An Intelligent Algorithm for Path Planning and Feed Rate Prediction in Shot Peening Forming of Complex Integral Panels
This paper proposes an intelligent algorithm for shot peen forming of complex integral panels that combines an extreme curvature-based path planning method with a linear interpolation model for feed rate prediction, ultimately achieving high-precision component forming validated through experimental testing on various stiffened structures.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine a massive, seamless metal skin that forms the wing of a modern aircraft. This is not a collection of riveted plates but a single, complex piece of aluminum, shaped to carry the immense weight of flight while remaining light enough to soar. To give this rigid metal the precise curve it needs, engineers do not use heat or heavy presses. Instead, they use a process called shot peening. In this method, a high-speed stream of tiny metal beads is blasted against the surface of the panel. Each bead acts like a microscopic hammer, striking the metal and leaving a tiny dent. Because the surface layer is compressed by these millions of dents, the metal underneath is forced to stretch and bend, slowly curving the entire panel into the desired aerodynamic shape.
The challenge lies in the precision required. The wing panel is not a simple curve; it is a complex surface with varying thickness and internal ribs that act like the bones of a bird's wing. If the stream of beads hits too fast, the metal will not bend enough. If it moves too slowly, the beads will hit too often, potentially damaging the surface or bending it too far. For decades, figuring out exactly how fast the nozzle should move at every single point on this complex surface has been a difficult puzzle, often relying on trial and error. A team of researchers at Northwestern Polytechnical University has now developed a smarter way to solve this, creating a system that calculates the perfect speed for the machine to ensure the wing bends exactly as designed.
The researchers began by breaking down the complex wing panel into a grid of tiny points, measuring the thickness and the curve at each location. They discovered that while the panel looks like a double curve from a distance, the curve along its length is so gentle that it can be treated as a single curve for the purpose of planning the machine's path. They mapped out the most efficient route for the nozzle to follow, tracing the lines of greatest curvature. However, if the machine were to stop and recalculate its speed for every single point along this path, the process would be incredibly slow. To fix this, the team developed a method to group these points into zones where the speed could remain constant. They merged points that required similar speeds into "equal-strength" regions, drastically reducing the number of times the machine had to adjust its pace.
To determine exactly what speed to use in each zone, the team needed to understand how the internal structure of the wing affected the bending. They built test pieces that mimicked the real wing, including flat plates and plates reinforced with different types of internal ribs, similar to the stiffeners found inside an aircraft wing. They blasted these test pieces with the metal beads at various speeds and measured how much they curved. They found that for the thinner ribs used in these specific panels, the internal structure did not significantly change the way the metal bent compared to a flat plate. This was a crucial discovery because it meant they could use data from simple, flat tests to predict the behavior of the complex, ribbed wing, saving time and resources.
With this understanding, the researchers tested three different mathematical approaches to predict the correct speed for any given point on the wing. One approach used a standard statistical formula, another used a complex computer model that mimics the way a human brain learns, and the third used a straightforward method of drawing lines between known data points to find the answer in between. The computer model and the statistical formula showed some instability, producing results that jumped around unexpectedly when the metal thickness changed slightly. The simple method of drawing lines between data points, however, proved to be the most steady and reliable. It consistently predicted the correct speed without the erratic fluctuations seen in the other methods.
The team then took their best method and applied it to a real, full-sized wing panel. They calculated the speed for every section of the path, merged them into efficient zones, and programmed a large industrial machine to perform the work. The machine moved along the panel, adjusting its speed only seven times instead of the dozens of times it would have needed to if it followed every tiny variation. After the process was complete, the researchers scanned the finished wing with a high-precision laser. The final shape matched the design with an average deviation of less than one millimeter across the entire surface. This level of accuracy confirms that their new, intelligent way of planning the path and predicting the speed works effectively, turning a complex, high-stakes manufacturing challenge into a reliable and efficient process.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.