ASTM E251-20a, published by ASTM Committee E28 and the current revision in the Book of Standards Volume 03.01, defines five mandatory test procedures for metallic bonded resistance strain gages: reference resistance, gage factor, temperature coefficient of gage factor, transverse sensitivity, and thermal output [S1].
The standard also states, in its scope section 1.4, that bonded resistance strain gages differ from extensometers: they measure average unit elongation (delta-L over L) over a nominal gage length instead of total displacement, which is the engineering reason each E251 parameter must be tested on a bonded, installed sample rather than on the raw foil [S3].
Five mandatory performance characteristics
ASTM E251-20a is structured as six sections: a general-requirements section (Part I) followed by five test parts numbered Part II through Part VI, each covering one characteristic that must accompany a strain-gage lot if the data are to be used with acceptable accuracy [S1]. The five performance characteristics are: (1) resistance at a reference temperature, (2) gage factor at a reference temperature, (3) temperature coefficient of gage factor, (4) transverse sensitivity, and (5) thermal output. Resistance measurement is the only non-destructive test; the other four are mechanical or thermal, so each consumed sample yields data that must be applied statistically to the rest of the lot [S1].
Practical values observed on metal-foil gages: gage factor in the 1.9-2.2 range, transverse sensitivity below 0.5 percent on precision grades, and thermal output inside plus or minus 20 microstrain per degree Celsius depending on alloy and backing [S5]. The typical service envelope runs from minus 269 degrees Celsius (cryogenic) to plus 300 degrees Celsius, set by the gage type and adhesive stack rather than by the foil itself [S5].
Why individual gages cannot be calibrated
Section 4.3 of ASTM E251-20a is explicit on a point that trips up new engineers: individual strain gages cannot be calibrated [S1]. Because four of the five E251 tests are destructive, a given sample is consumed in the test, and only its lot-mates inherit the resulting performance value. The standard therefore requires that this statistical fact be carried into the published uncertainty, and warns that if NIST-style calibration and traceability are mandatory, strain gages should not be used at all, since the chain cannot be closed on a single bonded device [S1].
This is also why the standard treats resistance at a reference temperature separately: it is the one parameter that can be measured on every shipped unit, and it is what balances the bridge circuit and anchors the change-in-resistance calculation that all other data ride on [S1].
Bonding, installation, and instrumentation error budget
Sections 4.1 and 4.4 of ASTM E251-20a list the practical variables that dominate field error and that the test methods explicitly do not eliminate: surface preparation chemistry, bond-line adhesive choice, protective coating, installer skill, and the instrumentation chain downstream of the gage [S1]. Test conditions can be severe enough that error signals from the strain-gage system exceed the structural deformation being measured, so E251-compliant data must always be published with documented uncertainty magnitudes rather than as bare readings [S1].
The standard makes clear that E251 outputs are inputs to a strain gauge selection decision, not a turnkey accuracy certificate. A properly characterized gage with bad bonding will still drift; a good bond on a poorly characterized lot will still carry the lot's spread into service. The five E251 parameters exist to expose that spread before the gage is committed to a structural test.
Comparison: the five E251 parameters side by side
The five parameters answer five different engineering questions, and an engineer should not treat them as interchangeable. A condensed comparison: reference resistance (Part II) balances the bridge and is non-destructive, so it is measured 100 percent on the lot; gage factor at a reference temperature (Part III) is the transfer function, typically 1.9-2.2 for foil, and is destructive; temperature coefficient of gage factor (Part IV) covers how the transfer function drifts with temperature; transverse sensitivity (Part V) is the unwanted response to strain perpendicular to the gage axis, held below 0.5 percent for precision grades; thermal output (Part VI) is the apparent strain generated by temperature change alone, expected inside plus or minus 20 microstrain per degree Celsius [S5].
Selection logic: if the test is a short-duration lab stress analysis at room temperature, Parts II and III dominate; if the test sits on an aerospace structure or pressure vessel element that sees thermal cycling, Parts IV and VI dominate; if the field strain field is genuinely biaxial, Part V moves up the list. All five are required by E251, but engineering weight shifts with the application [S1].
Failure modes and documented uncertainty
The most common reason E251 data is misapplied is the assumption that one lot-acceptance sheet covers every installation. Because each gage is bonded once and lives or dies on that bond, the lab value of gage factor, transverse sensitivity, and thermal output can shift in service by an amount set by the installer's surface prep, adhesive cure cycle, and protective coating. A related pitfall is ignoring transverse sensitivity in wide-gage or rosette configurations, where the nominally 0.5 percent specification can become a multi-percent error under biaxial load [S5].
For the insulation resistance tester and bridge-amplifier chain downstream, E251's uncertainty propagation is what sets the input bias the rest of the chain has to swallow. The standard's prescription is therefore conservative: publish the lot-mean and standard deviation, and treat every installed gage as a sample drawn from that distribution.
Revision history and related ASTM references
The historical trail inside ASTM's own catalog lists ASTM E251-74 as the earliest cited revision, then E251-92 (reapproved 2009 as E251-92(2009)), and the active standard as ASTM E251-20a priced at 104 US dollars for the single-user PDF [S1][S2][S8]. The standard's 2009 reapproval carried the same five-part test structure, so users coming forward from E251-92(2009) do not face a structural change in E251-20a, only tightened wording on the significance-and-use sections [S2].
ASTM E251 is paired with ASTM E177, which defines the precision-and-bias vocabulary that E251 leans on for its statistical treatment of destructive-test data [S4]. A lab quoting E251 to a client should therefore also reference E177 explicitly in the report header so the uncertainty language is unambiguous.
What an engineer should look for in an E251 report
Three things separate a usable E251 report from a marketing flyer. First, the report must name all five characteristics and report them at a stated reference temperature, not just gage factor in isolation. Second, the report must publish the sample size n and the resulting standard deviation per lot, since E251's destructive-test logic is statistical, not point-estimate [S1]. Third, the report must list the bonding adhesive, backing material, and protective coating actually used, because thermal output and apparent strain depend on the full stack, not the foil alone [S1].
For an RTD PT100 or load-cell builder sourcing gages for a measurement chain, the E251 data sheet is the closest thing to a datasheet the strain-gage world offers, and the same three checks apply. Lab service listings for E251 testing, such as the Infinita Lab and MaTestLab pages, list these same five parameters in their scope, which is the practical cross-check that the lab is running the full E251-20a protocol rather than a subset [S4][S5].
Track these signals next: any future ASTM E28 ballot that revises the transverse-sensitivity limit below the current 0.5 percent benchmark for precision grades, and any move to align E251's thermal-output limits with the wider temperature ranges now being specified for pressure transmitter and flow meter sensing elements on cryogenic and high-temperature process lines. Until either lands, E251-20a remains the working reference for metallic bonded resistance strain-gage characterization.
This topic is covered further in Channel vs Coreless Induction Furnace: Design, Application, Trade-Offs.