cancel
Showing results for 
Search instead for 
Did you mean: 

Simulate NC file with redundant kinematics

Experimenter
Experimenter

Hi all, 

 

Currently I am finishing the configuration of the kinematic model of a custom machine with redundant axes, having some sort of 4DoF head mounted on the Z-axis of a regular 5-axis machine (C-A-X-Y-Z). Practically the head has two translations (X' and Y') and two rotations (A' and B) and they need to be interpolated with the 5 axis of the machine (or at least with the X,Y,Z whereas A and C could be used initially only as an indexing table). Which is the best approach to define all the axes of the machine in order to able to simulate synchronized motion with respect to a specific NC file? What type of modifications are to be performed regarding the existing MCF/CCF and post processor files of the 5-axis machine to make it work ? Any suggestion will be greatly appreciated. 

 

Best Regards, 

Oliver 

6 REPLIES

Re: Simulate NC file with redundant kinematics

Siemens Phenom Siemens Phenom
Siemens Phenom
Hi Oliver,

I can tell you some points about the simulation.
Before I can tell you what modifications are needed I need some more essential information:

The first question is: How does the NC code look like to move the X, Y, Z etc. axis of the head and the machine or both together?

The initial setting for the machine is to have a channel where the machine axis are included. When mounting the head the axis getting assigned automatically to the channel. Now you should be able to command the available axis. But now we come back to my first question. Which command in the NC code moves the head or the machine axis? Is there an axis coupling? Is there one than more channel available?

Thomas

Re: Simulate NC file with redundant kinematics

Experimenter
Experimenter

Hi Thomas, 

 

Thanks for your reply. My initial idea was to simulate the motion of the axes starting from a regular NC file. That is to say the splitting of the motion between the redundant axes should be managed somehow internally, and not explicitly described in the NC file.  For instance a linear displacement commanded along X in the NC file needs to be split between the X (Gantry ) and X'  (Head) based on a predefined set of rules. Practically, as I see it, a new kinematic transformation embedding all the motion rules needs to be defined and called from the NC file.

Regarding the setting of the axes I have attached my current configuration. I have already defined a first channel and included only 5 axes defined as NC axes.  The reason for that was my initial thought that I have to group the axes  in two distinct channels and synchronize them somehow.

Am I supposed to define all the axes as NC axes and to group all of them under the same channel? Can I handle this configuration by calling existing transformations or is mandatory to define my own custom transformation ? 

 

Regards, 

Oliver 

Re: Simulate NC file with redundant kinematics

Siemens Phenom Siemens Phenom
Siemens Phenom
Hi Oliver,

sorry i need some more information regarding the NC code. With this I can give you the answers you need. If there is only one channel then you have to inlcude all available axis to this. But now it is necessary to know how to command the available axis in NC Code? At the end the simulation must reflect the real machine behavior. So it is essential to know how the real machine behaves with the nc code!? After that we can talk about transformations and the other stuff.

Thomas

Re: Simulate NC file with redundant kinematics

Experimenter
Experimenter

Hi Thomas, 

 

For the sake of simplicity I prepared and attached two files describing a simple linear interpolation to be performed by the translational redundant axes of the machine. The file master_tp contains only the coordinates of the master tool path (see path_section pic) whereas the file axes_tp contains the coordinates of the redundant axes (X and Y are associated with the gantry and X1 and Y1 with the head) after their were split according to the imposed motion rules. The objective is to perform the simulation by considering the master_tp file as input.

 

Regards, 

Oliver 

Re: Simulate NC file with redundant kinematics

Siemens Phenom Siemens Phenom
Siemens Phenom
Hi Oliver,

puuuhh ... interesting project.
I can tell you some of my thoughts but no general approach how this can be handled at the end.

1.) Command 2 axis with one command
This is very easy... you can add the SetJointValue command in a method or an adress as much as you want. There are no limitation. You just have to ensure that all axes are available in the channel.

2.) Motion rules
The key issue what I see is the motion rule. If you have already an idea how you will handle this then I´m very interested in that approach. What I could imagine is, when the X value is executed from master you have to check the sum of X1 and X value in axes_tp.txt. This file could be handled with openFile command and readFromFile. If the value matches you can go ahead with Nr. 1

Thomas

Re: Simulate NC file with redundant kinematics

Experimenter
Experimenter

Hi Thomas, 

 

Thanks for your reply. I will give it a try as per your suggestions. In the meantime I will be thinking about other approaches to handle the motion rule and if successful I will let you know. 

 

Best Regards, 

Oliver 

Learn online





Solution Information