Specimen density reproducibility testing is a critical quality assurance procedure that verifies whether the density of molded test specimens can be consistently replicated across different batches, operators, and laboratories. This comprehensive guide covers the sample types, testing items, standardized methods, statistical evaluation, and practical applications of density reproducibility experiments for polymeric and composite materials, with reference to internationally recognized standards such as ISO 1183, ASTM D792, and ISO 5725.
Technical Overview
Density is one of the most fundamental physical properties of polymeric materials, and its measurement on molded test specimens (commonly referred to as splines or test bars) serves as a sensitive indicator of both material composition and processing consistency. A specimen density reproducibility experiment is a structured, statistically designed study that determines whether a laboratory, a production line, or a group of laboratories can obtain consistent density values on nominally identical specimens under varying conditions such as different operators, different days, different molding machines, and different measurement instruments.
Reproducibility is formally defined in ISO 5725-1 as the closeness of agreement between independent test results obtained under reproducibility conditions, meaning conditions where test results are obtained with the same method on identical test items in different laboratories with different operators using different equipment. This is distinct from repeatability, which refers to results obtained by the same operator with the same equipment within a short time interval. A well-designed reproducibility experiment isolates the sources of variation contributed by specimen preparation, environmental conditioning, measurement technique, and instrumentation, allowing manufacturers to establish realistic tolerance limits for material datasheets, incoming material inspection, and process control.
The experiment typically combines specimen preparation standards such as ISO 294-1 for injection molding of tensile bars and ISO 293 or ASTM D4703 for compression molding of plaques, with density determination standards such as ISO 1183-1, ISO 1183-2, ISO 1183-3, ASTM D792, and GB/T 1033. The statistical treatment follows ISO 5725-2, which prescribes the estimation of repeatability limits and reproducibility limits from interlaboratory data sets.
Testing Samples
Injection-Molded Multipurpose Specimens
The most common sample type is the ISO 3167 multipurpose test specimen, molded as ISO 294-1 Type 1A or 1B tensile bars. These specimens are produced under controlled melt temperature, mold temperature, injection speed, and holding pressure, because density is strongly influenced by the degree of packing and the resulting void content and shrinkage. Multiple molding shots are retained and labeled so that shot-to-shot variation can be separated from measurement variation during the statistical analysis.
Compression-Molded Plaques and Sheets
For materials that are sensitive to orientation and packing effects, such as thermosets, high-viscosity engineering plastics, and compounds intended for property comparison, compression-molded plaques prepared in accordance with ISO 293 or ASTM D4703 provide a lower-stress specimen geometry. Plaque thickness, cooling rate, and applied pressure are documented as critical process parameters that directly affect measured density.
Filled and Reinforced Compounds
Glass-fiber reinforced, mineral-filled, and flame-retardant compounds are frequently evaluated in density reproducibility studies because filler dispersion and fiber distribution directly influence both the mean density and its scatter. Specimens are drawn from different positions within the molding to assess spatial homogeneity.
Material Families
- Commodity thermoplastics such as polyethylene (PE), polypropylene (PP), polystyrene (PS), and PVC compounds
- Engineering thermoplastics such as polyamide (PA), polycarbonate (PC), polyoxymethylene (POM), polybutylene terephthalate (PBT), and ABS
- High-performance polymers such as PEEK, PPS, and fluoropolymers where density changes of less than 0.002 g/cm³ can be significant
- Thermosetting materials including phenolics, epoxies, and unsaturated polyester molding compounds
Testing Items
The core deliverables of a specimen density reproducibility experiment include both individual density values and the statistical parameters that quantify their dispersion:
- Density of each specimen at the standard reference temperature of 23 °C, expressed in g/cm³ or kg/m³
- Relative density (specific gravity) as the ratio of specimen density to the density of water at the reference temperature
- Within-laboratory repeatability standard deviation (sr) and repeatability limit (r), derived from repeated measurements performed by a single operator on a single instrument
- Between-laboratory reproducibility standard deviation (sR) and reproducibility limit (R), derived from measurements performed across multiple operators, instruments, or laboratories
- Coefficient of variation (CV) of the density data set, commonly required to remain below 0.5% for well-controlled thermoplastic specimens
- Mandels h and k statistics for graphical consistency checking of laboratory bias and precision, as prescribed by ISO 5725-2
- Correlation of density variation with molding parameters such as holding pressure, melt temperature, and cooling time, when the experiment includes a designed process variable
Testing Methods
Specimen Conditioning
All specimens are conditioned prior to measurement in accordance with ISO 291 or ASTM D618, typically at 23 ± 2 °C and 50 ± 10% relative humidity for a minimum of 40 hours, unless the material standard specifies otherwise. Hygroscopic materials such as polyamides receive extended conditioning because absorbed moisture measurably reduces apparent density. Specimens are visually inspected and, where necessary, weighed to confirm the absence of voids and gross surface defects.
Method A: Immersion (Hydrostatic Weighing)
The immersion method described in ISO 1183-1 Method A and ASTM D792 Method A is the primary technique for solid specimens. Each specimen is weighed in air using an analytical balance, then re-weighed while fully submerged in a wetting liquid, usually distilled water containing a trace of surfactant or, for materials denser than water, a suitable replacement liquid. Density is calculated from the two masses and the known liquid density at the test temperature. The submerged weighing fixture is degassed carefully to eliminate buoyancy errors from attached air bubbles, and the balance is calibrated with a certified reference weight traceable to national standards.
Method B: Liquid Pycnometer
For pellets, powders, and small cut pieces of specimens, the liquid pycnometer method of ISO 1183-1 Method B offers a practical alternative. The technique compares the mass of the displaced liquid with the mass of the specimen and is particularly useful when whole specimens cannot be conveniently suspended for hydrostatic weighing.
Method D: Density Gradient Column
The density gradient column technique of ISO 1183-2 provides the highest sensitivity, resolving differences on the order of 0.0001 g/cm³. Calibrated glass floats establish a linear density gradient within a vertical liquid column, and specimen fragments settle at the height corresponding to their density. This method is frequently selected for crystallinity studies and for resolving very fine batch-to-batch differences in polyolefins.
Method C: Titration and Gas Pycnometry
The titration method of ISO 1183-1 Method C establishes density by observing flotation equilibrium in liquid mixtures of adjusted composition. Gas pycnometry using an inert gas displacing medium can be applied as a supplementary method for specimens with open porosity where liquid immersion would cause absorption.
Experimental Design and Statistical Evaluation
A typical reproducibility experiment follows a nested or fully crossed design: each participant measures the density of several specimens from each of several shots or batches, in duplicate or triplicate, on different days. Data are screened for outliers using the statistical tests specified in ISO 5725-2, and the remaining results are used to compute the repeatability limit r and reproducibility limit R, which correspond to the 95% probability limits within which two results may be expected to agree. Results are reported with the mean density, the standard deviations, the number of specimens, the test temperature, and the complete description of the measurement method so that the study can be reproduced independently by accredited third-party laboratories or by customers verifying supplier data.
Applications
- Injection molding process validation, confirming that machine settings and tooling produce specimens with stable, predictable density across production runs
- Compound development and formulation screening, where density tracks filler loading, phase morphology, and dispersion quality
- Lot acceptance and incoming material inspection, using statistically justified acceptance thresholds derived from reproducibility limits
- Generation and verification of single-point and multi-point datasheet values in accordance with ISO 10350 and ISO 11403
- Interlaboratory proficiency testing and comparison programs among accredited third-party laboratories, supporting laboratory accreditation and method validation
- Crystallinity and thermal history studies in semicrystalline polymers, where density correlates with the degree of crystallization
- Quality investigation of voiding, sink marks, and underpacking in molded parts, using specimen density as a diagnostic metric
- Support for regulatory submissions and material certifications where consistent, traceable density data across laboratories is required
FAQ
What is the difference between repeatability and reproducibility in density testing?
Repeatability describes the agreement between repeated density measurements performed by the same operator, on the same instrument, within a short period, using identical specimens. Reproducibility describes the agreement between results obtained in different laboratories or with different operators and equipment. Both are quantified in ISO 5725 through the limits r and R, and a complete reproducibility experiment evaluates both levels simultaneously.
How many specimens are required for a valid reproducibility study?
As a minimum practical design, ISO 5725-2 recommends participation of at least five to eight laboratories, each testing two to three replicate specimens per material level. For internal single-laboratory studies, at least ten specimens drawn from at least three separate molding shots measured on two or more days provide a reasonable basis for estimating within-lab components of variation.
What level of density variation is considered acceptable?
For well-molded unfilled thermoplastics, a coefficient of variation below 0.5%, corresponding to an absolute spread of roughly 0.005 g/cm³ around a typical density of 1.0 g/cm³, is generally regarded as satisfactory. Filled compounds and hygroscopic polymers may exhibit somewhat wider limits, which should be established from the reproducibility data itself rather than assumed.
Why is specimen preparation included in the experiment instead of testing only the measurement method?
Because injection and compression molding parameters such as holding pressure, melt temperature, and cooling rate directly determine the packing, void content, and crystallinity of the specimen, isolation of the measurement method alone would understate the true variability experienced by end users. A comprehensive experiment therefore treats molding as an explicit, controlled factor in the statistical design.
Which density method should be selected for a reproducibility experiment?
The immersion method of ISO 1183-1 Method A or ASTM D792 Method A is the default choice for solid molded specimens because of its speed, accuracy, and wide availability. The density gradient column of ISO 1183-2 should be selected when the expected differences between specimens approach 0.001 g/cm³ or finer, and the pycnometer method is preferred when specimens must be reduced to fragments or pellets.
How does temperature affect the results?
Density is temperature dependent because both the specimen and the immersion liquid expand thermally. All measurements are therefore performed at a controlled temperature, normally 23 °C, with the liquid density corrected to the actual bath temperature, and the test temperature is always reported alongside the result to ensure comparability between laboratories.