How it works
Agglomeration is a size-enlargement process: two particles of volumes u and v collide and merge into a single particle of volume u + v. The model follows the particle population by number, so the complete particle size distribution is resolved rather than a single mean diameter.
How often collisions lead to growth is set by two things: an agglomeration kernel and a size-independent rate constant. Both are explained in the next section.
Total solid mass is conserved — agglomeration only redistributes mass across size classes. The model deliberately stays at a shortcut level: it focuses on size as the primary distributed property and does not resolve attrition or secondary properties.
Understanding the agglomeration kernel
When two particles collide, whether they actually merge and grow depends on their sizes. The agglomeration kernel β(u, v) captures exactly that: it is a function of the volumes u and v of the two colliding particles, and it sets how frequently that particular size pair coalesces relative to others. A larger value means that pairing agglomerates more readily.
Different physical mechanisms make this size dependence look different — fine particles brought together by Brownian motion, particles swept into each other by shear in a stirred bed, or large particles colliding under gravity each follow their own law. Choosing a kernel is really choosing which mechanism you believe dominates your process; the rate constant then scales how intensely it all happens.
β₀ — the rate constant
A single size-independent number that sets how fast agglomeration proceeds. It lumps together the operating conditions — mixing intensity, binder, moisture — and is the value you calibrate against experimental data, since no formula gives it up front.
β(u, v) — the kernel
A function of the two particle sizes that sets which size pairs preferentially collide and coalesce. It encodes the dominant growth mechanism, and DyssolPro provides a standard library of kernels (constant, sum, product, Brownian, shear, gravitational and more) to choose from.
The model
DyssolPro solves a dynamic population balance for the number density n(v, t). Particles are born into a size class when smaller ones agglomerate into it, and die from a class when they agglomerate into something larger; the inlet and outlet streams complete the balance.
∂n(v,t)/∂t = B_agg − D_agg + ṅ_in − ṅ_out (with ṁ_out = ṁ_in)Birth and death rates B_agg and D_agg are evaluated by a selectable numerical solver, and a standard library of agglomeration kernels lets you match the dominant growth mechanism. The page stays at a non-critical level of detail and does not expose implementation internals.
Key inputs
- Agglomeration rate (β₀)Size-independent rate constant that scales the overall growth intensity with operating conditions.
- KernelSelects how collision frequency depends on the two particle sizes — e.g. constant, sum, product, shear, Brownian and other standard kernels.
- SolverNumerical method used to evaluate the agglomeration birth and death rates (e.g. cell average, fixed pivot, FFT).
Equipment this model can represent
Any process where particle collision and coalescence drive size enlargement.
Fluidized bed agglomerators
Bed agglomeration where suspended particles collide and bind into larger structures.
High-shear granulators
Intensive mixing where impeller-driven collisions promote rapid agglomerate growth.
Batch granulators
Drum, pan, and mixer granulators operated batch-wise for controlled size enlargement.
Continuous mixer granulators
Continuous units where coalescence governs the product size distribution.
Typical engineering studies
What teams investigate with the agglomerator model.
Homogeneity studies
Assess how uniformly agglomerates develop across the population under given operating conditions.
Moisture distribution
Study moisture in connected flowsheets when binder or liquid streams feed the growth zone.
Particle size distribution
Track the full PSD evolution from feed fines to the agglomerated product.
Agglomeration efficiency
Quantify growth and the conversion of fines into product across kernels and rates.
Process-parameter sensitivity
Map how rate constant, kernel choice, and residence time move the product size.
Model-based Design of Experiments
Plan and screen experiments in simulation before running them on the plant.
Scale-up and optimization
Carry validated parameters from lab to production and optimize against product targets.
Application example
FFT-based solver for dynamic agglomeration
A demonstration flowsheet (Agglomerator.dflw) ships with Dyssol, and an FFT-based population-balance solver for dynamic flowsheet simulation of agglomeration was published and validated.
Technical FAQ
How can I improve agglomerate strength without increasing binder usage?
DyssolPro does not compute mechanical strength directly, but you can run operating-window studies on residence time, growth, and binder distribution to reach a target size at lower binder input before committing to lab trials.
Why are my agglomerates breaking apart after drying?
Connect the agglomerator with the downstream dryer and handling units in one flowsheet to see how moisture and thermal load evolve, which helps localize whether the problem starts in growth, drying, or transport.
How do I choose the right binder for powder agglomeration?
Binder chemistry is a laboratory decision. DyssolPro’s role is to propagate a binder’s captured effect (through growth and nucleation parameters) across the process and predict the downstream PSD.
What causes oversized lumps in an agglomeration process?
With a population-balance growth model you can study how kernel parameters, liquid input, and residence time drive the coarse tail of the PSD and screen for operating points that suppress oversize.
How can I control particle size distribution in an agglomerator?
PSD is a first-class quantity in DyssolPro: the agglomerator tracks the full distribution, so you can run sensitivity and optimization studies on the parameters that shape it across the connected process.
What is the difference between wet agglomeration and dry agglomeration?
This is a process-design distinction rather than a single setting; in DyssolPro you represent the relevant mechanism through the chosen model and its parameters and compare process variants in one flowsheet.
How do impeller speed and residence time affect agglomerate size?
These map onto model parameters and holdup. You can vary them in sensitivity studies to see their effect on agglomerate size and throughput before plant trials.
How can I reduce dust formation during agglomeration?
Connect the agglomerator with cyclone and gas-filter units and study how much fines and dust leave with the gas stream under different operating points.
What process parameters influence agglomerate porosity?
If porosity is carried as a distributed particle property or via a custom model, DyssolPro can follow its evolution through the flowsheet and relate it to the driving parameters.
How do I scale up an agglomeration process from lab to production?
Fit model parameters to lab data, then run the validated model at production scale inside the full flowsheet to check feasibility before commissioning.
How can I prevent uncontrolled growth in an agglomerator?
Dynamic simulation of growth and recycle lets you identify stable operating windows and control settings that avoid runaway growth.
Why does my agglomeration process produce too many fines?
Study the fines fraction of the PSD against nucleation and breakage parameters and against recycle load in the connected flowsheet.
How does liquid spray droplet size affect agglomerate formation?
Droplet effects enter the model through nucleation parameters; you can study how changing them shifts the resulting PSD.
How can I optimize binder concentration for agglomeration?
Once the binder effect is parameterized, use the integrated optimization tool to search binder concentration against a target product property.
What causes agglomerates to be too soft or too friable?
Strength is not computed directly; couple a custom strength or breakage correlation through Model Maker and use the simulation to study which process drivers move it.
How do I model wetting and nucleation in agglomeration?
This is core to the unit: population-balance agglomeration represents nucleation and growth, and Model Maker lets you implement specific wetting and nucleation kinetics.
What is the best way to measure agglomerate quality?
Measurement is experimental. DyssolPro complements it by predicting size and PSD so you know what to expect and can reconcile model and measurement.
How does powder wettability influence agglomeration behavior?
Wettability is a material input reflected through nucleation parameters; its influence on growth and PSD can then be studied in the flowsheet.
How can I reduce recycle load in an agglomeration circuit?
DyssolPro simulates recycle loops, so you can study how operating points and screen cut sizes change the recycle load around the agglomerator.
What are typical control strategies for continuous agglomeration?
Dynamic simulation of startup, shutdown, and disturbances lets you test control strategies and operating windows before implementing them on the plant.