[overture] Re: rigid body motion

From: Bill Henshaw <wdhenshaw@xxxxxxxxx>
To: overture@xxxxxxxxxxxxx
Date: Fri, 14 Aug 2015 07:54:38 -0400

Junji,
I think that even in the two-drop case the constraint is not doing
what you think it should.

For the "savonius" case you should start with a simpler version of the
geometry where the
grid spacing is not so small so that you can see some motion, first
with no constraints.

I would not bother with parallel yet. This will only make your life
more difficult.

..Bill

On Thu, Aug 13, 2015 at 8:36 PM, Junji Huang <jhyt7@xxxxxxx> wrote:

Hi Bill,

I checked the twoDrop example and modified the grid and cgins command file
according to my problem. The grid is now shown as:
[image: Inline image 1]

with gravity set as 0, and velocity in -y direction in this case, and I
gave rigid body motion as:
specify grids to move
rigid body
density
$bodyDensity
moments of inertia
0.5
initial centre of mass
0. 0.
initial angular velocity
0.
position is constrained to a line
0 0 1
done
drop
drop2
done

Grid spacing is large in the case, so I get large time step and a quick
result. This rigid body system rotate clockwise and converge in the
vertical line. The constraint method works in this situation. For
'savonius' problem, I get small grid spacing at blade tip, so time step is
too small for a successful test. I decide to install the parallel solver to
reduce waiting time. In fact, I tried installing parallel solver two months
ago but failed(
https://www.freelists.org/post/overture/parallel-Overture-install,2). Is
there any instruction for v26 parallel installation? Thanks.

Junji

On Thu, Aug 13, 2015 at 1:33 PM, Bill Henshaw <wdhenshaw@xxxxxxxxx> wrote:

Hi Junji,

The first thing to check is that the bodies move if there is no
constraint.
You may also want to increase the density to start with to avoid added
mass instabilities.

It is likely the constraint does not do what you think it does. You
may need to change
the formula for the constraint in RigidBodyMotion.C, function
applyConstraints.
I have not used constraints very much yet.

Regards,
Bill

On Thu, Aug 13, 2015 at 12:48 PM, Junji Huang <jhyt7@xxxxxxx> wrote:

Hi Bill,

I tried to set up a case of rigid body(two savonius blades) motion that
is constrained to rotate about a point being driven by freestream flow, the
commands for moving body is as follows:

specify grids to move
rigid body
density
1.
moments of inertia
0.1
initial centre of mass
0. 0.
initial angular velocity
0.
position is constrained to a line
0 0 1
done
bottom-savonius
top-savonius
done

For these settings, I want the top and bottom blades rotate about the
point (0., 0.). However, the blades didn't seem to move. I have checked the
torque using aero.C, it produced a negative value that should make the
blade rotate. Would you give me some suggestions on this kind of problem?
Thanks.

[image: Inline image 1]

References:
- [overture] rigid body motion
  - From: Junji Huang
- [overture] Re: rigid body motion
  - From: Bill Henshaw
- [overture] Re: rigid body motion
  - From: Junji Huang

[overture] Re: rigid body motion

Other related posts: