Machining Invar CMM Calibration Bars and Reference Gauges

By Invar Machining editorial team· Published September 26, 2026 · How we write these guides

A CMM calibration bar, step gauge or ball bar is not a part the machine measures. It is the artifact the machine is checked against, so the relationship is reversed from ordinary work: the tolerance, finish and long-term stability of the artifact set the ceiling on what the verification can actually prove about the CMM. Invar 36 and Super Invar show up in this role because their coefficient of thermal expansion is small enough that a length change from an ordinary few degrees of room temperature drift stays well under the artifact’s own machined tolerance, something a steel or aluminum bar cannot promise at the same length. This guide covers machining Invar into that kind of artifact. It is a different job from inspecting an ordinary Invar part on a CMM, which is covered in reading an Invar CMM inspection report: that guide is about trusting a measurement of a machined part, this one is about building the reference the measurement gets checked against. The two jobs share an alloy and sometimes a room, and little else.

Why calibration labs specify Invar for reference artifacts

A CMM’s volumetric accuracy is verified against known lengths, most often with a step gauge, a ball bar or a set of length bars placed at different positions and orientations in the working volume. If the reference artifact itself grows or shrinks with the room, the verification is measuring the room as much as the machine. Invar 36 has a typical mean CTE of about 1.2 to 1.6 ppm/°C near room temperature and Super Invar runs lower still, about 0.3 to 0.6 ppm/°C over roughly 20 to 90 °C, figures collected on the Invar specs reference alongside Kovar and Alloy 42 for comparison. Steel runs about 10 to 12 ppm/°C and aluminum about 23 ppm/°C, both an order of magnitude higher. For a fixed length and a fixed uncertainty budget, that difference is the entire reason a calibration lab specifies Invar over steel for anything longer than a short gauge block.

The same property that makes Invar useful for optical benches and lithography stages, described in long-term dimensional stability, makes it useful here for a different reason: a calibration artifact is measured over and over across years, sometimes by different labs, and its expansion needs to be predictable and small every time it comes out of storage.

Common calibration artifact geometries

Most Invar reference artifacts a machine shop sees fall into a handful of shapes:

• Step gauges, a single bar with a series of parallel faces or shoulders machined at known nominal spacings, used to check a CMM’s linear accuracy along one axis at several positions.
• Ball bars, a tube or solid bar with precision spheres mounted at each end or at intermediate positions, used to check linear and volumetric accuracy by comparing measured center-to-center distances against a calibrated length.
• Length bars and gauge blocks, simple prismatic or cylindrical references with two flat, parallel measuring faces, used both stand-alone and as calibration standards for other gauges.
• Custom check fixtures, artifacts built to a specific machine’s working volume rather than a catalog size, common in aerospace and semiconductor metrology labs that verify a CMM against features similar to their actual production parts.

None of these are structural parts. Every feature on them exists to be measured, which changes how a shop machines, fixtures and finishes them compared with an Invar bracket or mount.

Machining and finish targets for a calibration artifact

The table below lists typical targets a shop works to on a precision Invar calibration artifact. These are typical figures for planning a quote, not a specification; the drawing or the calibration lab’s requirement governs the actual number.

Feature	Typical target	Notes
Length tolerance, 300 to 1000 mm bar	±0.002 to ±0.01 mm	Tighter bands need extra stress relief cycles and lapping
Flatness of a measuring face	0.001 to 0.003 mm	Checked after the full stabilization cycle, not right off the machine
Parallelism between step faces	0.002 to 0.005 mm over the step spacing	Depends on step count and spacing
Surface finish, measuring faces	Ra 0.1 to 0.4 micrometers	Ground or lapped, not left as-milled
Ball diameter and sphericity (ball bar)	Sub-micrometer, set by the ball supplier	Balls are usually bought finished, not machined in-house
Concentricity, ball mount to bar axis	0.002 to 0.005 mm	Affects the accuracy of the center-to-center length

A shop quoting this kind of work is quoting grinding and lapping time as much as milling time. The Invar machining tolerances guide covers how tolerance bands trade off against cost for ordinary Invar parts; a calibration artifact sits at the tight end of that curve on purpose.

The stabilization sequence for a reference artifact

The machining sequence is the same family of steps used on any precision Invar part, but a reference artifact gets it more thoroughly because the part is reused for years rather than fitted once and left alone:

1. Rough machining, leaving generous stock on every measuring face.
2. Stress relief, typically 315 to 345 °C for one to two hours with a slow cool, removing the bulk of the machining stress before any dimension is finalized.
3. Semi-finish machining, bringing the artifact close to its final size and re-exposing any stress that the roughing left below the surface.
4. A second stress relief or a full stabilization cycle, especially for artifacts intended to hold their calibration for several years, following the same anneal, quench, temper and low-temperature soak sequence described in Invar heat treatment to ASTM F1684.
5. Finish machining and grinding or lapping, taking light final cuts so almost no new cold work is added.
6. An extended soak at the measurement reference temperature, usually 20 °C, before any dimension is checked.
7. First article inspection, then delivery with the mill certificate and a record of the heat treatment cycles applied.

Skipping the second stabilization step is the most common shortcut that shows up later as a calibration bar that reads slightly different lengths a year after it was first certified, which is exactly the failure mode the alloy was chosen to avoid.

Worked example: temperature uncertainty on a 500 mm step gauge

Take a hypothetical 500 mm Invar 36 step gauge with steps at 100 mm intervals, going to a calibration lab that controls its room to 20 °C ± 1 °C. The question is how much of the gauge’s own machined tolerance the temperature uncertainty consumes, compared with the same gauge made from tool steel.

Length change = nominal length × CTE × temperature deviation.

Material	Typical CTE, ppm/°C	Change over 500 mm for a 1 °C deviation	Change over 500 mm for a 3 °C deviation
Invar 36	1.4	0.0007 mm	0.0021 mm
Super Invar	0.5	0.00025 mm	0.00075 mm
Tool steel	11	0.0055 mm	0.0165 mm
Aluminum	23	0.0115 mm	0.0345 mm

Values are typical and the example is illustrative. If the step gauge is toleranced to ±0.005 mm over 500 mm, a 3 °C swing in room temperature consumes over 40 percent of that budget in tool steel and less than half a percent of it in Invar 36. This is the calculation a metrology lab is actually running when it specifies Invar instead of steel for anything past a short gauge block, and it is why the room’s own temperature control still matters even with Invar: the alloy narrows the problem, it does not remove it.

Who actually certifies the artifact

This is the point worth stating plainly because it gets confused often: a machine shop machines, stress relieves and delivers the artifact. It does not calibrate it, and it should never claim the finished bar is NIST-traceable or carries an ISO/IEC 17025 accreditation, because traceability and accreditation describe a measurement process performed by an accredited laboratory, not a property a shop can machine into a piece of metal. What a buyer should expect from the shop is a mill certificate for the material, a record of the heat treatment cycles applied and a dimensional inspection report against the drawing, covered in more detail in Invar material certification. The certified length and the traceability chain back to a national measurement institute come afterward from the buyer’s own accredited calibration lab or a third-party metrology lab that measures the finished, stabilized artifact.

What to put on the RFQ

An accurate quote for a calibration artifact needs more information than an ordinary bracket drawing:

• The exact geometry and nominal lengths or step spacings, with the tolerance on each.
• Whether Invar 36 or Super Invar is required, and the specification the alloy should meet.
• The stress relief and stabilization cycles required, or a note that the shop should recommend a cycle for the intended service life.
• The surface finish on every measuring face, and whether grinding or lapping is expected.
• Where the artifact will be calibrated and by whom, so the shop can align its soak and first article inspection conditions with that lab’s expectations.
• Handling and storage requirements between machining and calibration, since a bar that sits in an uncontrolled warehouse can pick up new stress before it ever reaches the lab.

This guide is one of several under our metrology and calibration equipment overview, which covers the broader range of Invar parts these buyers order beyond calibration artifacts themselves. If you have a drawing for a step gauge, ball bar or custom check fixture, send it through our RFQ form and we match it with up to three shops that machine Invar and Super Invar to this kind of tolerance and can talk through the stabilization cycle before they quote.

Frequently asked questions

Reviewed against

• ASTM E228, Standard Test Method for Linear Thermal Expansion of Solid Materials
• Invar, Wikipedia

Values in this guide are typical handbook figures for planning. The mill certificate for your heat lot and the shop's quote for your geometry govern.

Keep reading

• Reading an Invar CMM Inspection Report: What Each Number Means · Process
• Invar Machining Tolerances: What a Shop Can Hold and What It Costs · Process
• Long-Term Dimensional Stability: Does Invar Age After Machining? · Materials
• Invar Mill Certifications: What a Certified Material Report Should Show · Choosing a Supplier

All process guides