "systems" level of engineering.

That's not to say I understand all the causes, effects, and relationships in tire / vehicle dynamics. That is far from the case. Read it again: it has taken me 5-6 years to figure out how to think about and understand things. How fucked up is that? 5+ years to learn how to learn. Admittedly this is on top of some specifics, and what I'm free to share in this blog probably amounts to 5-10% at best, and all public domain information.

As I had been alluding to in a previous post, to stay afloat in this business I think you need to be damn good at figuring shit out yourself - or at least having the balls to try. Anyway, the obvious question is - "What the hell takes so damn long?" For me - and I'm sure I've mentioned this before - it's been the lack of a teacher or an "all-knowing source" that has all the answers. Yes, even Google falls short sometimes. Unfortunately some engineers take what they hear for gospel (to a degree I suppose this is classroom mentality). As such, we come to some of the best words in motorsport engineering, and ultimately what has kinda become my MO. Stumbled across it when I opened up an old copy of Tune to Win, in the Preface of all damn things - which I probably never even read in college:

I am fully aware that much of what I have to say in this book is subjective. I wish that my knowledge and wisdom were such that this were not so. Many readers are going to disagree with my interpretations, conclusions, and recommendations. I offer no apology. In each case I will put forth my personal best shot on the subject at the time of writing. I reserve my right to change my thinking at any time.

Our knowledge of any field whose title includes the word dynamics should be constantly expanding. This is because, particularly in motor racing, we approach a complex subject from a base of abysmal ignorance and also because, in a field defined by compromises, knowledge gained in one area can and does modify our thinking in related areas.

IMO, that mentality is as important and relevant now in 2011 as it was in 1978.

It does make me think that it's a bit of a shame there's no definitive - or close to definitive - guide for all this stuff. Hell I don't think there's any publication that comes remotely close to touching tire data engineering well. This is good for my job security!! As much as I like the works of Smith, Milliken, and Rouelle... I feel as if each has their own strengths and weaknesses, but even in conjunction don't quite grasp it all. This is particularly true as I learn well with many examples anchored in hard data. I have yet to read over Neil Roberts' book.

At this point I am going on 4 years into a career in motorsport engineering, presumably with another 5+ ahead of me. Maybe one day I'll write a book - I'd enjoy it.

Until then, I should probably get back to designing this fucking F1000 / Formula B car, yeah?

Slow here the next days / weeks / ??

Let's talk about the future, shall we? Oh and pardon if I'm a little out of it today, drinking over the duration of the Super Bowl last night usually makes me a little foggy.

Anyway, got my move to North Carolina coming up, starting the new job, probably working a shit ton of hours (though thankfully not traveling too much - at least initially). May not have much activity here.

Additionally, given the nature of what I'll be working on, may have to cut out any of the vehicle dynamics / sim stuff and keep it a little more strictly to CAD. We'll see.

What you're taught in college vs. what you're going to work on

I have to admit, I enjoy perusing the interwebs. Sometimes you come across some really great gems of information (case in point). Other times, shit is just disappointing. Can't even make this stuff up, here is a recent post from FSAE.com

i am new member of fsae. i wud like to know what be the optimum values of
1.toe angle (rear and front for a rear wheel drive car)
2.caster angle
3.camber angle and shud all the wheels have same camber??
4.kpi
5.caster trail
6.scrub radius
plz reply soon
its urgent

There was also a thread on F1 Technical asking about benefits of a pull-rod suspension in comparison to a push-rod setup as well... not nearly as bad but generally oversimplified. Point of both of these is that there generally aren't all-encompassing answers to problems, even if problems in undergrad engineering tend to lead you to one precise solution, arrived at by some fixed process. I think most people grasp that things will be more difficult in industry, but perhaps the true scope of which isn't put in the right magnitude.

Allow me to illustrate!

Classroom Engineering Problem

"Find the minimum of this function."

Not too difficult! Illustrates a concept (minimization / optimization) in a pretty straight-forward manner. The problem is well-defined, and even if you're not 100% sure of the best way to solve it, you can fake it. Alternatively, you can just copy the solution off someone, or consult The All-Knowing Oracle.

FSAE-Level Engineering Problem

"Find the minimum of the function with respect to x- and y-. No worries, you have a few months to work on it."

Shit's starting to get real. Not quite as simple and straightforward if you're used to looking things in a purely one dimension manner of y = f(x). You'll get a different answer looking at it along one slice of x as another. Furthermore, if you ask different people how to go about analyzing it, you might start getting different answers. Lay a few line plots on top of each other at discrete input values? Plot a surface?

Harder to "brute force" these problems too. Takes some time to run through thousands, or tens of thousands of combinations of toe, camber, spring rate, ride height, etc. Overall it's manageable, and ultimately you're competing to be less of an idiot and have your shit held together by fewer zip ties than the competitor in the paddock stall next to you - who is asking for 1" chro-moly tube and a welder, after the first round of tech inspection.

Real World / Pro Motorsport Engineering

"Here's a rough idea of what you're looking at. Try and figure out what the problem is, because it's not really even defined. The number of variables you have to worry about is somewhere between 1 and infinity - depending on who you talk to - and the people who claim they have all the answers are usually full of shit. You have 5 minutes to find the global minimum, and at some point between 2 and 4 minutes in, the problem is going to change. Also, you will lose partial credit for every minute you take to solve the problem, even if you get it entirely correct at the end of your allotted time. Try to score the most points."

Ouch. The problems really are that challenging, open-ended, and undefined. There are generally dozens of theories and opinions on how to solve the damn thing, and the people who are probably the best source of an answer are the ones you're in competition with.

It's really challenging to prepare for this sort of work. Can attack the problem a number of ways and get different answers every time, depending on where you start and what you feel is important. Then there's always the feeling that's similar to what inevitably goes through your head during at least one exam you've taken...

Oh hell... I don't know any of this shit! Any confidence I had just flew out the window. Fuck. I wonder if anyone else is as clueless, let me take a look around... OK good we have some other blank stares... who are these people who are already furiously writing answers?! Oh well, guess I'll write something down and hope for the best. So long as everyone else fails out I will look brilliant by comparison!

The point of all this is what I feel is important to teach in engineering. Equations for principle stresses? Guess they're good to know, but you can look those up in a book at some point in the future. More so, how to approach problem solving when you don't know the answers, the process, or even the problem. That's what engineers and scientists get paid to do - pull answers out of chaos, and/or data that's typically shitty, insufficient, and can be interpreted in many ways. Some institutions, and seemingly even areas of the world, emphasize this more than others.

It's another reason why FSAE teaches you really good life skills - beyond just cooking with a blowtorch or heat treating oven (done them both!).

And it's official...

After almost 4 years of being a race tire data engineer at Goodyear, I am hanging up my hat. Next stop - Penske Racing.