Changes between Version 7 and Version 8 of Hand/280/MassByJoint

Timestamp:

Jul 31, 2012, 9:35:52 PM (12 years ago)

Author:

dc

Comment:

--

Hand/280/MassByJoint

@@ -9 +9 @@
  1. Inadequately modeled PCB and components
 
-Further investigation narrowed the source of error to subassemblies with dozens of strands of wire, all of which were omitted in the CAD model. To replicate the mass of these wires, slugs of the discrepant mass were created to fill the cavities in which wires could run. The base CAD model used now measured 664 g.
+Further investigation narrowed the source of error to electrical subassemblies with dozens of strands of wire, all of which were omitted in the CAD model. To replicate the mass of these wires, slugs of the missing mass were created to fill the cavities in which wires could run. The base CAD model used in the calculations below measures 664 g.
 
 In the following images, all parts are left transparent to display all parts of the mass calculations. Yellow denotes frame parts, red denotes electrical pieces, and blue is for motors and drivetrain components.
 
-The following frames were established on the hand. These inertial frames follow the same convention as on the BarrettHand BH8-262.
+The following inertial frames are established on the hand. In each of the following diagrams, coordinate systems are oriented with the Z-axis coincident to the axis of rotation of each joint. These inertial frames follow the same convention as on the BarrettHand BH8-262.
 [[Image(AllFrames_crop.jpg)]]
+
+Since no one is likely to wish to attempt computed-torque control on the Hand axes (since the frictions in the Hand are very high), the backdriven, reflected drive inertias of the drives were not computed. Since backdriven drive inertias have no effect on gravity terms, the frame inertias should give you good results for controlling the WAM arm and wrist torques as functions of hand-finger positions.
+