"Floating" an object in Structural FE

[amorrison]

Say you have a basic concept aircraft structural design.

In flying mode various loads will be put on the structure - by the flaps, elevator, rudder etc. which in the real world is taken up by rotational mass inertia of the aircraft structual materials.

FE support reaction member problem.

In an simple FE simulation hard reaction points are used to balance these loads - but this produces high reaction values at these local reaction points and "distorts" the local reaction member stresses.

Any solutions.
Thanks.

 

[jistre]

Just talking out of my posterior, but is the structure such that you can reverse the loads? What popped into my head is to put a distributed load on the support surfaces and fix the system at the control surface. Then, you increase the distributed load until the reaction at the control surface equals the known reaction that will result there.

Of course, it's early and I may not be thinking clearly, but it seems to me that it should work.

 

[GregLocock]

Either distribute the loads by applying pressures to area elements (or forces to many nodes), or apply the loads via an intermediate structure that replicates the theoretical stiffness of the interface.

The first method is the most sensible, but the loads will need to be recalculated for each loadcase.

For instance, with a pneumatic tire running on the road, the easy way to put the load into the wheel is to load each node across an arc at the bottom of the wheel, which varies with location.

The other option would be to model the tire - a lot of work.

Both approaches are used.


Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[rb1957]

it's all just loads and reactions.

in your first case (the airplane, yes, the plane's equilibrium is calculated, but a point on the plane (any point has to be "nailed down" to take out any imbalance loads and to provide the model with a reaction to the 6 rigid body freedoms.

same with your beam. if you don't like the very localised stresses near the hard constraint, then support your model on springs; this'll give the model some flexibility (instead of the infinite stiffness of a hard constraint).