The prototype's roof has 4 vertical support poles placed symetrically and these would need to counteract the turning moment of the lifting mechanism. So they would need to be constrained at both their tops and their bottoms, yet still be free to move up and down without imposing too much mechanical resistance. To achieve this I designed and printed a roof support frame which fitted the inside curve of the roof and with grooves running down each corner to take a piece of 0.8mm dia brass rod. The roof was fixed to this temporarily and the whole contraption positioned, then when I was happy, the four bottom corners were glued to the sides of the body, approx 5mm below the top edge. This ensured that I would have 4 perfectly vertical rods in due course.
- Screenshot 2022-10-26 06.40.48.png (205.05 KiB) Viewed 4775 times
The roof was removed from the frame and 4 lengths of brass rod cut. 2mm at one end of each was bent to an angle matching the roof curve, to sit in a small groove in the top face of the frame. Fortunately, at this point I realised a potential problem, which I'll come back to, but as a result I glued only three of the legs in place and omitted the one behind where the driver would be. Next he roof was glued to the frame and the frame was then carefully cut free of the 4 bottom corners, leaving them glued to the body. Finally the 4 legs of the frame were cut free from the roof, leaving the 3 support rods hanging down. Offering it up the body and inserting the rods into the fixed bottom 'bearings', I checked that the roof slid up and down quite freely.
Now came the bit I was very uncertain about, the driver needed to be able to move out of the way as the roof came down and given the open and cramped interior there was no real way of hiding a mechanism, but I did have a vertical rod behind the driver if I could somehow use that.
The driver was resin printed to sit on the stool, the dimensions of which I got from one of the drawings. I worked out that by slicing him in half diagonally from his waist to his bum and putting a hinge in his groin, he would just clear the lowered roof if folded forward, so all I had to do was work out how the rod coming down could tip him forward and then straighten him as it went back up.
The vertical rod movement is approx 29mm but from sitting to leaning forward the body vertical movement is only about 15mm so the body needed to be pushed up by a rod coming down twice as far! Physical space, physics and the laws of mechanics were not my friends at this point.
Somehow the idea of a Bowden cable wormed its way into my brain cell, followed by a bit of lever mechanics. To cut a long story short, I fixed a short length of brass tube to the bottom of the roof dropper that was still not fixed and to this I soldered a short piece of rod at right angles. I had to draw and resin print a restraint to force it to move vertically, this was basically a flat backed tube with a slot down the front face to allow the pin to travel up and down.
Next I fashioned a lever from a piece of 4mm scale bull-head rail. This had a hole at one end to fit over the dropper pin and near the other end I drilled/filed a slot to be a sliding fit on a 14BA screw.
This was all mounted on a piece of 12thou brass, cut and bent to exactly fit on the floor/internal wall of the loco at the drivers position and would allow me to work on it with having to be concerned about the rest of the model.
By coincidence the vertical tube was exactly behind the centre of the driver, so to get a push up as the rod came down I needed to reverse the direction and also reduce the travel by approx 50%. Going back to the Bowden cable idea, I resin printed a U-shaped tube with 1mm dia ID and then put a length of monofilament nylon fishing line through it, this is 0.55mm dia and slides easily. The lever had a hole drilled in it, to again take a 14BA screw and I clamped a small piece of brass aound the end of the fishing line, drilled it and fixed it to the lever with the 14BA screw. The position of this screw along the lever determines the amount of vertical movement and was located to give a 2:1 reduction.
The tube was printed at an angle, so that it was underneath the lever at one end but protruded past it at the drivers end and the fishing line stuck up vertically just under the back of the driver.
Originally I had thought to simply drill an almost vertical hole in his body and put the line into that but it proved to be too stiff, so I had to fashion a very small knuckle joint and let one side into his body and clamp the line to the other side and this works very well.
Hopefully these two pictures will show what I've just tried to explain:
- P1010005.JPG (124.62 KiB) Viewed 4775 times
- P1010007.JPG (172.69 KiB) Viewed 4775 times
This all worked fine..... on the bench! However, I had a small problem in that if I glued the operating rod into the roof structure I would be unable to get at all this mechanism and I knew that would be asking for trouble. So I drilled the mounting hole in the roof, to take a small piece of 1mmID brass tube soldered inside a nested piece of larger ID tube (to give it a thicker wall). I then drilled and tapped it to take a grub screw and glued it into the roof hole.
Put it all together and gave it a try - Fine to begin with but then about halfway down everything siezed up. Trouble is it is difficult to see whats going on inside but eventually I realised that the curved roof support is exactly in line with the drivers head and he was jamming on it. So I printed a separate curved support and glued it in about 15mm further along and then when all set, carefully cut out the original one. Problem solved...except that although he could go flat and I knew there was sufficient room, he kept getting stuck about 3/4 of the way down. Eventually it dawned on me that he was bending forward into the downward curve of the far side of the roof and there was just not quite enough room. Solved that one by carefully slicing the superglue holding his feet to the brass floor and then even more carefully filing about 1mm from the soles of his shoes, thus tipping him forward very slightly and reducing the angle between his head and the roof. As the video shows, that seems to have fixed it.
If anything doesn't make sense please feel free to shout!