Problem with connects and array components

[Christian Schubert]

Hi,

I tried to write a simple pendulum model in a very dense form

model Pendulum_N_MultiBody
  constant Integer N = 2 "Number of individual elements";
  inner Modelica.Mechanics.MultiBody.World world(enableAnimation=false);
  Modelica.Mechanics.MultiBody.Joints.Revolute revolute[N];
  Modelica.Mechanics.MultiBody.Parts.BodyCylinder bodyCylinder[N](each r={1,0,0});
equation 
  connect(world.frame_b, revolute[1].frame_a);
  connect(revolute.frame_b, bodyCylinder.frame_a);
  connect(bodyCylinder[1:N-1].frame_b, revolute[2:N].frame_a);
end Pendulum_N_MultiBody;

It seems, however, that there is a problem with the connect-statements, as the model is unbalanced. If the second and third connect is written for each individual component, then it works fine:

  connect(revolute[1].frame_b, bodyCylinder[1].frame_a);
  connect(revolute[2].frame_b, bodyCylinder[2].frame_a);
  connect(bodyCylinder[1].frame_b, revolute[2].frame_a);

A closer look at the equations reveals, that omc thinks that none of the connections to bodyCylinder[1] exists as

bodyCylinder[1].frame_a.t[:] = 0.0
bodyCylinder[1].frame_a.f[:] = 0.0
bodyCylinder[1].frame_b.t[:] = 0.0
bodyCylinder[1].frame_b.f[:] = 0.0
bodyCylinder[2].frame_a.t[:] = 0.0
bodyCylinder[2].frame_a.f[:] = 0.0

These quantities are also missing from the flow-balances they are supposed to be in.
which are the superflous equations.

I attached both models:

• Pendulum_N_MultiBody -> Array-Connects
• Pendulum_N_MultiBody -> Flat-Connects

[Per Östlund]

Fixed in r14813. Expansion of crefs didn't handle the case where a part of a qualified cref was a scalar with subscripts, so the connections weren't expanded properly.