Abstract Scope |
Introduction: Robotic directed energy deposition (DED) is a versatile, rapidly developing technology for “digital” welding, additive manufacturing (AM), and repairing parts. DED processes offer a unique alternative for producing complex structures in a high-mix environment as well as economic and schedule benefits compared to castings and other long-lead fabrications. Robotic DED systems are not readily available as off the shelf systems. DED systems are configured by modeling welding-based systems. These systems are converted into AM by utilizing CAD to path software such as Autodesk’s PowerMill. The robot system is modeled, and the deposition process and apparatus are added to create a DED AM capable cell. AM components are “digitally” manufactured from computer-aided design (CAD) files using computer-aided manufacturing (CAM) solvers. Multiple items need to be setup for a path planning software to communicate with the robotic system. This presentation will address these items, and how to setup software to model the digital twin and then calibrate the digital model with the physical robotic system. A range of best practices are recommended based on converting multiple robotic welding systems, which are commercially available, into DED AM systems.
Experimental Approach: There are multiple manufacturers of robotics around the world with an infinite number of configurations. A list of items that can be included in a robotic cell are:
• 6 to 7-axis robots
• 1 to 2-axis positioners
• 1-axis linear rails
• 3-axis gantries
Welding equipment (power supply, consumable(s) feeder, torch apparatus) are only limited by their ability to be integrated into a robotic system. In addition to the above, multiple peripherals can also be added to a cell. These items can include:
• Infrared (IR) spot sensors.
• IR cameras.
• Forced cooling systems.
• Wire clipping stations.
• Full torch cleaning stations.
• Automatic tool center point (TCP) calibration stations.
• Seam tracking.
• Height sensing.
• Automatic torch or contact tip exchange.
This project developed five robotic configurations across multiple robotic brands, power supplies, and peripherals. One of these systems is listed below as an example:
• Multi-Process Gantry configuration:
o Gantry (3-Axis)
o ABB IRB 4600 Robot (6-Axis)
Weld Guide IV System
ABB Bullseye Auto-TCP Calibration
ABB 1000kg Tilt/Turn Positioner (2-Axis)
o Welding Systems
Fronius TPS 500i CMT System with CMT Torch
Lincoln S500 System with STT Module and Binzel Torch with Exchange Station
Welding Torch Maintenance Stations
• Abicor Binzel TCS-FP Torch Station
• CM Industries Wire Clippers
o Auxiliary Equipment
MeltTools DART Puddle Camera
PushCorp 5HP BT30 Tool with Active Compliance and Tool Change Station
o Sensors
Micro Epsilon Infrared Camera
Abicor Binzel TH6D-GF Optical Seam Tracking Sensor
o Build Envelope Dimensions:
X-axis: 20 ft
Y-axis: 13.7 ft
Z-Axis1: 8.2 ft
Once the system was designed, integrated, and delivered, the digital twin and post-processor were developed in PowerMill. The digital twin is a digital replica of the physical cell; the post processor is a unique driver specific to a robotic system that converts CAM language to a language that the robotic controller can understand. Once the digital twin and post-processor were completed, they were calibrated to the real physical system.
Results and Discussion: The setup of a robotic DED AM system was grouped into three phases:
1. Building the digital twin
2. Developing the post processor
3. Calibrating and validating
The first DED system setup step was development of the digital twin. This started by defining the robot manufacturer’s functional data to create an accurate CAD model of the cell layout, external axis, and attached end effector(s) (i.e., welding torch). After this information is defined, the following items were located to understand how the robot can move, and what limitations it had:
• The robot zero position
• Axis directions
• Axis minimum and maximum limits
• End effector orientation
• Tool attach point
• External axis positions
Once this was all understood, the CAD model was completed and loaded into PowerMill. The robot system component CAD models can be developed in any off the shelf CAD suite where this project used SolidWorks. Once the component models were complete the individual entities were exported and configured in PowerMill to create an accurate digital twin for each system.
The second step was development of the post-processor for each DED AM system. Toolpaths were created independent of the machine tool or robot. Each toolpath was then processed through the PowerMill Robot Interface for a specific robot cell. The orientation of the tool or torch, collision avoidance, and singularity avoidance were edited in this step of the process. All this robot motion was recorded along with weld parameters, deposition feed rates, and other parameters and saved in the PowerMill’s RobSim file. The post-processor compiled the RobSim file and translated the build file into the specific language for each robot manufacturer system. The physical build file included weld commands for the DED process equipment which must be turned on and off at the appropriate points in the toolpath. The post-processor was configured to calculate and format everything to suit each robot manufacturer.
The third and final step was calibration of the digital twin with the physical DED AM system and validate the post-processor and digital twin. A simple path was drawn within PowerMill consisting of a square. Robotic code was generated by PowerMill and verified to match the real environment. This was repeated using each external axis as well to ensure coordinated motion between all components is correct. Once this was complete, each robotic DED AM cell was used to make builds of increasing complexity.
Conclusion: Five different robotic DED AM cells were developed with a range of commercial robot systems. These systems were configured with a range of arc DED process equipment, torch apparatus, support process, and monitoring sensors. This matrix of robotic DED systems configurations provide a wide range of capabilities for building different features and different scale from 1 to 155 cubic yards using a small platform to large-scale 9-axis gantry systems, respectively. |