Hello,
I am modelling a building made of very thin walls (10cm) reinforced by a single layer of welded reinforcements (D133: 15x15x5mm). Some of them can be relatively long (longest 5.7m), so they have a fairly slim appearance, and there is no real difference between cover and core concrete (of course no boundary elements either). When I get to the description of element classes, I notice that the fibers, notably in the cover concrete, have a quite high aspect ratio.
My question is the following:
I know that when modelling shell elements in FEM, it is adivsable to maintain the aspect ratio of the elements under a certain threshold for convergence (and continuity) reasons. Should the same principle be applied here? (I don't know exactly how those fibers work). In which case, what do you think would be the less computationally costly procedure? I thought about changing the value of the concrete cover thickness in the project settings. Would it create a problem if I set the concrete cover thickness to zero?
Thank you in advance for your answers
Matteo
Fiber's aspect ratio
Re: Fiber's aspect ratio
Matteo,
Setting the cover to zero should not create any problems.
The aspect ratio of the fibres also does not necessarily constitute a problem per se (section fibres discretisation and finite element discretisation are indeed different things).
It all depends on the strain profile in your section (e.g. if the strains were constant throughout the section, a single fibre for the whole section would be enough).
In any case, and as usual, if you carry out a sensitivity study that considers different modelling assumptions, you should be able to appreciate what matters and what doesn't.
Rui
Setting the cover to zero should not create any problems.
The aspect ratio of the fibres also does not necessarily constitute a problem per se (section fibres discretisation and finite element discretisation are indeed different things).
It all depends on the strain profile in your section (e.g. if the strains were constant throughout the section, a single fibre for the whole section would be enough).
In any case, and as usual, if you carry out a sensitivity study that considers different modelling assumptions, you should be able to appreciate what matters and what doesn't.
Rui