Kovar Alloy Expansion Performance Testing: Standards, Methods, and Applications

Kovar alloy expansion performance testing evaluates the coefficient of thermal expansion (CTE) and expansion behavior of iron-nickel-cobalt sealing alloys to ensure their compatibility with borosilicate glass and ceramic components in hermetic packaging. Using push-rod dilatometry and thermomechanical analysis in accordance with ASTM E228, ASTM E831, ASTM F15, and GB/T standards, accredited third-party laboratories verify that Kovar materials meet the tight expansion tolerances required for glass-to-metal seals in electronics, aerospace, and medical devices. This article explains the samples, test items, methods, and acceptance criteria involved in a complete Kovar expansion performance evaluation.

Technical Overview

Kovar is a vacuum-melted, controlled-expansion iron-nickel-cobalt alloy with a nominal composition of approximately 29% nickel, 17% cobalt, and the balance iron, standardized as ASTM F15 (UNS K94610). Its defining characteristic is a coefficient of thermal expansion (CTE) closely matched to that of borosilicate glass and several alumina ceramics over the temperature range of roughly 20 to 450 degrees Celsius, which makes it the preferred material for hermetic glass-to-metal seals.

Expansion performance testing measures how the linear dimensions of Kovar change with temperature. Because the reliability of a glass-to-metal seal depends on the expansion match between metal and glass across processing and service temperatures, even small deviations in CTE can cause seal cracking, leakage, or residual stress. The expansion behavior of Kovar is non-linear: below the Curie temperature (approximately 430 degrees Celsius), the alloy exhibits a low and stable CTE, while above this transition the CTE rises sharply. A complete test program therefore evaluates both the average CTE over defined intervals and the shape of the full expansion curve.

• Alloy family: Fe-Ni-Co sealing alloy, ASTM F15 / UNS K94610
• Key property: mean CTE typically near 5.0 x 10⁻⁶/°C between 20 and 400 degrees Celsius
• Primary function: expansion matching with borosilicate glass and alumina ceramics
• Testing basis: ASTM E228, ASTM E831, ASTM F15, ISO 11359-2, GB/T 4339

Testing Samples

Rod and Bar Specimens

Cylindrical specimens machined from hot-rolled or forged bar are the standard configuration for push-rod dilatometry. A typical specimen measures 50 mm in length with a diameter of 5 to 6 mm. Both end faces must be flat, parallel, and perpendicular to the specimen axis, because any geometrical error translates directly into displacement measurement error.

Sheet and Strip Specimens

Flat strip specimens are prepared from cold-rolled Kovar sheet, with the specimen axis aligned along the rolling direction unless cross-direction behavior is specifically required. Strip specimens must be straightened and, where cold-work-induced residual stress is suspected, stress-relieved before testing to prevent non-representative expansion readings.

Forged and Machined Components

Finished or semi-finished parts such as transistor headers, package bases, eyelets, lead frames, and feedthrough pins can be tested directly when a representative geometry can be clamped in the dilatometer. Component testing verifies that processing steps such as deep drawing, stamping, and hydrogen annealing have not altered the expansion characteristics of the base material.

Wire Specimens

Wire samples used for glass-sealed leads are tested as straightened sections with suitable end fixtures. Because drawing deformation can affect expansion behavior, wire specimens are frequently tested in both the as-drawn and annealed conditions.

Testing Items

A comprehensive expansion performance evaluation typically includes the following measurable items:

• Mean (average) coefficient of thermal expansion over specified intervals, most commonly 20–100, 20–300, 20–400, and 20–450 degrees Celsius
• Instantaneous (tangent) CTE as a function of temperature, derived from the slope of the expansion curve
• Complete expansion curve, expressed as fractional length change versus temperature
• Inflection or transition temperature, marking the departure from the low-expansion plateau associated with magnetic transformation
• Expansion hysteresis and dimensional recovery after repeated thermal cycling between room temperature and the maximum service temperature
• Comparison of specimen expansion behavior against the expansion curve of the mating sealing glass to estimate residual seal stress
• Verification of CTE homogeneity across different heats, product forms, or production lots

Testing Methods

Push-Rod Dilatometry (ASTM E228)

The push-rod dilatometer is the reference method for Kovar alloys. The specimen is placed in a fused-silica or alumina holder, and a spring-loaded push-rod transmits length changes to a displacement sensor, typically a linear variable differential transformer (LVDT) with resolution on the order of 0.1 micrometer or better. The specimen is heated at a controlled rate, generally not exceeding 5 degrees Celsius per minute, under a flowing inert atmosphere such as argon or nitrogen, or in vacuum, to prevent oxidation of the test piece. Temperature is measured with a calibrated thermocouple positioned adjacent to the specimen midpoint.

Thermomechanical Analysis (ASTM E831 / ISO 11359-2)

Thermomechanical analysis uses a smaller specimen and a lower-mass furnace, making it suitable for rapid screening and for sheet or component geometries. Heating rates of 5 to 10 degrees Celsius per minute are typical. TMA results are reliable for comparative purposes and for CTE measurement, provided the apparatus is calibrated with a certified reference material over the same temperature range.

Calibration and Reference Materials

Before testing, the displacement and temperature signals are calibrated using certified reference materials with known expansion behavior, such as sapphire (corundum), fused silica, or tungsten. System expansion of the holder and push-rod is measured with a zero-expansion correction run and subtracted from the specimen signal. This correction is critical for Kovar because the alloy CTE is of the same order of magnitude as that of the fused-silica parts of many dilatometers.

Data Reduction and Calculation

The mean CTE between two temperatures is calculated from the fractional length change divided by the temperature interval, using the specimen length measured at a reference temperature of 23 ± 5 degrees Celsius. The instantaneous CTE is obtained from the temperature derivative of the expansion curve. Results are reported as a table of mean CTE values for each specified interval together with the full expansion curve, heating rate, atmosphere, calibration reference, and specimen orientation.

Acceptance Criteria

Acceptance limits are defined by the material specification and purchase agreement. For ASTM F15 alloy, the mean CTE over the key sealing interval of approximately 20 to 300 or 20 to 400 degrees Celsius is commonly required to fall within a narrow band around 5 x 10⁻⁶/°C, with the expansion curve required to remain below a defined upper bound up to the glass sealing temperature. Specimens that show excessive hysteresis, non-reproducible curves, or premature inflection indicate improper composition, insufficient annealing, or excessive cold work, and the lot is normally rejected for sealing applications.

Applications

Kovar expansion performance testing directly supports quality control and qualification in every industry that relies on hermetic glass-to-metal sealing:

• Hermetic packages for integrated circuits, hybrid circuits, and microelectromechanical systems, where the Kovar lid and body must track glass insulator expansion through sealing and reflow cycles
• Power vacuum tubes, thyristors, and transistor headers requiring reliable glass-to-metal feedthroughs
• Aerospace and defense connectors, sensor housings, and optical windows exposed to wide temperature swings
• Automotive electronics, including pressure sensors and ignition components subjected to thermal cycling
• Medical implantable devices and pacemaker headers, where hermeticity over decades of service is essential
• Reed switches, lamp and LED substrates, and fiber-optic feedthroughs that depend on matched expansion joints

In each case, verified expansion data allow seal designers to calculate residual stress and select a glass grade whose expansion curve remains appropriately matched to the Kovar expansion curve between room temperature and the glass transition point of the seal glass.

FAQ

Why must Kovar be tested rather than accepted on composition alone?

Nominal chemical composition does not guarantee expansion behavior. Variations in melting practice, cold work, annealing history, and trace elements can shift the CTE and the inflection temperature enough to compromise a seal. Direct dilatometric measurement on the delivered product form is the only reliable confirmation.

What temperature range should be specified for testing?

Testing should cover at least the range from room temperature to the glass sealing or service temperature. For most borosilicate sealing applications, mean CTE values at 20–300 and 20–400 degrees Celsius plus the full curve to approximately 500 degrees Celsius are specified so that the inflection behavior is captured.

How accurate is a push-rod dilatometer measurement on Kovar?

With proper calibration using certified reference materials, a fused-silica push-rod dilatometer can achieve mean CTE uncertainty on the order of a few percent, which is adequate for the acceptance bands applied to sealing alloys. Instrument correction runs and controlled heating rates are the main factors governing accuracy.

Do finished parts need retesting after hydrogen annealing or deep drawing?

Yes. Processing can measurably alter expansion behavior through relief of residual stress and microstructural change. Retesting of finished or annealed components by an accredited third-party laboratory confirms that the material still matches the glass expansion curve expected by the seal design.

How many specimens should be tested per lot?

Common practice is to test a minimum of two or three specimens drawn from different portions of the lot or coil, so that both measurement repeatability and material homogeneity are demonstrated. Additional specimens are recommended whenever a lot fails marginally or when cross-direction behavior in strip is of concern.