• Registered: 10-02-2009
  • Posts: 79

Topic: Bounday condition and reaction output

Dear Borek,

I got confused by the boundary conditions and the corresponding reaction output. In total, there are three types of boundary conditions and two load time functions:

BoundaryCondition 1 loadTimeFunction 1 prescribedvalue 0.0
nodalload 2 loadtimefunction 1 components 3 0 0.0 -0.884
BoundaryCondition 3 loadTimeFunction 2 prescribedvalue 1.0
ConstantFunction 1 f(t) 1.0
PiecewiseLinFunction 2 nPoints 2 t 2 0.0 15.0 f(t) 2 0.0 1.5
With BC1, I want to apply zero displacement constraints in x, y and z directions. With BC2, I want to apply some nodal loads in z direction and keetp them constant. With BC3, I want to apply a ramp displacement in x direction with t=0~15. Deltat=1.0.

In the output file, the reactions are described. One example is shown as:

Node        1 iDof  1 reaction  6.4238e-01    [bc-id: 1]

My understanding is that bc-id=1 is corresponding to BC1. iDof=1 is in x direction.

when I summarized the reactions with bc-id=1 and idof=3, I got confused. I expect it to be the forces I applied with BC2 because that is the only force I applied and the ramp displacement is in x direction which caused bending, but it is not!

My case is similar to ■Simple bending of cantilever beam with plasticity in the gallery on oofem website. Only difference is I applied an axial load along the length of the beam. And I want to see whether the reaction is the same as the applied force.

I doubt that the load time function #1 accumulates the load with the increasing of time to 15 instead of keep it constant.

Thanks,
Xuejian

2 Reply by Mikael Öhman (edited by Mikael Öhman 15-07-2011 20:01:23)

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Bounday condition and reaction output

You say you expected it to be the forces applied with BC2, and that assumption is just false.  These are reaction forces, which arise from the external loading as well as the force displacements. Even without BC2, there will be reactionforces on all prescribed parts for this beam.

  • Registered: 10-02-2009
  • Posts: 79

Re: Bounday condition and reaction output

Thanks for the reply.

My case is similar to a cantilever beam with one end fixed in all directions and the other end free with applied displacement in x direction and axial load in z direction. The displacement in x direction only caused moment at the fixed end with equal tension and compression forces in z direction. Slight difference could happen for nonlinear static analysis, but should be small.

Xuejian

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Bounday condition and reaction output

I don't see why the sum of the reaction forces for bc-id=1 and idof=3 should  reflect the external load at all.
It would correspond to the total force applied on the "free" end in the x-direction, unaffected by your load in z.

  • Registered: 10-02-2009
  • Posts: 79

Re: Bounday condition and reaction output

Dear Mikael,

I am still unable to clear out the problem. Why did I get zero when summing up all the reaction forces? Thanks a lot.

Xuejian

  • Registered: 10-02-2009
  • Posts: 79

Re: Bounday condition and reaction output

Sorry, I just forgot to mention I applied the displacement control. I expected the reaction forces at the boundary conditions. Is there any problems with it?

Thanks,

Xuejian

7 Reply by bp

  • bp
  • Administrator
  • Offline
  • From: Prague, Czech Rep.
  • Registered: 02-03-2007
  • Posts: 1,047

Re: Bounday condition and reaction output

Dear Xuejian,
the prescribed displacement control can be applied in different ways, either by directly applying boundary conditions or indirectly. In the first case, the reaction forces are printed in reaction section of output, if applied indirectly in the solver (nrsolver) then they are reported independently.
Also in your first post, you express doubts that the load time function #1 accumulates the load with the increasing of time to 15 instead of keep it constant. Please see the "refloadmode" solver parameter (http://www.oofem.org/resources/doc/oofe … ode16.html).

Also when checking equilibrium, it makes sense only to sum up all forces (moments) in the same directions as well as reactions in the same directions. In this case, the bc-id is not relevant, just sum up all reactions forces in x-direction.
Borek

bp's Website

  • Registered: 10-02-2009
  • Posts: 79

Re: Bounday condition and reaction output

Dear Borek,

Thank you very much for the response. It seems I solved the problem and got pretty reasonable results. Thanks,

Xuejian