Model (not calculate) beam deflections

Hi,

I'm wondering if anyone has handled modelling beam / girder deflection
in SolidWorks?

I have had reasonable success with simple beams using the 'flex'
function; specifying the flex widget on the neutral axis and playing
with the angle / radius until the right deflection is achieved.

However in this instance I have a series of vertical deflection data
points between support columns, calculated by an external source. In
plan view the beam system is not linear; rather it has a couple of
angles. In elevation the whole beam is cambered upwards in the mid-section.

One approach to the problem of modelling this system would be to
calculate the length alterations and end face angles manually; and then
loft a single beam along a guide curve. However this has a number of
practical drawbacks; most notably that the data is more easily and
appropriately handled as an assembly of sub-beams.

Any clues on how this might be approached? Ideally some sort of
distortion / flexure driven by a control line in an assembly of
sub-beams; acting on each beam..... Eg. it would be nice to model the
beams 'as-built' and then get SolidWorks to show how the deflection
affects the overall shape.

Or perhaps a fresh approach?

Cheers.

 

Sweep the beam section down a spline and put spline points at a
distance that gives a result within your acceptable accuracy. If you
have external data you can import it to a curve.

 

There is no point or curve driven bend in SW that I know of. I would
just use a sweep through a curve driven by the imported deflection
points. You could unbend the beam by creating a configuration where
the data point coordinate values are of the undeformed beam.

 

The other thought that comes to mind is why not use FEA to do this?
Both Cosmos/M and NE Nastran do a nice job of modeling beams. Is it
that you are assessing the fit and function after loads are applied
instead of strength and stress?

 

Correct; I'm assessing the fit, rather than conducting a structural
analysis.

I did consider using FEA. However as usual there are mitigating
circumstances; namely that the data is only in 2-D form and I have a
very simple 3D representation from a survey of the as-built girders.

Having done some hand calcs I do not think that modelling the
deflections will be practical. Rather I'll use the deflection line to
calculate overall change of length in 3D; plus change in angle of each
end section. This I can then model.

Thanks for your input and suggestions. Normally I'm creating the data
from scratch and therefore would have full geometry and run it through
FEA - steel beam so pretty straight forward. However in this instance
I'm doing a 'quickie' and only have access to other's (awful) 2D data.

Cheers.