Tyrertecture
Home    Info    Ask

Contact     Friends     

Tyrertecture is the blog of designer and lecturer Nick Tyrer

Nick is currently working at Pattern Architects, and is Studio Tutor to the Abstract Machines MArch Unit, Leeds. His research explores computational fabrication and digital craft.

All work found here is Original

tyrertecture@gmail.com


Modelo - Experiments with vertex colour

Modelo is a new platform for viewing 3D models online. I have been trialling it over the last few weeks, and i am very impressed with the service and the freedom it gives, to access 3d models anyway without any software. Perhaps I have been using the service on different types of 3D models than what those the service was intended for. Instead of architectural models of many individual parts i have been using it for my complex organic sphere designs, that is a single mesh with many vertices.

I quickly found the for the moment the beta service limits the mesh complexity at 65,000 vertices, so I remodelled one of my recent prints to come in just under that total. 60,000. The design is a single woven topology, three layers of hexagons and pentagons thread themselves with a rotational randomness.

Default

When I first uploaded the above HexWeave model I loved the extra dimension. Navigating it really made a complex form a lot more understandable and even made the complexity more impressive. My only minor complaint was the visual impact of the model didn’t particularly pop, I am not sure whether its the shade of grey, or perhaps the lack of contrast, the model just does’t have the crispness that I wanted. As Modelo accepts mesh colour, I decided to experiment with creating my own colour/shading effects and painting them directly onto the base geometry to add to visual impact that I was looking for. After i started thinking on the subject, I started to see more application to this, how carefully colouring your mesh could enhance aspects of the design.

Colour Gradiant

My first idea was to brighten it by adding colour, but instead of a global colour, I could use a gradient that is determined by distance to centre point of the object. That way the colour would help to reinforce the sense of depth and layers in the object.

Gray-scale Gradient

I wasn’t a big fan of the colour, so I decided to stick with B/W. I coloured this mesh again according to mesh vertices distance to centre point. Black being the furthest away. An unexpected benefit of this style was on the white background the outer branches were amplified.

Ambient Occlusion

My 3D printed models are pure white, yet the complex forms are easy to distinguish, this is due to Ambient Occlusion. Shadows naturally occur where two objects meet, or are in close proximity. I set out to create my on version of AO. The method I ended up using was to project each mesh normal and tested for collisions, then coloured that mesh according to the distance to that collision (white if no collision). I really love this result, it really increases the contrast of the model, and enhances the weaving forms. The inside of the sphere end up much darker, obviously, but improves clarity of the weaving forms, it hides the pattern in the background.

Light External

After success with the AO method, I attempted creating a lighting method. This decision may have been too ambitious, without any real knowledge of lighting or rendering. My theory was to break the lighting into two methods. The first, a direct light pass, was too find which mesh faces were directly exposed to the light source without a collision, these should be the brightest point on the mesh and so should be white. After this process the mesh is high contrast, black and white. The second pass is to add some light to the est of the mesh. For this I measured the angle of each mesh face to the light source, the largest angles (meshes facing the opposite direction would be black, meshes facing the light source would be grey). A bit of tweaking was required with the gradients to get the result I wanted, while the method is totally fabricated nonsense, it gives a lot of flexibility to fine-tune.


Light Internal

The final model is the same process as the previous, but the light source is placed inside the mesh sphere. The results in this case doesn’t work as well as I had hoped. This effect works much better with a black background.

So in conclusion while I hope Modelo works on its shaders and visual settings, maybe adding some functionality in that area, but as a beta product it really is great. It has also given me this short but educational experience trying to create my own graphic effects, i will be continuing this research in future.

Marching Cube Studies
Implementing mesh creation algorithms in grasshopper. By manipulating the values of the point field i was able to make the mesh massing more organic, almost viscous, gathering at the nodal points and thinning across the spans.

Marching Cube Studies
Implementing mesh creation algorithms in grasshopper. By manipulating the values of the point field i was able to make the mesh massing more organic, almost viscous, gathering at the nodal points and thinning across the spans.

Standbeest
Video of this little guy walking in previous post.

Mini Strandbeest Model
Built this interesting little guy last night. The movement is hypnotising. When i get some free time, i would like to create my own. I have never attempted to design or 3d print joints before.

Interlocking Basket Lamp
Featured in London Design Week 2013

Interlocking Basket Lamp
Featured in London Design Week 2013

Interlocking Basket Lamp
Featured in London Design Week 2013

Interlocking Basket Lamp
Featured in London Design Week 2013

Hexagonal Origami Simulation

I decided to revisit an old project based on origami to learn some new skills. The hexagonal origami animation is simulated using the Kangaroo physics pluggin for grasshopper.

The main forces acting upon it are ‘hinge’ rotational forces to create Mountain and Valley folds. The ‘Planarize’ force to each hexagon ( as each hexagon is actually 6 triangles that need to be constrained to a plane). I added ‘sphere collide’ calculations to each vertices, just to stop any overlapping. Finally some rigid springs for each face edge, to mimic real world paper properties.

Animation
Unfortunately the animation/.gif won’t load properly. But you can see the simulation in action here: http://goo.gl/ZiC9Dn

Origami
This is the actual Hexagonal Origami piece i folded for the project. (It took about 8 hours).

There and back again
For me this exercise was very rewarding, it felt like i was coming full circle on my time as a parametric designer. It was the original origami project that set me on the path of parametrics. Desperately searching for a way to model its expansion. The only modelling tool i knew at the time was sketchup. The discovery of grasshopper really was a light bulb moment. And while i was very proud of that first definition (below), it was a very top-down system, that was built on purely geometrical transformations.

I have been very busy with work, and recently teaching. But i plan on being a lot more productive with my time this year, or at least document my research more. I have a slight back log of items i will upload soon.

Interlocking Basket
Selected by D2W for the London Design Festival. Small ornamental piece continuing my previous research on Erwin Hauer. The basket is made up of two independent but interlocking surfaces.

The two surfaces are hard to distinguish in the photos, while I appreciate the effect of the two surfaces blending, I think printing each in independent colours/shade would improve viewers understanding.

www.digits2widgets.com/
www.londondesignfestival.com/

Ribbon I

Ribbon I

Vascular Screen
Experimenting with transitions of solid to void.

Vascular Screen
Experimenting with transitions of solid to void.

Cutaneous

Cutaneous

Muscular

Spinal

Spinal

Pandora Plant

image

I was procrastinating, decided to make weirdest object i could. I feel my weirdness concept was achieved.

Kind of looks like a plant out of Avatar… Or some kind of acid trip!

www.tyrertecture.com

Mobius v3
The 3d printer in the office hadn’t been getting a lot of action recently. Used the occasion to set myself a speed challenge. This mobius based form took 3 minutes in grasshopper.