04/01/06 (updated 04/14) Busy day today. Besides the bike stuff and the dp1 stuff (see the respective pages) I also decided to do a bit of a cosmetic mod to the Mini. When I first changed the wheels and all the chrome was still on the car they didn't look quite right. Blacking out of the grille helped but the chrome stripe that goes all the way around the bottom of the windows still looked out of place. So I got some blackout tape from an auto parts store and decided to try it. Worked really well - you can't tell it's not OEM unless you look very close. The car now has a theme to it, like Darth Vader's helmet, or more like Lord Dark Helmet from the Spaceballs movie. This is the look I was going for so I'm pleased. Maybe it needs a DRKHLMT vanity plate... Nah.

UPDATE 4/13/06: Got the black side grilles from a regular Cooper (many thanks to John DaCrema for the suggestion and for sending me the parts!). The fourth picture is with the grilles installed. I also have some smoke-colored sidemarkers on the way from England, they should stand out less than the stock white ones. It's fun to tinker with non-consequential stuff :)

UPDATE 4/14/06: The smoke-colored side markers are in. See fifth picture. I think I'm done for a while now :)

As an interesting aside, Tracy (who is a car designer by trade) told me the story of why the Mini exhausts look like they do. Apparently when the design team made a clay model of the car they had nothing to use for the exhausts so they ended up just grabbing two Coke cans and using those. Everyone liked the look so the production version ended up looking exactly like that - two Coke cans. Cool.

04/01/06 Installed a strut brace from M7. This is more of a peace-of-mind thing than any real-life performance improvement. Recently on one of the online forums there was a thread on 'mushrooming' of the shock towers. I had never heard of this type of damage before and was rather skeptical, since both the tower and the strut top plate would have to bend. However, a link was presented to some pictures of just that type of damage. And indeed the strut plate in the pictures was bent. So I looked at the strut itself and it seems there's only about 3" of travel before the strut bottoms, at which point it's a solid path from the wheel to the strut tower. So a big-enough pothole would conceivably do it and if I were to lower the car later it would only make this more likely. The M7 strut brace comes with fairly hefty aluminum plates that go on top of the towers so it serves to reinforce them a bit. I suppose those can get bent too but it'd help. The strut is fairly inexpensive so I decided why not and just ordered it. On the website the description says the color is 'titanium' which I wasn't too thrilled with (too Bling) but I was pleasantly surprised to find that the one that showed up at my door was anodized black. Less pleasant was the discovery that the underhood insulation piece has to be removed so that the hood can close. Not a big deal, just unexpected.

A quick test drive revealed just what I thought it would - no discernible difference in the way the car feels and behaves. It is a little louder without the insulation but that's not necessarily a bad thing. The design of the strut has slots for the two attachment bolts on each side, so it probably can't transfer much force anyway. I did tighten them pretty well... Anyway, it's in there now, whether or not it does anything useful is not all that important. Looks cool, so there's that.

04/27/06 There was some discussion on the forums of the effectiveness of the strut brace in a Mini so I decided to do a bit of scientific evaluation. Namely, I disconnected one side of the brace by removing the two bolts and hooked up a 'travel gauge'. The gauge is a quick improvisation, consisting of a white zip tie with two black zip ties around it, snug but not tight so they can move easily. The locations of the two black ties are marked with electrical tape. I positioned the gauge so that any movement of the brace relative to the strut tower would move either one or the other black zip tie. The white zip tie is taped to the strut tower.

Since part of the discussion was that the brace only makes itself known under hard cornering, I needed to have a record of how hard "hard" is. So I hooked up the GPS datalogger and set out for my favorite twisty road. The conditions are dry and warm so optimum tire grip is available and the KDW2s don't need to be warmed up to work at their best. As an added bonus I drove over some rough pavement on the way to see if the body would flex. Subjectively, with the brace disconnected the Mini felt as solid as ever. After waiting for a bit by the side of the road to assure a clear run I took a few turns (about a mile's worth) with some enthusiasm, then pulled over and took some pictures of the gauge. Then I bolted the brace down and did the same run with it in place. Subjectively there was no discernible difference so whether or not the brace does anything at all will be determined by the data. First, the gauge:

It seems that during the run (including rough pavement part) the tops of towers moved less than 1/16" towards each other and did not move apart at all. So technically the strut brace does do something - it prevents 1/16" movement in the towers. That may have some long-term structural integrity benefits but there is no way it can affect suspension geometry or handling. The effect of distortion in rubber bushings and tires is many times greater than any effect of such small deflection at tower tops (the tire sidewall alone distorts by over 1" as seen in some track pictures). Just to verify that I did in fact corner "hard" I looked at the logged data for the run. Several sustained turns at 0.98g and a peak of 1.12g in both directions. Not bad for street tires and yes I'd classify that as "hard" cornering :) So there you have it. Now keep in mind that this only applies to the Mini which, with the strut towers very close to the firewall, is one of the less likely designs to benefit from a strut brace. Some other cars with longitudinal engines (like many BMWs) have towers pretty far away from the firewall and so are more subject to flex. Even then, however, to be benfitial a strut brace should also triangulate the towers to the firewall. I just looked at the new MZ4 roadster and that's exactly what BMW did there - put in triangulating braces to the firewall. Just tying the struts together is of relatively limited use and when I had installed a strut brace on my first M Coupe I did not notice any difference there either.

All that said, today I received an M7 underbody brace system so that will go in next. Again, I don't expect much effect on cornering but should help overall body integrity. It'll be too difficult to do a gauge similar to the strut one so I'll just have to make do with subjective impressions. Will probably do it this weekend. I just have to try stuff, what can I say.

05/03/06 A few updates in one. First, a couple days back I got and installed the M7 Under Strut System. It basically braces the underbody.

There are three parts to it - the rear rectangular plate, the middle straight brace and the most elaborate (and in my opinion the only thing that does anything), the front multi-link brace system. The latter bit serves to triangulate and brace the front subframe. The subframe consists of some bent tubes and plates and carries the front suspension control arms so it takes all the cornering loads. The tubes are bent and drilled to collapse progressively in a crash, which is cool. There is a downside to this however and that is the fact that they are not entirely rigid when it comes to taking cornering loads. The function of the USS is to stabilize the subframe (and yes, as a consequence it most likely affects what happens in a crash). Installation is fairly straightforward although I did have a problem with part of it - see below. Once installed, the subjective evaluation is that the car does feel a bit more solid and 'tighter' under high cornering loads. Unfortunately I haven't come up with a way to quantify this so subjective opinion is all I can offer. It's not a huge difference but I believe it's real.

Now to the installation problem. The procedure calls for inserting a pair of rivnuts into the floor of the unibody. You're supposed to drill out a couple existing holes with the supplied drill bit, then install the rivnuts with a special tool and washers (all supplied with the kit). The problem is this - the holes selected for the install have two layers of metal. There is a larger hole in the bottom layer and the smaller one (to be enlarged by drilling) in the upper layer. When one attempts to install the rivnut as suggested, it does not 'catch' the upper layer but instead expands below it and just attaches itself to the washer and not the car.

To fix this I got some new rivnuts and installed them without washers into the upper layer of metal alone.

That in itself went fine, but what I should have done is used a washer between the insert and the brace. Sadly I thought of it only in hindsight, after attempting to tighten the bolts securing the brace to the body. What that did, predictably enough, is pull the rivnuts down and distort the uppper layer of metal. So now I'm not at all comfortable with the durability of the install. I went ahead and ordered some larger rivnuts and once they arrive I'll remove the old ones, drill the hole out to the size of the larger one in the bottom layer and then install the larger rivnuts. Hopefully that'll fix it. So the verdict is - interesting concept, seems to actually do something (however subtle), needs fine-tuning.

The next item is that I finally put on the track wheels/tires in preparation for a trackday Friday. These are much lighter than the street set and are smaller, too. The wheels are 16" and the tires are a full 2" smaller in outside diameter, meaning that the car drops 1" when sitting on these. It also looks quite goofy, with huge gaps above and around the wheels.

David dropped by with his M Coupe and I thought the two cars side by side looked rather comical.

We'll see how well it works on the track. I looked at some GPS data comparing the earlier Mini run on street tires to Larry's laps in Spec Ford Racer. The SFR actually accelerates slower due to pretty modest power but it carries a lot more corner speed because of its R compound tires. The Mini on KDW2 did a best of 1.1g and sustained 1.0g whereas SFR did 1.5g best and 1.3g sustained. The racer was about 9 seconds faster around the track, with all of the difference being gained in the corners (Mini was actually getting a tiny bit of time back on the straights). So if the R compounds get me to something like 1.25-1.3g, which is entirely reasonable and if I can actually use them (remains to be seen) then a sub-1:30 lap should be within reach. And at least theoretically something in 1:28 range ought to be possible. If true, that would be about dead even with Ferrari Enzo on stock tires and faster than both F40 and F50. A tall order, but we'll see. I'm really looking forward to Friday.

The final bit also has to do with tires. Jist before swapping to the track rubber I happened to look at the KDW2s I have on there now and saw this:

Yep, that's a block of tread missing from the middle of the tire. On closer look it seems to have delaminated somehow and fallen off. I've never seen anything like this. E-mailed the pics to Tire Rack, we'll see what they say. I still think they're great tires but this is odd.

05/05/06 I think I'm just going to call the Mini DarkHelmet. It fits. So today was DarkHelmet's first outing on R compounds. A Cascade driving school event, it is run differently from many trackdays in that their emphasis is on the school part (including a 4-hour classroom session the night before). Those with track experience can skip the class, which I did, having taken it a couple years ago. So I got put into the single HPDE run group along with about 25 other people of varied skill levels. This resulted in very crowded conditions and most people were quite slow. There was one white STI that could hang with me in terms of overall lap times (he was faster on the straights due to nearly double the power, I had him in the corners because I was on R compounds and he wasn't). There was also a blue STI on hoosiers and with modified motor who was a second or two faster than me. The rest, including two C5 Vettes, were easy prey in the corners but tough to pass on the straights since DarkHelmet only has the stock 170 hp at its disposal. The first session was essentially a checkout run with an instructor riding along and no passing on the back straight. I started in the back and it took me the whole session to pick my way through the crowd with just a single clean lap at the very end. Session two I was the third car out, behind a Focus and a Vette. By the time I got by the Ford the Vette had pulled some distance so it took me two laps to catch and pass him - just as I caught up to the back of the pack. Frustrating. For whatever reason people don't lift to let the Mini pass so I ended up letting the Vette through and had him play icebreaker for me. He'd start the pass and the slow car would lift a little so I then slid through on the Vette's tail. Psychological drafting of sorts. Session three I lined up first and got one good clean lap before catching the back of the pack - that one ended up being the fastest for the day at 1:33.17 (here's a video of it - a 12.8M file which coincidentally is the 100th video on my site). This is barely faster than the best I've done in the Elise on street tires and not too bad overall, but still far short of what the car should be able to do (see below). Once I caught up to the pack my original plan was to just slow way down and let them pull away, then catch up again. It's the kind of 'rubberbanding' I often do on the street when passing is impractical. The mirrors were clear so I slowed to a crawl but about 10 seconds later a white STI came barreling down the straight so I decided to see if he can hang with me. Of course we caught up to the crowd in a few seconds but passing the backmarkers turned out to be not too problematic and we had a very nice session blasting through traffic. The following and final session we talked our way to the front of the line and thanks to two very kind STI drivers we had a couple clean laps to play. Once in traffic again I let the STI through to get the draft in passing. That worked pretty well until a Firebird let him pass but floored it before I could slip by. It took a lap of hanging on the Pontiac's bumper to convince him to let me through but by then the STI was gone and I never caught up again in traffic. Overall, a day of some fun mixed in with frustration and more tire issues... More on that later but first, a look at GPS data. Below are DarkHelmet's traces on BFGs and Kumhos.

The first thing to notice is that I'm finally improving my speed through turn 10 (at cursor), now carrying 85 mph compared to previous 79. Another interesting thing is the effect of the smaller/lighter wheel and tire combination on forward acceleration (bottom trace). Notice that the red line is consistently higher than the black. Max acceleration on 17" wheels is 0.27g, on 16" it's 0.37g. Also note the earlier shift points due to overall gearing reduction. So when people ask how much difference wheels/tires really make, the answer is 'quite a bit'. Overall, so far the car is exactly 3 seconds faster on track rubber than on the street set, but it should be more.... According to the data, consistently throughout the day the car was pulling 1.25g in left turns (vs 1.10g on street tires) but only 1.10g in right hand ones (vs 1.06g). This is very odd and frustrating because most of the turns at PIR are to the right and the speed through them really impacts overall lap time. If I could only get 1.25g in right-handers I'm certain a 1:30 lap is doable. I'm defiinitely sliding the car equally in all corners and the g loads are very consistent lap to lap so I'm pretty sure I'm getting all it has to give under the circumstances. So the potential causes are operator error, track surface or tire problem. Or just offset in the data which is the most likely explanation. But the tire theory started looking plausible when I examined the left front tire after the last session - the tread is delaminating!

There is of course still the possibility that I'm driving it wrong and the delamination is the effect rather than the cause. Regardless, it's a brand-new race tire so this is very disappointing. Especially in light of what happened to the street tires. On those, TireRack claims BFG won't warranty the street tire because it looks like road hazard damage to them. OK, whatever, I think that tire would be fine on the back and I'll just be running it on the street anyway so I'll let it go, at least for now. This one, however, is a different story. It's much more expensive, for one, and is purposely designed for track duty. The Mini is a relatively light car and I wasn't going all THAT fast. So on Monday I get to talk to TireRack again and see if they call this 'road hazard' as well. Actually, the 710 was initially recalled due to tread delamination problems but supposedly that got fixed. Of course there is no warranty on track tires... Not having much luck with tires lately, am I? Ah well, the price of having fun. I just need to sort it all out before the BMW trackday on the 19th.

05/09/06 Received and installed the new rivnuts. My suspicion was correct, the old ones were working their way out already. It took very little effort to remove them.

The new ones have 1/2-13 thread and are considerably larger. Most importantly they grip both layers of metal and the installation is very solid. I also used silicon sealer to prevent corrosion in the drilled hole. The 11/16" drill bit was quite pricey though - $22. But now I'm much happier about the install. The brace has much larger holes in it than the bolts originally supplied (again relying on friction to carry whatever loads it does carry) so I did not have to modify the brace at all despite going from an M10 bolt to 1/2" (equivalent to M13). Perhaps tellingly there was no evidence of the brace moving at all despite a pretty loose connection to the unibody so again one has to wonder if it really does much. I don't think I'm going to go to all the trouble of finding out for sure. It's on there, subjectively I think it may be helping, so I'll leave it at that. At least now it's on there good.

On the tire front, I took the wounded Kumho to ANT Tire, the local experts who deal with race tires a lot. The verdict - yeah, the tire is blistered from heat. Yeah, a Mini will do that to an R compound when driven hard. Guess I was driving hard. So I ordered a replacement from Tire Rack but now I'm faced with having to do something about this. Burning through one tire per trackday is not something I look forward to. Garth at ANT suggested a couple of things. One is more negative camber. Another is running some toe-out in front. Since camber plates cost $500 I'm not too eager to go that route just yet. What I did decide to do is get some springs. At $200 they're relatively cheap, will improve the car's looks by dropping it about 1.25" and will help the tire situation in several ways. First, the drop will result in a little bit of negative camber (not as much as I'd like but halfway there anyway). Second, both the lower CG and the stiffer spring rate should significantly reduce roll which should reduce the need for negative camber in the first place. As an additional step I'm going to drill the rear A/R bar and move the links to stiffen it a bit, just like I did on the Elise. It's an easy (and free) mod that should help some. So the springs are on order and should be here shortly. I should have those in and the replacement tire mounted before the next trackday. We'll see what the results are. One benefit is that it will be quite obvious what effect, if any, the springs have on both lap times and tire wear since that will be the only major mod between the trackdays and GPS will tell the tale in no uncertain terms.

UPDATE: just after posting this I got a suggestion to get camber plates from Ireland Engineering which are much more reasonable price-wise (thanks, Dennis!) so I went ahead and ordered a set of fixed plates for $180. No better time to add those than when I have the suspension apart for springs.