Hi community! I have been encountering so many problems while trying to recreate a surface following the steps from an article. From what they have mentioned in the article, the used grasshoper for parametric modeling, they have also used control points + nurbs curves to specify curves at each plane.
Where I seem to struggle the most is on creating this “central spine”, I’m not sure what the logic behind is and how to use it, from there I’m stucked, not even sure what command to use to make the surface, some suggestions I’ve read: loft, networksurface, rail, spline (none of these have worked so far).
Please I need help recreating this. So far I have: 3 curves for each plane (not sure if there is a rule to position/constraint them), central spine (which I’m not sure its working).
Parametric modeling of chiton girdle scales. We developed a
parametric geometrical model to reproduce the observed morphometrics of chiton scales to facilitate the subsequent modeling
of chiton-inspired scaled arrays. First, three principal sections
were made through the scale: one horizontal section through the
base (denoted as BASE) and two vertical sections running
transversely and longitudinally in the (YZ) and (XZ) planes,
respectively, both passing through the geometric center of the
scale (Fig. 5a, top). We then selected 20 spatial markers on these
principal sections. Using third-order polynomial interpolation,
we generated a spline through each set of points, creating parametric versions of the three principal curves (Fig. 5b; see Supplementary Fig. 3 and Supplementary Table 1 for more details).
We then generated a central spine running medially within the
YZ cross-section (Fig. 5a, bottom, red line). Constructing planes
normal to the spine along its length permitted the reconstruction
of the complete scale surface. Parametrically controlling the
relative positions of spatial markers along the principle curves
enabled variation of scale geometries across a large design space
(see Supplementary Video). Figure 5c illustrates the modeled
geometry for R. canariensis, which agrees well with the geometry
obtained from the native μ-CT data. The doubly curved geometry
that characterizes the L. mertensii scales could also be replicated
using the same parametric model (Fig. 5d), highlighting the
morphological diversity that can be accommodated with this
approach
The paper is contradictory. It says that, “three principal sections were made through the scale: one horizontal section through the base (denoted as BASE) and two vertical sections running
transversely and longitudinally in the (YZ) and (XZ) planes, respectively, both passing through the geometric center of the scale” and yet the supposed XZ “section” is not planar.
To me this looks like a very basic surface sweep… then again the XZ “section” could actually be the elevation profile and Curve2View could be useful? I dunno. Is there a reason you want this parametric, (other than for the reason that is says the form was “parametrically created” in the paper?)
On the file attached, I have replicated a similar geometry as in the paper (similar concept of 3 surfaces and central spine). From a previous comment, I learnt about the Curve2View command, which is what im using on my image below, however, results are not organic… I’m not sure what am I missing here, my lines are not closing as they should (I have added in red how they are supposed to look like), from some reason I’m not able to recreate it.
If I were to use surface sweep, could you please share any suggestions on how to group the 3 planes? and how to manipulate the central spine? I have watched some suggested videos for sweep command, but no sucess so far.
I need this to be parametrically so I can qucily iterate on multiple dimensions based on experimental data collected.
What we need clarified here is: are you looking to take existing 2D views and replicate a 3D object as, seemingly, in the paper they are doing, or are you making up the geometry to be “like” the chiton scales of the paper. In the former, Curve2View is useful. In the latter, it doesn’t matter, because you can just invent an arbitrary spine with widths at curve paramaters, and then with the spine, spine widths and a base profile, create a 3D form. With Sweep, or Network Surface or otherwise.
I’m trying to follow the same methodology from the paper (creating each section per plane and a central spine) but using my own parameters and dimensions (as described in the attached file earlier). Hope this makes it clearer. I will try surface network, I’m a new user, so I need to review the “rules" for this command. i really appreciate your response!
I’m sorry, but that does not answer my question. Do you have a physical object you can hold in your hands that you currently trying to replicate digitally, or are you starting from scratch on the computer?
I’m aware my explanation might not be the best without an image for reference… I dont have a physical sample so far or any images. I’m staring from scratch on the computer, trying to replicate a similar-looking shape as the one referenced by the chiton paper.
