Pipe-flow cross-checks re-run: 2026-10-22
Start with the question
Consider a circular pipe with a 10 mm internal diameter and a mean speed of 0.2 m/s. Comparing inertia with viscosity and comparing thermal advection with diffusion are different questions, so Re alone cannot answer the heat-transfer one. This note links three calculators, checks that their outputs agree, then repeats the work with tabulated water properties at two temperatures.
Example 1 uses rounded constants (ρ = 1000 kg/m³, μ = 0.001 Pa·s, cₚ = 4200 J/(kg·K), k = 0.6 W/(m·K)) that are easy to follow by hand; they are not a water-property table. Examples 2 and 3 take ρ, μ, cₚ and k from the IAEA table of saturated water (Table 3.3) at 20 °C and 60 °C, as an approximation for low-pressure liquid water. Diameter and speed are assumed in all three cases, and L is always the internal diameter.
1. Prepare the inputs in SI units
| Quantity | Input | Definition |
|---|---|---|
| Density ρ | 1000 kg/m³ | Same density for both diffusivities |
| Mean speed v | 0.2 m/s | Cross-section average |
| Internal diameter L | 0.01 m | 10 mm divided by 1000 |
| Dynamic viscosity μ | 0.001 Pa·s | Not kinematic viscosity |
| Specific heat cₚ | 4200 J/(kg·K) | Used to obtain thermal diffusivity |
| Conductivity k | 0.6 W/(m·K) | Property of the same fluid |
Entering 10 for diameter means 10 metres: Re and Pe become 1000 times too large, yet their relationship still holds, so an algebraic cross-check cannot replace a unit check. Likewise, 4.2 (kJ) typed instead of 4200 (J) makes Pr and Pe 1000 times too small while Re stays correct.
Three worked cases on the same 10 mm pipe
Example 1. Rounded constants: Re, Pr and Pe by two routes
Reynolds number
Check it with kinematic viscosity: ν = μ/ρ = 10⁻⁶ m²/s, giving Re = vL/ν = 0.002/10⁻⁶ = 2000. Entering ν into a field labelled μ changes the physical meaning because their units differ.
For fully developed flow in a smooth circular pipe, 2000 is in the usual laminar range. Entrance effects, disturbances and roughness still need separate consideration; a displayed regime label does not measure those conditions.
Open the Re calculatorPrandtl number and thermal diffusivity
Pr = μcₚ/k = 0.001 × 4200 / 0.6 = 7
α = k/(ρcₚ) = 0.6/(1000 × 4200) = 1.428571… × 10⁻⁷ m²/s
Independent check: ν/α = 10⁻⁶/(1.428571… × 10⁻⁷) = 7
Momentum diffuses seven times faster than heat for these inputs. Pr does not depend on speed, so if it changes when only v changes, another property or unit changed too.
Open the Pr calculatorPéclet number by two routes
The second route gives Re × Pr = 2000 × 7 = 14000, since (ρvL/μ)(μcₚ/k) = ρcₚvL/k = vL/α for the same length, speed and properties. Rounding α to 1.43 × 10⁻⁷ first gives about 13986; check input precision before calling that a calculator error.
The timescale interpretation offers another check: tadvection = L/v = 0.05 s and tdiffusion = L²/α = 700 s. Their ratio is 14000. These scales come from the chosen diameter; they do not predict outlet temperature or heating time.
Open the Pe calculatorExample 2. IAEA water at 20 °C in the same pipe
Given. From IAEA Table 3.3 (saturated water, 20 °C): ρ = 998.2 kg/m³, μ = 1.0000 × 10⁻³ Pa·s, cₚ = 4.185 kJ/(kg·K) = 4185 J/(kg·K), k = 0.5995 W/(m·K). The table also lists ν = 1.0020 × 10⁻⁶ m²/s, α = 1.435 × 10⁻⁷ m²/s and Pr = 6.981. Note the scaled column headers: 10.000 in the viscosity column is 10⁴μ and 599.5 in the conductivity column is 10³k. Assumed: L = 0.01 m, v = 0.2 m/s.
Re = 998.2 × 0.2 × 0.01 / 0.001 = 1996.4
Pr = 0.001 × 4185 / 0.5995 = 6.9808 (table: 6.981)
α = 0.5995/(998.2 × 4185) = 1.4351 × 10⁻⁷ m²/s (table: 1.435 × 10⁻⁷)
Pe = vL/α = 0.002/(1.4351 × 10⁻⁷) ≈ 13936.5; Re × Pr = 1996.4 × 6.9808 ≈ 13936.5
tdiffusion = L²/α ≈ 696.8 s
Interpretation. Example 1's constants are within about 0.4% of the 20 °C row, so Re, Pr and Pe barely move. Note that Re from the table's own ν is 0.002/(1.0020 × 10⁻⁶) = 1996.0, not 1996.4: μ/ρ = 1.0018 × 10⁻⁶, and the table rounds ν separately. A fifth-digit difference between the two routes is a rounding signature of the table, not a formula error. For a clean identity check, derive ν, α and Pr from one consistent set of ρ, μ, cₚ and k.
Example 3. Heat the water to 60 °C at the same speed
Given. IAEA Table 3.3 at 60 °C: ρ = 983.2 kg/m³, μ = 4.677 × 10⁻⁴ Pa·s, cₚ = 4183 J/(kg·K), k = 0.6510 W/(m·K); the table lists Pr = 3.005 and α = 1.583 × 10⁻⁷ m²/s. Assumed: the same L = 0.01 m and v = 0.2 m/s, with all properties evaluated at 60 °C.
Re = 983.2 × 0.2 × 0.01 / (4.677 × 10⁻⁴) = 4204.4
Pr = 4.677 × 10⁻⁴ × 4183 / 0.6510 = 3.0052
α = 0.6510/(983.2 × 4183) = 1.5829 × 10⁻⁷ m²/s
Pe = 0.002/(1.5829 × 10⁻⁷) ≈ 12635 = Re × Pr
tdiffusion = L²/α ≈ 631.8 s
Interpretation. Viscosity more than halves between 20 °C and 60 °C, so Re more than doubles at the same speed. Pe moves the other way, dropping about 9%, because thermal diffusivity grows by about 10%. Re = 4204 is also above the usual circular-pipe transition value of about 2300, so laminar results such as Nu = 3.66 or f = 64/Re cannot be carried over without checking the flow state. Bulk temperature changes along a heated pipe, so treat the 60 °C case as a bounding calculation.
Side-by-side results
| Case | Re | Pr | α (m²/s) | Pe |
|---|---|---|---|---|
| Example 1, rounded constants | 2000 | 7 | 1.4286 × 10⁻⁷ | 14000 |
| Example 2, IAEA water 20 °C | 1996.4 | 6.9808 | 1.4351 × 10⁻⁷ | 13936.5 |
| Example 3, IAEA water 60 °C | 4204.4 | 3.0052 | 1.5829 × 10⁻⁷ | 12635 |
Changing only v to 0.4 m/s in Example 1 gives Re = 4000, Pr = 7 and Pe = 28000. This shows formula sensitivity at fixed properties, not a solution for a heated flow.
Is the flow laminar and fully developed?
The value 2300 is a representative transition value for a uniform circular tube, not a sharp limit. The MIT physiological fluid mechanics notes state that flow in such a tube typically becomes turbulent above about 2300, while flow leaving a narrow orifice into a larger tube can become turbulent at Reynolds numbers as low as 10. Other references draw the line elsewhere: the TU Delft slurry-transport text uses 2320 to switch from 64/Re to the Colebrook–White friction factor, and the WIPP performance-assessment model treats Re below 2100 as laminar. Example 2 at Re ≈ 1996 sits below all three values; Example 3 at Re ≈ 4204 is above all of them.
Developed flow is a separate condition. The same MIT notes give a laminar entrance length of Le/D = 0.065 Re. For Example 1 that is 0.065 × 2000 = 130 diameters, or about 1.3 m in a 10 mm pipe. A 0.5 m test section would therefore be entirely in the developing region, where the thin wall boundary layer gives higher local friction and heat transfer than the fully developed values. Check the available straight length before quoting fully developed results.
Troubleshoot a mismatch
- Check whether one calculation uses diameter while another uses total pipe length.
- Distinguish μ from ν and convert specific heat from kJ to J if needed.
- Derive α from the same ρ, cₚ and k, or take every property from the same table row.
- Separate thermal Pe from species Pe. With a species diffusivity D, the corresponding identity is Re × Sc.
- Compare to the same number of significant figures. A difference in the fifth digit usually comes from rounded table entries, as in Example 2.
What these outputs cannot determine
These ratios give no pressure loss, heat-transfer coefficient h, wall temperature or heat rate. To evaluate Nu = hL/k, first obtain h from a measurement or a correlation suited to the geometry and boundary conditions. Record property sources, evaluation temperature and the definition of L so other calculations can be compared on the same basis.
Questions about cross-checking Re, Pr and Pe
Why does my Pr from μcₚ/k not match the table value exactly?
Tables round each column separately. With the IAEA 20 °C row, μcₚ/k = 6.9808, which rounds to the listed 6.981. A mismatch in the fourth or fifth significant figure is normal; a mismatch of 10% or more usually means a unit slip (kJ versus J, or ν entered as μ).
Should L be the pipe diameter or the pipe length?
For Re and Pe in a pipe, use the internal diameter. Pr contains no length at all. The pipe length enters separately through ratios such as length/diameter or the entrance length Le/D. Mixing diameter in one number and length in another breaks the identity Pe = Re × Pr.
Is Re = 2000 always laminar?
No. Sources use different reference values (2100, about 2300, 2320) for smooth circular pipes, and disturbances at the inlet, vibration or an upstream orifice can trigger turbulence much earlier. Treat 2000 as "laminar is plausible if the inlet is calm and the pipe is long and straight", not as a guarantee.
At what temperature should the properties be evaluated?
For a heated or cooled pipe, a common choice is the mean of inlet and outlet bulk temperatures. Example 3 shows why the choice matters: moving from 20 °C to 60 °C doubles Re and more than halves Pr. State the evaluation temperature next to every result.
Can the same cross-check be used for mass transfer?
Yes, with the species diffusivity D in place of α: the mass-transfer Péclet number equals Re × Sc, where Sc = μ/(ρD). Keep the thermal and species values in separate columns because they can differ by orders of magnitude in liquids.
References
- NASA Glenn Research Center: Reynolds Number. Re as the ratio of inertial to viscous forces.
- Penn State CE 574: 6.6 Dimensionless Numbers. Pe compares advection and diffusion time scales.
- Michigan Technological University, CM3120: Module 2 Lecture IV. Thermal Pe = Re Pr and mass-transfer Pe = Re Sc.
- IAEA (2008): Thermophysical Properties of Materials for Nuclear Engineering. Table 3.3, saturated water at 20 °C and 60 °C (ρ, μ, ν, cₚ, k, α, Pr).
- MIT OpenCourseWare, HST.542J: Physiological Fluid Mechanics. Transition near Re 2300 in a uniform tube, much lower for orifice flow; entrance length Le/D = 0.065 Re.
- Engineering LibreTexts (S. Miedema, TU Delft): 3.2 The Darcy-Weisbach Friction Factor. 64/Re below Re 2320 and the Colebrook–White equation above it.
- U.S. DOE WIPP: CRA-2009 Appendix PA. A model using Re = 2100 as the laminar/turbulent boundary.