How to correct the physical data of CAD models from Robotis
How to correct the physical data of CAD models from Robotis
by petercohen
Tue Apr 16, 2013 2:01 am
Sorry for the cross-posting. I did not get any reply from the DARwIn forum.
Hi, I downloaded the CAD data from rsourceforge
http://sourceforge.net/projects/darwino ... p/download
How come the weight of the parts are different from the real ones? Looks like Robotis did not set the material type. From SolidWorks, one could set the materials. Which materials should I choose for the frames and assemblies such as MX-28 (made of different materials) and frames+motor? Can I assume that even the weight is wrong, the center of mass of the part are correct?
Hi, I downloaded the CAD data from rsourceforge
http://sourceforge.net/projects/darwino ... p/download
How come the weight of the parts are different from the real ones? Looks like Robotis did not set the material type. From SolidWorks, one could set the materials. Which materials should I choose for the frames and assemblies such as MX-28 (made of different materials) and frames+motor? Can I assume that even the weight is wrong, the center of mass of the part are correct?
by Fritzoid
Tue Apr 16, 2013 1:17 pm
Hi petercohen,
I ran into the same problems with the CAD files for the Bioloid premium. My solution was to weigh each part then create a custom material that replicated the actual result. In doing so I found that even two otherwise identical parts can have slightly different masses due to the vagaries of the injection molding process.
Normally, the center of mass is calculated from the object's geometry and the software assumes that the material is homogeneously distributed through the part. A change in material does not affect this calculation so the Robotis values should be correct for the simple parts.
The servos are a special case. Here the mass is not distributed evenly and you will have to determine the actual COM by experimental means. Most of the weight comes from the motor so the COM will be near the center of the motor.
I ran into the same problems with the CAD files for the Bioloid premium. My solution was to weigh each part then create a custom material that replicated the actual result. In doing so I found that even two otherwise identical parts can have slightly different masses due to the vagaries of the injection molding process.
Normally, the center of mass is calculated from the object's geometry and the software assumes that the material is homogeneously distributed through the part. A change in material does not affect this calculation so the Robotis values should be correct for the simple parts.
The servos are a special case. Here the mass is not distributed evenly and you will have to determine the actual COM by experimental means. Most of the weight comes from the motor so the COM will be near the center of the motor.
by petercohen
Wed Apr 17, 2013 1:36 pm
Fritzoid wrote:Hi petercohen,
I ran into the same problems with the CAD files for the Bioloid premium. My solution was to weigh each part then create a custom material that replicated the actual result. In doing so I found that even two otherwise identical parts can have slightly different masses due to the vagaries of the injection molding process.
Normally, the center of mass is calculated from the object's geometry and the software assumes that the material is homogeneously distributed through the part. A change in material does not affect this calculation so the Robotis values should be correct for the simple parts.
The servos are a special case. Here the mass is not distributed evenly and you will have to determine the actual COM by experimental means. Most of the weight comes from the motor so the COM will be near the center of the motor.
Thanks Fritzoid for the suggestion. It will mean I have to take apart the entire robot. Since it is open source, why Robotis does not give us the correct data?
For determining the actual COM by experimental means, do you mean hanging the motor by a string in different orientations three times and get the intersection point?
by TiagoR
Thu Apr 18, 2013 10:50 am
Hello all
Robotis has available some drawings with inertial parameters for their servos that seem to be correct.
http://www.robotis.com/xe/download_en/26324 after the CAD files you have DXL INERTIA section with a PDF for each servo. Like this one for example: http://www.robotis.com/view/DXL-INERTIA ... NERTIA.pdf
In Solid Works you can change manually on the Mass Properties tab, the mass and center of mass of the object, but I'm not sure if it recalculates the inertia matrix or not.
I hope this helps.
Best regards
Robotis has available some drawings with inertial parameters for their servos that seem to be correct.
http://www.robotis.com/xe/download_en/26324 after the CAD files you have DXL INERTIA section with a PDF for each servo. Like this one for example: http://www.robotis.com/view/DXL-INERTIA ... NERTIA.pdf
In Solid Works you can change manually on the Mass Properties tab, the mass and center of mass of the object, but I'm not sure if it recalculates the inertia matrix or not.
I hope this helps.
Best regards
by petercohen
Thu Apr 18, 2013 12:47 pm
TiagoR wrote:Hello all
Robotis has available some drawings with inertial parameters for their servos that seem to be correct.
http://www.robotis.com/xe/download_en/26324 after the CAD files you have DXL INERTIA section with a PDF for each servo. Like this one for example: http://www.robotis.com/view/DXL-INERTIA ... NERTIA.pdf
In Solid Works you can change manually on the Mass Properties tab, the mass and center of mass of the object, but I'm not sure if it recalculates the inertia matrix or not.
I hope this helps.
Best regards
Thanks TiagoR. Don't know why Robotis removed the pdf, STEP and IGES for the MX-28 series. Anybody has the files? Not sure if the center of mass of the RX-28 is very similar to that of the MX-28. Any idea?
