Autocross Revolution 2.0

co-drivers in masks

Aspiring Autocross Car-Show Models

Newly returned from Solo Nationals 2019 is a good time to talk about the elephant in the room that we insist on buying drinks for. I’m not talking about electric cars or spotty availability of the hot new tire or clouds of cement dust. But I have to start with a little history.

Pre-Revolution

Once upon a time you had only your memory to tell you what you did wrong or right after taking an autocross run. If the back end came around 90 degrees that would be an obvious mistake. If a section “felt” faster then you must have done it better, right? Maybe not. It took a lot of experience to know for sure.

We walked the course as much as possible and with a certain discipline. What to do while walking, what to notice, where to look… the walking task was as much a part of the sport as the driving because every course was different, one of the key differentiators for autocross. Then came the mental planning part without any practice runs (all runs are officially timed), another key differentiator. What line to take in each corner, how much to brake, how soon to get on the throttle and much more had to be estimated, planned and remembered, not practiced over and over and refined like for time-trials or road racing. At a national event you had all night(!) to think and plan. Then, after a first run, you had only two chances to make changes where your plan was proven wrong by reality and to fix the inevitable mistakes. We’ve all heard the lament, “I can’t believe I coned away 27th!”

We learned various guidelines to help us, like Slow-In and Fast-Out, Maintain Momentum and Cut Distance and we could judge ourselves based on those guidelines (even if they were sometimes mutually inconsistent) and whatever the timer said. It was difficult to tell what was really going on from the timer, right or wrong, because being fast in one section is often negated by being slow somewhere else, especially in the beginning when we were very inconsistent. A few people set up single corners and used the stop watch to figure out which of several methods was faster. That was about as good as it got until Revolution 1.0.

 

Revolution 1.0: Portable Data & Video Electronics

The advent of GPS data devices allows us to analyze in detail exactly how we’ve driven a course, even during the event. We can compare doing the same corner different ways from run to run, either driving differently on purpose ourselves or compared to others who drove the same course with different choices. Did we gain on entry or exit or lose time everywhere? Where was it worthwhile to give it up to Save Time somewhere else? Gradually we proved out what actually worked and what didn’t. We learned that a lot of the guidelines carried over from road-racing were not applicable to autocross, at least not without some fine-tuning. Road-racing was going through a similar revolution and learning new things.

 

Here’s an example of how different autocross and road-racing can be. I’ve heard others tell a similar story. In my second year of autocross (before I had data or video) I asked a very successful former road-racer who owned a race shop at the time to co-drive with me. I had maybe 20 events under my belt. It took only a few seconds into the first run for me to be blown away by his car-control skills in a car he’d only ever driven on a test drive after he fixed it for me! He was immediately on the edge of adhesion, using the tires and controlling the car in a way I’d never experienced. That was eight years ago and I still remember my amazement.

But, I killed him on time.

When it was over neither of us had a clear idea as to how I was consistently faster over the day’s runs. He got faster run to run but I got faster even faster. I was too new to know why and he’d only done a couple of autocrosses in the distant past. With data it would have been obvious. Maybe something about not having track edges to set limits. (There’s another key differentiator in there somewhere.)

The portable high-resolution video cameras not only allow us to see our driving after the event but also allow us to video the course as we walk it before the event. Then we can speed it up and see the course at speed. This helps a lot with knowing the course before you ever drive it and determining where to look at each point along the way, a key aspect of being fast. When you can do this a day or even two before you run, like at Nationals, it’s even more helpful.

Early on I took still pictures with a digital camera as I walked my proposed driving line the afternoon before the national event. Then I would flip through the pictures many times before the first run. Later I used a smart-phone camera. Now I video with my Go-Pro as I walk. When viewing the movie I can stop it to get stills at any point, paying attention to the background as well as the immediate course. I think this helps you to unconsciously recognize your position, similar to the strong home-court advantage in basketball.

 

Revolution 1.5: Pre-Viewing with Virtual Reality Video

Now we have consumer-level devices that create a 3-D picture, really a 3-D visual model, of the entire course which you can later view along the course path wearing a Virtual Reality headset. You can turn your head left and right just as you would when driving to look ahead. If you run the last two days at Nationals, for instance, you can attach the camera to a car that runs the first two days. This gives the ability to visually tour the course at whatever speed desired many times before ever driving it. The old days of everyone walking the course, imagining in their head what it will be like at speed and planning how to drive it are long gone. I haven’t personally used the 3-D VR technology but others are. We’re no longer in Kansas, Toto.

But, wait. You ain’t heard nuthin’ yet.

 

Revolution 2.0: Pre-Driving with Virtual Reality Simulators

 

If I go ever back to Nationals I’ll take a complete driving simulator. That means a gaming PC, wheel, pedals, shifter, chair and VR headset. I will have already created a car who physics matches as close as possible to my actual autocross car. I will GPS-mark each key cone on the course and create the course in VR, maybe even combining it with the scene model from a VR camera. I will have developed a way to measure the surface characteristics of various sites (and tested it at previous events) and put that into the physics engine also. I can even change it to a wet surface (dump a bottle of water on the surface and remeasure) and use my VR rain tires. Then I’ll drive the course many times in VR before I ever run it for real. I will analyze what techniques and lines are fast and what are not. These capabilities are available now, if not all in one package, and will only get better and more accurate with time. Soon it will all be very cheap. Incredibly, this is all perfectly legal.

Autocross is about to become more like time-trials where you can practice the road course however many times you can afford to visit the track before the event, limited only by how much you can afford in tires, other consumables and time, or, you can VR drive it if it’s been mapped! Much of the previous mental key differentiator is no longer necessary, except at local events where the course is created the morning of and there’s little time to collect the data and drive the simulator. Does anyone doubt that soon it will take only a few minutes to collect the data, download it and drive the simulator (in the back of your pickup?) a few times in the paddock? Can you imagine what this will do to the sport at the local level the first time someone shows up with one of these?

This is actually really good for someone like me, an older and somewhat technically savvy person who has trouble walking National courses an unlimited number of times. I think I stand to gain. I can figure out the technical aspects. I can afford the time and at least some level of a VR driving simulator. Being older it’s become harder for me to pick out the key cones at speed. VR simulation driving may help a lot with that limitation and somewhat even the playing field between myself and the younger competitors.

Except I don’t want it. I probably won’t do it. The sport will have lost a lot of it’s interest for me.

Get it done in three, as the saying once went. Now it’s get it done in 100 VR runs in a simulator and then, oh yeah, drive it three times in RR (Real Reality) to prove the accuracy of your simulation.

I wrote a letter to the Solo Events Board on this subject this time last year. Thank you for your input. I see no indication that the technology tsunamis Revolutions 1.0, 1.5 and 2.0 are being given any thought. The elephant is about to step on our foot and it will be painful.

Can anything be done? It’s true that technology, left unabated, changes everything forever. There is no more potent force for change on the face of the Earth. So I think a lot of people are shrugging it off, thinking nothing can be done. Nothing can stand in the way of technology, right? Wrong!

I was a technologist by trade. It’s been proven over and over that Sport can preserve it’s essence in the face of technology. Even a highly dysfunctional sport like Formula 1 was smart enough to outlaw anti-lock brake systems, for example, in order to keep some driver skill relevant. Formula 1 outlaws all sorts of other technology as well. They do it to protect their sport and their revenues.

 

Automobile technology is changing autocross, of course, what with paddle- shifted automatic and dual-clutch transmissions, launch control, sophisticated traction and differential control, and the (smaller) elephant in the room everyone recognizes: electric cars. The SCCA is working hard to keep the sport relevant. Most seem to think we are headed for split classes: electric vs. internal combustion. This is where everyone is looking. We’re gazing at the wrong elephant.

I think we can and must preserve the mental aspect key differentiator of the sport. We do it with the ban hammer.

I don’t see why we can’t outlaw the problem at the root. The root is acquiring the course data. Don’t publish the course designs until after the event. No GPS or video electronics allowed on course walks. Use of any location-determining or picture taking electronics on course walks, including cell-phones, should be outlawed.

We don’t allow cells phones or other personal electronics in many other places, such as while driving or in construction zones or in classified work areas. Why can’t we ban their use during autocross course walks?

Any off-course use of technology (like LIDAR) to spot the cones and develop a virtual course map should also be banned as that would be an easy evasion of the course walk electronics ban. Finally, any attempt to use a racing simulator to predrive the course or a set of pictures or video to view the course, no matter the source, should be clearly outlawed as well.

Please, let’s have none of this I know, we’ll publish the course maps just before the event so it can’t(?) happen again but we don’t have the guts to clearly say that what happened last year was wrong ‘cause we’re PeeCeed, weak-kneed, chickenshits without any real principles. Stand up for our sport.

Only a clear ban on VR pre-driving will keep the mental aspect of the sport as challenging as it was when it started. Otherwise this key differentiator is certain to disappear.

Our new rallying cry: Protect our key differentiators! (Or something equally catchy.)

Could people cheat? Of course they could. People can cheat at anything. (I just read The Art of the Con. Wow!) But most won’t. Those that do cheat and get caught should be handed severe penalities.

 

One last thing: the SCCA should publish accurate electronic maps of copywrited as-run Nationals courses after the event is over. These can be obtained with GPS but more accurate methods are available. Why not negotiate a tie-in with one of the big VR driving simulator software houses? Isn’t this a no-brainer? We should be encouraging people to virtually drive previous Nationals courses on their racing simulators just like they VR-drive race-tracks around the world. Imagine a generation of kids that grow up doing VR autocross racing who can’t wait to get into the sport for real!

ed sitting with umbrella

Coolant-Spill Delay

Teaching Points For Novice Autocrossers

SONY DSC

Dixie National Tour 2012

Tomorrow I’m heading to Bowling Green to instruct at the Tennessee Region SCCA autocross school in the National Corvette Museum parking lot. I was thinking about what I would tell my students and decided I’d better write it down.

Now, some of this I was told when I started, but a lot of it I wasn’t. And I was told some stuff that turned out to be, shall we say, incomplete or possibly even misleading but certainly “problematic” which is a curious and pretentious word I learned when I was a Philosophy major a little while back. I’ve learned a lot from taking an EVO school or two and I really learned a lot from our Twickenham Automobile Club advanced autocross school that we put on each year with a curriculum developed by Steve Brolliar. (It’s coming up in a few weeks. Registration opens on 9/1 at teamtac.org)

So, below is what I wish I’d been told when I was a novice autocrosser, based on what I’ve learned over the last roughly 10 years.

Ed Fisher’s Teaching Points For Novice Autocrossers- 2019

Walking The Course

  1. Plan your driving line and initially drive to your plan but don’t get too wedded to it. Things may not be as they appear and conditions will change each run.
  2. Plan the entry and exit for each corner. A cone is usually (but not always) the apex location. The apex is the point where you will finish slowing and start accelerating. To plan the exit, stand at the apex cone and determine the correct angle (at your estimated minimum speed) to get the fastest possible exit that sets you up correctly for the next feature. Then figure out how to get to that apex location and angle as quickly as possible, i.e. where to initiate braking, how much to brake and where to initiate the turn. That’s your plan for corner entry.
  3. Autocross often contain a series of connected corners. The entire course may, in fact, consist of one long set of connected corners. (This differs from track driving that features mostly stand-alone corners.) Plan a compromised line and set of entries and exits to optimize a connected section in order to find the fastest average speed through it. Fastest average speed wins every time.
  4. As you walk think hard about where to be looking at each point, far enough ahead to see where you want to go. This may mean looking through the side window. 90% of autocross is just “get where you want to go as fast as possible.” Give yourself that command and give your mind and body the space to do it (by looking ahead) and you’ll do it mostly right and will get better with experience.

Driving The Course

  1. Drive in the moment and keep the car working as hard as possible. Forget about being smooth. You can do that when you’re old(er).
  2. Keep visually scanning far ahead from the time you leave your grid spot until you return. Use peripheral vision to aim at the key cones. Six inches from a key cone is four inches too far away.
  3. Constantly test for the lateral limit when turning. There’s grip on the other side of slip. Get comfortable over there. Only the car can tell you, when one end or the other begins to slide, that you’re driving at the limit. Every corner is different from the others. Each corner changes from one run to the next. Only by constantly testing the limit can you stay at the limit. (Don’t do this on the street!)
  4. Adjust your braking points earlier as each run gets faster or later if you’ve been too conservative in your planning.
  5. Minimize the distance travelled, but for short corners, i.e. < ~120 degrees, go wide on entry to get the best angle. This decreases time to the apex and will increase the minimum speed and shorten the time through any roughly symmetrical corner. (Road-racers call this “using all of the available track width.”)
  6. In long corners, i.e. > ~120 degrees, stay tight and limit how wide you go on entry. The shortest time around a circular path for all cars is on the minimum radius. Find the apex (the location and angle where you can throttle up and start opening the wheel) like any other corner.
  7. Really long corners, more than 180 degrees, may technically be double-apexes. There’s a special way to do those (see The Perfect Corner 2) but if you stay tight and keep the car working near the limit you’ll be close to optimum. Determine the exit apex like any other corner.
  8. The lower your car’s power the bigger the cornering arcs you must drive to keep your speed up. This is called maintaining momentum to maximize your average speed. It’s worth the extra distance travelled when you learn to optimize the balance between speed and distance. High power cars should run a smaller corner radius when given the choice, but the difference is not huge within Street classes.
  9. Learn to trail-brake. Now. This is the safe place to practice it. Initially brake hard and fast (ok to activate ABS) then release the brakes slower and smoother as you turn in. Perfectly done, braking ends at the apex.
  10. Mentally review each completed run and find two or three mistakes to eliminate or places to improve. The third run needs to be close to perfect if you have national event aspirations. Consistency will not happen overnight. In the beginning it’s mostly about eliminating major mistakes and minimizing minor ones.

The Last 944 – Part 2

in line at Bristol

At the end of Part 1 I mentioned doing well at this year’s first regional autocross event at Milton Frank stadium in Huntsville. After Alan McCrispen drove the car to 2nd in ES and 3rd overall and I took 3rd in ES and 5th overall, another driver, a national champion, came up to me and said, “I think you may have something with this car, especially on big national courses.” This was encouraging since the next week I’d be taking the car to it’s first national event, Dixie Tour (near Valdosta) and then to another one, the Charlotte Tour, the very next weekend.

But all was not perfect. For one, the car was difficult to drive, at least for me. I had an idea as to why but there was no time to make big changes. For another, the winner of ES (who was also 1st overall) was a Miata driven by Goofy Gomer #1 who I knew I would probably have to beat to ever win a national event trophy. 

The Goofy Gomers are two brothers that co-drive a Miata named Scarlett (painted red, of course) in ES. We often attend the same national events. They’re good friends and named themselves by calling their enterprise Goofy Gomer Racing. Friends refer to them collectively as the Gomers. Both have PhDs, by the way. They’re very smart and, yes, perhaps a little goofy, especially when they get together with brother number 3.

Winning a trophy at a national event was The Goal I set for myself and the Last 944. As PedalFaster, an accomplished autocrosser who’s run both a 968 and a Boxster, wrote to me on Rennlist, “If you win, or even place well, the sound of minds being blown across the country will be deafening.” I thought that if I could get the car competitive and learn how to drive it then I stood a chance at snagging a trophy. Coming from a high-power class I knew I had much to learn about racing a low-power car. The driver and the car would both have to get faster to claim a trophy at a national event.

Dixie Tour on March 16th and 17th was the first SCCA national autocross event of the year. I ran into one of the top drivers at breakfast the morning of the first day and he asked me, “What are you driving now, something weird I saw…?” 

“I’m driving a 944 in E-Street so I can learn momentum-maintenance,” I told him.

“Well, you sure picked the right car!” he said, looking away and shaking his head.

The event was a disaster. ES had 20 racers, including the Goofy Gomers, with six trophies to be awarded. The others didn’t know me and were probably a little suspicious of my odd car. I promptly allayed any concern by spinning on two of three runs the first day, landing myself in 14th and long out of the trophies time-wise.

I mentioned that the car was difficult to drive. One of the spins, right before the finish, was such a spectacular fail that it received an ovation from the crowd. The best I could hope for on the 2nd day, when we would run the same course in reverse, would be to climb a couple of places. That didn’t happen. It rained. I didn’t have rain tires. I dropped to 16th.

The necessity to have two sets of tires in Street classes, one set for dry and damp conditions the other set for heavy rain, is a new thing. What’s happened is that the tire makers have created specialized tires that have huge grip on dry pavement. You must use one of these specific types to be competitive. They’re about as good in the dry as race slicks of 20 years ago yet they have treadwear ratings of 200 and are perfectly legal for use on public streets. Unfortunately, they gave up some wet traction and hydroplaning resistance when they optimized these tires for warm (non-freezing) and dry conditions. Then they created different tires that work really well in standing water. I don’t have a set of those or a second set of rims to mount them on.

I watched the car in front of me in grid switch from dry to rain tires as it really began pouring. The driver went from last place, after three spins in a single run, to first. The Gomers had a set of rains and placed 7th (one place out of the trophies) and 10th. 

On to Charlotte, but only after a week of licking my wounds on St. Simon’s Island.

Charlotte was hot and dry. I finished lower mid-pack in 10th place out of 15. The Gomers were close behind in 11th and 12th. I managed to not spin and not hit cones but was discouraged by the result. The course should have been good for the 944 but I was one second per run behind 5th, the last trophy position. In autocross a second is a looooong time. I thought I’d driven well but it wasn’t nearly fast enough.

The 5th place trophy winner told me he was impressed by my times in the 944, not far off some good Miata and MR2 Spyder drivers. (Racers from Pennsylvania to Florida were there.) He had autocrossed a 924S back when they were competitive against the first generation Miata, but no one had ever been competitive in the heavier 944, much less later against the 2nd generation Miata. This was nice to hear but it was clear that being able to compete for a trophy, which was my definition of competitive, was not yet in the cards. I’d come to a tentative conclusion that the car was lacking steady-state front grip but I didn’t know why. 

Back in Huntsville I got together with the Gomers to compare data and it showed what was happening. I had a slight acceleration advantage over the Miata, but this was being equalized by a need to shift to 3rd gear when they could stay in 2nd. Shifting to 3rd usually means a downshift back to 2nd, sometimes twice per run, and extra shifts cost time even when done perfectly. Downshifting is especially tricky. Autocross requires ten times the number of driver inputs per time interval as compared to racing on a track and we don’t get to downshift while braking at the end of a straight. There are no straights, only curves where you are either accelerating or braking while turning. (If you try to “create” straights with a low-power car You. Will. Beeeee. Slooooooow.) 

Typically the downshift must happen while simultaneously trail-braking into a corner that’s either an increasing or decreasing radius but almost never a constant. It requires a high level of car control and coordination to a) not lose the rear end of the car when letting out the clutch in the middle of the corner, properly rev matching while continuing to trail-brake perfectly, b) keep the car within a foot or two of the correct line and hit your apex at just the right angle, and c) be back to full power at the apex not a moment late. A PDK transmission would make things so much simpler!

In transitions, like in a slalom, the data showed that the best I can hope for is to limit the time lost to my narrower, lighter, more nimble competition by driving the heck out of the car. When Goofy Gomer #1 remarked, after driving an autocross run in the Last 944, “You should never be able to beat me in this car” I think he was mostly reacting to a lack of agility. When I explained to him that Porsche had intentionally put the engine at the front and the transaxle at the rear to increase the polar moment of inertia and force the car to react more slowly than, say, a mid-engined car like the 914 or the Boxster/Cayman or every purpose-built Porsche race car in history he just looked at me like I was, uh, goofy.

“Why the heck would they do that?” he asked.

“I think it was so magazine testers who couldn’t drive could go fast without crashing and then they would write about how great it handled,” I said. “Chevy did the same thing with the C5.”

It was in the sweepers where I thought the car should be a match for the others or perhaps even have an advantage. Unfortunately, the data showed the Miata walking away from me in every long corner. It was simply generating more lateral grip. 

This was really bad. It meant that even if I was a better driver than many I was unlikely to ever beat enough Miatas and MR2 Spyders to climb into the trophies.

Now I did something I should have done long before: I calculated the stiffness of the front and rear suspension of the car with the steel springs, torsion bars and bump stops all acting together. The results were surprising.

You may remember from Part 1 that a key advantage for the Last 944 is the ability to use the bump stops to stiffen the car in roll. How stiff should the suspension be for an autocross car that competes on surfaces that, unlike a real race track, are often various levels of bumpy even at relatively low speeds? And how stiff was my car, anyway? The answers will require me to get a little bit technical, so brace yourself.

Turns out that a Canadian autocrosser and engineer named Dennis Grant (see his website Autocross To Win) did some testing and decided that the front end of an autocross car works best with a natural frequency of 2.2Hz and with the rear end a little stiffer at 2.5Hz. (People like to argue about these numbers, but Grant became an SCCA national champion in the car he developed.) Natural frequency is a measure of the stiffness of a spring-mass system, like the mass of one corner of a car sitting on top of a suspension spring. These numbers are lower than a typical road-race car, which might be 3Hz, and many autocrossers were then using and continue even now to use setups even stiffer. For reference, a modern sports car might be as much as 1.5Hz as delivered new. The stock springs on the non-modern 944 give a stiffness of only about 1Hz.

Elastomeric bump stops don’t really work like coiled steel springs (or torsion bars) but I was able to squint a little and derive an estimate for their effective spring rate over the distance they were probably being compressed. I made a spreadsheet and put these values in parallel with the steel, working through the correct motion ratios, along with estimated weights of the sprung mass at each corner.

The calculated rear stiffness was 2.6Hz. Almost perfect. Perfectly dumb luck.

The calculated front stiffness was 3.5Hz. OMG! as the kids text these days.

I had originally purchased bump stops simply based on the front stack of three being the same stiffness as the rear stack of two. I had utterly failed to take motion ratio into account. The bump stops on the front struts are very efficient with a motion ratio of 0.91. The bump stops on the rear shocks, due to their position, work through a motion ratio of only 0.63, a huge difference since it’s the square of these numbers that’s actually used in the equation.

The overly stiff front bump stops and the heavy shock damping they require for control are almost certainly what’s been impairing the grip.  If the spring rate and shock damping are too high then when the car hits a bump the impact energy is not well absorbed and it accelerates the car upward. This actually pulls the weight of the car off the tires (until it comes back down) so the tires lose grip. This not only happens on big bumps but also smaller irregularities where the pressure at the contact patch will vary so much that the grip is uneven. This lowers the effective grip and makes the car hard to drive. In a sweeper you only have as much grip as the weakest end.

To fix this I calculated what spring rates/durometers the bump stops needed to be to get 2.2Hz in the front and ordered new ones. At the next event with the new bump stops, TAC/TVR #2 on April 6th at Milton Frank, I won ES. I beat both Gomers by a second. The car had more grip and it was easier to drive. One event, though, especially at Milton Frank which has a weird surface, doesn’t mean much. I missed TAC/TVR #3. At TAC/TVR #4 I lost to both Gomers by a second, which muddied the waters.

What you need to realize is that I’m making constant changes to alignment, shock absorber damping and front sway bar stiffness in between and even during the events. Changing the front bump stops necessitated a cascade of other changes. To get to the top in autocross we treat all events, especially local events, as tests. If you’re not testing, making changes and evaluating the effects you can’t develop the car. (The same goes for the driver.) So, results with a new car are often inconsistent. Add to this the natural inconsistency of human performance and the varying character of different courses and surface conditions and… well, you get the point. I’m making excuses.

I made another major change before what might be my last and best chance at a trophy, the Bristol Match tour at Bristol Motor Speedway upcoming on July 5th and 6th. (In the parking lot, of course!) I mounted new front tires in 245mm width, up from 225mm. This is way “over-tired” for a 7” wide wheel and the same tire I have on the 8” rear wheels, but it worked. With all the other changes I now had so much more grip from the front than the rear that when I ran at our local Test and Tune event the week before Bristol the car was seriously unbalanced. Though I’d increased overall grip by improving the weak end of the car I was forcing the poor driver to adapt quickly or die.

Bristol was really hot. This was good for me because I’m using a tire that likes heat, but takes a run to warm up. Without a co-driver to warm the tires you know going in that your first run better not be your best. The driver needs to be consistent over the span of three runs, recognize and fix the inevitable mistakes and rarely throw runs away by coning. You must be confident that you can lay down your fastest run on the last one.

bristol start

Lining Up For The Start At Bristol

At a Match tour you get three runs in the morning then three more that same afternoon. The fastest times from each set of three are added together for your total time and that decides the class results. After the morning set I was 6th of 9 drivers. Trophies would be awarded only for 1st, 2nd and 3rd. 

I was way behind 1st (a Miata) and 2nd (an MR2 Spyder), but from 3rd on down was very tight so I had a realistic chance at the last trophy. I had an equally good shot at last.

My first run in the afternoon is faaaast, baby, and I vault into 2nd place. The MR2 that started out in 2nd cones his first run but I know he’ll probably clean it up, so to be honest I’m still racing for 3rd place. My real competition for 3rd is now an older guy (like me) in a Miata. We’re gridded with our cars right next to each other. He’s only 0.1 seconds slower so far. It’s clear that the one who improves more on the subsequent runs will likely snag the 3rd place trophy. We wish each other luck as we adjust tire pressures.

bristol showcase

In The Showcase Turn At Bristol

My second run is almost a half-second faster. I’m very pleased. Once all second runs are complete I’m still in 2nd place! The Miata next to me in grid spins and knocks a wall of four cones into the next county. The MR2 that started the afternoon in 2nd lays down a clean but slow run. He needs to get a lot faster or he won’t even trophy. Each of these two guys has one more chance to take 2nd and 3rd and drop me to 4th. The pressure is really on them to find some speed and finish off this pesky 944. Meanwhile, the Gomers are down in 8th and 9th but giving me encouragement.

I make a significant driving mistake on my third run (while downshifting) and go slightly slower, so I must stand on my second run. The MR2 finds his speed and goes 0.004 seconds faster than me to take back 2nd place. My competition for 3rd place in the Miata gets faster and runs clean but is still a tenth off my 2nd run. I win 3rd place by 0.054 seconds and take the last trophy. 

Goal achieved!

trophy2

With The 3rd Place Trophy

I think I’m finally getting the hang of this momentum-maintenance driving. It’s a complex equation, trying to figure out when going a few extra feet here will save time through there, or not, within the complex set of connected corners that constitute a modern autocross course. It’s a mental skill necessary to reach the highest levels of the sport. With a more stable car configuration I hope to focus harder on the driving going forward. That’s probably where the real time is to be found in the future.

But I’ve continued to make changes to the car. At Bristol the grip was unbalanced with a tendency for wicked lift-throttle oversteer that could be disconcerting. Pictures also show that it rolls too much thanks to the softer front bump stops. I’ve reduced the rear shock damping and added more front roll bar stiffness. These changes get tested tomorrow at Milton Frank.

I have one more national event, the Peru Tour in Indiana, before Nationals themselves the first week of September. The Gomers won’t be at Peru but many other racers in Miatas and MR2 Spyders are registered who represent the best in E-Street from the Midwest. 

The National Championships will be held in Lincoln, Nebraska. I’m registered in the 944 and the Goofy Gomers and my wife and I will be sharing an Airbnb rental. If anything interesting happens maybe there will be a Part 3 to this story.

Edit: I just became aware that Chuck Matthews is using bump stops in a similar fashion to tune his ES Miata. Rats! Seriously, go see his blog at mathews racing.com

The Last 944 – Part 1

 

bristol showcase

This year I’m autocrossing the last base-model 944 that will ever run in Street class in national competition with the Sports Car Club of America (SCCA).

I know this because the SCCA has a 30-year sunset rule in Street. 1989 was thirty years ago and was also the last year of 8-valve-engine 944 production. So all the other years of the base 944 have sunsetted except the 1989 model.

Street category is the lowest SCCA autocross preparation level, meaning the car has to be almost entirely configured as it left the factory. The sunset rule was implemented because after 30 years memories fade and documentation gets scarce so it can be difficult to agree on what is or is not a legal configuration.

In SCCA autocross the cars are separated into classes within preparation levels. The base 944 is classed by the SCCA in E-Street. E-Street is dominated by two cars: the 1999 Sport Edition Miata and the 2003 Toyota MR2 Spyder. No one seriously runs much of anything else in national competition. Both are 2200lb-ish cars. The 1989 Porsche 944 is, well, nowhere near that weight figure what with a heavy, balance-shafted 2.7 liter motor, a back seat, a huge glass hatch, a/c, cruise control, and power everything. It doesn’t have enough additional wheel and tire width to make up for the heft so no-one thinks it can be even remotely competitive. None have been run seriously in national competition in the 10 years since I started in the sport.

In my feverishly foolish imagination I hypothesized that the 944 has two quirks that might possibly allow it to be competitive in E-Street today where handling and cornering ability are king. The first is big negative tire camber made possible by the adjustability of the M030 sport suspension option. The second was Porsche’s use of tall bumpstops. While the M030 sport suspension option is well-known, none ever took advantage of the tall bumpstops as far as I can tell. Please allow me to explain, but I’ll have to get a little technical.

Camber is the angle of the tires as compared to the road. If the tires are perfectly vertical the camber angle is zero. If the top of the tire leans in toward the center of the car then that’s a negative camber angle. Most cars come from the factory with a slight amount of negative camber, like ½ degree or maybe as much as one degree. This allows the tires to wear evenly as long as miles spent in hard cornering are limited. Generally, more negative camber, up to as much as 4 degrees, is better for hard cornering, which is about all we do when autocrossing.

The rear of the 944 is sprung by torsion bars, almost exactly the same design as on the 911, and always allowed for infinite adjustment of rear ride height. Not an easy adjustment, but it can be done. The M030 option adds ride height adjustability to the front struts. With that you can lower the front and rear evenly while still using the stock springs required by the rules. Significant lowering of the car is what allows a big negative camber angle at the tires.

So, I gathered up the various parts that make up the M030 option and installed them. Some parts were still available new but many were not and I had to find them used. Now the car has three degrees of negative camber all around and is about 3/4” below the standard M030 ride height which, in turn, is somewhere below the stock suspension ride height. The car is now quite low, as in I’m getting to old for this low, especially with thin modern tires on 15” diameter wheels.

Three degrees is a serious camber number for a Street-class car. You can’t get three degrees in any standard Boxster, Cayman or Carerra. Only the GT cars now allow that much adjustability. Tires like that much camber in the corners. It makes them happy. When they’re happy they deliver more grip and the car corners faster. 

A fly just flew into the ointment, however. The low ride height has a negative side effect on the 944: it makes the front of the car roll more easily while making the rear roll less. Why it does this is very technical so I won’t bore you with it. (If you really want to know why, I explain it in this post.) The upshot is that the more the car is lowered the more the famous handling balance that Porsche gave the 944 gets upset. This roll balance problem can, in theory, be overcome by tuning the tall bump stops.

Bump stops are generally pieces of rubber that keep the shock absorbers from internally crashing metal into metal when fully compressed during a big bump, say, when you hit a deep pot hole on the interstate at 70mph like I did once. They are often only about 1/2” thick. 

What many people don’t know is that by the 1980’s Porsche was doing an innovative thing with bump stops: they were using tall elastic bumpstops as auxiliary suspension springs. The stock front bump stop, shown below, is 3-5/8” tall. It occupies almost all of the free shaft length at normal ride height. Notice the complex shape.

The idea was that while cruising down the highway the soft primary springs give a comfortable ride. When you crank the car over into a corner the bump stops come into play and progressively stiffen the suspension for responsive handling. This design philosophy was still in use when the 986 Boxster and 996 Carrera began production in the late 1990’s. 

The tall bump stop design method is important in Street-class because one of the few allowances is that bump stops are FREE!!! along with the shock absorbers. Some have called this a loop-hole in the rules but it’s a well-recognized and legal loop-hole. Miatas, for instance, also have relatively tall bump stops that need to be in good condition for best handling because they come into play when cornering.

Well, bump stops are almost free. You can make them from anything you want, including jello (not recommended) or solid steel (also not recommended) or get them in various rubber stiffnesses which is what I’ve done. What you cannot do is make them any taller than they were before. 

So you must be very careful because if you take a trophy from a Miata driver at a National Tour event while driving a 30-year old Porsche that everyone knows is slow you’d better be able to prove that you’re legal. I carry a notebook with relevant information for my competitors (or a protest committee) to peruse at their leisure.

This means that I’m free to install stiffer bump stops to stop the car from rolling over like a drunken sailor, which is the stock behavior of most 1980’s sports cars, including the 944 and especially a 944 that’s been lowered, as previous explained. Bump stops allow the roll stiffness of the front and rear to be tuned separately and keep the negative camber we obtained from totally disappearing in the corners due to body roll. The stiffness of the bump stops, if chosen to be stiffer than the stock springs, should also make the car transition from turning one way to turning the other way much faster. Maybe almost Miata fast. Transitioning quickly (nimbleness) is another thing that’s very important in autocross.

The picture below shows what the new bump stops look like on one of the front struts. The three yellow donut-like things are the bump stops sitting on top of the shock body and riding on the strut shaft. In a corner they will slide up the shaft and get squeezed against the top hat which is out of the picture.

frontbumps

The next picture shows the rear shocks with their new bump stops:

rearbumps

 

All is not quite so simple, however. I found out that some autocrossers have tried to use stiff bump stops before, if not on a Porsche. Apparently, they all gave up. At least the ones that are talking gave up. (Edit: I recently became aware of Chuck Mathews and his efforts in this regard.)

The problem they encountered is the abrupt transition from the stock soft suspension spring to the stiffer bump stop. Porsche solved this problem by using a tall, tapered bump stop shape that very gradually adds stiffness. I can’t do that. I need to stop the car from rolling right now, or at least almost right now. To explain this problem I have to tell you a super-secret engineering secret not normally revealed to those not initiated into the secret engineering lore known only by those who actually read and understand the textbooks: load is preferentially and proportionally attracted to the stiffer path.

Let’s imagine we brake the car and turn left. The right front spring and shock absorber will compress the most. The load into the tire increases gradually and proportionately as the spring compresses. The back of the car is following along, but later. When the stiff front bump stop is encountered there will be an immediate increase in the tire load. It’s as if weight from all around the car suddenly jumps into the front right tire. If this increase happens too fast the tire contact patch becomes overloaded and gives up grip. The right front tire begins to slide. Massive understeer is the result and the driver has to make a big correction which costs time.

In other situations it may be a rear tire that gets shock-loaded and calls in sick. The result in that case is oversteer. A car that oversteers one second and understeers the next is not fun. It is not fast. It can be, essentially, undriveable. This is the big fear, the big problem others have run into. I have only a few weeks and a handful of events to work the bugs out.

So how has it been working, you ask? Find out in Part 2. Here’s a hint: multi-time Nats champion Alan McCrispen remarked, after watching me take a run in the car at its first local autocross, “I’ve never seen a 944 corner so flat or change direction in a slalom so fast.” That was the good news. 

The bad news was that I found it a difficult beast to drive and it didn’t seem to have all that much cornering grip. Alan, a much more accomplished driver, was having fun with it and we both placed highly. But we were beaten by a well-driven Miata that took first place overall. Not a good omen since I know that at a national event the drivers will be even faster.

I think I’ve figured out the big issue and have a plan to fix it. Stay tuned.

Below is what the car looked like ready to compete at Dixie Tour, the first national event of the season.

dixie

Line Theory: Perfect Corner 1 & 2

In Perfect Corner 1 and 2, Adam Brouillard has made the most important contribution to racing line theory since the first technical book on the subject, Taruffi’s 1959 The Technique of Motor Racing.

A few months ago I started reading, studying and applying knowledge gained from these two books. They are an exposition of line theory, the theory of what driving lines are most efficient, i.e. the fastest, around a race track. Brouillard takes a physics-based approach.

With Perfect Corner 1 and 2 we can look at all the various permutations of line theory since 1959 and clearly understand what is right about each one and why and what is wrong about each one and why. When a new theory explains all the previous ones, that makes for a powerful theory.

Brouillard teaches that there are only three types of corners in all of road racing and autocross. (He tells me he began as an autocrosser.) The three are the standard corner, the chicane and the double-apex. That’s it.

A standard corner has enough space before and after to stand alone so that the entry and exit can be optimized without regard to the previous or following element. A chicane is defined as two corners in opposite directions that are so close to each other that they must be optimized together. (The autocrosser’s slalom is two or more chicanes end to end.) A double-apex is two corners in the same direction that are so close to each other they must be optimized together.

The books give rules for how to classify each corner you will encounter and rules for how to determine the most efficient line through each type.

My approach to autocross is now forever changed because of Brouillard. What I’ve realized in doing autocross events this year with his concepts and rules in mind is that many autocross courses are more complicated than most road-race tracks. Some autocross courses essentially have no stand-alone elements. Everything is connected in a series of chicanes and double-apexes with only the rare standard corner. So, applying the methods and rules he gives is not easy. Some of it is immediately applicable, some is not. He tells me he’s thinking of writing an autocross book. I hope he does. Soon!

In the meantime, get these books and start a new journey into autocross.