I am working on designing a tent using Kangaroo and would like to ultimately produce a set of cut patterns for the fabric panels to get the material cut out right when it’s time to sew it all up.
I can get a general kangaroo model of my tent working fine enough but I am having trouble figuring out the nuances of this to get it right and there are a couple of workflow questions I am stuck on:
-
Fabric has a shape when it is not stressed, and a second shape when placed under elastic deformation. Kangaroo is simulating the second stressed state, but what I really need is the first non-stressed shape so I can cut out the pieces from a roll of material. If I model the tent in kangaroo, export the patterns, and then cut out the panels from non-stressed fabric, once the material is placed under tension the fabricated structure won’t match the kangaroo model. Am I understanding this right? what is the solution to this?
-
In order to model the tent physics correctly I need to plug in mesh constraints that mirror the physical material in warp, weft, and bias stretch. I also need to have the mesh oriented correctly on the model in the same way the fibers of the fabric will on the physical tent so that the rules are applied correctly… this means I also need to know the shape of the panels before I plug them into kangaroo, as well as the direction of the mesh on those panels (since they will be shapes cut out of a rectangular mesh grid) I would need to have the end result that I am trying to solve for at the beginning. Again am I understanding this right?
If someone could set me straight on this it would be much appreciated! what is the best way to go about simulating fabric and getting a correct cut pattern at the end?
If your membrane has a cetrain size I would strongly recommend to use a dedicated app for that matter (there’s some available around but the best are used by Birdair and the likes). Alternatively you can collaborate with a membrane maker.
But in real-life the fabric is a very small portion of the puzzle (about 10% to be exact): if your anchors (et al) are not properly designed (and intergrated to the whole Topology from day one) you’ll get a very poor result. You need a solid modeler app for that kind of game (and most probably feature driven modelling). Rhino is NOT the app to use, mind.
Unless of course you are after a very small toy-like membrane where real-life tension parts have rather a very relative meaning. Anyway go for combos of anchors or hinged ones if the plate anchor angle is greater than, say, 50 degrees. Finally for parts and the likes nothing beats Ronstan.
The tent is going to be a four person mountaineering tent, so much much smaller than the type of applications in the photos you attached. Also the loads will be much much smaller by comparison.
The key for me though is getting the pattern just right so when I pitch it up it is smooth and tight like a drum all over. This is a pretty big deal when it comes to wind and weather performance. Also the material has about 2% thermal expansion and contraction to deal with depending on how it is designed which opens up some other cans of worms. If the tent expands and sags then wind and structural performance drops off significantly.
This is why I was asking. I guess i’m having trouble understanding the logic of kangaroo. As I understand it, it basically stretches a reference mesh onto anchor points and such and applies forces and rules. I am trying to figure out how you go from this to fabrication output.
In tensile structure design and engineering we start with 2 main fields, geometry and pre-stress. In Kangaroo the geometry is available but you should materialize it to get pre-stress and uncompensated geometry. For a small structure i would go with a simple approach, cut and flatten the panels and compensate for a small % to reach a 0.5 kn/m stress (or less) in both warp and weft and it should work fine in most cases.
Gd
Gerryark can you elaborate on uncompensated geometry? As I am thinking about it now, it would be the geometry of the tent with all the cut out panels sewn together, but without and tensile stresses applied. So the shape of the thing without any material stretching under tension. Is this correct?
The grasshopper workflow I am envisioning for this then is a two-step kangaroo simulation.
- simulate the tent in kangaroo to a generic shape approximating the final design
- cut and flatten the fabric panels
- re-define meshes for the flattened panels per the material fiber orientation, and define kangaroo rules for warp, weft, bias.
- re-run kangaroo simulation with redefined mesh panels fed back in to arrive at a simulation of the tensioned geometry (starting from the fabric panel shapes)
- if the final result is unsatisfactory, make changes to the original approximated simulated model and check the results.
Kind of a two-part guess and check process, does this sound about right per what you were saying?
You have a stretched 3d model ( think it made of rubber )
We need to build it smaller, mount it and stretch it so it reaches final geometry
The stress field clearly plays a central part in this process, more stress = smaller initial geometry.
Reason why we need a geometry AND a stress field in equilibrium to start with.
I have never played so much with kangaroo on this but I think it should be quite easy to get a stress field
out of the initial form.
Your steps seem ok, a bit tricky will be from step 4, it would require a finite element analysis, instead of kangaroo, we need to prestress a real mesh-material and let it assume the 3d geometry.
Maybe it can be done in kangaroo, I dunno, it’s a non-linear problem quite tricky.
gd
