The RAFS hardware capabilities were validated by the first successful coating of a production F-35 aircraft at RAFS in December 2008. Coating thickness control on the robotically coated unit was far superior to that of a hand-coated.
The RAFS applies to a special radar absorbing material (RAM) coating over all surfaces of the fully assembled F-35 except for the horizontal and vertical tails and various small parts that are coated in a separate Robotic Component Finishing System. RAFS comprises three six-axis robots mounted to auxiliary axis rails. All robots have X-and Y-axis rails, and the aft robot has an additional Z-axis lift to maneuver around the vertical tails on the top0 surface of the aircraft. Installation of RAFS was completed in June 2008. Coating process development was conducted using the fiberglass Finish Application Mockup of the F-35.
High positional accuracy is required to coat within exact part boundaries and to maintain the correct standoff distance, both essential for precise thickness control. RAFS is designed for a tool center point (TCP) positional accuracy of 0.08 in and a repeatability of 0.06 in a full payload. The system repeatability was verified in a independent metrology study using a Krypton infrared camera that tracked the motion and speed of LEDs mounted at eh TCP. This positional accuracy is most impressive for a system with auxiliary rails spanning the length of the F-35 and the 76-in-lon end-of-arm tooling.
This high positional accuracy is achieved by using mature commercial of-the-shelf technology wherever possible. Fanuc R2000iA 125L material-handling robots were selected instead of common “paint sprayer” robots because of their higher precision and payload capability needed to support the long end effectors. The material-handling robots were converted for use in a Class I Division 1 environment as defined by the National Electric Code. To further reduce deflection and vibration at the TCP, end effectors were constructed from lightweight, rigid composite tubing.
Each six-axis robot and its two to three auxiliary rails are controlled by a single off-the-shelf Fanuc R-J3iB controller. This allows seamless coordination of rail and robot motion when coating large surfaces. One of the primary challenges of the CASPER system used to apply F-22 coatings until 2003 was coordinating the motion of a six-axis robot and the scara arm to which it was mounted with independent custom controllers.
Repeatable positioning of the aircraft in the booth is equally as important as the positional accuracy of the robots. For this reason a custom paint cell dolly (PCD) was designed for a floor location (X and Y direction) to a repeatability of 0.010 in. The PCD also uses an in-floor track and guiding pins to help novice tug drivers position the aircraft correctly.
Because of the high positional accuracy of both the robots and the PCD, it is anticipated that there will be no need for complex user-frame adjustments for each individual aircraft as is required for the coatings systems used on the F-117A, F-22, and other aircraft.
The coating delivery system (CDS) controls all aspects of metered paint delivery to the high-volume low-pressure Kremlin gun mounted at the end effector. Several advances in the CDS enhance coating thickness control; for example, a Fluidic Systems reciprocating positive-displacement pump delivers the polyurethane-based paint at an exact volume rate, regardless of density of viscosity. Most legacy systems, including Lockheed’s F-22 and F-16 systems, instead use a fluid regulator device that relies on pot pressure to drive paint through the line. As the paint catalyzes and its viscosity increases over time, the material flow rate through the fluid regulator can deviate from the set point.
An Endress+Hauser Coriolis flow meter is used to measure the density and flow rate of the material, providing feedback to the pump’s closed-loop proportional-integral-derivative controller. The state-of-the-art flow meter helps to maintain a flow-rate accuracy of 3%± by volume and is not sensitive to vibration, allowing it to be mounted with the other CDS components on a cart, which moves with the robot along an auxiliary rail.
Booth temperature and humidity are controlled independently to optimize the coating cure rate, which can affect coating thickness. Large supply and exhaust fans above the booth ceiling drive airflow at a constant 100 ft per minute; faster airspeeds could distort the plume shape and reduce transfer efficiency. As a result of computational fluid dynamics analysis performed by Durr Industries Inc., a filter bank was installed at the front of the booth to create more diffused, laminar airflow.
Along with optimization of CDS parameters and robot programming, the hardware innovations resulted in much better coating coverage performance vs. the hand-spray ranged from 75 to 85%; for hand spray, the range was from 0 to 42%. Regarding thickness range (max-min), robotic spray realized decreases of 70 to 90% over hand spray.
This article is based on SAE technical paper 2009-01-3280 by Neal A Seegmiller, Jonathan A Bailiff , and Ron K Franks of Lockheed Martin Aeronautics Co. |
