From Open Source Ecology
(Redirected from The Embassy Elf)
Jump to: navigation, search


  • A vehicle design utilising Fiber Composite Reinforced Pipes. The design is focussed on efficiency.
  • The acronyms stand for Ultra-Efficient Libre Vehicles for Everyone - Myth Busters Car
  • Is a prototype for further iterations of vehicles using similar methods for various other vehicle types, as well as structures such as furniture, sheds and semi-permanent buildings.

The Embassy Elf (MBC UELVE) Introduction

The MBC (Myth Buster Car) is intended as a lightweight proof of concept for further development of additional larger (and smaller) 3 and 4 wheeled hybrid and tribrid electric vehicles such as solar vehicles, hydrogen-intake-injected and more. It's also intended to be eco-conscious, ultra-efficient, ludicrously high acceleration, low cost and eye-catching (hence the cross between a 2-seat sports car and a "Mad Max" off-road look), and intended for the average person to 3D print (or buy a pre-printed kit) and assemble in under five days.

Once completed the tools (including a specialist dedicated kevlar-crosshatching Fibre_Reinforced_Pipe_3D_Printer) and design techniques will be in place to expand the concept to other vehicles (as well as furniture and buildings). Other vehicles will not have the same "wow" factor however by going initially for a vehicle that is likely to be stressed to its limit (off-road, racing), design and materials flaws will be quickly found and fixed, long before more "ordinary" vehicles begin production.

Acronym decoding:

  • UELVEs: Ultra-Efficient Libre Vehicles for Everyone.
  • MBC: the Embassy, or the "MythBusters Car" (get it?), in honour of their discovery that a golf-ball texture on a car bodypanels significantly improves fuel economy. https://www.youtube.com/watch?v=VUiGhyHC-1A

Frame Concept

Uelves mbc.jpg

Blue pipes are 1m measure sticks. Note that the body takes whatever form we would like.

Minumum Viable Product

HintLightbulb.png Hint: Holy shit.

  • 4 wheels using mountain bike 19in rims and 2.75/19 motorcycle tires (26in outer diameter)
  • 100kg weight (full, complete vehicle)
  • 2 occupants
  • Twin 3-5kW Axial Flux BLDC chain-driven motors (no gearbox needed)
  • 1.1 x 1.5 x 3.4 metre dimensions with an "airbox" down the centre
  • Only a 1.1 m^2 front surface area (due to the "air tunnel")
  • 0-60mph in under four seconds (Sur-Ron "3" kW Motors can be pushed to 13kW for short durations, and there's two of them)
  • over 200mpg at 55mph due to low weight, low rolling resistance, reduced surface area and low drag coefficient.
  • capable of driving up a 50% gradient at around 25mph with 2 occupants.

Final Build Specifications (Post Build Tests)


  • While world records are held for under 3 seconds, under 4 seconds is close to world records. [1]. We can emphasize here that for "acceleration per horsepower", ours here is probably a world record. Luke - have you done the numbers on this one?
  • with the simulator i wrote a few years ago, yes. link at bottom of page. it's the reduced weight, reduced drag coefficient and the SurRon axial flux motors which can do 13KW for short bursts that do it. the world record is an MIT team EV with 4 hub motors. 0 to 60 in 1.9 seconds. in the slo mo the walls of the tires actually buckle under the acceleration.

Correction: that MIT team has been beaten by AMZ, who officially managed 0 to 60 (100km/h) in 1.513 seconds


Basic Design Concept

CAD Source Code


estimated, todo steering and suspension, estimated total under USD 4000 excluding tools and batteries.

  • 3D printed 15mm x 2mm kevlar/hemp continuous reinforced pipes (250 metres)
  • 3D printed "nodes" (120)
  • Kevlar twine (see gaboats)
  • Dacron (see gaboats)
  • QTY 2 Shimano Saint M820 mountain bike hydraulic brake sets (2 front, 2 rear)
  • QTY 4 19in diameter downhill mountain bike wheels
  • QTY 4 2.75/19in Shinko "Goldenboy" tires
  • QTY 4 Fox DHX2 rear suspension sets with orange spring
  • QTY 2 Sur-Ron 3kW BLDC Axial Flux Motors
  • QTY 2 Sur-Ron rear swingarm (or structurally strong 3D printed equivalent, TBD)
  • QTY 2 20kW 3 phase BLDC Motor Controllers (Vedder VESC)
  • 6 guage wire (appx 5 metres)
  • QTY 15 (appx) Open Source Battery Packs per kWh
  • Bus Bars connecting packs together
  • Bearings (lots)


Nozzle Choice for Printing With Abrasive Composites

Pipe Materials




Core Plastic Options

Kevlar Reinforced CFF

NylonX + Nylon K

found something that i think may work (worth investigating). NylonX and its newer cousin, NylonK.


the case studies are awesome. a skateboard bushing, and a mountain bike pedal.

these look like they have the strength to make both the ends of the pipes *and* the nodes *and* are worth exploring to make gears and full rack and pinion steering, and even the suspension arms.

now the only way i would be happy to make rack and pinion steering 3D printed is if they were massively oversized. the pinion gear a whopping 3in in diameter or 2in diameter but also a massive 1.5in depth, making the rack 1.5 to 2in deep as well.

the force of an emergency turn, or at speed trying to hold the steering wheel against friction, this is not something you mess with.

so we see how that goes?

but, i really like the idea of fibre reinforced *nylon* (not PLA) because it is already strong, and higher temperature.

Material Strength Test

the logical reasoning behind this test is to find out if, with fibre-reinforced plastic, the layer adhesion affects strength of the part under load when bolts are screwed directly into it.

  1. two M10 1.5mm pitch bolts (or close: 3/8 with 1/16in pitch)
  2. a 30 x 30 x 30mm cube made of fibre reinforced plastic with a smooth cylinder hole down the middle about 7 or 7.5mm diameter
  3. print that using 100% infill, fibre reinforced plastic, with a minimum 0.8mm nozzle, preferably 1.2mm, with the hole pointing UPRIGHT.
  4. screw the two M10 bolts in, using the thread to "tap" the holes, making them meet in the middle
  5. hang it from a bench by 1 bolt and keep putting weights on it until it breaks.

i need to know, how much weight it took. it might be a lot, before the cube breaks or the threads strip.

the next test is printing it where the hole is horizontal.

the "more advanced" version of that is for the hole to be at 45 degrees through the cube, i.e to test the effect of diagonal layers.

Testing Machine

  • User:Eric here again, it is a luxury, and not too necessary, but would lead to high quality measurements for things like advertising, and reputability
  • I have been conjuring up a Open Source Universal Material Science Destructive Testing Machine
  • it can basically measure all the various stresses. Compressive Strength and Tensile Strength are the main ones, but Shear Strength and Torsional Strength may be of use as well
  • It basically just a semi-fleshed out idea at this point but perhaps this could lead to it being developed a bit
  • Load cells could be used for high end, but a semi accurate hook scale (like a luggage scale) may work just as well
  • Then perhaps geared down servos?
  • Some Lab Grade Ones That I have seen use a pendulum that is lifted and dropped, this may be something to consider
  • Also some youtube channels (3D Printing ones mainly) do some of these tests every now and then so check those out if you haven't already
  • Also there are various Standardized Material Science Testing Forms of 3d prints and other objects (An S hook, a solid bar, your pipe, etc) so this could use those at least

Expected Issues

One key issue to watch out for is the block cracking as the M10 bolt is screwed in. Normally, a tap/die would be considered, here.

The problem is that when using 1.2mm nozzles, it is extremely unlikely to be able to 3D print a matching 1.5mm thread for the bolt to fit in, therefore tapping the hole is the next option, and seeing if the bolt itself will achieve that is logical to attempt.

One alternative ie to look for very coarse threaded M10 bolts, or even trapeoizoidal bar, or, if this really doesn't work, some alternative lock mechanism, or a much larger diameter hole and use epoxy glue. of course, epoxy glue you are never going to be able to remove the bolt.

MBC strength test.jpg


  • Simulation (playable game) using ODE (Open Dynamic Engine) which will be cool and also test real-world joint strength and drivability. See Hackrod. https://youtu.be/v__qoJsAs48 (note about Hackrod: everything about it is entirely proprietary software, based on cloud services. no source code is made available).

Notes from Julia log


  • When recycling fibre reinforced plastic, the fibres get shorter and shorter. (possible solution: add more back in?)
  • Carbon fibre is energy intensive to produce, is often produced using non renewable energy, and is sometimes processed, or coated with harmful substances. It is also rather hard to recycle ( see Enviromental Impact of Carbon Fiber (not going to use carbon fibre, use hemp fibre or wood fibre, see Da Ai Technology in taiwan) ( User:Eric here, have you considered semi-synthetics? You basically dissolve cellulose, undissolve it through a nozzle, and in turn get a natural fiber that has been processed for better properties (mainly consistency, i'm still trying to get hard facts on this (tensile strength etc) either way this project is going to be great so no pressure)
  • use detroit technique: 3 pieces, front middle and back. matches with what i suggested, given that big printer is aiming for 4x4x8 ft. roof and floor will probably need to be separate in this design (so, 4 pieces)

Questions and Comments

  1. what kind of fibre plastic? answer: don't know enough to say. ideally several materials should be tested and the results fully documented.
  2. what diameter hole? assuming using the bolt as self-tapping, an M10 bolt with 1.5mm pitch has around a 1.5 mm inset, therefore 10mmm minus 1.5x2 equals 7mm, so between 7 and 7.5mm seems sensible to try. bottom line: if it cracks on insertion, the hole was too small!
  3. how should the weights be hung? doesn't matter as long as side-twisting is avoided (we are not testing shearing or twisting, here, yet, just pulling). strong rope tied or clamped to both bolts, one end fixed to a table, the other has weights attached, it doesn't matter. doesn't have to be sophisticated, here.





Screenshots / Images

MBC doorframe geometry.jpg

See Also

Useful Links