Karamba - Simulate varying material property throughout shell

unhandled

(Charles) #1

Hi there,

I’d like to get some ideas regarding calculating the deflection on a shell that has some varying material property.

Specifically a varying percentage of carbon fibre throughout the shell. I know that tensile strength reduces but that Young’s modulus increase along with the percentage of carbon fibre.

Is this feasible with Karamba? If so could you share your thoughts? At this stage, I’m just looking to prove feasibility rather than getting accurate metrics.

Thanks,

Charles

Carbon Fiber (CF), Graphene nano-sheets (GN)

Screen_Shot_2018_07_18_at_11_57_56


(Luchavy) #2

Hi Charles,

As far as I can tell Karamba3D does not yet support orthotropic materials (or any kind of anisotropy), nor does it support parametric materials.

Yet if you stick to isotropic materials, your can discretize your material property variation : you can split the shell in smaller patches (or “elements”) to which you assign different properties. Ideally, if you reduce the size of the elements towards zero, your model will tend to a continuous gradient of structural properties. If you are just looking at global behaviour and not concerned about local values or max/min values this approach should be enough. Don’t trust any local value though as the discontinuities of material properties and the slicing of the mesh will introduce singularities in the FEA analysis.

If Orthotropy is an important concern, you can set up a makeshift solution by creating “stripes” with different material properties, which will give you a very rough first approximation of the global behaviour (again, nothing local).

I’ll recommand going directly to more suitable software for your problem, unless you specifically need the features of grasshopper.

Best regards,
Luc


(Charles) #3

Hi Luc,

Thanks for your input, I can see that the material component offers the ability to define Orthotropic materials. Perhaps you weren’t aware of this?

Further, the Isotropic materials takes as input both in-plane shear modulus and transverse shear modulus. Don’t isotropic materials behave in the same way regardless of the direction of the force vector.

Additionally, Clemens from Karamba pointed out that material properties can defined on a per-face basis simply by defining a equal number of materials per face.

Thanks,
Charles


(Luchavy) #4

Hi Charles,

Sorry for being negative then ! I didn’t know about this feature, I’ll check it out.

EDIT : I could’nt find it even after you mentionned it, it’s a new feature of the WIP version.
Had any good results?

Best,
Luc