Tuesday, July 24, 2012

(176) Tire shreddin'

The theme of the weekend this time at Golden Gate Fields for the 10th round was difficulty.

The first difficulty was the course. The front section was very fun and fast. The back section was very low speed, and you were basically cornering from the half way point of the track until the end. You mess up one corner, and the rest of the series is blown, all the way to the finish, unless you have some real remedial skill.

The second difficulty was the lack of runs. We were cut down to 3 runs due to time constraints. That meant we have to adapt more quickly and be more precise. I think this put a lot of pressure on Scott, and he suffered a couple big mistakes that put him in 5th spot.

The third difficulty was the competition. I ended up losing 1st place to Maxim Lukichev in the novice class by just 2 tenths of a second after handicapping took place. Our cars were in separate categories, so even though his time was slower than mine (by just a couple hundredths), the field leveling system resulted in Max getting a faster "indexed" time. I got 2nd. I might have retaken the top spot if I hadn't blown the final corner of my final run. It was neither of our best races.

The fourth difficulty was the tires. The Subaru WRX has Dunlop SP Sport 01 performance tires, and, well, take a look at the front ones.



That's called tread separation. Now, normally tread separation occurs between the steel belts of the tire and the rubber carcass. But this is different. This is like the tire is shedding off a top layer of rubber, revealing another layer underneath. Back when I was road racing, we would do this intentionally to our tires by shaving the tread off with a machine, revealing juicy, slick and sticky rubber underneath. There is still plenty of rubber left on the tire in this case, probably a good 5 or 6 32nds.

Even so, I think a new set of tires is required. We might talk to the dealer about it because this is usually a defect. But it might not be, because both front tires are doing this. These are the tires that came on the car. I wonder if they will ask about it when they see the grains... (Those are the little ripples in the tires that result from rubber rolling across the surface as the car corners hard. It's actually a decent pattern on these Dunlops which shows they handle the heat of autocross pretty well, just not the shearing factor obviously.)

In any case, it's not good for the grip of the tire. It should be an even surface. Not all of the tire is in contact with the road in this condition.

So, those were the foibles. Here's the fun bit:

SCCA Autocross Round 10 helmet cam footage.

The next round on the schedule was sadly canceled. That means there is only one round left, at Marina Airport in the middle of next month.

While I didn't win last weekend, I still carry a perfect score as far as the points go, with 1600 points. With two current drop races and no more being added, that means I'll have to finish in first spot one more time to secure a perfect championship score which would be 1800 points.

No matter what happens, though, I will still be novice champion.

Dad is sitting in second spot in the points race, and Michael Bajer is in third. Scott is far enough ahead of Michael that as long as dad earns points in Marina, he will keep second place. It looks like I will finish 1-2 with my dad, which would be incredibly cool.

Monday, July 9, 2012

(175) That's a lot of Mazdas

On Saturday I took a trip over to Sonoma Raceway (previously Infineon) to take a look at the first race of the SFR SCCA Spec Miata festival and Sonoma Sprints.

This race was something new for the region. Normally SCCA races are split into National and Regional weekends. If you want to go to the Runoffs to vie for a national championship, you have to do National races to earn the points to qualify. Regional series are their own championships and don't give you a spot at the Runoffs.

The Sonoma Sprints are a combination of National and Regional races. The regional racers benefit from longer races (the national races have to be 14 laps or 45 minutes, whichever comes first), and the national racers benefit from increased car counts. It's a win-win. Plus the region doesn't have to spend extra money on a separate event for the national drivers, as the national race at Thunderhill was never really popular with locals.

Add the Spec Miata Festival on top of that, and you've got a big field of cars. 73, to be exact. Standing in the middle of the grid, looking out over the sea of Miatas in all directions was a sight to behold. I've never seen that many cars on one track in person before. I rushed up to turn 2 to snap pictures because this was going to be a big one. This is what it looked like.

73 Miatas stream through turn 2. You'll notice my old yellow and silver car, now designated 61, in the final snap.

And no one got spun or wrecked. It was incredible. They were four wide at one point, which you can see in picture 3.

It looked to be a very successful event for SCCA. I'm just very sorry I couldn't be out there. Maybe next year!

On Sunday my dad and I took the Subaru to the 9th round of the SFR Championship solo series at the Golden Gate Fields horse race track.

The parking lot we had was in a beautiful location. It's right next to the water of the bay, with the horse race track just up the hill and behind a fence. There is a little beach near by, which seems to be a popular spot for kite surfing. The surface could use a repave, but it's really not too bad. Still, it chews tires a bit more than the other venues on the calendar.

I've been here before. Last month Scott took an Evolution autocross school where he got some very good instruction while I watched. I think it helped him a lot as he showed a lot of improvement this weekend.

The course we had was very enjoyable. Fast, but fun, and a good test. The layout encouraged you to make the best possible use of the racing line. This resulted in me grabbing a couple cones in the first two runs, and my dad hit his first cone in competition, ever! It's good to have courses like this that emphasize clean runs. Dodging cones is what autocross is all about!

I ended up picking up the pace in my 3rd and 4th runs, mostly because I wasn't incurring penalty cones. I grabbed another win, more narrowly this time (by 4 tenths), and Scott got a 4th in a pretty competitive field.

Round 9 helmet cam footage, now with dad-cam.

Not sure if this format will stay as swapping the camera back and forth takes time (we trade off runs so it's more fair). Scott is getting better at diagnosing his runs by himself, so I may not need to ride along for much longer.

Tuesday, July 3, 2012

(174) Driving enrichment center

I hate computers. I love computers. They can do many things, and I hate when they do not, but I love when they do. I even know a little bit of programming. That is not impressive, but it is enlightening.

A computer is somewhat like a human. It has a brain, it has memory (both short and long term), it has eyes and senses. It has a voice. A computer is faster than a human, but it is not as powerful. A human computes very slowly, but a human has the ability to do lots of computations at the same time, while the computer can only do a few (depending on how many processor cores it has). I often wonder if my computer has any sort of experience. The components seem to be there for it. But if it does, no doubt it is very different than ours.

Computers need humans to function. The web browser you are using to read this was caringly programmed, line by line, by a human, or even many humans. That program is installed on the operating system, such as Windows or Mac OS, which is a logic machine that operates the computer. Without that operating system, the computer does nothing. All the operating system does is flip electrical switches. It takes control of the switches and tells the switch whether to be on or off. By flipping switches very quickly, a program can send instructions to the processor via the operating system in order to solve a problem.

Those problems can be very simple or very complex. In cars, these problems tend to be complex. And the programming needs to be executed well, otherwise the problem will not be solved.

Computers can be found in some surprising places in cars. Of course, most people know about the engine control unit. In addition to keeping the engine running, the ECU can do other things like sending data to the gauges and calculating fuel mileage.

Another common one is the anti-lock braking system. The ABS senses when a wheel stops rotating relative to the other wheels, and adjusts the brake pressure of that wheel to get it rolling again. You also have traction control, which does the opposite for driven wheels in order to stop wheelspin. Stability control is the third common handling-oriented system, which senses when the car yaws relative to the direction of travel, and uses subtle brake pressure to adjust the yaw to something more optimal for the inputs the controller receives from the steering and the pedals. Basically it plays with the angle of the car in a turn.

Of course, none of these systems make the car handle better or worse. The magic is in the programming, which tells the computers how to react given certain conditions. The better the programming, the better the system works.

A lot of people are afraid of programming because it seems mysterious and delicate. Programming is, at it's core, not breakable. What you type is what you get. It is the flow of pure logic that dictates how the components of the system are used. It's not like an engine, which could work perfectly for hours and then blow up because some part reached it's stress point. If a program doesn't work, then it never has worked and it never will until the programming is altered.

This is one of the reasons why electronic throttle linkage (throttle by wire) is becoming standard. The programming is simple (even at my basic level I could write a program to do it 100% reliably), and as a result the whole system becomes much more reliable and safer than a cable. Multiple redundancies are put in place, and failsafes are built into the program to account for any failings of the mechanical parts. It's not the programming that breaks, it's the program that acts as a safety net for the mechanical parts.

I think this fear of the program comes from the use of desktop computers, which frequently have problems resulting from conflicts. Conflicts between programs or hardware. But a desktop computer is an extremely complex thing compared to a simple throttle control program on a closed system with no threat of conflicts.

The throttle by wire program reads a voltage from a sensor under the throttle pedal. That program basically looks like this:

[get the voltage of the throttle sensor and store it in a value];

[use the stored value to apply a new voltage to the motor that opens the throttle by an equivalent amount to the pedal depression];

And that's it. Depending on how the operating system of the ECU is set up and what programming language it uses, it probably is literally just two lines of text, and is probably less characters in total than what I wrote. Of course, supporting programs need to be written to do things like verify the opening of the throttle, and to build functions around detecting and applying voltages. But why am I even going into it? Because like anything, it's important to understand something before one says "yes, it's good" or "no, it's not good".

Of course, now we are getting into the whole issue of "driving purity". There are a lot of people out there who do not want computers in their cars for various reasons. I do not think they are valid reasons, and here is why.

The basic anti-computer argument goes like this: "the computer does things for me (changes gear, manages brakes, etc), and as such, I feel less involved with what the car is doing and I do not like it. I prefer to be in direct control of all of these systems."

The fatal flaw being that the car already does so much for you. Is a differential anti-driving-purist, because it manages the driven axle's traction better and gives more grip? Is a synchronizer gear anti-driving-purist, because it helps you change gear more smoothly and consistently?

Back in the day, all racing tires had tread on them, even in the dry. They didn't know that a slick tire had more grip. Is a slick tire anti-driving-purist because it provides more grip and helps the driver go faster? They also didn't have wings. Are wings also an abomination of driving spirit because they help the driver?

Even in the 1950s, they had plenty of tech to assist the driver. Differentials, of course, were one of those things. They also had automatic timing on the engine, whereas older cars had manual timing that had to be advanced or retarded while driving. Is that impure, because it assists the driver? Are disk brakes impure as well, because they don't fade as much as drums, and are safer and more powerful for the driver to reduce speed? A constant mesh transmission is another assistance device. They used to be sliding mesh, with all the gears resting stationary and they had to be manually spun up in between changing gears. A constant mesh transmission doesn't need that because all the gears spin. Is a constant mesh gearbox undesirable to driving purists?

Practically everything on a modern car has received upgrades, improvements, and sometimes even complete replacements to make them better, easier to use, safer, and faster.

Of course, the devices I'm mentioning are still mechanical. I think in light of this consideration, it is pretty clear that software is the culprit. But how different is it really?

In the typical examples of electronic handling programs such as stability control and traction control, people sometimes think of them being a sort of almost magic device that is separate from the mechanics of the car somehow. Of course, they're not. And there are better examples of this.

Consider the differential. It is in essence a purely mechanical computer. A clutch-type limited slip differential receives rotational data from the axles and adjusts the clamping force inside the diff via it's clutches to increase or reduce locking force between the axles. This response is highly tunable. It affects how "freely" the car can turn, drastically changing the way it responds and the level of grip in the driven wheels.

Racing cars, and even a lot of road cars these days, have electronically controlled tuning of this differential and have had so for many years. Not only can the driver tune, on the fly from the cockpit, the response of the diff, but a computer program can even do it for them, millisecond by millisecond, in the middle of a corner. Ferraris have included this type of differential for a number of years. The Nissan GT-R also has these types of diffs in it's all wheel drive system, as do other high end, very fast sports cars. These programs take away the inherent problems with "set and forget" mechanical diffs and allow it to adapt to new situations that don't quite fit the original mechanical tuning of the differential.

If you're an ardent racing fan you may remember the Williams Formula 1 team using another type of electronic system in 1991 and 1992. They took the fight very successfully to McLaren with their active suspension system in '92. The car could resist lowering it's nose while braking, or rolling in the corners, and even redistribute the weight of the car to help the driver.

Ayrton Senna was one driver who was adamantly against the tech at that time. He was driving for McLaren, and McLaren was getting hammered by Williams. Williams' cars were better. Ayrton famously came out against the tech by saying it took little skill to drive. Of course, he did not say this in 1987 when Lotus first tested the tech on his car for the season. In '87, Ayrton loved it. Being a driver, I know that we will complain about anything if things aren't going our way. Nigel Mansell, one of the Williams drivers at that time, loved the tech, and he said in an interview that it had greater possibilities to come. Nigel was the first driver to use active suspension in an F1 season in '83, driving for Lotus. He won the '92 season with the Williams system.

Nigel's prediction for great things might have been, assuming the system was allowed to be developed. Which it wasn't, because 1993 was the last time F1 raced with any kind of electronically controlled active suspension. Imagine what kind of technology we would have today, almost two decades later, had the rules not forbid its use, or the use of any other electronic system. The only car that I know of to have an active suspension system like in the old F1 cars is the McLaren MP4-12c, and the Pagani Huayra. Other cars have magnetic ride control (also amazing), but it is not the same as adjusting the attitude of the car with hydraulics.

Paddle shift technology was also being developed around that time in F1, and look at the amazingly fast and smooth systems we have on cars now. There are even paddle shift transmissions that do not have a drop in torque while changing gears. Of course, that has been banned in F1 too...

Which gets us right to the most prominent part of the argument. Paddle shift versus stick shift. The full, automatic, precise, fast experience versus the slow, imprecise, more involved system for swappin' cogs.

A paddle shift is faster, no doubt about that. Ignoring the seamless shift transmissions, a dual-clutch gearbox with electronically controlled, hydraulically actuated paddles can shift in less than 10 milliseconds. For reference, and blink of a human eye is usually between 300 and 400 milliseconds. And you can get this tech in Volkswagens. Now that's impressive. And that feels good to drive.

But does it make you less of a driver? I mean, does it make you less important as a cog in the machine of driver and car?


Think about it. If you're no longer focusing on downshifting, heel and toeing, while braking at 100MPH to slow for the next corner and be ready to accelerate out of it, don't you think you'll be able to take that same concentration you used for shifting and apply it to braking? This is, of course, assuming the program works well. A bad component is a bad component, whether electronic or mechanical.

Will it make you faster? Definitely. Just like getting better tires or a better diff or better whatever. Will it make you faster than the guy that was already faster than you using the old unassisted cars if you were to both race the new assisted car? No. Well, most likely no. If the old fast driver fails to adapt to the new systems, he wasn't really the fastest driver anyway. Assuming the cars are identical, of course. The only time a driver will not be critically important to the car is when a driver does not need to do anything to control the car. As long as the driver has control of the direction and speed of the car, this game will be all about skill. Naturally, we must also distinguish between a safety feature and a performance feature. Things like stability control and traction control can be tuned to be safe, rather than fast.

So, in this whole affair, have we solved the whole pure/impure driving thing?

I don't think there was a thing in the first place. Because we don't even know what "pure driving" is yet. So I'll tell you what I think.

Pure driving is driving any vehicle, in any condition, to the best of your ability given the situation. That means being the fastest driver. The smoothest driver. The most violent driver. The safest driver. The most ruthless driver. The most kind driver. Any kind of driver. Purity resides with the driver, not the car.

That also means driving all types of cars with skill and precision. Coupes, sedans, convertibles, trucks, SUVs, cars with different layouts and configurations, cars with and without electronics. A pure driver should do it all, and love it all.

So I don't think electronics infringe on driving purity. I believe the opposite - they enrich driving.