I have a good deal of automotive experience with production cars. I have little to none with specialty or one-off cars, although I did own a VW-based Devin roadster as my daily driver for a few years. Like the Locost, the Devin was a very basic car. It had a fiberglass shell bonded to a ladder frame, with VW front suspension welded on up front, and trailing arms locating the swing axles in back. Everything was molded into the fiberglass shell, including the trunk, seats, and dashboard. It had a bolt-on windshield from a boat, and no top. The electrical system was minimal and the car had two gauges. But it ran forever. I didn’t do a lot of work on the car because it didn’t have a lot of parts to break, and I couldn’t afford to do anything at the time anyway. But I did gain an appreciation for how simple a car can be.

I don’t think that building a car from a proven design is a terribly difficult task. I think the difficulty starts when you deviate from that design. Engineering any part of a car should go through several phases, like analysis, specification, design, and testing. No part should be installed on a Locost without at least some analysis and testing. For this reason, I think it’s always an excellent strategy to use a part that’s already been analyzed and tested by someone else. Preferably by many someone elses.

When Bill Devin set out to create a sports car from a VW Beetle, he stayed with the existing VW hardware everywhere he could. He bolted it all to a tried-and-true ladder frame, one that was probably beefy enough for a truck, but in any case was definitely going to work. He didn’t try anything fancy. His business was fiberglassing, and this is where he put all his design experience to work, on the sleek, Ferrari-like body. The only piece of mechanical engineering in the entire car was the rear trailing links, and those were the only parts that ever gave me any trouble.

Engineering a single part is not all that difficult. What starts to become difficult is having several parts that need to work together. The difficulty factor increases exponentially as the number of parts increases. Fortunately, a Locost can (and should) be much simpler than that. There are a lot of talented amateur engineers out there in the Locost community, but sometimes I think what they’re doing is not actual engineering, it’s more creative design. Fortunately, from a structural perspective at least, most people start from an existing, tested design, and tend to err on the side of safety. Still, performance can suffer, and the hit can be bad enough to create some real problems in terms of ride and handling. And sometimes a part that works perfectly as intended fails because it causes a problem for other parts.

No one builds a Locost just because they want a Locost. They build a Locost because they enjoy building. And they want a Locost. But many Locost builders enjoy the design work as much as or more than the build, and there’s no limit to the amount of design work you can do, everything from radical changes to the body or frame, to the choice of donor, to the drivetrain layout. I enjoy designing stuff too, but I want a Locost more than I want to design something new. I’m perfectly willing to stick to an existing design if it increases my odds of ending up with a running, fast, fun-to-drive sports car.

I guess this is my long way of saying I’m not going to deviate from the original design of the Locost any more than is absolutely necessary. Although the book doesn’t supply every detail and every design decision, enough working Locosts have been built that a lot of best practices have been developed over time. I’m not a mechanical engineer, so I’m going to be more like Bill Devin.

An M.G. Locost Build >> Building a Locost >> My Locost Build Philosophy