Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

3D pipe unwrap to 2D topology

  1. Jul 11, 2012 #1

    I am not very strong in maths, so sorry if these sounds simple. If I have a 3D geometry of a pipe which has its surface defined by triangles (such as that in Computational Fluid Dynamics or Finite Element Analysis) and I have the coordinate points for all the triangles, how can I represent the 3D object into a 2D plane.
    This would be like slicing the pipe through its centre and then unwrapping it into a flat piece. Would I use some kind of mapping? Is this achievable?

  2. jcsd
  3. Jul 13, 2012 #2
    Hey kyze,

    It is possible. The general way to go about it is to find a central axis of the pipe, assign an X,Y,Z coordinate system with Z along the length of the pipe. Then the conversion to flat is based on the radius of the point (vector from 0,0 to point) and the angle (in the form of total circumference).

    So for a point (X,Y,Z), the flat layout would be [(theta*2*PI*radius), Z] where theta is the angle of the vector relative to your coordinate system.
  4. Jul 14, 2012 #3
    Thanks athuss. Great idea. I guess it could also work for a square duct with a bend also? But this all leads to a final geometry which is absolutely irregular, say the carotid blood artery that need to be unwrapped. So do you think applying polar coordinates would work?
  5. Jul 16, 2012 #4
    I guess you could unwrap an irregular geometry. And I'd say the steps that I would take is to take z-slices along the profile, where the z plane is normal to the axis. And the axis is defined by the 'center point' (found by taking the average of the points in X-Y).

    But to make at any of the found 2D points valid you'd have to store a 3D coordinate transform (from say, world to your assigned) for each z-slice. Not sure what the final intent is, but unwrapping a pipe to 2D works to decrease the total information by setting the seam as an axis. In irregular geometry you can't make that assumption and need to store just as much information so staying in 3D might just be the simplest way to go? A square duct is halfway in between, but you would still need to have a packet of information to describe the bend - sort of like 2.5D?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook