wood engineering

wood ? why wood for the main frame and body ? scroll down...

- because that's cool and a nice material to work with ! - a structure making the most of wood is safer, stronger and lighter than a structure making the worst of "rocket science" materials - plus the usual green outcome and benefits (our birch plywood comes from Finland and the timber is locally sourced) - it's a difficult challenge for us, full of pitfalls, something very exciting compare with the other materials ! not yet convinced ? a trivial example: - the De Havilland "mosquito" bomber was the fastest bomber ever made during the world war II, airframe all made of wood

- still nowadays many light aircrafts, even acrobatic ones, are using wood as a structural material (including the main wing spar/longeron) -if wood performs well for an aircraft cruising at 350mph, why not for just a slow rolling velomobile ? all the challenge is to make the most of wood to design the lightest velomobile and yet road worthy: - by using the "state of the art" methods of structural analysis and detail design of wood airframes (we wouldn't ignore 80 years of modern engineering experience and literature would we?) and about not getting rotten ? same as above, plus , I suppose there must be some legacy to learn from 60 years of wood/epoxy yachts building and use. at the end, there is no need really to re-invent the wheel....just a bit of engineering. and still not yet convinced ? wood doesn't sound as sexy as berylium-boron-enriched-ultra-performance-high-modulus-carbon-revolution-core-matrix, fair enough :-)
our first step was to carefully examine how light structures made of wood are built. A lot was to be learnt from the past: how engineers solved the problem of load transfert from a concentrated area to a wood frame. Typically in our design, how to transfer cycling torque, brake effort, seat moment, etc... to the frame without introducing stress risers, in a durable manner ?
a second step was to understand how to design a light frame, to predict its stiffness, strength and safety margin. where a "Do-It-Yourself" bicycle just need to be stiff enough to withstand the abuse of its user and builder, a mosquito velomobile cannot be designed on the same basis. Being stiff is not a suficient condition, we need to make sure that the frame won't brake under ultimate loading conditions. first we have used the composites material theory to predict the stiffness and strength of any laminates and beams made with plywood and timber. This has been compared with the experimental results provided by the specialised literature.
	then we have been building simple finite element models to simulate the behaviour of various beams configurations (I beams, box beams, girders; trusses), those have been compared with the theoritical results provided by sound analytical solving. the intention here was to calibrate the relative (in)accuracy of a finite element model; it depends on the software, the computer, but most importantly the level of ignorance of the designer. Here, we have tried to minimize the last point, although being a lifetime mission. Analytical calculations, experimental design guidelines and FEA simulations have been compared all together with our real world. The left photo shows a box beam under torsionnal load, the angle deformation is being checked against our theoritical models.
	When it comes to computer simulations, we have learnt the hard way that "push buttons black box" softwares coming with CAD packages are just not doing the job. we have spent time following a more robust approach. in parallel it is necessary to know all the possible load cases scenario a velomobile will live in a 20 year period of time. We want our velomobile to be an utilitarian vehicle, the load cases are much more severe than for a racing bicycle, especially due to the large spectrum of weather and road conditions over such a long period of time. We built our first finite element model of the entire mosquito and tested it with various boundary conditions and loads. When happy with a nice version of the frame, it was time to work out all the details. From the model we can extract the efforts and moments and use them as loads to design all the connections and sub-components.
what is the point of calculating if we are not able to build what we have designed ? sounds like a silly question, but what about plies orientation tolerance ?, unconsistant material properties ? unconsistant material and joint thicknesses, unconsistant geometry ?, etc... in one word: discrepancy between the virtual world and the real one.
	the solution to minimize the unknown ? only materials of consistent properties have to be used; only "square" and simple beams have to be built, max 1 curvature, no funky shapes, compound curved panels and beams are banned from our workshop; only "state of the art" assembling methods have to be used. Design practices approved by the relevant industries are too often ignored in the bicycle trade. Those for wood are well documented in the literature. You will never see a bolt through soft wood assembly in our mosquito, no recessed socket heads, no abrupt change of stiffness by the use of thick metal fittings, no dodgy assemblies impossible to inspect or repair, no stainless steel bolts or inserts in carbon fibre components, etc...
	for instance, any plywood meeting the G.L. requirements exhibits certified minimum mechanical properties.
	apart from being carefully selected, all timber and laminated wood are being tested to evaluate their strength 3 point bending test according to an old french "air" standard short beam test and 4 points bending test are part of the plan as well. Ash tree is used were high stresses and large deformations are expected. All the beams are designed to fail under compressive load, it's a safer failure mode compare with a failure under tensile load (instant break into 2 pieces with sharp splinters).