Patch order sequence optimization improves lay-up time by up to 43%

Depending on the order in which patches are placed on a mold, the production time of the Fiber Patch Placement process may vary significantly. That’s why Cevotec has incorporated a new feature in ARTIST STUDIO to optimize the patch order sequence. Its goal is to find an optimal order in which patches are placed to minimize the production time.

For every patch placement cycle, two cooperating robots have to coordinate and position themselves correctly. The mold manipulation robot positions the mold surface for the next patch position vertically upwards. The patch placement robot approaches the mold perpendicular to the placement position and then applies the patch to the mold. If the robot movements exceed the cycle time of the feeding unit providing patches, the SAMBA production system slows down and productivity drops below 100%.

Typical patch laminates are complex structures, consisting of hundreds or thousands of patches in different orientations. The engineering software ARTIST STUDIO, consisting of a CAD and a CAM module, manages this system of patches and provides a clear user interface for setting the relevant parameters. The CAD module Patch Artist generates and optimizes laminates for mechanical performance. The CAM module Motion Artist simulates and optimizes the manufacturing process via a digital twin model of the production system.

Depending on the order in which patches are placed on the mold, the robot movements are more or less time-efficient. Placing the patches in the order of initial generation within the software will usually not lead to the most efficient machine program. In order to optimize the lay-up process, the Cevotec engineers now apply a more sophisticated approach, based on algorithms for search problems. “After investigating a large set of potential algorithms, we implemented two different types,” explains Dr. Marek Wieckowski, Team Lead Software Technology. “The evolutionary algorithm and the nearest neighbor algorithm.”

The evolutionary algorithm finds the best solution by iterating patch orders, including mutations and crossovers. At each step, the algorithm picks the best solution for further iterations. The nearest neighbor algorithm works differently. After simulating the placement of one patch, it selects the next patch that represents the shortest robot travel necessary. It therefore reduces manufacturing time on a patch-by-patch approach.

First tests of the algorithms on demonstrator parts have shown an improvement of up to 43% in production time. “This saving translates into faster lay-up cycles, reduced development time and machine preparation as well as less robot movements, which prolong their service life” claims Felix Michl, CTO of Cevotec. Engineers also benefit from improved manual adjustment features in Motion Artist, which bring advantages for programming the robots on the most complex segments of a part.

“While optimization is an art in itself which we will continuously improve here at Cevotec, the implementation of these algorithms in ARTIST STUDIO represent a major benefit for manufacturers. It not only reduces production time and therefore cost, but also increases the available capacity on a SAMBA system, which further improves margin for manufacturers,” states Dr. Wieckowski.

About Cevotec:

Munich-based automation specialist Cevotec offers one of the world’s most advanced production systems for complex fiber composites. At the intersection of composites, mechanical engineering and software, Cevotec develops production systems and software based on Fiber Patch Placement (FPP) technology: SAMBA and ARTIST STUDIO. The systems, which are particularly suited for automated fiber placement on complex 3D geometries, process carbon fibers, glass fibers, adhesive films and other technical fibers. They can be used for a broad range of applications. Such flexibility is particularly appreciated by customers from aerospace and med-tech industries, who realize cost and time savings of 20%-60% when switching from manual processes to FPP.