Stack of Donuts

[JoeStrout]

Permeable Ceiling pointed out that you could make a stack of toruses with windows, and light them all up with one mirror.

This turned out to be a fairly specific problem with the secondary mirrors on torus and barbell parts not shading neighboring parts as they should.  That's fixed now, so this particular exploit won't work in version 0.27.

However you can still make a stack of donuts if you want, by adjusting the hub radius so as to push the secondary mirrors out:




Why would you do this?  I'm not sure.  But you can!

[Permeable Ceiling]

The presented configuration of same size donuts would have different lighting conditions, in the vein of 60%/80%/100%, at least theoretically. Is that addressed in the upcoming version?
Also… Green/Red lights on Docking Port? Wonder what that’s about.

[JoeStrout]

(04-16-2017, 12:29 AM)Permeable Ceiling Wrote: The presented configuration of same size donuts would have different lighting conditions, in the vein of 60%/80%/100%, at least theoretically.

No, they would all have 100%.  (Remember the light rays are parallel; there is no occlusion here.)

EDIT: Just realized one thing that's probably causing some confusion... the light angle in design mode is not the "final" light angle, as deployed in orbit.  Your colony is always built with the spin axis perfectly perpendicular to the light; a torus is always edge-on to the sun.  But we found that when things were lit this way in design mode, it caused visual illusions that made it harder for people to perceive what they were doing.

The one exception, of course, is when you actually care about the lighting, in which case the current solution is deceptive.  It's a trade-off.

Quote:Also… Green/Red lights on Docking Port? Wonder what that’s about.

That's to guide the incoming spaceships.  Or at least, to give the passengers thereof something to focus on while the flight computers guide the incoming spaceships.

[Permeable Ceiling]

(04-19-2017, 04:15 PM)JoeStrout Wrote:

(04-16-2017, 12:29 AM)Permeable Ceiling Wrote: The presented configuration of same size donuts would have different lighting conditions, in the vein of 60%/80%/100%, at least theoretically.

No, they would all have 100%.  (Remember the light rays are parallel; there is no occlusion here.)

EDIT: Just realized one thing that's probably causing some confusion... the light angle in design mode is not the "final" light angle, as deployed in orbit.  Your colony is always built with the spin axis perfectly perpendicular to the light; a torus is always edge-on to the sun.  But we found that when things were lit this way in design mode, it caused visual illusions that made it harder for people to perceive what they were doing.

The one exception, of course, is when you actually care about the lighting, in which case the current solution is deceptive.  It's a trade-off.

I think I understood the lighting correctly, since my percentages weren’t taking into account the angle of lighting that’s displayed in design mode.


[Illustration of three toruses stacked on top of each other; view along the top axis (i.e. from the mirror); Green: stack of three toruses, superimposed; Grey: three rings of mirrors at 45° angle, reflecting light into the ceiling of their respective torus, radius increases with distance from point of view; Canteloupe/Salmon/Tangerine: hubs of increasing radius & spokes]

No, it’s more to do with the area of mirrors compared to area of torus. If the torus area is 1, then the top donut has a mirror area of something like 0.5, middle has 0.7, and the bottom donut of 0.9. So the bottom donut catches the most (of the parallel) light, and mirrors with smaller radius catch less. In this case, lighting percentage should be directly proportional with cross-sectional area the mirrors take up along the axis. 

[Permeable Ceiling]

To be fair, this is more of an engineering challenge, one outside the simulation and illustration purposes of HF.

One could address this problem by increasing the length of the mirror around the median radius (i.e. bottom mirrors would be radius x±20m, medium mirrors radius y±25m, top mirrors radius z±30m, with x>y>z & ‘±’ referring to the ‘inner’ and ‘outer’ radius of the mirror panes). This should increase the otherwise smaller cross-sectional area the torus mirrors would take up compared to their bottom neighbour. Best case the torus mirror cross section area for every torus in the stack would be the same.

[JoeStrout]

You make a good point. I guess I hadn't properly thought this through! The secondary mirrors aren't flat; they are either conic sections or (as illustrated in High Frontier) a series of flat panels. Either way, they take the incoming light and spread it out over the area of the torus. A smaller mirror is spreading less light over the same area, thus less intensity compared to the bigger mirror.

However, for game purposes at least, we assume that there is plenty of sunlight — as long as you're not occluded by other parts of your station, or passing through some nearby planet's shadow, or way out past Earth's orbit where the sun just isn't that bright anymore.

So in the situation illustrated, we'll imagine that the each torus has adjusted the mirror reflectivity, or the window tint, to produce a pleasant "100%" lighting (as long as none of the above conditions apply).