4J44 Kovar Alloy Coefficient of Thermal Expansion Testing: Standards, Methods, and Technical Guidanc

4J44 is a low-cobalt iron-nickel-cobalt controlled expansion alloy widely used for matched glass-to-metal seals in hermetic electronic packages. Accurate determination of its coefficient of thermal expansion (CTE) is essential to guarantee seal integrity and long-term reliability. This article provides a professional overview of 4J44 expansion coefficient testing, covering alloy characteristics, sample preparation, testing items, measurement methods, applications, and frequently asked questions.

Technical Overview

4J44 is a Chinese-designation iron-nickel-cobalt sealing alloy belonging to the Kovar family of controlled expansion materials. It was developed as a low-cobalt alternative to the classical 4J29 Kovar alloy, reducing the cobalt content from approximately 17% to below 10% while retaining a mean coefficient of thermal expansion (CTE) compatible with a range of sealing glasses and ceramic substrates. The alloy is supplied as bar, rod, wire, strip, and foil, and is primarily specified for glass-to-metal seal applications in vacuum electronic devices, semiconductor packages, and precision instruments.

The defining quality characteristic of 4J44 is its mean linear coefficient of thermal expansion over specified temperature intervals. Because the alloy is used in matched seals, even small deviations in CTE from the specified window can generate thermal stresses during processing and service, leading to seal cracking, leakage, or premature device failure. Consequently, CTE testing is a mandatory acceptance test in most purchase specifications, and it must be performed under controlled conditions using calibrated dilatometric equipment.

The typical chemical composition of 4J44 is summarized below:

• Nickel (Ni): 34.2% to 35.2%
• Cobalt (Co): 8.5% to 9.5%
• Iron (Fe): balance
• Carbon (C): no more than 0.05%
• Manganese (Mn): no more than 0.50%
• Silicon (Si): no more than 0.30%
• Phosphorus (P): no more than 0.020%
• Sulfur (S): no more than 0.020%

The magnetic and expansion behavior of 4J44 is governed by its Curie transition, which occurs at a relatively low temperature of approximately 330 to 360 degrees Celsius. Below the Curie point, the alloy exhibits anomalously low expansion due to magnetostriction effects; above it, expansion increases toward normal metallic values. Standard specifications typically require the mean linear CTE to fall between approximately 4.3 and 5.5 multiplied by 10 to the power of minus 6 per degree Celsius over the 20 to 300 degree Celsius interval, with corresponding limits for the 20 to 400 and 20 to 500 degree Celsius intervals. The exact acceptance limits should always be confirmed against the governing product standard and purchase specification.

Testing Samples

Bar, Rod, and Forged Samples

For push-rod dilatometry, specimens are machined from bar or rod stock to a length of 25 to 50 millimeters, depending on the instrument, with a diameter of 4 to 6 millimeters or a rectangular cross section of equivalent stiffness. Both end faces must be ground flat, parallel to each other, and perpendicular to the specimen axis, typically within 0.01 millimeter, because end-face errors translate directly into expansion measurement error. Machining should remove the deformed surface layer with low-stress grinding or turning, and sharp edges should be lightly chamfered to protect the dilatometer contacts.

Cold-Rolled Strip and Foil

Strip and foil products are tested using specimens cut along and, where required, transverse to the rolling direction to evaluate expansion anisotropy. Depending on thickness, multiple layers may be stacked and clamped to obtain a measurable length, or the material may be evaluated by thermomechanical analysis, which accommodates short and thin specimens. Strip samples should be degreased and deburred, and sheared edges should be re-ground when necessary to eliminate residual shear strain introduced during cutting.

Wire and Drawn Tubing

Wire specimens are straightened with minimal plastic deformation, cut to the required gauge length, and mounted so that axial alignment is maintained throughout the test. Drawn tubing is tested in the as-drawn or stress-relieved condition as specified. Because straightening and cutting can introduce residual stresses that distort low-temperature expansion readings, careful specimen preparation and, where permitted, a stress-relief heat treatment are recommended before measurement.

Finished Components and Seal Assemblies

Finished parts such as lead frames, pins, and seal housings can be evaluated by cutting representative coupons from the component or by using small-sample techniques such as TMA. Testing of finished components is valuable for verifying that forming, plating, and heat-treatment operations have not altered the expansion behavior of the base metal. Sample location should be documented, particularly for parts that have undergone localized welding or brazing.

Specimen Conditioning

Before testing, specimens are commonly stress-relief annealed to remove residual machining stress, for example by holding at 600 to 850 degrees Celsius for 30 to 60 minutes in dry hydrogen, argon, or vacuum followed by slow cooling. The conditioning treatment must be compatible with the delivery condition of the material and should be agreed between the purchaser and the testing laboratory, since an inappropriate treatment can shift the measured CTE.

Testing Items

A comprehensive 4J44 expansion testing program typically includes the following items:

• Mean linear coefficient of thermal expansion over specified intervals, commonly 20 to 100, 20 to 200, 20 to 300, 20 to 400, and 20 to 500 degrees Celsius
• Instantaneous (differential) coefficient of thermal expansion as a function of temperature
• Full thermal expansion curve, expressed as fractional length change versus temperature
• Curie temperature determination from the inflection in the expansion curve
• Assessment of expansion hysteresis and repeatability under thermal cycling between ambient and the maximum service temperature
• Anisotropy evaluation for rolled strip by comparing longitudinal and transverse specimens
• Chemical composition verification, usually by inductively coupled plasma optical emission spectrometry, to support interpretation of expansion results
• Metallographic examination of grain size and inclusions where abnormal expansion behavior is observed

The reported result for acceptance purposes is almost always the mean linear CTE over the temperature interval stated in the specification, calculated from the fractional length change between the reference temperature and the upper interval temperature divided by the temperature span.

Testing Methods

Push-Rod Dilatometry

Push-rod dilatometry in accordance with ASTM E228 or the equivalent Chinese national method GB/T 4339 is the standard technique for 4J44 acceptance testing. The specimen is placed in a fused-silica or alumina holder with a low-force push rod transmitting length change to a displacement transducer, typically a linear variable differential transformer with resolution on the order of 10 to 100 nanometers. The furnace is ramped at a controlled rate, commonly 3 to 5 degrees Celsius per minute, or operated in a stepwise equilibrium mode, under flowing high-purity argon or helium or in moderate vacuum to prevent oxidation of the specimen above approximately 300 degrees Celsius. The apparatus is calibrated immediately before or after the test using certified reference materials such as fused silica, sapphire, tungsten, or copper, and a correction run is performed to compensate for the expansion of the holder and push rod. Typical measurement uncertainty for mean CTE is on the order of plus or minus 0.1 to 0.2 multiplied by 10 to the power of minus 6 per degree Celsius.

Thermomechanical Analysis

Thermomechanical analysis per ASTM E831 is well suited to small specimens such as strip, foil, and coupons cut from finished components. Specimens 2 to 10 millimeters in length are tested under a small static force, typically 0.05 to 0.2 newtons, at heating rates of 3 to 5 degrees Celsius per minute under inert purge gas. TMA offers excellent throughput and small-sample capability, although its accuracy is somewhat lower than that of a well-calibrated push-rod dilatometer, so it is often used for screening and process monitoring rather than final acceptance.

Laser Interferometry

Laser interferometric dilatometry per ASTM E289 measures length change through the interference fringes produced by optical-quality specimen faces, achieving resolution approaching 0.01 multiplied by 10 to the power of minus 6 per degree Celsius. This technique requires specimens with polished, parallel optical surfaces and a more demanding setup, and it is generally reserved for referee measurements, certification of reference materials, or characterization of very low expansion increments where push-rod uncertainty becomes significant.

Measurement Quality Control

Regardless of the technique, laboratories should verify instrument calibration with certified reference materials at regular intervals, record the complete temperature-expansion curve rather than only discrete points, and test duplicate specimens when results approach specification limits. Heating rate, atmosphere, specimen geometry, and conditioning treatment must be documented in the test report, together with the calculation interval and the reference temperature. Where a dispute arises, testing should be repeated on the same specimen and on a second specimen from the same lot using a stepwise equilibrium method to eliminate heating-rate effects.

Applications

The combination of low and highly repeatable expansion, good glass wetting behavior, and reduced cobalt content makes 4J44 a cost-effective sealing alloy across the electronics industry. Typical applications include:

• Hermetic glass-to-metal seals in vacuum electronic devices, connectors, and feedthroughs
• Lead frames, pins, and bases for transistors, diodes, and integrated circuit packages
• Reed switch blades and contact support structures requiring stable dimensions
• Matched seals to selected soft glasses and metallized ceramic bodies
• Precision instrument components where low, uniform thermal expansion must be guaranteed
• Substitution for higher-cobalt 4J29 Kovar in cost-sensitive sealed packages with compatible glasses

In every application, verified expansion data provide the basis for matched-seal design, thermal stress analysis, and reliability qualification, which is why CTE certification accompanies material delivery in nearly all critical procurements.

FAQ

What is the typical acceptance range for the mean CTE of 4J44?

Most specifications require the mean linear CTE to be between approximately 4.3 and 5.5 multiplied by 10 to the power of minus 6 per degree Celsius over the 20 to 300 or 20 to 400 degree Celsius interval. The exact limits depend on the governing product standard and the glass being matched, so the purchase specification should always be consulted.

Which test method should be used for acceptance testing?

Push-rod dilatometry per ASTM E228 or GB/T 4339 is the preferred acceptance method because of its established precision for metals. TMA per ASTM E831 is acceptable for thin and small samples, while laser interferometry per ASTM E289 is used for referee and high-precision work.

What sample size is required for a push-rod dilatometer?

A typical specimen is 25 to 50 millimeters long with a 4 to 6 millimeter diameter or equivalent rectangular section, with end faces ground flat and parallel within about 0.01 millimeter. Exact dimensions depend on the instrument configuration and should be coordinated with the laboratory.

Why does 4J44 contain less cobalt than standard Kovar?

The reduced cobalt content, typically 8.5% to 9.5% versus about 17% in 4J29, lowers raw material cost and conserves a strategically important element while preserving expansion behavior compatible with many sealing glasses. The trade-off is a slightly lower Curie temperature and somewhat narrower glass-matching flexibility.

Which temperature intervals are reported?

Reports normally include mean CTE values for 20 to 100, 20 to 200, 20 to 300, 20 to 400, and, where required, 20 to 500 degrees Celsius, together with the complete expansion curve and the Curie transition if identifiable.

What causes results outside the specification window?

Common causes include chemical composition outside limits, improper final annealing, residual stress from machining or forming, excessive heating rates during measurement, uncorrected apparatus expansion, and specimen misalignment. A systematic review of specimen preparation, instrument calibration, and material certification is required to identify the root cause before retest or rejection.