FINDING OUT IDENTICAL POLYSURFACES

Hello Grasshopper users,

Can someone help me with the following problem.

I have a large scale 3D model.
And I need to find out the similar objects.

If one 3d geometry same of another then they must take the same names, else they must have different names.

Based on this lojic i would do fabrication drawings. and i dont want to end up with unnecessary work load in preparation of fabrication drawings.

Your helps are highly appreciated.

Thanks & Best Regards

Replies to This Discussion

Permalink Reply by Daniel González Abalde on February 23, 2016 at 4:39am

Check if all the different pieces have a different number of vertices. If they all have a different number, then it is very easy to group. If some pieces share the same number, see if these have different number of faces.

There would be many ways to evaluate the pieces, perhaps also by a section and adding angles or the amount of angles.

I recommend you upload the necessary geometry to help you.

Permalink Reply by Firat Aydin on February 23, 2016 at 5:30am

hey peter,
thanks for the hands.
At the moment i cannot use any software but rhino. its the only licence i have at the moment.
If you want to test, i can send you the 3D model.

Thanks

Permalink Reply by Firat Aydin on February 23, 2016 at 6:21am

Hi Peter can you accept my friend request please.
It is some difficult to follow the thread here.

Thankks

Permalink Reply by peter fotiadis on February 23, 2016 at 8:34am

Update: found a couple of minutes for the trad update on that one:

1. A known minor bug fixed (related with single objects).

2. Tree contains the "same" objects as well (OK, that was obvious) as follows: in branch {?;3} all the same objects are sampled. Branches {?.0} (the unique object) and {?;1} (how many same found) remain the same as in V2.

3. Text Dot (WOW) stuff is added in order to visual distinguish apples from sardines. Naming is user controlled.

4. Def runs 0.002% faster (that's a phenomenal boost on performance > Nobel is on hand).

5. 6 divisions by zero added as a free bonus (for the brave).

more soon (the V3, that is).

Permalink Reply by Firat Aydin on February 23, 2016 at 12:45am

hi all.
It is actually a steel structure model.
It contains plates and profiles, there are some drill holes and chamfers for weld time to time basically.

I used a software named TEKLA STRUCTURES.

The best thing with it was, it generates a numbering sequance, evaluating if the geometries are interusable.

If there are objects in different locations however at same geometries, then they are labelled same.
At the end the project team does not generate individual fabrication drawings for identical objects.
Im just wondering if someone who is good at programming, can help me with a small script or something like?

Permalink Reply by Raul Matagringo on February 23, 2016 at 1:42am

hi there,

i think it will be the best to use TEKLA and do part name = object name. Doing this you will have named objects in rhino. Comparing objects/solids isnt a small script. TEKLA has a good working algorithm for this.

Permalink Reply by Firat Aydin on February 23, 2016 at 2:38am

I know... but i dont have a licence.

Permalink Reply by peter fotiadis on February 23, 2016 at 12:07pm

Restart from node 0 (depth of answers is rather limited).

V3 does what I said plus more divisions by zero (In the name of Science, what else?)

If you observe what's happening in the tree (branches et all) you'll understand the naming schema used in the text dots:

Part_X refers to the unique thingy (where X is the first dimension branch where the thingy is sampled).

Part_X_Y refers to an Y clone (where Y is the corresponding index in the {X;2} branch - the clones are sampled there)

If there's no clone around ... there's no Part_X_Y text dots (and no {?;2} branches.

BTW: text dots are NOT GH things ... they are Rhino Doc objects.

For the test stuff load R file first.

Attachments:

object_equality_V3.3dm, 611 KB
object_equality_textDot_V3.gh, 122 KB

Permalink Reply by Firat Aydin on February 24, 2016 at 2:24am

Hi Peter,

Thanks for the help. Looks working good.

Permalink Reply by peter fotiadis on February 24, 2016 at 3:10am

Added some stuff more and a display thingy (spot the auto Zoom).

Works (demo data) with the V3 R file.

Attachments:

object_equality_textDot_V3A.gh, 131 KB

Permalink Reply by peter fotiadis on February 24, 2016 at 3:32am

BTW: since you have my e-mail > send some stuff of yours > I want to test some parallel processing approach (rather justified in similar cases - although tricky to implement).

Permalink Reply by peter fotiadis on February 24, 2016 at 4:37am

Yikes + Yikes + Yikes

THIS file received is structured man (what did I said some replies ago? what? you tell me).

Meaning that the only thing that you have to do is to "inquire" nested (or not) instance definitions ("you" is metaphorical > this is only doable via code). This requires 0.0034 milliseconds.

Remind me to kill you the trad way: slowly and painfully (I leave the medium for the job to you [gun. riffle, bazooka, nuke, rope, virus, sword whatever suits you])