Presets
Playback
Rig and instrument
Dynamics and view
Camera
Layers
Quality
Model check
The science
Strain and the strain gauge
A bonded foil gauge is an etched metal grid on a thin polymer film glued to the surface. When the surface stretches by a strain ε = ΔL/L, the grid gets longer and thinner and its resistance changes by ΔR/R0 = GF · ε. For constantan the gauge factor is GF ≈ 2.1, so a strain of 300 microstrain (0.03 %) changes a 350 Ω gauge by only 0.22 Ω. Nickel has GF ≈ -12, silicon about +120, which is large but strongly temperature dependent and nonlinear.
The cantilever (Euler-Bernoulli beam)
A tip force F = m g gives a bending moment M(x) = F (L - x) that is largest at the clamp. The face strain is ε(x) = M h / (2 E I) = 6 F (L - x) / (E b h²), positive (tension) on the top face and equal and opposite on the bottom face, with I = b h³/12. The tip deflection is w = F L³ / (3 E I) and the peak stress at the clamp is σ = 6 F L / (b h²); the page flags it when it exceeds the yield strength. The beam is drawn bent much more than in reality (set by the exaggeration slider); the numbers are the real ones.
The Wheatstone bridge
Four arms R1..R4 are fed with an excitation Vex. Node B sits between R1 and R2, node D between R4 and R3, and the output is Vo = Vex [ R3/(R3+R4) - R2/(R1+R2) ]. The page evaluates this exactly, not with the small-signal formula. With x = GFε:
Quarter bridge (one active gauge): Vo = Vex x / (4 + 2x) ≈ Vex x/4, slightly nonlinear (error about -x/2). Half bridge (top gauge in tension, bottom gauge in compression in adjacent arms): Vo = Vex x/2 exactly. Full bridge (two tension and two compression gauges): Vo = Vex x, four times the quarter bridge and exactly linear. Bending is the ideal case for the half and full layouts because the strain on the two faces is equal and opposite.
Temperature
A change ΔT alters the gauge resistance through the alloy coefficient β and through the mismatch between the expansion of the beam αs and of the gauge grid αg: ΔR/R = GF ε + [β + GF(αs - αg)] ΔT. In a quarter bridge this looks like a strain (the apparent strain, tens of microstrain per kelvin for general-purpose gauges). A dummy gauge on an unloaded piece of the same material in the neighbouring arm, or a half or full bridge in which all gauges see the same ΔT, cancels it in the ratio. Self-temperature-compensated gauges choose β to cancel the expansion term for one material, which is why the "steel" gauge on an aluminium beam still drifts.
Dynamics
The beam is reduced to one mode: meff w'' + c w' + k w = F(t) with k = 3 E I / L³ and meff = m + 0.1 kg + (33/140) ρ b h L. A weight released at once (step) overshoots to almost twice its static deflection; a sine force near f1 = (1/2π) √(k/meff) is amplified by 1/√((1 - r²)² + (2ζ r)²). Slow motion is set with the time-scale slider.
Instrument chain
The bridge output (millivolts) is amplified by the gain, corrupted by input-referred Gaussian noise, clipped at the ±2.5 V range of a 24-bit converter and converted back to an indicated strain εind = 4 Vo / (Vex GF n), with n = 1, 1, 2, 4 active gauges. "Balance" zeroes the offset caused by the resistor tolerances (initial imbalance).
Honest simplifications
Linear elastic beam and gauge (no creep, hysteresis, fatigue or yield behaviour: the page only warns); the same ε(x) shape is used in motion (quasi-static mode shape); transverse sensitivity, lead-wire resistance, gauge self-heating (the dissipated power is displayed but not fed back), thermal gradients and the adhesive layer are ignored; resistor tolerances follow a fixed pattern; the hanger mass of 0.1 kg is tared (it only adds moving mass). The two bottom gauges and the dummy are shown in 3D but only the wired ones act. The model is an educational one, not a calibration of a real load cell.