Learning Forum

[wilson1997]

[Mike Pettit]

I guess the right-hand ends of the horizontal links are connected to the same structure, or they're both connected to fixed points? What kinds of joint do you have at each connection in the linkage mechanism?
Aqwa is a rigid body solver, so it is not able to determine the distribution of forces between joints which form a closed loop. A simple example of this: given 2 ships rigidly connected to each other by 2 joints, Aqwa can calculate the total force acting between the ships; but it has no way of knowing how much of the total force acts in each joint, so it just divides the force in half. Aqwa would be able to solve this simple case (with the 'statically indeterminate forces will be distributed' warning) but for more complex linkages the solver may fail (with the 'principal minors are not non-zero' error).
It is also important to avoid redundant rotational freedom constraints. Consider four structures connected in a 2x2 arrangement, with one joint between each adjacent structure: if we wanted to model this in Aqwa, we would not be able to use a hinge for every joint, as the rotational freedoms of some structures would be over-constrained. Again you would expect to see the 'principal minors are not non-zero' error. However, you may be able to simulate this arrangement of structures if some of the joints were ball-and-socket type.
Hope this helps! Mike

[wilson1997]

[wilson1997]

hithank you very much for answering!
it's true that 2 right ends jointing to 2 fixed points, making the 4 bars a crank-and-rocker mechanism. the left end joint the float at center. all the joints are hinged. i want to know the movement of the float under the crank-and- rocker machanism. after reading your answer, i guess it impossible to do it. so i still want to ask how to do this. is it possible to use aqwa user force to read the force from other dynamic software working on machanism and supply the displacement back to dynamic part, making a combined sim?
looking forward to your answer

[Mike Pettit]

Try changing the 2 middle joints into ball-and-socket type. With all joints set as hinges, the rotations of the bars are over-constrained. I have been able to solve this test model with no problem:
Changing the joint types will not change the dynamics of the mechanism, as the bar freedoms are still constrained by the hinges on each side.
Cheers, Mike

[wilson1997]

hi i have tried and succeeded. thank you very much!

[Mike Pettit]

Great, very happy to help.
To answer your other question: you may be able to use the Aqwa user-defined force function to run a co-simulation with another software, but the function outputs are force/moment (rather than displacement). In other words, the rigid body motion is calculated by Aqwa, and the user-defined function can only provide an additional force to the equation of motion (rather than a solution to the equation of motion).
Cheers, Mike