Linton HODSDON Architect 0416 426 218

Anybody wishing the full copy of this report ( with ALL the photographs of each stage ) or any further details of the project can contact me thru this site.

My path to understanding the geodesic dome started with building a cardboard soccer ball ( 3v geodesic sphere ). After meeting Laurence at Babinda ( QLD ) in 1998 and seeing the program he had written to design geodesic domes, I built a 1m dia. 6v dome frame using kebab skewers, plastic tube and cotton. That dome moved around my lounge room for 8 years ( the period of my experiment to apply ‘ livingry ’ to local government politics ) before the next stage began.

6.2m dia. GEODESIC DOME @ 13 Mann Street, Margaret River.

A building license for a 6m x 6m shed was issued and earthworks started in September 2006. With the help of Laurence’s 6v dome printout, manipulation by photocopier and traced off at 3v, working drawings were drafted for engineering purposes. Roger Pateman engineer advised that he would have to inspect the finished building before he could certify the structure.

The concept for the structure was based on a design for a 3.6m dia cardboard dome found on the BFI website. Triangular panels with a 50 mm tab along each edge, a waterproofing gasket and 5 tekscrews per edge joint … with no structural frame … seemed viable for a 6m+ dia. dome in 1mm thick steel. A 6.2m diameter was found to be the biggest 3v dome possible with panels cut from 1200 mm wide colorbond steel sheet, which also suited window / door heads at 2.2m high. The waterproofing at the apexes was addressed by ss eyebolts thru the inevitable hole at the junction of the 5 or 6 panels, with rubber grommets inside and outside – the eye of the bolt located on the inside to support a tensioned fabric ceiling.

After securing a 1.2 tonne supply of steel sheets ( in any colour I wanted, providing it was Wilderness Green ) and finding a willing fabricator ( Warren @ BVA, Bunbury ), a schedule of panels was drafted and a sample panel was fabricated ( the very accurate zincalume panel built into the south face ). The production run of panels resulted in atleast half being outside of ( anticipated ) acceptable tolerance and were sent back for re-folding.

Meanwhile, the siteworks for the square shed took on a circular theme – the FRC pipe in the centre of the sandpad, that had been ‘ hanging around with no particular purpose ’ for 20 years, ‘ just seemed opportune ’ for the centre of the floor slab. Thought on how to assemble the panels at 4m+ above ground level found the first application for the hole formed for the middle of the slab. The hole was dug down to 1m deep and a 5.4m long 200mm dia. bushpole was set into it. That is after nearly getting it up by hand with ladders, and putting my back out in the process, then calling on the wonderful simplicity of employing Leeuwin Engineering’s crane hire.

At this stage, the dimension creep due to variation of panel size and gasket compression was not known. Similar uncertainty existed about how to line up the panels in the assembly stage – so a jig was made for each of the 3 different edges to pre-drill the holes for the tekscrews. Although this did not fully address the problem of variations in panel size, the holes were vital to adjust the alignment of the panel edges ( with the use of 38 mm nails ) to achieve a satisfactory joint. Panels were put together with the nails and vice clamps to allow the joint to be opened up to fit the gasket, the nails were then used to lever the panels to the best alignment, then clamp next to the hole, remove the nail and tek it off… checking to make sure the gasket has not moved. Panels were assembled in sets of 3 ( easily done while standing the panels on edge ) and then the assembly rack was used to join 2 sets of 3, with the gasket fitted from the top and the tekscrewing being done sitting on the ground underneath. This was all easy ( and good exercise ) for one person, up to a total of 9 panels.

Setting the guide holes for the tekscrews close to each apex ( 50 mm ) did not allow enough room to get the drill in to drive the tekscrews. A flexible shaft for the drill helped get the tek started but a ratchet ring spanner was used to tighten them up properly. After breaking 3 flexible shafts, I relied more heavily on that trusty ring spanner.

After assembling several sets of panels; allowance for such creep was made, slab formwork was setup ( thanks Mick ) and steel door frames were welded up ( thanks Richo – particularly for introducing me to the hinge weld joint, that was vital to being able to adjust the outer roller door frame to fit the reality of the panels in place ). With the complex roller door frame weighing nearly 170 kg, the crane was again employed to move and locate the frame into place.

It became apparent that the best quality panel assembly was done near ground level and that the top of the dome
( above 3m high ) would need to be assembled on a 2m high pedestal and lifted on by crane. So 5 sets of nine panels were assembled and then lifted into place ( by hand … luckily all still with 5 fingers afterward ) each with temporary timber framework to hold them in place while connections were made.

With a continuous ring of panels above door head height, all just sitting on the slab formwork, the slab edge thickenings were dug ( after a trench ‘ cave in ’, the opportune placement of the FRC pipe found another vital application – to wedge against to realign the formwork with the 400 kg of steel panels resting there on ), slab reinforcement was laid, the door and window frames were hung from the panels and the concrete slab was poured.

At this moment the unique acoustics of the dome became apparent – amplification of sound within the dome, with greatest amplification at a particular distance from the centre.

By this stage it had become patently clear that this building was different to the issued Building License for a conventional shed. After a quick inspection, the engineer was happy to certify the structure for an amendment to the building license. 2 hours later, the Shire’s principal building surveyor rang to enquire about the anomaly. A few additional notes on the working drawings together with the engineer’s certificate were submitted for the amendment to the Building License - that was a Friday. The next day, I left for a 3 month trip to Europe and on the Monday, apparently without Planning department input, the amended license was issued making the structure legal.

JUNE 2007

Upon returning from Europe and starting assembly of the top of the dome, I had a nagging suspicion about the durability of the ‘ sponge chord ’ gasket used between the panel tabs. So I found a vacuum hose that should offer better resistance to weathering, reassembled the top panels with the new gasket and dismantled the most exposed panels already erected to replace ( and better fit ) the gasket. In the process, I actually found that the ‘ sponge chord ’ ( that sealed the joint between panels very well ) had weathered well over 6 months ( both summer heat and winter storms ) and, after release, recovered from the compression. The vacuum hose required more pressure to properly seal the joints ( 4 additional tekscrews per joint ) but was easier to install and hopefully will better resist weathering. In the end, the 4 additional tekscrews were also used on the joints with the sponge chord gasket.

Next came fitting the eyebolts at the apexes. A 23 mm dia. grommet either side of the colorbond did not seal the junction … a larger 29 mm dia. grommet was found … but that did not work either. Eventually, each eyebolt was fitted with 3 grommets, 2 custom 40 mm dia. rubber washers, 2 nuts, 2 ss washers and a generous dollop of silicone. The intention to avoid adhesives / silicone sealants ( a philosophical desire to make the structure cleanly demountable ) was thereby compromised … but that made it easier to use silicone for some of the leaks found when water testing the top before it was lifted into position. Once lifted on, outside of the top of the dome would be inaccessible.

The crane lifted the top into place on a timber lifting frame. An automotive scissor jack was located on top of the bushpole to raised the top off the lifting frame and then progressively lower the top as it was stitched to the lower panels. Fixings to the door frames had to be removed to release the lower panels to flex to mate with the top – the resulting mismatch between dome and door frames was covered by the gutters and opening trim panels. Although the top did not mate perfectly with the bottom, the misclose resulting from the variation in panel sizes created an even 11 mm gap along the last edge ( instead of the 3 – 4 mm margin for the gasket ). Rather than have a new panel fabricated to close this gap, a flashing was fabricated to create a weatherproof permanent vent.

The construction role of the bushpole was now finished and the method of removing the 5.4m high bushpole from the 4.7m high high dome was designed to preserve the bushpole for new gate posts with only 200mm of wastage. The bushpole was removed and the hole in the floor was capped with a plywood cover to await another opportune purpose ! The space within the dome could now be properly appreciated and the sound of their own speech at the focus of the sphere has spooked some people.

Any flex evident before the top was fitted off disappeared when the structure was complete. With the structure now totally rigid, the standard panels truncated by the PA door and window openings were modified, and the door and window frame side panels could be measured and fabricated. The reveals to the roller door opening were lined with ply wood, the gutter fitted and the roller door installed. The PA door and window side panels, gutters and window were fitted. A fixed plywood panel was fitted to the PA door opening as temporary closure. 25 mm dia. PVC electrical conduit was used for rainwater down pipes to the 50 x 50 colorbond gutters over the window and PA door.

Installing the ceiling, a proper PA door and fitting a TARDIS drive mechanism into the hole in the slab will be fun for the future. The construction of shed 3 reused the timber used for the slab formwork for floor framing, the other construction timber was reused for the wall and roof framing and the last 2 sheets of colorbond were used for external end wall cladding. For the moment, the dome is used as a lock up garage.

The finished weight of the building, excluding 10 tonne slab, is approximately 1 tonne.
Project cost Au$17,000 incl slab & siteworks but excluding Richo’s and my labour.
All the construction support timber ( $1,000 ) was reused and therefore did not present as a cost to this project.
AUGUST 2008

Below : Dec 2007 and Finished