Cyclone Filter Cold-State Modeling Test: Similarity Criteria, Procedures, and Performance Evaluation

A cyclone filter cold-state modeling test is a scale-model experiment conducted at ambient temperature that uses similarity theory to predict the pressure drop, separation efficiency, and flow-field behavior of industrial cyclone separators under actual hot operating conditions. This cost-effective testing approach is widely applied in the design validation and optimization of circulating fluidized bed (CFB) boilers, fluid catalytic cracking (FCC) units, and industrial dust collection systems. This article explains the technical principles, sample requirements, testing items, methods, and typical applications of cold-state modeling tests for cyclone filters.

Technical Overview

A cyclone filter cold-state modeling test is an experimental technique in which a geometrically scaled model of a cyclone separator is tested using ambient-temperature air and representative solid particles to simulate the gas–solid separation process that occurs in full-scale hot equipment. Because operating cyclones in facilities such as CFB boilers and FCC regenerators handle flue gas or process gas at elevated temperatures—often between 600 °C and 950 °C—direct hot-state testing is expensive, hazardous, and difficult to instrument. Cold-state modeling overcomes these limitations by applying the principles of similarity theory, allowing engineers to measure key performance parameters on a transparent, accessible laboratory model and then extrapolate the results to the hot full-scale unit.

The validity of the cold-state test depends on satisfying appropriate similarity criteria between the model and the prototype. Full dynamic similarity is practically impossible to achieve simultaneously, so testing programs typically prioritize the dimensionless numbers that govern cyclone performance:

• Geometric similarity – the model must reproduce all relevant dimensions of the prototype (barrel diameter, inlet configuration, vortex finder geometry, cone angle, dipleg arrangement) at a reduced scale, typically between 1:5 and 1:10.
• Stokes number (Stk) – the ratio of particle inertial forces to drag forces, which is the dominant parameter controlling particle separation; matching Stk between model and prototype ensures comparable fractional separation behavior.
• Reynolds number (Re) – testing is conducted in the self-modeling (Re-independent) turbulent regime, generally above a Reynolds number of 10⁵ based on inlet dimensions, so that further increases in flow rate do not change the Euler number.
• Froude number (Fr) and Euler number (Eu) – evaluated where gravity-sensitive behavior (such as hopper drainage and re-entrainment) or dimensionless pressure-drop correlation is required.

By maintaining these criteria, the cold-state model accurately reproduces the vortex structure, particle trajectories, pressure field, and collection mechanism of the prototype, providing reliable data for design verification, retrofitting, and performance optimization before hot commissioning.

Testing Samples

Scaled Cyclone Filter Models

Test specimens are typically transparent scale models fabricated from acrylic (PMMA), polycarbonate, or glass. Transparency enables flow visualization, dust-rope observation, and laser-based velocity measurement. Models include the complete flow path: the tangential or volute inlet duct, the cylindrical barrel, the conical section, the vortex finder (exit tube), and the dust hopper or dipleg. Both single cyclones and multi-cyclone arrays can be modeled.

Test Particulate Media

Solid media are selected to match the particle size distribution and density of the actual process dust. Commonly used materials include fly ash, quartz sand, FCC equilibrium catalyst, limestone, coal powder, and standardized test dusts such as those specified in ISO 12103-1. Particle size distributions are characterized before testing, generally by laser diffraction in accordance with ISO 13320 or by sieve analysis.

Ancillary Flow Components

Inlet ductwork, expansion joints, diplegs with flap valves, and loop-seal or fluidized-bed hopper geometries may be included in the model when their influence on separation performance or material reinjection is part of the evaluation scope.

Testing Items

A comprehensive cold-state modeling test program typically covers the following measurement items:

• Total pressure drop – the static pressure difference between the cyclone inlet and gas outlet, expressed dimensionlessly as an Euler number for extrapolation to hot conditions.
• Overall separation (collection) efficiency – the mass ratio of collected particles to fed particles, measured gravimetrically over defined operating periods.
• Fractional (grade) efficiency – separation efficiency as a function of particle size, obtained by comparing inlet, captured, and escape particle size distributions.
• Cut size (d₅₀) and critical diameter (d₉₉/d₁₀₀) – characteristic particle diameters at which the cyclone captures 50% and 99–100% of particles, respectively.
• Gas flow field mapping – tangential, axial, and radial velocity profiles at multiple cross-sections, measured with five-hole probes, hot-wire anemometry, laser Doppler velocimetry (LDV), or particle image velocimetry (PIV).
• Inlet concentration effect – performance variation across a range of solid loading rates, from dilute-phase (a few g/m³) to dense-phase conditions (several kg/m³).
• Turndown characteristics – efficiency and pressure-drop behavior across the expected operating load range, typically 50–110% of design flow.
• Re-entrainment and hopper performance – assessment of dust leakage up the dipleg, vortex core instabilities, and gas leakage from the collection hopper.

Testing Methods

Test Rig Configuration

A representative test rig consists of a variable-speed centrifugal fan, a flow-conditioning section, calibrated airflow measurement devices such as orifice plates or Venturi tubes installed in accordance with ISO 5167, a calibrated screw-feeder or fluidized-bed dust feeder for dosing particles at controlled rates, the transparent cyclone model, a fabric filter or baghouse to capture escaped fines, and instrumentation for pressure and temperature monitoring. Duct velocity verification is performed with Pitot-static tubes in accordance with ISO 3966.

Pressure Drop Measurement

Static pressure taps are installed upstream of the cyclone inlet, at the vortex finder outlet, and at intermediate sections along the barrel and cone. Measurements are taken with calibrated U-tube manometers, digital micromanometers, or differential pressure transducers at multiple flow rates to establish the Euler-number relationship and confirm operation in the Re-independent regime.

Separation Efficiency Measurement

Overall efficiency is determined by the gravimetric balance method: the mass of test dust fed over a timed interval is compared with the mass collected in the cyclone hopper, with escaped dust recovered in the downstream filter. Escape concentrations may be independently verified by isokinetic sampling of the outlet gas stream following ISO 9096, ASTM D3685/D3685M, or the applicable national stack-sampling procedure such as GB/T 16157. Each operating point is repeated to demonstrate measurement repeatability and quantify uncertainty.

Grade Efficiency Determination

Representative samples of the feed dust, hopper catch, and escaped dust are analyzed by laser diffraction particle size analysis in accordance with ISO 13320. Fractional efficiency for each size band is calculated from the mass balance across the three streams, generating the grade-efficiency curve from which d₅₀ is derived.

Flow Field Measurement

Three-dimensional velocity fields inside the model are mapped using a calibrated five-hole spherical probe traversed along defined diameters at multiple elevations, or non-intrusively using LDV or PIV systems with seeding particles. The resulting data validate computational fluid dynamics (CFD) models and identify flow anomalies such as vortex eccentricity or excessive short-circuit flow through the vortex finder.

Extrapolation to Hot Full-Scale Conditions

Cold-model results are converted to full-scale hot predictions by applying the similarity relationships. Pressure drop is scaled through the Euler number and gas density ratio, while separation performance is scaled through the Stokes number, accounting for gas viscosity and density differences between ambient air and hot flue gas. The extrapolated values are reported with stated assumptions and uncertainty bounds.

Applications

Cold-state modeling tests of cyclone filters serve a broad range of engineering and industrial purposes:

• Circulating fluidized bed (CFB) boilers – validation of high-temperature cyclone designs that control the circulation of bed material, where separator efficiency directly determines combustion efficiency, temperature uniformity, and limestone utilization.
• Fluid catalytic cracking (FCC) units – evaluation of reactor and regenerator cyclone systems, including dipleg performance and catalyst loss reduction.
• Industrial dust collection and air pollution control – sizing and optimization of single and multicyclone collectors used upstream of baghouses and electrostatic precipitators, supporting compliance with particulate emission limits.
• Pneumatic conveying and spray drying – verification of product recovery cyclones in bulk-solids handling and chemical process industries.
• Design optimization studies – parametric investigation of inlet dimensions, vortex finder diameter and insertion depth, cone angle, and roof-lip or volute configurations to maximize efficiency at acceptable pressure drop.
• Retrofit and troubleshooting – diagnosis of excessive pressure drop, high emission, hopper re-entrainment, or erosion problems in existing installations, and verification of corrective modifications before plant implementation.
• Independent verification – performance confirmation of manufacturer designs by accredited third-party laboratories, or in-house testing supporting product development and customer acceptance documentation.

FAQ

What is a cyclone filter cold-state modeling test?

It is a laboratory experiment in which a scaled, usually transparent model of a cyclone separator is operated with ambient-temperature air and test dust. By applying similarity criteria—principally matched Stokes numbers and Reynolds-number-independent flow—the test reproduces the separation physics of the hot full-scale unit and allows its pressure drop, efficiency, and flow behavior to be measured safely and economically.

Why is the test performed cold instead of at operating temperature?

Hot-state testing at 600–950 °C requires specialized materials, combustion or heating infrastructure, and difficult instrumentation, making it costly and slow. Cold modeling provides full optical access for flow visualization and laser measurement, allows rapid parametric studies, and still yields quantitatively scalable results when similarity rules are respected.

How are cold-model results converted to full-scale hot performance?

Pressure drop is extrapolated using the Euler-number relationship together with the actual gas density and velocity at operating temperature, while collection performance is extrapolated through the Stokes number by adjusting for gas viscosity and density differences between air at ambient conditions and the hot process gas. This yields predicted full-scale pressure drop, grade efficiency, and cut size with documented uncertainty.

What similarity criteria must be satisfied?

Geometric similarity is essential for all internal dimensions. Dynamic similarity is achieved by matching the Stokes number for particle separation behavior and ensuring the flow Reynolds number lies in the self-modeling turbulent region so that the Euler number becomes independent of Re. Froude number matching is applied where gravitational effects such as hopper drainage are significant.

What particle media are used in the test?

Media are chosen to replicate the density and particle size distribution of the actual process dust. Typical materials include fly ash, quartz sand, FCC catalyst, limestone, coal fines, and standardized reference dusts conforming to ISO 12103-1. All media are characterized by laser diffraction or sieve analysis prior to testing.

Which standards govern the test?

There is no single mandatory standard for cold modeling itself; however, supporting measurements reference recognized procedures, including ISO 5167 for flow measurement, ISO 3966 for Pitot-static velocity traverses, ISO 9096, ASTM D3685/D3685M, and GB/T 16157 for particulate sampling, and ISO 13320 for particle size analysis by laser diffraction.

How long does a typical test program take?

Duration depends on scope. A single operating-point verification may be completed within a few days, while a full parametric program covering multiple geometries, load levels, and dust loadings—complete with flow-field mapping and CFD benchmarking—typically requires several weeks of laboratory work plus time for data reduction and reporting.