Historically, assembly of large aerospace structures has required large, heavy-duty, expensive machines designed and built with (and for) high accuracy over the entire work envelope. The machines are generally very complex, and it is normally financially and physically impossible to build them with more than one spindle/assembly tool.
But Exechon AB says there is a solution using a platformless design of a parallel kinematic machine (PKM) to deliver high dynamics and accuracy while dramatically reducing cost and eliminating the restriction of one spindle/assembly tool.
Exechon’s goal with its PKMs—including the latest “platformless” XT300S, XT700S, and XT1100S—is to give aerospace manufacturers a solution that uses the flexibility and cost benefits of articulated arm robots, the performance of CNC machines, and the efficiency of special machines to fasten together large airplane exterior sections.
Exechon was founded around this dream in 2004, and the first “ball-jointless” X700 machine saw daylight in summer 2006. Since then the technology has been licensed to 21 manufacturers and integrators worldwide.
But the company was out to make even more improvements by eliminating the heavy platforms or structures holding and supporting the legs and/or actuators of PKM machines. It had been the case in PKM technology that all legs and actuators needed to be mounted in some kind of structure to enable it to resist the forces from each leg individually and in combination, resulting in a very heavy design not at all agile and mobile.
It was first decided to reduce moving mass. Between the design of the X700 PKM and that of the X700S, moving mass was reduced by 40% (300 kg). The savings were accomplished by moving the motors for axes one to three from the moving lower platform to the neutral inner gimbals. To further improve dynamics and temperature stability, the motors for axes four and five were put on the outside.
Then, company engineers designed out the previously mandatory platforms and structures supporting the legs and actuators. This was done by connecting the outer gimbals of actuator one and three, eliminating the need for central support of these gimbals and at the same time making it possible to add on extra material in the center (impossible in the platform design due to cable issues).
This design also prevents actuator one and three from twisting in relation to each other (when a side force is applied) on the machine, resulting in further increased stiffness and improved accuracy. In addition, the outer gimbal of actuator two has been turned 90°, eliminating a center support for actuator two as well. This design puts the holding points of actuator two right on top of the side structure.
Among other improvements was devising a way to calibrate the machines without using expensive, high-tech equipment such as laser trackers and computerized adaptive systems. Even more important was the ability to calibrate the machine tool vector—and not only the tool center point, which is a serious problem in most existing calibration systems.
Exechon developed a solution using a FARO Arm for automatic calibration. Programming is done such that a pattern of points can be checked in one tool vector and then again multiple times in other tool vectors. The pattern can be adapted to an existing jig or fixture to avoid disassembly of any parts in the system during calibration.
The collected x, y, z, and vector values are uploaded onto an external Internet-based calculation computer, which performs a large amount of mathematical parameter calculations offline. After 30 min, a complete set of 31 machine parameters and 5 reference positions are ready to be downloaded and implemented into the machine. Volumetric accuracy is about ±10 micron.
Another improvement was in the fastener assembly process itself. In cooperation with one of its licensees, Exechon developed a simplified method using fastener “discs” that are loaded off-line, where jamming of feeders and other related problems does not affect the production nor the quality of the automated assembly.
Another in the many improvements relates the challenge of adapting multiple assembly heads to an unknown airplane surface without using time-consuming probing and/or other methods. To overcome this problem, Exechon developed a new probing system in cooperation with Boeing and Meta U.K. The system combines two linear lasers into one laser head—a so-called cross laser. It sends four points of measurements to the system, which in a few seconds calculates surface angle and position, as well as hole and edge positions, to within 50 micron.