Kovar alloy (ASTM F15 / UNS K94610) is an iron-nickel-cobalt sealing alloy whose precisely controlled coefficient of thermal expansion (CTE) makes it the material of choice for glass-to-metal and ceramic-to-metal hermetic seals in electronics. Evaluating the thermal expansion performance of Kovar requires standardized dilatometric testing, careful sample preparation, and comparison against alloy specification limits and target glass expansion curves. This article provides a technical overview of Kovar alloy, describes typical test samples and testing items, reviews the principal measurement methods and standards such as ASTM E228 and GB/T 4339, and answers frequently asked questions about acceptance criteria and applications.
Technical Overview
Kovar is a low-expansion iron-nickel-cobalt alloy containing nominally 29% nickel, 17% cobalt, and balance iron, specified in ASTM F15 and known in the Chinese alloy system as 4J29. Its defining characteristic is a mean coefficient of thermal expansion that closely matches that of borosilicate and sealing glasses over the critical temperature range from room temperature to approximately 450 °C, which enables the fabrication of stress-free, leak-tight glass-to-metal seals.
Between 20 °C and 300 °C, Kovar typically exhibits an average CTE in the range of 4.6 × 10⁻⁶ to 5.5 × 10⁻⁶ per °C, which is comparable to common sealing glasses used in electronic packages. Below its Curie temperature of approximately 430 °C, the ordered magnetic structure suppresses normal lattice expansion through the magnetovolume (Invar-type) effect. Above this transformation temperature, the alloy loses its low-expansion behavior and the CTE rises toward values typical of ordinary steels. For this reason, the expansion curve must be characterized not only at room-temperature service conditions but also across the entire sealing and cooling cycle, since residual seal stress is governed by the integrated mismatch between metal and glass.
Thermal expansion performance evaluation therefore serves several purposes: verification of chemical and metallurgical conformity of the alloy, qualification of heat treatment (including hydrogen annealing and stabilization), assessment of compatibility with a specific sealing glass, and quality control for incoming material in hermetic package production. Accredited third-party laboratories routinely perform these evaluations against ASTM, ISO, EN, and GB standards.
Testing Samples
Raw Material Forms
Thermal expansion specimens are prepared from the delivered product forms, including cold-rolled strip and sheet, hot-rolled plate, bar and rod, wire, tube, and deep-drawn foil. Strip and sheet are the most common forms for lead frames and package bodies, while rod and bar are typical for machined feedthrough pins. Specimens are cut with their length parallel to the rolling direction unless transverse anisotropy is specifically being evaluated.
Finished and Machined Components
In addition to raw material, expansion performance can be verified on machined parts such as lead frames, eyelets, lids, transistor caps, and glass-to-metal seal components after their final annealing and degassing treatments. Testing components in their delivered metallurgical condition is important because deep drawing, machining, and heat treatment can alter the expansion behavior relative to the mill-annealed condition.
Sample Preparation Requirements
For push-rod dilatometry, specimens are typically 25 mm to 50 mm in length with a cross-section of 3 mm to 10 mm, flat and parallel end faces, and a clean, oxide-free surface. The following preparation practices are essential for valid results:
- Machining must be performed gently to avoid introducing residual stress, followed by stress-relief or stabilization annealing where specified.
- Hydrogen-annealed material must be degassed to remove absorbed hydrogen, which can distort expansion readings.
- Specimen length is measured at room temperature to an accuracy of at least 0.01 mm, since the length value directly enters the CTE calculation.
- Identification of material heat, condition, and orientation should accompany each specimen to ensure traceability.
Testing Items
A comprehensive thermal expansion evaluation of Kovar alloy generally includes the following measured and derived quantities:
- Average (mean) coefficient of thermal expansion over defined intervals, most commonly 20–100 °C, 20–200 °C, 20–300 °C, 20–400 °C, 20–450 °C, and 20–500 °C, with 20–300 °C and 20–450 °C serving as the primary acceptance intervals.
- Instantaneous (differential) coefficient of thermal expansion as a continuous function of temperature, revealing the transition at the Curie point where the low-expansion regime ends.
- Full thermal expansion curve (fractional length change versus temperature) for direct graphical comparison with the expansion curve of the mating glass.
- Curie (magnetic transformation) temperature determination, which marks the upper limit of the useful low-expansion range.
- Expansion hysteresis and repeatability under multiple heating and cooling cycles, which indicates metallurgical stability of the alloy.
- Seal compatibility assessment, evaluating the integrated expansion mismatch between the Kovar specimen and a reference sealing glass through the glass transition and annealing points.
Testing Methods
The principal methods for measuring the linear thermal expansion of Kovar alloy are push-rod dilatometry, thermomechanical analysis (TMA), and, for the highest precision, interferometric dilatometry. Method selection depends on the required temperature range, specimen geometry, and target measurement uncertainty.
Push-Rod Dilatometry (ASTM E228 / GB/T 4339)
A push-rod dilatometer measures the change in specimen length with temperature using a fused silica or alumina rod transmitting displacement to a precision transducer. Testing is conducted from room temperature to 600 °C or higher at a controlled heating rate, typically 3 °C/min to 5 °C/min, in an inert atmosphere such as argon or nitrogen, or in vacuum, to prevent oxidation. The instrument is calibrated immediately before testing using certified reference materials such as platinum, tungsten, or certified low-expansion standards, and a baseline correction is performed with the holder alone. The mean CTE is calculated as the specimen length change divided by the initial length and the temperature interval. Well-maintained systems achieve CTE uncertainties on the order of ±0.1 × 10⁻⁶ to ±0.3 × 10⁻⁶ per °C.
Thermomechanical Analysis (ASTM E831 / ISO 11359 Series)
TMA uses a smaller specimen, typically a few millimeters in length, contacted by a low-force probe, and is well suited to finished components, thin sections, and localized measurements on package details. Heating rates of 5 °C/min are common, and the method provides rapid screening of the expansion curve up to approximately 600 °C, with slightly higher uncertainty than classical dilatometry due to the shorter gauge length.
Interferometric Dilatometry (ASTM E289)
Independent of the method used, the reported result should include the temperature intervals, heating rate, atmosphere, specimen orientation, heat-treatment condition, and estimated measurement uncertainty. Comparison against the acceptance limits of the applicable material specification, such as ASTM F15, completes the evaluation.
Applications
The thermal expansion performance of Kovar directly determines the reliability of components in which a metal must be joined to glass or ceramic and survive thousands of thermal cycles without loss of hermeticity. Key application areas include:
- Hermetic packages for microelectronics, hybrid circuits, MEMS, and optoelectronic devices, including flat packs, transistor-outline (TO) cans, and dual-in-line packages.
- Glass-to-metal seals in connectors, feedthroughs, battery headers, and pressure transducers for automotive, aerospace, oil and gas, and medical implantable devices.
- Lead frames and lead wires for power semiconductors, LEDs, relays, and vacuum devices, where matched expansion prevents glass cracking during sealing and soldering.
- Instrument components such as optical housings, waveguide windows, and sensor envelopes requiring dimensional stability across wide temperature ranges.
In each of these applications, an expansion curve matching the sealing glass within the specified tolerance minimizes residual stress at the seal interface, suppresses crack initiation, and ensures long-term leak tightness under thermal cycling required by qualification standards.
FAQ
What is the typical acceptance criterion for Kovar thermal expansion?
Most specifications for Kovar (ASTM F15 and equivalent national standards) require a mean CTE of approximately 4.6 × 10⁻⁶ to 5.5 × 10⁻⁶ per °C over 20–300 °C, with a slightly higher but still controlled value over 20–450 °C. The exact limits depend on the specification level and the mating glass, so the purchase order or drawing should always state the applicable intervals and limits.
Why are the 20–300 °C and 20–450 °C intervals so important?
These intervals bracket the service temperature range of hermetic assemblies and the region in which the glass is rigid and residual stresses are locked in during cooling from the sealing operation. Matching the metal and glass expansions in this region controls the final seal stress; a mismatch here is a leading cause of glass cracking and leak failure.
Does the Curie temperature affect the test result?
Yes. Above the Curie temperature of roughly 430 °C, the Invar-type magnetostrictive contribution to low expansion disappears and the CTE rises sharply. The measured mean CTE over intervals extending beyond this point, such as 20–500 °C or 20–600 °C, will therefore be noticeably higher, and comparisons between laboratories are only valid when identical intervals are used.
Can finished hermetic packages be tested instead of raw material?
Yes. Thermomechanical analysis permits small specimens cut from finished parts or measurements on local features such as lids and pins. Testing components in the final metallurgical condition is recommended for qualification, because forming and hydrogen annealing can shift the expansion curve relative to mill-annealed strip.
What factors most influence measurement accuracy?
The dominant factors are specimen geometry and end-face parallelism, residual stress from machining, completeness of the calibration and baseline correction, control of heating rate and thermal gradients, and the test atmosphere. Using an accredited third-party laboratory with calibrated dilatometers and certified reference materials ensures defensible, comparable results.