WO2006110963A1

WO2006110963A1 - Monitoring of the v1sco-elastic properties of gels and liquids

Info

Publication number: WO2006110963A1
Application number: PCT/BE2006/000038
Authority: WO
Inventors: Flip Bamelis; Josse De Baerdemaeker
Original assignee: K.U.Leuven Research And Development
Priority date: 2005-04-21
Filing date: 2006-04-21
Publication date: 2006-10-26

Abstract

The present invention provides a method and apparatus for the measurement of the visco-elastic properties of a liquid or gel, whereby said parameters are determined by measuring the vibrations generated by an air puff on the surface of a liquid or gel. The apparatus and method are particularly suited to measure the evolution of the visco- elastic properties during the coagulation of milk during the production of curd.

Description

MONITORING OF THE V1SCO-ELASTIC PROPERTIES OF GELS AND LIQUIDS

FIELD OF THE INVENTION

The present invention provides a method and apparatus for the measurement of the visco-elastic properties of a liquid or gel, whereby said parameters are determined by measuring the vibrations generated by an air puff on the surface of a liquid or gel. The apparatus and method are particularly suited to measure the evolution of the visco- elastic properties of milk during its coagulation in the production of many kinds of dairy products.

BACKGROUND OF THE INVENTION

A considerable amount of research has been undertaken to develop a measurement technique for monitoring the coagulation process of milk. These techniques are mainly used in studies aiming at the understanding of the processes associated with the coagulation process. In order to allow the use of these measurement systems in commercial cheese making, the measurement system should comprise following characteristics: (i) operate online on a cheese maker's vat, (ii) be non intrusive and (iii) be up-to-date with dairy hygienic requirements (O'Callagahn et al., 2002). Because of these requirements, many of the methods used for research on the coagulation of milk cannot be used in commercial cheese production processes. Lucey (2002) lists 11 difficulties to resolve for automated measurement techniques of the coagulation process. The existing techniques can be divided in 7 groups, based on their measurement principle.

Mechanical systems

In general, the mechanical systems to monitor the coagulation process are registering the drag force from the milk on a moving body. A well known reference technique used by lots of research institutes is the Formagraph, which is based on this measurement principle (McMahon and Brown, 1982). Other techniques are developed from which differ mainly on the part inserted in the milk, as there are the Curd Firmness Tester (Kowalchyk and Olson, 1978), the pressure transmission system (Hatfield, 1981), the Gelograph M (Reddy, 1992) and the Amelung coagulometer (Mehra et al, 1994). The disadvantage of this kind of devices is the very carefull cleaning procedure to maintain the correct operation. Moreover during the measurement, the curd becomes more or less damaged and this can alter the measurements done in the same curd later on. For these reasons, these techniques cannot be used in online applications. Vibrational systems

These techniques are less intrusive than the mechanical techniques and can be adapted in commercial cheese vats. Here, a vibrating probe is immersed in the coagulating milk and the viscosity of the curd is measured by damping and phase-shift effects on the probe. Moreover, since these probes have no mechanical moving parts inside, they are less subjected to mechanical wear. Research has been done with different frequencies of the excitating probe. Senge et al. (1997) and Schulz et al (1997a); Schulz et al (1997b) report the use of the Paar Physica Rheoswing at 8500 Hz, Sharma et al (1989) studied the Nametre torsional vibration viscometer probe at 660Hz and Lagoueyte et al (1995) reported the use of a very low frequency vibrating probe (10-50 Hz). Although the lower frequency applications seems to damage the curd less, the disadvantage of these measurement systems stays the damage and hence the interaction of the measurement technique with the coagulation process. Ultrasonic systems Ultrasonic waves with wavelengths smaller as the size of the casein micelles in the coagulating milk (i.e. frequencies larger as 1MHz), can be used to monitor the agglomerating casein micelles and hence the coagulation process. Several researchers investigated different frequencies (1-400 MHz) of the ultrasonic waves (Ay andGunasekaran, 1994; Benguigui et al., 1994; Cosgrove, 2000). In 1992, a patent was awarded in which the use of ultrasonic system is patented for the monitoring of phase shifts in foodstuff as there is the coagulation of milk (Beudon et al, 1992).

In general, the ultrasonic systems seem to be promising, but certainly further investigations are needed since different compositions of the milk can cause difficulties with these types of devices. Electrical conductivity Dejmek (1989) reported an experiment in which the electrical conductance of the coagulating milk is shown to change with 0.5-1.0%. However, a large natural variation can be expected of the electrical conductance due to temperature variations and changes in the concentration of electrolytes. Therefore, the application of this measurement technique seems to be limited. Hot wire probe

This technique is based on the principle that the viscosity of the fluid governs convective heat transfer from a heated probe immersed in a fluid. The fall in viscosity of the coagulum can be observed by measuring the heat transfer rate, or in practice by measuring the temperature difference between a heated probe and the milk. Van Hooydonck and Van den Berg (1988) reported the hot wire as the first technique that is useable for in line measurement of the coagulation process.

However, other research (O'Callaghan et al, 2002) reported the hot wire technique to measure the gelation point right, but it does not measure the curd firming phase of the coagulation process. Optical systems

It has been known for many decades that the coagulation of milk is accompanied by changes of its optical properties. And because of the decreasing costs for optical technology and its easiness in use, lots of efforts are done to develop a suitable optical technique to monitor the coagulation process. The coagulation process can be monitored by both colour measurement techniques (Hardy and Fanni, 1956; Dybowska and Fujio, 1996) and by absorbance, transmission or reflectance techniques. The latter technique can be applied in the visible range (McMahon et al, 1984a&b) or in the Near Infrared Range (Banon and Hardy, 1991 ; Scher and Hardy, 1993; Ustunol et al, 1991). Results from this research is showing clear optical changes during the coagulation process, but since non linear relationships between time and optical changes are reported, the application of these techniques is more difficult. Dynamic light scattering technique (DLS) Here, the measurement principle is based on the Brownian movement that a particle in a suspension describes. Light that is scattered by this moving particles experiences a Doppler shift. During the coagulation, this movements are reduced due to the increasing viscosity of the milk. The application of this technique is called the Diffusive Wave Spectrosopy (DWS) and was introduced by Home and Davidson (1990). They found that this technique could follow the entire gel firming stage to which most on-line techniques are not well suited. O'Callaghan et al. (2002) concludes that this technique is of great potential, but still needs to be compared with other systems.

EP0442667 discloses a method and apparatus for the measurement of the viscosity of paints using an array of nozzles suspended above and directed normally to the surface of a sample of said liquid, having a known depth. Periodically, a burst of air is forced through the nozzles to force a harmonic surface wave in the liquid having a wavelength equal to the periodic spacing of the array of nozzles. One light beam, placed in an angle with the direction of the nozzles, is reflected from the surface of the liquid and is optically detected to measure the damping of the surface wave and the surface wave frequency. This damping and frequency is then used to determine the surface tension and the viscosity.

The present invention provides a method for the measurement of the visco-elastic properties of a liquid or gel, whereby said properties are assessed by measuring and analysing the vibrations generated by one air pulse on the surface of said liquid or gel. However, the method of the present invention does not depend on the availability of information on the depth of the liquid or gel sample nor does it require the generation of a wave of a known wavelength. The apparatus and method are particularly suited to measure the evolution of the visco-elastic properties of milk during coagulation. The apparatus has the advantage that it can be easily be installed on a cheese makers vat and that the measurements do not interfere with the curd formation process.

SUMMARY OF THE INVENTION

The present invention provides a method for the measurement of the visco-elastic properties of a liquid or gel, whereby said properties are assessed by measuring and analysing the vibrations generated by a gas pulse on the surface of said liquid or gel. In a second object the present invention provides an apparatus allowing a practical implementation of said method.

DETAILED DESCRIPTION OF THE INVENTION

Legends to the figures

Figure 1: Interpretation of the parameters of a sigmoid curve. Values of the parameters in this example: A=1 B=-0.75 C=5 k=0.5 Figure 2: Measured time signals on milk, 2, 5, 15 and 20 minutes after the start of the coagulation process. Figure 3: The measured total deformation [V] (dots) with the best sigmoid fit (full lines) for each experimental trail.

Figure 4: Evolution of the elastic modulus G', measured by the rheology- setup for the normal milk (full line) and the CaCI₂ enriched milk (dashed line). Mean curves are presented.

Figure 5: Evolution of the gelograph output for the normal milk (dashed line) and the CaCI₂ enriched milk (full line). Each trail is presented. Figure 6: The different signals registered at different temperature. At each temperatures, two measurements are done Figure 7: Determination of the intrusion and jump back values calculated from each registered time signal.

Figure 8: The evolution of the damping with decreasing temperature. The quadratical fit (full line) fits best the measured values (crosses). R²=0.84 Figure 9: Schematic presentation of an apparatus of the present invention also referred to as a Foodtexture Puff Device. (1) nozzle, (1a) top view of the nozzle, (1 b) bottom view of the nozzle, (2) Laser displacement sensor, (3) outgoing laser beam, (4) reflected laser beam, (5) gas pulse, (6) air valve, (7) liquid or gel surface, (8) inner tube of the nozzle, (9) outer tube of the nozzle and (10) computing unit. Figure 10: Calculation of maximal and minimal deformation values from the registered deformation curve by the FPD. The example is the taken from the coagulation process from milk with an initial pH of 6.29 just after adding the rennet (left) and 20min later (right). The range of the FPD is the summation of maximal value and minimal value. Figure 11: The kinetics of the maximal and minimal deformation and the deformation range measured by FPD (upper plot) and the rheological parameters G' and G" (lower plot) during the coagulation process of the sample with milk pH 6.29. Measurement points and fitted model are presented. Figure 12: The fitted model of the evolution of the total range measured by the FPD with the physical interpretation of the fit parameters. 2C is the start point of gelation, A is the initial total range, A-B is the final total range and 2C-E/D is an estimator for the speed of the gelation process. Figure 13: the effect of the initial milk pH on the start of the gelation process (T=33°C), as measured by G' (o), G"(D), the minimal puff value (D), the maximal puff value (x) and the range of the puff (•). For each dataset, the exponential model is also presented. Figure 14: the relationships between the speed of the gelation process estimated by 2C-E/D for G' and G" (left plot), for the minimal puff value (middle plot) and the maximal puff value (right plot). For each parameter, the exponential model as well as the datapoints are presented.

Description

In a first object the present invention provides a method for the assessment of the visco- elastic properties of a liquid or gel. Said method comprises the steps of (i) generating a gas pulse, for instance an air pulse directed towards the surface of a liquid or gel, (ii) measuring the deformation of said surface during and/or following the gas pulse in order to record the vibrations generated on said surface by the gas pulse, (iii) digitising the measured signals, (iv) analysing the digitised signals in order to determine selected vibration parameters which correlate with the visco-elastic properties of the liquid or gel, such as the power spectrum, the resonant frequency, the amplitude, the maximal deformation and damping. In a particular embodiment the method of the present invention comprises a fifth step (v) wherein a value indicative for the visco-elastic properties of the liquid or gel is calculated based on the determined vibration parameters and using information on the relation between said vibration parameters and the visco- elastic properties of such liquid or gel. In a preferred embodiment of the method of the present invention the gas pulse is emitted perpendicular to the plain of the surface of the liquid or gel. Furthermore, it is preferred that the deformation of the liquid or gel surface is measured within the zone of the said surface excited by the gas pulse. The duration of the gas pulse should be longer than 10 μs, preferably this duration should be between 30 and 60 μs, for example between 45 and 50 μs. The pressure of the gas pulse is selected between 0.1 and 8 bar, depending on the viscosity of the liquid or gel. Typically the pressure of the gas pulse is between 0.1 and 6 bar and 0.1 and 2 bar for liquids and gels, respectively. In a second object the present invention provides an apparatus for the measurement of the visco-elastic properties of a liquid or gel. The apparatus of the present invention comprises (i) a means for generating a gas pulse, which can be released on the liquid or gel, (ii) a means for measuring the vibrations induced on the surface of the gel or liquid during and/or following the gas pulse, (iii) means to digitise the measured signals and (iv) means to analyse the digitised signal and to determine the vibration parameters correlating with the visco-elastic properties of the monitored liquid. The apparatus may further comprise means to calculate a value describing the visco-elastic properties of said liquid or gel based on the determined vibration parameter and using information on the relation between said vibration parameters and visco-elastic properties of such liquid or gel. Preferably, the means for generating the gas pulse is a gas nozzle being operationally linked to a valve and a pressurised gas source. The means for measuring the vibrations induced on the surface of the gel or liquid can either be a laser displacement sensor comprising a signal digitising unit or another contact less measurement technique, such as targeted ultrasonic waves. Preferably the means for analysing the digitised signal is a computing unit.

When using a laser displacement sensor, the measurement dot of said sensor is preferably pointed on the zone where the gas puff excites the gel or liquid surface. A laser displacement sensor can either be positioned in line with the gasflow or in an angle with the gasflow. However, mounting the laser displacement sensor in line with the gasflow has the particular advantage that there is no need for adjusting the angle of said sensor when changing the distance between the gas nozzle and the surface of the liquid or gel. The laser displacement sensor generates a voltage on its output poles, which is linear with the measured distance to the surface of the liquid and hence, during an excitation, a vibration signal can be recorded between these two poles. Using a data acquisition board, the digitised signal is sent into a computing unit for further treatment. In a preferred embodiment the apparatus starts measuring the vibration of the liquid or gel surface at the moment a gas puff is generated. The laser displacement sensor preferably continues measuring the vibration after the end of the gas puff until said vibration is extinguished. The measurement board digitises the analogue signal from the laser displacement senor with a sample rate of at least 10 Hz, more preferably the sample rate is higher than 100 Hz, for example 1 kHz. In a particular embodiment the sampling frequency was 100 kHz. The method and apparatus of the present invention can be applied to non-destructively test the visco-elastic properties of certain food products such as cheese, butter, yoghurts and puddings as an element in the quality control of these products. For instance the relevant vibration parameters as measured for a given food product can be compared to those parameters as determined for a reference food product having the desired visco- elastic characteristics. A large deviation from the reference vibration parameters will then indicate that the tested product has suboptimal visco-elastic properties.

Furthermore, the method and apparatus of the present invention are particularly useful for monitoring the evolution of the visco-elastic properties of a liquid or gel during processes associated with a change of the said properties. Examples of such processes are coagulation or melting processes or the blending of liquids or gels having different visco-elastic properties. To monitor the evolution of the visco-elastic properties, the liquid or gel surface is exposed to a gas pulse at given time points during a such process in order to estimate the visco-elastic properties at said time points. The analysis, preferably on-line, of the evolution of the obtained values in function of time allows both the monitoring of these properties during such a process as well as the prediction of the further evolution of such a process. Therefore, the method and apparatus of the present invention are valuable tools assisting an operator in taking operational decisions in the course of such processes. The skilled person will understand that when using the system of the present invention in the monitoring of the dynamic behaviour of the visco- elastic properties of a liquid or gel, it is beneficial to maintain the distance between the gas nozzle and the liquid/gel surface constant in order to allow an optimal comparison of the respective measurements.

The method and apparatus of the present invention are particularly useful in the monitoring of the coagulation of milk following the addition of a coagulation-promoting agent, such as rennet. The formation of such coagulum is an important step in the production of cheese and other dairy products. During this process several interventions are required in order to obtain curd having the desired characteristics. Moreover, it is important that these interventions are taken when the firmness of the coagulum has reached certain critical threshold values. Given that the start as well as the speed of the coagulation process vary significantly according to different process parameters, such as the milk composition and that it is very difficult to standardise some of these parameters, it is clear that the automatisation of cheese-making is a challenge. On the other hand, it is shown that the method and apparatus of the present invention allows to non- destructively monitor the visco-elastic properties of the coagulum by repeatedly exposing the surface of the coagulating milk to a gas pulse and by measuring and analysing the induced deformation of said surface. In particular, it was found that the analysis of either or all the maximum (a) and minimum deformation (b) and the total range of deformation (|a|+|b|) (Fig. 10) as derived form the measurements during and after the exposure of the surface to a gas pulse correlated with the firmness of the coagulum. Moreover it was shown that the mathematical relationship between the data obtained for the respective measurements in function of time accurately described the evolution of the firmness of the coagulum. Therefore, in a third object the present invention provides a method and apparatus for the monitoring of the firmness of the coagulum during curd formation. However, it will be clear to the person skilled in the art that the present invention can be used in many different processes, which require the monitoring of the evolution of the visco-elastic properties of a liquid or gel involved in the process. Examples of such processes are melting down of foodstuffs such as cheese and butter and the extraction and concentration of soluble compounds such as sucrose. Furthermore, is was found that the method and apparatus is of interest for the continuous monitoring of the visco- elastic properties of honey during the blending of different types honey at high temperature. The visco-elastic properties of the warm honey allow a good prediction of the firmness and spreadability of the bottled recrystallised honey. The present invention can also be used in the monitoring of the gel time of reaction resin, such as unsaturated polyester resins, epoxy resins, polyurethane resins, Acrylic resins and Silicone resins. Moreover, the method of the present invention can be used to to monitor the visco-elastic properties of liquid paint.

The invention is further illustrated by way of the illustrative embodiments described below.

Illustrative embodiment

EXAMPLES Example 1: Description of an apparatus of according to the present invention

An apparatus according to the present invention was developed and is schematically presented in Figure 9. Said apparatus comprised an air nozzle (1) from which an air pulse (5), for instance a pulse of sterile air, can be released towards the liquid or gel surface (7) and a laser displacement (2) sensor allowing the measurement of the level of the liquid or gel surface. The device was constructed such that it can be installed on a recipient containing a liquid or a gel, whereby the air nozzle (1) is mounted such that an air pulse (5) can be released towards the liquid or gel and perpendicular to the liquid or gel surface (7). To generate such air pulse the air nozzle was operationally connected to a fast valve (Parker PM 139FV, Parker Hannifin, Nivelles, Belgium) (6). Only when the valve is opened, air passes through the opening of the nozzle (1) resulting in an air pulse (5). During a typical activation of the liquid, the valve is opened for 40 μs. To open this valve (6), a pulse is sent automatically by a computing unit (10) through a measurement board (PCI-6024 E, National Instruments, Zaventem, Belgium) to the valve. The air pressure is variable between 0 and 8 bar. A typical set point for liquids is 1.5 bar. The measurement dot of the laser displacement sensor (Acquity AR200-50, Portland, USA) is pointed towards the center of the zone where a released air pulse (5) excites the liquids surface (7). The laser displacement sensor (2) was mounted in line with the airflow (5). Hereto the air nozzle comprised an inner cylindrical (8) tube allowing the transmission of the laser light through the nozzle. The walls of said inner cylinder were closed throughout the nozzle such that the air pulse (5) is released from the outer tube (9). This provided the particular advantage that the measurements of the laser displacement sensor are not disturbed by optical effects associated with air compression surrounding the outlet of the air nozzle. The laser displacement sensor is generating a voltage on his output poles, which is linear with the measured distance to the surface of the liquid and hence, during an excitation, a vibration signal can be recorded between these two poles. Using the data acquisition board, this voltage signal can be sent into a computing unit for further treatment.

In a typical set-up a measurement program comprised in the computing unit (10) being operationally linked to the above described apparatus will at a specified moment open the valve (6) for 40 μs releasing an air pulse (5) inducing a vibration of the surface of the liquid or gel (7). This vibration is registered by the laser displacement sensor (2) and converted in an electronic signal, which is sent via the measurement board in a computing unit (10). Preferably the laser displacement sensor measures the vibration until it is totally extinguished. This can keep up to 15 seconds. The measurement board digitises the analog signal from the laser displacement senor with a sample rate of 100 kHz. From the digitised signal, the amplitude can be measured directly from the time signal or after calculation of the power spectrum. The other vibrational parameters, resonant frequency and damping, can be calculated too. These values are correlated with visco-elastic properties of the measured liquid or gel.

Example 2: Monitoring the coagulation of milk

Experimental set-up

To test the method and apparatus of the present invention, following experiment was carried out. During the experiment, the coagulation of milk was followed from 1.5 till 30 minutes after the addition of rennet. Each 30 seconds, measurements were carried out with the pressure of the air nozzle at 2.5 bar. During the measurements, the recipient containing the milk was placed in a water bath of 33⁰C to maintain the temperature constant. The experiment was repeated four times. For the four experiments, reconstituted 'low heat' 12% milk (INRA Rennes, France) was prepared with a homogenisator (Polytron PT 2000). This milk has a protein content of 4.1% and a total fat content of 1.1%. The pH of this milk is corrected with lactic acid to 6.4. 250μl/l rennet (520mg/kg) was added after warming up the milk to 33⁰C.

For the first two measurement trails, no additives were added to the milk. For the two other trails, 2OmM CaCI₂ was added at least one hour before the start of the measurements.

Treatment of the signals and statistical analysis For each measurement, the maximal amplitude was calculated. On these data, a sigmoid curve was fitted for all four trails separately, using the procedure for non linear regression (proc nlinfit) of the SAS 6.12^® for windows software package. During the fitting process, the statistical program will estimate values for the parameters A, B, C and k. Each of these parameters can be interpreted from the signal, as can be seen in figure I.For each trail, these parameters were calculated.

Reference technique: Rheology

The rheological properties of the milk were measured using Rheostress RS 150 equipment in combination with a thermostation (Haake F6).

Reference technique: Gelograph

A gelograph provided by Gel lnstrumente AG, using Almeno 8990-1 a measuring cell was used.

Comparison of the results obtained using the three technologies To be able to compare the three technologies (the method of the present invention, rheology and the gelograph), the starting point of the coagulation process is calculated from the measured data and the point at which 90% of the coagulation is measured.

Results Figure 2 presents some examples of the surface deformation signals, measured by the laser displacement sensor after the exposure of said surface to an air pulse.

In figure 3, the evolution of the maximal amplitude is presented for each of the four measurement trails. The sigmoid plot is presented too (full lines). The parameters A, B,

C and k can be found back in table 1. Since all the four monitored processes start from the same product (not coagulated milk), the A value is estimated with the total of all data and is taken constant for all fitted sigmoids. Besides the classical sigmoid values A, B, C and k, the A+B value is calculated. This value indicates the maximal amplitude at the end of the monitored period.

It is clear that these data are well fitted by the sigmoid curve. Obviously, the CaCI₂ enriched milk started to coagulate earlier than the normal milk. The speed of the coagulation process (indicated by the k-value) was higher in the CaCI₂ process.

Moreover, at the end of the process, the values of the maximal amplitude in the CaCI₂ enriched milk were smaller compared with the normal milk. In table 3, the time points of the start and 90% end of the coagulation process are presented. Here again, it is clear that the CaCI₂ enriched milk started earlier to coagulate. Moreover, the coagulation process ended earlier. The results of the first reference technique, the use of rheology for monitoring the coagulation are presented in figure 4 and table 3. In this figure, the mean curves for standard milk and CaCI₂ milk are presented. Obviously, just before the coagulation started, the G' value decreased. The coagulation process started earlier in the CaCI₂ enriched milk. Finally, the elastic modulus increase more in the CaCI₂ enriched milk. In table 3, the start point of the process as well as some other characteristics of the evolution of G^J can be seen for the 5 repetitions (2 with normal milk and 3 with CaCI₂ enriched milk) of this measurement.

The results of the second reference technique, the gelograph, are presented in figure 5 and table 4. In the figure, all the three measurements are presented. Here again, it is clear that the CaCI₂ enriched milk started to coagulate earlier than the untreated milk. Contrary to the other two techniques, there is no difference between the two types of milk at the end of the measurements. Obviously, in the time period before the start of the coagulation of the milk, a higher variance on the measurements could be noticed. In table 5, the characteristic values of the evolutions of the Gelograph output are presented. A concluding comparison of the three techniques for both the normal and the CaCI₂ enriched milk is given in table 5. For both milk types the starting point of the coagulation as determined with the method of the present invention corresponded to the starting point as determined using the rheograph. On the other hand, the starting point of coagulation as determined using a gelograph was consistently later.

Discussion

The presented data show that the maximal amplitude of the vibration measured during and after excitation of the coagulating milk surface with an air puff decreased sigmoidal from the moment that the coagulation process starts. Moreover, the calculated sigmoidal parameters, accurately described the coagulation process. The coagulation of the CaCI₂ enriched milk is known to start earlier and this process result in more solid curd. This earlier start of the process was reflected by lower values for the C parameter of the sigmoid fits in the experiments with CaCI₂ enriched milk, compared with the normal milk. At the end of the coagulation process, the B parameter in the CaCI₂ enriched milk is more negative and hence the A+B parameter is lower than the one obtained for the coagulation process in normal milk, indicating a relatively lower final destination of the sigmoid curve at the end of the coagulation process. This is in agreement with the more solid CaCI₂ curd at the end of the coagulation process. Whether one considers the evolution of the total sigmoid or only the values at the start and at 90% coagulation of the milk as presented in table 1 , the same conclusions can be drawn about the start point and the endpoint of the process. Looking at the results of the first reference method, the measurement of the elastic modulus with the rheometer (figure 4 and table 3), the same conclusions could be drawn. The results of these measurements detected an earlier increase of the elastic modulus of the milk, caused by the coagulation of the milk proteins in the CaCI₂ enriched milk. After finishing the coagulation process, the G' values of the CaCI₂ enriched milk were larger, indicating a more viscous curd, what is in accordance with data found in the literature and the results from the method of the present invention.

Contrary to the maximal amplitude measured with the method of the present invention, the G' value measured by the rheometer is not describing a sigmoid course, but a steadily increasing curve, due to the logarithmic Y-axis looking as an logarithmic curve. Therefore, the end of the coagulation process could not be clearly detected and hence, it was more difficult to estimate the time point for the 90% coagulation process. Probably, this may complicate the use of a rheometer in the daily practice of monitoring the coagulation process.

The data obtained using a gelograph (figure 5 and table 4), also allowed to observe the earlier start of the coagulation process in the CaCI₂ enriched milk. However, at the end of the coagulation process the output data of the gelograph revealed no differences between the two milk types. Whereas the evolution of the measured values by the other two techniques could be described by one mathematical function (a sigmoid in the method of the present invention and a polynomial in the rheology), the data measured by the gelograph are not suitable to fit using a single mathematical function. This may indicate that during the time before the onset of the coagulation process measured by the gelograph, this device is not accurate enough to detect the small changes in the visco-elastic properties of the milk starting to coagulate and hence, the gelograph will detect the onset of the coagulation process only if the process is already ongoing. Comparing the values obtained at the start of the coagulation process and at 90% coagulation for the three different measurement techniques (table 5), it is clear that the start of the coagulation process is detected at the same moment by the rheology application and the method of the present invention. Both for the normal milk and the CaCI₂ enriched milk, the gelograph detected the start of the coagulation later which may be due to a lack of sensitivity of this measuring technique. When looking at the 90% there were no significant differences between the three techniques. Here, the time points calculated from the measurements with the rheology application were the highest, but this may be caused by the difficulty to estimate an endpoint of the coagulation here. As a conclusion, the method of the present invention accurately measures the changing visco-elastic properties of coagulating milk.

Example 3: Measurements of the melting Cancoillotte Maison Bernard

Materials and methods

Using the food texture measurement technique of the present invention, the behaviour during heating down of Cancoillotte (Maison Bernard) was investigated. Therefore, an amount of Cancoillotte cheese was heated up till 75⁰C. During the heating down period, 7 measurements were done. At each measurement point, two puff measurements (resulting in two deformation curves) were done on the sample and the temperature of the cheese was registered. The temperature at each time point is presented in table 6. From each deformation curve, the initial maximal intrusion value and 'jump back value' were calculated (as presented in figure 7). The sum of both values, i.e. the total span was calculated too. Thereafter, from the time signals the frequency spectrum was calculated using an FFT-approach with a 4096 broad window. From the frequency spectrum, the amplitude, the resonant frequency and the damping of the main vibration were calculated. These calculations were done with the Matlab software. All these characteristics were used as dependent variables in statistical models with the temperature of the cheese as the independent variable. Both the linear and the quadratical fit were tested for significance. Hence, for each characteristic this approach resulted in two P and R² values for each model. The statistical analysis was done using the SAS 6.12 software. Results and Discussion

In figure 6, the time signals are presented. It is clear that a vibrational signal is registered by the laser displacement sensor after excitation of the heated product. For the measurement after 2 minutes, no meaningful signal could be registered hence this time point was skipped from further analyses.

Table 7 presents the results of the statistical fits on the data. The model with the highest correlation factor was the quadratic model of the damping as dependent variable (R²=0.93). Another good model was the linear model with the jumpback-value as dependent value. Thereafter, the linear model of the amplitude in the frequency domain and total span were significant, but the correlation decreased. The intrusion value revealed a model with a seriously decreased significance, whereas the resonant frequency didn't show any temperature influence.

The most significant model is presented in figure 8. It is clear that the damping of the vibration in the heated cheese decreased quadratically with increasing temperature.

Conclusion

Due to the heating of the Cancoillotte cheese, its viscosity parameters changed. The change of these parameters can be found back in the vibrational parameters, which also changed with the temperature. The main change was in the damping of the main vibration excited by the puff in the heated product. Theoretically, a changing damping of a vibration will influence the jumpback value, and therefore, a significant effect can be found in this value with changing temperature. Moreover, since the total span value is calculated from the jumpback value, a significant changing damping value might be translated in a changing total span value. A significant correlation between the amplitude of the vibration and the temperature was also found. Since the intrusion value can be seen as an estimator in the time domain of the amplitude, it is normal that also this value shows a significant relation with the temperature.

Example 4: The Effect of Milk pH on Coagulation Process measured by the Non Destructive Foodtexture Puff Device Introduction

The first step in the production of cheese and other dairy products is the production of a coagulum by adding rennet to the standardized milk. During this process, κ-casein is hydrolyzed by rennet enzymes which cause the protein micelles to aggregate. As a result, the milk is transformed into a gel matrix that surrounds fat globules and bacteria: the coagulum. When a certain coagulum firmness is reached, it is cut mechanically into small pieces to facilitate the expulsion of whey (syneresis). This is further accelerated by the acidification of the mix by bacteria and an increase of the process temperature (Payne et al, 1993). The exact moment of cutting the coagulum is critical for the quality of the final curd and therefore on the cheese that is produced from this curd. Cutting too late will reduce the syneresis and the moisture content of the cheese will increase. Cutting too early may causes the loss of fat and curd fines (Mayes and Sutherland, 1984; Riddell-Lawrence and Hicks, 1988).

Due to the subtle variations in milk properties and the wide range of standardization procedures that plays a role in the coagulation process (e.g. protein and fat content, initial pH, temperature, enzyme concentrations, CaCI₂ content and bacterial load (Lopez et al, 1998; Landfeld et al, 2002; Najera et al, 2003)), the exact start point of the gelation process and the speed of the process are variable. Therefore, for automation of cheesemaking, a measurement technique is needed for the prediction of the cutting time or to adjust the milk towards a standard product in that way that always the same cutting time can be used (Lucey, 2002). O'Callaghan et al. (2002) lists the requirements for such a device. First, it has to monitor the curd firmness which is not accessible to current thermal or optical sensors. Second, it has to operate on a cheese vat, instead of separate measurements on samples. Third, the measurement should not interfere with the coagulation process and fourth, the device has to be up-to-date with dairy hygiene design requirements. Lots of research has already been performed to develop a suitable device to monitor the formation of the coagulum and to predict the right cutting time. The current status of the existing technology is reviewed by O'Callaghan et al. (2002) and Lucey (2002). They both describe different technologies that were tested as a prototype for online measurements based on mechanical, vibrational, ultrasonic, electronic conductivity, hot wire or optical measurements. From these only the hot wire presented by Hori (1985), the fibre-optic reflectance (Ustunol et al., 1991) and the RheoLight sensor developed by Nizo (the Netherlands) became commercially available. Recent research however pointed out that the hot-wire technique is able to measure the onset of the gelation process, but is not able to measure the rate of the curd firming and even not the final firmness (O'Callaghan et al., 2002). The optical methods have been shown to be very powerful, but since they need a continuous calibration process, their actual use is yet quite limited. The RheoLight sensor seems to be useful, but its price is rather high.

In the present manuscript, a Foodtexture Puff Device (FPD) is used to monitor the rennet induced coagulation process of milk. The technique is based on the measurement of the dynamic deformation of the surface of the coagulating milk after excitation with an air puff, an idea presented by Prussia (1995) to test the firmness of fruits. As advised by research of Lopez et a/. (1998) rheologic measurements are used as a reference technique. In the described experiments, the effect of the initial milk pH on the coagulation process is monitored using both techniques.

Materials and method

Materials: To standardize the experiments, 71 reconstituted milk was prepared from commercial Low fat-Low heat skimmed milk powder (Belgomilk^®, Langemark, Belgium) at 12% (w/v) in deionized water at room temperature. After reconstitution, the milk was stored in a refrigerator at 4⁰C. An hour before the beginning of a measurement a sample of 11 was poured in a glass beaker and equilibrated at room conditions. Subsequently, the pH of the sample was adjusted with lactic acid (88% solution, VWR International, Haasrode, Belgium) and controlled with a digital pH meter (WTW pH-meter lnolab pH level 1). The pH was adjusted to different values, such that the range between 6.61 and 6.01 was covered with 7 samples. Once the pH was stabilized, the sample was equilibrated in a water bath to 33°C. The preparation of each sample took not longer as one hour. After the milk was equilibrated at 33°C, 250μl/l single strength rennet (80% chymosin, 20% pepsin, BMS, Kuurne) was added and the time recorded to synchronize the rheology and FPD measurements. From the 11 sample, 100ml was taken to be used in the rheology measurements. The remaining 900ml stayed in the water bath where the FPD measurements took place. Each sample was monitored for one hour with the FPD and the rheometer. All measurements were performed on the same day.

The Foodtexture Puff Device (FPD) measurements: The FPD generates during 50ms an air puff of 0.10 bar nozzle pressure that is directed towards the surface of the coagulating milk using an air nozzle. The end of the nozzle is placed 4cm above the milk surface (see figure 9). From the moment that the air puff is generated, the deformation of the milk surface is recorded by a laser distance sensor during a period of 5s after the excitation of the surface. The deformation data are sent to a PC via a data acquisition board (E-6024 Low cost Data Acquisition board, National Instruments^®, Zaventem, Belgium) and stored on the hard disk for later analysis. The used measurement head constructed by LET nv (Deinze, Belgium). In figure 9, a drawing of the device is shown. When analyzing the data, the maximum and minimum deformation and the maximum range of deformation were calculated from the deformation as a function of a wave (see figure 10), since the deformation shows up as a wave. In fact, the minimal value in the deformation curve corresponds with the maximal deformation of the milk surface (= maximal intrusion) under a constant force of the air puff. The maximal value in the deformation curve range corresponds with the height at which the milk bounces back after removal of this constant force. The total range is the summation of both values. The software that was used to communicate with the measurement device was written in the graphical programming language Labview 5.3 (National Instruments^®, Zaventem, Belgium). The calculation of maximal and minimal values and the range of the deformation curve were done with Matlab 6.1 (The Mathworks, Inc). The coagulation process was monitored with this device during the first hour after renneting. Four measurements were done per minute.

Rheological Measurements: Gel formation was monitored using a controlled stress rheometer (TA Instruments^®, AR1000-N, Delaware, USA) with concentric cylinder geometry (rotor diameter: 25mm, stator diameter: 30mm, gap height: 6418μm). The temperature of the sample was kept constant at 33⁰C by a controlled temperature water flow through the envelope of the outer cylinder and a Peltier temperature controller in the bottom plate on which the concentric cylinder system was placed. The storage modulus (G') and the viscous modulus (G") were recorded at a frequency of 1 Hz under 0.015 strain amplitude during the first hour of the coagulation process. This was found to be within the linear viscoelastic region for rennet milk gels (Lopez et a/., 1998; Herbert et al., 1999). A measurement was done each 10 seconds. The rheological parameters were calculated using the TA Instrument software and stored on the hard disk of a PC for further analysis.

Data treatment: To illustrate the different steps in the treatment of the recorded data, the data of the coagulation process at pH=6.29 are used as an example (see figure 11). For both the FPD and the rheological parameters, it can be seen that the recorded time signals can be split up in two parts: a constant value before the gelation process starts followed by a hyperbolic decrease for the total range and the maximal value of the FPD measurements and a hyperbolic increase in the minimal value of the FPD measurements and the G' and G" registered by the rheometer. On these data, equation 1 was fitted with the SAS software (SAS 6.12, Cary, NC, USA). This equation is a combination between a time constant value (the first term) followed by a hyperbolical change (the numerator of the second term). The denominator of the second term realizes the shift from the constant part to the hyperbole part at the time point C/k. k is the rate at which the shift from the first term towards the second is realized. For this work, k is chosen constant and rather high (0.5) to realize a fast shift from the constant towards the hyperbolic regime.

Eq. 1

With: A : the initial value of the measured property [mm]

B :the change of the initial value at the end of the process of the measured property [mm]

C: an estimator of for the time shift of the point at which shift /s made

D & E: estimators for the asymptotes of the hyperbolic phase

[mm/s] and [mm^'1] k: the rate at which the shift is made from the first (constant) phase to the second (hyperbolic) phase of the curve, here chosen 0.5 [mm^'1] t: time [s] f(t): the registered characteristic that is fitted

The advantage of this approach is that the estimated parameters of this equation have a descriptive interpretation (see figure 12). The time point of the start of the gelation process is estimated by 2*C. The initial value of the measured characteristic is given by A whereas the final value is given by A-B. The rate of change of the modeled parameter during the gelation process is estimated by (2*C-E/D)^"1. Since the correlation coefficients of the fits on the measured data were quite high (see table 8), these parameterizations were accepted.

This fit is calculated for all five monitored characteristics (i.e. minimal and maximal value and total range for the FPD measurements and G' and G" for the Theological parameters) for each of the seven monitored gelation processes at different initial milk pH.

Statistical analysis: For the fitting of equation 1 on the registered data, a non linear iteration approach (proc nlinfit) was used in the SAS software. Not more as 30 iterations were needed to find the right model parameters. Pearsons correlations coefficients between the model parameters of the different measured characteristics were calculated with the SAS software. Finally, relationships between the calculated model parameters and initial milk pH were investigated with a general linear model approach of the SAS software.

Results

In table 8, the estimated process parameters after fitting eq. 1 on the evolution of the different registered characteristics are given for the 7 monitored gelation processes at the different initial milk pH. Except for the G" measurements on the milk of pH 6.61 , the R² of each fit was found to be higher as 0.97 and hence, they were accepted. On the evolution of the G" of the milk at pH 6.61 , no good fit could be made. A high correlation could be found between the start points (2C) of the gelation process as calculated from the three characteristics of the FPD measurement and those registered by the classic rheology parameters G' and G" (see table 9). Moreover, for each characteristic separately, an exponential increase with the pH for the 2*C value (see test statistics in table 8, right hand side) was found. These exponential models are presented in figure 13. From this figure, it can be concluded that all the registered characteristics start to change from the same time point during the gelation process, dependent on the milk pH. Neither for the new characteristics that are calculated from the FPD measurements, nor for the classic rheology parameters G' and G", a good statistically significant relationship between their value at the end of the gelation process (i.e. the A-B value) and the initial pH of the milk could be found. This means that the gelation processes at all the different pH evolve towards the same state at the end of the process, at least for the characteristics measured here.

The rate of the change of the measured properties during the gelation process could be estimated by (2C-E/D)^"1. Here, an exponential increase of this rate was found with the increasing pH for all characteristics except for the range of the FPD signal. This can be concluded from table 8 as well as from the correlation table presented in table 9. The exponential increase is presented in figure 14. Remark that the value of the (2C-E/D)^"1 process estimator (rate of the process) is different for each measured characteristic, whereas the 2C estimator (start of the gelation process) gives an absolute value that is the same for each characteristic. However, since the (2C-E/D)^"1 parameter of the modeled changes in maximal and minimal deformation values are well correlated with G' and G" (see table 9), they make the estimation of the rate of the gelation process possible. Discussion In the described experiments, the gelation process of milk at different initial pH between 6.61 and 6.01 is monitored with a newly developed technique named FPD and with a classic rheometer. From the FPD measurements, the course of the maximal and minimal deformation and the total deformation under a constant force provided by an air puff is recorded. The measurements of this novel technique were compared with the classical rheology parameters G' and G".

A non linear equation was fit to these different variables to analyze their kinetics after rennet addition to the milk. This approach is different from the classic analyses where a coagulation process may be characterized by only one value as there are the gelation time where G'=G" (Gastaldi et al., 2003), the flocculation time or rennet coagulation time (RCT) (Gunasekaran and Ak, 2003) or the time point for maximal deformation of G' (Lopez et al., 1998). In other research of Landfeld et al. (2002) or by fitting the Scott-Blair and Brunnett model (Scott-Blair and Brunnett, 1963; Daviau et al., 2000), an equation is presented that may be used to fit the course of G' and G" after the start of the gelation point. In our research, a new non linear model was used that is able to fit the course of G' and G", as well as the evolution of the parameters given by the novel FPD, with high precision (see figure 3 and table 8) from the moment that the rennet was added to the endpoint of the gelation process. This model divides the course of the registered parameters in two different parts. The first part is the lag phase during which the registered property stays constant. The second part is the part with a hyperbolic change of the registered property. Another advantage of the proposed model is the direct descripive interpretation of the model parameters. The start point of the gelation process, the process speed and the end point can be directly calculated from the estimated model parameters. The three characteristics measured by the FPD (maximal and minimal deformation values and deformation range) were modeled very precisely using the proposed model, and this for all the gelation processes observed in the presented research. The time point of the start of the gelation process was estimated by the 2C-value in eq. 1. With decreasing pH of the milk, the coagulation process started earlier. The same exponential relationship between initial milk pH and start point of gelation was found for all the monitored properties (see figure 14). This exponential relationship between the rheology parameters was presented before in research carried out by Daviau et al. (2000). In our research, the 2C-E/D value increased exponentially with increasing pH for the G', G", the FPD minimal value and the FPD maximal value. Since the 2C-E/D value can be seen as an inverse estimator for the gelation rate, it may be concluded that the gelation rate decreased exponentially with increasing pH. Again, this is consistent with previous research (Daviau et al., 2000; Lopez et al., 1998). The values of the measured properties at the end of the coagulation process, estimated by A-B in the model, was not influenced by the pH of the initial milk. Also in the literature no evidence is found for such a relationship.

Altogether, the newly developed technology is able to determine the time of coagulation, the firming rate and the final firmness of the coagulum, that is, all the relevant characteristics of the formation of coagulum from milk in a reliable way consistent with the most rigorous rheologic approach. Whereas other sensors were shown able to estimate the start of coagulation process too, the FPD is the first device that is able to monitor properly the firming process kinetics in a non destructive way. The estimation of coagulation characteristics is made in a straightforward manner by plain analysis of the signal and does not require specific calibration or the building of a data base such as techniques based on neural network analysis (Acuna et al, 1999). Moreover, since this novel technique is non-destructive, no sampling is needed when placed over a industrial cheese vat and it can be constructed conform with the today's hygienic design requirements. The criteria formulated by O'Callaghan et al. (2002) for an automated measurement technique are therefore fulfilled. In addition a model is proposed that can describe model the total gelation process. This makes the FPD a technique of interest for the development of a fully automated sensor head to monitor the coagulation process in the dairy industry.

References

Acuήa, G., F. Cubillos, J. Thibault, and E. Latrille. 1999. Comparison of methods for training Grey-Box Neural Network Models. Comp. & Chem. Eng. Suppl. 23:S561-S564.

Ay, C. & Gunasekaran, S. (1994). Ultrasonic attenuation measurements for estimating milk coagulation time. Transactions of the ASAE, 37:857-862.

Banon, S. & Hardy, J. (1991). Study of acid milk coagulation by an optical method using light reflection. Journal of Dairy Research, 58:75-84.

Benguigui, L., Emery, L., Durand, D. & Busnel, J-P. (1994). Ultrasonic study of milk clotting. Le Lait, 74: 197-206. Beudon, D., Audidier, Y., Frouin A. & Innocent, J-P (1992). Device and method for the detection of changes of phase of a product. FR92 2 693 271.

Cosgrove, N (2000) Design and development of an ultrasonic measuring system to monitor milk coagulation. Master Thesis, National University of Ireland, Dublin - Ireland;

Daviau, C, M-H. Famelart, A. Pierre, H. Goudedranche, and J-L. Maubois. 2000. Rennet coagulation of skim milk and curd drainage: effect of pH, casein concentration, ionic strength and heat treatment. Lait. 80:397-415.

Dejmek, P (1989) Precision conductometry in milk renneting. Journal of Dairy Research, 56:69- 78.

Dybowska, B. E. & Fujio, Y. (1996). Effect of temperature and clucono-D-lactone (GDL) concentration on milk aggregation and gelation process as revealed by optical method. Milchwissenschaft, 51:557-560.

Gastaldi, E., N. Trial, C. Guillaume, E. Bourret, N. Gontard, and J. L. Cuq. 2003. Effect of controlled κ-casein hydrolysis on Theological properties of acid milk gels. J. Dairy Sci. 86:704-711.

Gunasekaran, S., and M. M. Ak. 2003. Cheese Rheology and Texture. CRC Press, London. Hardy, J. & Fanni, J. (1956). Application of reflection photometry to the measurement of milk coagulation. Journal of Food Science, 46:1956-1957.

Hatfield, D.S. (1981). A new instrument to measure cheese curd rigidity, an preliminary trials in cheesemaking. Journal of Society of Dairy Technology, 34:139-142.

Herbert, S., A. Riaublanc, B. Bouchet, D.J. Gallant, and E. Dufour. 1999. Fluorescence Spectroscopy investigation of Acid- or Rennet-Induced coagulation of milk. J. Dairy Sci. 82:2056- 2062.

Hori, T. 1985. Objective measurements of the process of curd formation during rennet treatment of milks by the hot-wire method. J. Food Sci. 50(4):911-917.

Home, D.S. & Davidson, CM. (1990). The use of dynamic light-scattering in monitoring rennet curd formation. Milchwissenschaft, 45:712-715.

Kowalchyk, A.W. & Olson, N. F. (1978). Firmness of enzymatically-formed milk gels measured by resistance of oscillatory deformation. Journal of Dairy science, 61 :1375-1379.

Lagouette, N., Lagaude, A. & de Ia Fuente, BT. (1995). Rheological properties of renneted reconstituted milk gels by piezoelectric viscoprocess: Effects of temperature and calcium phosphate. Journal of Food Science, 60:1363.

Landfeld A., P. Novotna, and M. Houska. 2002. Influence of the amount of rennet, Calcium chloride addition, temperature and high pressure treatment on the course of milk coagulation. Czech J. Food Sci. 20(6):237-244.

Lopez, M. B., S. B. Lomholt, and K.B Qvist. 1998. Rheological properties and cutting time of rennet gels. Effec of pH and enzyme concentration. Int. Dairy J. 8:289-293 Lucey, J.A. (2002). Formation and physical properties of milk protein gels. Journal of Dairy Science, 85:281-294.

Mayes, J.J., and B.J. Sutherland. 1984. Coagulum firmness and yield in Cheddar cheese manufacture. Aust. J. Dairy Technol. 39:69-73. McMahon, D.J. & Brown, R.J. (1982). Evaluation of Formagraph for comparing rennet solutions. Journal of Dairy Science, 65:1639-1642.

McMahon, D. J., Brown, R.J. & Ernstrom, CA. (1984a). Effects of calcium, phosphate and bulk culture media on milk coagulation

Mehra, R., Ozimek, L., Gilligan, M. & Connolly, J. F. (1994). Assessment of the Amelung- Coagulometer, Formagraph and hot wire technique for measurement of rennet clotting time. Proceedings of IDF Seminar on Cheese yield and foactors affecting its control. Cork, Ireland April 1993. International dairy federation special issue, 9402:333-338.

Najera, A.I., M. de Ronabales, and L.J. R Barron. 2003. Effects of pH, temperature, CaCI₂ and enzyme concentrations on the rennet-clotting properties of milk: a multifactorial study. Food Chem. 80:345-352.

O'Callaghan, D. J.; O'Donnell, CP. & Payne, F.A. (2002). Review of systems for monitoring curd setting during cheesemaking. International Journal of Dairy Technology, 55(2):65-74.

Payne, F.A., CL Hicks, S. Mandagopal, and S.A. Shearer. 1993. Fiber optic sensor for predicting the cutting time of coagulating milk for cheese production. Trans. ASAE. 36(3):841-847 Prussia, S. 1995. Non-destructive firmness measurement device. World patent WO 095/07772

Reddy, R. (1992). Monitoring cheese milk setting by Gelograph. PhD thesis, Mississippi State University, USA.

Riddell-Lawrence, S., and CL. Hicks. 1988. Effect of curd firmness on stirred curd cheese yield. J. Dairy Sci. 72:313. Scher, J. & Hardy, J. (1993). Study of the variations in particle size distribution after renneting by means of quasi-electric light scattering. Australian Journal of Dairy Technology, 48:62-65.

Schulz, D., Senge, B. & Krenkel K. (1997a). Beitrag zur Beschreibung des Strukturierungsverhaltens fon Milch bei der Labgerinnung. 1. Mess- und Auswerteverfahren. Milchwissenschaft, 52:243-246. Schulz, D., Senge, B. & Krenkel, K. (1997b). Beitrag zur Beschreibung des Strukturierungsverhaltens von Milch bei der Labgerinnung 1. Einfluss verschiedener Gerinnungsparameter und Strukturierungsmechanismus. Milchwissenschaft, 52:303-306.

Scott-Blair, G.W., and J. Burnett. 1963. An equation to describe the rate of setting of blood and milk. Biorheology. 1:183-191. Senge, B. Schulz, D. & Krenkel K (1997). Structure formation in milk during rennet coagulation. Applied Rheology, 7:153-160.

Sharma, S. K., Hill, A.R., Goff, H. D. & Yada, R. (1989). Measurement of coagulation time and curd firmness of renneted milk using a Nametre viscosimeter. Milchwissenschaft, 44:682-685.

Ustunol, Z., Hicks, CL. & Payne, F. (1991). Diffuse reflectance profiles of eight milk-clotting enzyme preparations. Journal of Food Science, 56:411-415.

Van Hooydonck, A.C.M. & Van den Berg, G. (1988). Control and Determination of the Curd Setting during Cheesemaking. IDF Bulletin No 225. Brussels: lnternation Dairy Federation. Table 1: Values for all 4 coefficients after fitting the sigmoid curve on the data for each experimental trail.

Table 2: Start of the coagulation process and 90% finishing moment of the coagulation process, calculated on the individual sigmoid fit on the data of each process.

Table 3: Time point of start of coagulation, the time at which G'= 20Pa, the maximal G' value and the 90% coagulation time point as measured by the rheology technique.

Table 4: Time of the first detection of coagulation using the gelograph for both the normal milk and the CaCI₂ enriched milk).

Table 5: Mean start time and 90% coagulation time as measured by the three different techniques in minutes. Means marked with the same letter are not significant different (p<0.0001).

Table 6: Timepoints of measurements and the registered temperature during the measurement period

Table 7: Results of statistical fits on the data. For each characteristic, the 1^st and 2^nd order P-value and correlation (R²) with the temperature is shown. For the significant model with the highest order, for each characteristic the model parameters are presented

Table 8: Estimated process parameters after fitting equation 1 on the evolution of the different registered characteristics for the 7 measured gelation processes. In the bottom line, the R² of the fit is given. At right, the test parameters for the effect of pH on the characteristics are given. 2C= start of the process, A-B: final value of the characteristic, 2C-E/D: speed of the gelation process

G¹

PH 6.01 6.11 6.21 6.29 6.41 6.51 6.61 P value R²

2C*** 392 499.8 590.4 792.2 1140 1401.6 2160 O.0001 0.984 A-B* 2.5594 2.5497 2.5668 2.5621 2.5021 2.4546 2.2842 0.0228 0.67 2C-E/D** 148.2439 117.6922 168.8848 210.3538 218.2292 355.8308 510.1786 0.0007 0.958

R² 0.99 0.99 0.99 0.98 0.97 0.96 0.98

G"

PH 6.01 6.11 6.21 6.29 6.41 6.51 6.61 P value R²

2C*** 413.6 541.6 623.6 840.8 1282.2 1499.6 <0.0001 0.988 A-B 1.9876 2.0055 2.0049 2.0221 2.0024 1.9374 - 0.3546 2C-E/D** 189.5406 206.6562 234.3051 321.231 465.7 550.9043 0.0019 0.876

R² 0.99 0.99 0.99 0.99 0.99 0.97

FPD Minimal deformation value pH 6.01 6.11 6.21 6.29 6.41 6.51 6.61 P value R²

2C*** 385.6 470 590.6 832.6 1200 1379 2009.6 <0.0001 0.989 A-B -0.2016 -0.1242 0.038 0.1154 -0.443 -0.6237 -0.4349 0.124 2C-E/D*** 270.475 316.7164 402.5643 656.0286 931.1111 1018.167 2105.055 <0.0001 0.961

R² 0.99 0.99 0.99 0.99 0.99 0.99 0.99

FPD Maximal deformation value pH 6.01 6.11 6.21 6.29 6.41 6.51 6,61 P value R²

2C*** 409.4 500 585.8 780.4 1193.8 1369.2 2020 <0.0001 0.9850 A-B 0.0765 0.0342 0.0484 0.0321 0.0434 0.045 -0.0239 0.0679 2C-E/D** 9.055647 25.12667 24.27977 36.35238 67.80651 78.7 148.8517 0.0002 0.9530

R² 0.96 0.96 0.97 0.96 0.98 0.97 0.96

FPD Deformation range

PH 6.01 6.11 6.21 6.29 6.41 6.51 6.61 P value R*

2C*** 426.6 503.8 536 802.8 1118 1300 2009.6 <0.0001 0.968 A-B* 0.3554 0.3173 0.1604 0.117 0.975 1.4084 2.0504 0.013 0.510 2C-E/D 217.9136 205.6735 199.9831 416.2091 338.5128 252.0646 359.6 0.148

R² 0.98 0.98 0.99 0.99 0.99 0.98 0.98

Table 9: Pearson's correlation coefficients between the FPD parameters and G'

10 and G".

G¹ G" Physical

Mini Maxi Range Mini Maxi Range interpretation

2C 0.996 0.998 0.999 0.998 0.999 0.997 start of process A-B 0.644 0.828 -0.956 0.821 -0.279 -0.820 texture at end 2C-E/D 0.964 0.959 0.461 0.989 0.986 0.399 speed of process

Claims

1. A method for the measurement of the visco-elastic properties of a liquid or gel comprising the steps of: (i) generating a gas pulse directed towards the surface of a liquid or gel,

(ii) measuring the deformation of said surface during and/or following the air pulse in order to record the vibrations generated on said surface by the air pulse, (iii) digitising the measured signals, (iv) analysing the digitised signals in order to determine vibration parameters correlating with the visco-elastic properties of said liquid or gel.

2. The method according to claim 1 comprising the additional step of:

(v) calculating and presenting a value reflecting the visco-elastic properties of . said liquid or gel using said determined vibration parameters and information on the relation between said vibration parameters and the visco-elastic properties of such liquid or gel.

3. The method according to claims 1 or 2 wherein the selected vibration parameters are selected out of the group consisting of the power spectrum, the resonant frequency, the amplitude, the maximal deformation and damping.

4. The method according to any of the preceding claims wherein the gas pulse is emitted perpendicular to the plain of the surface of the liquid or gel.

5. The method according to any of the preceding claims wherein the pressure of the gas pulse is selected between 0.1 an 8 bar.

6. The method according to any of the preceding claims wherein the vibration of the liquid or gel surface is measured using a laser displacement sensor.

7. The method according to claim 6 wherein the laser displacement sensor is mounted in line with the direction of the gas pulse.

8. The method according to claim 6 wherein the laser displacement sensor is mounted in angle with the direction of the gas pulse.

9. The method according to claims 6 to 8 wherein the measurement dot of said laser displacement sensor is pointed on the zone where the gas pulse excites the gel or liquid surface.

10. The method according to claim 9 the measurement dot of said laser displacement sensor is pointed on the centre of the zone where the gas pulse excites the gel or liquid surface.

11. The method according to claims 1 to 10 wherein the gas pulse is an air pulse.

12. The method according claim 11 wherein the gas pulse is a sterile air pulse.

13. The use of the method according to claims 1 to 12 to assess the visco-elastic properties of a food product having a liquid or gel nature.

14. The use of the method of claims 1 to 12 according to claim 13 wherein the food product is a dairy product.

15. The use of the method of claims 1 to 12 according to claim 13 wherein the food product is honey.

16. The use of the method according to claims 1 to 12 to assess the visco-elastic properties of paint.

17. The use of the method of claims 1 to 12 to monitor the visco-elastic properties of a liquid or gel in the course of a process, which is associated with a change of the visco-elastic properties of said liquid or gel.

18. The use of the method of claims 1 to 12 according to claim 17 wherein the liquid or gel surface is exposed to repeated gas pulses in the course of said process and wherein the vibration parameters of the induced surface vibration are determined for each gas pulse.

19. The use of the method of claims 1 to 12 according to claim 18 wherein the relation between said determined vibration parameters is determined in function of time.

20. The use of the method of claims 1 to 12 according to claim 18 wherein the relation between said determined vibration parameters is determined in function of temperature.

21. The use of the method of claims 1 to 12 according to claims 17 to 19 wherein said process is coagulation or gelling process.

22. The use of the method of claims 1 to 12 according to claim 21 wherein said process is the coagulation of milk following the addition of a coagulating agent.

23. The use of the method of claims 1 to 12 according to claim 22 wherein the determined vibration parameters are the maximum deformation, minimum deformation and the total range of deformation.

24. The use of the method of claims 1 to 12 according to claims 17 to 19 wherein said process is a melting process.

25. The use of the method of claims 1 to 12 according to claim 24 wherein said process is the melting of cheese.

26. The use of the method of claims 1 to 12 according to claims 17 to 19 wherein said process is a blending process.

27. The use of the method of claims 1 to 12 according to claim 26 wherein said process is the blending of different types of honey.

28. The use of the method of claims 1 to 12 according to claim 17 to 19 wherein said process is an extraction process.

29. The use of the method of claims 1 to 12 according to claim 17 to 19 wherein said process is a concentration process.

30. The use of the method of claims 1 to 12 according to claim 17 to 19 wherein said process is a gelling of a reaction resin.

31. An apparatus for the assessment of the visco-elastic properties of a liquid or gel comprising of

(i) a means for generating a gas pulse towards a liquid or gel surface, (ii) a means for measuring the vibrations induced on the liquid or gel surface during and/or following the air pulse,

(iii) means to digitise the measured signals and

(iv) means to analyse the digitised signal in order to determine the vibration parameters correlating with the visco-elastic properties of said liquid or gel.

32. The apparatus according to claim 31 further comprising means to calculate and present a value reflecting the visco-elastic properties of said liquid or gel using said determined vibration parameters and information on the relation between said vibration parameters and the visco-elastic properties of such liquid or gel.

33. The apparatus according to claim 31 wherein said means to generate a gas pulse is a nozzle operationally linked to a valve (6) and a pressurised gas source.

34. The apparatus according to claims 31 to 33 wherein said means for measuring the vibrations induced on the liquid or gel surface is a contact less measurement technique.

35. The apparatus according to claim 34 wherein said means for measuring the vibrations induced on the liquid or gel surface is a laser displacement sensor (2).

36. The apparatus according to claim 35 wherein the laser displacement sensor (2) can be mounted in angle with the direction of the gas pulse.

37. The apparatus according to claim 35 wherein the laser displacement sensor (2) can be mounted in line with the direction of the gas pulse.

38. The apparatus according claim 35 comprising a nozzle (1) having an inner tube (8) allowing the transmission of the laser beam.

39. The apparatus according to claim 38 wherein the walls of said inner tube (8) are closed throughout the length of the nozzle (1).

40. The apparatus according to claims 32 to 39 comprising a computing unit (10) controlling the opening and closing of the valve (6) in accordance to a given measurement programme.

41. The apparatus according to claims 31 to 39 comprising a computing unit (10) allowing the analysis of the digitised output of the means for measuring the vibrations induced on the liquid or gel surface.