Pulsating Fluid Flow Through a Heater: Do Not Size from Mean Flow Alone

System boundary

A thick-film-heated fluid passage supplied by an intermittent pump, including retained fluid, wall thermal capacity, inlet and outlet measurement planes and the heater-control sequence. Calculations here assume a defined single-phase fluid unless otherwise stated.

Integration interfaces

Integration risks

Describe the pulse in mass and time

Record the flowing interval, stopped interval, pulse mass and actual shape of the flow transition. A positive-displacement pump can still have delivery ripple, valve delay, compression and leakage. Its electrical drive signal is therefore supporting context, not a substitute for the fluid observation at the heater.

Identify whether the passage stays filled during the pause. A filled stagnant passage, a drained passage and a passage containing a gas pocket have different heat-transfer and pressure behavior. The pulse model developed here does not turn a no-flow interval into an authorized dry-heating interval; wetting and protection remain separate application requirements.

Use average power for the average energy question only

Over a repeating cycle with no net change in stored energy, the electrical input minus losses equals the net energy carried away by the delivered fluid. For a single-phase liquid with approximately constant heat capacity, a mass-weighted average temperature rise can be estimated from that cycle energy and delivered mass.

For an illustrative 100 W net input and 0.005 kg/s average flow, taking a fluid heat capacity of 4180 J/(kg·K) gives an average rise of approximately 4.78 K. That value says nothing about the order in which the energy reaches the fluid. It is not necessarily the temperature reported by a time-averaging sensor, especially when no fluid moves during part of the cycle.

Budget the energy accumulated while flow is stopped

For a first bounded check, integrate net heat input over the pause. If 100 W continues for one second with negligible heat leaving the chosen wall-plus-retained-fluid boundary, the additional stored energy is 100 J. A three-second pause gives 300 J under the same assumptions. Actual losses and temperature-dependent power must be included when they are significant.

If the effective participating capacity is assumed to be 20 J/K, these energies correspond to average node rises of 5 K and 15 K. The capacity must belong to material that participates over the interval. Assigning the entire appliance mass to a small heated wall can conceal a faster local rise. These calculations do not establish safe wall temperatures or an allowable pause.

Compare two equal-mean-flow cycles

Assume flow is 0.010 kg/s whenever the pump delivers. A one-second flow interval followed by a one-second pause has the same 0.005 kg/s mean as three seconds flowing followed by three seconds stopped. Continuous 100 W input gives the same cycle-average fluid energy per kilogram in the ideal loss-free periodic balance.

Their stopped-flow storage increments differ by a factor of three. This is why lengthening both portions of a cycle can preserve average delivery while worsening wall excursions or leading-portion variation. The result is not obtained by changing average pump flow; the distinguishing variable is how the delivery is grouped in time.

Do not spread restart energy uniformly over the whole dose

A lumped dose average distributes energy across all delivered mass, but the first portion may encounter the hottest wall or displace hotter retained liquid. As a screening allocation, 100 J assigned entirely to 0.002 kg of the example fluid corresponds to approximately 11.96 K of additional rise. The same energy spread over 0.010 kg gives only 2.39 K.

These are energy-allocation comparisons, not predictions that all stored heat instantly enters the first portion. Finite heat transfer, wall temperature, mixing and losses constrain the actual distribution. A calculated fluid temperature above the available hot boundary would invalidate that assumed transfer. Use the comparison to choose temporal and mass resolution for measurement, then determine the real leading-portion behavior experimentally.

Measure the phase of each thermal observation

Acquire actual flow or delivered-mass timing, heater voltage and current, inlet temperature, representative wall temperatures and outlet temperature on a common clock. Define where fluid time zero occurs relative to pump command and first movement through the heated region. Transport between the heater and sensor can shift the apparent peak into a later part of the pulse.

Assess sensor response and mixing volume. A large receiving cup gives useful average energy information but can erase a brief hot leading portion. A fast probe can still read its own stem or a wall rather than the moving fluid. Where appropriate, compare sequential collected portions with the continuous trace, preserving the exact mass and collection interval for each portion.

Compare control changes without changing the hydraulic question

A reviewed strategy may coordinate electrical input with verified flow, reduce pause energy or alter the delivery profile. Evaluate it against the same actual flow trace before attributing improvement to control. A pump command synchronized to heater duty is insufficient if valve opening or fluid arrival remains delayed.

Keep protection for prolonged no-flow, loss of prime and abnormal pressure independent of the desired dose-temperature behavior. An average-energy controller cannot guarantee local wetting or prevent every transient wall excursion. Any proposed pulse sequence must remain within the qualified mechanical, electrical and thermal boundaries of the assembled system.

Specify both the delivered dose and the within-dose extremes

Define acceptable average delivered temperature together with the relevant leading-portion, peak, minimum and response-duration requirements. State the initial condition, pause range, fluid identity, inlet state and flow profile covered. Include the longest credible pause and the first delivery after idle, not only a convenient stabilized repeating cycle.

Retain pulse-resolved traces and the calculation boundary that selected their resolution. Changes to pump mode, valve timing, retained volume, wall capacity or sensor position can invalidate the earlier result even at unchanged average flow. The completed review explains where stored energy goes during each delivery phase and prevents a favorable average from hiding an unacceptable portion.