Currently, the members have different length. Is it possible to have same member size. Meanwhile. minimize the deformation of the original model?
The BIG challenge for trusses (of type W like yours or some other) on double curvature Surfaces is the required clash checks between real-life members (i.e. NOT axis … but cones, sleeves, tubes and/or bolts). These are carried over solely via trigonometry (using obviously classic Connectivity [ VV, EV, VE ] because any truss is in fact a Graph).
For instance for this “linear” rnd truss the reds indicate nodes and angles that yield clash events.
Other than that makers (like MERO) charge the same regardless the tube size diffs since the process is 100% automated (and waste material is more or less not a consideration [VS the cost of the truss … where preparing/drilling the balls is about 70% of the whole])
Are you using some sort of plug in for that model? I am using lunchbox.
It is good to know the clash checks which I have thought of how to solve the issue.
However, the model demonstrate is for building a physical one with paper straw and eyelet to hold them together which there will have some sort of overlapping between paper so I dont think clash checks is the main issue in this circumstance.
I see there is a min length in the input para. Can I strictly control the length instead of length in range while keeping the model relatively close to original one? Varies in length is not efficient for making model.
This is 100% C# code. The scope is to feed BIM related aps (like AECOSim) with Instance Definitions and their Connectivity. Any similar C# that I write targets solely AEC oriented real-life trusses (in most of cases MERO TSK Type KK).
Indeed a paper model is NOT the sort of thing for similar checks.
BTW the min Length is related with min node to node distance - obviously an unknown quantity if the nodes are random (and the truss/Graph is done via Proximity and the likes). Imagine a random result like this …
… where nothing is known. So for any solution/variation you should fix issues on the fly (on a per node basis) etc etc.
PS: You can attempt some sort of Length/Angle Goal “optimization” via K2 (leave enough space for expansion - if the relaxation yields an expansion).
PS: Another option is to thicken (result: a closed Mesh) your Graph using some add-on (ExoW/Dendro etc) and then 3d Print the whole Model “as a single unit”.
do you have any suggestion of how can I make the all member size restrict to certain parameter while keeping a relatively similar shape?
I guess it is quite a difficult definition to work on, which i dont have any clue on that.
holy mary, I thought what I was trying to do is actually impossible right?
the only way to have the regular member size is a 2d curve
Here come to a more practical problem. Can I have all the member in integer length while maintaining a relative similar shape?
Right.
But hope dies last: try K2 (but what is the meaning of this in real-life?)
K2? what is that
Can you show me how to do it exactly?
I have searched for a while they are all mesh.
Though some file use the component of equal length but the length wasnt really equal
Space truss Structure.gh (23.6 KB)
Do you want the Edges around A to have the same length as the Edges at B without deformation of the shape?
isnt that impossible? cause if the A and B unroll the length difference should be huge. But yes I do want. if it is impossible to do so, I prefer to have edges that are in integer. btw it is space truss oriented not srf.