Skip to content

Commit

Permalink
docs celling
Browse files Browse the repository at this point in the history
  • Loading branch information
nortikin committed Sep 17, 2024
1 parent 8c23b78 commit 86e8314
Show file tree
Hide file tree
Showing 8 changed files with 302 additions and 223 deletions.
274 changes: 60 additions & 214 deletions docs/tutorials/Celling_00/celling_02.rst
Original file line number Diff line number Diff line change
@@ -1,237 +1,83 @@
***************************************
Celling 02. Preparing geometry for work
***************************************
*********************************
Celling 03. Flattening and offset
*********************************

Subject
~~~~~~~

Importing initial shape (railing lines) for celling, checking of consistency

Workflow
Node Tree
--------

|image0|

1. **Import geometry**:

|image1|

1. **Importing DXF**:
1. ObjectGet from scene - Small area' boundary with incet faces at 0.003 m value;
1.1. Crop mesh (3) with that boundary;
2. ObjectGet from scene - Large area' boundary with incet faces at 0.003 m value;
2.1. Crop mesh (3) with that boundary;
3. Getting Rhomb Pattern manually created;
4. Attempt to create pattern parametrically, it fails only because simplicity of initial pattern to be manual;
5. Separately boundary for between-columns spaces;
6. Lines to manually determinecorner bended plates. It is diagonal for rhomb;


2. **Making surface**:

|image2|

1. ObjectGet from scene - lower and ipper line for surface construction$
2. Making surface. In some cases needed natural paramereisation.

That step convenient to be in math surface, but far there we will go mostly mesh.

3. **Projecting pattern to surface**:

|image3|

1. Raycast on surface main area;
2. Raycast on surface space between columns;
3. Most intriguing part - MANUAL angustment. Look for third part point 4. Dissolve;
4. Joining all geometry to one;
5. Convenient to make such reroute points (NodesMenu->layout->Reroute);
6. Additionallychecking all vertical dimentions to match initial line. We are using manually created lines in scene to adjust that Z-values. Maybe in some cases this checkout should be after flattening stage.

4. **Separating bended corner plates**:

|image4|

1. Raycast pattern on surface corner area (edges);
2. Section matrices to separate left and right plate sides;
3. Select corner faces (not edges);
4. Bisect and separate left and right sides;
5. Join mesh;
6. Drop and flatten left and right sides separately;

5. **Flattening every plate**:

First of all we import DXF. Last 15 years using Blender I did not checked correctness of geometry. Reason for that was usage of blender, mostly for vizualisation. But now checked for production we found that Blender' DXF importer works as shit in circles case. Deviation can achive 200 mm, that is not suits any needs of production ever in the world. So, only way to avoid such dissapointments was to write short importer by myself for only circles to replace original bezier. On vertical it have deviations because of different rails levels. It required little triangulation, that not calculated, but manually esteblished and checked on viewer index node.
6. **Sew corner splitted plates**:

|image2|
7. **Tipisation of plates**:


.. code-block:: python
:caption: Код узла DXF импорта кругов.
'''
in resol s d=128 n=2
in path FP d=[[]] n=0
out vers v d=[[]] n=0
out edges s d=[[]] n=0
'''
8. **Drop plates to layout**:

'''
Принцип работы -
брать данные и в dxf
'''
9. **Testing and gathering**:

10. **Output**:

import bpy



def make():
''' ЗАГОТОВКА ДЛЯ БУДУЩИХ ОТДЕЛЬНЫХ УЗЛОВ
DXF ИМПОРТА.'''
.. |image0| image:: celling_02_images/00_tree.png

def import_dxf(fp,resolution):
import ezdxf
from ezdxf import colors
from ezdxf import units
from ezdxf.tools.standards import setup_dimstyle
from mathutils import Vector
.. |image1| image:: celling_02_images/01_tree.png

dxf = ezdxf.readfile(fp)
lifehack = 50
ran = [i/lifehack for i in range(0,lifehack*360,int((lifehack*360)/resolution))]
#print(ran)
vers = []
edges = []
for a in dxf.query('Arc'):
#a = dxf.query('Arc')[1]
#arc = sverchok.utils.curve.primitives.SvCircle
#arc.to_nurbs()
vers_ = []
for i in a.vertices(ran): # line 43 is 35 in make 24 in import
cen = a.dxf.center.xyz
vers_.append([j/1000 for j,k in zip(i,cen)])
vers.append(vers_)
edges.append([[i,i+1] for i in range(len(vers_)-1)])
edges[-1].append([len(vers_)-1,0])
return [vers], [edges]
.. |image2| image:: celling_02_images/02_tree.png

.. |image3| image:: celling_02_images/03_tree.png

# MAKE DEFINITION BODY HERE #
if self.inputs['path'].is_linked:
fpath_ = self.inputs['path'].sv_get()[0][0]
resol_ = self.inputs['resol'].sv_get()[0][0]
verts_, edges_ = import_dxf(fpath_,resol_)
if self.outputs['vers'].is_linked:
self.outputs['vers'].sv_set(verts_)
if self.outputs['edges'].is_linked:
self.outputs['edges'].sv_set(edges_)
make()
2. **Surface**:

Surface created from two lines with ordered vertices. I naturally reparametrized them only for better view, that is eye-checked quality. Separately created flat part of surface, it is the same as curved.

|image3|


3. **Raycasting on surface**:

Raycasting pattern on surface. For cropping pattern i used other countour - offseted from shape-definition curves. dimensions are defined by farthere offset for plates. So, offset for plates is half of 14 mm, 7 mm. That means, i needed to offset other 7 mm here.

|image4|

4. **Dissolve**:

Dissolve accures with list of edges. Manually iterate all 2500 edges. So, i created edges generator, that choose edges by plate area threshold and boundary analising. Output is flat list in texts, that i need to edit manually. So, need to finish at some stage and not return back for autogenerate list. That means, shape boundary and pattern need to be fixed by design. In practic it was several times changed. After that, i get text with text in node and dissolve it. Than check border elevations (heights) to manually esteblish correct surface shape.

|image5|

|image6|

|image7|


.. code-block:: python
:caption: Код узла нахождения лишних рёбер.
'''
in vers v d=[[]] n=0
in edges s d=[[]] n=0
in pols s d=[[]] n=0
in areas s d=[[]] n=0
in border s d=[[]] n=0
in cutarea s d=0.1 n=2
'''
'''
Принцип работы -
снаружи:
определить площади
определить границы
снутри:
определить граничащие полигоны
определить граничащие рёбра
далее:
вывод индексов рёбер
'''
import bpy
self.make_operator('make')
def ui(self, context, layout):
cb_str = 'node.scriptlite_custom_callback'
layout.operator(cb_str, text='B A K E').cb_name='make'
def make(self, context):
from mathutils import Vector as V
def do_text(out_string):
if not self.name in bpy.data.texts:
bpy.data.texts.new(self.name)
datablock = bpy.data.texts[self.name]
datablock.clear()
datablock.from_string(out_string)
def main_border(vers,edges,pols,areas,border,cutarea):
j = 0 # номер полигона короткого
found = []
used = []
for pol,bor,ar in zip(pols,border,areas):
if ar < cutarea and bor:
fou = [i for i,po in enumerate(pols) if any([e in po for e in pol]) and po != pol]
# i номер полигона большого
for i in fou:
#i = fou[0]
if i not in used:
used.append(i)
# v это искомые индексы вершин, нужны индексы рёбер
v = list(set(pols[i]) & set(pol))
if len(v) < 2: continue
# записать индексы рёбер, в которых совпало два индекса вершин
eds_ = [i for i,e in enumerate(edges) if len(set(v) & set(e))==2]
a = lambda x: (V(vers[edges[x][0]])-V(vers[edges[x][1]])).length
eds = sorted(eds_,key=a)
#print(v,eds_)
found.extend(eds)
j += 1
foundout = sorted(list(set(found)))
#print(foundout[:5],len(foundout))
#edges_out = []
#for i,e in enumerate(edges):
# if i in foundout: edges_out.extend([True])
# else: edges_out.extend([False])
#print([edges_out])
return [foundout]
if self.inputs['vers'].is_linked:
vers = self.inputs['vers'].sv_get()
else: return {'FINISHED'}
if self.inputs['pols'].is_linked:
pols = self.inputs['pols'].sv_get()
else: return {'FINISHED'}
if self.inputs['edges'].is_linked:
edges = self.inputs['edges'].sv_get()
else: return {'FINISHED'}
if self.inputs['areas'].is_linked:
areas = self.inputs['areas'].sv_get()
else: return {'FINISHED'}
if self.inputs['border'].is_linked:
border = self.inputs['border'].sv_get()
else: return {'FINISHED'}
cutarea = self.inputs['cutarea'].sv_get()
if type(cutarea[0][0]) == int:
cutarea = cutarea[0]
#print('\n'.join([str(i) for i in (vers[0][:5],edges[0][:5],pols[0][:5],areas[0][:5],border[0][:5],cutarea[0][0])]))
edges_out = main_border(vers[0],edges[0],pols[0],areas[0],border[0],cutarea[0][0])
do_text(str(edges_out))
return {'FINISHED'}
.. code-block:: python
:caption: Результат полуавтоматического списка индексов рёбер.
[[3, 7, 14, 21, 24, 30, 33, 37, 41, 43, 45, 48, 54, 57, 60, 63, 66, 70, 78, 82, 88, 95, 104, 106,108, 112, 115, 119, 126, 130, 134, 145, 149, 152, 156, 160, 164, 168, 172, 176, 180, 184, 188, 192, 196, 200, 204, 208, 212, 216, 220, 224, 228, 232, 236, 240, 244, 248, 252, 256, 260, 264, 268, 272, 276, 280, 284, 288, 292, 296, 301, 306, 310, 315, 321, 330, 333, 337, 341, 347, 351, 391, 400, 405, 415, 417, 423, 427, 529, 591, 599, 661, 684, 708, 710, 728, 730, 732, 752, 762, 765, 767, 771, 777, 788, 793, 795, 797, 799, 802,805, 808, 810, 814, 816, 818, 822, 826, 830, 834, 838, 842, 846, 850, 854, 858, 862, 866, 870, 874, 878, 882, 886, 890, 894, 898, 902, 906, 910, 914, 918, 922, 926, 930, 934, 938, 942, 946, 950, 957, 965, 969, 975, 981, 987, 996, 999, 1003, 1007, 1013, 1017, 1278, 1298, 1300, 1313, 1315, 1317, 1325, 1337, 1340, 1965, 1966]]
5. **Separate corner plates**:

To avoid jumping plates' corners, happening far from original surface, i created part of triangulated plates. Defined by lines, snapped to projected pattern corners, i filtered plates by lines location, later that lines helped me to define left and right side of plates' triangles, because, farthere offset happened only to two sides. Here I need to say, that all plates are completely flat. So, flattaning happens here in step forward. To flatten triangulated plate I separated triangles, flatted and joined pairs after.

|image8|

|image9|


.. |image1| image:: https://github.com/nortikin/sverchok/assets/5783432/10a4fcef-1eb9-421a-8863-f5057b5e4f84
.. |image2| image:: https://github.com/nortikin/sverchok/assets/5783432/696ec148-69b8-4436-b506-c6eff419e582
.. |image3| image:: https://github.com/nortikin/sverchok/assets/5783432/4ee4f3e7-9cef-4eaa-9aec-13f9d442b579
.. |image4| image:: https://github.com/nortikin/sverchok/assets/5783432/e625e033-cece-4166-8a1d-1789bf4b343f
.. |image5| image:: https://github.com/nortikin/sverchok/assets/5783432/4dfe924a-fca8-4dc8-be28-0f749257a06d
.. |image6| image:: https://github.com/nortikin/sverchok/assets/5783432/289bc089-ec0e-4f56-b0fe-8e9a9445c988
.. |image7| image:: https://github.com/nortikin/sverchok/assets/5783432/a7aeca59-f8dc-4308-98d1-da1a6068af29
.. |image8| image:: https://github.com/nortikin/sverchok/assets/5783432/5e7bea86-a6a6-48a9-9294-24df20655781
.. |image9| image:: https://github.com/nortikin/sverchok/assets/5783432/2b4fde10-8bcf-458f-9683-64959ed12083
.. |image4| image:: celling_02_images/04_tree.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading

0 comments on commit 86e8314

Please sign in to comment.