A slight tweak to my code for the Twisted Drop, with a stretch and a Boolean merge with a cone to give a trunk and remove some of the weirdness inside the tree.
Printer:
Eaglemoss Vector3
Rafts:
Yes
Supports:
Yes
Resolution:
0.2 mm
Infill:
20%
I started with some code I give my students to help them start 3D graphics projects. This helps them create models in Vertex and Index Buffers programmatically. I then grabbed the Marching Cubes function from here, and used it so that I could feed in an equation and get out a Vertex Buffer containing a given isosurface of that equation, display it and then spits out a basic obj file.
My wife commented that the twisted drop looked a bit like a Christmas tree, so I slightly tweaked the code for it to produce this.
The model is in a cubic region 2 Pi to a side and centred at the origin, with a grid of values 100 x 100 x 100 and the isosurface value is zero.
This left a few odd bits of other geometry junk in that region, so when I pulled it in to Netfabb to stitch it all together I also removed all of them apart from this piece which I also slightly stretched.
The geometry got a bit strung out and tortured on the inside of the "tree", so I Boolean added a plain cone to cover it up and act as a trunk
This is the code which calculates the function / equation value for a given point in space
float VBMarchCubes::function(Vector3 _pos)
{
_pos *= 2.0;
float w = _pos.z *1.2f;
float x = (_pos.x *cos(w) - _pos.y * sin(w))/(0.25*(_pos.z+XM_PI)), y = (_pos.x*sin(w)+_pos.y*cos(w)) /(0.25* (_pos.z + XM_PI)), z = _pos.z*cos(w)*cos(w);
float x1 = _pos.x, y1 = _pos.y, z1 = _pos.z;
return ((x + 1)*(x + 1) + (y + 1)*(y + 1) + (z - 2)*(z - 2) - XM_PIDIV2*XM_PIDIV2)*((x - 1)*(x - 1) + (y - 1)*(y - 1) + (z - 2)*(z - 2) - XM_PIDIV2*XM_PIDIV2)
*((x - 1)*(x - 1) + (y + 1)*(y + 1) + (z - 1)*(z - 1) - XM_PIDIV2*XM_PIDIV2)*((x + 1)*(x + 1) + (y - 1)*(y - 1) + (z - 1)*(z - 1) - XM_PIDIV2*XM_PIDIV2)
*((x + 1)*(x + 1) + (y + 1)*(y + 1) + (z + 1)*(z + 1) - XM_PIDIV2*XM_PIDIV2)*((x - 1)*(x - 1) + (y - 1)*(y - 1) + (z + 1)*(z + 1) - XM_PIDIV2*XM_PIDIV2)
*((x1*x1 / (XM_PI*XM_PI) + y1*y1 / (XM_PI*XM_PI) + (z1 - 4)*(z1 - 4) / (XM_PIDIV4*XM_PIDIV4)) - 1);
}
