Modelling of linseed oil expression curves. Impact of variety and maturity on the identified parameters of the model
Special Symposium - Innovations in Food Technology (LMC Congress)
Flexible Production, PAT & Modelling (Food-3b)
Keywords: linseed, oil expression, modelling, maturity
The aim of this study is to propose a method to identify a model for linseed expression curves in order to predict the expression ability of linseed on screw press. The model parameters are identified from data obtained on a small scale uniaxial press. This method gives an evaluation of the pressability of linseed from few seeds (3 to 10 g) and would be particularly interesting during the seed varietal selection process.
Seven linseed varieties (commercials or under development) are tested at seven different dates. Maturity stage of seeds is evaluated by calculation of the growing degree-day, the seeds are harvested between 839 and 1593 degree-day. Before processing, seeds are crushed and, in order to avoid any influence of the seeds humidity, all varieties are dried at 43°C down to a humidity of 4.1% (db). After conditioning, crushed seeds are stored at 4°C until expression.
Linseed is expressed using a micropress that allowed pressing of up to 10 g of seeds. Expression is performed at a constant pressure of 10 MPa during one hour. The temperature is regulated at 50°C. After expression, oil yield - calculated as a percentage of total oil content - is determined by weighting the recovered oil. This yield is considered as an overview of the ability of the seeds to oil expression (pressability).
The micropress is coupled with a universal testing machine (texturometer) and allowed recording of applied force, time and piston displacement during expression process.
After the so-called oil-point, the displacement versus time curves during constant pressure expression are modelling using a four exponentials model designed for cellular materials pressing. This model was build using a four Kelvin-Voigt rheological model (i.e. viscoeleasticity of the press cake is represented with four elements composed each of a spring and a dashpot connected in serial). The Kelvin elements represent the different stages of the cake consolidation. Each element is characterised by two physical constants (tr: characteristic time and Gi the compressibility modulus).
he corresponding equation of the model is written as: h/h∞=[1/∑(1/Gi)]*[∑(1/Gi*(1-exp(-vit))] for i=10,1,2 and 3 and where h is the piston displacement, h∞ the maximum displacement which would be reaching if all the extractible oil is removed and vi is the inverse of the characteristic time.
The mathematical identification process of the 9 parameters model starts with h∞ which is determined using a semi-empirical method. For cakes saturated in liquid, displacement h can be written as: h=t/(1/v0+at) where v0=λh∞ and a=λ/v0. By plotting t/h versus h, the determination of the slope gives the inverse of h∞. For the given h∞, parameters Gi and tri are then identified using an appropriate software (Tablecurve 2D).
The expression yield is related to the mechanical characteristics of seeds obtained by the model. The compressibility moduli appear to be the most relevant parameters because they evaluate the stress needed to extract oil from their corresponding cellular compartment. Then, the correlation between the oil yield, the model parameters, the variety and the maturity of the seeds is discussed.
See the full pdf manuscript of the abstract.
Presented Wednesday 19, 17:15 to 17:20, in session Flexible Production, PAT & Modelling (Food-3b).
