Cylinder geometry warping during Piston motion

geniusjimmy92geniusjimmy92 Member
edited September 22 in Fluids

I want to perform a cold-flow analysis in a simple 2-valve IC engine with the 'In-cylinder' option turned off in dynamic mesh. I have defined the piston motion using UDF and applied it to the bottom surface plane of the cylinder. The thing is my geometry is a little different. It is a curved cylinder with angular sinusoidal piston velocity as shown in figure.

2 parts in geometry. Port1 is the upper cylinder head, Port2 is the lower cylinder. I named selection piston, inlet, outlet and cylinder wall. Wall-port1, Wall-port2 and Cylinder are 'deforming' while Piston remains rigid body (UDF). Minimum and Maximum length scales in Remeshing are equal to the initial element mesh size.

Here's the results:

The piston attached to the cylinder edge but the cylinder is compressed because of piston motion and the sides of cylinder are seen warped because of it (as shown in fig) instead of being rigid. I need the cylinder wall to be re-generated (bigger or smaller) at each timestep due to the piston motion instead of being compressed and expanded, just like in a normal straight cylinder engine, the air is compressed but the cylinder wall doesn't, only the cylinder wall interface increases and decreases.

Please help what dynamic mesh settings should i select so that the cylinder end edge moves with the piston with continuous remeshing for the cylinder at each timestep.


Answers

  • kkanadekkanade Forum Coordinator

    You are using tet mesh. So no need for layering.

    Please cross check your udf for giving motion to piston.

    Also please check the dynamic mesh zones defined. Please see help manual for more details about these commands. 

    Regards,

    Keyur

    Guidelines for Posting on Ansys Learning Forum

    How to access ANSYS help links

  • Thank you for your reply.


    I understand now that if I use the tet mesh, I do not need layering option turned on.

    My UDF for piston motion is very simple. Kindly check it out if there's any mistake there. The piston motion works fine, its just the warping of cylinder during piston motion.


    #include "udf.h"

    DEFINE_CG_MOTION(PistonMotion, dt, cg_vel, cg_omega, time, dtime)

    {

    real omega;

    omega = 617*sin(3.14159265/180*time/.000005);

    cg_omega[2] = omega;

    }


    As for the dynamic mesh zones definition, I am facing problem here because I have tried almost every possible combination to go.

    I have tried to keep the cylinder wall 'deforming' but that made the whole cylinder 'running away' along with the piston as can be seen in the fig.

    I have tried to keep the cylinder wall 'stationary' but that made the piston-cylinder joining edge to stretch, deforming the geometry. The cylinder wall remains stationary but piston surface moves causing the 'negative cell volume error' something like that (fig below):


    Please let me know which settings will be suitable so that the cylinder wall interaction with the piston will be usual as in case of straight cylinder engine that, the air is compressed but the cylinder wall doesn't, only the cylinder wall interface increases and decreases according to piston position and piston-cylinder edge.

    Waiting for your kind reply. Thanks!

  • kkanadekkanade Forum Coordinator

    This looks UDF issue.

    You need to check UDF. Please see customization help manual for the same.

    Regards,

    Keyur

    Guidelines for Posting on Ansys Learning Forum

    How to access ANSYS help links

Sign In or Register to comment.