Saturday, March 7, 2015

The value of mock-ups and a fundamental design compromise

Doomed - Photo: Darren Kall
While designing the sculptures I did for the Electra C. Doren library there were lots of open items to research, track down, and resolve. There were unknowns and decisions that would impact my final designs and, if unresolved, create risks to success. That made me feel uncomfortable about quality and it made me anxious that it would mean rework for me because the issues would reveal themselves too late in the process to be mitigated - as they often do.

I was working with a whole new crew of fabricators, installers, and others. I tried to resolve open items through asking questions and proposing approaches but even that didn't resolve all the issues with the design or with the process steps between all the partner companies. We needed… an experiment!

All the players needed to walk through the process of building a sculpture to discover the "gotchas". We needed to build a physical mock-up by the same process we planned to use for the final sculptures. It needed to result in something we could see and touch, so that we could decide about open issues.

For example; I'm familiar with unpainted steel surfaces but not with unpainted sheet aluminum surfaces. What would they look like? What would the ground surface treatment I designed look like? I was told that the aluminum should be treated to preserve it since exposure to weather would impact the aesthetics of the aluminum over time. Would a clear powder coat give me the luster of steel I wanted? There were stakeholders who questioned the combination of fence color, sculpture color, pavers, concrete, and brick that will make up the terrace. We brought samples together for an approximation and it convinced me that the fence coating called "bronze" worked well with the paver colors and raw aluminum. But there are stakeholders who were not there and reserving judgment until they could see it. Another stakeholder recommended that a black coating might work better and others started to question my choice of raw aluminum.

I had a ton of other questions like how much visible space is taken up by the mounting equipment? How much of a border do I need to leave uncut in aluminum to have it support my design? What would 1/4 inch aluminum look like in cutouts compared to 1/8 inch steel? Would I have to change my design to accommodate the new thickness? What format did the fabricator need my final artwork in? How accurate would the CAD interpretation of my final artwork be? What does the cut edge of water-jet cutting aluminum look like compared to laser cutting steel? Will it have burrs? Will it be too sharp for safety? … I was boring everyone with questions they couldn't answer.

I proposed that we create a mock-up and walk through the whole build process end to end. In the process I learned I wasn't the only one with questions. Stakeholders jumped on the idea. The fabricator, the architects, and the library also wanted some open items resolved. Mock-ups to the rescue for all of us!

Most notably Ultra Aluminum, my fence fabricator partner, was very concerned about my unsupported ~ 4 foot lengths of rib patterns that made up the sky in my designs. 

In steel they would be no problem, but in aluminum the fabricator couldn't guarantee that these ribs would be covered under warranty without adding in supports.  They were concerned that the aluminum would be susceptible to breaking if someone put too much weight on the aluminum. But where to put supports in the design? It made the most sense to put the supports between the ribs at the points where I was creating Sheeler-like sections of the fractured sky. I was creating an illusion of a line but the supports would be actual lines between these sections.

Adding supports to my ribs would be a fundamental compromise of my design. It would be a step backwards. I had earlier sketches with supports and having supports was just not as innovative a design as the one I created through illusion. I made ribs that vary in thickness as they transition over the different Sheeler-like sky fractures. By creating a transition between thicknesses along an imaginary line it appears to observers that there is a line there. Observers "construct" the line without thinking about it. With supports the viewer won't have to "construct" the lines in their perception; it will be easier for them to see the sections of the fractured sky. To me this makes it less challenging and less interesting as a piece of art. But this is not just art, it is also a fence.

Ultra Aluminum is a subject matter expert and they know their material better than me, but I was attached to my design. So together we decided that one of the mock-ups had to be several 4 foot ribs of aluminum that could be tested for durability with supports. I gave them a paper cutout of a handful of 4-foot ribs. They converted this to a machine-readable CAD and checked the metal specs. It wouldn't work without supports. If this sculpture were hanging on a gallery wall or suspended in a window, there would be no need for supports. It is because it is an accessible fence that people might hit accidentally that the ribs might fracture. The library couldn't get the warranty they wanted unless there were supports added. I compromised and gave into the design change. Next I needed to mitigate the issue and reduce the impact on my design.

The first step was for the Ultra Aluminum CAD specialist to simply add straight line supports along my illusion lines. We started with 1/2 inch wide supports connecting my ribs. It was a hard compromise but I went along to see the result. The 1/2 inch didn't look too bad in the CAD drawing so the first mock-up was cut in metal. Immediately after it was done, and before I could even see it, the fabricator rejected it and cut another one but with 1/4 inch wide supports. The 1/2 inch was just too thick aesthetically.

For this fundamental compromise and all the other open items we designed two mock-ups and assembled them into representative railing parts.

Mock-up of "sky bars" with 1/4 inch supports, edges not sanded, black powder coat
The first was 1 ft H by 4 ft W with several 4 ft long ribs including 1/4 inch wide supports, with a black coating, with no sanding of the edges of the water jet cut openings.

Mock-up of shapes, clear powder coat, sanded edges, hand-ground surface

The second was 1 ft H by 1 ft W with a region showing a variety of cutout shapes but no ribs, with a clear coating, and sanded cut out edges and a hand-ground surface finish.

When the mock-ups arrived I answered all my remaining open questions, made my final design decisions, and prepared to present them to the Library Art Board to get their final approval to produce the final artwork. More on that meeting in another post.

Are we better off by having walked the process and created a mock-up? Absolutely. Could we have reached the same conclusions with the same level of risk reduction and increased confidence without the mockups? No.

End-to-end process-created mock-ups are the best time-saving and cost-effective way to resolve open risks. 

No comments:

Post a Comment