Eventos Anais de eventos
COBEM 2021
26th International Congress of Mechanical Engineering
Computer vision system for automatic evaluation of images and thickness measurement of dry layer of wind blade protection paint
Submission Author:
DARLAN ALMEIDA BARROSO , CE
Co-Authors:
DARLAN ALMEIDA BARROSO, JÚLIO CÉSAR CAPISTRANO ESTÁCIO, Marcello Reis, Jéssyca Almeida Bessa, Auzuir Ripardo Alexandria
Presenter: JÚLIO CÉSAR CAPISTRANO ESTÁCIO
doi://10.26678/ABCM.COBEM2021.COB2021-2290
Abstract
The search for inexhaustible sources of energy has been caused the wind energy market to have a great expansion for 10 years. The blades are considered the main components in the design of a wind turbine and have the function of capturing wind energy and later converting it into rotational energy. The enormous structures that make up the blades make it difficult to manufacture and inspect, and this leads to the search for innovative methods capable of ensuring product quality. Failures in wind blades during operation can generate environmental risks and significant loss of capital due to unexpected downtime, and studies show that most failures in wind turbines in the last 5 years are related to blades defects. The quality of a blade during the manufacturing process depends essentially on the people who manufacture it, so manufacturers must have well-refined process controls to avoid failure. Several controls undergo visual inspection and manual measurement to validate items that are critical to the manufacturing process and product quality. Due to the still shy amount of work found in the literature using computer vision, and with the work found limited to inspections during the operation of small wind turbines, proposing corrective solutions for surface defects, this work proposes the application of an automatic inspection system for dimensional and quality control of wind turbine blades in a factory environment by means of computer vision and pattern recognition algorithms, applied to the evaluation of images captured during the dry paint thickness measurement process of the protective paint that guarantees the integrity of the blade surface for the life span for which it is sized. The developed system evaluates images that are captured using a digital microscope whose focus is standardized and parameterized using a graduated ruler with a resolution of 0.01 mm. The evaluated images correspond to a destructive test that uses a hand drill with a drill that has a known cutting angle and makes a hole in the painted surface of the blade. With the development of the system, the thickness measurements that were previously verified by an inspector visually and individually evaluating each captured image to guarantee the paint thickness on a huge painted surface of a blade, will be evaluated automatically, increasing the reliability measurements and the traceability of possible defects that may appear during the operation of the wind turbine.
Keywords
Computer Vision, pattern recognition, wind turbine blade, automatic evaluation
