How to Test and Grade Raw Silk for Quality Assurance

Raw silk doesn’t reach the loom by accident. Before a single thread enters a weaving shed, it passes through a rigorous testing and grading process designed to confirm its size, strength, cleanliness, and consistency. This quality control step is the backbone of fair trade in the silk market – it protects buyers from substandard products, gives producers measurable feedback for improvement, and ensures that every lot of silk delivered meets the specifications demanded by the textile industry. Understanding how raw silk is tested and graded reveals just how precise and science-driven sericulture truly is.

Why raw silk testing matters 🔗

The silk textile industry operates on precision. Handlooms and sophisticated power looms produce a wide range of fabrics – each requiring a specific thread thickness, tensile strength, and surface quality. According to the FAO Silk Reeling and Testing Manual, raw silk must be tested and classified so that the weaving industry can select the right material for the right product, and so that buyers can purchase at internationally accepted grades. Without standardized testing, there would be no reliable way to price silk fairly or maintain quality across supply chains. Testing also generates the kind of structured feedback that helps reelers identify and fix problems in their process – from water quality and cocoon preparation to reeling speed and machine calibration.

The international framework for silk testing 🔗

Raw silk testing is not an informal process. It follows protocols established by the International Silk Association (I.S.A.), whose standardized methods were first codified in a bulletin issued in 1949-1950. The I.S.A. convened its landmark international conferences in Zurich (1949) and New York (1950), bringing together silk-producing countries like China, Japan, Italy, and France alongside major consuming nations to build a globally accepted testing framework. The methods have been revised over the decades – most notably in 1961 and 1974 – and electronic testing systems have since been incorporated. China also follows its own parallel standard, GBT 1797-2008, while India operates under Bureau of Indian Standards guidelines. The core principles, however, remain consistent worldwide.

Testing is applied to defined consignments. A standard testing unit for export raw silk is typically 5 to 10 bales of 60 kg each, or 20 bales of 30 kg each. Two major tests are carried out on every consignment: the conditioned weight test (a quantitative test to determine moisture-corrected weight) and the quality test, which examines all the physical properties of the silk thread.

The quality test: what inspectors actually examine 🔗

The quality test begins with a full-lot visual inspection, conducted in a well-lit inspection room. Three elements are assessed during this stage: uniformity of colour, lustre and feel; general finish (covering defects in re-reeling, arrangement, and physical damage); and the nature of the silk – including colour depth (light, medium, or deep), lustre (bright, medium, or dull), and hand (smooth, medium, or rough). After the lot-level inspection, individual sample skeins are drawn and subjected to a battery of precise tests.

Size and size deviation (denier) 🔗

Size is the most fundamental measurement in raw silk testing. The denier system is the internationally used method to measure thread size – it is expressed as the weight in grams of a 9,000-metre length of yarn, or more practically, as the weight of a 450-metre skein. A thread weighing 1 gram per 9,000 metres is 1 denier; a 21-denier silk is considerably thicker than a 14-denier one. The size of the raw silk determines which fabric constructions it is suitable for, and any inconsistency in size from one skein to another signals problems in the reeling process.

The size deviation test measures precisely this inconsistency. For raw silk of 33 denier or finer, 200 sizing skeins of 450 metres each are drawn and weighed; for coarser silk (34 denier and above), 400 skeins of 112.5 metres are taken. Each skein is weighed individually and in grouped lots on a calibrated quadrant scale. A statistical formula is then applied to derive the average size deviation – the spread of individual skeins around the mean denier – and the maximum size deviation, which captures the extremes. A lot with high size deviation will produce uneven fabric; therefore, low deviation is a marker of high quality.

Evenness 🔗

Evenness refers to the consistency of thread diameter along the length of the yarn – not just from skein to skein, but within individual threads. Under standard illumination, changes in silk diameter affect how light passes through or reflects off the thread, allowing inspectors to detect fine passages (thin sections) and coarse sections by observing brightness variations. Test samples are wound onto inspection panels called seriplane boards and compared against Standard Variation Photographs – reference images that define the allowable degree of variation. Better evenness results in smoother, more uniform fabric with more balanced colour absorption during dyeing – a major concern for premium textile manufacturers.

Evenness defects are classified into three levels of severity (Evenness Variation I, II, and III), and their frequency across 100 panels is recorded as a percentage.

Cleanness 🔗

Cleanness measures the presence of physical defects in the silk yarn – foreign matter, gum lumps, split ends, loops, and other structural flaws. Defects are classified as super major, major, or minor, and each category carries a different penalty: a super major defect deducts 1.0 percentage point, a major defect 0.4 points, and a minor defect 0.1 points from a starting score of 100. The result is a cleanness percentage – a lot scoring 95% cleanness has fewer and less severe defects than one scoring 80%. This score directly influences the final grade assigned to the lot.

Neatness 🔗

Neatness covers imperfections that are smaller than those classified as cleanness defects but still affect the yarn’s surface quality. The five neatness defect types include nibs (small thickened spots less than 2 mm), loops (small open places less than 10 mm), and similar minor irregularities. Like cleanness, neatness is scored in percentages by comparing panels against standard photographs at a fixed distance of 0.5 metres. Both the average neatness percentage and the low neatness percentage (average of the lowest 20 panels) are recorded, as these are used directly in grade determination.

Tenacity and elongation 🔗

The strength of silk is measured through the tenacity and elongation test, conducted on an instrument called the Serigraph. The typical tenacity of a raw silk bave (filament) ranges from 3.6 to 4.8 grams per denier, while its elongation – the amount it can stretch before breaking – is between 18 and 23 percent of its original length. The test is carried out at a controlled temperature of 20°C and 65% relative humidity to eliminate the effects of moisture on results, since excess moisture increases elongation while reducing tenacity. Strong, elastic silk performs better during weaving and produces more durable end products.

Cohesion test 🔗

The cohesion test measures how well the individual cocoon filaments within a raw silk thread adhere to one another. Using a Duplan cohesion tester, the number of friction cycles required to split the silk thread is counted – higher friction cycles indicate better cohesion between filaments. Poor cohesion results in weak, easily fraying yarn that causes problems in weaving. The test is performed under controlled humidity to ensure consistency.

How raw silk is graded 🔗

Once all test results are compiled, the raw silk lot receives a formal grade. Under I.S.A. standards, grades are expressed as 4A, 3A, 2A, A, and B; the Japanese grading system extends this to 5A, 4A, 3A, 2A, A, B, C, and D. China’s GBT 1797-2008 Raw Silk Quality Standards use an even broader scale, running from 6A at the top through 5A, 4A, 3A, 2A, A, B, C, D, E, and F.

For silk of 33 denier or finer (Categories I and II), the grade is determined by the lowest result among six parameters: size deviation, evenness I, evenness II, cleanness, average neatness, and low neatness. For coarser silk (Category III, 34 denier and above), maximum size deviation is added as a seventh parameter. If any single parameter falls below the minimum threshold for a grade, the entire lot is moved to the next lower grade – a rule that enforces genuine quality consistency rather than allowing a strong result in one area to compensate for weakness in another.

What the top grades look like 🔗

Grade 6A represents the pinnacle of quality, accounting for only about 5-10% of global silk production. It is characterized by long filaments, minimal size deviation, high evenness and cleanness, strong tenacity, and a pearly white colour with a natural luminous shine, free from impurities. Grades such as 3A and 2A represent solid commercial quality, widely used in premium textile manufacturing, while B, C, and lower grades are typically used in industrial applications or blended products.

Electronic testing: the modern shift 🔗

The traditional testing approach – visual panel inspection, manual weighing, and the Serigraph – remains the backbone of ISA-certified testing. However, modern silk testing facilities have introduced electronic systems that significantly improve speed and precision. ISO 15625:2014 specifies an electronic test method using capacitive and optical testers for detecting defects and measuring evenness of raw silk, applicable to yarn sizes between 12 and 69 denier, whether in skein or on cone.

Capacitive testers detect mass variation in the silk thread as it runs through a sensor, while optical testers capture cross-sectional area variation using photometric analysis – the two methods complement each other, as the optical tester can identify defect shapes that the capacitive tester may miss, and vice versa. These electronic systems reduce operator subjectivity, enable higher-volume testing, and generate detailed statistical data on defect frequency and distribution across a lot.

Testing as a tool for quality improvement 🔗

Raw silk testing is not only about classifying what has already been produced – it is a diagnostic tool. When a lot consistently fails on size deviation, the problem likely lies in the reeling speed or cocoon selection. Repeated cleanness failures point to issues with the reeling water, machine calibration, or cocoon cooking. Testing and classification methods based on evenness, cleanness, and neatness provide the primary basis for both grading and price fixation in the raw silk trade. For producers aiming to compete in international markets, achieving and maintaining high grades is a direct economic objective – not just a quality aspiration.

India’s sericulture sector, in particular, has been working to close the quality gap with Chinese raw silk by strengthening testing infrastructure at reeling units, investing in better cocoon grading, and training reelers to interpret test results and adjust their processes accordingly. The data generated by systematic testing is the clearest guide available to anyone seeking to move up the grade ladder.

What do you think? If a lot of raw silk fails on just one parameter – say, neatness – but scores exceptionally well across all other tests, should the grading system allow for any flexibility, or is the strictness of the “lowest score decides the grade” rule essential to maintaining market trust? And how might wider adoption of electronic testing methods change quality standards and pricing dynamics in India’s silk industry?

References

1. https://www.fao.org/4/x2099e/x2099e12.htm
2. https://www.ajol.info/index.php/ijest/article/download/210284/198230
3. https://fyitester.com/method-of-testing-for-raw-silk-2/
4. https://www.fao.org/4/x2099e/x2099e03.htm
5. https://ursilk.com/silk-grades-quality-guide/
6. https://mayfairsilk.com/pages/grades-of-silk
7. https://www.iso.org/standard/55447.html
8. https://cdn.standards.iteh.ai/samples/55447/875632c72ed24b489074006642cfd3c0/ISO-15625-2014.pdf