How can the scale-up from 20-litre laboratory batches to 2,000-litre production be secured without losing mixing quality?

Question

Accepted Answer

How amixon® secures the scale-up from the pilot plant into production: the mixing principle is identical across all scales

The core of the amixon® approach is that trial and production mixers work with the same mixing system. The SinConcave®/SinConvex® mixing tool conveys the material upwards at the periphery and lets it flow downwards again in the centre. This creates a three-dimensional total flow with the same characteristics across all sizes.

Scaling is based on the mixing task and on the results achieved in the small batch or pilot-plant trial, with the ideal mixing quality always the target:

the same mixing time or cycle time,
the same texture,
the same particle stress,
the same bulk density,
the same flow behaviour,
the same particle size distribution,
the same filling and packaging behaviour,
and further product-specific target variables.

Scaling is based not on rotational speed or circumferential speed alone, but on the conveying characteristics of the helical mixing tool.

All sizes are operated within a similar speed window, as a rule at low speed. Flow profile, mechanical particle stress and the decisive similarity conditions are thereby preserved.

Mixing quality independent of fill level

amixon® mixers achieve very high mixing qualities even at a low fill level. This is a substantial advantage in scale-up, since process quality does not stabilise only at a narrow full-load limit. Batch sizes can therefore be varied in the same mixer without the homogeneity being fundamentally in question each time.

A fine size range instead of jumps in scale

amixon® manufactures mixers in a very fine size range. The step from the validated trial scale to the target size can therefore be planned without hard jumps in the catalogue. It is precisely this that reduces the risk of a similarity condition being lost through too large a jump in size.

Validation before the investment

35 test units in various sizes are available at the amixon® pilot plant. Trials run with the original product, at real fill levels and within the intended temperature and pressure range. The trials are carried out, evaluated and documented together with amixon® experts. In addition there are amixon® pilot plants in Japan, India, Thailand, China, South Korea and the USA.

The same basic formula for calculating the restratification frequency applies to all amixon® machines, that is the number of revolutions of the mixing tool for complete circulation of the entire batch in the mixing chamber:

I_V = (π/4) · (D² − d²) · Φ · S · n · ζ

I_V: displacement or conveying volume flow
D/d: outer diameter/inner diameter of the helix
Φ: fill level, referred to the working volume
S: pitch of the helix
n: rotational frequency, in revolutions per second
ζ: velocity coefficient, tool-specific

At the pilot plant, amixon® also always has large apparatus available for tests, up to 3 m³ batches and beyond. This is important where questions about dynamic pressure as well as particle rounding and the generation of fines are to be answered.

Mixing time, circumferential speed and further process parameters can thus be determined in such a way that they can feed robustly into the design of the production machine. This markedly reduces technical and economic risks before the investment.

Manufacture in Paderborn as the quality foundation

amixon® develops and manufactures exclusively at the Paderborn works with a high depth of manufacture and uses components from Germany exclusively. As a certified welding company with European, Japanese, Korean and American qualifications, amixon® designs every apparatus as a one-off on the basis of the operator's URS. Quality control remains entirely in-house without gaps, and every specification is verifiable down to component level.

Service across the entire life cycle

After commissioning, amixon® continues to stand by the operator. Regular inspections and preventive maintenance secure availability, on request also with predictive maintenance. Selected wear parts are supplied together with the initial delivery; most spare parts are held at the Paderborn site and at the service bases in Japan and the USA.

Many amixon® machines have been in daily use for more than 30 years. Modernisation and retrofitting keep them at the latest state of the art.

Answer

A validated scale-up rests on the combination of geometric similarity, the selection of suitable process-relevant characteristic numbers and verification by measurement and modelling at intermediate and production scale. The aim is to maintain the mixing quality reproducibly despite the increase in scale by a factor of 100.

Ensuring geometric similarity

Vessel, agitator and internals geometry should be executed as proportionally as possible at both scales, so that flow patterns, dead zones and circulation behaviour remain comparable. This includes in particular ratios such as vessel diameter to filling height, agitator diameter to vessel diameter and the position of internals and inlets.

Defining the right scaling variable

There is no universal scale-up criterion; what matters is the purpose of the process. Typical starting points are the specific power input P/V, agitator tip speed, mixing time, Reynolds number, Froude number and, in dispersive or reaction-dominated processes, local energy dissipation as well. Not all characteristic numbers can be held constant simultaneously, so the dominant target variable must be chosen process-specifically.

Analysing the mixing regime

The Froude number relates to rotating free-fall, drum and throw mixers as well as to systems with a horizontally mounted mixing tool. In vertical mixers with forced restratification it is not a design or scale-up criterion. What is decisive there is that the mixing principle remains the same across all sizes and that the restratification covers the entire mixing chamber independently of the fill level; transferability is secured in the pilot plant with the original product.

Before transfer to the production scale, it should be clarified whether the process is limited above all by micromixing, mesomixing or macromixing. With fast reactions, precipitations or sensitive suspensions, the local shear and dissipation situation near the agitator is often decisive, not just the average mixing time.

Using an intermediate scale and reducing risk

A direct jump from 20 litres to 2,000 litres is technically possible but risky. A pilot or pilot-plant scale in the region of approximately 100 to 300 litres makes it possible to check the chosen scaling rule, to detect deviations early and to verify parameters such as rotational speed, power consumption and mixing time.

Using CFD and DoE

CFD simulations help to assess flow profiles, dead zones, short-circuit flows and shear gradients at the target scale before cost-intensive plant modifications are made. In combination with a design of experiments, critical influencing variables can be investigated systematically and robust operating windows defined.

Measuring mixing quality objectively

The actual mixing quality should be verified not only by calculation but also by process analysis, for example through in-line measurements of conductivity, turbidity, spectroscopic or other PAT signals. The coefficient of variation and the measured mixing time, for example, serve as characteristic values for assessing the state at production scale quantitatively.

Taking thermal and operational boundary conditions into account

As the volume grows, the ratio of heat transfer area to volume deteriorates, which makes heating and cooling more sluggish. In addition, dosing points, gassing, fill level, start-up and shut-down behaviour and rheological changes in the product must be included in the design, since they can markedly influence the mixing quality at large scale.

Using uniform definitions of terms and processes

For a robust transfer, the terms, target variables and acceptance criteria used should be defined unambiguously and applied uniformly throughout the development and production context. This reduces scope for interpretation and facilitates technical release of the scale-up.

A robust scale-up therefore arises not from a single characteristic number but from the coordinated combination of geometry, choice of characteristic numbers, intermediate scale, simulation, PAT and clear acceptance criteria.