The coefficient of linear thermal expansion (CTE) is one of the most critical thermomechanical properties of PCB substrates, directly governing plated-through-hole reliability, solder joint life, and resistance to delamination during thermal excursions. This article explains PCB CTE testing methods in detail, covering test standards such as IPC-TM-650 2.4.24 and ASTM E831, specimen preparation, TMA and dilatometer procedures, data calculation, and typical acceptance criteria. Whether you are qualifying copper-clad laminate, evaluating high-Tg or lead-free-compatible materials, or performing failure analysis on multilayer boards, this guide will help you select and execute the correct CTE test methodology.
Technical Overview
The coefficient of linear thermal expansion (CTE, commonly denoted as α) describes how much a material elongates per unit length for each degree of temperature change. It is defined as CTE = ΔL / (L0 × ΔT), where L0 is the original specimen length, ΔL is the dimensional change, and ΔT is the temperature change. For printed circuit board materials, CTE is expressed in parts per million per degree Celsius (ppm/°C) and is a primary figure of merit in material datasheets and procurement specifications such as IPC-4101.
PCB laminates are anisotropic composites. In the X and Y (in-plane) directions, the woven glass fabric constrains the resin system, so typical CTE values for FR-4 type materials range from 12 to 18 ppm/°C below the glass transition temperature. In the Z direction (through-thickness), expansion is dominated by the resin, producing values of roughly 40 to 70 ppm/°C below Tg. Above the glass transition temperature, the resin softens and the expansion rate rises sharply, frequently reaching 200 to 400 ppm/°C in the Z direction.
Because of this behavior, CTE results are reported in two segments: alpha 1 (α1), measured below Tg, and alpha 2 (α2), measured above Tg. Excessive Z-axis expansion during reflow soldering or thermal cycling stresses copper-plated via barrels, causing cracking and interconnect failure, while CTE mismatch between the board and attached components accelerates solder joint fatigue. Accurate CTE measurement is therefore essential for material selection, reliability prediction, and process qualification.
- X/Y CTE (in-plane): typically 12–18 ppm/°C for FR-4 below Tg
- Z CTE (through-thickness): typically 40–70 ppm/°C below Tg
- Alpha 2 (above Tg): 200–400 ppm/°C in the Z direction for standard resin systems
- Total Z expansion from 50 °C to 260 °C: commonly limited to 3.0–4.0% in IPC-4101 slash sheets for lead-free-compatible laminates
Testing Samples
Copper-Clad Laminate (CCL) Specimens
For raw material characterization, specimens are cut directly from laminate panels. In-plane (X/Y) specimens are commonly 6.4 mm × 6.4 mm (0.25 in × 0.25 in) squares or rectangular strips with a well-defined gauge length, while Z-axis specimens use the as-laminated thickness, ideally between 0.5 mm and 2.0 mm. Depending on the specification, the copper foil may be chemically etched away so the dielectric composite itself is characterized; tests with the foil retained can also be performed to reflect the as-supplied composite behavior.
Finished PCB and Multilayer Board Specimens
Specimens taken from finished printed boards evaluate the material after lamination presses, drilling, and thermal processing. Z-axis specimens are typically cut as small squares containing only dielectric between copper planes, free of vias, pads, and resin-starved areas. Specimens must be taken away from panel edges, tooling holes, and scoring lines to avoid edge effects, and all cut faces must be smooth, flat, and parallel to ensure uniform load transfer to the measurement probe.
Specimen Conditioning
Moisture strongly plasticizes epoxy-based resins and shifts both Tg and apparent CTE, so conditioning is mandatory. Common practice includes baking specimens at 105 ± 2 °C for 1 to 2 hours, cooling in a desiccator, and testing within a defined window thereafter. Where end-use evaluation of a moisture-conditioned board is required, specimens may instead be conditioned at 23 °C / 50% RH per IPC guidelines before testing. The conditioning route must always be reported together with the results.
Testing Items
A complete CTE characterization of PCB materials normally includes the following measured items, all of which can be obtained from a single controlled temperature ramp:
- In-plane CTE, X direction (alpha 1 and alpha 2), ppm/°C
- In-plane CTE, Y direction (alpha 1 and alpha 2), ppm/°C
- Z-axis (thickness) CTE, alpha 1 and alpha 2, ppm/°C
- Glass transition temperature (Tg) by TMA, °C, determined from the inflection of the expansion curve
- Total thickness expansion over a defined range, for example 50 °C to 260 °C, expressed in percent
- Time to delamination or expansion excursion behavior where required for qualification testing
Testing Methods
Thermomechanical Analysis (TMA)
TMA is the industry-standard method for PCB CTE measurement, specified in IPC-TM-650 2.4.24 (laminate), 2.4.24.1, 2.4.24.2, 2.4.24.3, and 2.4.24.4 (printed boards), as well as ASTM E831. A quartz probe or flat-tipped expansion probe rests on the specimen with a small static force (typically 0.05 N to 0.3 N), and an LVDT sensor records dimensional change as the furnace heats the specimen. Typical test conditions are a nitrogen purge of 20–50 mL/min, a heating rate of 5 °C/min (some specifications allow 10 °C/min), and a temperature range from ambient up to 260–290 °C or at least 40 °C beyond the expected Tg. The instrument is calibrated for temperature and expansion using reference materials such as aluminum and sapphire prior to testing.
Push-Rod Dilatometry (ASTM E228)
A push-rod dilatometer measures expansion of longer specimens, typically 25 mm to 50 mm in length, using a quartz or alumina rod transmitting displacement to a precision sensor. Dilatometry offers excellent precision for in-plane CTE of rigid laminates and thick samples at modest heating rates of 1–5 °C/min. Its longer specimen requirement limits its use for thin Z-axis measurements, where TMA is strongly preferred.
Optical and Interferometric Methods
Laser interferometry per ASTM E289 provides very high resolution (sub-micron) and is suitable for ultra-low-expansion materials such as ceramic substrates and low-CTE package cores. Optical extensometry and digital image correlation (DIC) techniques offer non-contact, full-field measurement across entire board panels, capturing local variations caused by copper distribution, resin-rich zones, and woven-fabric architecture. These methods are increasingly used in research and advanced reliability engineering.
Strain Gauge Method
Bonded high-temperature strain gauges can measure apparent in-plane expansion of actual circuit board sections, including the influence of patterned copper. This approach is useful when the effective CTE of a populated or patterned region must be known for warpage simulation, although gauge reinforcement and adhesive creep must be carefully corrected.
Data Calculation and Reporting
CTE is computed from the slope of the dimension-versus-temperature curve using linear regression over defined intervals: alpha 1 is normally taken over a range such as 50 °C to 150 °C (or another range below Tg specified by the customer), while alpha 2 is taken from a regression just above Tg. The first 20–30 °C of the ramp is usually excluded to avoid thermal equilibration artifacts. The Tg is identified at the intersection of the extrapolated alpha 1 and alpha 2 regression lines. A complete report should include specimen identification, axis tested, conditioning history, copper-etched condition, temperature program, heating rate, alpha 1, alpha 2, Tg, and total expansion percentage.
Applications
CTE test data are used throughout the PCB supply chain for material qualification, design, and reliability engineering:
- Copper-clad laminate qualification and incoming inspection against IPC-4101 slash sheet limits, including maximum Z-axis CTE and maximum percent expansion between 50 °C and 260 °C
- Verification of high-Tg, mid-Tg, and lead-free-compatible material families for multiple reflow excursions at 260 °C peak temperature
- Plated-through-hole and via reliability assessment, since Z-axis expansion drives copper barrel fatigue during thermal cycling
- Solder joint life prediction and coefficient-of-thermal-expansion matching between board, BGA packages, ceramic components, and heat sinks
- Warpage and stress simulation input for HDI boards, chip package substrates, and rigid-flex constructions
- Failure analysis, supplier comparison, and evaluation of process changes such as press-cycle modifications or resin chemistry updates
FAQ
What is the difference between alpha 1 and alpha 2 CTE?
Alpha 1 is the linear expansion coefficient below the glass transition temperature, where the resin matrix is rigid and expansion is comparatively low. Alpha 2 is the coefficient above Tg, where the softened resin expands much faster, typically three to five times the alpha 1 value in the Z direction. Both must be reported because board assemblies experience temperatures above Tg during reflow and rework.
Why is Z-axis CTE more critical than in-plane CTE?
In-plane expansion is constrained by the glass fabric and stays close to the CTE of copper (about 17 ppm/°C), minimizing stress on solder joints. Z-axis expansion is resin-dominated and much larger; during heating it stretches plated through-holes and via barrels, causing copper cracking, inner-layer separation, and delamination in multilayer boards.
Which test method should be chosen, TMA or dilatometry?
TMA per IPC-TM-650 2.4.24 and ASTM E831 is the preferred method for PCB laminates and finished boards because it accommodates the small, thin specimens typical of the industry and simultaneously yields Tg. Dilatometry per ASTM E228 is better suited to longer, thicker in-plane specimens where its superior length resolution can be exploited.
What temperature range and heating rate are typically used?
A common program ramps from room temperature to 260 °C or 290 °C at 5 °C/min under nitrogen, which covers the Tg of most FR-4 systems and the lead-free reflow regime. Deviations from the specified heating rate should be avoided because a faster ramp shifts the apparent Tg and alpha 2 values upward.
Should the copper foil be removed before testing?
For characterizing the dielectric material itself, copper is normally etched away so results reflect the resin-glass composite, enabling fair comparison between laminates. For evaluating the composite behavior of a copper-clad product or finished board, testing with copper retained is acceptable and often specified; whichever condition is used must be clearly stated in the report.
How does moisture affect CTE test results?
Absorbed moisture lowers the effective Tg and increases apparent expansion of epoxy-based systems, which can cause significant underestimation of thermal margin. Proper pre-test baking and desiccator cooling, combined with documentation of conditioning, are essential for reproducible and comparable results.
What are typical acceptance criteria for PCB CTE?
While limits vary by slash sheet and application, common qualification targets for lead-free FR-4 materials are an alpha 1 Z-axis CTE of 60 ppm/°C or less, an alpha 2 below approximately 300 ppm/°C, and total Z expansion from 50 °C to 260 °C not exceeding 3.0 to 4.0%. High-reliability and package substrate applications impose stricter requirements and often require low-CTE cores or filled systems.