TOMMY HAAKE

Now available for sale here: https://www.shapeways.com/product/LXNXBUPGY/naca-duct-for-first-generation-miata-na

Being an openly gear-headed member of society, I own and occasionally race a first generation Mazda Miata. It's a fantastic little car with a surprising amount of race-minded engineering behind it and a huge following in the car community. 

It does however have a shortcoming in the cooling department as it uses the originally transverse engine from a Protegé and coverts that to an inline 4 cylinder. In that conversion, the exhaust and the intake end up on the same side of the engine compartment heating the intake air and making it less dense. In a car that had 132hp brand new, we need all the punch we can get. That's where NACA's aerodynamic intake comes in. 

NACA (which eventually became NASA) originally developed this intake in 1945 for use on jet aircraft and eventually de-classified the design tables which allowed me to calculate and build my own. The objectives were to build an efficient, functioning NACA duct that could bolt onto a Miata without further modifications to the car and provide cool air to the air filter. 

The Headlight lid was an obvious location for the project as it is an isolated body panel and small and easy to work on. The space was measured and the duct was designed in Solidworks and sent to Shapeways for 3D printing (click here to purchase the raw duct). The hexagonal pattern on the back of the printed part is to save on cost as they charge by material volume. I purchased an extra headlight lid and temporarily mounted the duct with bolts for it's proof-of-concept drive. I took to the track and with the very sophisticated temperature sensing instrument attached to my arm determined that the under-hood temperatures had moved from "Hey, it's hot. I probably shouldn't stick my hand in there!" to "I mean, it's a bit warm." which I consider to be a large improvement (actual numbers forthcoming). 

Upon a successful test, the bolts came out, the bondo went in and hours and hours of sanding begins. In hindsight, bolting the trial run to the lid was a mistake for a couple of reasons: 1. Holes. I made a lot of them.  2. The bolts flattened the body panel and I had to fix that warping with lots of sanding.

After all of that was fixed, the seam between the duct and the panel needed to be addressed and sharpened. The sharp 90deg corner is important aerodynamically for good intake (so is the spline curve of the bottom and the airfoil at the back). 

Once finished with many weeks of sanding up to 2000 grit and polishing with rubbing compound, it was time to pull a mold. I used Smooth On's Vytaflex 20 Urethane rubber compound and had a great and easy time with it. I made a small mistake by leaving the product out too long and got a few errors in the mold that you will see a bit later. Totally my mistake. Check those expiration dates, folks!

Now that I had a flexible rubber mold I could start my fiberglass layup. I used 4 and 6 OZ cloth with West System epoxy resin from a local supplier. I found epoxy resin is so very much worth the extra expense as it doesn't smell up the neighborhood up and makes a fantastically hard and quality surface. Pictured here is the first prototype layup out of the mold and it's finishing process including mounting.

With the prototype on the car finally I put together a little Arduino temperature sensing circuit and stuck the sensor in the air path just behind the air filter. I was excited to see the peak difference between the hot and cool air at 30.76°F and the average difference at 10.16°F. This is before moving the air filter closer to the duct and before building an isolation box for the cool air. I couldn't be more pleased!

Next up a new mold is being made, I'll be employing gelcoat and some other tricks to improve the finish, and final products will start rolling out!

Star tours was always my favorite ride at Disneyland and now with the phasing out of the old ride, I've set out to build a working replica of the Disneyland version Captain Rex. His almost Clouseau-esque bumbling humor always struck a chord with me. This will be a project I take my time on and get right.

I've enlisted the help of the replica prop forum and have collected mountains of reference photography, behind the scenes pictures and known measurements. I have made countless fun discoveries about the construction of Rex, including the fact that his mouth is a C3PO eye. It would seem as is often the case, if it was on hand, it was used.

Feeling confident I had a pretty good grasp of the scale and feel of Rex in general, I set out to make a CAD model of his head. I got about half done when a comment by Adam Savage on a video about his ZF-1 replica made me stop in my tracks. He said that without fail you'll always build replicas from photos about 10% too big. At that moment a dimension seemed to jump out at me. 5.2 inches for the diameter of the ear. 5.2 inches seemed awfully large. Something was wrong. I was working off of reference photography but perspective and lens distortion is always going to be an issue. So I figured, why not just print out an ear, take it to Disneyland and hold it as close to Rex as possible and see how it feels? As soon as it got on the 3D printing table it felt too large and I printed two more in half inch smaller intervals. Upon taking them to Disneyland, it was clear the 4.5" ear was correct and I trudged home, not sure where to go for accurate dimensions.

Defeated for the moment, I turned to procuring found Items and greeblies from local scrap and surplus stores. Marman clamps, Hartwell latches, and a grey neck coil as close as I could get it. I also took the time to start thinking about the mechanical and electrical side of things. I collected two lead-screw linear actuators, started learning Arduino and found a servo easing library, reverse engineered the visor levers, and made and shared some paper props of stickers seen around Rex in his current state of disrepair. 

Getting back on the horse of a proper CAD model in pursuit of a nice set  of blueprints, and armed with a great profile shot of Rex, I made a line drawing of his head in illustrator. This proved to be much more useful than trying to do this on the fly in Solidworks as decisions on edge locations can be made, then tweaked much easier. The front view line drawing is next and a full solid model in progress after.

I still yearn for the moment I get in contact with an imagineer who opens a drawer in his roll top desk and says,"Here ya go, kid. The plans for Rex!". Until then, I march on!

    I've been fortunate to be able to work in a few effects shops since early 2015. Many projects I can't share yet but those that I can are to the left.

    The first project I can share is a four-color LED light suit for practical lighting effects. Being able to selectively color  fade LEDs on set gives production a great deal of control and flexibility. This suit is capable of hour plus run-times and can be wirelessly controlled via DMX. The suit remains comfortable and safe for the actor while outputting a tremendous amount of light. For this showroom display I programmed an Arduino to handle the fades and button presses and built a controller board with MOSFETs to handle the high load power switching (around 30A for the whole suit).