film simulation

[cacciatorino]

I am trying to use code-aster to simulate simple articulated systems (e.g. an articulated quadrilateral). the purpose would be to have the tensions and deformations of the members during the trajectory, which software of type "rigid body motion" could not give me.

code-aster gives me meaningless results, but to be honest I have not yet understood whether the fault is mine or code-aster, so I ask fem experts:

finite element software can do this type of analysis even under the hypothesis of Big deformations? or should we always remain in the field of small deformations?

 

[Onda]

The two operations are normally separated.
a kinematic analysis of the system, and in the configuration of maximum load a static analysis of the component with the (even inertial) loads in that particular moment.
Solidworks has a procedure to extract from the cinematic analysis the loads and to load the component in cosmos automatically to perform the analysis.

I believe that ansys workbench has a system to consider flexible components in cinematic analysis, but not using it I don't know how to say much.

with nastran, in my opinion it is not feasible. the analyzed parts must always be in balance and the structure must be at least isostatic but never labile.
you could recreate the structure in its conformation of maximum load, insert a bond instead of the strength resistant to make the isostatic structure, and then analyze it in its overall, so as to have also the total deformations.
wave

 

[cacciatorino]

The two operations are normally separated.
a kinematic analysis of the system, and in the configuration of maximum load a static analysis of the component with the (even inertial) loads in that particular moment.

I imagined something like this, I did with code-aster a simulation of the insertion of a flexible hook on a match, and I can see the deformed and tensions during the motorcycle, but with the cinematisms I still did not bring out anything good.
http://www.youtube.com/watch?v=ypcxc4py6hkfor lability, you actually have to have some force at stake otherwise the energy criterion with which the fem determines the balance does not work (there comes a 0/0 between internal and external forces and he goes into crisis), but the problem you can solve with soft-spring and imposing instead of forces.

 

[Onda]

to simulate bodies in balance but not bound, such as an airplane on nastran there is a particular system called inertia relief.
with this nastran simulation calculates the forces and moments missing to balance and 'spalm' them on all knots.
better than creating a system of weak springs.

you can try in your case to impose a constraint node a shift known to the configuration you intend to study and then apply forces correctly. with the large shifts activated the solutor recalculates the matrix of stiffness in the new configuration dictated by the shift imposed and then calculates the deformed according to the new configuration.

to be clearer, assuming that your quadrilateral is controlled by a piston and that has a load that balances it to the piston effort, you can move the piston node to the desired point and apply the balance load.
repeating the analysis with different points several times you can recreate cinematism and their structural analysis for each point of interest.
just do different load cases, each with the correct constraint and load.

wave

 

[MacGG]

Hello hunter,

if you are interested in multi-body analysis and also struutural, the best way is to use adams where you can insert fe bodies that then in addition to better describing the behavior of the multi-body model, if in phase of generation of the flexible body for adams; also calculate stress, you can see these results directly.

the generation of the descriptive model of the flexible body, you do it with a fe code (msc.nastran, nx nastran, ansys, abaqus, strauss) going to consider it linear (i.e. in its interior, small deformations, linear materials, etc.) but if you use md nastran, you can also consider the nonlinear body (as well as moving degrees from the multi-body).

today the simplest platform for you to turn this is simxpert.

In this way, in addition to the excellent results you get, the overall simulation times (model construction times, visulaization results, etc.) are very low.

Hi.

♪