In the manufacture of ingredient blends, the optimal duration of the blending process is key to ensure that the final product is homogeneous and has an adequate contribution to the product to which it is added.
During the manufacture of the blend it is a priority to ensure that the raw materials are present in the right quantities and with uniform blending, as this will influence the performance of the product, its texture, flavor, etc.
When it comes to very high-value ingredients, such as vitamins, minerals or amino acids, product homogeneity becomes particularly important to ensure the correct contribution to the whole product.
However, in the industry, the optimum blending time is calculated on an estimated basis, based on theoretical knowledge.
To calculate this parameter, a number of points are evaluated, particle size, type of material and product density; and based on these results, an estimate of the blending duration is made. Fluid dynamics is difficult to evaluate, therefore, the results of this type of analysis are imprecise and this can lead to defects in the homogeneity of the products.
What is new is that it is now possible to use NIR technology to assess the homogeneity of a batch by comparing the chemical fingerprint of the product with samples taken at different points in a batch.
For example, Blendhub applies Chemometric Brain's qualitative models to determine the optimum blending time. The procedure consists of 4-, 6- and 8-minute blends with stops at the end of each of these times. At each stop, samples are taken from 5 different points in the mixer, analyzed with NIR and can then be returned to the batch, as this technology does not destroy the sample. After repeating these tests for the 3 blending times, a pattern is obtained that allows determining the optimum blending time for the product.
In this fast and effective way, a manufacturing procedure is established that ensures homogeneous batches. This is key to avoid customer complaints about problems that could be caused by variability in a batch.