Updated: November 7, 2026

Table of contents

Two transparent bottles comparing particle distribution after controlled powder filling, with segregation visible in one bottle.

A powder blend can leave the mixer within specification and become nonuniform before it reaches the packaging line. The problem may develop during chute transfer, hopper filling, conveying, or discharge. Once particles separate, a sample taken from one location may no longer represent the material entering the next processing stage.

For process engineers and laboratory specialists, the first question is practical: does a particular transfer create conditions that allow segregation?

A simple bottle test can provide an initial indication. It requires little equipment and can expose visible differences between filling conditions. Turning those observations into reliable compositional measurements, however, requires more careful sampling and analysis.

Why Screen for Segregation?

Segregation occurs when particles or components separate because of differences in properties such as size, density, shape, cohesion, or aerodynamic behavior.

The mixture may be uniform after blending but change during subsequent handling. Muzzio and colleagues demonstrated that characterization inside a blender does not necessarily describe the mixture after further processing. Their work also identified substantial errors associated with conventional powder sampling [1].

For pharmaceutical blends, stratified sampling of dosage units provides one way to investigate composition changes throughout production [2]. Similar principles apply to food powders, chemicals, minerals, and other products in which component distribution determines quality.

A bench screen can identify a potentially troublesome transfer before a larger investigation and help determine which filling conditions deserve closer examination.

Where the Bottle Method Fits

The bottle method imitates one event: a powder stream entering a receiving container. It is particularly relevant to relatively free-flowing blends containing distinguishable components or particle-size fractions.

Sifting segregation occurs when smaller particles move through spaces between larger particles during motion. ASTM D6940/D6940M-20 identifies four conditions necessary for this mechanism: a difference in particle size, sufficiently large mean particle size, adequate flowability, and interparticle movement [3].

The standard provides useful guidance on particle size:

  • The size ratio between components can be as low as 1.3:1.
  • Sifting can occur with a mean particle size around 50 µm.
  • Sifting can become a dominant mechanism when mean particle size exceeds 100 µm.
  • The standard limits maximum particle size to 3 mm, primarily to prevent binding of the apparatus slide gate.

These are conditions that make sifting possible, not universal thresholds for segregation. In strongly cohesive fine powders, particle mobility may be insufficient for sifting, even when the particle-size ratio is large. A bottle test focused on pile formation may then reveal little about the main segregation risk.

Fluidization segregation behaves differently. Fine particles can remain aerated while coarser particles settle. ASTM D6941-19 addresses fluidization segregation in powders capable of being fluidized [4].

For coarser blends outside D6940’s particle-size range, or for processes dominated by aeration, a different test arrangement may be necessary.

For more detail on these mechanisms, see Powder Segregation: Causes, Mechanisms, and Control.

The Quick Bottle Screen

The initial visual comparison can take approximately 20 minutes once the materials and equipment are ready. Replicate testing, sampling, and compositional analysis require additional time.

Step 1: Prepare the Equipment

Use two identical transparent containers, such as 1-liter bottles, and a funnel or controlled feeding arrangement.

Select a funnel throat appropriate for the material. A 10 to 15 mm opening may work for some free-flowing blends, but the opening must not restrict flow through bridging or particle blockage.

Mark the intended fill level on each bottle. Keep the powder mass, receiving-container geometry, and outlet position consistent between comparisons.

Position the funnel above the receiving container and record the initial distance between the outlet and the receiving surface. Where possible, choose a drop height representative of the transfer under investigation.

Record the material identification, temperature, humidity, and relevant observations about its initial condition.

Safety: Do not conduct open pouring tests with hazardous, potent, or combustible powders unless appropriate containment and exposure controls have been established. Use an enclosed arrangement where required.

Step 2: Compare Two Filling Conditions

Prepare two equal, representative portions of the same starting blend. Use a spinning riffler where available to reduce the risk of segregation during sample division.

Pour the first portion through the funnel as a relatively fast, continuous stream.

For the slow condition, meter powder into the funnel at a controlled rate, for example with a vibratory feeder or small timed increments, so the outlet never runs full.

Keep the funnel opening and receiving geometry unchanged. If using timed increments, record their frequency and quantity because intermittent feeding may influence pile formation differently from continuous flow.

For each test, record:

  • Powder mass and final fill level.
  • Filling duration and average mass flow rate in g/s.
  • Initial drop height and any important changes during filling.
  • Stream position and whether a central pile forms.
  • Bridging, dust formation, splashing, or other unusual behavior.

Calculate average mass flow rate by dividing the transferred mass by the filling time.

Change only the feed rate between conditions. Keep stream position and drop height the same, so any difference in the pattern can be attributed to rate.

Do not tap or shake the bottles after filling. Vibration can induce segregation independently of pouring. Research on vertically tapped granular mixtures has demonstrated both conventional and reverse Brazil nut effects, depending on particle properties and excitation conditions [5].

Conduct the test away from drafts, since air movement can introduce additional separation.

Step 3: Examine the Powder Without Disturbing It

Look for visible differences between the two filled bottles.

These may include:

  • Concentration of coarse or fine particles near the container wall.
  • Distinct bands or layers.
  • Differences between the central pile and surrounding material.
  • Accumulation of visibly different components near the top or bottom.
  • Changes in the pattern between fast and slow filling.

Photograph each bottle from consistent positions and under similar lighting.

A uniform-looking bottle does not demonstrate blend uniformity. The transparent wall provides only a limited view of the internal bed, while components that look alike may have different concentrations.

When the Bottle Screen Needs Quantitative Measurements

If the visual comparison suggests segregation, or if composition uniformity is a product requirement, analytical measurements can help determine its extent.

Three issues deserve particular attention: the starting blend, the spatial sampling pattern, and measurement repeatability.

Establish the Starting Blend Composition

Before testing, establish whether the prepared blend is reasonably uniform. Use a suitable representative sampling and splitting procedure to obtain baseline measurements of the component or particle-size fraction being investigated.

Where possible, measure multiple baseline samples to estimate initial variability. Document the sampling method and sample mass so that later observations can be compared with the starting material.

Research by He and colleagues illustrates the importance of distinguishing initial mixture quality from segregation introduced during subsequent handling [6].

If the starting composition is unknown, a difference between two bottle samples cannot automatically be attributed to pouring.

Sample the Pattern That Matters

Top, middle, and bottom samples may reveal vertical concentration differences. However, not all segregation produces a vertical pattern.

Sifting segregation frequently produces radial separation.

As material forms a pile, finer particles can concentrate beneath the impact point while larger particles move toward the perimeter. ASTM D6940 identifies center-to-periphery segregation as a characteristic concern [3].

A bottle could therefore show little top-to-bottom variation while containing substantial radial segregation.

For quantitative analysis, choose the sampling arrangement according to the expected pattern:

  • Vertical segregation: Compare defined upper, middle, and lower regions where they can be recovered without substantial disturbance.
  • Radial segregation: Compare central and peripheral regions using a method that preserves their spatial identity.
  • Discharge segregation: Collect defined sequential discharge fractions when the question concerns composition entering downstream equipment.

These sampling approaches are not interchangeable. Sequential discharge samples, for example, cannot automatically be treated as samples from the original upper, middle, and lower layers.

Taking a core through the bed with a thin tube introduces additional uncertainty. Such a procedure resembles thief sampling, which may disturb the bed and collect different components unevenly [1].

Where reliable spatial sampling is impractical, a purpose-built segregation apparatus or process-specific sampling arrangement is preferable.

Validate Any Tracer

A colored or otherwise distinguishable tracer can make segregation easier to observe, but it must behave sufficiently like the actual component being investigated.

Particle size, density, shape, surface properties, and cohesion may all influence its movement. A tracer with substantially different properties may show a convincing pattern without representing the real component.

Where tracer results support a quantitative conclusion, compare the tracer distribution with an independent assay of the relevant component.

Repeatability and Statistical Interpretation

For an initial quantitative comparison, test at least three independently prepared bottles per filling condition. Comparing fast and slow filling therefore requires at least six test bottles.

Analyze the same defined regions in each replicate using a consistent method.

Where target-component concentration is expressed in weight percent, one simple descriptive measure is the vertical concentration difference:

Δc(vertical) = c(top) − c(bottom)

For radial segregation, use:

Δc(radial) = c(center) − c(periphery)

Report the vertical difference when examining top-to-bottom segregation and the radial difference when investigating center-to-periphery separation. If both patterns matter, report both. The signs indicate the direction of enrichment, and the values are expressed in percentage points.

For example, a top sample containing 20 wt% of a component and a bottom sample containing 14 wt% gives a vertical difference of +6 percentage points.

For each condition, report the mean difference across replicate bottles and its standard deviation. Compare the results with baseline blend variability, sampling variability, and analytical repeatability.

Three replicate bottles provide an initial estimate of variability. If the differences between conditions are small relative to the observed spread, increase the number of replicates before making a quantitative decision.

On a dedicated spectral tester, Johanson reported test-to-test errors of about 0.5% for a moderately segregating material and 7% for a badly segregating one [7]. Those figures describe that instrument, not a bottle. Set an internal threshold only after measuring the repeatability of your complete procedure.

How to Interpret the Results

The purpose of the screen is to identify conditions that deserve closer investigation.

ObservationPossible interpretationNext step
Coarse particles concentrate around the perimeterSifting, trajectory, or rolling segregation may be contributingInvestigate radial segregation using suitable sampling or ASTM D6940
Fine particles concentrate near the surfaceAeration or fluidization may be involvedExamine air entrainment and consider ASTM D6941
Fast and slow filling produce different patternsSegregation is sensitive to the tested filling conditionsRepeat the comparison and assess other changed variables
Both bottles appear uniformNo visually detectable separation occurredUse analytical methods if composition uniformity matters
Replicates produce inconsistent resultsThe material, procedure, or measurements may be variableInvestigate starting blend uniformity and test repeatability

A radial concentration pattern may result from sifting, particle trajectories, rolling, or a combination of mechanisms. Likewise, a surface layer enriched in fines may indicate aeration, but it does not prove that fluidization is the only cause.

The practical question is whether the observed differences are consistent enough to justify further testing and relevant to the actual transfer.

Why Drop Height and Fill Rate Have No Universal Optimum

Reducing drop height is often proposed as a simple segregation remedy. Published results show why that advice requires qualification.

In a study of pharmaceutical powders falling through a vertical chute with a closed bottom, Jaklič and colleagues found that segregation patterns and their magnitude depended on the material [8].

An earlier die-filling study by Lawrence and Beddow produced a different observation. Increasing the fall height promoted mixing and decreased segregation under the investigated conditions. Faster filling also reduced the opportunity for segregation [9].

Shinohara and Golman examined mixtures during filling of a two-dimensional hopper. Their segregation indices increased at lower feed rates [10].

These findings involve different materials, geometries, and flow conditions. Changing the drop height can alter impact energy, particle trajectories, and air entrainment. Changing the filling rate affects pile formation, residence time, and particle movement.

Consequently, a shorter drop or slower stream is not automatically better. The useful operating range must be established for the specific material and transfer geometry.

Process Changes Worth Testing

Once the suspected segregation pattern has been identified, several modifications may be worth investigating. Each should be evaluated against the mechanism it is intended to control.

Change the Filling Geometry

A sock, spout, baffle, or alternative chute arrangement can change the path of the powder stream and the way the receiving pile forms. Reducing free fall may help in some systems, while changing the landing position or distributing the stream more evenly may be more effective in others.

Where sifting creates radial segregation, reducing the formation of a steep central pile is one possible approach. Jenike & Johanson describes how fines can accumulate beneath the pile impact point while coarse particles move toward the perimeter [11].

Any proposed change should be evaluated against the actual transfer, including possible increases in wall contact, attrition, or dust formation.

Examine Hopper Discharge

Mass flow can reduce side-to-side segregation in the discharged stream because material across the bed moves rather than leaving large stagnant regions. This is particularly useful for radial segregation during near-steady filling and discharge.

It does not automatically correct top-to-bottom gradients, such as fines concentrated near the surface by fluidization, because material from different heights may discharge in sequence.

Even radial segregation can persist during batch operation. Johanson found that conventional mass-flow designs may not adequately suppress segregation when a hopper is filled and then completely emptied. In such cases, the discharge velocity profile and hopper geometry become decisive [12].

The appropriate intervention therefore depends on the segregation pattern and whether the process operates continuously or in batches.

Control Aeration and Air Movement

Where fine particles remain suspended or move with displaced air, investigate receiving-bin venting, air release, and the way the stream enters the container. Fluidization segregation can cause finer particles to remain near the surface while coarser particles settle [4, 11].

Changes to venting or inlet geometry may reduce this behavior. However, they should be selected according to the actual gas-solid flow conditions rather than applied as general remedies.

Review Particle Properties

If process changes do not adequately reduce segregation, investigate whether particle-size differences, density, shape, surface properties, or cohesion are contributing.

Possible formulation changes include narrowing selected particle-size fractions, modifying particle surfaces, or granulating an ingredient. For example, making particles more cohesive may reduce one form of sifting while introducing difficulties in blending, feeding, or discharge. Evaluate such changes against the product’s complete processing and quality requirements.

Validate the Findings in the Real Process

A bench observation becomes useful when it helps explain a measurable problem in production. The next step is to establish whether the same segregation pattern occurs under representative operating conditions.

Evidence from pharmaceutical direct-compression blends illustrates the importance of process context. Kong and colleagues compared sifting and fluidization segregation tests with tablet content uniformity for three formulations. Although fluidization testing produced pronounced composition differences, those differences were not reflected in the tablet results. Instead, the tablet data correlated more closely with the sifting test, apparently because pilot-scale handling minimized fluidization [13].

Once a promising change has been identified, evaluate it under representative operating conditions. Establish the blend composition before transfer, record the operating settings, and collect samples during filling or discharge using a sampling plan capable of detecting the relevant spatial or time-dependent pattern.

Compare the production results with the bottle observations. If the pattern and improvement are reproducible, the modification can be considered for formal process qualification.

For a more detailed diagnostic route, see Segregation and Blend Inconsistency.

Quick Checklist

Before interpreting a bottle screening result, confirm the following:

  • ☐ The material and test arrangement are suitable for safe handling.
  • ☐ The starting blend condition and composition are known.
  • ☐ Representative portions have been prepared using an appropriate splitting method.
  • ☐ Powder mass, filling rate, drop height, and container geometry are recorded.
  • ☐ The filling procedure does not introduce uncontrolled changes between conditions.
  • ☐ The powder has not been tapped or vibrated after filling.
  • ☐ Vertical and radial segregation patterns have both been considered.
  • ☐ Intrusive sampling is avoided unless its effect has been assessed.
  • ☐ Any tracer has been checked against the actual component.
  • ☐ Quantitative comparisons include replicate data and measurement variability.
  • ☐ The results are confirmed under representative process conditions before significant equipment or operating changes.

Conclusion

The bottle method provides a quick way to compare powder-filling conditions and identify visible segregation patterns. It is especially useful when the engineering question concerns stream position, fill rate, or pile formation.

Its main limitation is the narrow view it provides of the powder bed. Radial segregation may remain hidden, visually similar components can separate without obvious banding, and intrusive sampling can alter the pattern being measured.

For quantitative investigations, establish the starting composition, match sampling to the expected segregation mechanism, and determine the repeatability of the complete procedure. Use ASTM D6940 for sifting segregation and ASTM D6941 for fluidization segregation where their scopes apply.

The final decision belongs to the real process. A useful bottle result identifies a variable worth testing, but reproducible production measurements determine whether the proposed change improves blend uniformity.

Frequently asked questions

No. A bottle test exposes the blend to a limited set of filling conditions and reveals only separation that can be observed or measured with the chosen procedure. Radial segregation may be hidden inside the bed, while components with similar appearances may separate without visible banding. Other handling stages can also introduce segregation even when the bottle result appears uniform.

No. Lowering drop height changes particle impact, trajectories, and air entrainment, but its effect depends on the material and equipment geometry. Published experiments have found that increasing fall height can reduce segregation under some conditions [9]. Compare representative heights while keeping other variables controlled.

Start with at least three independently prepared bottles per condition. For a comparison between fast and slow filling, this means at least six bottles. Three replicates provide only an initial estimate of repeatability, so additional runs may be needed when differences between conditions are small relative to measurement variability.

Inserting a tube disturbs the powder bed and may preferentially collect particles of certain sizes or properties. The sample can also include material displaced from other regions during insertion or withdrawal. Consequently, the measured composition may differ from the original local composition. Use an appropriately verified sampling or sectioning method when spatial concentration measurements are required.

Use ASTM D6940 when the decision requires a standardized evaluation of sifting segregation and the material falls within the method’s scope, including its 3 mm maximum particle-size guidance. ASTM D6941 is intended for evaluating fluidization segregation in powders capable of being fluidized. Where neither method adequately represents the handling conditions, develop a process-specific sampling and validation approach.

References

  1. Muzzio, F. J., Robinson, P., Wightman, C., & Brone, D. (1997). Sampling practices in powder blending. International Journal of Pharmaceutics, 155(2), 153–178. DOI04865-5).
  2. Boehm, G., et al. (2003). The use of stratified sampling of blend and dosage units to demonstrate adequacy of mix for powder blends. PDA Journal of Pharmaceutical Science and Technology, 57(2), 64–74. PQRI document.
  3. ASTM International. (2020). ASTM D6940/D6940M-20: Standard Practice for Measuring Sifting Segregation Tendencies of Bulk Solids. ASTM.
  4. ASTM International. (2019). ASTM D6941-19: Standard Practice for Measuring Fluidization Segregation Tendencies of Powders. ASTM.
  5. Pica Ciamarra, M., De Vizia, M. D., Fierro, A., Tarzia, M., Coniglio, A., & Nicodemi, M. (2006). Granular species segregation under vertical tapping: Effects of size, density, friction, and shaking amplitude. Physical Review Letters, 96, 058001. DOI.
  6. He, X., Han, X., Ladyzhynsky, N., & Deanne, R. (2013). Assessing powder segregation potential by near infrared (NIR) spectroscopy and correlating segregation tendency to tabletting performance. Powder Technology, 236, 85–99. DOI.
  7. Johanson, K. (2014). Review of new segregation tester method by Dr. Kerry Johanson, P.E. Powder Technology, 257, 1–10. DOI.
  8. Jaklič, M., Kočevar, K., Srčič, S., & Dreu, R. (2015). Particle size-based segregation of pharmaceutical powders in a vertical chute with a closed bottom: An experimental evaluation. Powder Technology, 278, 171–180. DOI.
  9. Lawrence, L. R., & Beddow, J. K. (1969). Powder segregation during die filling. Powder Technology, 2(5), 253–259. DOI80036-7).
  10. Shinohara, K., & Golman, B. (2002). Segregation indices of multi-sized particle mixtures during the filling of a two-dimensional hopper. Advanced Powder Technology, 13(1), 93–107. DOI.
  11. Jenike & Johanson. Cure Segregation and Improve Blending. Technical guidance. Source.
  12. Johanson, K. D. (2019). New mass flow limiting lines based on segregation pattern and magnitude. Powder Technology, 357, 240–247. DOI.
  13. Kong, A., Anwar, H., & Narayan, P. J. (2006). Assessment of the segregation potential of pharmaceutical direct compression blends. AIChE Annual Meeting, paper 677b. Conference abstract.

Read more

Close-up photo of a segregated powder heap showing fine particles on top and coarse granules around the base, illustrating particle size separation.

Segregation Mechanisms and Prevention

Understand how particle properties and handling conditions drive separation in powder blends.

Particle size distribution interpretation graphic showing D10, D50, D90, fines, oversize, span, and curve shape in a powder PSD curve.

Interpreting Particle Size Distributions

Connect distribution tails and size differences with the behavior of powder mixtures.

Assessing Powder Packing Density

Powder Packing Density

Explore how particle size, shape, and packing affect bulk powder behavior.