Testing and Conclusions

Testing with water

We used two watering cans to imitate the effects of rain, we continuously filled them up and poured them onto the dome over five minutes and recorded the results via video.

As the video shows, the mud cladding was not very waterproof to say the least....

Why did the cladding not resist water?

The mud cakes, although smoothed at the joints, shrunk and receded from each other to create cracks and holes in the cladding (we should have foreseen this as it was extremely obvious to us after it had happened).  We believe that the straw needed to be cut much more finely as to integrate with the mixture more thoroughly. In hindsight we probably wouldn’t include the straw unless we were doing two layers of cladding, as it is purely an insulative measure. The sand that we mixed with the clay and straw was granitic sand....in other words, the wrong sand. The sand wasn’t fine enough and therefore made the mixture dry inconsistently and hard to apply.

The mud isn’t dry!!!

Over the few days that we allowed the mud to dry it had rained, but never fear, it was covered by a big farm shed..... Little did we know that the shed leaked right where the dome was situated and therefore the mud cladding didn’t dry fully, although it was still a lot drier than when applied.

How did we try to rectify the situation?

Upon realisation that the mud had receded and shrunk during its interrupted drying period we decided to make a mixture that consisted only of clay and water, apply it to the cracks and holes and attempt to dry it under high heat. This would, in theory, stick to the structure more successfully than the previous mixture and was able to be easily poured into cracks to fill them. This worked to an extent but again, we underestimated the drying time and some of it washed away with the application of water.

If we had our time again

If we were to change a few major things about this assignment they would be:

-        - The mixing and application of the clay/mud cladding. We would use finer sand, like the sand used with cement and we would (if using straw at all) chop or blend the straw into a much finer mix.

-      - The mixture would be applied to the outside of the inside skin of chicken wire and then the outer layer of chicken wire added, followed by another layer of mud on top of that, this would create a much stronger skin and would be much more waterproof.

-         - The entire dome, once the two layers of mud cladding had set, would be covered with a thinner (with the         addition of more water), more cement-like mixture of clay sand and water, omitting the straw.

-        - We would have made a ridge (similar to the one around the window opening) around the edge of the cladding at the halfway mark where we stopped cladding to show the structure beneath and create a cut away effect, so that the water did not drip over the edge and instead ran down the sides and onto the ground.

In conclusion, we believe the structure held up beautifully, it was of sound design and assembly and it looked great, it lent itself to a number of different claddings and was lightweight, enabling it to be moved around with ease once constructed. The cladding was a bit of a failure but the results needed to construct a much sounder model were obtained and recorded for later use or use by other people looking to make a mud clad geo-dome. All in all it was a success in that we discovered why and how things did and didn’t work. 

Archive photos for the day 

The Window!

The window was an issue as we couldn’t source a piece of glass in the right shape and setting anything into the mud was also an issue. The frame, now covered with mud was also hard to access and cylindrical struts are hard to attach anything substantial to anyway. We eventually decided to continue along our ESD themed cladding selection path and chose to use recycled polypropylene bags, not environmentally friendly in themselves, but better to use them for something than to leave them lying around polluting the place. We opted for an operable window, one that opened (as the material wasn’t translucent). The challenges that this presented were how were:

1. How do we affix the window to the frame while still allowing it to open?

We ended up cutting small sections of fencing wire (approx. 150mm long) and punching them through the bag and the mud cladding and plying them around the aluminium frame beneath, allowing for a circular movement, effectively making them into a primitive hinge.

2. How do we make it as waterproof as possible without using a waterproof spray?

We melted the openings of the bag shut and them taped them over with a foil tape (creating seems), we also layered the bags as much as possible to minimise permeability.

3. How do we give the bag some kind of structure so that it holds its shape and doesn’t pool in the event of a downpour?

To create some form of structural resistance and shape within the bag we inserted between the polypropylene layers the chicken wire that was cut from and previously occupied the window space. This stopped water from pooling in the middle of the bag and allowed it to run straight off. We also applied fencing wire around the edges of the triangular window piece with a small loop of wire extended out the front to allow it to clip down and hold itself shut (refer to photos).

4. How to we stop water from seeping in the sides of the window?

When applying the mud cladding we created raised edges around the window opening (shown in photos) effectively creating impassable hills to deflect incoming water, the window material folded down around these rises to further support this effect.

Making and Applying the Cladding

Preparation

To prepare for the mud cladding, we wrapped three quarters of the dome in chicken wire. This would provide something for the mud to grip onto. We used cable ties to attach two layers of chicken wire to the aluminium frame – one external layer, and one internal. We thought the mud was likely to press through the first layer, so a second layer would provide additional support.

We chose to clad just half of the frame. This meant we were able to see all the internal elements, which was particularly important when it came to testing. It also made for easy access during the cladding process.

Mixing the Cladding

As mentioned earlier:

“To make the mixture, we will need 2 parts clay-rich soil to 1 part sand. We will estimate the amount of water and straw we need. All components can be sourced from parts of Julian’s farm. We have found a particularly rich deposit of clay, so will most likely use less than the 2:1 ratio suggests.”

After gathering all the components, we started to mix up the cladding. Having watched a video about the process (shown in our previous post, Sourcing Materials: Cladding [07/06/11]), we decided to make the mixture on a tarp. This allowed us to pick up one edge of the tarp and fold the mix over itself, making sure everything was thoroughly combined.

To start with, we blended the clay-rich soil and sand. Stomping on the mix helped break down any lumps into finer grains. This was also a good opportunity to remove any impurities such as rocks or grass.

Next, we created a well and poured in a small amount of water. Using shovels, we turned the mixture over itself, then ‘chopped’ through it with the edge of the shovel. After a few minutes of shovelling the mix, we stamped it down. It was really important to use the tarp to fold the mix over every now and then to ensure all parts were properly combined. We repeated this process a number of times until the mixture was a nice consistency – it was malleable but also held its shape, even when thrown to the ground.

We then started to add some straw. This required a similar process to the one we used to blend in the water – we stamped the straw into the mix, then used the tarp to fold it over itself. We repeated these steps until we were satisfied that the straw was thoroughly dispersed.

Applying the Cladding

Instead of just slapping the mud on, we rolled up handfuls of the mix and threw them on the ground to make cakes. This made it much easier to press the mud onto the chicken wire.

Instead of making the cakes all the same size, it was more useful to have varying widths. Larger cakes were better for areas over the aluminium frame, whereas smaller ones were more valuable in the areas between struts.

To help the cakes adhere to the chicken wire, they needed to be massaged and pressed in. It was easy to apply too much pressure, pushing the mud too far into the wire. We resolved this by having someone inside the dome press gently outwards against the mud, while the person outside massaged it in.

To avoid gaps between the mud cakes, we overlapped them and tried to smooth them into each other.

It seemed obvious to start from the ground up. As we worked our way up the frame, it became easier to fix the mud cakes. This was partly due to us gaining experience as we went, but was mainly because the more horizontal each cake was, the more gravity would hold it in place.

We left a triangular opening at the front of the dome as our window. To divert water around the opening, we formed the mud into a ridge around the edges.

Archive of Photos - Cladding

Assembly of Structure

Flattening the ends of the struts:

To accommodate the holes for the bolts at the nodes, we had to flatten the ends of the struts. We began to do this with a vice but after about 2 hours abandoned this idea as it was immensely hard to close the vice on the strut ends. This was something we thought we would accomplish more easily with the use of aluminium rather than steel......oh how wrong we were!

Click the picture to enlarge 

We then decided to use a hammer to flatten the ends, this required about 10 or so hits at each end of a strut, produced inconsistent results and tired the arms quite quickly. We had to devise yet another way to flatten the ends. Finally we realised there was a sledge hammer leaning against the wall at the other end of the shed in plain view, almost mocking us.

The sledge hammer worked perfectly! It took only one or two hits and produced the same nice flat result every time.

Drilling the ends of the struts:

We marked about a cm in from the ends of each flattened strut end and drilled a 7mm hole to accommodate our 6mm bolts. We had to rest the strut ends on a flat piece of wood to ensure consistent and neat holes.

Bending the ends of the struts:

The ends of the struts had be to bent in particular directions and at particular angles (depending on their place in the dome) so that they would all fit together nicely when added to a node. We did this simply by hitting it with the hammer. It could have been done, probably with more finesse, in the vice.

Putting the struts together:

A step by step guide to the assembly of the structure

Step-by-Step Frame Construction Photos

Archive Photos - Constructing the frame

Problems

Bending and flattening the ends

We realised, while putting the struts together, that they would not all fit (especially at nodes with 6 vertices coming together) unless a longer section of the struts was flattened at each end so as not to interfere with one another at the nodes. This required some disassembly and re-flattening.

The dome instructions made no sense

The instructions and research we had gathered on the actual assembly of the dome and integration of each strut were relatively clear before we commenced assembly. This was not the case however, when we began assembly. Absolutely no sense could be made from the backwards and poorly explained assembly process diagrams and descriptions so we went about making our own (above, under Putting the struts together).

Sourcing Materials

Structural

Having looked at various methods of construction (including timber and paper, as shown below), we decided that using aluminium conduits would be the most structurally sound option, as well as the most durable. The lightweight properties of aluminium will allow the dome to be constructed and transported with ease.

 

Julian’s Dad, a construction manager, suggested we try Capral – a cheaper option than Bunning’s, as well an independent Australian supplier. They were nice enough to cut the conduits down to the sizes we needed (specified below).

Cladding

We considered a few different cladding options.

One possibility was to cut triangular segments out of MDF, join them, then seal the joints with silicon.

Another possibility was to cover the structure in chicken wire, then cover it with papier mâché. We then would have needed to add a sealant and an internal insulative layer.

We also considered glad wrap – this would look fantastic, however it would not be an appropriate response to an ESD brief. Glad wrap itself is not a sustainable product, nor would it provide any insulation.

We kept coming back to the idea of using mud and straw, and have settled on this as our final choice.  We will attach chicken wire to the aluminium structure using cable ties. This will provide something for the mud to stick onto.  The following video helped us figure out how to make the mud/straw mixture:

To make the mixture, we will need 2 parts clay-rich soil to 1 part sand. We will estimate the amount of water and straw we need. All components can be sourced from parts of Julian’s farm. We have found a particularly rich deposit of clay, so will most likely use less than the 2:1 ratio suggests.

Calculation of Parts

To minimise waste, we are using a measurement of 500mm for Strut B as our starting point. This results in the following measurements:

Radius: 914.9mm

Strut A [or Longs (L's)]: 565.4mm

Strut B [or Shorts (S's)]: 500mm

Connections: 26

Total parts needed:

Strut A: 35

Strut B: 30

Bolts/nuts: 26

Washers: 52

To achieve these results, we used this dome calculator.  This particular calculator cannot work out figures using Strut B as the starting point, so we entered different radii until we reached the desired outcome (500mm for Strut B).

Considerations of the Structure

We will be constructing a '2V' dome, that is a dome with two strut lengths, struts A and B. The lengths, obviously, depend on the diameter of the base of the dome. 

These lengths will be calculated with the help of the dome calculator at http://www.desertdomes.com

Strong Members 


As the dome needs to take the load of a covering/cladding (in a real situation this would be the roof and walls and therefore rather weighty) the members making up the structure must be strong. For this reason we have chosen to use 20mm steel electrical conduit. 

Stiff Joints 


The dome must avoid structural movement and for this the joints between the members must be stiff and strong, similar to the needs of the joints of a portal frame. For this the conduits will be flattened at the ends and holes drilled through the flattened sections to accommodate bolts to tie the members to one another.