Class-based physical properties of air-classified sunflower seeds and kernels

Publication Type

Journal contribution (peer reviewed)

Authors

Munder, S; Argyropoulos, D; Müller, J

Year of publication

2017

Published in

Biosystems Engineering

Pubisher

Elsevier

Band/Volume

164/

DOI

10.1016/j.biosystemseng.2017.10.005

Page (from - to)

124-134

Keywords

air-separation, Classification, high oleic, mechanical, physical, Sunflower

High oleic sunflower seeds, with higher oil content and greater oleic acid content than other sunflower seeds, are gaining in economic importance. However, there are still major gaps in knowledge concerning their post-harvest handling. This study was carried out, to establish a simply implemented method of classifying the seeds via air-separation, which allows rapid and precise assignment of high oleic sunflower seeds into different quality classes. Physical properties were evaluated and compared to earlier studies. A low hull to kernel ratio of 0.26 ± 0.04 was found, with seed masses ranging from 0.024 to 0.108 g. A high positive correlation between seed mass and kernel mass (r = 0.993) was observed but negligible correlations between seed length, width, thickness and mass were found. Seed mass analysis from the air-separation classes indicated highly precise (F1 score = 0.966) allocation into mass based classes, with below 5% false positives. Thus, a significant difference between hull to kernel ratio, bulk density, porosity, sphericity, the angle of repose, and rupture force in vertical and transversal orientation was found for the different classes. Rupture force in the horizontal orientation, true density and static friction did not reveal any class differences. Threshold values for air velocity where therefore established for air-separation that can classify seeds of different qualities and optimise postharvest seed handling operations.