Three variables decide the right thickness gauge: substrate material, presence (or absence) of a coating layer, and the tolerance the downstream quality call requires [S1][S2]. Pick wrong on any one and you get either a falsely thin elastomer reading or a misread on coated steel, both of which will fail a customer audit.
The measurement family splits into three working groups: ultrasonic thickness gauges (UTG) for one-sided access on metals, plastics, and composites; digital coating thickness gauges for paint, plating, and dry film on F or N substrates; and dead-weight dead-foot gauges for compressible rubber and elastomers under ISO 23529 / ASTM D3767 [S2][S4][S5]. A hand micrometer is not a substitute for any of these when the spec demands controlled contact pressure.
Match the gauge family to the substrate and access condition
For metallic tanks, ship hulls, and process piping where you only have one-sided access, an ultrasonic thickness gauge is the practical instrument: a probe fires a pulse, the pulse reflects off the back wall, and thickness is calculated from velocity × time ÷ 2 [S2]. Because the pulse-echo principle is independent of operator feel, UTG is the dominant choice for in-service corrosion surveys.
For a coated metal substrate (zinc, epoxy, powder coat, plating), a digital coating thickness gauge is the correct tool. The first sub-decision is ferrous (F) probe, non-ferrous (N) probe, or dual FNF that handles both magnetic and non-magnetic substrates [S5]. Use F on steel and iron, N on aluminium, copper, and stainless, and FNF when job site conditions alternate between the two within the same shift.
For soft elastomers, silicone gels, O-rings, and weather stripping, a hand caliper is unsuitable because operator grip introduces variable compression; the standard-compliant approach is a dead-weight dead-foot gauge that delivers a calibrated pressure in kilopascals through a precision-machined presser foot [S4]. Refer to the dedicated thickness gauge reference for the family-level taxonomy across these three groups.
Decision criteria: range, resolution, accuracy, and probe geometry
For UTG selection, the four engineering numbers to lock in are measurement range, resolution (down to 0.001 mm / 1 μm on high-precision models), sound-velocity adjustability, and calibration points (one-point vs. two-point with built-in reference block) [S2]. A standard UTG is sufficient for general metals and plastics; a high-resolution variant (≤1 μm) is justified for thin-wall tubing, precision machined parts, and laboratory use; a multi-layer UTG is required when you must read the coating and the substrate in a single pass.
For coating thickness gauges, the criteria are substrate compatibility (F / N / FNF), probe scale (0-500 μm, 0-1500 μm, 0-5 mm, 0-13 mm, and up to 25-31 mm on heavy-duty scales), probe shape (standard straight, 90° or 45° angled for bores, mini probes for small-diameter features down to 3 mm sample diameter, telescopic up to 122 cm for rope-access work), and high-temperature rating (PINIP-style probes up to 250 °C / 480 °F for hot substrates) [S5]. Match scale to coating thickness: a Scale 0.5 probe is more accurate on 0-500 μm paint films than a Scale 3 probe used outside its optimal band.
For rubber and elastomer gauges, the criteria shift to applied foot pressure, foot diameter, and the governing standard. The S4 product map lines three configurations against exactly these criteria: a 22.0 kPa presser foot at 4.0 mm for standard rubber (≥35 IRHD) under ISO 23529 / ASTM D3767, a 10.0 kPa foot at 16.0 mm for soft rubber (<35 IRHD) under ASTM D3767, and a modular head with swappable foot and weight for multi-standard labs. See the coating thickness gauge entry for probe-by-substrate decision logic.
Options lined up against duty: UTG vs coating gauge vs dead-foot
Three realistic options compared on four decision criteria. Use the comparison to shortlist rather than as a final purchase: (1) UTG pulse-echo: works on bare metal, plastic, glass, and composites, single-sided access, typical resolution 0.01-0.1 mm, requires couplant gel; (2) digital coating gauge with F or N probe: works only on coated metal substrates, contact or non-contact magnetic/eddy-current principle, resolution 1 μm on dedicated low scales, no couplant; (3) dead-weight dead-foot gauge: works only on compressible elastomers, delivers controlled kPa pressure, resolution 1 μm in lab use, requires flat parallel sample. None of the three is interchangeable with the others; specifying a coating gauge to measure bare pipe wall, or a UTG to certify an O-ring, will produce a number that looks plausible and is wrong. [S2]
For corrosion and remaining-wall-life work on pressure equipment, the ultrasonic path is the only one of the three that reads the back wall through the wall itself, which is why ultrasonic thickness gauge selection is a separate sub-discipline with its own dual-element vs single-element probe choice. For paint-line QA, a coating gauge with a probe scale matched to the expected film build is recommended, as coating thickness probes are designed for specific thickness ranges where they deliver more accurate readings.
Who should NOT pick the mainstream option
Buyers who should not default to a standard UTG: any lab certifying soft rubber under ISO 23529 or ASTM D3767, because the pulse-echo reading is meaningless on a material that deforms under the probe and is governed by kPa contact pressure, not time-of-flight [S4]. Buyers who should not default to a hand caliper for rubber: any QC station whose data is referenced in an automotive or aerospace PPAP package, because operator-dependent compression produces falsely thin measurements that have caused joint-binding failures in service [S4].
Buyers who should not default to a dual FNF coating probe: a paint line that runs only cold-rolled steel and only aluminium on separate shifts, because the FNF compromise reads both but is rarely as accurate on either as a dedicated F or N probe matched to substrate. Buyers who should not default to a standard straight coating probe: anyone measuring inside-diameter pipe, narrow recesses, or edges under 3 mm diameter, where a mini right-angle or telescopic probe is required to seat the tip correctly [S5].
Real use cases and the gauges that fit them
Corrosion monitoring on a refinery pressure vessel: a standard or high-precision UTG with a 5 MHz dual-element probe, paired with a known sound velocity for carbon steel (≈5920 m/s) and a two-point calibration block; this configuration reads remaining wall through the external surface without entering a confined space. Dry film thickness on a structural steel bridge: a digital coating gauge with an F probe, Scale 0.5 or Scale 1 (0-1500 μm), supplied with a Test Certificate and calibration foils matched to the scale [S5].
Lab certification of an automotive O-ring: a dead-weight dead-foot gauge at 22.0 kPa with a 4.0 mm foot for standard nitrile ≥35 IRHD, or 10.0 kPa with a 16.0 mm foot for soft silicone <35 IRHD, per the S4 selection table. High-temperature coating inspection on a hot-dip galvanizing line: a high-temperature PINIP probe rated to 250 °C so the substrate can be read in-process without waiting for cool-down [S5]. Rope-access paint inspection on a tank exterior: a telescopic probe extended to 122 cm, or a PINIP plug-in probe that turns a separate gauge into an integral single-handed unit. For related QA decisions around coating lines and surface treatment, see OEM vs ODM for industrial coatings: a 2026 spec-by-spec decision map.
Limitations, failure modes, and what to verify before purchase
UTG failure modes are dominated by couplant failure, surface roughness, and unmatched velocity. The instrument will return a stable number even when the pulse is reflecting off a near-surface lamination rather than the back wall, so a B-scan or A-scan display is the practical defence. Coating gauge failure modes are dominated by substrate mismatch (using F on aluminium gives nonsense) and scale mismatch (using a 0-13 mm probe on a 200 μm film). Dead-foot gauge failure modes are dominated by sample parallelism, foot cleanliness, and calibration of the dead-weight stack; verify the kPa reading against a traceable load cell annually. [S5]
Before purchase, verify three things on the datasheet: a stated accuracy figure in μm or mm with the test conditions, a list of included calibration foils or test blocks with traceability statement, and the operating environment rating (IP65 for field use is common on UTG; IP rating matters less for lab dead-foot gauges). For buyers cross-referencing sheet metal gauge charts to physical thickness, the S6 guidance to "choose the thinnest material that meets the required thickness to minimize cost" only applies when the thickness gauge is a downstream verifier, not the spec source itself. Buyers comparing thickness instruments against other dimensional metrology such as height gauge or force gauge selections should keep the families separate: each tool answers a different physical question.
Shortlist logic and final selection signals
Shortlist by writing down substrate (ferrous, non-ferrous, elastomer, composite), access (one-sided vs two-sided), coating present (yes/no), and required tolerance in μm before looking at any catalogue. That four-line spec cuts the field from dozens of models to two or three. Confirm the chosen model ships with a Test Certificate and calibration foils or reference blocks appropriate to its scale; this is the audit trail that survives a customer or notified-body review [S5].
Two trackable signals to monitor over the next procurement cycle: published updates to ISO 23529 and ASTM D3767 for elastomer methods, and any manufacturer move to extend high-temperature PINIP probe ratings beyond 250 °C for in-process hot-substrate inspection. Either shift will reopen the shortlist on rubber and coating gauges respectively. Buyers looking at adjacent dimensional QA, including gauge block selection for instrument calibration, should treat the thickness gauge shortlist and the gauge-block calibration chain as one connected decision rather than two separate purchases.