Once the roof structure was built Tom used superglue to attach the roof structure to the beams, this ensured we got a strong bond between the two as it was a very fragile structure. I had produced the windows that are located in the roof from the cad data, if these were used as a guide for where to stick the plates to the beams then I wouldn’t have had to amend their sizes which I now had to do as they had been stuck together randomly, I had originally extracted the template for the windows from the CAD data. I clad the rest of the windows with the flashings required from white styrene, the zed flashing was made from one of the bought styrene channels cut in half, this proved to be a much more effective solution than making our own.

Roof Structure

We opted to make the roof structure from 'off the shelf' styrene channel http://www.plastruct.com/Pages/OnlineProductDetail.lasso which I sourced from http://www.terrainwarehouse.co.uk. I used the full size drawings to make sure that the ends of the channels were cut to the right angle and length to look like the angled structure.The lengths of channel were glued together using dichloromethane, the internal angles were a full channel cut in half.

Further to the concrete foundations not being made from thick enough plastic, we opted to design and cut a lattice structure from 2mm thick acrylic on the laser cutter, this will ensure that the profiles will be accurately cut. We will then clad it in 2mm thick styrene to ensure that any deviations in the alignment of the profiles will not be picked up as it was on the previous model.







The concrete profiles which were cut out from 2mm acrylic on the laser cutter. Assembly was crisp for one the structure which I assembled, for the other which Tom assembled the laser settings were slightly different so the profiles required a piece of 0.5mm styrene to pack them.

Concrete foundations.

Both myself and Tom have spent an unacceptable amount of time on the cast concrete foundations and still not finishing them to a satisfactory standard. Firstly we looked into making them from machined MDF. This idea was superseeded because we thought that they would take too long to make that way.



We then opted to extract the data from the 3D Cad model I have constructed and print cutting templates which could then be used to cut out styrene profiles.



We opted to make them from 1mm styrene which is too thin! It doesn't hold its rigidity very well at all and will not remain straight.



This then has an impact on the skin that we are using to clad over the profiles which in itself is too thin at 0.5mm. This creates a faceted appearance instead of a smooth curve.



After discussion between ourselves we are going to look at the possibility of laser cutting the profiles using the same CAD data from 2mm thick acrylic sheet, this will remove the tolerances that inevitably appear when cutting profiles with a scalpel. The material we are going to use to skin the profiles will be 2mm thick also, this will allow any deviations in the profiles to be overcome as the material should hold the curve required by itself.

Starting the model- beams!

The 3D model allowed us to extract template data for the beams which could then be used to cut them out. It was decided that chemiwood would give us the best finish. Although the architects drawing shows the beams with sharp corners, each of the beams would in fact be rounded off to a radius so we decided to show this to remain true to the look of the building once constructed.

Scale Change!!

When making the backboard it gave us a true feeling as to how big the model was going to be at 1:10, too big. On further discussion we have opted to make the model at a scale of 1:20.

Drawing Issues

Whilst I can convert the data contained within a PDF to work within a CAD package the information contained within a PDF is not as reliable as CAD data in a DWG. Converting the data is taking the line image and converting it into useable data, however dimensions can vary greatly from what they are actually supposed to be. As such the group was working from the same drawings, but the information contained within them could vary between each group member. We decided to produce one set of CAD drawings that we could all work from. I modelled the building in 3D to ensure that the information I was reading would enable us to build the model correctly and that everything would work together and fit. Producing a 3D model allowed a 'dry run' at ironing out any problems with the drawing information and also ensured that all team members were working to one set of data and drawings.