It is usually assessed by measuring the product’s viscosity, defined as the resistance of a fluid that is being deformed by either shear or tensile stress, but usually thought of as its “thickness.” Companies are looking for consistency above all else. Quality control means ensuring consistent product texture. A successful process will generate a product that is always the same.
In reality, controlling texture is difficult and there are frequently fluctuations. Especially for non-Newtonian products, it is a real challenge to effectively measure viscosity in-line. Flow rate, temperature, and flow history have an effect on viscosity, and managing viscosity of such products is not always an easy task.
Since direct measurement can be challenging, quality control technicians have frequently had to infer viscosity from other measurements and simply hope for the best. Quality control labs are looking for ways to control viscosity more tightly so they can eliminate off-spec product. One method being used more frequently is a new type of in-line viscosity measurement device, which uses patented technology based on mixing principles applied to pipes. Bulk viscosity is measured on the whole fluid flow, which is re-homogenized by the action of in-line static mixers.
At a yogurt factory, the device was designed to control a smoothing valve on the white mass after maturation. The instrument is specially designed so that the shear rate applied to the yogurt is not higher than that applied in the empty pipe, thereby creating a totally non-intrusive measuring process. Texture fluctuations have been reduced to a tenth of what they were and protein consumption is optimized by reducing texture fluctuations and by attaining the minimal required specifications.
Viscosity measurements were used at a brewery in Montreal to detect the specific point at which the transition from yeast to beer takes place during the purge of a fermentation tank. Accurate measurements led to significant reductions in loss of beer production time and maintenance.
Measuring viscosity of non-Newtonian fluids is tricky
It has always been difficult to measure the viscosity of non-Newtonian fluids, because their viscosity depends on the precise flow conditions in effect during measurement. The same fluids flowing at two different flow rates in a pipe can give two different viscosity measurements.
To understand why this is, one must look at the microscopic level at the differences in how molecules are arranged depending upon the flow. For example, as the shear rate increases, the molecules tend to align themselves in the flow direction, with less resistance to flow. This molecular rearrangement can take some time to occur.
In some instances, the rearrangement time can be very short, compared to the observation time, but in other cases it can take much longer. In those cases, what happens to a fluid before the moment of measurement can have a considerable impact on the actual measurement.
Processes with non-Newtonian fluids passing through a pump, a filter, instruments with a smaller diameter than the pipe size, or those flowing through a sampling valve may be significantly affected. For example, a yogurt sample removed from a process pipe or through a sampling valve may take several minutes (or more) to stabilize, and only after this time can you have a stable and repeatable viscosity measurement on a laboratory rheometer.
The requirement to have the precise flow conditions in effect during measurement of viscosity is why the use of an in-line viscosity measurement device for measuring non-Newtonian fluids is so useful in process quality control.
Yogurt manufacturing facility – Existing quality control measurements not adequate to ensure consistency
For yogurt, viscosity measures whether the product is lumpy or smooth, watery, or too thick. A company may test a product and then adjust it to build the texture they want. But if not handled properly in the processing line, for example if the product is sheared too much, it will take much more energy and expensive raw materials to achieve the final desired texture. Also, the product’s texture can be broken and the manufacturer might experience huge variability. This happens during the curd change of phase, when the water and synerese (whey expelled during cooking) may break down. Operators may lack sufficient information to determine precisely how and why the variability came into the process and why it led to a product defect.
Take the example of a large yogurt factory that uses batch processing to mature its product. Once mature, operators empty the batch and pump the white mass to a packaging plant where it is cooled down before packaging. The high temperature is about 40°C in the tank, and it is cooled to about 20°C. Maturation stops when the product is cooled down. If it is too cool, the live enzyme will be killed and the quality of the product is affected, so maintaining the target temperature is extremely important.
Quality control personnel found that the quality of the white mass yogurt was not as stable as they expected, particularly after the maturation and cooling processes. Viscosity measures were done by sampling several production batches and the results showed significant fluctuations, particularly between the beginning, the middle, and the end of a batch’s transfer process.
Data at the plant showed that the texture varied by plus or minus 25 percent for the same product produced in one month. QC personnel did not expect such a wide variation before actually measuring it, and it then became a serious concern.
The measurement method used involved taking a sample out of the line and bringing it to a laboratory to analyze. However, the very act of taking a sample out of the line affects the product. The product goes through the line in a pressure condition, so when passing through the sampling valve, the product is submitted to a shear rate that affects its texture. In addition, the time needed to get the product to the laboratory and conduct an analysis has an impact on the results.
The instrument generally used for the measurements is a viscometer, which measures under a particular operating condition. For products like yogurt, viscosity depends upon flow, and to measure you must define what flow condition you are using. Therefore, when you take a sample and make an analysis at the lab, you have to define viscosity at a particular flow condition. When you use a viscometer, flow conditions are not defined, so flow patterns will differ if there are two different fluids at the same operating conditions. They are not measurements you can compare. While an excellent device for measuring Newtonian fluids that flow in a pipe, like oil or sugar syrup, a viscometer does not work as well for yogurt.
At the yogurt facility, the target value is 1000 centipoise (CP), and the range of spec of the measurements went from 500 to 1500 CP. With such a wide range, quality control personnel are not able to really control a process using this measurement. Also, they can’t accurately replicate the measurements, which may vary by several hundred CP. Analyses performed using two different offline instruments couldn’t give less than 10 percent difference between the two instruments.
In addition to the viscosity measurement, the analysis includes an instrument penetration test, in which a cylinder is pushed into the product with a given load and displacement is measured along with how fast the cylinder penetrates. This is done on the final product when it is packaged, and could be conducted days after it is produced. A penetrometer measures the gel firmness of the final products. The measurement consists of measuring the load required for a cylindrical plunger to penetrate into the sample at a given speed. This measurement is usually accomplished about two days after the product has been packaged. The means that it takes at least two days before a lot can be validated.
A smoothing valve is sometimes used after the white mass has been cooled down to smooth the product, or break the protein aggregates that may occur. A smoothing valve is usually used at one setting for a whole batch transfer, based on offline measurements of previous batches. Operators found that a different setting would have been required for better optimization of this operation, especially at the beginning and at the end of a transfer.
Quality control group seeks better information to eliminate inconsistencies
The quality control group at the yogurt plant was seeking more information and in-line measurements to understand the factors that generate defects and inconsistencies so they could avoid them. Rather than the off-line measurement process, they sought to continuously monitor viscosity during a batch transfer, after its cooling process, on a 3-inch diameter main production line. They installed a Viscoline CVL030S, made by KROHNE, Inc. Viscoline is made with stainless steel construction, no moving parts and no in-situ calibration required. It features continuous measurement with analog outputs. The unit is extremely reliable, with a repeatability of 0.2 percent, resolution to 0.1 CP while meeting government policy on metrological traceability.
With this device, the fluid flows through a continuous pipe containing two low pressure drop static mixers. The sensor device measures the pressure drop at both static mixers by means of two differential pressure measurements: ?P1 and ?P2. Pipe flow rate, required to accurately determine the flow regime to which the viscosity is to be calculated, can be obtained from either an external data measurement or with a flow meter.
From the two pressure drop measurements and the flowrate reading, the fluid flow parameters are processed in the system, and the pipe line viscosity is determined. Every viscosity measurement is referred to a corresponding operating temperature. Thermal corrections can be applied, when a measurement is required at a reference temperature. Such correction requires laboratory characterization.
With Viscoline installed in the mainline, the smoothing valve can be opened or closed in an automated fashion to achieve constant viscosity in real time. Leveling the texture fluctuations, the recipe can be adjusted upfront to get closer to the minimum specifications required. This reduces the consumption of expensive ingredients like proteins and results in a consistent, stable, uniform product that costs less money.
In addition, with the Viscoline installed right after the cooler, operators observed that 10 percent of the product was out of specification for texture after the transfer. This was happening at every start and stop during a batch transfer, due to a lack of heat exchanger control during these transient operations. Prior to using the Viscoline, there had been no way to measure the effect of these transitions.
The new Viscoline device gave operators the means to better control their process, as well as measure the impact of their corrective actions. They can standardize operations with all the operators on the line. This is allowing them to save money and get a uniform product that meets their quality standards.
Beer manufacturers use viscosity to separate yeast from beer
Once the fermentation process is complete during beer brewing at this Montreal brewery, the yeast used in the process settles and is deposited at the bottom of the tank. The tank is then emptied by pumping the yeast out and sending it to out to be reclaimed and recycled. A portion of what is left is directed to a centrifuge to separate the small amount of remaining yeast from the beer, and then the rest of the batch goes directly to filtration. The problem is to determine at which specific point during this pumping process all the yeast has passed, leaving only the beer.
Since both yeast and beer are brown, it is a tricky operation to determine with certainty which is which. An operator would view the pumping process through a glass section of pipe, spot the subtle color change between yeast and beer, and stop the process at the point he guessed the separation occurred.
In addition, operators had to guess the quantity of yeast in the tank and add more or less based on temperature, yeast type, and pH. Since yeast production is never stable from one batch to another, the brewer also had to pump different quantities from each batch, depending on the type of beer, and send it to recycling, while sending a different quantity to the centrifuge. The centrifuge process also took too much time and did not filter away the proteins that regularly clog up the filters.
If operators guess incorrectly and miss transition, they can end up either sending beer to the yeast recycling operation, which means losing beer (and also getting penalties from the company buying their reclaimed yeast), or sending yeast to the filtration operation, where it stresses the filters, quickly plugging them. When this happens, operators have to stop to clean the filters, which costs time and money. Using the centrifuge to separate yeast from beer can take as long as three hours per batch. In addition to wasting time, the manual system led to a great deal of waste of beer – they estimated about $100,000 worth of product was lost per year.
The brewery was looking for a way to detect the specific point at which the transition from yeast to beer takes place during the pumping process. They installed a sanitary 2-inch Viscoline on the main line, just below the main transfer pump. A solenoid valve was also installed, which is controlled using the viscosity reading in real-time. Concentrated yeast is much more viscous than beer, so it measures at very high levels, and this is directed to the recycling operation. When the measurement drops down, it is beer, which goes directly to filtration. This completely circumvents the centrifuge process, thereby saving time and beer.
The brewery saved about three hours per operation per batch. They are no longer assessed penalties from the yeast recycling company, and no longer lose more than $100,000 worth of beer each year. Because the filters no longer have to handle high yeast concentrations, there is two times less filter plugging.
The brewery installed three units, and in their 3-month summer season their savings equaled three times their investment, giving them a one-month payback.
The brewery was so pleased by the successful results that it decided to go for an even more precise yeast detection threshold, less than 1 percent, and they are now able to use the indicator to detect a viscosity variation of one tenth of a CP, which represents a significant drop from 2 percent down to 0.25 percent yeast content in each batch of beer. The new process means the brewery saves $1 for every 4 barrels of beer produced.
Source: Food Processing – IML Group PLC