Abrasion testing of anti-reflective coatings under various conditions

Abstract

PV modules can quickly accumulate a light hindering layer of dust on the front glass, especially in arid regions. This soiling needs to be removed regularly, often using dry brushes as water is usually not easily available to wash the panels. Repeated cleaning negatively affects the anti-reflective coating (ARC) on the glass, leading to permanent transmittance loss and therefore a reduced energy yield. Within this work, two different ARCs and a reference glass are investigated regarding their resistance against harsh abrasion with a dry linear brushing abrasion cycle tester for laboratory and mobile use. The reduction of the spectral transmission is studied as a function of brushing cycles (0, 100, 1000) and integral relative transmittance loss has been calculated. Furthermore, a sensitivity analysis is performed for additional test parameters being type of dust, brush speed and weight. In particular, the type of dust shows a significant influence on the abrasion behavior. Less influential are the brush weight and the speed in the range used here. The dusts used in this setup were Feldspar, Calcite, Corundum and Arizona test dust in different particle size distributions. Using the hard Corundum dust, 1000 brush cycles lead to a relative loss in transmittance of about 3%. The application potential of the presented linear abrasion test for quality assessment and benchmarking of commercial glass coatings will be discussed. In particular, the implementation of region-specific test dusts in laboratory test procedures is of high interest for further standardization activities. • Linear brush abrasion test setup was used to test two different Anti-Reflective Coatings (ARCs). • Huge differences in transmittance loss between the coatings detected. • Investigation of the impact of test parameters brush speed, weight, brush cycles and type of dust. • Dust type and hardness are the most important influencing factors on the transmittance loss while brush speed and weight show no significant impact. • Method can be used for the development of site-specific ARC.