Focusing efforts for a 2nd year team

[Goody]

I’ve read through dozens of threads on this forum and I am just curious to hear some opinions and comments from those that have been through a team’s early days. It seems the amount of experience here is quite vast!

We started the FSAE program at UNT a little more than one year ago. Last March the team met for the first time, designed through Summer/Fall and had a driving car the following April. We went to Lincoln and ended up 31st out of however many, and completed every event. Our goal for the first year team was met, we’re very happy with those results! [Yes, the team went from not existing to completing all events in about 13 months with some help http://fsae.com/groupee_common/emoticons/icon_smile.gif]

Onto this coming year we have a substantial amount of new members, in fact, only 3 of us from last year remain. While the old members are nearby and willing to help, I would like to hear some input regarding where you would put your efforts. The car we have is still 100% together and will remain that way through building the next car. While we met and exceeded our first year goals I feel our tuning and driver training has a substantial amount to grow.

Honing in on our suspension team, they will be going through the previous kinematics and will try to verify the previous design. The last year group spent a substantial amount of time on the suspension, with some guidance from other teams. My thoughts are that we should focus on working with the car we have and driver training, I feel our car was better than our tuning and drivers. Given we have no pro drivers, didn’t have a substantial amount of seat time and have extremely little experience/instruction with suspension tuning…It’s a fair assumption to make. My feeling is we should get some cones out, a stop watch and start working with the car we have. See if we can get rid of mid-exit pushing and how we can work on getting the car to do what we want, rather than going far beyond what the last suspension team did and going into crunching numbers or using equations we hardly comprehend at this point.

This process will take up limited “design time”. However if the plan is to verify the current design, and if it looks reasonable, recreate it so we can have the new car on the ground sooner, more seat time and an idea of how to tune it…it is time well spent.

Thoughts?

http://farm8.staticflickr.com/7366/9130626640_70d75abfa3.jpg (http://www.flickr.com/photos/73218401@N05/9130626640/)
IMG_8928 (http://www.flickr.com/photos/73218401@N05/9130626640/) by GoodmanPics (http://www.flickr.com/people/73218401@N05/), on Flickr

[mdavis]

First off, congrats on your finish at Lincoln. I think a few of your team members came over to talk to us (Cincinnati) at some point, but I can't remember exactly (we had a ton of people around our car after Endurance).

Anyway, to answer your question (I had a similar discussion with one of our team members last night, so there's a lot of stuff here), I would say pick 2-3 things that you didn't like about last year (get input from the guys that were on the team last year for this as well) and work on fixing those. You say you didn't have sufficient testing/driver training time, so that is 1. Now think about why you didn't have that time. Is it that the car wasn't finished from manufacturing in time? If so, this would be another thing that needs fixed. Look at why manufacturing took so long. Was it resource limited? Planning? Figure out the reason behind each of those, and formulate a solution to that problem as well. In Texas, there's no reason you can't drive in January (I've been in Texas in January, it's plenty warm for driving), so there's no such thing as finishing too early. But as you drill deeper and deeper into the actual cause of something, you will find more things that need fixed. If you have something to fix from last year, do not implement new things on the car. Focus on further optimizing what you have before you add more things onto it.

If your team last year did a good job of organizing CAD files and the like, you should have no trouble picking up where they left off and iterating their design. Figure out what they did, but more importantly why they did what they did, and if you like the reasoning, go with it. Otherwise make a change, but don't let the number of changes add up to be too many for your team. The younger/less experienced the team, the less changes you can make safely.

This is a suggestion I've given to a couple of teams (ours and a local startup team) and I'll pass it on to you: Get the design close to what you think you want, design it to be adjustable, build early and test, test, test. IMO, no choices you make should sacrifice testing time.

As for the fall driving/testing/development, that is absolutely required, in my opinion. The new guys coming in can't fully appreciate what the application is that they're designing parts for until they've seen an FSAE car run. Videos are one thing, seeing in person is something completely different. Get the best driver you can find (even if it's one of the old team members) put them in the car and show the new guys what an FSAE car is capable of.

As a note, the Cincinnati team is in a similar situation this year, and most of this advise applies to them as well (and I've already told them this). Unfortunately (or maybe fortunately, I haven't figured it out yet) I'm on the outside looking in this year since I graduated in April.

-Matt

[Zac]

For a first year effort, that's a tremendous accomplishment. It looks like you have a good base to build from. The first year is about getting a workspace and manufacturing tools/processes in place, building a functional car, and going to competition. The second year is about refinement and knowledge transfer.

Most of this is covered in Big Bird's infamous thread, but generally, if you're scoring less than 700 or so points you don't have a car problem (from a concept perspective). It's more likely an issue with driver training, design and manufacturing process, time management, etc. I've seen several teams put up a good first effort, and then get distracted by adding more complicated systems onto their car for the second attempt only to show up at competition with a worse car than the year prior. It probably isn't the lack of aero, a single, carbon fiber, 10" rims, or even suspension design that is holding you back.

I took a quick look at your results. You're giving up some points in design and getting creamed in efficiency. Those are two areas where you can make up a lot of ground without giving up a lot of additional time or $$$.

Judging from the event scores, the real killer does look to be setup and driver training. You have pretty big lap time disparities between your drivers. You're also giving up too much time per lap in autocross and endurance for it to be just the car. In my experience, 3 or so seconds on a minute course is the difference between a great and a mediocre car on the same course with the same driver. 3 seconds is also about the gap between something that's well setup and a basket case driven by a good driver. You shouldn't be giving up 10 seconds per lap.

I'm not sure that design refinement and driver training is an either or proposition. You need to do both. In your position I would be more trying to address the known issues (raised by judges or in testing) with your current designs and knocking off some weight than I would fiddling with the suspension kinematics themselves. I'd also be spending as much time as I could out testing and training drivers. It's something you can use to keep the project fun and help teammates stay focused.

if there are other teams located nearby see about setting up some combined test days. You can do car swaps and help each other identify strengths/weaknesses.

[Goody]

Thank you for the responses!

I will go more in-depth on many of the questions or concerns that came up a litle later, but quickly:

Lack of testing time was primarily a 1st year problem, we had to start with a chassis table, new jigs, finding tools to use, space, etc. Also just plain not knowing how to do what we needed to do. This year we plan on getting started sooner, chassis laser cut instead of hand notched and a more firm set of plans to manufacture along with a schedule.

We got killed in efficiency because that basically saved our engine. We had an over-heating issue on the last lap of endurance, our ECU was programmed to dump fuel in the event of that to help the engine. It did, but at the expense of those points. Not ideal, but we finished nonetheless! [the goal]. Those should be easy to somewhat recoup next year permitting it's not running on the brink of destruction over something silly.

Design we were clearly unprepared for and disorganized. We really weren't sure what to expect, that will improve. The same with the business case.

As far as design changes, there won't be anything substantial. There is a significant amount of weight in the car, an honest 50lbs won't be a problem without further complicating things much at all. Wet layup carbon instead of fiberglass for the same body panels is included, but shouldn't really complicate the process. Sticking with 13s, no fancy shifting mechanisms, using another F4i, we're sticking with what "we know" [or what we think we know http://fsae.com/groupee_common/emoticons/icon_biggrin.gif].

We also would like to spend time testing the current car so we can see potential issues that need adjusted with the suspension. A big one is the ARBs, CAD and real-life got mixed up and they could not be implemented without major changes during manufacturing. The decision was made that they wouldn't keep us from competeting, so they were dropped in favor of moving forward. I have a feeling the setup is quite a bit off, within the next week or two we will revist that and start with a new baseline.

As always, thank you for the input!

[mdavis]

Originally posted by Mean Green Racing:

This year we plan on getting started sooner, chassis laser cut instead of hand notched and a more firm set of plans to manufacture along with a schedule.

Sticking with 13s, no fancy shifting mechanisms, using another F4i, we're sticking with what "we know" [or what we think we know http://fsae.com/groupee_common/emoticons/icon_biggrin.gif].

We also would like to spend time testing the current car so we can see potential issues that need adjusted with the suspension. A big one is the ARBs, CAD and real-life got mixed up and they could not be implemented without major changes during manufacturing. The decision was made that they wouldn't keep us from competeting, so they were dropped in favor of moving forward. I have a feeling the setup is quite a bit off, within the next week or two we will revist that and start with a new baseline.

As always, thank you for the input!

A couple points here:

1. We hand fishmouthed tubes this year and put a frame together in 2.5 weeks, from nothing but a jig to a complete frame, with somewhere around 8-10 people working on the frame. It wasn't necessarily easy (I was seeing weld beads when I closed my eyes to go to sleep), but we did it. If you can get laser cut tubes, go for it, but you may be able to spend that money elsewhere for a bigger advantage.

2. Good thinking.

3. Tire temperatures will be your friend for looking at the suspension geometry. In terms of ARB's, we finished 9th in Lincoln without them. Understanding why you do or don't need them may help more in the Design event than having ARB's will help on track. Your call though.

-Matt

[Goody]

Originally posted by mdavis:
A couple points here:

1. We hand fishmouthed tubes this year and put a frame together in 2.5 weeks, from nothing but a jig to a complete frame, with somewhere around 8-10 people working on the frame. It wasn't necessarily easy (I was seeing weld beads when I closed my eyes to go to sleep), but we did it. If you can get laser cut tubes, go for it, but you may be able to spend that money elsewhere for a bigger advantage.

2. Good thinking.

3. Tire temperatures will be your friend for looking at the suspension geometry. In terms of ARB's, we finished 9th in Lincoln without them. Understanding why you do or don't need them may help more in the Design event than having ARB's will help on track. Your call though.

-Matt

Thanks for your input, 8-10 people is quite a bit for just chassis work for our small team. The way we've looked at it is, chassis was a long project for our car, talking with CRD it's supposed to be about $450-500 + material + shipping for them to cut the tubing. Say it's $800 total over DIY fitment by the time it's all over [figuring we buy material regardless]. Say 3 guys work for 80hrs to save $800 [240hrs], or have 1 guy spend 10 hours finding the $800 in sponsorship somewhere to make up for the cost? Obviously that can't be done for everything, but if we make the most out of what we can raise for the car to cut manufacturing time, it's money well spent and it's more time spent for testing.

There aren't too many splurges we'd like to do, but I think getting the tubes cut is a good one to do. I've talked to several teams, all agreed.

As far as the ARBs go, that's a very good point. It looks like that is not uncommon either, which would be another reason why we should spent time testing/tuning the car we have now as much as we can. My, and a couple other's gut reaction is to think not having the rear ARB is promoting our understeer...actual suspension tuning and more consistent driving may prove otherwise!

Thanks-

[mdavis]

Originally posted by Mean Green Racing:

Thanks for your input, 8-10 people is quite a bit for just chassis work for our small team. The way we've looked at it is, chassis was a long project for our car, talking with CRD it's supposed to be about $450-500 + material + shipping for them to cut the tubing. Say it's $800 total over DIY fitment by the time it's all over [figuring we buy material regardless]. Say 3 guys work for 80hrs to save $800 [240hrs], or have 1 guy spend 10 hours finding the $800 in sponsorship somewhere to make up for the cost? Obviously that can't be done for everything, but if we make the most out of what we can raise for the car to cut manufacturing time, it's money well spent and it's more time spent for testing.

There aren't too many splurges we'd like to do, but I think getting the tubes cut is a good one to do. I've talked to several teams, all agreed.

As far as the ARBs go, that's a very good point. It looks like that is not uncommon either, which would be another reason why we should spent time testing/tuning the car we have now as much as we can. My, and a couple other's gut reaction is to think not having the rear ARB is promoting our understeer...actual suspension tuning and more consistent driving may prove otherwise!

Thanks-

If you can get tubes cut that cheap, then yes, go for it. It was going to cost us over twice that to get them done, and the lead time was 3 weeks from the time we were willing to place the order (after passing SES). As for the number of people working on the frame, there were probably 3-4 there a lot (1 welding, 2-3 cutting tubes), the others just kind of floated in and out, making 1 or 2 tubes here and there then leaving. If it makes more sense for you guys to laser cut the tubes, then go for it. For our situation it didn't. I'm just trying to point out that even if a lot of teams suggest something, it may not be the best option for your team.

You guys are fairly local to UTA (you did a testing day with them, if I remember hearing correctly?) so you could probably get one of their drivers to drive your car around their lot and give you very consistent feedback. Having 1 driver for the full day as the car is changed is a good idea. Eliminating possible variation from one test to the next is always a good test process. Without ARB's to quickly adjust car balance, we used spring rubbers cut down to fit our springs. It took a sophomore ~1 hour with the bandsaw to modify all 3 durometer spring rubbers to 2 different heights (different gap in the springs front and rear), then a couple of man hours quantifying the change that each durometer bumper made to the base spring. Take that information and look at some suspension quantity typically associated with car balance (we used anti-roll torque distribution, but there are a number of them that you can use) and make a chart of what each possible setup is. Then go driving. Try the base setup out, try adding a spring rubber to one end of the car. Look at the suspension quantity change, compare with driver feedback, and take an educated guess at the next adjustment (was the spring rubber change you made better or worse for the balance of the car, and does that mean you need more in that direction or to switch directions completely?). Make another adjustment and send the car back out. Repeat until you have a better setup, or have exhausted all of the possible combinations and have a ton of driver feedback (and ideally data that you've taken) to sort through and make changes for the next test session.

That's where I'd start.

-Matt

[jlangholzj]

Originally posted by mdavis:
A couple points here:

1. We hand fishmouthed tubes this year and put a frame together in 2.5 weeks, from nothing but a jig to a complete frame, with somewhere around 8-10 people working on the frame. It wasn't necessarily easy (I was seeing weld beads when I closed my eyes to go to sleep), but we did it. If you can get laser cut tubes, go for it, but you may be able to spend that money elsewhere for a bigger advantage.


typically 2-3 guys (usually 2) can have our frame together on a table in about a week. I'm not saying that it cannot be done by hand but laser cutting tubes is one major resource that we picked up that has saved us lots of time, money and resources.


Everyone seems to be commenting about the car, and everything so far is great but here's a couple other things that you should be looking at:

-documentation. When you're verifying designs, make sure to write up a quick paper on it. Otherwise when those three founding members disappear, you can put the team in a very sticky situation.

-static events. How did you guys do in cost, design and business? The above recommendation will help volumes in design. I've been the dedicated "cost" guy for the last three years. Luckily this year I had a couple other guys help out more so I only had to do about 45-50% of the report. As much as we're always RACECAR!!!! about competition keep in mind that there's other points to be had as well

-sponsors. Especially for a young team. I don't know what you school involvement/funding is like but this is always a good idea. This year we had a bit of uncertainty with one of our main sponsors..which led to some worrying. Also the more stuff you can get for free (services, material, etc) the better! Kind of going along with this is making a pamphlet/booklet to give to potential sponsors. Makes it a little more professional and will work to your advantage.


Remember that building a fast car is important but if you don't have the resources to do it...you're stuck in the water! Some of these things are items that my team is struggling a bit with and had we tackled them sooner, would lead to a much more productive team AND a faster car.

Good luck!

[Pete Marsh]

Nice first year effort!

I think you will struggle to get ahead with your combination no car set up, no set up experience, and no test driver. A dificult place to work your way out of!

Here in Aus we have driver swap days with other teams from time to time. I think this has been a huge benifit for teams in your position and I would recomend trying to arange something similar for yourselves.


The idea would be to have your test driver and race (tuning) engineer go along and have a drive of a good car, so you know what one CAN be like if it's good, and have their driver drive your car and tell you what is wrong with it. If your lucky they might even tell you how to sort some things, or at least a direction to go in. Check the results for a team with fast lap times straight away in the enduro and little difference between drivers, and no drop off towards the end and ask if they will give you a hand to get on the right track. Such a team has a good car and a good test driver. It's probably best if they have a similar configeration to your car as well if possible, or at least no wings.


In my experience driving these things, (I have driven a few) most bad cars have usually have at least one huge problem that totally dominates the driving experience, and if you have no concept that it is wrong, or what right feels like, changing all the other things around the car will generally just confuse you. I don't just mean suspension things either, every part of the car has the potential to ruin the car, and quite often there is a problem an inexperienced driver will never spot.


Pete

[Charles Kaneb]

I was in your shoes with Case Western after the 2005 competition.

The #1 piece of advice I can give you is that the second car is a lot harder than the first. There is a massive, massive temptation to make the Perfect Car, to use all the neat stuff you saw at competition, to attend an OptimumG seminar and do all the analysis Claude can show you before cutting a single tube, to immediately get down to the weight that the $50k+ programs hit with your car concept.

You have to get the second car DONE ON TIME. Since inadequate testing time was a problem this year, the design freeze and rollout date have to be earlier than last year. A lot of your suppliers, fabrication help, and sponsors will have helped you as a one-off deal - your manufacturing capabilities won't grow as fast as you think they will and nor will your budget.

Think about things you can leave out of next year's car, rather than things to include. Case Western's 2006 car wasn't bad at the 2007 competition until it broke!

[Luniz]

First of all, congratulations on your strong and impressive debut!

I'd like to steer my advice away from the technical aspects of the car onto some management stuff. As you mentioned, only 3 people will remain in the team. This will happen all the time in your FSAE program so what I'd suggest is that you try to think of some system to pass on the knowledge (not just technical, also team management, event strategies, everything)to new members. This works best if you set up some kind of routine or process for that, something that has well defined "todo's" which people can follow. Just make sure no information or experience will get lost.

From my point of view, this is the single biggest, most complicated and most rewarding task to build up a strong FSAE program.

[Z]

MGR,

Lots of good advice above. Perhaps the most important is Charles's when he suggests that you DO NOT let this first year success go to your head and then try to build the PERFECT CAR next year.

Here are some other points (sort of summary of above):
~o0o~

1. In the next few weeks do a very comprehensive and very CRITICAL review of the current car. Put your grumpy hats on (I have lots of them http://fsae.com/groupee_common/emoticons/icon_smile.gif) and find every useless, PITA, difficult-to-make, expensive, unergonomic, gets-in-the-way, @#%*+++, part on the car, and resolve that you are "... never going to make that mistake again".

That is, and VERY IMPORTANTLY, try to find things you can delete, rather than dreaming up things to add.
~o0o~

2. As part of the process of passing on knowledge to future teams, I would suggest a comprehensive "quantity survey" of last year's car be done now. It would be good to get ALL the new team members involved in this (it keeps them busy, and gets them interested).

So all the major sub-assemblies are weighed, giving the masses of 4 x wheel-assemblies (includes suspension and spring-dampers), engine-drivetrain, chassis, and everything else. Also include CG heights, Yaw-MoIs, and the relative position of the CG of each of these sub-assemblies. Then add up all the numbers for the whole car. So now you have a good idea of the car's most important "mass distribution" numbers.

Next do a similar survey of structural stiffnesses. Torsional stiffness of the chassis, and toe and camber stiffness of each wheel assembly (apply load at wheelprint, measure deflection), are most important. Add torque/power curves for the engine, and importantly these days, its Enduro-typical fuel consumption figures. You may as well throw in an estimate of "cost and time taken" for each sub-assembly. Finally, add the best Skid-Pad and Acceleration times the car has ever made (these being the most objective performance numbers, but you should also note track conditions, etc.).

At the end of each year (or start of next), all these numbers should be printed on a big laminated poster that is pinned to the workshop wall. Every team member, from noob up, should know all these numbers better than they know their family member's names and birth dates. http://fsae.com/groupee_common/emoticons/icon_smile.gif
~o0o~

3. Getting specific now, regarding the "mid-exit pushing" you mentioned. As Pete said above, one small, difficult to find, but easy to fix, problem can often cause 99% of your problems. In fact, depending on the driver, there might NOT be a problem at all. The following is very common:

Inexperienced driver, "Aaarghh! It's a pig!!! Just as I get to the apex it starts to move around all funny, and I have to back off... The roll-centres must be migrating all over the place! Gotta fix the kinematics... And we definitely need those four way adjustable dampers, 'cos it keeps doing that wobbly-thing on entry..., err... and on exit..."

Experienced driver in EXACT SAME CAR, "Yeah, it goes good. Gotta stomp on the right pedal just before the apex, and keep the boot in... But then the tail comes around real sweet... Could do with a few more horses, though ..."

I am just guessing, but your "push" problem might be fixed with a few tenths of a degree of static toe-out at the rear. The problem may be that you have a bit too much compliance toe-in when under power, which then feels like understeer?

Anyway, I strongly suggest you follow Matt's advice above (U of Cinc) and don't bother with ARBs. Most handling problems (on FSAE type smooth tracks) can be fixed by adjusting tyre-types, tyre-pressures, wheel-toe-angles, wheel-camber-angles, and wheel-rim-widths (roughly in that order of importance). None of these adjustments needs any sort of suspension at all. The last big adjustment is LLTD (aka RMD), and in a sense this can be done with the right sort of chassis stiffness, again with no suspension at all (eg. go-karts). But in FSAE a simple change of spring and/or bump-rubber stiffnesses (as Matt suggested) should be enough.

Z

[Kirk Feldkamp]

I think that's one of your best posts Z. All very good, practical advice with a minimum of grumpiness. Haha. http://fsae.com/groupee_common/emoticons/icon_wink.gif

-Kirk