Home » Meet The Guy 3D Printing A Widebody Kit For His Mazda Miata

Meet The Guy 3D Printing A Widebody Kit For His Mazda Miata

3d Printed Widebody
ADVERTISEMENT

Body kits were once exclusively something you’d pick out and buy from an aftermarket supplier. Then you’d have a body shop fit the pieces to your car, do some paint and panel, and call it good. You can still do it that way of course, but you can also DIY the whole thing yourself, from design to manufacturing, if you’ve got the skills, tools, and know-how. Architecture student Brent Foster does, and he’s doing exactly that.

Brent’s a Mazda Miata man, and he’s been customizing his ride for some time. He’s already given it a sweet pink paint job and a dope set of wheels, but now he’s in the process of taking things up a notch. He’s designing and building his own widebody kit from scratch.

Vidframe Min Top
Vidframe Min Bottom

The trick to Brent’s build? He’s doing it all with the aid of 3D printing.

ADVERTISEMENT

Brent started his Miata build with some basic customization. But now he’s going all out on a widebody kit.

“I first got into 3D printing back in my sophomore year of high school,” explains Brent. “I took a class called Design Tech, and a large chunk of the curriculum was learning Autodesk Inventor and 3D printing designs made in that program.” Right away, he was hooked, and he quickly saved up to buy his first printer—an Afinia H480. “It was a very very small, very unadvanced printer at the time but I thought it was the coolest thing ever,” he laughs.

Now at college, Brent found himself inspired to leverage these skills in pursuit of his thesis of architecture. Having grown tired of designing buildings that would only ever live on paper, he had a thirst to create something that would actually exist in the real world. “I have had a passion for cars for years now, so I figured designing a body kit for my Miata would be a fun way to challenge and further develop my design skills while still being able to build something in real life,” he explains.

437564610 1569856120469914 3452302131137504394 N

435123133 380643864975232 4531421275709583469 N
Brent created a striking curvaceous widebody design that also includes a nice ducktail spoiler at the rear.

While 3D printing has become popular in some parts of the automotive world, it’s mostly used for small trim pieces or simple mockups. Brent wanted to do something bigger. “I wanted to show that full-scale models can be made with standard 3D printers, and anyone can do it if they want to,” he says. “Far too often 3D printers are just used as an iterative tool, where people print their designs at a small scale to help visualize their design and make improvements.”

ADVERTISEMENT

His goal was to produce a full-size body kit with the aid of 3D-printed parts. “In a perfect world, the end goal would be to take these parts off the printer and use them on the car,” he explains. Currently, that’s not really practical. The PLA and PETG plastic parts he’s able to print aren’t really suitable for use in the rough-and-tumble automotive environment. Vibration and weather take their toll, and the 3D-printed parts wouldn’t hold up too well in the long term.

437258571 352983221099081 5694303932597595361 N

Screenshot 2024 05 02 At 7.29.49 am
Brent’s design drew inspiration from the Live To Offend kit for the FD Mazda RX7. Credit: LTO

Instead, he’ll rely on a standby of the custom auto community. “For the time being the plan is to use these printed parts as a mold to fiberglass over to create the final functional part,” he says. This method allows him to still gain the benefits of 3D printing, while still having a body kit tough enough to last. His Creality K1 Max printer was able to turn his digital designs into real parts—he’s just going to fiberglass over them as needed.

Brent picked the Creality unit for its large build volume and fast print speed. He estimates he’s gone through 30 spools of filament just to print the pieces for the kit.

F41a83cf9d3d456e4b1a1d8f3697537d
Brent relies on a Creality K1 Max FDM printer, which creates parts by depositing layers of molten plastic. Credit: Creality
437348108 1366579081402107 3449691683776295086 N
The printer has a decent but limited build volume, so parts of the kit are printed in several pieces and assembled.

437204394 462678229450182 7191159433311847426 N  437370611 7637418159638221 4364532761037290987 N

ADVERTISEMENT
440913324 958815222893427 552914732631770963 N
Note the round magnets—these are used to hold the pieces of the kit to the Miata’s body during test fitting.

Perhaps the coolest thing is that this sick widebody is Brent’s own creation. “I did design the kit myself, and damn was that a brutal process,” he says. “I’m a serial procrastinator so I found myself in a position where I had 2 weeks to get this thing designed to finish in time to print everything.” His inspiration came from a multitude of Google searches for rad widebody kits, and those for Miatas in particular. “I didn’t really find anything I loved after hours of scrolling, but I’m sure subconsciously there are some elements of my design from those searches,” he says. “The main design inspiration is this kit was the LTO kit made for the FD RX7.” Taking that as a guiding light, he applied his own personal flair as he worked to create something that would fit the Miata body.

435209009 735141935165879 5650081863260777219 N

434676039 2761487550675811 7746907257503984572 N (1)

434677423 2761487560675810 205937286357122194 N (1)
Early test fits of the bodykit.

Brent says he picked up his 3D design skills in Fusion 360 and Blender from plenty of online study at. “I want to give the most massive shout out to Matt Perez,” he says. “He is the only reason I was able to do this, so please do go give his channel a follow and check out his website.” Perez is the man behind the Learn Everything About Design channel on YouTube. He regularly posts videos on designing parts for cars and motorbikes.

He’s eager to point out, though, that one piece of this kit is not his own work—and that’s the fastback hardtop. “All credit for the design goes to Ethan at Hutchins Racing,” he says. “The hardtop is beautiful and he put an incredible amount of work in.” For anyone wanting the same roof, the parts are available for sale on Etsy.

ADVERTISEMENT

Il 1588xn.4917514411 94x6

Il 1588xn.4869249476 Trtu
The Hutchins Racing fastback hardtop can be 3D-printed in multiple small pieces, and then assembled into a greater whole. It’s intended to be used as a base to create a carbon-fiber or fiberglass hardtop for use on the vehicle.

The 3D printer saves a lot of hard work to produce fiberglass molds, but it’s not a magic wand. There are still plenty of challenges in building a kit this way. Brent notes that having an accurate mesh model of the base car is key. Otherwise, whatever you design won’t fit properly. “If your mesh isn’t borderline perfect you’re going to have fitment issues to work out when everything is printed and you mount it to the car,” he says.

The printing process also has its own limitations. Brent found his printer couldn’t produce super thin sections, for example, so he designed around this. Maintaining the printer is important, too, if you’re racking up tens or thousands of hours printing lots of large parts. “The print quality had a noticeable drop off towards the end of the process,” he explains. “The belts got loose, the chamber was dirty, things needed to be greased… The quality of your parts will directly reflect the state your printer is in, so make sure it’s in perfect working order!”

437167564 955406875876099 2569529732297662584 N

437590512 324037727099165 6520677178503345493 N

ADVERTISEMENT

437193634 444434261600462 7337116998222237842 N

While it’s a huge project, and a daunting one, Brent reckons just about anyone could tackle it with the right attitude. “I designed the kit in Fusion360, in which I had absolutely zero experience when starting this design,” he says. Regardless, hitting YouTube and just having a go was enough to make it all happen. “If you’re thinking, it’s gonna take forever for me to learn everything and make it, you’re correct,” he says. “But the time will pass either way—so why not make something rad?”

Right now, Brent has the parts fitted up to the car, and they look awesome. Next up, he has to fiberglass and paint. “I’m currently trying to finish everything up to graduate and move, but this summer I’ll be fiberglassing and painting if everything goes to plan,” he says. We can’t wait to see the finished results.

Image credits: Brent Foster except where stated

Share on facebook
Facebook
Share on whatsapp
WhatsApp
Share on twitter
Twitter
Share on linkedin
LinkedIn
Share on reddit
Reddit
Subscribe
Notify of
41 Comments
Inline Feedbacks
View all comments
thejewosh
thejewosh
7 months ago

He should’ve gone with an X1C and just printed straight to carbon.

Sklooner
Sklooner
7 months ago

I would think that using a CNC to machine foam for a mold would be better cheaper and easier

Eslader
Eslader
7 months ago
Reply to  Sklooner

CNCs are pretty huge, even the consumer-grade ones they sell at Rockler. I had to build a ridiculous 5′ tall bench to store mine because it takes up so much room that it has to straddle the cabinet saw when not in use or there wouldn’t be room to park the car in the garage. And that’s just for a 3-axis with a relatively small work area. To make car part molds you’d be better off with a 6-axis, and now you’re getting into “forget parking any cars in the garage, upgrade the building’s power, and BTW you could have bought a Porsche for less” territory.

ADDvanced
ADDvanced
7 months ago

Aight, maybe Adrian will chime in, but I’ve been an industrial designer for almost 20 years and been using 3d printers for like the same time period. Point is, they aren’t new, and I’ve used them a bunch.

So this is… definitely a way to do it. BUT, fiberglassing OVER the 3d printed parts is going to be a huge challenge; you need to lay out the resin and glass very evenly and thinly, and when it comes time to sand things to prep for paint, it’s going to be really easy to cut too deep and expose some of the cloth. So, to get around that, you need to spray with a heavy build primer, and since he’s using fiberglass (polyester resin most likely), he’ll want to use a compatible primer, which would be something like Featherfill G2 (polyester). This will allow you to add enough thickness that you can actually sand it properly… which.. is still going to be a challenge.

The parts look pretty good right now because they’re matte/not reflective, but I have a feeling that once he starts sanding, he’ll realize some of these squares are not glued up perfectly; the surfaces aren’t tangent or G3 to each other, they might have a slight crease or valley.

Then, at the end of the day, you have a fairly heavy 3d printed part, with a bunch of resin, fiberglass cloth, and TONS of heavy build primer… and… yeah.

I think he’d be better off fiberglass from the BACK, to strengthen the panels, then secure them to the car temporarily, and support them with spray foam or something, and then do build primer and get them perfect.

THEN, make molds. Yeah it’ll be more of a PITA, but once you have the molds made, you can pop out replacement parts if you damage one, they’ll be way lighter, stronger, more durable, and you could evne make them out of CF or Kevlar if you want. AND you could sell the kit!

Brent Foster
Brent Foster
7 months ago
Reply to  ADDvanced

That’s actually a damn good idea! My game plan was going to be to body filler over the panels, get them flat and smooth. Fiberglass the front side, fiberglass the back side, then join them at the seams. My research is minimal and experience is 0 but that’s what I’ve seen done before to avoid having the heavy 3D printed part inside of the fiberglass. What do you think?

Black Peter
Black Peter
7 months ago
Reply to  Brent Foster

So the 3D printed parts are just the Buck? Because that’s what I assumed..
Once you glass the parts the PLA (or whatever) isn’t staying with the car..

ADDvanced
ADDvanced
7 months ago
Reply to  Brent Foster

I think bodyfiller is the wrong product; use Featherfill G2 instead. It sprays so thick you can bury a penny if you want to; it is quite literally sprayable body filler. The bondo filler stuff is for dents; you want full surface coverage, and if you sand into the printed parts themselves the paint will look weird. Feel free to email me if you want to talk more about it; just search ‘addvanced’ and there’s a contact page on my youtube channel.

Holvey
Holvey
7 months ago
Reply to  ADDvanced

This is the way we do it for custom motorcycle builds at my work. Obviously scale is the issue, we can typically build a tail section in 2-4 parts as opposed to a 15 piece fender, but we usually glass up the inside of a fairing or tail section and high build primer the PLA on the outside. We also do more graphic wraps instead of gloss paint and we aren’t making molds to sell. The PLA holds up fine with the Glass backing, and the few heat problems we’ve had have been from trapped heat from headers and not from the elements.

I think it really comes down to the end use expectation. It would be fine for a show car, or a garaged track day car, but its a lot of work to get messed up when someone on their phone clips it in a parking lot.

Adrian Clarke
Adrian Clarke
7 months ago
Reply to  ADDvanced

You are correct, this is A way to do but the challenges Brent is facing are WHY it’s not done this way. We’ve spoken about design software before I think, but worth having the conversation again for newcomers – Blender and Fusion are not the tools for getting good bodywork surfaces. That patch layout shown doesn’t look great and I can see that kit is going to take A LOT of handing finishing and flatting as you say. Ok for a fun/hobby build, but probably too much for a pro-build (although I’m outside of my knowledge zone on how good pro-builds actually are, quality wise).

Last edited 7 months ago by Adrian Clarke
Millermatic
Millermatic
7 months ago

Ok… I’ve got to ask… What is all of this stuff likely to do to the driving dynamics?

Manuel Verissimo
Manuel Verissimo
7 months ago
Reply to  Millermatic

Drag. A fuckton of it.

Drew
Drew
7 months ago

Re: the topshot text, I’ve been slowly printing my own wide body via unhealthy foods and beverages. This one is a lot cooler.

Mark Jacob
Mark Jacob
7 months ago
Reply to  Drew

I’m in this picture, and I don’t like it.

Ranwhenparked
Ranwhenparked
7 months ago

Man, it’s a shame you can’t buy a ready made hardtop like that, I’d order one today and then run out and buy a Miata to put it on. Had the Miata been around during the golden age of fiberglass in the ’70s, you’d have been able to get something like that right out of the JC Whitney catalog

Turbotictac
Turbotictac
7 months ago
Reply to  Ranwhenparked

There are several companies producing fastbacks for the NA/NB

A. Barth
A. Barth
7 months ago

*looks at topshot*

I wondered what Snow ended up doing after the ‘Informer’ craziness died down.

Drew
Drew
7 months ago
Reply to  A. Barth

With how quickly the dude could sing, it seemed like he had a future as an auctioneer or a political commentator using the Gish gallop.

Andrew Daisuke
Andrew Daisuke
7 months ago
Reply to  A. Barth

came here IMMEDIATELY for this. it’s uncanny

Pupmeow
Pupmeow
7 months ago

This is really cool and the sexy hood pictures are *chef’s kiss*

DaChicken
DaChicken
7 months ago

Pretty slick. I love my 3d printer. I just need to get better with the CAD end of things. All of the stuff I make is very rudimentary and nowhere near as complex as actual body panels.

If PLA and PETG aren’t up to the task (not surprised), I wonder if he’s tried using PA (nylon). Not important if he’s going to use them for molds, which is probably a better plan for body panels, anyway.

My less-fancy Ender3 S1Pro does a pretty good job with PA-CF (nylon, carbon fiber filler). I assume a K1 Max would do better. I love the material. I did a life-size model of a Tesla small drive unit for a friend to do mock-ups in his EV conversion and it is quite sturdy.

ADDvanced
ADDvanced
7 months ago
Reply to  DaChicken

You need to learn how to do surface modeling, not parametric.

Toecutter
Toecutter
7 months ago

I have a friend with a 3D printer that can do 1 m^3. We’ve thought of doing something similar, except instead of this widebody trend, we’ve talked about making aero bodies.

What we don’t have is CFD software or a wind tunnel. While I know some things about aerodynamics from what I’ve read out of many SAE papers at my university’s library as well as a number of books, I’m not an aerodynamicist.

Manuel Verissimo
Manuel Verissimo
7 months ago
Reply to  Toecutter

Reminder: I could get you some CFD analysis.

Toecutter
Toecutter
7 months ago

I need to settle on a final design for my electric velomobile. When I do, I’ll have the drawing scanned and send it to you to model.

Buzz
Buzz
7 months ago

This might be one of those “everything starts to look like a nail when your only tool is a hammer” type situations, but I feel like a CNC would be better suited for this application.

It’s really cool though, regardless of how it was made.

StillNotATony
StillNotATony
7 months ago
Reply to  Buzz

When your only tool is a hammer, every problem looks like a nail.

When your only tool is a 3D printer, every problem looks like whatever you want it to!

Millermatic
Millermatic
7 months ago
Reply to  StillNotATony

Don’t knock hammers. They can do a _lot_ more than pound nails. This kit could be made out of metal with the right hammer!

(But my resin printer is pretty amazing…)

Hangover Grenade
Hangover Grenade
7 months ago
Reply to  Buzz

I think this is actually a great use of 3D printing since printing is an additive process. If he were to use CNC, you’d have to start with big block of foam and cut away. And I think you’d need a 5-axis mill for this sort of cutting, and those aren’t cheap.

Alexk98
Alexk98
7 months ago

And with the sorts of curvature we’re looking at, even if broken into sections, this is multiple tooling changes, which from a setup and programming perspective is very expensive, especially on a nicer 4 or 5 axis mill.

Alexk98
Alexk98
7 months ago
Reply to  Buzz

CNC is so unbelievably expensive for things like this, even if only being machined from plastic. For prototyping and one-off stuff 3D printing is infinitely more practical. CNC really only comes into play when the product run is large, must be extremely precise, and cannot be replicated other ways. Having dealt with that manufacturing tradeoff at work as an engineer, trust me that attempting to CNC this body kit even from plastic would be easily in excess of 15k, if not more, while 3D printing is cost of a printer, filament and your own time.

Buzz
Buzz
7 months ago
Reply to  Alexk98

Home hobbyist CNC carving machines are a few grand. An expensive resin printer is probably in the same cost ballpark while having a smaller build platform. SuperfastMatt is CNCing body molds with the equivalent of hardware store insulation and glue sticks.

I’m not saying that 3d printing is the wrong way, but a CNC seems like it would be much faster and the same amount of complexity.

https://youtu.be/rgooQLz1ofM?si=yP2YmKUyovPr6zex

Last edited 7 months ago by Buzz
Alexk98
Alexk98
7 months ago
Reply to  Buzz

That’s a fair point, and I am a huge fan of his channel, but that is still several thousand dollars versus a few hundred for a 3D printer, which have become awfully competent in recent years. Given Brent is a student, money and especially space are at a premium, being able to adhere together multiple printed parts is far more cost effective than milling and then molding from a foam base, even if that part would eventually be far stronger.

We have a larger format resin printer at work and its successor has already halved pricing, but there are also significant differences and tradeoffs versus a more affordable mill.

Hangover Grenade
Hangover Grenade
7 months ago
Reply to  Alexk98

I looked it up and the printer he used is $750. I’ve seen some CNC machines for under $1000, but I’m sure they aren’t smart enough to do complex 3D milling, just simple cut outs.

Buzz
Buzz
7 months ago
Reply to  Alexk98

I’m assuming his school has an industrial design program and he has access to a CNC there or at a library or makerspace of some kind. I’ll stop talking about CNCs now, since it doesn’t really matter as the work has already been printed.

Alexk98
Alexk98
7 months ago
Reply to  Buzz

I actually attended the same school (I could tell immediately from the pictures) and I’m not aware of any available CNC machines generally available to students, particularly in Architecture. Some of the engineering labs have basic mills and lathes, but they all have restricted access. I know there are a large amount of resources for things like 3D printing, and other makers type spaces, but they’re always in short supply. Plus getting large things to and from campus is needlessly hard at our school, and there are a lot of guys with heavy Miata modify experience in the area that helped out, so that probably played into decision making as well.

All that said, I think schools should absolutely have more CNC availability and training to students, There was one lab I did which had some very basic 3-axis programming, but the machine was very locked down, and not the most helpful in teaching me what to do from start to finish.

Brent Foster
Brent Foster
7 months ago
Reply to  Alexk98

Heyo! The architecture school/industrial design now has CNC machines, but they are very basic sadly. Figured I would work with a tool that I already have!

Eslader
Eslader
7 months ago
Reply to  Buzz

I have a home hobbyist CNC carving machine. It’s great for certain applications but I have no idea how I’d manage to make a complicated-curve car body part with it. It could do the one side just fine, but then how’s it going to do the off-vertical stuff?

These things are basically drill presses on a movable gantry. They point straight down. If your drill press couldn’t position the bit where it needs to be in order to cut out part of the shape, the home hobbyist CNC can’t cut out that part of the shape. Yes, you could move the part after the first-side pass but now you run into potential issues with positioning errors.

You can step those machines up to a 4th axis by adding a rotating spindle to hold the part. Now your point-down bit can carve out shapes you’d see come off of a lathe, but it’s limited by the spindle’s clearance from the bed. If you try rotating a big part like a fender, it can only rotate so far before it whacks the bed, so those are mainly useful for making small things – think valve bodies or chess pieces.

Once you start getting beyond those capabilities and move to a CNC that does all the regular home hobbyist moves plus can gimbal the head around so the cutting bit can point somewhere other than straight down and do so with larger parts, you start getting into serious money territory (and keep in mind the hobbyist machines are usually in the $5k neighborhood for a halfway decent one, plus $800 and up for the 4th axis).

You also start getting into more intensive power requirements. Get a machine that can position the bit in any orientation you need to make any part you can think of, and in a useable size for car body kit work, and you’re looking at a 5 or 6-axis, and now you usually need 3-phase 240 which would involve significant electrician expenses if you could even get the city to allow you to do it in a house.

You’re also increasing the size of the machine significantly if you want it to make your car body parts. It’s no longer a 3×5 platform with a gantry on top of it. Now it’s about the size of a van. There are smaller 5-axis machines, but unless all you want to do is make very small parts or maybe tabletop gaming pieces, they won’t do you much good.

Buzz
Buzz
7 months ago
Reply to  Eslader

You can CNC materials on two sides with indexing holes and dowels. That ensures that the part stays exactly where it is meant to when you flip it.

Complex 3d parts need supports when printed, it’s not like they come out of the printer ready to install. This is no more complicated than that.

Eslader
Eslader
7 months ago
Reply to  Buzz

True, but a body kit’s likely to need 3 or maybe even 4 sides, which makes things quite a bit more complicated. Looking at that back 1/4, with those really deep horizontal parts top and bottom, you’d probably need to approach that at an angle to get the spindle where it needed to be, which you could do with the dowels and holes but you’d also need to set up support for it and all of a sudden you’re getting really complicated and hoping your Z gantry is tall enough to handle all of it (I know neither of mine would be).

Rad Barchetta
Rad Barchetta
7 months ago
Reply to  Buzz

Having been messing around with such a hammer making car parts for a few years now, I can assure you, there are a lot more nails out there than you might think. Even if it’s just for prototyping.

41
0
Would love your thoughts, please comment.x
()
x