30 Sep 2009

The great performance of 321 stainless comes at a very steep price, but it shouldn’t fail, ever. Part of the reason it’s so expensive is having two 2-1 collectors fabricated, with double-slip joints no less.

Been doing a lot of thinking about how to solve the cable drag problem with the shifter. After talking it through with some engineer buddies, it seems the best way is to remove the offender, the cable handling lateral shifter motion. But, doing so means admitting I messed up, so it becomes an issue of how to fix it and not let my ego get in the way.

With Kimini, and now Midlana, I’ve done pretty well getting things right the first time, so in the rare case it goes wrong, it doesn’t sit well – at all. A sense of failure, along with the lost time and wasted money, is a bitter pill to swallow. Part of me wants to put blinders on and just say, “It’ll be fine,” moving on to other parts of the car – you know, ignore it and it’ll go away. But for this particular problem it’s enough of a concern it has to be dealt with. All it takes is one single mis-shift, going from second back into first instead of third, and it’ll grenade the engine… what’s prevention of <em>that</em> worth? More testing this weekend.

27 Sep 2009

Didn’t accomplish a lot; for some reason my mind is unsettled and wandering. Messed about with the shifter springs but don’t yet have a good solution – push-pull cable drag isn’t helping.

Did get the fuse block mounted – it’s going at one end of the glove box. The various power relays are mounting to the opposite side of the same panel. With those in, wiring can commence.

Did a mock-up of the brake and clutch master cylinders… they’ll work – just. The master cylinders are just high enough to properly feed fluid to the cylinders, but with nothing to spare. Better is to use the physically shorter pedal assembly to gain a valuable inch of vertical space, but I’m using a standard-height assembly – because I have it… and because I’m hard-headed.

24 Sep 2009

Received several sets of springs; the “winning couple” will serve as the shifter centering springs.

Made the big purchase today for what I hope is all the 321 stainless tubing necessary to make the turbo exhaust manifold. To save time, the supplier is making two two-into-one collectors to fit my divided T03 flange. They’ll have double-slip joints to prevent cracking of the header assembly. Yes I could make them myself, and as much fun as it would be, it takes a long time to get all the angles right. I’ll derive as much fun putting the rest of it together 🙂

Being a divided turbine means running two wastegates. Thinking it through I’m almost sure one big wastegate can work, running a divided tube right up to its valve face. However, big wastegates are very expensive, much more than two smaller units with an equivalent larger total valve area. Of course, had I included the additional tubing into the cost calculations, it might have been smarter just to use the large one…

Anyhow, running two wastegates means running a tube out each side of the turbo flange, with the forward one having to do an unfortunate U-turn to get it to the rear of the turbo, but oh well. If this isn’t clear it will be once the build starts. Until the tubing shows up, there’s a shifter to finish and virtually all of the electrical.

In other news, I saw several car trailers heading to San Diego for this weekend’s Vintage Race weekend. I took Kimini to that several years ago and several people asked if I’m going this year… nope. If I’m there it’ll be bugging me that I’m not working on the car because Sunday is Build Time.

20 Sep 2009

As another example of how everything is interconnected, the arbitrary goal is to start the engine, even if for just a few seconds. To do that requires all the engine wiring, which means shortening the harness and placing the ECU, but it would be good to have the fuse block in place, which is to go in the dash somewhere, dictated by where the instruments are not. That requires figuring out where the instruments go, since their placement is more important than the fuse box, and on and on.

Laying out the instruments consumed the entire day. It’s pretty cool how little room the flat dash consumes. In addition are the usual switches: lights, turn signals, wipers, ignition, hazard, horn, fan, fuel pump, and a couple spares. The cross-hatching is area either blocked by the steering wheel or unavailable due to the support frame. It ensures everything is both visible and easy to reach. It’s hard to see but most of the switches are in pencil to the right of the steering wheel. Not sure where to mount the big heavy duty battery switch. It may go just below the dash and ahead of the shifter.

The right side of the dash is left blank for a real, actual, working glove box. Inside it will be the fuse block, relays, and space for glasses, papers, etc.

My buddy Cecil dropped by and asked how I was going to install the dash cover without risking cracking the paint. I don’t know. Worst case, the split in the dash, which currently allows slipping it around the tubes, can be extended all the way across, making the dash three pieces. That makes it trivial to install, but leave two odd-shaped end pieces that I’m not sure how well will integrate.

One solution is make the dash cover in steel and weld it in permanently, not sure why I didn’t think of that. It avoids needing rivets at all, but also means the dash will be the same color as the tube-frame chassis and I’m not sure builders want that – I don’t. Still, it would make things much easier, though there’s always concern about paint coverage in the corners and overlaps.

Anyway, in other news, the A-arm patterns are being made – that’ll save a bunch of time on setting up the fixtures.

17 Sep 2009

Thank you for the generous offers of help regarding the CAD situation. One of the beta-builders is inputting them into AutoCad, in effect, redrawing them. Thankfully there’s only four drawings and they’re pretty simple. As for SketchUp, as long as it’s quarantined to only providing pictures for the book, it’s tolerable… I just wish Google spent as much time debugging it as they do touting their self-proclaimed awesomeness…

The exhaust manifold design is progressing, mentally if not physically. The turbo has moved a bit further from the cylinder head and closer to the intercooler to provide a straight shot to the intercooler. Tipping the turbine inlet downward allows the primaries to sweep upward into the inlet while the wastegate feeds will sweep downward, as good design dictates.

The nature of the manifold has changed somewhat, with the primaries being longer to minimize lag. Haven’t decided upon primary tube diameter yet; 1.625″ appears to be the “right” size, yet because 1.75″ 321 stainless is used in aircraft, some vendors offer it quite a bit cheaper, so we’ll see. Unlike the Kimini manifold, this one will use double slip-joints from the primaries to the collector. The time is getting near for ordering tubing.

The solution to the lateral centering for the shifter has been worked out; it’s just a matter of ordering springs and making the mounts.

16 Sep 2009

I’ve finally lost all respect for SketchUp, the “CAD” software from Google. Had I stuck with their free version I’d have written it off as a cute drawing program. However, since $12,000 for SolidWorks isn’t happening, I bought the Pro version of SketchUp because it provides output converters to allow generating files in various formats, including PDF for the book and DXF files in case I ever choose to have parts CNC-generated.

This week, DXF files were created from the A-arm drawings. The idea was to create templates to make it much easier to build up the A-arm fixtures. So the drawings were made, dimensioned, and sent off. Get a call the next day saying they won’t work, that the dimensions don’t match the drawing. What? In SketchUp, you draw something, then use the dimensioning tool to do just that. The point is, the line being measured already exists; the dimensioning tool measures it and adds the length text with lines and arrows… only, the value it comes up with doesn’t match the actual length of the line.

Example: draw a line that is 1.000″ long. Now dimension it and it will say 1.000″. Run it through their DXF converter and send it to a CNC shop, where the guys find that the actual line (which is what will be cut) is something like 0.987″ or so. WTF? I don’t know if the bug is in the core drawing package or the DXF converter, but either way it makes it useless for anything requiring precision. (Consider how a complex drawing will have these little errors accummulate…) For the book this doesn’t matter. Builders go by the dimensions listed, but it’s a complete no-go for CNC jobs.

Another very annoying bug: their PDF converter fails miserably, removing lines and adding lines as it sees fit. Completely unacceptable for book use which sucks eggs because the high-resolution looks really good. I recently read something about Google moving into products that they charge money for. One comment stuck in my head, something like “Google’s good at creating free apps but not so good at bug-free apps that people count on.” My advice is to stick with the free version which is fine for making garden sheds and such, but that’s about it.

I know this comes across as yet another rant from a grumpy pissed-off guy, but is it so unreasonable to expect people and products to do as they promise? I didn’t promise these things, they did. It’s especially irritating to have given them hundreds or thousands of dollars. All the above aside, SketchUp’s not a total loss. It serves its purpose of making drawings for the book, but that’s about it because it’s more of a toy than a serious tool.

15 Sep 2009

You may recall we were going to have our roof redone and it’s finally underway. It’s a sad reminder of the level of workmanship which can occur even from people you’re giving money to. For example, today they delivered shingles in a truck with a conveyor belt that puts them on the roof. Of course they wanted to get as close as possible so they wouldn’t have to carry them far on the roof.

What were they thinking? Are they thinking? Do they care? They’d probably laugh to learn they’ll never be allowed to work on Midlana with their demonstrated lack of care and workmanship. Running over a plant and cracking the cement… I’d be embarrassed if I did that, but apparently with this line of work comes the attitude, “f*ck it – not my house.” What a sad matra to live by. Is it a big deal? No, it’s just a bit depressing that the people who care seem to be the exception to the rule.

13 Sep 2009

Finished the shifter. Works great, though it needs a spring (like in the OEM shifter that center’s the shifter laterally, on centerline between 3rd – 4th gear. Need to figure how what type of spring to use and what strength.

With input from various Honda forums, turbo placement is settling down. I finally weighed it, expecting something like 25 lbs or so but was surprised it’s only 14 lbs. I think the heavy weight expectation’s due to holding it up for long periods of time during mockup… no complaints. Regarding the exhaust, received two 1.5″ flex joints for the wastegates. I’m not sure whether to use heavy-but-cheap 304 stainless weld-els or light-but-expensive 321. If the turbo was to go right near the exhaust manifold I’d use weld-els, but the further away it gets, the more concern there is about the weight of the manifold. It’s not trivial since the weld-el manifold would weigh 3X as much… but 321 costs about 5X as much.)

Other odds and ends: Regarding the big sheet of steel picked up last week, there is a potentially much easier way for fixturing the arms. Research is going on behind the scenes to see if it’s doable and if so I’ll describe in more detail later. Received the remote dual oil filter assembly but it’s uncertain where that’s going just yet.

10 Sep 2009

Picked up a 24″ x 24″ sheet of 0.25″ thick steel. It will serve as the base for all the suspension arm fixtures. At $50 it wasn’t cheap but builders have the option of using wood; it just means having to add screw-down tabs to all the fixture bits. Since I’m pushing hard on getting things done I went this way, the steel being much faster since I can use TIG welds to “super-glue” the parts down 😉

Also picked up enough 1/8″ material for all the suspension bits – I think. The nice thing about the steel place I use is that they have 12″ sections of this material in about a dozen different widths. Yes it’s more expensive that way, but how many people really need 20 feet, never mind having no way of getting it home.

9 Sep 2009

Worked a bit on the shifter. Drilled out the rod-ends, fabricated threaded press-in plugs to reduce the rod-ends down to match the push-pull cables, pressed them in, and was able to at least test one axis of the shifter. Works great, though it showed the transmission-end mount for one push-pull cable sheath needs “adjustment.” Oh well.

Received a remote oil filter adaptor, ordered a remote dual oil filter mount, two 1.5″ stainless flex lines for the wastegates, and an ODB-II connector. The connector is because the engine will no doubt have its share of error codes while getting it running. Even though the Hondata KPro can clear the codes, it’s nice to be able to read them without needing a laptop.

7 Sep 2009

A busy couple of days spent working on the shifter. It’s just short of done; just ordered a couple rod-ends for the transmission end. Learned some good tips which are going into the shifter chapter before I forget.

Midi is a real character. He seems to be part snake, bending into really crazy shapes. One thing he does all the time is sit at our feet facing away from us, then looks at us upside down, I’m only scratching his neck. His flexibility is really demonstrated here with him asleep – I’ve never seen a dog do that. Then there’s the standing on his hind legs, looking up into the fruit trees for rats. Haven’t managed a picture yet but he’ll do it for as long as five seconds. And finally, today I heard a great commotion outside and caught him running through what used to be a nice neat pile of leaves. Round and round he was going, a crazy dog have a great time. I guess I should have yelled at him but he was having so much fun I just took a picture. Just like a child.

5 Sep 2009

Here’s an interesting Arial Atom review performed by Edmunds. Not that an Atom test is anything new, but because the Atom in question has an especially hot 375 hp engine. Reader Raphael sent me the article, suggesting that because the car has similar weight and power as Midlana will have, the results might help validate or negate the times CarTest2000 produced. Turns out it’s <em>very</em> close: 0-60 in 2.6 sec and the quarter in 10.6 sec, and that’s on street tires.

That said, there are several big caveats that bear mention. #1: The power comes from a supercharged high-compression engine that demands 110 octane race gas (how much is that per gallon?) #2: It’s loud as hell, the consequence of high rpm, high compression, and race muffler. #3: The cost of this particular car is plain stupid: $160k? (For that much I’d buy one of Dennis Palatov’s 4WD creations.) #4 is the worst, from their summary: Something simply isn’t right with this car’s handling as it’s difficult to detect its limits not what we expected from an Atom. Understeer is substantial and difficult to overcome on the skid pad. Balance is not good. And it’s just plain scary through the slalom, where even slight throttle application in 5th gear made it want to spin. Let’s hope I can do better.

1 Sep 2009

My brother gave me grief over my calculated numbers (no doubt due in part to them being faster than his car.) His does have a point though, that I’m being overly optimistic about what Midlana’s going to weigh. As a data point, Kimini weighed 1595 lbs wet. Okay, so where’s all the weight-savings going to come from on Midlana? No composite shell, that’s about 100 lbs. No doors, there’s another 65 lbs. The engine’s 70 lbs less (very surprising.) BUT some things are heavier. There is some composite, let’s say 25 lbs. There’s a heavy turbo system, maybe 50 lbs, and the wheels and tires are quiet a bit heavier, perhaps another 50 lbs. Another 10 lbs saved by going to a flat dash. So I removed 120ish pounds, so Midlana will be around 1475. Interestingly, had I paid attention, the spreadsheet says it’ll be 1460 lbs, which is pretty close!

So the car is heavier but not by much… Running the numbers again shows a 2.7 sec 0-60 and a 10.7 quarter. Still crazy, though I expect to only see numbers like that with drag racing slicks.

This number doesn’t sit well with my brother since his Super Stalker weighs around 1680 lbs wet. He thinks I’m living in La La Land on my weight estimate – and maybe I am. His guess is Midlana will weigh in at around 1600 lbs or so. I know my drivetrain’s 170 lbs less than his, not including his 150-lb rear axle, but I have more steel in the chassis. We’ll see 🙂