Josse De Baerdemaeker

Light penetration properties of NIR radiation in fruit with respect to non-destructive quality assessment

Abstract Some issues related to the non-destructive measurement of apple quality attributes by means of NIR reflectance spectroscopy are addressed. A comparison was made between two optical configurations, which can be used to perform NIR-spectroscopic measurements: the bifurcated and the 0°/45° optical configuration. A relationship was established between the reflectance spectra (880–1650 nm) and the soluble solids content by means of the partial least squares technique. Depending on the data pre-processing method, correlation coefficients between 79 and 91% were obtained. The results obtained with the bifurcated fibre were only marginally better than those obtained with the 0°/45° configuration. The apple skin reflectance and skin transmission properties with regard to NIR radiation were also investigated. The intensity of the light source was high enough to penetrate through the apple skin and gather information about the apple parenchyma tissue. A method was developed to calculate the light penetration depth for each wavelength in the range from 500 to 1900 nm. This method was applied to measure the light penetration depth in ‘Jonagold’ apple fruit tissue. The penetration depth is wavelength dependent: up to 4 mm in the 700–900 nm range and between 2 and 3 mm in the 900–1900 nm range.

Computational fluid dynamics modelling and validation of the temperature distribution in a forced convection oven

Abstract This paper discusses the validation of a Computational Fluid Dynamics (CFD) model to calculate the heat transfer in an industrial electrical forced-convection oven. The CFD model consists of the continuity, momentum and energy equation with the standard k–e approach to model the flow turbulence. Density effects are accounted for through a weakly compressible formulation. Time-dependent boundary conditions and source terms are derived from a simplified lumped model, which results in a good qualitative agreement of the calculated oven temperatures and the measured temperature distribution. The average oven temperature difference between measurements and predictions is 4.6°C for a set point of 200°C. The heating uniformity of PVC bricks in different configurations was calculated with the CFD model, but the wall functions in the k–e model limit the accuracy to a qualitative agreement. A correlation was established between the calculated flow field variables and measured surface heat transfer coefficients.

Acoustic impulse-response technique for evaluation and modelling of firmness of tomato fruit

The firmness of tomato fruit was monitored using an acoustic impulse-response technique. This technique provides a stiffness factor S, based on the first resonance frequency and the mass of the intact fruit. A logarithmic relation was found between objective stiffness measurements on tomatoes by the acoustic impulse-response technique and subjective firmness measurements, suggesting that subjective measurements distinguish firmness differences more easily in soft than in firm tomatoes. Accuracy and repeatability of the acoustic impulse-response technique and of subjective scores by market experts were equally high. However, the advantage of using S is that it is objective and no scaling problems between different experts occur. In soft fruit only a stiffness difference as large as 2.5×106 Hz2 g2/3 could be distinguished with satisfactory accuracy by all experts. Tomatoes with a stiffness of 2.0×106 Hz2 g2/3 or less had more than 50% chance of being rejected. The acoustic impulse-response technique was also used to study firmness changes of tomatoes during storage, and the influence of variety, producer, season, production method, maturity at the time of picking, and storage conditions. Change in stiffness as a function of time could be expressed as a decreasing exponential function. Tomatoes stored at 20°C had the same deterioration constant, regardless of maturity at harvest. However, when fruit were stored at 12°C, the deterioration constant of the three tested varieties was greater for red than for orange fruit. The deterioration constant was much higher for fruit harvested in spring than in the other seasons. The influence of temperature could be expressed in the form of an Arrhenius equation. The model was validated for tomatoes subjected to fluctuating temperatures.

Computational fluid dynamics modelling and validation of the isothermal airflow in a forced convection oven

Abstract This article discusses the application of computational fluid dynamics (CFD) to calculate the three-dimensional isothermal airflow in an industrial electrical forced-convection oven. The governing fluid flow equations were expanded with a fan model and a turbulence model. The standard and the renormalisation group (RNG) version of the k – e turbulence model produced comparable results. The performance of the CFD model was assessed by means of point measurements of the velocity with a directionally calibrated hot-film velocity sensor. From the validation it was found that important aspects of the model are the fan head-capacity relationship, the fan swirl and the oven geometry. The calculation error was on an average 22% of the actual velocity, caused by the limitations in turbulence modelling and numerical grid density.

Eggshell Penetration of Various Types of Hens' Eggs by Salmonella enterica Serovar Enteritidis

Egg weight, shell thickness, number of pores, cuticle deposition, eggshell strength (dynamic stiffness and damping ratio), and the ability of Salmonella enterica serovar Enteritidis (SE) to penetrate the eggshell were determined. Penetration was assessed by filling the eggs with a selective medium that allowed viewing of Salmonella growth on the inside of the shell and membrane complex. After inoculation of each shell with on average 2.71 log CFU, the eggs were stored for up to 14 days at 20 degrees C and 60% relative humidity. Commercially available eggs were used. At 14 days of storage, only 6.0% of the eggs from free-range hens and 16.0% of the generic (i.e., eggs from hens in conventional battery cages that were given standard feed) white eggs were penetrated. The generic brown, organic, and omega-3-enriched eggs were penetrated at a frequency of 30 to 34%. In a second experiment it was shown that the layer strains of the hen (ISA-Brown Warren versus Bovans Goldline), which were kept in furnished cages, did not affect eggshell penetration by SE. For Bovans Goldline hens, the housing system (furnished cage versus aviary) did not affect penetration, while a trend was visible toward a higher fraction of penetrated eggshells when hens were fed corncob mix rather than standard feed. Eggshell penetration was observed more frequently in the absence of cuticle spots and for eggs having lower dynamic stiffness values. Shell contamination at the end of storage was highly correlated with SE penetration.

A neural network based plant classifier

The Self-Organizing Map (SOM) neural network is used in a supervised way for a classification task. The neurons of the SOM are extended with local linear mappings. Error information obtained during training is used in a novel learning algorithm to train the classifier. The proposed method achieves fast convergence and good generalization. The classification method is then applied in a precision farming application, the classification of crops and weeds using spectral properties. The proposed method compares favorably with an optimal Bayesian classifier that is presented in the form of a probabilistic neural network. The classification performance of the proposed method is proven superior compared with other statistical and neural classifiers.

Predictive microbiology in a dynamic environment: a system theory approach☆

The main factors influencing the microbial stability of chilled prepared food products for which there is an increased consumer interest-are temperature, pH, and water activity. Unlike the pH and the water activity, the temperature may vary extensively throughout the complete production and distribution chain. The shelf life of this kind of foods is usually limited due to spoilage by common microorganisms, and the increased risk for food pathogens. In predicting the shelf life, mathematical models are a powerful tool to increase the insight in the different subprocesses and their interactions. However, the predictive value of the sigmoidal functions reported in the literature to describe a bacterial growth curve as an explicit function of time is only guaranteed at a constant temperature within the temperature range of microbial growth. As a result, they are less appropriate in optimization studies of a whole production and distribution chain. In this paper a more general modeling approach, inspired by system theory concepts, is presented if for instance time varying temperature profiles are to be taken into account. As a case study, we discuss a recently proposed dynamic model to predict microbial growth and inactivation under time varying temperature conditions from a system theory point of view. Further, the validity of this methodology is illustrated with experimental data of Brochothrix thermosphacta and Lactobacillus plantarum. Finally, we propose some possible refinements of this model inspired by experimental results.

The starch gelatinization in potatoes during cooking in relation to the modelling of texture kinetics

Abstract A dynamic model for texture changes during cooking treatments of potato samples was developed. The model includes the kinetics of the gelatinization process and was compared with a simpler model lacking this feature. Model parameters were derived from texture measurements on potato samples. The experimental results and simulations show that the gelatinization process contributes only to a limited extent to the texture.

Effects of process conditions on the pH development during yogurt fermentation

Continuous pH measurements were performed during the yogurt fermentation process. Various conditions of dry matter fortification, heat treatment of the base milk, starter culture, and incubation temperature, were included in the experiments. The pH profiles with incubation time were described on the basis of the modified Gompertz equation for bacterial growth. The four parameters of the equation were used to evaluate the effects of the different process conditions. Dry matter fortification did not affect the pH development. Sterilization of the base milk enhanced the pH development. Incubation temperatures that were lower than the optimal one, caused the pH development to slow down. Inoculation with a ropy strain starter culture yielded a different shape of the pH profile with time. The lag time for pH decrease and the maximal pH decrease with time, reflected the intended acidification rates of the involved starter cultures.

The local surface heat transfer coefficient in thermal food process calculations: A CFD approach

The surface heat transfer coefficient during thermal processing of foods of different shapes and for different heating conditions has been calculated using Computational Fluid Dynamics techniques and compared to experimental results obtained from the literature. The calculated product-averaged coefficients are a maximum of 14% smaller than the experimental values. For Reynolds numbers lower than 10,000, the calculated local distribution corresponds to the experimental pattern. Variations in the product temperatures are evaluated using a finite element model for heat conduction. For the cases studied, the simulations indicate that the product temperature change as a result of local variations of the surface heat transfer coefficient is slower than that obtained under the assumption of a homogeneous surface heat transfer coefficient. Further, the coldest spot is no longer at the geometric centre.