In industrial solid/liquid separation, the removal of fine particles in a suspension with a significant amount of solids content can be a challenging task. Cake filtration in particular is a useful principle for separation of such suspensions. This separation occurs in a wide array of industrial production applications and has some advantages compared to other separation principles. To achieve the required performance, different types of filters can be used. One commonly used type is operated semi-continuous and use backwashing to discharge the filter cake. The regeneration of the filter cloth is as important as the actual filtration step. In the case of a properly regenerated filter cloth, the filtrate flow at the beginning of each step is still the same, while the initial filtrate flow decreases for and inefficient regeneration.This regeneration of the filter cloth depends strongly on the interaction between the particle system and the filter cloth. In the case of backwashing filtration, a clear difference could be observed between multifilament and monofilament fabrics.For multifilament filter cloths, it is possible to calculate the minimum required cake thickness for a complete discharge by means of the total resistance, i.e. the product of the specific cake resistance multiplied by the cake thickness. For 3 investigated fabrics with mesh size ≤ 20 µm, a complete cake discharge could be achieved by all 3 model particle systems. It is noticeable that the minimum required cake thickness decreases with increasing specific resistance. As soon as the cake thickness with the respective specific cake resistance of the particle system exceeds the value 5∙1010 m−1, a complete discharge can be observed. The cake thicknesses found within the same particle system were very similar between the 3 fabrics investigated. Based on this observation, the release of cakes from multifilament fabrics appears to be dominated by cohesion of particles in the cake. The adhesion from the first particle layer to the fabric does not seem to be a significant influencing factor.The situation is different for monofilament fabrics. This behaviour is different from the multifilament fabrics in the cake thickness. Here, smaller cake minimum required cake thicknesses were observed for small specific cake resistances. This is due to the particle deposition mechanism within the fabric. A lower specific cake resistance is associated with a larger particle size and/or a different particle shape (greater deviation from the ideal sphericity). Larger particles relative to the mesh size of the filter fabric preferably form a clean interface between the particle layer and the filter fabric (surface deposition). This leads to a lower interference and better cake discharge. This behaviour could be estimated by dividing the filter media resistance (determined by a particle system whose particle size is significantly larger than the mesh size of the fabric), by the specific cake resistance multiplied by a particle shape-dependent factor (Eq. (2)). The shape-dependent factors were determined to be 0.03 for orthorhombic, 0.15 for flaky and 0.46 for needle-shaped particles and thus increases with increasing deviation from the ideal sphericity. By including the filter medium resistance, the influence of the fabric and the adhesion of particles to the fabric is to be expected. Validity of this model was proven with 6 filter clothes with ≤ 22 µm mesh (3 twill and 3 plain weave).Similar behaviour between multifilament fabric and monofilament fabric could only be observed with the required minimum backwashing volume for a complete discharge. The required necessary backwash volume is always ≤ 2 l/m2. In summary, this study shows the regeneration performance of different filter cloth types (multifilament and monofilament) can be predicted by different models.
Read full abstract