Rolling a 'tri-arc' or 'cove' roof.

[John Candy]

On my 'to do' list are several horse boxes and other pre-group npcs items.
Any tips on successfully rolling variable radius roofs in brass / ali?
I am not certain slip rolls would cope with the tight curves of the edges.
Perhaps styrene sheet is better for this type of roof?
Any suggestions appreciated.
Thanks,
John.

[MikeWilliams]

I don't possess slip rolls alas, but suggest thinner brass - maybe about 10 or 15thou - and laminate it.  If your rolls are too large you could roll that my hand over a dowel.

I have made 4mm roofs using narrow strips of styrene flooded with solvent and sanded to chape.  I've also used a solid wood former sanded to shape and covered with styrene, but much prefer some solid metal in there somewhere.

Do your slip rolls curve right to the edge anyway?

Mike

[Andy B]

What are the actual radii you need to achieve?

A couple of ideas I have (neither of which I have ever tried, unfortunately):

1) Could the the roof be made from 3 sections, soldered (butt joint) together, using a support / former?
If the outer radii are smaller than you can roll, could a length of suitable tube be split lengthwise into suitable sized arc sections?

2) Another option, if you can find someone with a good sized fly-press, could be to make a press tool. For a limited run, using thin, well annealled brass, I would have thought that a close-grained hardwood press tool would be adequate.

Andy

[Andy B]

P.S. Mike - if you need something rolled, drop me a mail.
We can do an exchange once my MR W-irons are cut!

Andy

[John Candy]

Thanks for those ideas, all of which are possibilities.

My slip rolls should roll to a min. dia. of 1" (I have never tried that small) but do roll to the edges of the material. I would guess that rolling first across the entire sheet with the largest radius, then removing and setting to the smaller radius before rolling just the edges of the sheets should produce the correct shape but it could be a 'trial and error' job to get the profile exactly right to fit the ends.

Looking at drawings which do not have the radii marked, some (e.g. LNWR and LSWR) appear to be very much tighter than the approx. 11 inch radius which the 1" diameter scales down to in G3. Others, such as the NER have a much less pronounced curvature and may be possible with a 1" roll.
Other alternatives I have considered :
1) My original idea was to make a wood pattern and vacuum mould in thick styrene (I would need to construct the vacuum moulder which sounds simple according to the book I have but it is American and the parts required may be more difficult to find over here).
2) Making a mould and and producing the roof in GRP (messy but produces a strong but light structure).
3) Moulding in resin, using a semi-rigid RTV rubber mould made from a wooden pattern.

I am not yet ready to start on the project so have some time to mull over the options and try out a few techniques.

John.

[MikeWilliams]

John,

If you are thinking of moulding a batch in resin, rather than a one-off, them I would certainly laminate with a thin metal layer.  The thin brass could be rolled by hand at the edges round a small tube and if slightly uneven covering with 10 thou plastikard either side would even that up.  You may need a couple of layers.  I'd not use that for a one-off because it could tend to delaminate, but if you only need it long enough to make a mould it should last.

Mike