Hello everyone,
I would like to know whether it is possible to have a varying young modulus, and/or shear modulus along a beam. And for a varying cross-section ?
If yes, how does it work ?
Thank you in advance for your consideration,
Regards
Hey,
a varying cross section will be possible in the next release. The Young’s modulus cannot be changed.
You could try if the changing cross section already works in the old release with the additional input text but no guarantee that it will work (correctly):
EL-PROP 1 : BEAM_3D_NURBS_KLRod
MAT= EL-MAT 1
AREA = 0.02
IY = 6.66666666666667E-05
IZ = 1.66666666666667E-05
IT = 4.5775E-05
PRESTRESS = 0
PRESTRESS_BEND_1 = 0
PRESTRESS_BEND_2 = 0
PRESTRESS_TOR = 0
INT_TYPE_BEAM_3D_NURBS_KLRod = FULL
!U n1 n2 n3
AXIS 0 0 0 -1
AXIS 0.25 0 0 -1
USE_VAR_CROSSSECTION = 1
!U r h w
VAR_HEIGHTWIDTH 0 0.2 0.1
VAR_HEIGHTWIDTH 0.25 0.5 0.1
!==================================
The only other way is to split the curve and assign different cross section / E to the subcurves. However, if you split the curve in many small elements, you have to take care of the penalty factor for the coupling as this acts like a spring inbetween the elements. It has to be larger than your beam/cable stiffness.
I have a problem where I am trying to do exactly this by splitting the curves into subcurves. How / where can you input and control the penalty factor for the coupling as currently deflections are significantly higher than 1st principles calcs, and are a function of number of elements…
hey,
the global penalty factor can be changed in the analysis model compont. Right-click-> penalty factor.
In case that you only want to change one penalty factor locally, you have to duplicate these lines in the input text and set a new id:
EL-PROP 32 : NURBS_BREP_POINT
BREP_TYPE = COUPLING_DISP_ROT
BL_FACTOR_DISP = 1.0000E07
BL_FACTOR_ROT = 1.0000E07
and change the id in the respective line here
DE-BREP-EL 17 1 32 DE-EL 6 DE-EL 26 TOL=0.0001 [0.000001907309][0.000000000000] !Coupling-1D: Crv|Crv
Great thank you!
Setting the global penalty to approximately the EI of the sections locally does give fairly good correlation to the closed form solution of a uniform beam, however might not be accurate for the highly tapered beam (as you hinted!). There also appears to be a dependency on setting ‘strong supports’ on the SupPt input initially for this to take effect - but perhaps that is another topic?
Regarding the solution for the local penalty, apologies for my ignorance but where am I to input these lines? Would you have a simple example available?
Thanks again!
The strong supports should in general result in better results as no penalty factor is used. However, they can only be applied to curve ends.
There is a component called DeconstructModel. It returns the input text for the solver (Also in the file NAMEANALYSIS.txt next to your GH file). This text can be manipulated and returned to the model by the component ModifyModel (category utilities). It is important to know that once you have this component on the canvas, the file will be modified, also if the component is not attached to the solver component.