This is the penultimate blog post in this series titled Representation and culminates in the documentation of a physical model I have made in the workshop at The University of Greenwich. I chose to develop a model rather than the alternative options of an animation or three renders. This is because I enjoy the opportunity to work away from a computer screen and I wanted to work with different materials to better understand the tidal flows of the river Thames – a variable in my studio design project.
Photographs of the Final Model can be seen in this post.
My main priority when looking for precedent models was to find a suitable method of representing moving water, however this presented a challenge. Most models were pristine and static or ‘messy’ with motion i.e. the materials and viewer were going to have to get wet.
I decided to combine the topographic approach of Cary Foster with the sense of wave movement achieved by Norman Diaz, rejecting – for the purpose of a 3-week project – the inclusion of actual water. I did purchase hydrogel (which LCLA office used to ‘flood’ their work, by heating with a lamp from underneath) and will test the suitability of this later in my studio design project.
Process – six components
Site: Impounding Station, Isle of Dogs; River Thames
- As the twice-daily rise and fall of the Thames inform my design project, I wanted a section of the river to be the main focus of the physical model and therefore modelled this digitally in Rhino.
- In reality, the Thames and the Isle of Dogs (my studio brief) is a relatively flat area, with heights varying from -5m to +5m yet the river spanning 550m in width at this point. I exaggerated the topography of the river bed by x10 fold to create visible yet still proportionately accurate variation at a model scale of 1:1000.
- Sourcing of timber was the next challenge, as I wanted a hardwood with attractive grain yet with a depth of 90+mm (length 600mm x width 200mm). I had a piece of London Plane cut to size by specialist timber merchant City Wood based in Bromley-by-Bow.
- The 3D .stl file and timber were then supplied to the workshop to be cut on the CNC machine (computer numerical control). The timber had a minimum +10mm excess on the three ‘exposed’ edges. Waiting to access the machine has been the biggest challenge due to long queue times.
- Other key elements of the .stl file include 9 slots – where the CNC machine cut through the timber entirely – and grooves to hold my proposal and any existing buildings (which would be created using 3D printing).
- To finish the topography, I sanded any rough edges and added an oil to the wood to enhance the grain.
- As per my original sketches, I wanted to create a handle mechanic – also from wood – that could be manually moved back and forth at a 90-degree angle to the topography to mimic the tide. I modelled this in Rhino to plan the construction and dimensions.
- This was cut from the same London Plane sourced from City Wood. The 480mm top section and small end pieces (30mm x2) I cut by hand on the Band Saw, including the diagonal edges. The larger 860mm bottom section was cut on the Circular Saw, which was a lot more accurate and produced a better end result.
- Using an AutoCAD file edited in Adobe Illustrator, I transferred an image of the Thames on to the 480mm section. This was achieved using the Laser Cutter and the raster engraving option.
- I created two finger grooves on each end piece by using the Pillar Drill on a double length piece of wood (60mm) and then cutting this in half, again on the Band Saw, to create two identical pieces of 30mm in length.
- The four separate wooden pieces were then glued together with wood glue, clamped and left to dry overnight.
- To finish the handle, I sanded any rough edges and added an oil to the wood to enhance the grain and match the topography.
- As a variation on the design by Diaz, I wanted to incorporate nine unfixed sheets of acrylic that could move up and down inside the topography. By placing them on the wooden handle and then sliding this back and forth, the model would attempt to represent the rise and fall of the tide.
- Again, I modeled the concept in Rhino but then created the final files in AutoCAD before editing in Adobe Illustrator. I sourced 5mm gloss white acrylic and cut the 9 ‘tidal waves’ on the Laser Cutter using vector cutting only.
- As per the above component, I modelled the concept in Rhino, prepared file in AutoCAD and edited line weights and colour in Adobe Illustrator. I imported text from an Adobe InDesign file.
- The base has been Laser Cut on transparent light grey 3mm acrylic using both vector cutting and vector engraving. The engraved text represents quarterly tide times and heights from 2017.
- I added a finger joint to the edges of the base primarily to give the structure additional strength, as I was concerned that the acrylic might struggle to hold the weight of the timber topography if it was flimsy in any way. It has also been glued.
- The design proposal of my site have been modeled in Rhino and then supplied as an .stl file to be printed on the Formlabs 3D Printer in grey resin. The Form 1+ is a desktop 3D printer that ‘prints’ onto a liquid resin, solidifying each layer of the model with an ultraviolet laser.
- The model was printed with a support structure that held the model together during the printing process. I removed this with a scalpel.
- The existing buildings of my site have been modeled in Rhino and then supplied as an .stl file to be printed on the Ultimaker 2+ 3D Printer in white plastic.
- The Ultimaker creates tough, solid ABS/PLA plastic models ‘toothpaste style’ by printing each layer with a melted plastic, known as FDM (fused deposition modeling) printing.
Photographs of the Final Model can be seen in this post.
4 / 5 A short diary of five introductory software sessions, covering Rhino (two parts), ArcGIS, 3D Scanning and Adobe After Effects. Delivered via The University of Greenwich as part of the Landscape Representation module within the Master of Landscape Architecture program.
2.5 hour demonstration workshop of how a 3D scanner works. There are two components to the result: firstly a scan of the physical geometry, and secondly a scan of the colour information i.e. a photo of the space, taken in full panoramic.
A 3D scanner:
- Scans what you can see…. that is to say, it travels through glass, but can’t see through objects. Always do more than one scan of any space for this reason. The best place to hide is behind a wall, or under the scanner!
- Requires at least three scans, with correct setup needing to be able to see scan site #1 from #2, then needing to be able to see #2 from #3, etc.
- Set at the maximum resolution it will take about 2 hours per scan and it’s highly unlikely a computer will be able to process multiple files: so reduce the settings. Low scan in comparison will take about 5 minutes and the capture is still very high resolution. Only use the highest settings on very small spaces.
Scene software allows you to:
- Make either a point cloud model (which you can’t edit in 3D software), OR make a Mesh (which is really easy, and means you can then manipulate the file further).
- Follow a clear workflow bar that sits across the top of the software program. The ambition is to make all 3 boxes in ‘Project’ (tab 1) turn green, and at that stage, the process is finished.
- Select the ‘Registration’ tab (see pic), where you can stitch all of the scans together. Here there are 3 options, being ‘Auto’ – which may or may not work; ‘Manual’ – which is very easy: matching viewable elements; ‘Visual’ – which is organising parts. We tested ‘Manual’.
- In the ‘Manual’ option, pick the top image from each scan > mark targets (see pic) > mark a point or a plane > keep marking > the top button goes green. Points and corners of things tend to work better than planes.
- Navigate around a rough model in the ‘Explore’ tab, which will be a lower resolution on screen than what is actually captured. Use an auto-clipping box to quickly tidy up the file: move and rotate this to fit. You can use as many clipping boxes as needed.
- Finalise your master image via the last option on the ‘Create’ tab. Typically keep the settings on the default options, then this stitches all the scans together.
Export considerations include:
- File types are typically .stl – for 3D printing; .ply – not very commonly used; .obj – the universal file, can be used in any software (obj = object)
- Point cloud is good for background fills. In last ‘Export’ tab, export project point cloud. Studio Max doesn’t work with .pts (point cloud) file, but export this then open up Recap, scan project, open and save as .rcs (Revision Control System) file.
- Point cloud option can’t be used in Rhino, but can be used in other 3D software like Studio Max (Autodesk 3ds Max) and Blender.
- The settings when making a Mesh have the option to make it watertight, which would then be suitable for 3D printing…. But remember it’s 1:1 and will likely morph some of the shapes. Mesh turns your scan into lines and triangular faces. Mesh Selection > mesh clipping boxes > will turn yellow. Meshes > right click > export > save.
- A Mesh should import into Rhino easily, however, don’t apply lighting as this is already baked in from the photographs. When viewing the model in Rhino Render mode, it will retain the photo information. Note: this is if you have scanned with colour. You can turn colour on/off, and without it will give whitey grey overall finish. Turn Mesh Wires on/off in display settings of Rhino.