- June 18, 2014
- Force Group Automation
April 25, 2014 - Force Robots introduces Touch Robot that is able to perform precision grinding and machining to polish, deburr, and deflash cast and forged parts. Metal finishing with hand tools is a dirty and difficult job that carries known repetitive stress injury risk. It is also a process that has proven extremely difficult to automate. Force Robots' system combines the precision of a machine with the finesse of the human hand. It can “feel” existing part contours, match it to a CAD reference, and autonomously work to remove material to specification. Consolidated Precision Products' Cleveland foundry is first to use it to remove excess material on difficult-to-reach areas of tough, precision-cast turbine engine components. The Touch Robot is a compact, self-contained system requiring only 120VAC and shop air to operate. It consists of a 4-axis material removal arm and a 2-axis part positioner mounted to a 1.2m x .8 m portable work table. Dividing the system's 6 degrees of freedom between the two coordinated mechanisms preserves the deft touch of the tool arm while permitting manipulation of heavy castings up to 0.4 m long. It has a simple and rugged construction suitable for a finishing department environment. No fragile force or optical sensing is required to perform its work: all needed feedback is derived from the motor encoders. Its inherent safety features allow it to be deployed along side manual metal finishing cells. Joint limits restrict its .5 m reach largely to the work table on which it is mounted. Low motor supply voltage (24Vdc) keeps joint speeds (max 300 deg/sec) below where the arm would develop dangerous levels of energy. With modest force capacity and low-friction, back-drivable joints, it is easily overpowered by a human.
Turbine blade finishing presents a particular challenge because the parts require perfect smoothness on the micro level, while tolerating the macro-level dimensional variations of the casting process. That, plus the lack of easily referenced datum surfaces for locating, rules out machining to fixed coordinates. The Touch Robot begins operation by locating the workpiece with contact measurements, using the material removal tool as the probe. Machining and grinding passes are then performed with a well controlled, optimal contact force. With part geometry determining the tool path, the control software can identify the location and amount of excess material by comparing the tool trajectory to a CAD model reference. Automatically generated tool trajectories focus effort on the needed areas until the measured surface contour matches the CAD specification. This closed-loop mode of operation means the system delivers robust results despite numerous process variations, including part fixturing, the amount of material needing to be removed, and the changing size and efficiency of the material removal tool. Despite initial geometry variation and uncertainty of up to several millimeters, the accuracy of the material removal is within a few tenths of a millimeter of the desired. This is because of the common datum reference between measurement and material removal. The absolute location of the workpiece is not important, only that the robot can discover it in the context of its material removal tool. Accuracy depends on the low error of differential kinematics. While part geometry varies widely, the types of casting artifacts requiring finishing are relatively few, e.g. parting lines, pin blips, and core exit flash. The programming software uses a parametric approach that allows predetermined strategies for these feature types to be invoked on a wide array of parts. Part programs are generated by filling in blanks on sequentially arranged function blocks within a web browser interface. Material removal tool poses and transition waypoints are identified by manually guiding the robot and part to the desired positions. Areas requiring finishing are designated by annotations on the part's CAD model within the provided software. Tool trajectories, contact forces, and success metrics are dynamically generated at run time by the strategy using the user-provided parameters, geometry extracted from the CAD model, and self-acquired part measurements. The key to the system's success at working by force is its duplication of the dynamics of the human arm: low-friction joints, smooth actuation, and low-inertia links. Each joint is driven by an anchor- shaped linkage with cables that attach at the tips and wrap across its arc-shaped perimeter. A brushless motor drives a capstan that winds the cables, imparting a magnifying torque to the linkage. The drive has no backlash and is torsionally stiff. Without rubbing or sliding components, it requires no lubrication, is long-lived, and low in friction. The robot arm uses serially-connected cable drives for the first 3 joints and a compact harmonic drive actuator for the 4 th tool joint. The part positioner uses a cable drive for its pitch axis and a direct-drive rotary stage for the yaw.