This simulator models a thermowell immersed in a stirred calibration bath, with an interchangeable Pt100 resistance sensor or thermocouple insert, connection-head terminals, 2/3/4-wire cabling and a transmitter. Step the bath temperature, choose the sensor and wiring scheme, dial in lead resistance and cold-junction compensation, and inject open- or short-circuit faults, then watch thermal lag, wiring error and reference error separate a true process temperature from what the transmitter displays.
• A real-time 3D cutaway of the stirred bath with heater coil and impeller, protective thermowell, RTD winding or thermocouple hot-junction insert, connection head and terminals, 2/3/4-wire cable, and transmitter/recorder, with home view, focus-selected-part, full-enclosure/cutaway toggle, exploded view, auto-rotate, expand and show/hide numbered labels matching a companion diagram. • Eleven live controls: process bath temperature, terminal/initial probe temperature, probe time constant, sensor insert (Pt100 resistance sensor or thermocouple approximation), RTD wiring (two-wire, three-wire compensation, four-wire Kelvin), each lead resistance, three-wire residual mismatch, RTD excitation current, probe dissipation coefficient, a thermocouple cold-junction-compensation checkbox, and sensor condition (healthy / open sensor / shorted sensor). • Play/pause, single-step and larger-step time controls, plus a playback-speed selector. • Start/stop trial, step-bath-to-120°C and step-bath-to-25°C actions, with a live "what is happening" sequence narrative, operating state, switch-state tokens and a scrollable cell-readings table. • Ten live metrics: process bath temperature, actual probe temperature, Pt100 equivalent element resistance, RTD channel measured resistance, thermocouple relative EMF, transmitter temperature estimate, transmitter signal current, valid process indication error, estimated self-heating rise, and sensor-valid flag. • A Curves & measurements tab with two charts (bath/probe/indicated temperature, and loop current), the full model equations, and snapshot measurements. • An Experiments tab with four guided scenarios (thermal step, two-wire lead resistance, thermocouple without CJC, open element), a model-verification bench of independent automated checks, and a timestamped event log with a copyable trial report. • A Learn & assess tab with guided lessons, a knowledge-check quiz and a written scope/reference statement.
The probe approaches bath temperature with a finite first-order time constant (τ) rather than jumping to it instantly — after one time constant, the probe has completed roughly 63% of the total temperature change, and the thermal-step experiment demonstrates this explicitly on an 8-second time constant. On top of that lag, self-heating from RTD excitation current adds its own small rise (I²R divided by the dissipation coefficient), and this simulator uses the prior step's resistance when estimating that rise rather than solving it simultaneously.
Wiring scheme changes what the transmitter actually measures at the resistance terminals: a two-wire connection adds both lead resistances directly onto the sensed element resistance (Rmeasured = R(T) + 2·Rlead), producing a bias that looks exactly like a temperature offset. Three-wire wiring cancels matched lead resistance but leaves any residual mismatch uncorrected, while four-wire Kelvin sensing is treated as ideal, measuring only R(T) itself.
The Pt100 element follows the positive-temperature Callendar–Van Dusen approximation R(T) = 100[1 + 3.9083×10⁻³T − 5.775×10⁻⁷T²] Ω for T ≥ 0 °C, while the thermocouple channel uses a deliberately linear teaching approximation, E = 0.041(Thot − Tterminal) mV — meaning thermocouple output depends on the *difference* between the hot junction and the terminal (cold junction) temperature. Without cold-junction compensation, the transmitter has no way to know the terminal temperature and reports an estimate that is biased low by roughly that terminal temperature, exactly as the no-CJC experiment shows (~25 °C low with 25 °C terminals).
Open and short sensor faults are modeled as invalidating conditions rather than plausible-but-wrong temperatures: the transmitter flags the reading invalid and drives a representative fail-high 21.5 mA diagnostic current, so a fault is distinguishable from a real, valid low or high temperature. The stated scope notes the thermocouple's 41 µV/°C figure is a teaching approximation rather than a full Type K reference table, thermal lag is lumped rather than distributed, and resistance/EMF channels stay available for comparison even when the other insert type is selected.
The probe approaches the bath temperature with a first-order time constant (τ) rather than responding instantly. After one time constant has elapsed, the probe has completed about 63% of the total temperature change — the simulator's thermal-step experiment demonstrates this on an 8-second time constant as the probe moves from 25 °C toward an 80 °C bath.
Two-wire RTD wiring adds both lead resistances directly onto the measured resistance (Rmeasured = R(T) + 2·Rlead), which the transmitter misreads as extra temperature. Three-wire wiring cancels matched lead resistance but leaves any residual mismatch as error. Four-wire Kelvin sensing is modeled as ideal, measuring only the true element resistance R(T) with no lead-resistance contribution.
A thermocouple's EMF depends on the temperature difference between its hot junction and its cold junction (here, the connection-head terminals): E = 0.041(Thot − Tterminal) mV. Without CJC, the transmitter has no way to know the terminal temperature and reports an estimate biased low by roughly that terminal temperature — the simulator's no-CJC experiment shows about a 25 °C low bias with 25 °C terminals.
Open and shorted sensor conditions are modeled as invalidating faults: the transmitter flags the reading invalid and drives a representative fail-high 21.5 mA diagnostic current rather than reporting a plausible but wrong temperature. The model uses the Callendar–Van Dusen Pt100 equation over its stated range and a simplified 41 µV/°C linear thermocouple approximation rather than a full Type K reference table, with lumped (not distributed) thermal lag and self-heating computed from the prior time step's resistance.