Rehydration characteristics of milk protein concentrate powders monitored by electrical resistance tomography

Electrical resistance tomography (ERT) is a robust and low-cost method offering real-time visualization of processes. In this work, we developed an ERT-based method to characterize the rehydration behavior of milk protein concentrate (MPC) powders. Circular-type and linear configurations were used to achieve high resolution in the radial and axial directions, respectively. To evaluate the rehydration profile, MPC powders were reconstituted to 2.5% (wt/wt) total solids at room temperature, and the rehydration behavior of the MPC powders [MPC with 85% protein (MPC85) and milk protein isolate with 90% protein (MPI90)] was monitored for a dissolution time of 30 min using the ERT system. The MPC powders were characterized in terms of overall mean conductivity, area under the mean conductivity curve, slope at a dissolution time of 3 min, and the relative dissolution index. Additionally, the focus beam reflectance measurement (FBRM) was used as a reference method to follow rehydration characteristics. Particle count changes from the FBRM measurements showed that MPI90 had higher larger particle counts and more resistance to dispersing in water. As the dissolution time proceeded, mineral ions and proteins were released and consequently increased the overall conductivity, confirming the transfer of water into MPC particles. At lower protein contents, the particle dispersion rate was higher and an increase in overall mean conductivity was observed, indicating better powder dissolution. Both configurations were able to effectively monitor differences in the dissolution behavior of MPC powders. In the ERT circular configuration, MPC85 and MPI90 showed maximum conductivity of 0.201 ± 0.006 and 0.162 ± 0.001 mS/cm, respectively. In the linear probe configuration, MPC85 and MPI90 showed maximum conductivity of 0.161 ± 0.001 and 0.136 ± 0.001 mS/cm, respectively, suggesting increasingly inhibited water transfer as the protein content of the powder increased. In this study, we demonstrated the capability of ERT using the circular and linear probe configurations to offer, in addition to qualitative tomographic images, reliable quantitative data by which to characterize the dissolution behavior of high-protein dairy powders.

A Novel Ensemble Model on Defects Identification in Aero-Engine Blade

Machine learning-based defect identification has emerged as a promising solution to improving the defect accuracy of the aero-engine blade. This solution adopts machine learning classifiers to classify the types of defects. These classifiers are trained to use features collected in ultrasonic echo signals. However, the current studies show the potential number of features, such as statistic values, for identifying defect reaches a number more than that offered by an ultrasonic echo signal. This necessitates multiple acquisitions of echo signal and increases manual effort, and the feature obtained from feature selection is sensitive to the characteristic of the classifier, which further increases the uncertainty of the classifier result. This paper proposes an ensemble learning technique that is only based on few features obtained from an echo signal and still achieves a high accuracy of defect identification as that in traditional machine learning, eliminating the need for multiple acquisitions of the echo signal. To this end, we apply two well-known ensemble learning classifiers and simultaneously compare three widely used machine learning models on defect identification of blades. The result shows that the proposed ensemble learning models outperform machine learning-based models with an equal number of features. In addition, the two-feature-based ensemble learning model reaches an accuracy close to that of multiple statistic features-based machine learning models, where features are obtained from multiple collections of the signal.

Monitoring Viscosity and Total Solids Content of Milk Protein Concentrate Using an Inline Acoustic Flowmeter at Laboratory Scale

Control of milk concentrate viscosity and total solids (TS) content prior to spray drying can improve dairy ingredient manufacture. However, the availability of hygienic and appropriately pressure rated process viscometers for inline monitoring of viscosity is limited. An acoustic flowmeter (FLOWave) is an inline process analytical technology (PAT) tool that measures changes in acoustic signals in response to changes in liquid properties (i.e., acoustic transmission (AT), acoustic impedance (AI), temperature and volume flowrate). In this study, an acoustic flowmeter is evaluated as an inline PAT tool for monitoring viscosity of milk protein concentrate (MPC85), protein and TS content of (MPC85), and standardised MPC (sMPC). Laboratory scale experiments were carried out at 45 °C for five different concentrations (4–21%) of MPC85 and sMPC. Results showed that AT decreased with an increase in MPC85 viscosity (e.g., AT was 98.79 ± 0.04% and 86.65 ± 0.17% for 4% and 21% TS content, respectively). Non-linear regression was carried out to develop a relationship between AT and offline viscosity (R² (coefficient of determination) value = 0.97 and standard error of prediction = 1.86 mPa·s). AI was observed to increase at higher protein and TS content which was dependent on protein to total solid ratio (P_TSR). Multiple linear regression was carried out to develop the relationship between AI, protein content, TS content and P_TSR. Results demonstrated that AI could be used to monitor the protein and TS content of milk protein concentrate (R² > 0.96). Overall this study demonstrated the potential of an inline acoustic flowmeter for monitoring process viscosity, protein and TS during dairy concentrate processing.

Application of front-face fluorescence spectroscopy as a tool for monitoring changes in milk protein concentrate powders during storage

This study investigated the feasibility of front-face fluorescence spectroscopy (FFFS) to predict the solubility index and relative dissolution index (RDI) of milk protein concentrate (MPC) powders during storage. Twenty MPC powders with varying protein contents from 4 different commercial manufacturers were used in this study. The MPC powders were stored at 2 temperatures (25 and 40°C) for 0, 1, 2, 4, 8, and 12 wk. The front-face fluorescence spectra of tryptophan and Maillard products were recorded and analyzed with chemometrics to predict solubility of MPC powders. The similarity maps showed clear discrimination of the MPC samples stored at 25 and 40°C. Partial least squares regression models were developed using the fluorescence spectra of tryptophan and Maillard products to predict the solubility index and RDI measurements of MPC powders, and the prediction models were validated using an independent test set. Coefficients of determination (R²) of 0.76, 0.84, and 0.68 were obtained between fluorescence spectra (tryptophan emission, Maillard emission, and Maillard excitation, respectively) and solubility index. The R² values for the RDI predictions were 0.58 and 0.60 for the data set of tryptophan emission and Maillard emission, respectively. The ratio of prediction error to standard deviation was >2 for Maillard emission fluorescence spectra and solubility index measurements, indicating good practical utility of the partial least squares regression prediction models. The results indicated that the solubility and dissolution behavior of MPC powders were related to their protein content and storage conditions that could be measured using FFFS. Hence, FFFS can be used as a rapid nondestructive analytical technique to predict the solubility and dissolution characteristics of MPC powders.