for what it is worth, here’s the slope angle:
Arch 9a.skp (203.3 KB)
Yes, this is very close. My measurements show it is just 2" off a hypothetical floor I created between the two legs - 874’ 10", but that may be just my placement of the floor being off the ground plane by that much too. The measure across the top similarly is probably distorted by the SU snap to default, but it is within an inch of 400’
The only real problem is the inner arch still bows out slightly. See the guide I created for the first 270’ on the lower left corner inside arch. The inner arch deviates outside the guide before coming back in again. The structure should go up very slightly curving inward from the start, then increase the angle as it ascends towards 875’ in a non-concentric fashion relative to the outer Catenary arch.
Thanks. This will help as I advance up the floors. The curving sides are obviously more tricky to calculate per floor (10’ each…mostly). I’m using a rule-of-thumb of 2’ in for every 10’ up = 200’ for a 1000’ radius.
The outer shape started as semicircular. I think you want the inner one semi elliptical. Will try that this evening - this same method will work
Try this. Inside is now a half circle, stretched up so its top is at 875’, semi-elliptical.
The inset is exactly 12’6 deep, and the offset is approximatelly 21’ - because I rotated the circles half a segment to get the end segment upright. So the notch is ‘off’ by the difference between the radius to an endpoint of a segment and the radius to a midpoint.
Tradeoff - do you want that distance exact, or the the first segment vertical? If both, I’ll need to think how to do that - probably a very small scaling adjustment to the radius, which might have to be calculated rather than drawn.
I’ve left a few guide lines and points in for your reference.
Arch 10.skp (235.4 KB)
This would be perfect except that the channel dimensions should be reversed - the 21’ is the green axis and the 12’ 6" is in the red axis (at least on the ground plane).
Arch 8c.skp (2.6 MB)
See the way the elevator disappears into the arch channel when it is too small in this previous version of the arch. The elevator will fit in a 21’ wide space easily and the 12’ 6" allows for sufficient “play” in the floor height at the top of the inner arch when the elevator stops near - but not quite at - the midpoint so that the floor will not have to be adjusted to meet the elevator landing level.
OOPS! Sorry, and I should have known better. Will try to redo, soon, and get back to you.
Does the ‘approximately 21’ bother you? And do you care if the first (half) segment is vertical or not quite?
The latest way of constructing it gets the recess dimensions constant all the way round.
It should be 21’ exactly so that the elevator entrance will be within tolerance limits all the way up the channel respective to the Arch floor.
I’m not sure what you mean by the first half segment being vertical or not quite. The depth of the channel does have to be consistently 12’ 6" however, and it doesn’t seem to be that yet. It is quite a challenge…
It’s the difference between left and right in this image (easier to see with fewer segments in the circle but the same principle even with 360sided circle).
I rotated the RH circle by half a segment.
or in close up:
I thought the depth was a constant 12’6 on my Arch 10 (within small rounding error - less than an inch - you get a different result?
I’ll wait to hear back from you before starting again. Will be busy soon for an hour or two this evening,
And by the way, the 21’ will vary slightly between the middle and the ends of each arc segment - though not by much. I have the same number of segments in each arc, but because one is elliptical, the joins will not be exactly radial to the segments.
A further thought. At the moment, the inner face of the channel is sloped, parallel to the outer face, while the upper surface is part of a horizontal cylinder.
So the inner angle at the top is about 78.7degrees (90 - 11.310)
Do you need it to be vertical? How else will the inner face of the elevator match the static structure?
That would be a horse of a different colour, and needs a different method of drawing.
Or will passengers get on and off through doors in the cylindrical face, like the inner elevators?
I got a much lower distance when measuring the depth near the top than 21’ - more like 13.’ The purple color tells me I am perpendicular to the inner curve, though even if I wasn’t it ought to be more, not less anyway.
I am using 180 degrees in my arcs, both for smoothing and calculation purposes.
It ought to be 21’ all the way up, though I realize the actual length of the side leaning closest to the center of the arch will be very slightly shorter than the side of the channel at the outer edge. The two elevator tracks of the longer elevator would have to be adjusted and maybe the gearing ratio too, to accommodate this in real life, but for illustration purposes, they can be the same.
You were measuring an older version of the arch, I think. v10 is parallel all the way, I think.
Does the 21 clearance need to be at the bottom of the elevator, or the top?
External Elevator 2.skp (3.0 MB)
As you can see here, I tilted the tracks relative to the elevator 10 degrees. They will rotate as the elevator ascends, around the central point of the rings, while the cylinder rotates INSIDE the rings. I realize I have a bit more work to do to make the motor blocks realistically rotate about a curved surface beneath them, but the principle is the same.
Arch 10a.skp (229.9 KB)
I was using Arch10 (now 10a here). When I measured the distance at top, it was about half that at the bottom. In any case, the 12’ 6" and 21’ measurements have to be reversed in the channel.
The 21’ is for the length of the elevator, another words, the horizontal, and the 12’ 6" is for front to back at the bottom of the channel, but top to bottom when at the top, and in between on the way up the curve. Since the total diameter of the elevator never changes no matter its position - 9’ 10 3/16" to be precise, including the depth of the concrete mounting on the inner arch itself - the distance is invariant. The reasons I have for making the total depth almost 3’ more are:
- the curve makes the distance greater than if the elevator was traveling along a flat surface
- More importantly, I need some “play” so that I don’t have to land the elevator at the precise pinnacle of the arc in order to meet the 87th floor at 875’ 10" (all floors are 10" thick). By having more “depth” on top, I can stop several feet short, TBD, of the absolute high point, leaving room for an observation deck in between the two stop points of the elevators coming from each leg of the Arch (the elevator on the upriver side will span the entire arch).
Could we Skype briefly? I’ve got about 10-20 mins before I need to go and get supper.
Arch 8c.skp (2.6 MB)
I think this will work. I pasted in a second elevator. The tracks are not angled right. They have to be turned in the opposite direction. But they are placed flat on the plane of the overhead arch.
At the top of the arch, the 21’ “green” plane is flat relative to the ground and so the elevator will not need to tilt, and the tracks out to align parallel to the rings at that point. I anticipate the motor mounts to twist in the opposite direction as the elevator descends on the other leg.
If you want the inner face of the channel to be approximately vertical, it has to be helical - or it would be if the arc were a circle instead of an ellipse.
But it can be scaled to stretch from circular helix to elliptical.
I have a plugin (available on Sketchucation) to drawn helical shapes. Will try it later tonight.
Arch 8c.skp (2.6 MB)
I placed 3 elevators in different positions on the channel and aside from rotating the rings so the tracks would lie flat on the green wall, I didn’t have to change the orientation of the elevator. Of course, the tracks need to be rotated about their pivot points on the rings too, and I didn’t do that in this quick and dirty model. But I can’t see any point where this won’t work, at least with the green wall the way it is here.
I also did a pushpull to push the wall back to apx. 21’ so I could see the placement of both tracks. This is imprecise because of the slant of the wall I was pushing was 11 degrees. But the original green wall allows the tracks to lie flat against it, so I think this is merely a matter of how the wall is created from the start.
I don’t want to complicate the elevator too much as that will make it less reliable, and I think the two rotation points should be sufficient, at least as arch 8c (formerly 8b) was (mostly) drawn. This arch has other problems, so it is a matter of combining the best features of arch 10 and 8b together.
Just working on that at the moment. Will take a while - maybe an hour.
Well, I still think you need to - this is a plan view from the top, in parallel projection, of one of the higher elevators.
And have you doubled up on the end rings? What are the off-square rings?
At the very top, the angle of the tracks is 90° to the elevator long axis. At the bottom, it’s around 11° off square.