I'm at a point in my depot project when I'm getting ready to cut the tile roof panels. Does anyone have any tips to cutting them accurately? My strategy right now is to measure 5 times, cut slightly oversized, and then file down until it fits right. Getting the angles right is going to be tricky. I've got 8 pieces to cut, the roof is 'C' shaped with one main long section and 2 shorter sections on each end - see the picture below, it shows one of the shorter ends . . At $35 per pack for the Volmar roof panels, I don't want to make too many mistakes. . . . Here's my mockup showing more of the roof...
I usually use taped up card stock mock up to get the angles the way I want them. Then I use the card stock pieces for templates. For tile roofs you don't have to be exact because the joints will be covered up with the ridge row tiles.
Thanks Russ - I'll give it a shot. You're right about having a little wiggle room - I built a Taco Bell kit awhile back and the ridge rows do a good job of hiding small imperfections.
Well, that blows my contribution out. I was gonna suggest you figure out how much material you'll need and buy 3X that amount. :tb-wink: I'd do a mockup, using cardboard that's roughly as thick as the tile material. I've done some rough work on one of my ideas using the backside of cereal boxes. I bring home empty air filter boxes from work if they are white on the inside, they're about .020" thick.
If you have front and end elevation drawings, you can measure the height and width of each panel as well as the location of panel intersections, Do this on cardstock and cut trial panels before hacking into styrene. Even though there are no ridge tiles (i.e., butt joints only), the roof of my Colfax depot was done this way, and (with minimal filing) it worked out fine.
Just thinking about this. If you have these drawings, you can measure the angular slope of the roof, do a little trigonometry, and compute the exact plan view size of each roof panel.
Pretty much my point, but I found that the measurements could be made directly from the drawings (no math!). For example, the heights for the front roof panel (both center and for the "wings" was measured off of the end elevation, and the height of the end pieces was measured off the front elevation. After some mental contortions (remember those spatial recognition aptitude tests?), I found that there was always some plan view that would yield a direct measurement.
Oh yeah, right. This old brain has not done any of that in about 40 years. I just do the old cut, file and fit and somehow get it to work. In 1:1 scale I do the same thing. When I was helping my Dad design and build my parents retirement house, he was going through all kinds of mental gyrations and slide rule exercises trying to get the roof lines to work. I just got up on the frame and started putting up rafters, nailing down plywood sawing things off where they stuck out. He was very concerned and somewhat scheptical but pleased when it ended up looking like he invisioned it should.
A good old carpenter's square will do the trick, if you know how to read one. The trig functions are more or less engraved in the steel. Now to find one in N scale!:tb-biggrin: I've designed an built a few 1:1 houses with hip roofs and lots of valleys. I can still do the trig, which is helpful for ordering the correct length lumber. But, if a corner is, say, 1/4-inch off, the hip rafter will be 1/4-inch or more off, which could be too much, especially if it's on the short side. So I still ended up measuring. Or cutting the top end and then measuring. In N scale, the precision becomes pretty difficult. I've done the trig and cut all four sides of a hip roof from one piece of styrene, then folded it. Sometimes it fits; sometimes not! So I resort to Russ' method. For the compound curves on a ship, it's about the only way. On the Beavercove, the front of the superstructure is curved, and the deck is both crowned (curved from one side to the other) and swept (curved from bow to stern). It's fairly easy to figure this out on a CAD package: you draw parts as if they were straight (a chord), then use a little trig to figure out what the lengths of the curve above the chord would be. Then you just stretch the parts to that length. Simple--except it doesn't work! Who says I got the crown correct? Or the sweep? The method does give me a starting point, but it's still easier to snip paper than .020 or thicker styrene. Here's what I'm writing about: The front of the white superstructure is not so apparently curved in the top image; you can see it better in the bottom image.
