Complete signal chains for mechanical strain measurement and temperature sensing, from physical transduction through calibration and error analysis.
A foil strain gauge changes resistance proportionally to mechanical strain. The gauge factor Gf relates fractional resistance change to strain.
G_f = (ΔR/R) / ε (typical: G_f ≈ 2.0 to 2.1 for metal foil)
Solving for strain from bridge output (quarter-bridge):
ε = (4 · V_out) / (V_ex · G_f) (linear approximation)
To determine principal strains and their directions, three gauges at known angles must be used. The rectangular rosette (0°/45°/90°) is most common. Given strains εa, εb, εc at these angles:
ε_1,2 = (ε_a + ε_c)/2 ± (1/√2) · √[(ε_a - ε_c)² + (2ε_b - ε_a - ε_c)²]
tan(2θ) = (2ε_b - ε_a - ε_c) / (ε_a - ε_c)
The Seebeck effect: when two dissimilar metals form a junction at different temperatures, a voltage proportional to the temperature difference is generated. The Seebeck coefficient is specific to the metal pair.
Most widely used. Range: −200 to 1350°C. Sensitivity: ~41 µV/°C. General purpose, good stability.
Range: −40 to 750°C. Sensitivity: ~52 µV/°C. Lower max temperature, susceptible to oxidation above 550°C.
Range: −200 to 350°C. Best for cryogenic and low-temperature measurements. Good accuracy.
Highest Seebeck coefficient (~68 µV/°C). Best sensitivity. Non-magnetic.
V_corrected = V_measured + V_CJC(T_ref)
where V_CJC(T_ref) is the thermocouple voltage at the
reference (ambient) temperature relative to 0°C.
T_junction = T_inverse(V_corrected) (using NIST tables or polynomial)
Platinum RTDs (Pt100, Pt1000) offer excellent accuracy and stability. Resistance increases linearly with temperature over a wide range.
R(T) = R_0 [1 + A·T + B·T² + C·T³(T-100)]
For Pt100: R_0 = 100 Ω at 0°C
A = 3.9083×10&sup-₃ °C&sup-¹
B = −5.775×10&sup-⁷ °C&sup-²
C = −4.183×10&sup-¹² °C&sup-⁴ (only below 0°C; C=0 above 0°C)
4-wire connection eliminates lead resistance error.
Thermistors are semiconductor devices with a large, nonlinear resistance-temperature relationship. NTC (negative temperature coefficient) types decrease resistance with increasing temperature. High sensitivity makes them excellent for narrow-range precision measurements.
1/T = A + B·ln(R) + C·[ln(R)]³
where T is in Kelvin, R in ohms
A, B, C are empirically determined constants
Simplified B-parameter equation:
1/T = 1/T_0 + (1/B)·ln(R/R_0)
V_corrected = V_measured + V_CJC = 18.516 + 1.000 = 19.516 mV
T ≈ 19.516 mV / 0.041 mV/°C ≈ 476°C (using linear approximation; precise answer requires NIST tables)
T_junction ≈ 476°C
138.50 = 100(1 + 0.00385·T) → 1.3850 = 1 + 0.00385·T → 0.3850 = 0.00385·T
T = 0.3850 / 0.00385 = 100.0°C
T = 100.0°C