Randomly Systematic, Systematically Random

Last time I checked in, I was on the verge of buying an aftermarket fuel tank. The tank itself wasn't hideously expensive, but it would not have been compatible with the Miata fuel pump and sender - and by the time I did an aftermarket inline fuel pump and then having to chase down the different fuel lines and trying to make an aftermarket pump make friends with a Mazda fuel pressure regulator, the hassles and expenses were starting to add up. 

Scott Graham of ANC/MEV Replicar Builders Group fame had a fuel tank fabricated for his car with good results - and I have the support of Pops Racer who has fabricated a large number of fuel tanks over the years, so I decided to make my own tank. I can't copy Scott's design exactly because my frame has some additional strengthening members that weren't present on the early Replicars. So my first designs in SolidWorks had a shelf in the bottom to stay within those pesky frame rails.

Looking up from below. The deep sump sticking out the bottom is necessary if I want to use the OEM Miata fuel pump and fuel level sender. Keeping the sump just in the middle, with "shelves" on either side, was necessary to clear the frame rails - but adds a lot of hassle to the build. Solidworks said this was a 13+ gallon tank, so way bigger than I need anyway.

My next idea was to lop off the far side shelf. Shelves are a pain, so just one shelf is half the pain of two shelves. I built a cardboard mock up of this, and also built a surface for it to sit on.

Yes, I'm aware that it's plywood. It's marine grade plywood, actually. I think this will be a good material for the trunk floor (and maybe sides) because it's light and relatively inexpensive and easy to work with. When this is done, I will reinforce all the joints with fiberglass tape and coat the whole thing in fiberglass resin. I could paint it, but I'm tempted to leave it in the clear resin so people can look in the trunk and say "Augh! You made your car out of wood! Can you do that? Is it legal?"

The fuel tank mock up above does have the "shelf" for added fuel capacity on the near side, but loses the shelf on the far side. This will leave a nice spot to put the battery as well as simplifying the fuel tank build. But while I'm on a simplifying tear, I thought it better to eliminate the shelves on both sides. So this is the current version of the design:

Why are all the corners angled off? Well, not all of them. The corners that are bends are just bends. But for the welded corners I'm doing one of Pops' favorite tricks. The main challenge of welding sheet metal is that it wants to expand, contract, and warp when you apply the heat to it. There are a number of ways to deal with this, but one of the better ones is to bend the metal parallel and proximate to the weld. This way you get a butt joint instead of an angle joint, and the bend resists warping (much the same way a piece of angle iron is stiffer than the same material flat). So the welds will run down the middle of all the angled-off corners. Yes, a really good welder can execute a difficult angle joint weld in sheet metal. But I'm not a really good welder. Pops is an extremely good welder, but even for someone at his skill level, if you set up the job to make easy welds the quality will be better than if you make him execute difficult welds all the way around.

But then before I started actually fabricating the tank I was gently reminded by Sir John of the Royal Order of Jedi Kit Car Knights Templar that it would be extremely disappointing to make a gas tank and then find out that it doesn't fit under the bodywork. Test fitting the body opens up a whole new set of projects, including making a different harness for the bodywork that will be out of the way when lowering it onto the frame. And I have a bunch of other jobs halfway completed, and all of which are easier with the car jacked up as high as my jack stands will go, which is where it is now, whereas for the body test fitting it will be easier to have the car on wheel dollies. So, it probably makes sense to finish those first. 

The version of the tank above is just slightly more than 8 gallons... which is small... but is it too small? I'm not sure. When I lower the body on, I'll see how this fits and may find it possible to add a little to the height or to the anteroposterior depth and sneak in another gallon or gallon and a half. 

A Systematic Approach

One of my favorite YouTube project car guys is SuperfastMatt. His projects are different than mine but I learn a lot by seeing how he solves problems, as well as his sense of humor when he creates problems. One of my favorite SuperfastMatt lines is, roughly, "I learned my lessons so I don't make the same mistakes. Instead, I make completely new mistakes."

I've previously linked to the video above. It's more about the psychology of car projects than the engineering or mechanical aspects, and one of the things he recommends is trying to stick with one system at a time. This way your task list seems more manageable and you get the satisfaction of finishing a system, before you open up the next can of worms... I mean, system. 

Unfortunately I have found this to be almost impossible to do. I want to get the wiring done, but I can't do that until I have the gauge clusters ready to go and I can't finish the gauge clusters until I figure out how deep of a cutaway I need in the dash for the steering column, but I can't figure that until I know how high I will be mounting the steering column and I won't know that until I can sit in the seats. And so I have a large number of systems partially completed, all waiting on each other. 

The upside of this is that sometimes the stars and planets all line up and a bunch of things get done in rapid succession. In a previous post I referred to this phase as "deconstipation." This can be a really satisfying time because I go out to the garage and work for a relatively short amount of time and make what seems like big progress.

One "System" Down

This isn't much of "system" - it's really a small part of the electrical system - so, charitably a "subsystem" - and an unnecessarily complicated one at that. But one of the things I was going to get picky about was how the horn sounds and there are a lot of vehicle horns I do not like the sound of. In my last post I showed a picture of the air horn installed - but an air horn also requires an air compressor so I guess it does qualify as a "system."

I needed to mount the air compressor somewhere so I designed this part in Solidworks and then 3d printed it in ABS plastic. 

The question always comes up: how strong are 3d printed ABS parts? It depends on what kind of strength you're talking about. This mounting flange is plenty strong enough to support the weight of the compressor, even with some bouncing around. But you can see where I cracked the part a little bit by over-tightening the upper mounting bolt. There are ways I could adjust the Gcode to make the bolt hole stronger - basically by making the walls of the bolt holes thicker. But to borrow another SuperfastMatt thing, I think this is Good Enough^TM  for this job.

Overall, I'd say 3d printed ABS is very handy for irregular parts that don't have to be very strong. If I wanted the same part stronger, I could send it out for 3d printing in metal, or mill it out on Pops' CNC machine.

It was kind of satisfying to get this "system" done, though, because it can be made to do something.

The video does not do justice to this horn. In fact, it does injustice in two ways. One; it's a heck of a lot louder than you'd think from the video. Two; despite being very loud indeed, it's not at all unpleasant. I attribute this to it being an actual air horn, creating sound by blowing compressed air over a diaphragm, much like a woodwind instrument. I wouldn't exactly call it "musical," but it's a lot more musical than the shrill square wave nastiness produced by a piezoelectric horn.

Actual Systems

After getting the radiator installed I was close to having a complete cooling system. Or at least cooling was the one that felt closest to being done. All that was left was installing the heater core... but the heater core would obstruct access to the upper forward part of the transmission tunnel so that had to be installed first.

(above) Underside of the transmission tunnel with heat/sound insulation placed. There are significant sources of heat and noise under the tunnel and large aluminum sheets are really good at transmitting both. The passenger side in particular has the exhaust downpipe running just inside the tunnel and although my lovely wife is remarkably longsuffering, burns to her right leg would probably detract from her willingness to ride with me.

This (above) is the kind of thing that's easy to do in the Mazda factory, but not so easy in the garage - making a circular hole with two notches on the sides, all just the right size for the speedometer cable grommet assembly to fit through and lock into place (left). Making large holes neatly in sheet metal is kind of tricky to begin with, and especially in odd sizes. I think this hole had to be 1.167 inches diameter to be big enough for the housing to fit through, but still small enough for its retainers to not fit through. Fortunately Pops Racer has an astonishing collection of hole-making tools and one was within a couple thousandths of the size hole I needed. Then drilling two smaller holes and squaring them off with a file for the final Saturn-shaped hole. Worth the effort, though, because now the speedo cable can pass through the tunnel into the cockpit, but heat and fumes and weather can't. Unfortunately the side nobody can see, underneath the trans tunnel (lower right) is quite a bit more attractive than the plastic housing that protrudes into the car interior and holds everything in place (upper right). However this is at the very forward edge of the trans tunnel, up under the dash, so it shouldn't be very visible.

Then I ran into another problem. With the insulation inside the trans tunnel and the trim I wanted to use around the edges, the trans tunnel flange built into the Replicar frame was too short. There was enough material (just barely) that I could have semi-permanently riveted the trans tunnel to the original flange, but I thought it would be better if the tunnel could be removed. See, I've gotten spoiled being able to access the transmission and the exhaust and a bunch of other stuff from above, working on things with light and comfort and a jaunty attitude like I own the place. So I wanted to use riv-nuts and screws to hold the trans tunnel in place and the original flange wasn't tall enough for that. Solution: a taller aluminum flange added to the original flange, and install the riv-nuts in that. In the picture above you can see the aluminum strips that run all the way around the trans tunnel opening. I suppose I could have taken a picture with the riv-nuts installed. I love riv-nuts.

To attach the pieces of my new taller flange together, I tried that funky aluminum brazing rod product. It's not as terrible as you'd expect for something advertised on social media, but it also isn't very pretty. You can see a joint in the picture above, where the horizontal aluminum strip joins its upward-angling neighbor. I could have taken this out to Pops' shop and TIG welded it that would be a lot of hassle for a joint that probably didn't really need to be welded or brazed anyway since it's held together by solid pieces of metal on both sides.

So here's the trans tunnel AND the heater core fully installed. Which completes an Actual System: the Cooling System. Unlike the "horn system," which I tongue-in-cheek called a "system" in order to award myself a participation prize, the cooling system is legitimately a system. Also, the sharp-eyed reader will notice that the steering column is no longer "installed" with bungee cords, but actually installed.

The VERY sharp-eyed reader may notice that most of the button head screws holding the trans tunnel in place have a rivet just above or just below (some, both above AND below). This, annoyingly, is the result of my process of discovering that the trans tunnel flange built into the frame wasn't tall enough for my install. So, mistakes were made, and then covered up by a conspiracy of extra rivets. More annoyingly, the rivets had to sit flush on the back so as not to create a gap between the trans tunnel and my extended flanges - which meant tapping the mandrel out of each rivet, then using a sheet metal hammer and a hand dolly to flatten them. All of them. Every. Single. One. 

Most annoyingly, this is the kind of thing that serves as a perpetual reminder of my fallibility. Every time I see these extra rivets, I will remember the frustrations of this process. They will haunt me like evil taunting ghosts for the remainder of my days on Earth. At least, that's how it's gone with other projects in the past. For whatever reason, this isn't bothering me too much. Yet.

Probably more likely to haunt me is this little tangle of hose clamps, straight connectors, T-connectors, hoses, and fittings. There is some unnecessary complexity here because I decided to add a bypass to the heater core. One way to manage this installation is to allow hot engine coolant to always flow through the heater core; the idea being that it won't actually heat the cockpit unless the heater core fans are engaged. I got mixed opinions about this - ranging from "if the fans aren't on you won't notice at all" to "you will broil your feet in the Summer." Interestingly many of the "you won't notice" crowd live in England or Scotland, which share the same latitude as Moscow and are known to be a bit damp at times. Indiana people tended more toward the "broil your feet" school of thought. In the end, what decided me is that a heater core bypass gives me a good excuse to have a pull cable control in the cockpit. 

Ahh, more 1960s pickup truck goodness. The steering column flange (background, in focus) and the pull cable actuator (foreground, out of focus) are both strongly reminiscent of pickup trucks that were made in the same prehistoric era I was. I haven't actually seen anything like either of these in an old British roadster, but from my own distorted perspective these things feel like period correct vintage nostalgia.

Random tangent on period correct vintage nostalgia: these are the tires we got for Taylor the Mighty's go-kart resurrection project. For those not from hereabouts, "Hoosier" is a nickname (or epithet, in some circles) for people from Indiana. When I was a kid, Pops Racer used Hoosier tires on some of his race cars. So Hoosier is a very nostalgic hometown roots brand for me. If Hoosier made street legal tires that fit my car I certainly would have bought them instead of the Continentals. It would be possible for an unscrupulous faker to get a Hoosier stencil and make the unmarked Contintentals look like Hoosiers. This would be period correct for the car, because Hoosier has been around forever, but it would be very genre incorrect. Historically, Hoosier tires were associated with USAC sprint car racing; subsequently branching out into other kinds of racing... but never anything close to the kind of racing Aston Martin DBR-1s were designed for. So now I have to decide between Personal Vintage Nostalgia Correct and Period Time and Place Correct. 

Desperately Trying to Relocate my Long-Lost Topic

Ah, here it is, I found it: we were discussing self congratulation over the completion of an Actual System.

So here is the engine bay with its one gloriously completed system. The choice for white coolant hoses was driven by Unpopular Color on Sale considerations but I've decided I like it. In fact, I like it enough to be a little bit disgruntled by the not-white vacuum lines. The two coolant lines leading through firewall to the heater core (behind the engine, toward the right side of the picture) look so neat and tidy, in sharp contrast to the aforementioned tangle on the other side. 
Tidy Engine Bay is in fact one of my mission goals for this project. It's just such a satisfying thing to pop the hood and see everything looking neat and orderly and not too tightly crammed together. It's even satisfying just to say. Tidy Engine Bay. Tidy Engine Bay. Aaahh: bask in the peaceful goodness of it.
The current level of tidiness will be challenged once I start adding wiring.

Also, the sharp eyed reader will notice that there is no coolant overflow tank anywhere in the picture above. So it isn't really a Completed System. 
But that's easy. I already have the tank. All I have to do is figure out where to mount it, come up with some kind of mounting bracket (another good candidate for 3d printed ABS), finish the bracket, install riv-nuts to mount the bracket, install the overflow tank, run the hose from the radiator overflow to the tank, and secure the hose. Easy. 
So, here it is: my first Completed System.*

Starting a New System

Last weekend we had a very welcome visit from Long Lost Miata John, who ran almost all of the brake lines. So this is not far from being a Completed System. There are just a couple of short lines to run, and then all the lines have to be anchored down so they don't flop all over the place, gradually self-destructing, to catastrophically fail on a steep downhill curvy road with cliffs. So yeah, I'll need to secure all the brake lines.

The Long-Awaited Seats

Lots of jobs were waiting on the seats. When last I checked in, I was contemplating custom made seats versus similarly styled seats ordered from BadBoy Classics in the UK. Cost, and Custom Seat Guy's apparent unwillingness to get back to me with an actual quote, finally decided the matter. I placed the order and they showed up four days later.

Pops Racer's admonition was "don't get cheap seats with no suspension springs." So I got cheap seats with suspension springs. 

Actually the quality of these seats seems really good - much better than I had any right to expect at this price. They're very pleasant to sit in, a least, for the short duration sit tests I have undertaken so far.

Here's the driver side seat, not really installed, but just sitting in position. I think these seats will be a good fit for the styling I'm going for.

When I was initially communicating with Custom Seat Guy I took this interior styling preview picture, showing the various materials and colors and textures. I like the way this is coming together.

It does raise the question what to do with the aluminum floor panels. Pristine aluminum, when polished and clean, looks fantastic. However pristine, polished, clean, and aluminum cannot be said in the same sentence without adding "maintenance hassle" and "Argh! Another scratch!"

For the trans tunnel, I took suggestions from the Replicar Builders Group and did a deliberately de-shined finish. Random orbital sander and Scotch-Brite pad wound up achieving the look I liked best. The rearward part of the trans tunnel and the forward part are different aluminum alloys. Actually the forward part came from ANC in the UK, so it's aluminium and the rearward part came from Metal Supermarket Beth's scrap pile here in Indianapolis, so it's aluminum. So, being completely different metals (or at least different alloys) they will never look exactly the same. However, being de-shined covers a multitude of scratches, mismatches, and other sins. 

I didn't think de-shining would be adequate for the floors, because you can still scratch de-shined aluminum, creating a jagged shiny thing that is no longer de-shined. And, being on the floor, I think that would be almost impossible to avoid. 
I explored a lot of options and decided to go with a Custom Designed Polymaterial Laminate Automotive Flooring. Otherwise known as heat/sound insulation I already had laying around applied to the floor, and inexpensive marine anti-slip surface applied to that. 

Stylistically, the marine anti-slip surface looks perhaps more appropriate for a modern vehicle but I think this will be cheap, effective, and not really very noticeable with all the shiny wood and de-shined metal and black leather and machine-turned shiny metal in the cockpit to distract from it. So, to once again borrow a SuperfastMatt trademark, I think this will be Good Enough^TM