Building a new dome shutter
Index
Introduction
Fibreglass is a very strong and lightweight polyester plastic composite material you can 'lay-up' or mould in the garage.
When i started this project I had some minor previous experience with this materials and techniques but none with moulding or building large scale moulds.
I also had trouble finding 'noddy' guides answering the practical questions of what materials to use and their
preparation for releasing mouldings from moulds. So I wrote it down . Hope its useful.
On the way I discovered a local GRP one-man-band while I was part-way through construction. He had a mould segment hanging from his workshop roof
of a 24' mosque onion dome. He had also made a 3 metre diameter hemisphere early that year and thrown the mould away since due to lack of storage.
Morale : if you want to make your own observatory - find your local supplier and ASK!. He also said that the hemisphere had cost £4500 GBP (April 2005 ) to the customer.
I could have had it for nothing had I asked at the right time.
the reason for this page is that I needed to make a new fibreglass shutter for the observatory dome because the original blew out in a
storm the previous year. Actually it blew in. The lightweight shutter rides on rollers front and rear and along flanges
on the dome sides to keep it aligned.
The shutter I replaced it with was entirely satisfactory
- it was 12mm plywood arc sides with thin MDF surfacing and foam filler for rigidity between the front and back faces.
However after saturating with thinned resin and cladding with GRP for long-term
waterproofing, it weighed a ton, metaphorically speaking. So much so that it distorts the rest of the grp dome depending
on its deployed position. Then after moving house and repairing the dome halves and replacing the joining section
it became clear that the width of the joining section had increased by a few millimeters sufficient to make the shutter
no longer slide or roll - at all.
Back to top
 |
The ridges the shutter rides over - under the lip of the shutter you can see the folded aluminium bracket that carries the axle and small wheels at either end. |
|
 |
Cladding the plywood shutter in fibre-glass. Its already absorbed resin thinned with 15% more styrene. This picture shows the two
side curves clad with fibreboard and the whole outer resin-wetted and covered with fine glass tissue. Then a thicker layer of mat is now being applied to provide the main strength. |
|
 |
The plywood dome shutter seen on the dome - there are two shutters - a removable lower white grp one and the self-colour
grp-clad plywood one. The plywood one rolls on wheels at either end. The white one lifts out. |
|
 |
Detail of the roller on the shutter that keeps the shutter from blowing off. There are two wheels - one either side of the shutter.
the bracket is made from folded 2mm Aluminium sheet.The long steel bar stretches between the two brackets and rolls in pressed-in bronze bushes
liberated from an old printer. This axle carries theouter wheels. In theory driving this axle will cause the shutter to rise up.
In practice the shutter is too heavy and the plastic wheels I used don't grip slippy. But it does make the shutter very easy to roll back.
In fact - too easy- it will happily drop under its own weight.
|
|
![]() |
After popping it out of the mould and transferring the hardware - pic to come! |
|
Back to top
Requirements
So I needed a new shutter for the observatory - to be able to observe at all.
The key requirements were :
Back to top
Making the mould
I made a mould in plywood based on the size of the clad plywood shutter, allowing extra width forthe
the increased width of the joining section and incorporating a moulded lip to allow sealling over the existing
lower shutter.
The mould was made from 3/4" plywood with a 4mm plywood inlay providing the curved shutter flat surface.
The depth of the inlay controlled the depth of the shutter edges. The curves were
marked out using a long compass ( spar of wood, pencil and nail at either end ). A curve for the inlay was routed out using
a 6mm straight cutter on a home-made compass arm for the router. The two sides of the mould were then assembled using
battens to hold the sides the correct distance apart while the flat sheet was slid down the router groove from top to bottom.
This bit is not easy. I've done it twice now - once for each shutter. The first one I used candle wax to grease the grooves for
easier sliding but I couldn't do that for the second one because I wanted to paint over it to seal it.
There are really only two options, feed the sheet into to a rigid frame and hope it fits all the way or spring it into a less rigid frame
and stiffen the frame after to prvent it springing out again.
I found the former the only way to proceed. Anything that can spring in can spring out again before you can apply stiffness.
Using a combination of pushing the sheet along the routed grooves using a wooden mallet and spare spar to protect the end, and pulling it using a a g-cramp to sandwhich the
end and provide a handle for pulling worked very effectively. I also found that I had to run a spar between the sheet and
frame stiffeners to prevent the sheet from bowing and springing of its own accord.
The end of the mould with the overlap moulding required the use of some lightweigth battens fastened across the end
of the sheet and some more hand-routed curves into which these lightweight battens could be sprung in sections. These were then
nailed to the support spars of the framework underneath and the curves filled out with bodyfiller.
The entire woodwork of the mould surface was then sanded back before giving it a good varnish with Ronseal exterior hard varnish.
This I assumed was a one-part polyurethane varnish. I followed that up with a surface sanding before coating again, sanded again, final coating of varnish and final sand.
 |
The latest dome shutter - mould manufacture. This mould is made from two large pieces of 3/4 inch plywood and a melamine-covered fibre
piece slide along a channel cut with a router on a bar-trammel. Stiffeners and cross-braces keep the shape stiff. Blobs of foller on the back surface retain the fibreboard. |
|
 |
Detail of the rebate along the mould lip where the grp panel needs to meet and ride over the lower grp panel ion the dome. The mould has been painted in a two-part yachting epoxy at this point
as a suitable finishing surface to apply release agent to. |
|
Following further discussions about release agents and the practicalities of removing the GRP component from the mould,
I then applied a further layer of 2-part polyurethane gloss paint, as often found in marine chandleries. Nasty isocyanate hadeners and more... but finishes to a great hard gloss coat.
The reason for this was that the varnish when tested reacted against the polyester resin used in GRP by wrinkling. If the GRP and varnish react then
all my efforts would be for nought.
This two-part paint is strange stuff, in the tub its quite runny, as soon as its applied its very stiff to work and then when you think you have it suitably spread out
it starts running again. I had a great finish in some areas and some major runs in others.
However my paint brush when applying the varnish hadn't been as clean as it ought and so there were parts of the varnish
poisoned by a different chemical which subsequently reacted with the two-pack gloss coat and caused serious wrinkling and
bubbling where the poisoning had occurred. Everywhere else the paint had a good hard grip on the mould and varnish. Where the varnish was poisoned, it was a soft skin you could peel off and bubbles you
could rub off the paint with your thumb back to the wood.
The only solution was to sand back those posioned areas and major runs to wood and body fill to match the varnish/paint levels.
After that I gave up messing about and applied two coats of PVA release solution, the first by brush and the second by sponge. The sponge layer comes up much better
- its more even if you can avoid bubbles and also is much less prone to streaking.
Back to top
Laying Up
Then it was time to lay up the glass fibre. I mixed a litre at a time mostly. I did a coloured gel-coat first consisting of 500ml resin-A, 50ml polyester colourant and catalyst.
this resin went on as thickly and evenly as I could get it to try and get a thoroughly opaque and even white coloouring on the shuter outsides.
Then after this had catalysed, ( took all night it was so cold ) a layer of glass tissue was resined in place with some medium-weight CSM glass on top.
As I worked through the CSM layers I started with 4% catalyst and eventually moved to 2% to achieve working time in 1 litre pots, even in the 5°C working temperature.
After 3 layers of 350g/m2 CSM I then finished with another layer of tissue - running out of tissue and CSM as I went.
Back to top
Important lessons learnt
A ribbed consolidating roller worked very well, kept in a pot of acetone between uses. Applying a width of CSM across the shutter at a time and staggering joins worked very well at keeping time per layer down. I'd already body-fillered into the corners to provide a smooth chamfer curve When laying up so I had no problems with CSM in corners. When CSM extends beyond the mould edges, cut to about an inch protrusion and resin it all the way up so it joins the previous layers. This means the
edges are well supported and makes a throughly resined edge for later trimming which is far easier to handle as a single surface rather than scrappy ends of glass mat.
It also means you trim and sand the edge back to your mould edges later. Do rough trimming at the 'green' stage and full trimming prior to removing from the mould.
I did the rough trimming with the glass scissors and the neat trimming after cure with a hacksaw blade. the blade should be pulled towards ou against the support of the
mould edges to prevent break-out of glass fibres.
Back to top
Removing from the mould.
The mould is all flat surfaces. This either makes it quite hard or quite easy depending on circumstance. There's large areas to grab but no complex shapes to 'pop'.
Starting from the end I pushed with a piece of wood to start lifting the centre and spread it to the edges. Then I went along the
edges of the shutter pushing off the grp from the mould with a wooden chisel and block hammer.
Repeating this all the way around and using a long chisel for lifting the shutter surace all the way down the mould eventually lifted it after half and hour.
The PVA surface was like a layer of cling film starting to peel off.
A little of the gloss paint had come off the mould in places but could be rubbed off the shutter with a thumb.
Finally, I went around all the edges of the shutter sanding with a coarse sander back to the line of the gel coat.
this gives a nice clean finish all the way around and cleans off all the extra glass left by the neat trimming.
And the result ? - Wow its light!. Its also a bit more flexible than I expected and thinner - about 3mm wall-thickness.
If anything this needs beefing up a bit. By the time I fix the roller hardware on it will be much stiffer. All in all a very useful experience .
![]() |
The finished item. |
|
Back to top
Fixings
In the end I fixed the roller hardware on to the shutter using car body-filler applied to a freshly sanded metal surface. The surface was sanded usng #60 grit
. The body-filler was one that described itself as having a strong bond with metal and GRP.
My key points for a next project would be :
Varnish IS fine as a mould sealant - use a one pack polyu and ample PVA release. I applied PVA release direct to the porous body filler and it still came off fine. I guess even the hardware store ordinary varnish is fine as long as you have a good application of PVA. Keep your brushes clean and single-use only. ( Don't mix chemistries.. ) Think about stiffness and use foam sheet. ITs very light but stiff, takes resin and it's easy to laminate into the finished product. Apply two layers of coloured gelcoat. Mine was too thin in areas. A layer of glass tissue on both outer surfaces makes handling the finished product a lot nicer afterwards.
Back to top
Next Projects
My next projects are going to be :
Making a geodesic dome in GRP using 2mm insulation polystyrene sheet (roll ) as the former. This should be fine with a lightweight cladding
of tissue left to go cure before applying the heavier CSM and further tissue. The only issue here is to protect the polystyrene with emulsion paint or latex
to prevent resin attack by the polyester resin.
Make a GRP mould for a 4' dome based on the same design as the skybadger dome. This would be purely for robotic small telescopes not requiring human intervention.
Back to top
Suppliers of GRP in the UK
In no order and not being exclusive
Glasplies - supplies a lot of others. Based in Southport, Lancs. This is where I get most of my advice and supplies from. CFS - based in Redruth, Cornwall. A great printed catalogue with lots of help and reasonable prices with fast delivery. Search Google The Marine Store
Back to top
|
