LAUSR

Abstract Undesired agglomeration of powders, known as caking, results in the loss of product quality and is unacceptable in many powder industries. Information on how ambient conditions influence the cake strength is essential to prevent caking. Although there are a lot of existing methods to form caked samples and even more to characterize them, there is still no consensus regarding these methods. Repeatability is rarely evaluated and the comparative studies remain infrequent. As a deliquescent powder, sucrose is characterized by a threshold Relative Humidity (RH), called Deliquescence Relative Humidity (DRH), above which it turns into an aqueous solution. Partially deliquesced crystals are linked by liquid bridges of aqueous solution. When the RH is decreased below the DRH, these liquid bridges recrystallize into solid bridges to form a hard cake. In this study, an accelerated caking device was developed to obtain homogeneous caked samples of sucrose under controlled conditions. Three mechanical tests were compared on the basis of their sensitivity and repeatability: the uniaxial compression test, the diametrical compression test (tensile test) and the shear test. The varying parameter was the amount of water uptaken before drying, and the data were fitted by cross-validated local linear nonparametric regression. The best mechanical test was found to be the shear test, giving a sharp failure of the cake along a predetermined failure surface and a linear relationship between the cake strength and the amount of water uptaken.