原创 Lessons from my four-legged failure

You may already be aware of Theo Jansen, a remarkable engineer who has created a herd of wind-powered machines that he calls "StrandBeests," which literally translates into "beach beasts."

He designed them to roam the sand dunes, powered by the wind, walking on multiple legs consisting of a series of complicated-looking linkages. His goal is to release a herd of these beests, and as he puts it, "Let them live their own lives." If you haven't seen them, they are a marvel to behold.

Jansen's designs inspired me to build my own version of a StrandBeest. I have a BSME degree from Clemson University, and have worked on and designed process equipment for nearly 10 years. This experience led me to believe that mimicking a simple mechanical linkage would be a breeze.

I also have a well-equipped garage for building, well, whatever mad idea I come up with in my mind. I hoped that this education and experience would be enough to produce a working beest. Instead of designing in the seemingly endless number of legs on some of his beests, I decided on what I thought would be a more manageable four legs.

After briefly attempting to design my own linkages using CAD software, I realised I could save a lot of time and frustration by simply copying his design. But any simple four-bar linkage that I could come up with wasn't sufficient to produce a good walking system. After reading more on Jansen's device, it turns out that it took an abundance of computer simulation time to come up with what he calls the "holy numbers." Apparently this device is more complicated than I first assumed.

Fortunately, I discovered that a guy named Dominique Studer had already created a four-legged walker based on Theo's design. So I thought to myself, "Why should I have any problems if some other guy was able to make it work?"

One thing I think we as engineers sometimes do is assume that we know what we are doing, and that we actually can do it better than the person who created the original design. That does not always hold true.

My first mistake was that I decided to just start building the legs, even though I wasn't quite sure how I would actually make the gears that Dominique incorporated into his design. But I figured I could always compensate for that part later.

Theo used PVC pipe for nearly all of his linkage design. I decided to go with wood and PVC pivots, mostly for aesthetics. Unfortunately, the closest drill bit to the pipe I used produced a somewhat loose fit.

Once I mounted the legs on a frame that I'd set up, it was quite evident that the little bit of slop in each of the joints was going to add up, causing the legs to shift back and forth nearly a foot in each direction. With my beest at a standing height of close to 3 feet (.91 m), this kind of sway was clearly going to be unacceptable. Further, when I tried linking the four legs directly together, my beest wobbled uncontrollably during its "test push." The gearing in Dominique's design (that I blithely neglected) cleverly times the legs to mostly eliminate this wobble.

I haven't given up on my beest. By going to a much wider leg assembly I believe I can eventually make it work." That being said, I think there are some early lessons to be learned from my experience so far:

1. Don't let your ego get in the way and assume you know more than the guy who originally built the thing you are designing. You need to try and fully understand the general operating principles of the original design, as well as the trade-offs and choices made in the engineering process.
2. Obsess about the details. A few parts out of tolerance or a little looseness here and there in your design may not seem like a big deal, but when added up over an entire design those little errors may spell disaster.

Obviously there are more lessons to be learned. I'd also love to hear about the experiences you've had in designing something, whether for work or for fun, as well as lessons you've learned the hard way yourself.

About the author
Jeremy Cook is a manufacturing engineer with 10 years experience as his full-time profession, and has a BSME from Clemson University.