26 Aug 2008

Changed the rear suspension layout yet again. The shocks have moved to the back corners where they’re more accessible and the suspension arms now consist of lateral and trailing links – basically, really big A-arms – all to give the drivetrain plenty of room. Once the decision was made to have the engine tray removable, it’s natural to attach all the suspension links to it, and maybe even the rocker-arm pivots. Though things are still in flux the design is settling down, really.

25 Aug 2008

Yesterday was materially non-productive but a fundamental decision was made; the bottom engine frame/tray will be removable. It allows running tubes as needed to produce a strong structure without compromising them to get the engine in and out. The tray will bolt across the base of the main roll hoop and to the rear down tubes – just like Kimini. This stuff is the hardest part of the project, where all the decisions will come back to bite me if they aren’t thought through very carefully. Everything that comes after this is icing on the cake!

22 Aug 2008

Being back at my day job gives time to reflect on various design decisions – during break time of course. One consideration is whether to push on with a traditional A-arm rear suspension, or change over to what was used on Kimini, long lateral links and even longer trailing links. Doing so would eliminate one of the two bulkheads behind the engine, freeing up room and simplifying the layout. My only concern is the trailing links, long tubes heading back from the main roll hoop plane to the suspension uprights. In the case of an accident, they might be driven forward into the gas tank(!), or even into the passenger compartment. One solution is to use tall U-shaped channels to attach them to, which also act as “catch fences” for broken or bent trailing arms. The large bracket catches the failed tube and prevent it from passing forward of the main roll hoop plane.

Even if the above is done, it doesn’t avoid the need for one transverse bulkhead behind the engine to attach the lateral suspension links to. The bulkhead is a prime issue when it comes time to install or remove the drivetrain. The trick is to have few – or no – tubes that have to be removable; I’m not sure yet if it’s achievable.

In other news, philanthropist/actor/race car driver/race team owner Paul Newman is expected to pass away within a few weeks – cancer. I had the honor to see him in person once in 1984 at the Los Angeles Times Grand Prix. I was walking through the pits and there he was, sitting on the end of a bench all by himself, race suit around his waist, kicking his feet back and forth like a little kid. I was to shy to walk talk to him but in hindsight I know he’d have liked to talk about the Datsuns we both once owned. Anyhow, that’s how I’ll remember him, someone who obviously really enjoyed what he was doing. When he was out on track he was just one of the guys, and a very competent driver in his own right. With the millions donated by his “Newman’s Own” food brand, he’s done a lot to help people. I read one time that he said that he acted in order to fund his racing and I don’t doubt it. I enjoyed him most in “Butch Cassidy and the Sundance Kid” and “The Sting” – a simpler time in Hollywood. Yes, he’s had a pretty good ride. I think that’s the best we can do here, where when it’s over, people say, “you done good.”

To end on a higher note, Kimini is virtually sold; the deposit has been received and the buyer has through September to pick her up. Her new home will be Salt Lake City, Utah, where the new owner said that I can come visit her. I might just take him up on that.

16 Aug 2008

Heard from Mitchell Software… my mistake – big surprise there. I somehow changed the rotation point, an obscure parameter I had a hard time finding in order to set back to zero. How I managed to accidentally get into that menu, set it wrong, and forget doing so beats me.

15 Aug 2008

Between catching the flu and awaiting an answer on strange results I’m getting from the Mitchell suspension software, I’m starting on the CAD drawings – may as well do something productive. Doing the drawings is already proving educational, forcing the need to address tubes that aren’t even on the mockup yet. One thing to decide is how to pass the coolant lines, shifter and throttle cables, brake and clutch lines, and wires from the front of the car to the back. Some people have suggested running the coolant lines outboard, but that doesn’t really simplify anything. They still need to be covered to avoid burns or having boiling coolant spraying about in case of a cracked line or accident. Even if they were run there, the throttle and clutch lines lend themselves best to running down the center of the car anyway, so may as well run everything that way.

14 Aug 2008

Today the decision was made on where and how the rear shocks will mount.

There’s going to be rocker-arms at the rear. Not because it’s ubersexy, just that it’s the most appropriate solution. Why? The Miata rear upright has its upper suspension pickup cushioned in a big rubber bushing. Putting a screwdriver through the bushing and pulling it from side to side shows that it’s fairly soft. I was going to mount the bottom of the shock to it, but applying an offset load of ~400 lbs to it isn’t a good idea. A solid bushing could be substituted for the rubber, but one design goal is to avoid lathe-work. On top of that, the steeply-inclined shocks gave too low an installation ratio, requiring spring rates higher than what’s commonly available in 10-12″ springs. They were also too close to the engine and exhaust for comfort, and that’s on my drivetrain. Who knows how close they’d be to whatever drivetrains builders are going to install. Tilting the shocks to be more upright wasn’t an option, either, because they’d stick through the engine cover with nothing to attach to. I would have attached the shocks directly to the lower suspension pickup point on the upright, but they’d would have to hang off a bolt in single-sheer about an inch out… won’t go there.

The best solution appears to be to keep the shocks away from the drivetrain and keep them low. While putting them behind the drivetrain makes them accessible, the exhaust is back there, too, cooking anything close by. For these reasons they’re being placed vertically, about a foot forward of axle centerline.

Speaking of the rubber bushing, if a single rod-end is bolted to one end, it’ll twist the bushing. If a U-shaped bracket is used so a bolt passes all the way through the assembly (like a Miata) then it’ll both bind slightly when toe is adjusted, and move “some amount” during cornering. I’ll probably leave it as-is until I figure out how much of an issue it is.

13 Aug 2008

With the front end sorted and the axles having arrived, attention returned to the rear suspension. Push-rod rear suspension was almost considered a requirement, but after spending hours staring at it and trying different solutions, an old-school shock layout was found. However, it’s very tempting to modify the Miata upright to make it work even better, but that has to be thought about a bit more.

It’s not easy routing the chassis around the engine, and while a de Dion axle layout makes things simpler, I don’t want to settle for almost-great, may as well do the best I can. A true independent suspension is also the most flexible, for those builders who want to tinker with things – or just leave it as-is.

One thing I’ve been ignoring is how to remove the drivetrain; haven’t decided if it’s coming out the top or the bottom. I kind of like the bottom-exit solution because the drivetrain doesn’t have to be lifted in and out, scratching up chassis tubes. Just wheel something under the car for the engine to sit on and lift the car up off it. Either way it means some of the tubes need to be removable, just haven’t decided which ones yet.

I have two more days to get everything sorted (before the wife comes home) and I think it’s doable, at least by the end of the weekend. Well, “sorted” in the sense that tube placement has been decided, then it’ll take weeks to put it all into CAD. I’m concentrating on only one side of the mockup; no point doubling the work of making the other half.

I realized that it’s impossible to go straight to steel without completely recording every dimension. Since I don’t have space to set the mockup aside once steel construction starts, it has to be chopped up for firewood before I can. If I missed some critical dimension, well, that would be unfortunate. For that reason it’ll go to CAD first. Now that I’m nearly out of the hardest part of the project, I’m starting to change my mind on how to do mockups. That is, I now think it would be easier and maybe even faster to do it all in CAD (“next time”) though detailed dimensions are still needed of all major components.

While taking Cooper for his morning walk, we had another encounter with an animal. Much like the ferret from a few years back, we were walking along a trail and out of a bush came a light tan and white rabbit. It must have been someone’s escaped pet, due to the coloration and also because he walked right up to us. I thought of picking him up, but then what? Cooper, who thinks rabbits taste wonderful, had never had one actually walk right up to him – he didn’t know what to do. When he finally decided that it might really be a rabbit, it took off when he barked at it. Hope the little fellow finds his way home; being in the wild with white markings is like having a bulls-eye on your head.

11 Aug 2008

Here’s the axles, twin-disc clutch, radiator fan, and a poor shot of the new chassis layout. Connecting the axles to the Miata uprights and Honda drivetrain for the first time made me very nervous, wondering if I’d screwed up the specs somehow, but it all went together fine (whew!) They’re rated at 450 hp – to make sure there are no issues with the turbo engine!

Changed the side tubing placement slightly; I’d been trying too hard to achieve some sort of “look” instead of pure function. I finally relented and just let the car design itself – like what happened with Kimini – and put the tubes where they want to go. What it means is that the car might end up looking “like an engineer designed it.” Be that as it may, all the tubes are right where they need to be, making the chassis lighter, cheaper, easier, and faster to build. Over the years I’ve been listening to what builders don’t like about Locost chassis fabrication and designing mine to avoid the issues. For example, builders complain about the PITA (pain-in-the-ass) compound cuts at the front of the Locost chassis. I made sure mine only requires single-plane cuts 🙂

One slight contribution to looks is that I raised the base, and lowered the top, of the windscreen to make it look less dune-buggy-like. It also had the side benefit of moving several tubes right to where they are needed, which avoids additional tubing. I’m trying very, very hard to make the chassis as simple as possible. It’s easy to make a <em>complicated</em> chassis… but very challenging to make a simple one!

I took this week off to push through the toughest part of the design, the front suspension/radiator/master cylinders/steering rack area. I’m thinking that in a day or so it’ll be time to start transferring numbers into CAD.

10 Aug 2008

I may be backing away from using push-rod suspension; “simple, cheap, and easy” might win out. The more suspension motion that’s needed, the more volume is consumed in the chassis for the rocker-arm, and with 6″ suspension travel, the arm gets too large, never mind the weight. If the shocks are located down low, they cook in radiator exhaust air and crowd the master cylinders. Moving them up out of the air duct, raising them so that they’re visible to look cool is poor form due to raising the combined 14 lb weight so high. And, placing them parallel the centerline of the car means increased side-loading on the rocker assembly due to the push-rod changing angles. A rocker setup weighs more, costs more, and takes longer to build. Eh, I think I want to go simple on this car.

Push-rod setups do look really sweet, but the reason they’re used on F1 cars has nothing to do with looks; it’s to minimize the shock’s aero drag – a situation we need not worry about. (F1 cars also have suspension motion on the order of 1″, not 6″ like we have.) The one valid argument is that rockers can greatly improve installation ratio, where the shock can be made to move 1″ when the wheel moves 1″. With a traditional outboard arrangement, the shock may move much less than the wheel so stiffer springs have to be used. Since the shocks don’t move much, there can be issues with stiction of the shock valving. However, this isn’t a problem here because of how much suspension travel we have.

After moving the parts around in CAD, it’s not bad. The installation ratio ends up about 0.6, so the spring rate has to be about 2.8 times as much (installation ratio squared.) It means that while the wheel moves 6″, the spring moves about 3.6″. A 1″-thick bump stop brings total stroke to 4.6″, and since it’s good to have some overhead, the same shocks will work fine. Checking the effective wheel rate shows a gradual rising-rate, just as desired.

In other news, I took the transaxle and Wavetrac LSD to a shop to have the LSD installed. I don’t know how many of you’ve seen the movie, “The Fast and The Furious”, but this place and the customers were right out of the movie. That said, the mechanic was amazing to watch. He had the tranny apart in five minutes and it’s good he checked it because there were several things wrong. First, one of the shift-rod ball-detents was broken. Second, transaxles have a magnet in the sump to trap metal debris. The previous builder apparently didn’t know how to reinstall it and guessed wrong. When we opened it up it had attached itself to the final drive gear! Good thing it hadn’t been run. Anyhow, he completely disassembled the gear clusters, closely inspecting each gear and syncro. They all looked good except for third gear which had sharp burrs on it from overzealous shifting. He said reusing the parts could cause it to jump out of gear – no thanks. I left it to be repaired properly.

3 Aug 2008

The decision’s been made, out came the saw and off came the bottom radiator pipe. The decision was make easier after realizing that there are other radiator modifications that need welding, so what’s one more? To fit the radiator under the nose, the filler cap flange is being replaced with a low-profile water bleeder. (The cooling system will get filled through a header tank back in the engine compartment.) Mounts are also needed, too. All the welding can be done at one time and transform a nearly-impossible-to-work-with-awkward-assembly into something that fits really well. It’ll be well worth the bother.

Worked all day on the front chassis area, probably the most complicated part of the entire design. Kind of like juggling a bowling ball, flaming torch, chicken egg, frog, and a running chainsaw, there’s the nosecone, steering rack, cooling fan, duct-work, and suspension pickup points. It’s easy to deal with these one at a time, dedicating tubes to deal with them separately. It’s about 100 times harder to use one tube to do two or three different tasks at the same time, very tricky! It’ll be worth it – after it’s done – in the form of fewer tubes doing the work. No pictures; it’s just a bunch of half-finished wood pieces at various angles – not much to look at. I plan on taking a week off soon and will dedicate the entire time to finishing the mock-up.

After the mockup’s complete, is the task of getting it into the computer. It’s the only way to force myself not to cheat, taking measurements off the wood and going straight to steel without writing the numbers down. Come to think of it, that’s not all that bad… and might actually be a good idea. If I go to CAD first, then build the steel chassis, I risk following mistakes and wasting material – and forgetting to make the corrections on the CAD drawings. If I build the steel chassis first, then take measurements, it guarantees that what’s there is really what was built, not numbers that were later changed and forgotten about. Hmmm, maybe cutting steel is not as far off as I thought…

2 Aug 2008

Ordered a muffler to make sure there’ll be enough room for it and its associated pipes. Because a turbocharger acts as a muffler, too, a smaller muffler can be used, but since most people will run a normally aspirated engine, I have to save room for a larger unit. Since neither the muffler nor axles are here yet, the rear suspension area is on hold, so attention moves to the front suspension.

Installed the radiator and – as many Locost builders discover – the lower radiator pipe interferes with the steering rack in a big way, like, going right through it. The choices are to move the rack, move the radiator, or move the radiator outlet. Car design is all about compromise, but moving the rack is a really bad idea since it messes up both Ackerman and bumpsteer. Another way to improve clearance is to move the rack (and wheels) back, shortening the wheelbase. It’s currently 96″ and while there’s nothing magic about that number, it gives a warm fuzzy 1.6:1 wheel-to-track ratio. I have to think about it.

Then there’s moving the radiator pipe, and since it’s a big 1.75″ diameter that’s already pointing the wrong direction, it’s very tempting to move it. I’m reluctant to do this because it means builders will have to weld aluminum, or pay to have it done, something I’m trying to avoid. And finally, the nose and radiator can be moved forward to free up space… but I find this solution offensive, lengthening the entire car by 4″ just to avoid one aluminum weld. I just can’t bring myself to do that.