This is the third installment of the Aerolithe project, the recreation of one of the great lost Bugatti's from the 1930s.

In beginning a project such as the Aerolithe's recreation requires a lot of planning. The same can be said for any comprehensive restoration of a piece of machinery as complex as a car, but in restoring the Aerolithe we have a lot less information. A lot of assumptions must be made as well, although those assumptions have to be firmly grounded by a comprehensive knowledge of the cars of the thirties and the procedures and the often eccentric engineering of the Bugatti Company in particular.

Some of the project is quite easy. The chassis being used is mostly complete and the engine and driveline original and easily rebuilt.

The body of the car however is a different kettle of fish and to make life even more difficult (as if only a handful of old photographs as reference for the build wasn't bad enough) Ettore Bugatti at the time of the 1935 Paris Auto show announced that the body of the Aerolithe was made from elektron, a magnesium alloy.

We decided very early into the build that we needed to research the use of magnesium first, to determine if the Bugatti craftsmen of the time could in fact use it to create a car body and next, to determine if we could, within reason, do the same.

Obtaining a sheet of magnesium was our first challenge. Many of our normal suppliers of raw materials had no idea where to get it or no interest in supplying such an exotic material. Once found we then had to wait a month or more for the first sheet to come in the door. It finally did arrive, an eight by four foot sheet of dull grey metal that looked just like a sheet of aluminum. With it came an invoice for over thirty five hundred dollars. That was a bit of a shocker, as a quick calculation determined that just the raw material for the Aerolithe body would cost in excess of thirty to thirty five thousand dollars and was definitely something that we had to take up with our client for whom we were building the car.

Because magnesium is difficult to weld, the designers used rivet flanges which were found on the outside of the cars fenders and centerline rather than hidden and some critics panned the design. I had often thought that this was a styling cue perhaps copied from the Bugatti Grand Prix cars, the Type 59s whose body shells were riveted together despite being made from aluminum.

Rivetting elektron made sense as it is hard to handle and could have been a plausible excuse for using flanges to join it. No matter, our research lent credibility to Ettore's proclamation about the car the world soon knew as the elektron Coupe. That firmed up our client's resolve to have it recreated in Magnesium.

So, now that we had a sheet of magnesium to work with we decided that the creation of an entire front fender outer shell would tell us almost all of the problems that we might face with the material. The fenders have large flats sections, rivet flanges, complex compound curves and rolled edges.

So first we discovered that magnesium does indeed burn and great care has to be taken with filings, dust and the ragged edges of the material. It is also difficult to extinguish. Next we found that it only wants to bend about fifteen degrees before it snaps like cheap plywood. It is also incredibly light. The whole sheet can be picked up between two fingers and lifted so imagine our surprise when we also discovered that it is almost impossible to dent it with a forming hammer and has a memory which makes it want to re-assume the shape that you are trying to make it give up.

One thing was for sure, it was so expensive that we could not just mess about trying to discover the methods that would best serve us.

Three weeks of careful handling later we had it, an Aerolithe front fender that could be picked up between a thumb and a finger and would resist being dented with a ball peen hammer.

We planned for about a square foot of waste from the first panel, not much more. In the next installment I'll tell you exactly how we went about it.