Capillary Flow: Pressure Drop Is Viscous, Not Static

Capillary flow delay comes from viscous pressure loss through a small-bore tube. Treat the tube as a hydraulic resistance, calculate its flow at the actual differential pressure, and then calculate how long that flow takes to displace the chamber volume.

Reject the usual quick fixes

Do not start with capillary-rise equations. Those equations describe the static height produced by surface tension and wetting. A hydraulic delay line normally operates because a pressure difference drives liquid through a tube, which is a dynamic pipe-flow problem.

Do not substitute static head for the operating differential pressure. Elevation can contribute to the pressure balance, but the required input is the net pressure difference between the two tube ends while flow is occurring.

Do not pick a turbulent friction factor from a normal pipe chart without checking Reynolds number. Small diameter and high viscosity commonly put capillary service in the laminar region, where viscous forces control the pressure loss.

Do not use simplified long-pipeline correlations for a very short or very small tube. Their fitted assumptions may not represent entrance losses, fittings, developing flow, or the low Reynolds numbers encountered here.

Separate dynamic pressure drop from static head

Define the capillary differential pressure as Δp = p_in - p_out, with any elevation contribution handled consistently in the pressure balance. It is the pressure available to overcome tube resistance at the instant being calculated.

Static liquid head and pressure are related by:

Δp = ρgh

Here ρ is density, g is gravitational acceleration, and h is head. This conversion does not make head loss and pressure drop different physical losses; it expresses the same energy loss in different units. The dimensional role of g matters, so never enter an unspecified “constant” without matching the unit system.

For a horizontal tube with negligible velocity changes outside the restriction, use the measured end-to-end pressure difference directly. For different elevations, include hydrostatic pressure with its proper sign. Also subtract chamber backpressure, spring force per effective area, downstream restriction loss, or other opposing load when those forces are present.

Confirm that laminar flow applies

Calculate mean velocity from V = Q/A, where A = πD²/4. Then calculate Reynolds number:

Re = ρVD/μ = VD/ν

D is tube inside diameter, μ is dynamic viscosity, and ν = μ/ρ is kinematic viscosity. Keep dynamic and kinematic viscosity separate; using one in an equation written for the other creates a density error.

Hagen-Poiseuille flow requires steady, fully developed, laminar flow of a Newtonian, incompressible fluid through a straight circular tube with no slip at the wall. Because Q is initially unknown, solve with the laminar equation, calculate Re from the resulting flow, and check whether the assumed regime is valid.

In laminar flow, the Darcy friction factor is f = 64/Re. The equivalent head-loss expression is:

h_f = 32νLV/(gD²)

Substituting Δp = ρgh_f and ν = μ/ρ produces the same pressure-flow relationship as Hagen-Poiseuille. If the Reynolds-number check does not support laminar flow, switch to a standard pipe pressure-drop method with an appropriate friction factor and include local losses.

Calculate pressure, flow, and ideal delay

For the laminar assumptions above, calculate volumetric flow with:

Q = πD⁴Δp/(128μL)

Rearrange the equation for the quantity needed:

• Δp = 128μLQ/(πD⁴)
• L = πD⁴Δp/(128μQ)
• D = [128μLQ/(πΔp)]^(1/4)

Use one coherent unit system throughout. An older inch-based form from the supplied engineering data is:

Q_in = 0.0245 d⁴Δp/(μl)

This form returns Q_in in in³/s when d is tube inside diameter in inches, Δp is differential pressure in psi, μ is mean dynamic or “absolute” viscosity in lb·s/in², and l is length in inches. Do not mix that coefficient with SI viscosity or length units.

If chamber displacement volume V_c, viscosity, and differential pressure remain constant, the ideal delay is:

t = V_c/Q = 128μLV_c/(πD⁴Δp)

The fourth-power diameter term dominates tolerance. A small inside-diameter error produces a much larger flow error, while length and viscosity affect delay linearly. Use the actual bore, not nominal size or outside diameter.

Account for a changing chamber load

The simple expression t = V_c/Q works only when flow is effectively constant. A chamber often develops changing backpressure as a piston moves, a spring compresses, a gas pocket changes volume, or the source pressure droops. In that case, calculate the delay incrementally:

dt = dV/Q(Δp, μ)

At each chamber position, determine inlet pressure, outlet pressure, fluid temperature, viscosity, and the net Δp. Apply the flow equation, calculate the time for that volume increment, and sum the increments over the stroke.

Do not treat a trapped gas volume as liquid displacement without accounting for gas compression. Do not apply the incompressible equation across flashing, cavitation, or two-phase flow. Air trapped in the tube or chamber also adds compliance, causing the observed response to depend on compression and expansion as well as capillary resistance.

For a very short tube, entrance and exit losses may be a material fraction of total loss. Represent them as local losses only after establishing the applicable flow regime. Fittings, valves, drilled passages, and abrupt area changes must be added separately; the straight-tube equation covers the capillary length itself.

Run the calculation from shelf data

1. Identify the liquid and record its operating temperature. Obtain dynamic viscosity at that temperature; convert kinematic viscosity using μ = ρν if density is known.
2. Measure the tube inside diameter and effective straight length. Inspect for flattened sections, partial plugs, fittings, and drilled transitions that add restriction.
3. Determine displaced chamber volume from measured travel and effective area, or measure the delivered volume directly.
4. Measure p_in and p_out while flow occurs. Calculate Δp = p_in - p_out, including elevation and changing chamber load where applicable.
5. Calculate laminar flow with Q = πD⁴Δp/(128μL). Calculate velocity and Reynolds number to validate the assumed regime.
6. Calculate t = V_c/Q for constant conditions. If pressure or load changes, divide the stroke into volume increments and sum dV/Q.
7. Add separately calculated local losses when the capillary is short or connected through restrictive fittings.
8. Compare the predicted delay with a timed test from the same initial pressure, temperature, and chamber position.

Get production moving with a clean, correctly sized tube and known fluid condition. Then correct the model with measured pressure and temperature data instead of trimming length repeatedly around an unidentified variable.

Verify the restored delay

Bleed the circuit before testing. Run several strokes from the same initial condition and record delay, inlet pressure, outlet pressure, and fluid temperature. Stable pressure and temperature with repeatable timing indicate that the hydraulic resistance is stable.

Change one controlled variable at a time. Increasing tube length should increase laminar delay in direct proportion. Increasing viscosity should also increase delay directly. Increasing differential pressure should reduce delay inversely, while increasing bore should reduce it according to the fourth power.

If those trends fail, inspect for a changing load, tube deformation, contamination, non-Newtonian fluid behavior, leakage, trapped air, or a non-laminar flow regime. Stop here if the required pressure approaches an equipment rating, the fluid cavitates, the tube cannot be identified, or measured pressure and flow cannot be reconciled without unsupported assumptions.

FAQ