History is a nightmare from which I’m trying to awake…
But at least it’s better than geometry!
– Lil’ Stevie Daedalus
I’ve been doing a lot of work on design lately, and a lot of thinking about the design process as well. So I’d like to get back to the planemaking part of this blog with some discussions about that process, and how I approach designing planes. If you’re not the sort of person to appreciate such discussions, please feel free to skip the words and just have a look at the pictures. I won’t be offended. Having said that, though, I do think some readers might find it interesting, and perhaps it will spark some comments on design, or even inspire some thought about your own process of designing projects.
Here we go:
I should begin by saying that my general preferences in infill planes tends toward the traditional designs. My favorite antique planes are the Spiers and Mathieson models of the late 19th century. If you look at panel planes specifically, there is a typical sidewall profile that was established by Spiers, and later adopted almost wholesale by nearly every major infill maker since. I take this as proof that it’s just that good a design. I’ve yet to see a variation that even ‘works’ especially well, much less improves upon it — and this includes my own attempts.
As a starting point, I’ve borrowed a couple of images from Australian goldsmith/jeweler/planemaker and old tool collector Peter McBride. (Peter’s website is highly recommended, by the way. It’s loaded with a wealth of information and photography of his collection of tools, as well as excellent writeups of his own excellent planemaking.)
Here they are:
Using these two photos, we can see the prototypical sidewall profile. This basic design is remarkable for just how ubiquitous it is for infill panel planes; other than the craftsman-made traditions (particularly in Scotland), the vast majority of panel planes adhere pretty closely to it. First, the major elements: starting from the blade and moving forward, there is a rather iconic double-sine-wave curve traversing the throat area. This curve terminates in a concave upswing, where it meets the front infill just ahead of the throat. From that point, the sidewall drops in a subtle concave curve to a fairly low height at the front of the plane. Moving backward from the blade, there is first a convex slope that shifts to concavity as it meets the rear tote, and then continues in an elongated, graceful curve all the way to the heel of the plane, mirroring of the front termination.
There is a degree flexibility here – such as the final sidewall termination (note the curved ramp on the Spiers plane vs. the perpendicular drop on the Mathieson). Overall, though, there is not a tremendous amount of room for radical innovation – at least not that I’ve managed to find yet. What interests me so much about that even with all these ‘constraints’, if one pays attention to infill planes, it’s still relatively easy to identify the major makers’ designs from across a room.
Anyway – pressing on now (no going backwards – you have to go forward to go back.)
I have been sketching ideas and designs, both radical and relatively traditional, off and on for a couple years now, and have built four or five panel planes with varying degrees of success. Over the past couple of months, I’ve settled down into a range of designs that I’m rather keen on. So a few weeks back, I settled on a design that I thought had enough possibilities to merit a prototype – which is the only real test of the viability of a design. Here is the AI drawing I worked from for this plane:
There are a couple of aspects here that are relatively significant departures — most especially in the addition of an additional peak at the rear tote, and in the transition at the bed between the rear of the plane and the throat. The overall effect of these changes is a plane that looks, to my eye, fairly modern, and which had some nice echoes of my smoothing planes. Some of the features were a bit ‘risky’, but at the end of the day the only way to know where I’d overshot and what worked was to have a plane in front of me. The finished prototype is pictured at the top of this entry, and also below:
In general, one of the rules of thumb I use is that really good designs look better and better to the eye with each viewing; bad designs may look pleasing at first, but they tend to become tedious and less pleasing over time. So, having spent some time looking at the plane, and discussing it with a couple of very design-heavy friends, there were some parts of this design that I liked , and some that I decidedly did not. The ‘nots’ in this case were significant enough that the design on the whole doesn’t work for me. My friend Konrad expressed the problem perfectly with the comment that the profile looks more like a series of strung-together curves than a unified design. While most of the individual parts of the profile might be fine in isolation, the ‘flow’ of the overall profile is severely interrupted in a couple of places, and the profile ends up with a somewhat disjointed and jarring ‘feel’. In this case, there were two major issues. Not surprisingly, they were also the two major deviations from the traditional design.
First, the arris at the rear tote: while I did like the way it sort of mirrors the front of the tote, it’s too tight a curve. In the end, I decided to keep the feature, but to soften it somewhat, hoping it blends into the entirety of the plane better.
Second is the throat area of the sidewall, especially the transition at the blade. The more I looked at it, the more it gave me a headache – never a good sign. I tried some subtle changes, but at the end of the day I think it simply does not work. In this case, I ended up scrapping that part of the profile and reverting to a much more typical version of the traditional design. As I’ve mentioned before, I do think there is plenty of room in that design for the designer to express themselves, and I tried to stay toward the more modern end of the spectrum than either Spiers or Mathieson’s designs (at least those above) — which is part of the ‘signature’ I like to shoot for in my planes.
Here is the revised drawing:
Overall, having spent several days with the redesign, I have a fairly good ‘feeling’ about it. So tonight I’ll move back into the Lab with the templates and get to work. I’m hopeful that I’ll be happier with the redesign – but it will be another week or so until I can be sure.
It is worth noting that this particular process is quite typical for me – my first pass at a finished plane tends to be two steps away from the traditional designs, and the next one takes one step back toward convention. I’ve become quite comfortable with this pattern, and it generally works well for me. For the first design, I try to push the envelope as far as I think is possible without overstepping, and then I execute it to get a sense of what works and what doesn’t. From there, I almost always end up revising back toward the traditional design, and usually end up with a plane that I’m happy with — one treads the line between the canonical design and a more updated and modern one. Knowing this as ‘my’ process is immensely helpful to me, as it gives me a pretty good framework to understand when I can stretch, and when I should be conservative… where my instincts tend to be on, and when they tend to overstep their bounds. It also means
I’ll post more on this plane in a couple of weeks, when I’ve made some significant headway on the next plane.
Until then, please consider the possibility that Iggy, now well into his 60s, apparently laps a lot of planes: