SM&T Section 3
Standards Measurement & Testing Project No. SMT4-CT97-2165
UNCERT MANUAL: 2000
Manual of Codes of Practice for the Determination of Uncertainties in
Mechanical Tests on Metallic Materials
SECTION 3
Typical sources of uncertainty in materials testing
F A Kandil
National Physical Laboratory
Queens Road
Teddington, Middlesex TW11 0LW
UNITED KINGDOM
Issue 1
September 2000
SM&T Section 3
Standards Measurement & Testing Project No. SMT4-CT97-2165
UNCERT MANUAL: 2000
Uncertainty values associated with standard instruments used in
mechanical testing
Important Note
The following uncertainty values are typical and given for guidance only to the
authors of the UNCERT Codes of Practice. The uncertainty values relating to the
instruments actually used in the measurement must be obtained from either valid
calibration certificates, Standards specification, manufacturer’s information or other
means approved by the laboratory’s accreditation authority.
3.1 Uncertainty in force measuring devices (for static and low cycle fatigue
testing machines)
Class of Machine † Expanded uncertainty, U, % ‡
0.5 ±0.44
1 ±0.88
2 ±1.75
3 ±2.61
† According to ISO 7500/1 “Metallic materials - Verification of static uniaxial testing machines -
Part 1: Tensile testing machines”, First edition, 1986.
‡ Assuming: (a) normal distribution with a coverage factor, k, of 2,
(b) no temperature effect,
(c) no significant drift, and
(d) no end-loading effects.
3.2 Uncertainty in strain measurement using extensometers
Class of extensometer † Expanded uncertainty, U, ‡
Relative, % Absolute, µm
0.2 ±0.2 ±0.6
0.5 ±0.5 ±1.5
1 ±1.0 ±3.0
2 ±2.0 ±6.0
† According to EN 10002-4: 1994 “Tensile testing of metallic materials - Part 4. Verification of
extensometers used in uniaxial testing”, CEN, November 1994.
‡ Assuming: (a) normal distribution with a coverage factor, k, of 2,
(b) whichever value is the greater of the relative and absolute errors, and
(c) relative errors on the gauge length are not exceeding the limits specified in
the standard.
Page 1 of 2
SM&T Section 3
Standards Measurement & Testing Project No. SMT4-CT97-2165
UNCERT MANUAL: 2000
3.3 Typical errors in length measuring devices
Instrument used Resolution, mm † Accuracy
Micrometer (mechanical) ±0.01 ±0.002 mm
Micrometer (digital) ±0.001 ±0.002 mm
Vernier caliper (digital) ±0.01 ±0.02 mm
Steel rule (up to 1000 mm) ±0.5 ±0.5 mm or 0.1% ‡
† Resolution is defined as follows:
For an analogue scale, the resolution is the centre-to-centre distance between two adjacent scale
graduation marks (scale interval).
For a digital scale, the resolution is considered to be one increment on the numerical indicator,
provided that the indication does not fluctuate by more than one increment.
‡ Whichever value is greater.
3.4 Uncertainty in thermocouple readings
Type of thermocouple Temperature range, oC Uncertainty †
K -40 to +375 ± 1.5oC
375 to 1000 ± 0.004 T
T -40 to +125 ± 0.5oC
125 to +350 ± 0.004 T
J -40 to +375 ± 1.5oC
375 to 750 ± 0.004 T
N -40 to +375 ± 1.5oC
375 to 1000 ± 0.004 T
E -40 to +375 ± 1.5oC
375 to 800 ± 0.004 T
R 0 to +1100 ± 1.0oC
1100 to 1600 ± [1 + 0.003 (T -1100)]
S 0 to +1100 ± 1.0oC
1100 to 1600 ± [1 + 0.003 (T -1100)]
† For Class 1 thermocouples according to IEC 584.2, 1982 (BS EN 60584.2:1993). The
uncertainty tolerance is expressed either as a deviation in degrees Celsius or as a function of the
actual temperature assuming (a) no drift, and (b) no effects related to wire diameters and
conductor insulating materials.
Page 2 of 2