Kovar alloy (ASTM F15, equivalent to Chinese grade 4J29) is a controlled-expansion iron-nickel-cobalt alloy whose linear thermal expansion behavior must be precisely verified to guarantee reliable glass-to-metal and ceramic-to-metal hermetic seals. Linear expansion testing measures the mean and instantaneous coefficients of thermal expansion (CTE) across critical temperature intervals using push-rod dilatometry, thermomechanical analysis, and interferometric methods in accordance with ASTM and GB standards. This article explains the testing principles, specimen requirements, procedures, acceptance criteria, and typical applications of Kovar alloy CTE inspection.
Technical Overview
Kovar is a vacuum-melted, controlled-expansion alloy containing approximately 29% nickel, 17% cobalt, and a balance of iron, standardized as ASTM F15 and known in Chinese material systems as grade 4J29. Its coefficient of linear thermal expansion (CTE) is engineered to closely match that of borosilicate glass and alumina ceramics over the range from room temperature to approximately 400 °C, which makes it the material of choice for hermetic sealing in electronic, optical, and vacuum devices.
The low expansion behavior of Kovar arises from the magnetovolume (Invar-type) effect of its ferromagnetic matrix. Below the Curie point, at roughly 430 °C, the alloy expands at approximately 4.6 to 5.2 × 10⁻⁶/°C (mean value from 20 °C to 300 °C). Above the Curie point this mechanism disappears and the CTE rises rapidly toward 9 to 11 × 10⁻⁶/°C or higher. Because sealing stresses, crack initiation, and long-term hermeticity failures are directly governed by the mismatch between the alloy and the glass or ceramic, accurate verification of the linear expansion rate is a mandatory incoming-material and outgoing-quality inspection step.
Linear expansion testing therefore determines whether a heat, strip, wire, or finished component lot meets the specified CTE window before it enters sealing production. Testing is typically performed by accredited third-party laboratories or in-house metallurgical laboratories using calibrated dilatometers traceable to certified reference materials.
Testing Samples
Bar, Rod, and Billet Specimens
For push-rod dilatometry, a representative specimen is cut from the lot, typically 25 mm to 50 mm in length with a diameter or square section of 4 mm to 6 mm. The two ends must be flat, parallel, and perpendicular to the longitudinal axis, because end-face errors translate directly into length-change measurement error. Specimens are taken in both the longitudinal and transverse directions when anisotropy of wrought material must be evaluated.
Strip, Foil, and Flat Products
Thin Kovar strip used for lead frames and stampings can be tested in dilatometers with flattened specimen holders. When the strip is too thin for a rod configuration, specimens are stacked, edge-prepared, or tested by thermomechanical analysis (TMA), which accepts short specimens from approximately 2 mm to 10 mm in length. Specimens must be flat, free of burrs, and cut cleanly to avoid stressing effects that distort the dilatometric curve.
Wire and Finished Components
Kovar wire for glass-seal pins, as well as finished parts such as transistor-outline headers, feedthroughs, and packages, can be evaluated using straightened wire segments or sectioned features prepared from the component. Sampling follows the applicable lot-definition rules of the purchase specification so that the tested specimen statistically represents the delivered material.
Specimen Conditioning and Preparation
- Perform the stabilizing or final hydrogen anneal required by ASTM F15 or the purchase specification before testing, so that residual stresses from machining do not superimpose on true thermal expansion.
- Deburr, degrease, and dry the specimen; verify flatness and parallelism of the measurement faces.
- Measure the reference length at 23 °C (or 20 °C) with a calibrated micrometer or length-measuring instrument to an accuracy consistent with the dilatometer resolution.
- Identify the material heat number, direction of sampling, and heat-treatment condition on the test report for full traceability.
Testing Items
The scope of Kovar linear expansion inspection typically includes the following measured and derived quantities:
- Mean (average) coefficient of linear thermal expansion over specified intervals, most commonly 20–100 °C, 20–200 °C, 20–300 °C, 20–400 °C, 20–450 °C, and 20–500 °C, reported in units of 10⁻⁶/°C.
- Instantaneous (differential) CTE as a function of temperature, which reveals the steep expansion increase at the ferromagnetic transition.
- Continuous dilatometric expansion curve (ΔL/L₀ versus temperature) for engineering design and finite-element modeling inputs.
- Expansion percentage over the full test range, used for dimensional compensation in sealing tooling design.
- Transition-temperature assessment near the Curie region, confirming that the low-expansion plateau extends through the sealing temperature window.
- Repeatability and hysteresis checks between heating and cooling cycles, indicating microstructural stability and complete stress relief.
- Anisotropy comparison between rolling-direction and transverse-direction specimens for wrought products.
Testing Methods
Push-Rod Dilatometry (ASTM E228 / GB/T 4339)
The horizontal or vertical push-rod dilatometer is the primary method for metallic controlled-expansion alloys. The specimen, pressed by a low-force rod against an inert holder, is heated at a controlled rate of typically 3 °C/min to 5 °C/min in an inert or slightly reducing atmosphere, while a calibrated displacement sensor records length change. The instrument is calibrated immediately before or after the test run using certified reference materials such as sapphire or fused silica, and correction for the expansion of the sample-holder system is applied. Typical measurement capability for mean CTE is on the order of ±0.1 to 0.2 × 10⁻⁶/°C, adequate for resolving specification windows of 0.3 × 10⁻⁶/°C or wider.
Thermomechanical Analysis (ASTM E831)
TMA employs a small specimen, usually 2 mm to 10 mm long, heated under a programmable furnace with a low contact load. It is preferred for thin strip, foil, coated sections, and limited-quantity samples where rod specimens cannot be prepared. Because of the shorter gauge length, displacement resolution requirements are more demanding, and the method is best used for comparative screening and trend verification of Kovar heats.
Optical and Laser Interferometric Dilatometry (ASTM E289)
Interferometric methods measure length change through fringe displacement with nanometer-level resolution, providing the lowest uncertainty for mean CTE, often better than ±0.05 × 10⁻⁶/°C. This absolute technique is applied when certifying reference materials, resolving disputes between supplier and purchaser, or characterizing premium lots for high-reliability aerospace and photonics programs.
Test Conditions and Quality Controls
- Heating rate, atmosphere (argon, nitrogen, helium, or vacuum), and soak schedule are fixed in the test plan and reported, since rate-dependent effects can shift apparent CTE values.
- A baseline run with a certified reference specimen is performed to validate system calibration for each test session.
- At least one duplicate specimen per lot is recommended to confirm reproducibility within the stated uncertainty.
- Results are compared against the acceptance window of the governing specification, for example a mean CTE of approximately 4.6 to 5.2 × 10⁻⁶/°C between 20 °C and 300 °C for typical Kovar and 4J29 purchase specifications, with exact limits defined by the customer drawing or ASTM F15 class.
Applications
Verified linear expansion behavior of Kovar is essential wherever the alloy must track the expansion of glass or ceramic through processing and service thermal cycles, including:
- Hermetic packages and lids for hybrid microelectronics, RF devices, and optoelectronic modules.
- Transistor-outline (TO) headers, lead frames, and glass-sealed lead arrays in discrete semiconductor devices.
- Hermetic feedthroughs and connectors for sensors, batteries, and pyrotechnic initiators in automotive systems.
- Vacuum tube, traveling-wave tube, and high-power microwave component envelopes and internal supports.
- Lithium and thermal battery glass-to-metal seals requiring long-life hermeticity.
- MEMS and photonics packaging, fiber-optic feedthroughs, and infrared window mounts in aerospace and defense hardware.
- Matched-seal brazing and ceramic-to-metal adapter structures where CTE gradients must be controlled to limit residual stress.
FAQ
Why is the 20 °C to 300 °C mean CTE the key acceptance parameter for Kovar?
Most glass-to-metal sealing operations and subsequent service excursions occur below 400 °C, within the low-expansion plateau of the alloy. The mean CTE from 20 °C to 300 °C therefore captures the property that controls sealing stress and hermeticity, and it is the interval cited in ASTM F15 and in typical 4J29 purchase specifications.
What specimen size is required, and how much material should be submitted?
For push-rod dilatometry, a rod of approximately 50 mm length and 6 mm diameter gives the best accuracy; a minimum of 25 mm length is generally acceptable. For TMA, a 5 mm to 10 mm segment of strip or wire is sufficient. Laboratories usually request two to three specimens plus retain material so that duplicate and referee testing can be performed.
How accurate are the results, and what is the typical turnaround?
Push-rod dilatometry typically achieves an uncertainty of about ±0.1 to 0.2 × 10⁻⁶/°C on the mean CTE, while interferometric methods can reach better than ±0.05 × 10⁻⁶/°C. A single programmed run to 500 °C takes two to three hours; with calibration runs, duplicate specimens, and reporting, results are normally issued within three to five working days.
Can finished parts be tested instead of raw material samples?
Yes. Sections can be cut from headers, packages, or pins, provided the specimen ends are prepared flat and parallel and the part history (especially any final anneal) is known. Testing a finished part verifies the combined effect of composition, processing, and heat treatment on the delivered component.
What causes an out-of-specification result, and how is it handled?
Common causes include compositional drift in nickel or cobalt, improper final annealing, residual cold-work stress, or test artifacts such as poor end-face preparation and excessive heating rate. A failed lot is first retested on a duplicate specimen after verifying instrument calibration with a reference material; if the deviation is confirmed, the material is rejected for sealing applications and the supplier heat record is reviewed for corrective action.