Strain gauges and rosettes
How resistance strain gauges measure strain and how rosette layouts (0-45-90 and 0-60-120) resolve principal strains/stresses and their directions using bridge circuits and strain transformation.
Drafted with Aria, reviewed by the AiCanCode.org team · Updated . How we write and check our notes. Spotted an error? Use Give Feedback at the bottom of the page.
Why it matters
Strain gauges turn tiny deformations on real parts into voltages we can measure, letting us validate designs, find hotspots, and keep machines safe. Rosettes (three gauges at known angles) let us recover the full in‑plane strain state and hence principal stresses where directions are unknown. This bridges analysis (Mohr’s circle, Hooke’s law) with testing (bridges, instrumentation) in labs and on structures.
Key ideas
- A strain gauge is a bonded electrical sensor whose resistance changes with mechanical strain along its grid direction. When the surface elongates by a small normal strain ε (tension positive), the gauge’s resistance increases; under compression it decreases.
- The gauge factor (GF) defines electrical sensitivity: GF = (ΔR/R)/ε. For common metallic foil gauges (e.g., constantan), GF ≈ 2 (datasheets give the exact value). Origin: change of length and area (geometry, giving 1 + 2ν term) plus a small piezoresistive contribution (material dependent).
- A Wheatstone bridge converts small resistance changes into a differential voltage. With small ΔR, the normalized bridge output scales linearly with ΔR/R. Configurations:
- Quarter-bridge: 1 active gauge; simplest but less sensitive and more temperature/error prone.
- Half-bridge: 2 active arms; can add outputs (tension/compression pair) and cancel temperature and lead resistance.
- Full-bridge: 4 active arms; highest sensitivity and best compensation when all gauges respond.
- A strain rosette is three strain gauges at known in‑plane angles at the same point. Standard types:
- Rectangular rosette (0°–45°–90°): easy algebra; common on flat plates.
- Delta rosette (0°–60°–120°): equal angular spacing; more rotationally symmetric error behavior.
- We assume a plane strain state at the gauge area is uniform (gauge length small vs strain gradient), so the measured strain equals the average surface strain along the gauge axis. Usually, thin plates under in‑plane load are modeled as plane stress for stress recovery.
- Strain transformation (plane strain) with engineering shear strain γxy (positive if a positive τxy would cause the +x face to move in +y) states: εθ = (εx + εy)/2 + (εx − εy)/2 cos 2θ + (γxy/2) sin 2θ. This underlies all rosette formulas.
- Principal strains ε1, ε2 are the extrema of εθ as θ varies; they and their directions (θp) follow from Mohr’s circle for strain or by formula. Once ε1, ε2 are known, principal stresses follow from Hooke’s law with a model assumption (plane stress or plane strain) and known E, ν.
- Real vs ideal:
- Temperature effects: mismatched coefficients of thermal expansion (CTE) between gauge and structure produce apparent strain. Bridges (dummy gauges), self‑temperature‑compensated (STC) gauges, and instrument correction mitigate this.
- Transverse sensitivity: gauges respond weakly to strain perpendicular to the grid; small but non‑zero (manufacturer specifies). Matters in high Poisson effect fields.
- Strain transfer: adhesive layer and backing filter and attenuate strain (loss at high gradients); gauge measures an average over its length.
- Lead resistance, self‑heating, drift, creep: wiring and excitation cause offsets and slow changes; minimized by 3‑wire/4‑wire techniques, low excitation, and stable adhesives.
- Sign conventions used here (state clearly in your working, as textbooks may differ):
- Tensile normal strain ε positive; compressive negative.
- Engineering shear strain γxy positive if a positive τxy would rotate a material element counter‑clockwise (consistent with most Mohr’s circle derivations used in GATE).
- Angles θ measured counter‑clockwise from the x‑axis (here, the 0° rosette gauge).
- Standard rosette relations (measured strains εa at 0°, εb at 45°/60°, εc at 90°/120°):
- Rectangular (0°, 45°, 90°):
- εx = εa, εy = εc, γxy = 2(εb − (εa + εc)/2)
- Principal strains: ε1,2 = (εa + εc)/2 ± sqrt([ (εa − εc)/2 ]² + [ εb − (εa + εc)/2 ]²)
- Principal direction: tan 2θp = (2εb − εa − εc)/(εa − εc)
- Delta (0°, 60°, 120°):
- εx = εa, εy = [2(εb + εc) − εa]/3, γxy = (2/√3)(εb − εc)
- Principal strains: ε1,2 = (εa + εb + εc)/3 ± (2/3) √(εa² + εb² + εc² − εaεb − εbεc − εcεa)
- Principal direction: tan 2θp = [√3(εb − εc)]/(2εa − εb − εc)
- Rectangular (0°, 45°, 90°):
- Converting strain to stress (Hooke’s law):
- Plane stress (thin plates): σ1 = E/(1 − ν²) [ε1 + ν ε2], σ2 = E/(1 − ν²) [ε2 + ν ε1]
- Plane strain (thick bodies): σ1 = 2G ε1' + λ(ε1' + ε2' + εz'), etc. (usually not required in rosette problems; if needed, use εz' = 0 and 3D relations). In most surface rosette use‑cases, plane stress is assumed.
- From stress analysis topics you already know (do not re‑derive here): Mohr’s circle for plane strain/stress gives the same principal values and directions; von Mises stress for ductile failure (Theories of Failure topic) is σv = √(σ1² + σ2² − σ1σ2) under plane stress.
- Instrumentation practice (what examiners may ask conceptually):
- Why half/full bridges double/quadruple sensitivity and cancel thermal effects.
- How a “dummy” gauge provides temperature compensation when the structural point strain is needed with quarter/half bridge.
- Choosing gauge length and alignment w.r.t. expected principal directions; when to prefer rectangular vs delta rosette.
Formulas
ε = ΔL / L
Normal strain; ΔL is change in length (m), L is original length (m); applies to small deformations.
R = ρ L / A
Electrical resistance; ρ resistivity (Ω·m), L length (m), A area (m²); base for gauge resistance.
GF = (ΔR / R) / ε
Gauge factor; ΔR/R is relative resistance change (dimensionless), ε is strain (dimensionless); manufacturer provides GF for a given gauge.
ΔR / R ≈ (1 + 2νg) ε + (Δρ/ρ)
First‑order relation for metallic gauges; νg gauge Poisson’s ratio (dimensionless), Δρ/ρ piezoresistive term (dimensionless). For constantan, Δρ/ρ is small; then GF ≈ 1 + 2νg. Use datasheet GF in practice.
Vout / Vs ≈ (1/4) (ΔR / R)
Quarter‑bridge small‑signal sensitivity; Vout differential output (V), Vs excitation (V); valid for |ΔR| ≪ R.
Vout / Vs ≈ (1/2) (ΔR / R)
Half‑bridge with two active arms changing equally and in the same additive sense; small‑signal.
Vout / Vs ≈ (ΔR / R)
Full‑bridge with four active gauges changing equally and additively; small‑signal.
εθ = (εx + εy)/2 + (εx − εy)/2 cos 2θ + (γxy/2) sin 2θ
Strain transformation (plane strain); εx, εy normal strains along x, y (dimensionless), γxy engineering shear strain (dimensionless), θ angle from x to the measurement axis (rad or deg); small strain, sign convention as stated.
ε1,2 = (εx + εy)/2 ± √( [ (εx − εy)/2 ]² + [ γxy/2 ]² )
Principal strains from components; dimensionless; plane strain.
tan 2θp = γxy / (εx − εy)
Principal strain directions relative to x; θp in radians or degrees; beware of quadrant using atan2.
Rectangular rosette: εx = εa, εy = εc, γxy = 2(εb − (εa + εc)/2)
Mapping from measured rosette strains (εa at 0°, εb at 45°, εc at 90°) to components; dimensionless.
Rectangular rosette principal: ε1,2 = (εa + εc)/2 ± √( [ (εa − εc)/2 ]² + [ εb − (εa + εc)/2 ]² )
Principal strains directly from 0°–45°–90° measurements; dimensionless.
Rectangular rosette angle: tan 2θp = (2εb − εa − εc) / (εa − εc)
Principal strain direction w.r.t. 0° gauge; use atan2 for correct quadrant.
Delta rosette: εx = εa, εy = [2(εb + εc) − εa]/3, γxy = (2/√3)(εb − εc)
Mapping from measured 0°–60°–120° strains to components; dimensionless.
Delta rosette principal: ε1,2 = (εa + εb + εc)/3 ± (2/3) √( εa² + εb² + εc² − εaεb − εbεc − εcεa )
Principal strains directly from 0°–60°–120° measurements; dimensionless.
Delta rosette angle: tan 2θp = [√3(εb − εc)] / (2εa − εb − εc)
Principal strain direction for delta rosette; use atan2.
σ1 = E/(1 − ν²) (ε1 + ν ε2), σ2 = E/(1 − ν²) (ε2 + ν ε1)
Principal stresses from principal strains (plane stress); σ in Pa, E in Pa, ν dimensionless. Applies to thin plates/surfaces.
σv = √(σ1² + σ2² − σ1 σ2)
Von Mises equivalent stress under plane stress; σ in Pa; useful for comparing with yield (Theories of Failure topic).
Worked examples
Example 1 — Direct use of a quarter-bridge to get strain and stress (uniaxial) Given: Quarter-bridge, Vs = 5.00 V, measured Vout = 1.25 mV, GF = 2.00, material E = 200 GPa. Assume: one active gauge aligned with the uniaxial stress, small-strain, temperature compensated.
Strain from bridge output. Formula: ε = 4(Vout/Vs)/GF Numbers: ε = 4(0.00125 V / 5.00 V) / 2.00 = 4(0.00025)/2.00 = 0.001 / 2.00 = 0.0005 Result: ε = 0.0005 = 500 µε.
Stress (uniaxial, plane stress is fine here). Formula: σ = E ε Numbers: σ = 200×10^9 Pa × 0.0005 = 100×10^6 Pa Final: σ = 100 MPa (tension).
Check: Quarter-bridge sensitivity 4×1.25 mV / (2×5 V) = 0.0005, consistent.
Example 2 — Rectangular rosette to principal strains/stresses and direction (GATE level) Given: Rectangular rosette (0°, 45°, 90°) at a point on a thin steel plate (plane stress). Measured strains: εa = +300 µε, εb = +600 µε, εc = −100 µε. Material: E = 210 GPa, ν = 0.30. Assume sign convention and formulas as in these notes; uniform strain over gauge area; temperature effects negligible or compensated.
Convert to strain components. Formulae: εx = εa; εy = εc; γxy = 2(εb − (εa + εc)/2) Numbers:
- εx = 300×10^−6
- εy = −100×10^−6
- Average (εa + εc)/2 = (300 − 100) µε / 2 = 100 µε = 100×10^−6
- γxy = 2(600×10^−6 − 100×10^−6) = 2×500×10^−6 = 1000×10^−6 = 0.001
Principal strains. Formula: ε1,2 = (εx + εy)/2 ± √( [ (εx − εy)/2 ]² + [ γxy/2 ]² ) Numbers:
- (εx + εy)/2 = (300 − 100) µε / 2 = 100 µε = 100×10^−6
- (εx − εy)/2 = (300 − (−100)) µε / 2 = 400/2 = 200 µε = 200×10^−6
- γxy/2 = 0.001/2 = 0.0005 = 500×10^−6
- Radius = √(200^2 + 500^2) µε = √(40,000 + 250,000) µε = √290,000 µε ≈ 538.5 µε = 538.5×10^−6
- ε1 = 100×10^−6 + 538.5×10^−6 = 638.5×10^−6
- ε2 = 100×10^−6 − 538.5×10^−6 = −438.5×10^−6
Principal directions. Formula: tan 2θp = γxy / (εx − εy) Numbers: tan 2θp = 0.001 / (300×10^−6 − (−100×10^−6)) = 0.001 / 400×10^−6 = 2.5
- 2θp = arctan(2.5) ≈ 68.20° → θp ≈ 34.10° (from the 0° gauge, counter‑clockwise)
Principal stresses (plane stress). Formulae: σ1 = E/(1 − ν²) (ε1 + ν ε2); σ2 = E/(1 − ν²) (ε2 + ν ε1) Numbers:
- 1 − ν² = 1 − 0.09 = 0.91; E/(1 − ν²) = 210×10^9 / 0.91 ≈ 230.769×10^9 Pa
- ε1 + ν ε2 = 638.5×10^−6 + 0.30(−438.5×10^−6) = 638.5×10^−6 − 131.55×10^−6 = 506.95×10^−6
- ε2 + ν ε1 = −438.5×10^−6 + 0.30(638.5×10^−6) = −438.5×10^−6 + 191.55×10^−6 = −246.95×10^−6
- σ1 ≈ 230.769×10^9 × 506.95×10^−6 ≈ 1.17×10^8 Pa = 117 MPa
- σ2 ≈ 230.769×10^9 × (−246.95×10^−6) ≈ −5.69×10^7 Pa = −56.9 MPa
Von Mises stress (optional, for design check). Formula: σv = √(σ1² + σ2² − σ1σ2) Numbers: σv ≈ √(117² + (−56.9)² − 117×(−56.9)) MPa ≈ √(13689 + 3239.6 + 6657.3) ≈ √23585.9 ≈ 153.6 MPa Final: ε1 ≈ 639 µε; ε2 ≈ −439 µε; θp ≈ 34.1°; σ1 ≈ 117 MPa; σ2 ≈ −56.9 MPa; σv ≈ 154 MPa.
Common mistakes
- Mixing microstrain and strain: forgetting that 500 µε = 0.0005.
- Using plane stress formulas for thick parts (plane strain) or vice‑versa without stating/justifying the assumption.
- Sign errors in γxy and tan 2θp; not using atan2, hence choosing the wrong principal direction quadrant.
- Plugging εb directly as shear strain in rectangular rosettes instead of using γxy = 2(εb − (εa + εc)/2).
- Ignoring Poisson’s effect when converting strains to stresses (i.e., taking σ = Eε1 for principal stress under plane stress; the correct relation includes ν).
- Taking GF = 2 exactly without using the gauge’s datasheet value; small differences matter.
- Neglecting temperature and lead resistance in quarter‑bridge setups; large apparent strains result.
- Poor alignment of gauges; a few degrees error rotates the axes and biases ε1, ε2.
- Using long gauges in regions with steep strain gradients, leading to underestimated peaks.
For GATE ME
- Direct computations:
- Convert bridge output to strain (quarter/half/full bridge sensitivities).
- Rectangular and delta rosette formulas to find εx, εy, γxy; then ε1, ε2 and θp.
- Convert principal strains to principal stresses under plane stress using E, ν.
- Conceptuals:
- Gauge factor meaning and why GF ≈ 2 for common metallic gauges; geometric vs piezoresistive contributions.
- Why rosettes are needed when principal directions are unknown; when to prefer 0°–45°–90° vs 0°–60°–120°.
- Temperature compensation strategies: dummy gauge, half/full bridge, STC gauges.
- Effects of transverse sensitivity and strain averaging.
- Practise:
- Derivations of rosette relations from εθ transformation (one‑page outline suffices: substitute θ = 45°, 60°, 90°, 120° and solve for εx, εy, γxy).
- Using Mohr’s circle for strain to check computed ε1, ε2 and θp.
- Careful unit handling and atan2 for angles.
Quick check
- A quarter‑bridge with Vs = 10 V and GF = 2 outputs Vout = 2.5 mV. What is the strain (in µε)? State assumptions.
- For a rectangular rosette, εa = 200 µε, εb = 350 µε, εc = −100 µε. Compute γxy (in µε).
- Given ε1 = 800 µε, ε2 = −200 µε, E = 200 GPa, ν = 0.3 (plane stress), find σ1 (in MPa).
Answers: 1) ε = 4(0.0025/10)/2 = 0.0005 = 500 µε (small‑signal, linear bridge). 2) γxy = 2(350 − (200 − 100)/2) = 2(350 − 50) = 600 µε. 3) σ1 = [200×10^9/(1 − 0.09)](800e−6 + 0.3×(−200e−6)) ≈ 219.8 MPa.
Finished this topic? Mark it so your progress, study plan and readiness keep up.
Stuck on something here?