Sampling Strategy and Acceptance Criteria for PTFE Blend Homogeneity (Upper/Middle/Lower × Center/Edge)

Jun 11, 2026

Leave a message

Below is a professional English translation with consistent engineering language and a fluid technical writing style:

Assessing the Homogeneity of a PTFE Powder Blend: Sampling Procedure and Acceptance Criteria (Filled PTFE Systems)
1. What is understood by "homogeneity" in PTFE blends?
The homogeneity of PTFE powder mixtures is commonly described by one or more critical quality characteristics (CQAs) or marker components:

1.1 Filler Content (wt % or vol %)
The most popular indication, especially for systems that contain;

Fibreglass

Graphite

Bronze

Carbohydrate

Molybdenum disulphide (MoS2) etc.

These fillers represent a major source of segregation danger.

Filler content can be measured by:

Loss on Ignition (LOI)

Pyrolysis under an inert atmosphere

Thermogravimetric analysis (TGA)

ASTM D4745 gives advice on the determination of filler content in filled PTFE using combustion/residue-based procedures including sample mass, temperature control and inert atmosphere conditions.

1.2 Lubricant Content (for paste extrusion systems)
Lubricants for paste extrusion grades are often Isoparaffinic oils, mineral oils or alcohol based systems .

Typical amounts of lubricant are approximately 15-20 wt %.

Reasons for poor mixing:

Variations of extrusion pressure

Nonuniform density preform

Abnormal porosity and shrinkage after sintering

1.3 Further Optional Indicators
Depending on the risk profile of the product.

Colour uniformity (ΔE of pigmented systems)

Ash content / elemental composition (XRF / ICP)

Tapped density / bulk density

For formulations with considerable changes in density (e.g. bronze fillings) or large variances in morphology (e.g. glass or carbon fibres), filler content and elemental analysis are the preferable indicators since they are particularly sensitive to segregation (top/bottom or radial variation).

2. Principles of Sampling: The Choice of Dynamic Flow Sampling
2.1 General Principle
Powder sampling is quite prone to bias. Two principles are at work here:

Sampling of flowing material streams (discharge or transfer flow) is preferred.

Don't only sample from static powder beds

"Each sample should be cut across the whole cross-section of the stream and taken over multiple increments, rather than one scoop,"

ISO 3954 (which is a widely used standard for powder sampling principles in powder metallurgy) recommends:

Sampling from the continuous flow of discharge whenever possible.

At least three periods, including early, middle and late discharge stages

2.2 Issues With Bed Sampling
In case a sampling has to be done from a static container or mixer:

The sampling depth, insertion angle and withdrawal speed need to be properly standardised

Static sampling is susceptible to electrostatic effects, wall friction and particle segregation

Operator training and procedure consistency can help minimise unpredictability

3. Sampling Layout: Top / Middle / Bottom × Center / Edge (6 Point Model)
3.1 Definition of Coordinate (Standardisation Needed)
For cylindrical bins or mixing vessels:

Vertical Levels
Upper (U): ~15-20% below surface

Middle (M) ~45-55 % depth

Lower (L): ~80-85% depth
You can't only sample the dead zones at the absolute bottom. The lower level still needs to reflect areas prone to segregation.

Positions Radial
Center: r \leq 0.2R (near axis region)

Edge: r ≈ 0.7R (recommended typical radius, no wall scraping)

ISO 3954 also permits the use of ~0.7R as a typical sampling position in cylindrical containers.

3.2 Sample Locations for Standard 6
Minimum required configuration:

UC: Upper Campus

UE: Upper Edge

MC: Middle Center

ME: Mid-Edg

LC: Low Center of Gravity

LE: Low-Edge

This layout is allowing to recognise:

Vertical segregation trends (U>M>L)

Radial segregation (Centre vs Edge disparities)

3.3 Static Sampling Procedure
Insert closed port sampler to target depth

Open sampler to acquire material then close before withdrawal

Avoid contamination from top layers during removal

Less unpredictability with numerous increments (≥2) at each place integrated into a composite sample.

Typical sample weight: 20-50 g per site

Manual mixing should not be used for homogenising and reducing composite samples, rotating sample dividers should be used.

4. Improved Sampling Methods for High-Volume Production: Discharge Flow Sampling (Recommended)
In many systems segregation can diminish apparent homogeneity after mixing during discharge.

Therefore, the addition of discharge sampling is strongly recommended:

D1: First third of the discharge

D2: Mid emission

D3: Last third of discharge

ISO 3954 also allows multi-increment sampling over the whole of the discharge.

Major Qualification:
Each increment must be a complete cut through the whole powder stream cross section, not a partial scoop.

5. Evaluation Metrics & Acceptance Criterias
5.1 Statistical Measures
For sample outcomes (x_i):

Mean: ( $barx$)

Standard deviation (SD)

Relative standard deviation (RSD %):

[RSD = fracSDbarx times 100%]

Also worth a read:

Maximum relative deviation :
[max left| fracx_ibarx - 1 right|]

5.2 Framework Proposed for Acceptance
A) Process Qualification/New Formulation/Equipment Change (Stringent)
Typical sample size:

≥10 data points (e.g. 6 static + 3 discharge + 1-3 more points)

Replication of robustness suggested

Industry guidelines (e.g., ISPE mixing uniformity concepts) generally recommends several sampling locations and replicates for validation.

Acceptance criteria:

SD < 3% of desired value ⇒ Uniformity good

3% < SD <= 5% --> Acceptable, but requires inquiry

SD > 5% => Non-conforming, process change needed

B) Routine Batch Release (Practical Control)
Typical sampling:

Minimum: 6 points static OR 3 points discharge

Preferred: 6 pt + 3 pt discharge (9 in all)

Acceptance criteria:

RSD < 5% (common industrial limit)

Optional tightening to <= 6% depending on product risk.

Set single-point content restrictions depending on specification (e.g. +/- 5% relative window as a starting point, refined by SPC capability analysis)

6. Diagnostic Interpretation ("Segregation Fingerprints")
Sample Patterns
6.1 Heavy Bottom/Edge Enrichment (LE high, UC low)
Common reason:

Density driven filler settling

Re-segregation during Discharge or Vibration

Action Taken: Corrective

Minimise mechanical vibration in transport

Improved sampling of whole stream discharges

Optimise the mixing and resting period before discharge

6.2 Edge enrichment at all levels
Normal cause:

Radial segregation caused by wall effects

Electrostatic or fibre-orientation effects

Action to be taken:

Standardise edge sampling at 0.7 R (no wall scraping)