Topic: [request] Radiant Pipe Layout

[JohnK]

My jobs typically have some radiant heating involved (about 7,000 square feet of the buildings) and for the last year or so, I've been doing my designs the old fashion way with excel, charts, and just drafting the layout in AutoCAD. The design is easy, but the layout is very annoying to have to draw and redraw. Does anyone have any code to do this type of layout already in their arsenal?

[ribarm]

@BIGAL states that there are plenty links on www just by google-ing, but AFAIK I have only this link where I was involved...
https://www.cadtutor.net/forum/files/file/47-floor-heating/

HTH.
M.R.

BTW. It is very delicate to code for it, but we tried - I mean me and L.M. (Lee Mac)...

[JohnK]

Oh, that's too bad. I guess I'll just have to chip away at the problem in my spare time then.

BTW, that's a "spiral"; not sure if I'm correct or not but I've always used a sort of "reverse return" layout to hopefully get a more even heat distribution on the panel (aka floor). So, given my luck, my pattern will be impossible to code. :]

[kirby]

Commercial software 'LoopCAD' that includes automated layout, hydraulic & hydronic calcs (and code checks with the Professional version):

Code - HTML5: [Select]

1. https://www.avenir-online.com/AvenirWeb/LoopCAD/LoopCADHome.aspx

It looks like your layout is known as a serpentine, vs. the spiral layout discussed on links.  Serpentine layout should be easier to adapt to L-shaped rooms.

[JohnK]

I just evaluated LoopCAD and put in a purchase request. It was denied because apparently, they would not even discuss "security" and other things with Legal and IT. My request was denied.

[JohnK]

At lunch I wrote some concept code to draw a center line. I calculate the points, but the line drawing portion draws a zig-zag pattern so, my theory needs a little adjustment. Also, I keep typing C code (while (int i=0... so that's a bit annoying. At this rate, I'll get a concept coded up by next month. :]

[kdub_nz]

Hi John,

Do you have a min/max centers and edge-cover for pipes ??

I imagine you'd prefer to maintain the default centers ( 12" in this case ) as much as possible.

For interest, what are the pipe sizes ( assumed dependant on area ) and bend details ( assumed threaded elbows )

added:  ( or 'welded' copper ?? )

added 2: are the floors precast or in-situ slabs ?

 

[BIGAL]

Have another Google Like Ribarm I am sure there were multiple versions provided using different layout methods, finding the key words to use in the search is the hard part. I think even Lee-mac had a go. Its not in my field of interest so never saved.

https://www.cadtutor.net/forum/topic/43738-a-challenge-how-to-draw-this-floor-heating/page/2/

[Lonnie]

I just evaluated LoopCAD and put in a purchase request. It was denied because apparently, they would not even discuss "security" and other things with Legal and IT. My request was denied.

I feel your pain.

[JohnK]

Hi John,

Do you have a min/max centers and edge-cover for pipes ??

I imagine you'd prefer to maintain the default centers ( 12" in this case ) as much as possible.

For interest, what are the pipe sizes ( assumed dependant on area ) and bend details ( assumed threaded elbows )

added:  ( or 'welded' copper ?? )

added 2: are the floors precast or in-situ slabs ?

 

Pipe type is: 5/8 Pex with oxygen barrier. The bend radius they say is something like 7 inches, but I wrote my specs to use bend supports for all that small bend situations (I keep my bends large[er] if I can).

The centerline distance is either 9, 12, 16 or 18. Based on how hot you need the floor (radiant panel).  Here is a snip from my spreadsheet; these are simple calculations/designs the drawing is what is a PITA.

Before I got to finish my first cup of coffee, I fixed my concept code and it now calculates the points in the correct order, so I added the ability to draw a polyline. The concept code doesn't scale (will not handle rectangles at odd angles or different directions but I am trying to find the best way to calculate points so...). My next goal is either to try and think up a way to divide a shape up into rectangles to account for odd shapes or to work out drawing in different directions (vertical/horizontal/left/right). I'll just keep picking away at these problems.

[kirby]

Have been thinking about this because I'm easily side-tracked...

How about using Marko Ribar's routine to create the centreline

Code - HTML5: [Select]

1. https://www.cadtutor.net/forum/topic/48627-filling-up-one-area-with-a-quotorientedquot-polyline/

(but increasing the 'end' offsets by half the pipe spacing)

then another routine to create a double line from the centerline

Code - HTML5: [Select]

1. https://www.cadtutor.net/forum/topic/50819-draw-double-polyline/

Convert Mline to polyline

Code - HTML5: [Select]

1. https://lee-mac.com/mlinetopline.html

Fillet radius

and lastly capping the opposite end from the start point with a semi-circle

[BIGAL]

I dont do heating coils but had a think about it maybe a semi manual approach create parallel lines join ends to make loops left and right, not sure how to get back to start other than offset more on one side so can draw a line back to start. It would be pretty easy to do a close 1 end to all lines using a drag over selection. No code just an idea.

What shapes are we talking about ?

[JohnK]

@kirby,
I'll take a look at links in a bit.

@bigal
Imagine a serpentine pattern but toss in a rectangle (representing a floor drain, maintence pit, elevator, etc) in the middle of your layout so you need to either go around or on both sides while trying to maintain a total length of about 500 feet (you cannot go over 500 feet because that is what a spool of pipe is and you don't want to go too far under either. 500 feet of pipe is one loop.). ...This problem requires several different iterations to get the proper coverage and actual number of loops.

In my spreadsheet screenshot, I used 300 feet per loop because my loop counters are automated and I needed to look at my flows and pressure drops for the number of loops I wanted against pump selections (I was sort of working backwards from a total flow against a pump selection)

I have concept code that will generate a horizontal serpentine pattern in a rectangle already (drawing/calculating a back and forth pattern is easy). The hard part will be addressing the NO-FLY zones and whatnot.  Layouts are fun for one or two loops but when you have 16 it's not fun anymore (it typically takes me about 3 days to layout and draft 16 loops and 4 zones. When the architect starts changing on you, is typically when your boss will ask a question like "why are you so far behind?!").

[JohnK]

Kirby,
I glanced at code in the first link quick, and I just cannot maintain code like this:

Code - Auto/Visual Lisp: [Select]

1. (command "_.pline" (mapcar '+ (polar ip 0.0 ls) (list 0.0 (* r 2.0))) (polar (mapcar '+ (polar ip 0.0 ls) (list 0.0 (* r 2.0))) pi lr) "a" "s" (mapcar '+ (polar (mapcar '+ (polar ip 0.0 ls) (list 0.0 (* r 2.0))) pi lr) (list (- r) r)) (mapcar '+ (polar (mapcar '+ (polar ip 0.0 ls) (list 0.0 (* r 2.0))) pi lr) (list 0.0 (* r 2.0))) "l" (polar (mapcar '+ (polar (mapcar '+ (polar ip 0.0 ls) (list 0.0 (* r 2.0))) pi lr) (list 0.0 (* r 2.0))) 0.0 lr) "a" "s" (mapcar '+ (polar (mapcar '+ (polar (mapcar '+ (polar ip 0.0 ls) (list 0.0 (* r 2.0))) pi lr) (list 0.0 (* r 2.0))) 0.0 lr) (list r r)) (mapcar '+ (polar (mapcar '+ (polar (mapcar '+ (polar ip 0.0 ls) (list 0.0 (* r 2.0))) pi lr) (list 0.0 (* r 2.0))) 0.0 lr) (list 0.0 (* r 2.0))) "")

No offense, Marko.

It would be faster for me to just write my own. Sorry.

[dexus]

This topic isn't about this specific code, but I was intrigued to figure out what this string of code did.
But even with indentation some more variables would have been nice.

Code - Auto/Visual Lisp: [Select]

1. (command
2.   "_.pline" ; Start polyline
3.  
4.   (mapcar '+ ; First point of line segment
5.     (polar ip 0.0 ls)
6.     (list 0.0 (* r 2.0))
7.   )
8.  
9.   (polar ; Second point of line segment
10.     (mapcar '+
11.       (polar ip 0.0 ls)
12.       (list 0.0 (* r 2.0))
13.     )
14.     pi
15.     lr
16.   )
17.  
18.   "a" ; Arc
19.   "s" ; second point of arc
20.   (mapcar '+ ; Second point of arc
21.     (polar
22.       (mapcar '+
23.         (polar ip 0.0 ls)
24.         (list 0.0 (* r 2.0))
25.       )
26.       pi
27.       lr
28.     )
29.     (list (- r) r)
30.   )
31.   (mapcar '+ ; end point of arc
32.     (polar
33.       (mapcar '+
34.         (polar ip 0.0 ls)
35.         (list 0.0 (* r 2.0))
36.       )
37.       pi
38.       lr
39.     )
40.     (list 0.0 (* r 2.0))
41.   )
42.  
43.   "l" ; Line
44.   (polar ; next point of line segment
45.     (mapcar '+
46.       (polar
47.         (mapcar '+
48.           (polar ip 0.0 ls)
49.           (list 0.0 (* r 2.0))
50.         )
51.         pi
52.         lr
53.       )
54.       (list 0.0 (* r 2.0))
55.     )
56.     0.0
57.     lr
58.   )
59.  
60.   "a" ; Arc
61.   "s" ; second point of arc
62.   (mapcar '+ ; second point of arc
63.     (polar
64.       (mapcar '+
65.         (polar
66.           (mapcar '+
67.             (polar ip 0.0 ls)
68.             (list 0.0 (* r 2.0))
69.           )
70.           pi
71.           lr
72.         )
73.         (list 0.0 (* r 2.0))
74.       )
75.       0.0
76.       lr
77.     )
78.     (list r r)
79.   )
80.   (mapcar '+ ; End point of arc
81.     (polar
82.       (mapcar '+
83.         (polar
84.           (mapcar '+
85.             (polar ip 0.0 ls)
86.             (list 0.0 (* r 2.0))
87.           )
88.           pi
89.           lr
90.         )
91.         (list 0.0 (* r 2.0))
92.       )
93.       0.0
94.       lr
95.     )
96.     (list 0.0 (* r 2.0))
97.   )
98.   "" ; End of polyline
99. )