Confluence: complete

Confluence_180817_88523 (Large)

Skip to the end for making details.

They would have been well within their rights to have thought the day would never come, but after eight months from finishing the concept, ‘Confluence’ is complete. To say the reception to the piece was positive would be putting it mildly. To quote the new owner:

“…Paul has completely blown us away with this beautiful new piece that we commissioned him to do… Confluence celebrates every single aspect of our lives and she will continue to do so for generations to come… We are more in love with it by the minute as we keep seeing more and more depth and character… Paul’s deep personal perception of us, his making and exceptional artistic creativity really do resonate from every angle.”

This slideshow requires JavaScript.

The making

Like most worthy pursuits, the making was both maddening and a joy. Without a single straight line in the design, I had to be meticulous in the planning of every step, much more so than normal as most tasks resulted in the removal of a flat working surface, leaving fewer and fewer reference surfaces to work off.

Usually straightforward procedures – such as following the general rule of bringing timber down to final dimension over time to allow for wood movement – were also complicated and time consuming. Having said that, however, the design’s organic form and little use of jointing meant that any unplanned wood movement that did occur was imperceptible in the end.

The size and shapes of the components meant I couldn’t use the jointer for flattening, having to resort to using a router ski that slid along on top of the torsion box. As I didn’t have a spindle moulder, this router ski was also used to template-trim the frame’s width, but resulted in at least an additional week’s worth of exhausting work.

Having documented the detailed design in CAD meant I could get the MDF templates, forms and guides cut by CNC down the road, however it became clear quite quickly that my curves that looked ok in CAD were in some places not perfect sweeping arcs in the finished templates. It wasn’t the CNC’s fault – I just hadn’t smoothed out some of the nodes enough, leaving me to spend a day reworking the router-cut edge by hand.

By far the biggest challenge on this piece has to be the modest concave curve that sweeps through the face of both frames (this curve is imperceptible on photos except for the below top-down perspective shot, look closely at the junction of the frame and table).

Confluence_180817_88538_c (Large)

Initially I was planning on using a router cradle swinging on a pendulum that was limited to movement in one plane but after several attempts the rig didn’t feel safe enough to use for long periods, so in the end I decided to use it only for establishing the reference depth at points along the frame and then take the bulk off using an Arbortech planner cutter on an angle grinder. The risk with this tool is the speed at which you can take off material as you can quickly go past your finished dimension. As it is a freehand power tool, the quality of the surface is limited by how controlled your sweeping movements are. Even my lightest use required days of follow up using a short flat rasp and sanding using 40 grit. A compass plan could potentially have helped had I bought one, but this would only help in the horizontal plane across the concave, not along the length of the concave.

This slideshow requires JavaScript.

The rebating of the mirror was straight forward enough, again using the router ski to freehand route up to a pencil line trace off the mirror. In hindsight, I’d cut the mirror after having the frame nearly complete, as there are some parts where the tolerance was less than 5mm between the inside of the rebate and the outside of the frame. The three metal brackets I had cut and made up by a local fabricator, and while I think  their work was excellent, for the cost I think it was overkill and I could’ve made these up myself.

The frame’s finished thickness is up to 70mm thick, which allows for a subtle contrast with the thinner parts of the frame towards the centre and bottom. Standard commercial thicknesses typically maxes out at around 50mm in Australia, which meant I had to source a slab, as I didn’t want to laminate up to get the required thickness. After a long search I found a 250kg slab of 100mm spotted gum. Most slabs are from small scale specialists who find big ‘paddock’ trees that, while enormous, have more ‘shake’ (cracks) than forestry derived trees as they are more prone to wind stress. However, considering the natural/organic themes of the brief and the final design, the shake (once filled with dark epoxy) actually complimented the piece, adding to other striking features such as the the gum lines that run radially through the timber (you can see these come to the surface at a tangent in one of the photos of the frames detail.

I’m very satisfied with the piece but had underestimated the time to complete some of critical steps due to the size of the components. However, in the end, the proof is in the client’s reaction, and I hope they’ll agree eight months is a short time to wait for a piece that will last generations.