SPRINKLER SPACES
@dvir860 wrote:
I want to be able to choose an area not just in square shape, but with more complex ones. Something in the style of the right shape.
....
Speaking as one involved in development of that routine....
I can almost imagine a way to place a grid pattern within an irregular shape, by some approach involving drawing temporary Lines and finding their virtual intersections with the shape, and spacing along the Lines within the range between the intersections. But there is so much in the routine that is based on the size and corners of the rectangle, and on its being a rectangle, that won't translate with an irregular shape, that it could be largely a matter of starting over.
Should the spacings just be specified directly? Presumably they can't be calculated based on the dimensions of the shape, as they are with a rectangle in that routine. Also, where should the spacings be calculated from? Presumably a starting location can't be calculated based on some location on the shape, as that routine does from a lower left corner. Should there be one placed in the middle of the shape's extents, and others spaced from there? I can't imagine that would always result in the best distribution. And it can't use Array, but would need to place each individually. [The "shadow" issue is a whole order of magnitude more difficult to imagine.] Can you provide some realistic examples?
A no-routine-required suggestion, if the spacings should be the same in both directions : Draw a User-defined both-directions Hatch pattern with that spacing in the shape, tweak its Origin for what looks like a good distribution, and Explode it. There are routines you can Search for that will find all intersections between selected objects, which you can use to place sprinklers at the intersections of the Lines resulting from Exploding the Hatch pattern.
I mean that shape:
A sprinkler splashes water during a fire so there should be no area for the sprinkler water to reach. Attached Example (But it doesn't matter to me now to solve this problem):
These are the conditions that the sprinkler set should be (according to US standard nfpa 13):
Maximum distance between sprinkler and sprinkler - 460 cm.
Minimum distance between sprinkler and sprinkler - 180 cm.
Maximum distance between sprinkler and wall - 230 cm.
Maximum area for light hazard - 21 square meters.
Maximum area for ordinary hazard level - 12 square meters.
Maximum area for extra hazard level - 9 square meters.
Possible distances to light hazard:
@dvir860 wrote:
These are the conditions that the sprinkler set should be ....
....
Maximum distance between sprinkler and wall - 230 cm.
....
There seem to be some contradictions. The diagrams show several situations with sprinkler heads at 2.7 m [270 cm] from a wall, which exceeds your 230 cm limit.
Also, are the maximum spacings between sprinklers, and between sprinklers and walls, really the same for all hazard levels, or do they vary as the coverage-area limits do?
Should the locations of heads in an irregular shape like that be in a grid pattern, or could they be in lines paralleling major edges or something? [Not that I have any idea how to automate such a placement pattern....]
And finally, what is your drawing unit? Millimeter? Centimeter?
@Kent1Cooper wrote:
....A no-routine-required suggestion, if the spacings should be the same in both directions : Draw a User-defined both-directions Hatch pattern .... place sprinklers at the intersections of the Lines resulting from Exploding the Hatch pattern.
Here's an example in a shape similar to the one marked, whether actually accomplished by way of a Hatch pattern or by calculation somehow:
Of course, the size of the spacing grid relative to the room is arbitrary, but the same issues could occur at any relationship between room size and sprinkler spacing.
I put the Hatch pattern origin at the middle of the shape's bounding box, and it goes from there. There are some heads ridiculously close to walls, and some areas around the perimeter that are probably not covered adequately. The grid positioning could be moved around in various ways to try to improve on those faults, but "fixing" one issue is likely to cause others elsewhere.
I can imagine that the best you'll be able to do is a routine that gives a first-pass layout similar to this, and you will need to make adjustments.
I tried rotating the pattern to be parallel to the longest edge, thinking that might give a more efficient layout. Given the same size grid, still generated from the middle of the bounding box:
Obviously, the middle of the bounding box is not always going to be the best place to put the Hatch pattern origin. A routine could perhaps be made to put the first row of heads a half-spacing off the longest edge, and go from there, but the positioning in the other direction would be a challenge, and the same kinds of "issues" will no doubt occur along other edges.
@Sea-Haven wrote:
... get bounding box of shape, ... array sprinklers …. get all the inside sprinklers … erase anything [outside] ….
Have a look at image prior post.
That's too simplistic, in my opinion. For example, in that image as drawn, if the sprinkler spacing is at their maximum 460 cm [Message 8], the ones up the middle will be farther from the right-side wall than their maximum 230 cm -- the area along that right edge doesn't have coverage from what would be the next row of sprinklers, as the continuation of that area going downward does, where that next row is present. It's the same issue as the supposition with my images that there are areas around the perimeter that would not be covered adequately.
