Sunday, July 20, 2008

1995 M3 Issues

There are a few important maintenance items that are really must-do on a 1995 M3. They are all actually pretty difficult to do but if you don't, and they fail, you will wish you did.

1. Weak Valve Retainers

Early 1995s (prior to 10/95) have improperly heat-treated valve retainers that can fail causing a catastrophic engine failure.

More details here


2. Loose Oil Pump Nut

The nut attaching the drive pulley to the oil pump can come loose causing a total loss of oil pressure and a few seconds later, catastrophic engine failure.

More details here

3. Loose Rear Diff Bolts

The bolts attaching the LSD cover to the carrier may not have had locktite applied properly at the factory allowing bolts to fall out and be ingested by the ring/pinion gears causing catastrophic differential failure.

More details here

So, go check your nuts, bolts and retainers...

Saturday, July 19, 2008

Miller Photos

Better late than never.

I finally connected with the photographer from the Miller NASA event. Here are a couple of really nice shots.

Wednesday, July 16, 2008

More about roll cages

Ranked by importance, my criteria for optimum cage design is

1. A safer cage is better

2. A stiffer cage is better
3.
More driver space is better
4. Easier driver egress is better
5. A lighter cage is better

The following series of charts is from a thread on bimmerforum.com about cage design. One of the members did a series of mathematical simulations of a roll cage designs using ProE to build the model and Mechanica to run the simulations. These simulations estimate what happens to a roll cage when a load is applied to one corner. The goal is to understand the stiffness of each cage design and the trade-offs involved in different designs.

This approach has a few weaknesses, primarily the lack of a similar model of the unibody chassis in order to evaluate how the two structures work in concert.

That said, this is really interesting work and bears thorough review.

This first picture shows a basic 6-point rollcage with “X” style door bars, harness bar, and a rear diagonal.

In this simulation, a 500 Lb. upward force is applied to the driver’s A-pillar leg and the strain on the rest of the cage tubes is represented graphically.

Notice how large the windshield and door openings are in this design – probably not good for stiffness.

Here is a second chart showing what happens if you add an A-pillar down-bar, change the driver’s door-bar to allow easier access, add a bar from driver’s shoulder to rear bulkhead, and small upper windshield gussets. The stiffness of the cage increases by over 60% with the addition of only a couple of strategically placed bars.


Here we have added a bar from the A-pillar straight to the floor and dash bar gussets. This is 76% stiffer and weighs only 23 Lbs more then the original design.

All the models so far have used 1.5” .120 wall DOM tubes. What happens when you go to 1.75 .090 wall DOM? Cage weight goes down 2.5% and stiffness goes up 7%.

This chart shows the difference in stiffness between two different A-pillar configurations. The design on the left is about 10% stiffer than the design on the right. Clearly there are trade offs to be made here, as the new A-pillar bar on the left is pretty close to the driver.

Bryan Watts has this design (with an additional gusset) in his E36 M3 and says it doesn’t interfere with driver’s vision at all.


The door bar configuration shown here – typically called “Porsche style because it is derived from the a Porsche GT-3 Cup car – is about as stiff as the stiffest designs and makes entry/exit easier and because they bow outward they provide more room in the drivers “office”.



This chart shows the stiffest design to date without crazy braces in the windshield area. It has a fairly radical A-pillar bar, small roof and dash bar gussets and a roof diagonal all in1.75x.090 DOM.

At this point I’m planning a cage like the stiffest shown above but with either Porsche-style door-bars or with a standard X. Both of options are pretty close to the ultimate in stiffness. The standard X would allow me to keep the windows in the car. Unless you have an enclosed trailer (not in my plans for a few years) then windows are really practical. The Porsche door bars can be made to just fit while keeping the window glass, but removing the door panels (which are pretty thick on the E36).

I will also extend the cage forward from low on the A-pillar bar to connect to the firewall. This provides some additional protection for the drivers feet.


The final key feature I'll add is a jacking tube integrated into the door bars. This creates a reinforced point from which to lift the car. A short "spike" attaches to a floor jack effectvely preventing the car from slipping off the jack.

Monday, July 14, 2008

Pueblo #2

Saturday it was back to Pueblo for the NASA Summer Beach Party Event. The weather was a little cooler than the 95s we had been seeing last week. It was a nice change. The car ran great, no major problems.

I pitted with a large group of BMW drivers who regularly attend NASA events. I met some of them at the BMWCCA School at Pueblo a couple of months ago. It was nice to feel like part of group for the first time. I think many open-track drivers are reluctant to admit new attendees into the clique until they have show some kind of driving proficiency.

The new HeroCam needs a little debugging. The camera eats batteries at a disturbing rate. The Users Guide recommends buying a couple of sets of rechargeable batteries and now I understand why. The camera seems to get about 90 minutes of recording out of a pair of AAAs. The problem is that the SD card holds 56 minutes of video. As a result, I need to install new batteries for every session. Home Depot sells 16 packs of AAAs for about $10. Rechargables are about $16 for 4 and the charger is about $25. So, for a typical track day with 4 sessions, I'd need 8-10 batteries. Call it 10 for $40 + $25 for the charger = $65.

Non-rechargables cost about $6.20 for the same 10 batteries so it takes 10 track days (or 5 weekends) to recoup the cost of the rechargables. I'll do that many next year for sure so I think I'll get the rechargables.

I also need to upgrade the Traqmate software from TraqView to TraqStudio ($199) so I can integrate video with the Traqmate data on a single screen.

Here are a couple of shots of the camera mounted in the car. For forward looking stuff I used the suction cup mount on the inside of the windscreen with the camer mounted updide down.

I also added a mount to the rear of the car using the supplied plastic mount that attaches to the car with a super adhesive. The camera and case weigh only a few ounces so this is actually a pretty rugged mount.

When I actually mounted the camera, I added a little safety wire between the camera and the license plate frame just in case.

This shows the complete camera mounted on the windscreen. The camera is mounted upside-down here and has a setting to record upside-down so I don't have to flip it in post

The picture quality is acceptable and very usable to YouTube and other web publishing. It even looks OK on a big-screen if you are sitting far enough back.

Because of the battery problems I only captured two sessions. I'll add some video here soon.

Tuesday, July 08, 2008

Monday, July 07, 2008

Video

I spent most of a day Sunday trying to mount a video system in the car. Being in the video industry I planned the ultimate in-car video system. I have a Sony HVR Z1 U 4:4:4 1080P HD camera looking forward and a Canon XH-A1 1080i HD camera pointing aft. Using our Red ONE 4k camera seemed like overkill.

The problem is that pro- and pro-sumer cameras are large and heavy, each of the HD cams weighs something over 5 pounds and is large enough to make them all but impossible to mount safely in a car without a cage. They won't fit in the rear window becasue they are too tall. I fabed an arm the attaches between the stock head-rest mounts but found it too flexible for even a single camera. Hard braking dropped the point of aim by 6 inches. After a bunch of trial fitting and fabrication, I have decided on another, very different, path.

I gave up on the HD setup and ordered a GoPro HeroCam. $179 + free shipping. It is solid state, light-weight, comes with lots of mounting options and is designed for just this application. Plus it's in a hard plastic case to allow it to be mounted outside the car. Something quite out of the question with someone elses $5k HD cam.

I'll revisit the full 2-screen HD setup when the cage is in.

...and the Big Brown Truck just dropped off a new Nady PMC-2x Intercom so driver and passenger can chat in comfort.

Counting the days till NASA on Saturday -- lots of new things to try.

Saturday, July 05, 2008

Chunking Tire

I finally unwrapped my tires from the Miller event. Here's a photo of my passenger front tire showing significant chunking both on the center band and the edges of inner tread block.

This is suprising because the stock suspension lacks enough negative camber for track work. This is confirmed by some significant wear on the outside shoulder of front tires. The opposite side showed very minor chunking -- 2-3 small chunks total.

I'm not sure when this occured -- I suspect it was saturday afternoon. It was really hot, 98 F, and blazing sun, the track was probably close to 120 F. I was pushing pretty hard as well. There was a little push in a couple of corners but nothing out of hand.

I think the centeral nature of the chunking indicates too much air pressure. Except -- you can clearly see the mold line separating the left and right halves of the tread. If the air pressure is really too high wouldn't you see enough wear on the center rib to obliterate the mold line?

Saturday afternoon I dropped pressures and ended up at 35 Lbs hot in front. I guess I'll drop to 33 but I'm concerned about anything that might increase wear on the outside shoulder.