31 Jan

When the new springs were installed, the shocks were set full-soft to ease installation. For no particular reason, they were then set to 4-clicks as a starting point, even though they were at 8-clicks with the old springs. A quick test over speed bumps showed that the combination was too soft, giving a ride even softer than with the original springs. The shocks were increased to 6-clicks, which felt about right, firm but not harsh. It does make me wonder whether the shocks might have contributed to the problem; stiff shocks coupled with soft springs make the suspension act stiff during transitions, but not once stabilized. That is, in a long enough turn, the shocks could have had enough time to slowly compress until eventually the soft springs allowed the suspension to bottom – instant spin. It’s just a theory, but it might explain in part why the car did what it did.

Speed bumps aren’t really a valid test of whether the spring rate is now enough to prevent bottoming, but it’s a start. The bump stops were pushed down to serve as travel indicators and over the next few weeks they’ll be monitored. Still, it won’t be until the car’s back out on-track that total travel can be measured and quantified. No testing could happen today due to rain.

26 Jan

The springs arrived.

A buddy wants me to make a bolt-on undertray for track events. Basically a huge venturi, the theory is that it would provide some good downforce without much drag. The unknown is how to deal with air flowing in from the sides; I don’t care to mess with sliding skirts. Anyway, that’s on the back burner for now; I rather make incremental changes because the results are easier to quantify. I’ll start with the springs, then new tires, and go from there. That said, if the undertray is bolt-on, it would be easy enough to do some back-to-back track tests. One thing at a time.

24 Jan

In preparation for replacing the springs the shocks were removed and disassembled. Being an engineer, I wanted to calculate the forces on the rear suspension, but realized doing so was pointless – I already know the suspension was bottoming. The new springs should be here Tuesday or Wednesday and it’ll be back together by the weekend. A test drive will show whether it’s tolerable – though the answer will have to be “it’ll do.”

Sized up the proposed mods for the rear lower panel, probably cutting it back to right below the license plate and replacing it with stainless mesh the same as the existing mesh above. Also, I’m seriously considering making a true diffuser for the bottom back edge of the car. It would be the full width of the chassis and trail behind by 46-61 cm (18-24″). Haven’t decided what to do about the dented and scratched side panel; some have suggested proudly leaving it like a scar from a bar fight. We’ll see.

23 Jan

I forgot that the installation ratio of the rear springs is 0.75, not 1.0. This is especially significant because that value is squared when calculating wheel rate, so the current 400 lb/in springs give roughly 200 lb/in at the wheel – same as the front suspension. Due to the rear of the car carrying twice the weight of the front however, when the car hits a dip, the rear suspension will compress twice as much. As the pictures show, the rear tire was moving as much as 4″, which corresponds to 3″ at the shock. While the upper-most marks may have been caused by the off, there were marks there before and nearly as high. Chalk it up to being distracted by the engine nonsense for why I didn’t pay more attention to this.

The question becomes: how much should rear spring rate be increased? I don’t know how much the rear suspension was trying to compress; it stopped only because it hit the bump stop, so how far would it have gone if there was no bump stop? Since the rear of the car has 4″ of ground clearance, that’s the lowest it’ll ever go, but it still doesn’t answer the question for the same reason. I’m going to guess that worst case it might have compressed 5″ (3.75″ at the shock), so that needs to be reduced to no greater than 3″ (when it hits the bump rubber). That works out to changing the existing springs from 400 lbs to roughly 650 lbs.

In a related matter, Midlana only exhibits understeer (push) at very low speed (slower than 30 mph). Because it never exhibited this on-track, it means the front spring rate needs to be increased as well, more than the rear. This gets into the consequences of picking springs that might be best for the track, but result in an intolerable ride on the street. That wouldn’t be any fun, yet neither is spinning off the road due to the rear suspension bottoming out!

I know some people will say just increase shock stiffness. That’s only a solution in transient conditions; the situation in which my off occurred was in a long right-hand turn, after the shocks came to equilibrium, so they don’t play a part in this. Other people may suggest stiffening the anti-roll bars. Well… I don’t have any – for now. I’ll start with increasing the spring rates and go from there. For what it’s worth, I had heavy duty front and rear bars on my Datsun 1200 that was used for both autocross and track day events. For autocross, the rear bar was absolutely necessary in order to get the car to rotate. Conversely, that same rear bar at high-speed track events caused the car to be a real (slow) handful, the rear of the car always wanting to come around. I ended up disconnecting the rear bar for track events, so if a bar ends up necessary for Midlana, it’ll be at the front only.

21 Jan

Arrived Friday afternoon and set up immediately. After dinner, started the heater and played around with the new camera a bit. The dark picture of the track was taken with a 30-second exposure in total darkness other than the lights on the building – I couldn’t see the mountain at all. The shot of Orion was 20-seconds. Saturday morning was cold but fine with the good sleeping bag and heater. The pretty red cars are an Alfa 4C and an older Alfa GTV. I’ve always thought the old Alfa engines were works of art. Scott went out first in his very fast appliance and you can see how his tires wore well. In comparison, mine look like they aren’t being worked and/or never came up to temperature. My brakes look fine though, front and back working about the same. And then there’s the spin/off damage. I wasn’t kidding when I say I picked about 10 pounds of dirt out of the car – plus some stones jammed between the tires and wheels. I’m currently considering what to do about the damage, whether to leave some as-is or rework those areas with changes that were already on the back burner

Regarding why I spun off, my brother came up with a very compelling theory – that the rear suspension may be bottoming due to being overly soft. Early on he had a devil of a time getting his Stalker to corner before realizing that his was bottoming and causing an instant spin. He raised his car 1/2″ and the rest is history. I’ll take some measurements and run some numbers, but the car’s handling matches that theory, that it handles great – until it doesn’t, then the rear instantly goes around and can’t be caught. Its behavior does seem to fit the evidence, but we’ll see.

19 Jan

Friday my brother Scott and I took very different routes to Willow Springs raceway yet arrived within 10 minutes of each other, even though I took I-15, the route that everyone uses on Fridays to head to Vegas or skiing, plugging up the freeway. We both arrived about 4 pm which allowed time to set up everything, as night falls quickly this time of year. At the same time, temperatures drop fast in the desert, so I brought along a propane heater intended for construction sites – it worked great, with the 20-lb bottle lasting two nights and two mornings. In addition, some years back I bought a highly-rated sleeping bag, wondering if was worth it. It worked awesome even on the morning with ice. It’s amazing how well a 2-lb sleeping bag can work.

Saturday morning we were up at dawn, eating scrambled eggs, sausage, apples, and hot chocolate – no coffee on race days. Driver’s meeting at 7:30, then Scott headed out in his group which allowed passing anywhere in the straights and with point-bys in the turns. Not wanting to get myself in trouble I signed up for the other group, which allows passing only in the straights with a point-by, but no passing in the corners. I say “not getting in trouble” because even though Midlana has been done for several years, I’ve never had a safe environment where it could be pushed to find out what it would do at high speed.

Next it was my turn and everything went fine, I did a lot of passing and no one passed me. When pointed by, well, it was as if they put their brakes on. I should have run in Scott’s group because a few people either didn’t see me or just decided they weren’t going to move over – I’m not sure which is worse. Whatever issues Midlana might have, sufficient power isn’t one of them. I even caught up to a really nice Acura NSX, a car which I’ve always liked; I guess short of owning one, catching one on the track gets pretty close. Over the day I gradually built up speed, getting down to 1:38, aware though that Willow is famous for its soft dirt, rocks, ruts, and embankments off the paved course.

The datalogs showed that speed down the back straight was about 130 and 135-140 down the front straight. One thing really cool about the AIM dash was how seamlessly it timed laps after the track map was loaded into it. It was one of those rare moments where a complicated product “just worked”, so I have to take back some of my negative points it gets due to its poor CAN bus interface.

Anyhow, the excitement came after lunch when I entered the high-speed decreasing radius right Turn 8-9 combination that is famous for eating cars, and unfortunately it was my turn. Part way through Turn 8, the back end stepped out slightly, I corrected by steering slightly left, let it settle – I thought – and turned back to the right, but the slide had apparently never stopped and kept getting worse. Once it was clear I was no longer in control, I put both feet in and just held on. As you’ll see in the video, I ended up back on-course, having done a full 360 spin. Once I realized where I was I quickly got going for fear that someone might not see me. Even though there’s good visibility, the thought of a 100 mph car coming through was a little scary.

It was a good thing that I came straight in after the off. In three wheels, small rocks had become lodged between the tire and wheel, allowing the tires to rapidly lose air. Within about 15 minutes, three were flat, so a trip to the track tire shop was worth the expense, otherwise it meant removing all four tires and heading into town to find a tire shop.

Thinking about what happened, plus asking around, I came up with the following:
1. I may have contributed to the spin by doing the classic “Porsche 911 reverse parking maneuver”, where when the back of a mid-engine car steps out, the driver lifts off the gas. You can usually get away with doing so in a front-engine car, but in a tail-heavy car it can make the situation worse. However, knowing what to do is one thing, doing it is another; it’s very counter intuitive to keep your foot down when things start to go wrong.
2. It was suggested that I may have helped the slide develop by first making a small correction and then turning in again, which may have caused the back to break loose because the small correction increased lateral motion and momentum, then abruptly halting it at the front with steering. I’ve watched the clip a dozen times and I’m still not sure.
3. My tires are literally about half as sticky as my brother’s (they’re supposedly the same compound) and our cars weigh about the same. Turns out that my tires were manufactured six years ago(!), so maybe a big part of this is that they’re old and the compound has hardened.
4. Aero lift may have caused the rear of the car to become light, making an oversteer condition more likely and being less able to control it. This seems unlikely since there was no wind.
5. A previous car may have dropped oil or coolant, or kicked sand onto the track. Also seems unlikely.
6. Everyone complained about the bumps in the track in that area, so that didn’t help.
7. Not enough rear camber, probably, which is currently at -1.0 degree
8. Not enough rear tire pressure, maybe, which is currently 16 psi cold (though tire wear was even).
9. Possibly need a front anti-roll bar – but I consider it the last thing to do; bars lower traction at the end that’s working better.
10. Some combination of #1-9.

Other interesting tidbits:
– The days were cool, around 18 C
– Oil temp peaked at 90 C (194 F)
– Coolant stayed at 80-85 C (about 185 F)
– Midlana used roughly 5 gallons of ethanol per 20-minute session; my brother used about the same amount of gas.
– The ECU datalogs showed that Midlana was rarely at full throttle. There are only three areas where it’s foot-to-the-floor; the rest of the time was spent partial-throttle managing anything that wasn’t a straightaway.

The entire track was run using only 4th and 5th. In hindsight, I gave up time by not downshifting to 3rd for the upper loop of the track. That area of the track has a lower risk of damage since it’s the slowest section, but I was lazy, thinking that it wasn’t worth the bother – maybe I was wrong.

On Sunday, four more sessions were scheduled, but due to the time needed to pack up, the 3.5-hour drive home, and returning the trailer to U-Haul before they closed, we ran just the first session and called it a day. Not surprisingly I went slower than on Saturday due to being spooked by my off; the session was instead spent practicing proper lines through the turns and staying behind slower cars to study their lines. (I thought it was funny that just loafing around the track during the cool-off lap produced a 2:01, which was faster than some cars get when trying hard.

Lastly, my brother and I were extremely impressed by a class of cars referred to as “Super Miata.” It’s a bit pretentious since the mods are very slight, consisting of engine intake, exhaust, and ECU, which results in maybe 140 hp. Tires are limited to 200 wear index. I ran a 1:38, my brother ran a 1:29, and the fastest of those Miatas turns a 1:34 – very humbling. They claimed to corner at 1.7-1.8 G, which seems almost impossible (videos online seem to support 1.2-1.4G) That’s about the only way they can get such a fast lap time in a 140 hp car – you just never brake. No, I don’t plan to sell Midlana and buy a Miata.

The GoPro was once again a disappointment; I recorded four sessions yet it contained six when it was downloaded. It just decides on its own when to start and stop recording. Fortunately it was running when I had my off and interestingly it recorded the impact and spin just fine, no interrupted recording, which it seems to save for other random times. Sometimes I’d go to turn it on and it was an absolute brick – dead as a door nail, and had to fuss with it for about 10 minutes, then it would suddenly start working like nothing was ever wrong. Another thing it’s done several times is forget that it’s been configured for “one-button recording.” Digging down into the configuration menus showed that indeed, it had just forgotten how it was set up… that’s nice. I’m currently looking at alternatives.

The next event is at the Streets of Willow, a smaller tighter track next to Willow Springs. What’s really timely is that they have a large skidpad, perfect for assessing handling. I want to know what happens when the back starts coming out, can I catch it, can I control it, can I drift it controllably? When the back starts coming around, what happens if I take my foot off the gas; what happens when I keep my foot down or even increase throttle?

Years ago my brother took a BMW driving school and found what helped him the most was driving a figure-8 while trying to keep the car sideways the entire time. The purpose was to understand how a car feels while sliding in a safe environment and how to be comfortable controlling it. I want to do the same and am working with the organizers to set aside time for that, as the skidpad is normally ignored and is part of the active track. Since my tires are already half-worn and are likely in poor condition, if they end up bald at the end the weekend, that’s okay since they’ll be replaced anyway. Best of all, the whole thing will provide website visitors with something they keep pestering me for: shameless videos of “cocking about.”

Alright, enough typing. I’ll add pictures and video soon…

17 Jan

This weekend my brother and I attended a track day event at Willow Springs. I have a lot of pictures as well as video but it’ll take time to compose everything, so check back over the next week for updates. I had a big off at Turn 8-9, an area famous for eating cars. I’m fine but Midlana now proudly sports some battle damage. More soon.

7 Jan

The start of a new year and a new diary – like always, the old diary has been moved to the Old Diaries page, use the above link.

Finally got time to tick some things off the to-do list, fortunately which didn’t require driving the car, what with all the rain we’ve been (finally) getting.

First was checking and setting rear camber and toe. When the rear rod ends were upgraded, the new ones were screwed in the same number of turns as the old ones, so I figured they’d be plenty close and didn’t bother checking alignment. During later drives though, the car definitely felt less stable – not a good feeling when your hands involuntarily tighten on the steering wheel. Another related issue is that the car had always required a constant bit of left-steer which seems a little odd. Regardless how the front toe is set, the steering wheel should be able to be turned to balance out any differences left-to-right, so there shouldn’t be any consistent pull in the steering (caster was fine). This is brought up because if the rear toe isn’t identical left and right, it applies a constant steering input that has to get countered by the steering wheel.

The first thing was to measure rear camber; the right rear was about -0.7 degrees which is about right, but the left rear camber was slightly positive… Not sure how it could have migrated so far. The lower A-arm was adjusted outward about 0.25″, which by chance resulted in the same camber as the right side.

Next was checking and setting toe, which is deceivingly simple. How hard can it be: measure the distance between two similar grooves on the tire, both on the forward and rear side of the tire. Turns out that it’s not that simple at the rear of the car. Depending the sequence of steps, rear toe may end up exactly “correct”, yet be totally wrong. How? Unlike at the front of the car where steering balances left and right toe, the rear is fixed. This means there are actually three dimensions that have to be set: total toe AND left and right toe off centerline. Unless toe from centerline is correct, the rear suspension will forever apply a steering vector which has to be corrected by steering. What a pain; due to the nature of suspension in general, I came this close to taking the car to an alignment shop. Not because the actual adjustments were difficult, but because the results were so squirrel. Setting total toe took hours because it would be adjusted, yet didn’t appear to change, so it was adjusted a bit more, then it changed too much. I’m fully aware of stiction and play, but the rod ends are all new and have no discernibly play, plus the car was rocked and the suspension resettled after each adjustment, but it was still elusive. Anyhow, at the end of an annoying day, rear total toe is 0.125″ and equal left and right.

Next was mounting the fire extinguisher. While I always intended to have one onboard, where to put it was a problem; every place was either deemed too accessible for seedy people, or it fowled my feet, or it was too inaccessible to be of any immediate use. I finally settled on attaching it to under the glove box. Haven’t decided whether to leave it there all the time or only for track events. Have to see how much it interferes with passengers’ legs.