Exocet XP-3

Hey guys, I figured I'd make a build thread for my "Exocet done right" coming up for the UTCC. It's going to be sparse now due to the timeline, but I'll flesh it out eventually.

What the berkeley is an Exocet? It's what happens when you look at a Miata and say "I like it, but it's 1000 lbs heavier than it needs to be". More info here.

Goals for this Car:

Have some fun, and spread the gospel of Exo-Cars

Demonstrate that you can do the Miata -> Exocet conversion very quickly

Make the ultimate driver's car for <$20k

Build something responsive enough and fast enough to be a FToD machine at the autocross course.

This car is not built to be a road course car, but it sure will be fun in an HPDE environment. 0-60 should be in the low 4 second range. The expected top speed is rear-end limited (4.30) at 145 mph; I suspect we will become pretty familiar with the rev limiter down the straights of VIR. We could go faster with a different ring and pinion, so don't tell anyone involved with the build that my daily driver has a 3.9 Torsen...

If any of you guys in Atlanta want to get involved, check out the MX5Atlanta thread on the build HERE. It's a crazy timeline, we'll take all the help we can get.

June 11 Picked up the donor. Fully loaded '99 with a Torsen and 140k-something miles. Owner spun it into a guardrail, did some bodywork damage. Didn't want to fix it. Clean title, $900. You're jelly, I know.

June 19, 2013 After about a day's worth of Miata stripping, we pulled the unibody from the PPF. Flyin' Miata claims to be able to strip one of these things in 90 minutes! The only difficult part is noodling the harness out of the car. I like to call Miatas "Lego Cars".

http://www.youtube.com/watch?v=YHdNfuqr7ho

Forgive the dirty shop, we were crunching, and we have NO space (our new building has been delayed)

June 20-30 Prepping for the first 10-car production run, manufacturing stuff.

July 1, 2013 The car is a computer model. UTCC is in 18 days. We got this. The Exocet went through a healthy refresh after our first three prototypes. Due to manufacturing reasons, every tube was going to be re-programmed through our laser partner's CAM software. Ever build something and then at the end have a list of "if I did this again"s? Well, I did, and so I fixed all of them. The changes are subtle, but they make a big difference for our assembly/welding. Biggest feature is laser-etched part numbers on all tubes.

This is Flyin' Miatas planned Exocet. Ours will look the same, but either bright blue or green.

July 2, 2013 We started production of tubes at 6pm. We have to work around our laser partner's schedule, this is a fun side project for them. The round tubes (except for the 1.75x0.095 DOM) came in a huge cardboard tube.

After just three hours of FIRIN MAH LAZER (alright, not actually my multi-million dollar machine):

July 3, 2013 And just like that, 10 cars worth of steel is cut out. These guys don't mess around. We've got to wait until Monday to bend the tubes and sheet metal parts, but we did some test pieces and know a lot from the last time around. We also have a few more 1.75"x0.095" tubes to cut, they'll get those Friday. Our new custom-tooled 6" CLR bending die is a beast.

There's another box like this, but filled with round tubes.

It's a series of tubes.

Holy part yield Batman! That's one of the two sheets of 4mm DOMEX 700 MC, basically 3x the strength of mild steel, even as-welded. It's a microstructured alloy, and there are no downsides to this material other than cost. As far as cost, let's just say there are benefits to working with a company that does a lot of military contract work (am I allowed to say that?).

The other sheet of DOMEX being shot at with frikkin laser beams (I'll never get tired of this laser business).

Here's a crappy-quality video of the front subframe mounts being cut. Yes, they're made from tube, yes they have socket cutouts. Awesome, I know.

July 4, 2013 Go 'Murca! I'm celebrating by starting the bodywork for this thing and Flyin' Miata's car. Both sets are black gelcoat, which is difficult to get right since all you see are imperfections and scratches. The last set came out pretty good, I have high hopes.

Warren v wrote:

I don't think another KTM X-bow is going to sell after this thing comes out! I wasn't that interested in Sevenesque cars...until now. And I have a buddy with a couple of wrecked Miatas, hmm...

GameboyRMH wrote: I don't think another KTM X-bow is going to sell after this thing comes out! I wasn't that interested in Sevenesque cars...until now. And I have a buddy with a couple of wrecked Miatas, hmm...

That thought process is EXACTLY how I got into this mess in the first place.

Got the subframes and suspension bits from the powdercoaters. Nice solid glossy black.

Also, for those that missed my edit on the first post:

Atlanta-based GRM'ers, we're opening up the assembly of this thing to you guys. Check out this post on MX5Atlanta for details:

http://www.mx5atlanta.com/forums/showthread.php?21200-Wanna-help-me-build-a-UTCC-turbo-Exocet

I can't be the only one sorta hoping to wreck his miata now, can I?

SnowMongoose wrote: I can't be the only one sorta hoping to wreck his miata now, can I?

Heh, here's my daily. Got backed into by a truck while it was parked overnight.

Stop, just stop. This, my wrecked 95 miata, and the LS1 on an engine stand in my garage are all conspiring against me...

Warren,

Are you laser etching "alignment marks" on these tubes to make them easier to tack in place before welding?

SnowMongoose wrote: I can't be the only one sorta hoping to wreck his miata now, can I?

Nope... I'm hoping to get my turbo'ed '94 running soon, just so I can hit something solid with it.

Then I can use the parts to build something much cooler!!!

OHSCrifle wrote: Warren, Are you laser etching "alignment marks" on these tubes to make them easier to tack in place before welding?

Better than that, actually. We have physical tabs and slots at just about every intersection.

The actual tube walls are cut perpendicular to the tube face for maximum weld strength. That's programmed in the machining package, it's much easier to model the intersection in CAD with sharp tube edges.

The whole thing goes together like a jigsaw puzzle on our jig table. We use the Miata subframe mount points as clamp points, so you know they won't drift during welding. The first three prototypes went together really well, the system worked better than expected. It took half a year of CAD and about 2200 digital sketches, but I'm really proud of the result.

The main swept tubes in this picture are NOT tacked in, just held in place by the tabs. You rotate the top one and progressively insert the triangulation tubes from rear to front:

My favorite feature is the floor. The front, inside pieces, and rear are all one self-aligning hand-folded tube.

Sorry if it seems like I'm bragging or paddling my canoe, but this is kinda my baby.

Been enjoying all your post on this subject as you come along with the project. As someone who hires and enables creative individuals, I recognize a person who is operating in their niche in full happy mode.

Warren v wrote: Sorry if it seems like I'm bragging or paddling my canoe, but this is kinda my baby.

Dude, brag away. It is awesome. The tabs & slots are an even better solution than registry marks. Color me impressed. The curve geometry on the 'corners' of your bent square tube reminds me of a cabinet making wood router bit profile (more surface area for glue than a simple butt joint) and I'm sure thats no coincidence.

Regarding the razor edge tube cuts: when you cut such a thin point on a tube, won't the welding burn through the point, and potentially result in a weak joint? I am an architect, not a welder or machinist or mechanical engineer - but it seems like the sharp edge needs to be chamfered back a little so the weld will fillet between full thickness tube on both sides of the joint.

It's killing me that I will be out of town during your build. I have nearly zero "mechanic" experience, just enthusiasm.. and I live in Roswell. I hope to see these running all around me soon.

In reply to OHSCrifle:

You're totally right about the razor edges. They don't exist on the parts, only the CAD model. Perpindicular cuts are a checkbox in the CAM package (software that tells the tube cutter what to do). To make perpendicular cuts in the CAD package, I'd need to convince SolidWorks that the tube is actually a rolled up piece of sheet metal, unroll it, do the perpendicular cuts, and then roll it back up, fill up the slot, then export that tube. I prefer the checkbox method. That picture is in the CAD package.

"The actual tube walls are cut perpendicular to the tube face for maximum weld strength. That's programmed in the machining package, it's much easier to model the intersection in CAD with sharp tube edges."

The little features on the bent square tube are indeed there for more weld area. They also help align the tube when hand-bending to make sure it ends up perfectly straight. I have seen birdmouthing and bending on square tube before, but the "wiggles" were an idea I had during the design of the car. I was really happy to nail the k-value on the first shot (0.3):

Sounds good. I was staring at the images and not grasping all the words. Did you seriously not copy the "wiggle" design?

OHSCrifle wrote: Sounds good. I was staring at the images and not grasping all the words. Did you seriously not copy the "wiggle" design?

It's an evolution of this idea that I copied and applied back when I was designing a lot of sheet metal parts a few years back:

The search term is "metal origami". That particular method with the straight lines and curves looks cool and is useful on large parts, but it has a certain amount of inherent inaccuracy that sucks on intricate parts. Basically the planes of each face don't quite intersect where you want them to, as each little tab decides to bend in a different place. I would design side tabs with chamfered angles on the edges that end up touching after bending, and the wiggles were a version of that idea that would work on a lasercut tube, keeping in mind the limitations of the laser.

Warren v wrote: In reply to OHSCrifle:

Brilliant.

At the risk of muddying this thread with general questions about the kit, how much welding is required for the kit as purchased by a home builder? I'm pretty interested in this idea but have exactly zero welding experience.

dyintorace wrote: At the risk of muddying this thread with general questions about the kit, how much welding is required for the kit as purchased by a home builder? I'm pretty interested in this idea but have exactly zero welding experience.

None whatsoever, unless you're doing a crazy swap or something.

Gotcha. Thanks. I got nervous seeing the comments made above regarding welding. I guess you guys do that before the kit is delivered to the customer?

tuna55 wrote:

Warren v wrote: In reply to OHSCrifle:

Brilliant.

Indeeeeeeeeeeeeeeeeeeeeeeeeeeeeeed!

dyintorace wrote: Gotcha. Thanks. I got nervous seeing the comments made above regarding welding. I guess you guys do that before the kit is delivered to the customer?

He's documenting the process of turning a pile of tubes into an Exocet kit at this point, and bragging about his clever techniques and big industrial lasers that make it easy for Exomotive to build a kit for such a low price with repeatable quality. As he should

Warren, might want to draw a line when the time comes and say "if you order an Exocet kit, you will start here"

dyintorace wrote: Gotcha. Thanks. I got nervous seeing the comments made above regarding welding. I guess you guys do that before the kit is delivered to the customer?

Yup. You get a welded frame, gelcoated bodywork, hardware, trim, aluminum panels pre-bent with pilot holes, and a few brackets and knick-knacks. There are options such as having us powdercoat the frame in any color you'd like. The bodywork whatever RAL color you want on the Sport and Race. The Base is any color you want, so long as it's black...or white. The idea is that we supply you with everything you need to get a working car that doesn't come on the Miata donor. One of the first orders of business when we get our new shop is to set up a white product photo booth that shows everything you get in the kit, and everything you need from your donor.

The slots, tabs, etch marks, bend documentation, etc, are things that the customer won't get to see on their finished product. The benefit to the customer is quality and price. Other tube frame cars are usually assembled with a lot of fabrication and fitment required. Our goal is to be able to assemble and weld the car without ever touching a grinder (except for the specifically ground-flat parts). We were shockingly close on the prototypes, the production run should be even better. That labor reduction, simplification, and extensive reuse of donor components is how we can sell such a good chassis and kit for a fraction of the other single-donor kits out there.

Now that it's over with, I want to be clear that this is a build thread, not an advertising thread, so I'm actively trying to avoid advertising the kit as a product.

Je ne suis pas un canoe.

Fueled by Caffeine wrote:

tuna55 wrote:

Warren v wrote: In reply to OHSCrifle:

Brilliant.

Indeeeeeeeeeeeeeeeeeeeeeeeeeeeeeed!

Very cool concpet, and practice.

What kind of welding are you doing? I had thought that one needed some space to properly weld that together with penetration all the way through. But that's more of a butt together without a V.

I would imagine that TIG will be slow, and not needed, where as MIG will be quick, and strong enough (since ships are welded togther quite often with MIG welders).

In reply to alfadriver:

A good Miller MIG with a skilled operator. With penetration confirmed, the difference in this application is mostly aesthetic. Heat management is also easier with a TIG, but we are pulling these cars off the table and they are staying very straight, so it's not a problem. The Exocet is incredibly stiff and light because it uses a lot of thin tubes loaded axially rather than a few thick tubes with a lot of loadpaths in bending. While this is ideal from a finished-product perspective, it is labor intensive, which is why you don't see it very often. The amount of intersections would mean a week's worth of work if we TIG'd it. MIG is more like two full man-days, including chassis assembly.

That tube is 0.065. With our Millermatic, full penetration butt welds are not difficult until you get up to 3/16" and thicker. We set up the machine and wrote settings for the different tube thicknesses after checking penetration. Having a cold saw is nice for that stuff. Either way, the floor is not completely dependent on those particular butt welds; nearly all of the shear stiffness is resolved through the riveted aluminum floors or welded-in steel floors. Also, those intersections are welded, and then tubes are intersected on those points. There are a lot of loadpaths in that area, which is important to protect the driver from any large debris that may impact the front of the floor. That's one of the things that really scares me in the Atom, you're hanging in a brittle carbon tub with no impact protection.

In reply to Warren v:

Cool- I just thought it would be hard to get good penetration when it's that cleanly butted together. Just thought a V would make it easier. It's a very cool concept- one of the many things that allows assembly/welding to fitting/assembly/adjusting/assmebly/welding. As far as I can see- you are spending one expensive step to save quite a few more expensive steps. Easy trade off if you have the right tools to do what you are doing.