Site banner

What works is what works


The car-friendly guide to fixing human-made greenhouse emissions.

Free Download!
Sample Chapters
Get the real book from
Accurate Stories
Or, get the ebook from

The Truth about the Shiners Cover

Buy it for your Smartphone, Tablet, Computer, or eBook Reader, from
Barnes & Noble
Diesel eBook Store
Kobo eBook Store

Free Download
Sample Chapters

What works best works best

Next month my friends at Friendship Systems are holding an international users meeting. That means I have to put together a poster showing how I use 3D parametric modelling to speed up the yacht design process, and why the kspline is the best hull design tool a yacht designer will ever meet (except for a multi-billionaire yachtie).

That got me thinking about a comment by Giovanni Belgrano, Emirates Team New Zealand's Principal Structural Engineer, at Auckland University’s Fourth High Performance Yacht Design Conference in March 2012. He said, "Nothing happens unless it goes in the 3D model.”

Fair enough. Whether you're designing an America's Cup yacht or a ballpoint pen, you must make sure everything fits together and does exactly what it is expected to do. Three-dimensional modelling helps. Every drawing is built around data extracted from the 3D model. You won't get problems such as a dimension on one drawing being different to the corresponding dimension on another drawing.Keelboat hull

Parametric modelling takes this one step further. It's a lot easier to change the model as the design evolves. My last hull design project, for example, went through several iterations. Moving from one iteration to the next involved changing only a handful of parameters. With conventional modelling, I would have had to tear down the model and rebuilt it each time I wanted to change it.

Thinking about Giovanni Belgrano's comment reminded me of an idea I picked up from  Robert Woodbury's book, “Elements of Parametric Design.” I'm pretty sure he said that parametric modelling is messy and contingent. If he didn't, he probably should have. My main parametric model, the one I used for my last hull design, has been ripped apart and rebuilt several times. It can still draw an Open 60, which is what the first version was used for. But the latest version is a lot more versatile and easier to use than the original. If you look under the hood, you'll find parts that look like professionally-written code. I've tried lots of things that I've chucked out, which is what Woodbury means by contingent. You use what works. But you always end up adding something you didn't need on the previous project. The newest parts are always the messy bits. As soon as you get them working, you move on.

The newest part of this hull model is the sheerline (the top edge of the hull). It's the most important line on any yacht. Only in the last year or so have I been really happy about the way my computer handles this bit. People who have seen the prototype of this boat all comment on how good it looks. That's because of the sheerline. You've probably heard people say that Emirates Team New Zealand's AC72 looks a bit agricultural. That's because its sheerline is absolute rubbish. Fortunately, an ugly sheerline can be just as fast as a pretty one.

Three-dimensional modelling can be incredibly time-consuming and expensive. That's not such a problem with a hull model. A well-designed parametric hull model can be used for dozens of individual designs. You don't even have to build one from scratch. I'll give you one for free if you buy a commercial subscription to the Friendship Framework.

The decision is not so simple when you come to things like keels and rudders. A simple conceptual change might mean you have to rip up your 3D model and start all over again. I've already done that three times with the keel for my current project. I might as well have drawn it in two dimensions. I didn't because my 3D modelling system happens to have a library of aerofoil shapes. I could have extracted the foil sections and dropped them into my 2D CAD. But like a mug, I built a 3D model. And another. And soon I'll probably go and do another one.

Computer tools force us to use computer tools. No matter what you're designing, a 3D model reduces risk. It's the hub of your system. Every drawing taken from the 3D model should fit with every other drawing taken from that 3D model. The 3D model is the master programme for any CNC machined or 3D printed parts. Your competitors do it. If you don't, you'll end up making a mistake your competitors would never make. (That's the theory, anyway.)

The trick is deciding when to go ahead and build a 3D model. I could have drawn six keels on my 2D CAD system in the time I spent drawing two 3D keel concepts. Situations like this are not uncommon. Sometimes you need to get all the way up to the first or second prototype before you put together production-quality documentation, such as a 3D model. Other times, you're better-off building that 3D model right from the start. Whatever works best is the best thing to do.

I'll be happy if my poster shows how 3D modelling helped speed up this hull design, and how to get a decent sheerline. As for deciding when to commit to a 3D model, and how much effort to put into it, that can only come from long experience. Which you can only get from building 3D models and then converting them into reality.


Kevin Cudby is a Wellington-based Freelance Writer and Technologist who loves writing about cool new technology. Email him to discuss your communication requirements: hello {a}

+ Text Size -
Original generation time 1.2409 seconds.