Which process sensors (e.g. torque, temperature, moisture) are useful for the reproducibility of mixing processes?
For reproducible mixing processes, a combination of recipe monitoring, plant parameters and product-related measurement technology is useful. The mixing state often cannot be reliably assessed from a single signal. With powder mixtures in particular, a distinction should therefore be made between indirect process variables such as torque, temperature and mixing time, and direct or product-related measured variables such as moisture, concentration and spectrally determined homogeneity.
Precise dosing is the basis of every reproducible batch. Load cells, dosing systems, mass flow meters and batch-related recipe management ensure that all components are added in the correct quantity and sequence. With liquids, Coriolis mass flow meters are often useful because they directly capture the mass actually dosed. Level measurement and mass balancing supplement the monitoring of raw-material addition.
Speed, direction of rotation, mixing time and, where applicable, the position of internals or tools must be stored as fixed process parameters. Measuring torque or active power at the drive shaft provides additional information on mechanical energy input and on changes in product state. If power consumption rises or falls markedly, this can indicate, for example, a change in moisture, agglomerate formation, a viscosity change or an altered fill level. Torque and power, however, are not direct proof of a homogeneous mixture. They can support a validated process fingerprint, but must be secured for the particular recipe through product analyses.
Product temperature is important because it influences viscosity, flow behaviour, reaction rate, moisture distribution and, where applicable, crystallisation behaviour. Temperature measuring points in the product area are useful, along with additional measurements at the jacket, heating or cooling medium and, where applicable, at critical mixing-in points. In thermally controlled processes, product temperature, jacket temperature and heat-carrier flow should be captured together. This makes it easier to detect temperature gradients and control deviations.
Moisture is often a critical variable with powders, hygroscopic substances, granulates and drying processes. It influences flowability, the tendency to dust, agglomeration, mixability and storage stability. Depending on the product and installation situation, capacitive, microwave-based or spectroscopic inline moisture measurement methods come into consideration. A moisture measurement is only reliable if it is calibrated with representative samples of the original product and checked against a reference method.
For the direct assessment of mixing homogeneity, spectroscopic inline moisture measurement and Raman spectroscopy are particularly interesting PAT methods. With suitable calibration, they can monitor concentration differences and the progression toward a homogeneous mixing state in real time. Spectroscopic inline moisture measurement, for example, has been used to determine mixing endpoints and assess blend uniformity. Raman spectroscopy can likewise be used for inline monitoring of homogeneity and the mixing endpoint. Its informative value, however, depends on the recipe, the measuring window, particle movement, chemometric evaluation and product-related validation.
With suspensions, emulsions and dispersing processes, particle size measurement, turbidity, conductivity, pH value or viscosity can additionally be relevant. These variables do not necessarily capture mixing homogeneity, but they can describe product state, degree of dispersion and process stability. Pressure and differential pressure are important above all with closed, vacuum-loaded or gas-flow plants. They help to detect leaks, filter loading, blockages or deviations in vacuum and gas management at an early stage.
Vibration and structure-borne sound sensors serve predominantly for condition monitoring of the drive, bearings and gearbox. They can detect mechanical anomalies at an early stage, but are normally not a primary measuring tool for assessing product quality or mixing homogeneity.
For robust reproducibility, all relevant values should be recorded in temporal synchronisation, batch-related and at a sufficient measuring frequency. A proven process fingerprint can, for example, combine dosed quantities, dosing sequence, fill level, speed, mixing time, torque profile, product temperature, moisture and a validated signal from spectroscopic inline moisture measurement or Raman spectroscopy. The mixing endpoint should not be derived solely from a fixed time or from torque, but from a set of criteria validated for the particular recipe and plant.
How amixon® integrates process sensor technology and endpoint determination in the mixer
For reproducible mixing processes, it is not a single sensor signal that is decisive, but a validated interplay of recipe management, machine parameters and product-related measured values. amixon® plants can be equipped with suitable sensor technology on a project-specific basis. Which measured variables are actually useful depends on the product, mixing task, batch size, moisture, temperature management and quality requirements.
The central machine parameters are mixing time, speed, direction of rotation, fill level, dosed quantities and dosing sequence. These values are stored in the PLC and documented on a batch basis. They form the basis for reproducible operation.
Torque, temperature and moisture
Measuring torque or drive power can indicate changes in product state. It provides indications of altered flow properties, moisture, agglomerate formation, viscosity or a change in consistency during mixing. A torque profile, however, is not a universally valid proof of mixing homogeneity or a granulation endpoint. Its informative value must be established through trials for the particular recipe, fill level and chosen mode of operation.
Product temperature and the temperature of the heating or cooling medium are important variables in thermally controlled mixing processes. They support the consistent management of viscosity, reaction rate, crystallisation or drying. Several temperature measuring points can be useful where different temperature zones, liquid additions or large batches are present.
Moisture measurement is particularly relevant with powders, hygroscopic products, moist granulates and drying processes. Depending on the task, capacitive, microwave-based or NIR-based methods can be used. An inline moisture measurement, however, requires product-related calibration against suitable reference samples so that a reliable moisture value can be derived from the measurement signal. NIR and Raman methods can additionally be used, with suitable calibration, to assess concentration and mixing homogeneity.
Monitoring vacuum drying
In vacuum mixer-dryers and mixing reactors such as VMT and AMT, system pressure, product temperature, heating or cooling medium, condensate quantity, vapour profile and drive power supplement process monitoring. In vacuum contact drying, the combined profile of product temperature, pressure and condensate removed can indicate an advancing drying state.
The drying endpoint, however, should not be derived from pressure and temperature alone. For the specific product it must be secured via a target residual moisture, reference analysis and validated process data. An integrated sampler can supplement the inline readings with a direct product sample.
Sensor technology to the URS
amixon® apparatus is designed according to the operator's User Requirement Specification. Sensor technology, measuring points, automation, sampling, interfaces and the scope of documentation are therefore established on a project-specific basis. The PLC can manage mixing programmes, temperature profiles, dosing sequences and alarm limits. Connection to a control system, MES or ERP takes place via the interfaces defined in the project.
In regulated applications, requirements for user rights, audit trail, data integrity and electronic batch documentation can be taken into account. The specific execution depends on the operator's validation concept and regulatory requirements.
Validating the endpoint in trials
A once-validated parameter set can enable a high degree of batch reproducibility. This includes, for example, mixing time, speed, fill level, temperature profile, order of material addition, dosed quantities and, where applicable, vacuum profile. Reproducibility, however, presupposes that raw-material quality, grain size distribution, bulk density, initial moisture and batch size also remain within the defined process limits.
In the amixon® pilot plant, it can be examined with the original product which measured variables represent product state most reliably. Mixing quality, torque profile, product temperature, moisture, drying time and residual discharge, among other things, are assessed. This trial data forms the basis for the sensor design, limit values and process control of the production plant.