Key Takeaways
- Frame the question before choosing the instrument. Each technique answers a narrow question well and every other question badly.
- Visual and dimensional examination comes first — it is free, it is non-destructive, and it determines which instrumental tests are worth running.
- Identify, then compare. Nearly every conclusion depends on a known-good control run under identical conditions.
- Preserve samples before testing. Most analytical methods are destructive and the failed part is usually irreplaceable.
Analytical instruments do not solve failures; they answer questions. The difference matters because every technique in the plastics laboratory answers a narrow question well and most other questions badly, and a failure investigation that begins by sending parts out for “testing” without a defined question usually returns a folder of data and no conclusion. The efficient sequence is always the same: observe, hypothesise, then select the instrument that discriminates between the surviving hypotheses.
Before Any Instrument
Documented visual examination is the highest-value step in the whole investigation and it is free. Photograph the part as received, from multiple angles, with a scale. Record where the failure sits relative to gates, knit lines, ribs, bosses, holes and any secondary operation. Measure critical dimensions and compare them with the drawing. Examine the fracture or defect at 10–40x under a stereo microscope and classify what you see.
This step usually determines the entire subsequent plan. A crack originating at a molded-in stress concentration in a chemically exposed region suggests environmental stress cracking and points at DSC and chemical exposure testing. A clean interfacial separation with no polymer transfer suggests a surface problem and points at FTIR and surface analysis. A brittle fracture in a material that should be ductile suggests degradation, contamination, or a resin substitution, and points at TGA, DSC and melt flow measurement.
FTIR: What Is It, and What Is On It
Infrared spectroscopy is the general-purpose workhorse. It identifies polymer type by functional group signature, which confirms or refutes the assumption that the part is made of what the drawing says. It detects surface contamination directly — silicone, oils, release agents, residues from cleaning — which makes it the definitive test for the most common adhesion failure cause. It shows oxidation products that indicate thermal or UV degradation, and it can be run in reflectance mode on a fracture face without destroying the sample.
Its limits are worth knowing. It identifies chemistry, not quantity, without careful calibration. It cannot distinguish grades of the same polymer. And it reports what is in the sampled volume, which for a surface technique may be a few micrometers — useful when the question is about a surface layer, misleading when the question is about bulk composition.
DSC: Thermal History and Cure
Differential scanning calorimetry heats a few milligrams and measures the heat flow required, which exposes melting, crystallization and glass transition. In failure work its most common uses are confirming material identity, detecting a resin substitution that a supplier has not disclosed, quantifying crystallinity in semi-crystalline materials where cooling rate determines properties, and measuring residual cure in thermosets and UV-cured coatings.
Residual cure is a frequently overlooked cause of delayed failure. A coating or adhesive that continues to cure after assembly changes modulus and shrinks, and both effects load the interface. DSC quantifies how much cure was left, and a comparison against a properly cured control converts a suspicion into a measurement.
TGA: Composition by Mass
Thermogravimetric analysis heats a sample and records mass against temperature. Because polymer, plasticizer, filler and residue decompose or volatilise in distinct temperature ranges, the resulting curve is effectively a composition breakdown: volatiles and moisture, plasticizer content, polymer fraction, carbon black, and inorganic filler or glass residue.
This is the test that answers “is the glass loading what we specified?” and “how much plasticizer is left in this part?” It is also the standard way to check thermal stability, which matters when a material is being processed near its degradation threshold — a common root cause of brittle behavior and of the weak surface layers that produce substrate-mode adhesion failure.
SEM and EDS: Topography and Elements
Scanning electron microscopy provides depth of field and magnification that optical microscopy cannot approach, which matters when the question is where a crack started and how it traveled. Fracture surfaces carry directional features that point back to the initiation site, and identifying that site — a void, an inclusion, a gate mark, a machining score — frequently ends the investigation.
Energy-dispersive X-ray spectroscopy, usually available on the same instrument, adds elemental identification. It finds inorganic contamination, identifies metallic inclusions, maps filler distribution, and confirms silicon where silicone contamination is suspected. SEM/EDS is more expensive than the other techniques and it is destructive in the sense that samples must be sized and often coated, so it belongs later in the sequence rather than earlier.
Mechanical and Rheological Methods
Melt flow rate is cheap and disproportionately informative: a shift against the specification indicates degradation, contamination or a different grade, and it is often the fastest way to catch a material problem. Tensile, flexural and impact testing quantify whether the material meets its data sheet, which matters when the argument is about whether the part or the load was wrong. For bonded joints, lap shear and peel testing to a defined standard turn subjective assessments into numbers that can be compared across lots and suppliers.
Sequence and Sample Discipline
Two rules prevent most wasted testing. Preserve before you consume: most of these methods destroy the sample, and the failed part is frequently the only one that exists. Photograph, document, and set aside untouched material before the first test is run. Always run a control: almost every conclusion in this work is comparative. A spectrum, a thermogram or a melt flow number on the failed part means little without the same measurement on a known-good part, ideally from the same tool and preferably from a lot that performed correctly.
Investigations that observe both rules tend to close in days. Investigations that skip them tend to produce data that supports several competing explanations equally well, which is the practical definition of an inconclusive report.
Related Reading
- Adhesive Bond Failures
- Reading a Fracture Surface
- Environmental Stress Cracking
- Sabreen Engineering Services
- What a Defensible Failure Analysis Report Contains
Need help with this?
The Sabreen Group provides independent engineering support for analytical failure analysis and root-cause investigation of plastics components. Our engineering services team works with manufacturers on process development, material qualification and production troubleshooting. Contact us to discuss your application.
Frequently Asked Questions
Which test should we run first on a failed part?
None of the instrumental ones. Start with documented visual examination, dimensional measurement and low-magnification microscopy of the fracture or defect. That establishes the failure mode and location, which determines whether you need material identification, thermal history, contamination analysis or mechanical property data. Running instruments before framing the question is how testing budgets are consumed without conclusions.
What does FTIR actually tell us?
Fourier transform infrared spectroscopy identifies chemical functional groups, which means it identifies polymer type, confirms additives and fillers, detects contamination on a surface, and shows oxidation or degradation products. It is the most broadly useful single technique in plastics failure work, and it is fast and requires very little material.
When is DSC the right choice?
Differential scanning calorimetry measures thermal transitions, so it reveals melting point, glass transition, crystallinity and degree of cure. Use it to confirm material identity against a specification, to detect a resin substitution, to assess whether a semi-crystalline part was cooled correctly, or to check whether a thermoset actually finished curing.
How is TGA different from DSC?
Thermogravimetric analysis measures mass loss with temperature, so it quantifies composition: filler and glass content, plasticizer and volatile content, carbon black loading, and thermal stability. DSC tells you about transitions and cure; TGA tells you what the material is made of and how much of each component is present.
Do we need SEM for most investigations?
Not for most. Scanning electron microscopy earns its cost when fracture initiation site and crack propagation mechanism matter, when features are below the resolution of optical microscopy, or when elemental analysis via EDS is needed to identify inorganic contamination or filler distribution. For a straightforward adhesion or contamination question, optical microscopy plus FTIR usually settles it.
