A next-generation approach to pipe friction and hydraulic surface characterization
Most conventional turbulent pipe-friction calculators reduce the wall to a single equivalent roughness. The 2P framework introduces two independent hydraulic coordinates, k and q, so measured friction can retain information associated with roughness scale and additional morphology-dependent resistance.
Why a next-generation 2P formulation beyond Colebrook–White?
Limitations of a one-roughness-parameter description
Colebrook–White represents wall condition through one equivalent relative roughness, ε/D. This is very effective when the surface can be represented adequately by an equivalent sand-grain roughness. However, real engineered surfaces can differ in element shape, spacing, density, orientation and form drag even when they have similar characteristic roughness heights. A single ε/D cannot independently retain both a roughness length scale and this additional morphology-dependent hydraulic information.
The consequence is practical: two surfaces may have comparable equivalent roughness heights yet exhibit different friction-factor curves. In such cases, forcing the entire Reynolds-number dependence into one Colebrook roughness parameter can produce systematic residuals or an apparent roughness that changes when different portions of the data are fitted.
What the second 2P coordinate adds
The 2P model uses two independent hydraulic surface coordinates, k and q. In this calculator, k = ks/R represents the roughness-scale coordinate, whereas q provides an additional degree of freedom associated with the hydraulic influence of surface morphology and additional resistance not captured by roughness height alone. Thus, the model can distinguish surfaces that a one-parameter equivalent-roughness description may collapse onto the same ε/D.
Once Re, k and q are specified, the 2P friction factor is evaluated explicitly. This is computationally convenient, but the more important advantage is physical and descriptive: the second coordinate permits the measured friction curve to carry information beyond a single equivalent sand-grain roughness.
Why engineers may prefer 2P when suitable data are available
For multi-point hydraulic measurements, the 2P approach can be calibrated against the complete friction curve and compared directly with the best Colebrook fit. In the experimental database used by this research calculator, several calibrated rough-surface cases show lower 2P MAPE than the corresponding globally fitted Colebrook correlation. For example, the Taylor cone A5 case gives 1.42% versus 9.08%, and the Taylor small-hemisphere B3 case gives 3.13% versus 7.01%. These examples do not establish universal superiority, but they demonstrate why a second hydraulic coordinate can be valuable for morphology-sensitive surfaces.
Validate before you adopt
2P is intended to complement established engineering practice where a single equivalent roughness is not sufficient. The calculator therefore keeps Colebrook–White as a visible benchmark and reports comparative fit information where experimental datasets are available. A second parameter should be retained only when the measurements support it.
Smooth-wall consistency
The physically smooth limit corresponds to k = 0 and q = 0. This provides a direct baseline for checking that additional surface coordinates disappear when no roughness contribution is required.
Experimental comparison
For rough surfaces, judge the model from the complete friction curve—not from a single operating point. Compare 2P and Colebrook using residuals and error metrics over the measured Reynolds-number range, and examine whether the inferred parameters remain physically and statistically meaningful.
Talk to us — test 2P on your data
Are you working with a pipe, coating, corroded or scaled surface, manufactured roughness, or experimental friction data that cannot be represented satisfactorily by a conventional equivalent roughness? We are interested in discussing engineering applications, validation datasets, research collaborations and industrial case studies involving the 2P methodology.
What you can bring
Useful information includes Reynolds number and friction-factor measurements or, alternatively, flow rate, pipe diameter, pressure drop, test-section length and fluid properties. Surface measurements, photographs, profilometry, nominal roughness, corrosion/scaling information and operating history can further strengthen the hydraulic interpretation.
What we can investigate
The L'CADAME/UFRJ team can evaluate whether a two-parameter description is justified, estimate k and q, compare the result with Colebrook–White, assess parameter identifiability and uncertainty, and examine whether the inferred hydraulic signature is consistent with the available information about the surface.
Quick 2P calculation
Explicit 2P equation
Compare with Colebrook
Colebrook–White equation
Interactive friction diagram
Interactive 2P friction chart — graphical map of (Re, q) → X → (k, f2P)
Use the controls below exactly as in the 2P simulator, or click directly on the chart. The lower panel selects the pair (Re,q) and determines X(Re,q); the upper panel then uses that same X with k to obtain the Darcy friction factor.
Computational effort: 2P versus Colebrook
The 2P friction factor is evaluated explicitly once Re, k and q are known. The Colebrook–White equation is implicit and requires iteration because the friction factor appears on both sides of the equation. For a single engineering calculation the absolute time difference is usually small, but repeated evaluations can matter in pipeline-network solvers, transient simulations, optimization, Monte Carlo studies, and digital-twin applications.
Structural comparison
| Feature | 2P model | Colebrook |
|---|---|---|
| Surface description | k, q | ε/D |
| Friction-factor equation | Explicit | Implicit |
| Iteration required | No | Yes |
| Per-evaluation structure | Direct formula evaluation | Repeated logarithmic updates until convergence |
| Repeated large calculations | Potential computational advantage | Higher numerical overhead |
Live browser benchmark
Pressure-drop calculation
Parameter status
The parameter database distinguishes the epistemic status of each pair. k and q are hydraulic descriptors of a tested rough surface configuration; they are not universal material constants.
Calibrated Exact tested surface, well-populated curve, both parameters resolved away from the numerical lower bound.
Indicative Useful engineering starting estimate, but limited generality; recalibration is recommended.
Research-only One parameter is poorly resolved, the curve is sparse, or the fit is not strong enough for a calculator default.
Independent validation Dataset not included in the original 53/649 calibration database. Used to test the model independently; fitted parameters may carry specific identifiability limitations.
Fit your experimental friction data
Additional independent validation — Botros
The Botros data are kept separate from the core 53-dataset / 649-measurement database used in the engineering manuscript. This prevents the independent test from being confused with the data used for the main model assessment.
Botros Pipe #1
20 individual high-Reynolds-number measurements are available for Pipe #1. The measured interval is narrow and already close to a high-Re friction plateau, so the friction curve can be validated accurately while k and q remain less strongly identifiable individually.
Colebrook comparison
Core 2P parameter database — 53 datasets / 649 measurements
ℓt=kR is the turbulent roughness length and ℓd=qR is the dynamic roughness length. Colebrook ε/D and ks,C=D(ε/D) are fitted independently.
| Status | Source | Surface class | Case | N | Re range | D or Dh (mm) | R (mm) | q | k | ℓd=qR (mm) | ℓt=kR (mm) | Colebrook ε/D | Colebrook ks,C (mm) | MAPE 2P | MAPE CB | Dimension status |
|---|
Contribute an experimental dataset (optional)
If you calibrate the 2P model using your own experimental data, you may optionally identify the material/surface and volunteer the dataset for possible inclusion in the L'CADAME research database. Nothing is shared automatically. Contribution requires an explicit opt-in.
Surface identification
Attribution and consent
What is included in a contribution?
The package includes the pasted experimental Re,f points, fitted k and q, fitted Colebrook ε/D, fit-error statistics, the material/surface information entered above, and the explicit consent choices. In this offline prototype, clicking “Prepare contribution” downloads a JSON package to your computer. When the L'CADAME page is deployed with a secure backend, the same button can submit that package directly to the research database.
Theory, references and use
What is new about the 2P model? The 2P formulation is not another explicit approximation of the implicit Colebrook equation. It is a distinct rough-wall friction model that represents the hydraulic action of the surface with two independent parameters, k and q, rather than collapsing the wall condition into the single equivalent roughness ε/D used by Colebrook.
The 2P formulation distinguishes an effective roughness-induced vortex/dissipative contribution, represented through k, from an additional geometry-dependent/form-drag-related contribution represented by q. The parameter pair should be determined for a surface configuration and retained over its operating range. The hydraulic coordinate k = ℓ(r)/R is not automatically a physical roughness height or equivalent sand-grain roughness; q represents an additional dimensionless resistance. This calculator evaluates turbulent single-phase wall friction, not a multiphase flow model.
Primary theory: Cruz, D. O. A., Anbarlooei, H., Santos, C. M. M., and Celis, G. E. O. (2026), “An improved two-parameter model for turbulent rough-wall flows: Addressing the limitations of Colebrook and Gioia–Chakraborty type models,” Physics of Fluids 38, 065155, DOI 10.1063/5.0325227.
Near-wall/friction basis: Anbarlooei, H. R., Celis, G. E. O., Santos, C. M. M., and Cruz, D. O. A., “On the turbulent friction and the near wall structure of pipe flows.”
Engineering validation: manuscript in preparation. This public research edition provides the current model and calibration database for technical evaluation.
Usage statistics and contributed datasets: a true laboratory-wide user counter and direct dataset submission require a small secure server/database endpoint. This release uses a calculation counter stored only in your browser and downloads contribution packages only when requested. No calculator inputs or contributed datasets are sent to a server. Your browser may clear or block local storage.