# Wing analysis

The wing designer's Analyze tab, also at [/wing/analyze](/wing/analyze), runs the wing you drew through a 3D solver and answers what the section pages cannot: does the wing make the lift at take-off, and what does it cost at cruise? Planform, dihedral, twist and the root and tip sections go in together, so the answer is for the finite wing, not the airfoil it was built from. The running example is the Freeride wing foil carrying 95 kg from 4 to 10 m/s in salt water, on AeroBuildup.

## Two modes

![The top of the Analyze tab: the two mode cards, α sweep and Constant lift, each ringed, under the Wing under analysis summary.](https://media.foil.tools/docs/shots/wing-analysis-modes.b900c590c637.webp)

*The Analyze tab opens on two modes. α sweep is selected here; Constant lift is beside it.*

### α sweep

One speed, a range of angles of attack: the wing's polar. Set the velocity, then α min, α max and α step; one row per angle. The mode for CL(α), CM(α), the drag polar and the stall.

### Constant lift

A load, a range of speeds: at each speed the solver finds the angle at which lift equals the target load times g, on the pre-stall branch of the lift curve. Target load is in kg; negative is downforce, for a stabiliser. A speed too slow to carry the load before the wing stalls comes back infeasible, marked × on the chart. For a foil or a glider this is the question that matters: 95 kg at 4 m/s needs CL about 1.1, and the run says whether the wing gets there. Constant lift always runs on AeroBuildup, because trimming near stall is a viscous question.

## Two methods

- **AeroBuildup** walks NeuralFoil section polars across the span and assembles them with finite-span corrections. Stable through stall, about 10–20 seconds, 1 credit.
- **VLM** tiles the wing with horseshoe vortices and solves for the strengths that make the flow tangent at every panel. It resolves how sweep, taper and twist reshape the loading, is attached-flow only, takes 60–240 seconds, and costs 2 credits up to 10 spanwise panels or 3 above. The solver clamps the panel count at 16, and the price follows the clamp.

If VLM produces a non-physical lift curve the solver falls back to AeroBuildup, says so above the results, and charges the AeroBuildup rate.

## Speed unit and fluid

The Speed selector shows and reads every speed on the tab in m/s, km/h or knots: inputs, results header and plot axes alike; the solver works in m/s. Fluid sets density and viscosity for salt water, fresh water or air, the same constants the Design tab uses, so the Reynolds number shown there is the one analysed here.

![The α-sweep parameters with the speed unit buttons, the fluid buttons and the method buttons, AeroBuildup and VLM, each ringed.](https://media.foil.tools/docs/shots/wing-analysis-method.83e27ecd4bed.webp)

*Under the modes: the speed unit at the top right of the card, then the fluid and the method.*

## Running it

**Recording: One AeroBuildup run** (30 s, https://media.foil.tools/docs/wing-analyze-run.8239b270ef69.mp4)

> Constant lift trims the wing at each speed to carry the load: 95 kg from 4 to 10 m/s in salt water.
> AeroBuildup answers in seconds; VLM is the more careful solver for a few credits more.
> Run. The tiles give lift, drag, the best lift-to-drag ratio and span efficiency; the Checks card under them says where the wing stalls first and what moves it.
> The freeride wing as drawn stalls tip-first, at 72 % of the span. The Checks card names the fix: more washout, or a tip section with more maximum lift. That is the arrow back to stage 5.

Analyze Wing sends the wing as 20 stations sampled from the drawn splines, root and tip sections blended between them. The log under the button ticks the elapsed time. Changing the wing or the sweep afterwards marks the results stale until you re-run.

## What it costs

Sign-in is required. A free account has 10 wing-analysis credits a month, spent at 1 per AeroBuildup run and 2–3 per VLM run; after that a run is paid from your credit balance. The counter under the button says what is left, and you are charged for the solver that actually ran. See [pricing](/pricing).

![The constant-lift parameters filled for the freeride foil, target load 95 kg, 4 to 20 m/s in steps of 1, salt water, with the Analyze Wing button ringed.](https://media.foil.tools/docs/shots/wing-analysis-run-button.ebe54e5611ec.webp)

*Constant lift asks for the load and a speed range. The Analyze Wing button is the full width of the panel, under the method notes.*

## The results

The tiles report the starred point, the row with the best L/D, ranked by magnitude within the lift sign the run is about, so a stabiliser at downforce reads correctly. **CL** and **CD** are whole-wing coefficients on S_ref; **L/D max** is the ratio and, in constant-lift mode, the speed it occurs at; **CM** is the pitching moment a stabiliser will have to balance; **Span eff. e** is the Oswald factor fitted to the spanwise loading, 1 for an elliptical loading. S_ref, b, AR and c_ref are projected, the wing seen from above, so they agree with the Design tab. The Checks card re-reads the run against requirements; n/a means not measured here, not failed.

![The Results at best L/D card after the run: CL 0.325, CD 0.0110, L/D max 29.6, CM −0.199, span efficiency 0.820, stall margin 1.178, first stall at 72 % tip-first. The L/D max and First stall tiles are ringed in orange and the Download CSV, Copy chart block and Save analysis buttons in cyan.](https://media.foil.tools/docs/shots/wing-analysis-results.082857c2fe75.webp)

*The run's headline numbers, at the speed of best L/D. The buttons on the right save the run or take the polar away as CSV.*

## Where it stalls first

The spanwise loading is compared at two angles of attack, which splits it into a part set by twist and camber and a part that scales with total lift. Each station's own Clmax comes from NeuralFoil at its local Reynolds number, lower at a tapered tip. Raising the lift until the loading first touches the local Clmax gives the station where stall begins: the **First stall** tile, as a fraction of the half-span, with a verdict. **Stall margin** is the smallest Clmax(y) − Cl(y) along the span at the cruise point, as ΔCl. Root-first is the pattern you want; tip-first wants washout or a higher-lift tip section. A run saved before this readout existed shows Δα to the 3D polar peak instead, and says so.

![The Checks card under the results with its first row, FAIL: stall starts at 72 % span, tip-first, ringed in orange.](https://media.foil.tools/docs/shots/wing-analysis-checks.b8365b28d261.webp)

*Checks, under the results. The freeride wing as drawn fails the first one: it stalls tip-first, and the row names the fix.*

## The plots

An α sweep draws CL, L/D and CM against α and the drag polar. A constant-lift run draws trim α, CD, L/D and the required CL against speed, in the selected unit. The spanwise chart under them is the local lift coefficient along the half-span, from a vortex-lattice solve on either method: solid is the computed Cl(y), dashed the elliptical reference, dotted the local Clmax, and where dotted meets solid is where the wing stalls first.

![The Constant-lift polars card: four plots against speed, trim angle, trimmed drag coefficient, L/D and required lift coefficient, with the L/D plot ringed in orange.](https://media.foil.tools/docs/shots/wing-analysis-plots.7a47cab59b4d.webp)

*The plots, against speed. L/D peaks at 29.6 near 7 m/s on the example.*

## The polar CSV

Download CSV writes the run as `wing_polar.csv`, columns `alpha,CL,CD,CM,L_over_D`, one row per point, coefficients for the whole wing on the S_ref reported. A constant-lift run writes the trimmed α of each speed but no speed column; read the speeds off the α-vs-V chart. [File formats](/docs/file-formats) has the rest.

## Not enough at take-off?

The arrow points back. If the run is infeasible at 4 m/s, or the first stall is at the tip, the fix is upstream: a section with more Clmax at the tip's Reynolds number, found on [Compare](/docs/compare) or with Find → from the Design tab, or more area, less taper or a few degrees of washout on the planform. Change it, re-run, read the same tiles.

## What this hands on

CL, L/D and span efficiency at cruise, the first-stall station, and a CM for the stabiliser to balance, which is what [Trim](/docs/trim) takes next. Save analysis keeps the run and, with the box ticked, the wing, in your Dashboard, where a connected agent can read them too.

## Controls on this page

| Control | What it does | Units |
| --- | --- | --- |
| α sweep / Constant lift | Two kinds of run. α sweep holds one speed and walks angle of attack: the wing's polar. Constant lift holds a load and solves the angle at each speed: what carrying it costs across the envelope. |  |
| Velocity | The one speed an α sweep is run at, in the unit selected above; sent to the solver in m/s. With the fluid it sets the Reynolds number every section polar is read at. | m/s, km/h or knots |
| α min / α max / α step | The angles of attack an α sweep walks, one polar row per step. Past about ±20° the section model is outside the range it is trusted in, so a wider sweep mostly shows where the model stops. | deg |
| Target load | The mass the wing must carry; the solver turns it into lift = mass × g. Negative is downforce, for a stabiliser. At each speed the angle is solved so the lift matches, on the pre-stall branch of the lift curve. | kg |
| V min / V max / V step | The speeds a constant-lift run walks, in the unit selected above, one row per step. A speed too slow to carry the load before the wing stalls is reported as infeasible, not trimmed. | m/s, km/h or knots |
| Speed | Shows and reads every speed on this tab in m/s, km/h or knots: the inputs, the results header and the plot axes. The solver always works in m/s. |  |
| Fluid | Air, salt water or fresh water: density for the dynamic pressure, viscosity for the Reynolds number. ρ = 1025 kg/m³ salt, 999.1 fresh, 1.225 air; the same constants the Design tab uses. |  |
| Method | AeroBuildup walks NeuralFoil section polars across the span with finite-span corrections: 10–20 s, 1 credit, stall included. VLM solves a vortex lattice: 60–240 s, 2–3 credits, attached flow only. |  |
| Spanwise panels | Strips across the half-span for a VLM run: 10 or 16. The solver clamps at 16 and the price follows the clamped count (2 credits up to 10, 3 credits above), so asking for more buys nothing. |  |
| Analyze Wing | Sends the wing as 20 stations sampled from the drawn splines, root and tip sections blended between them. Sign in first. A free monthly allowance of wing-analysis credits is spent before your balance; the counter says what is left. |  |
| CL | Whole-wing lift coefficient at the starred best-L/D point: lift over dynamic pressure and S_ref. In constant-lift mode it is the CL the load demands at that speed. |  |
| CD | Whole-wing drag coefficient at the same point: induced drag from the spanwise loading plus each section's profile drag from NeuralFoil at its local Reynolds number. |  |
| L/D max | The highest lift-to-drag ratio in the run, and the angle (or in constant-lift mode the speed) it occurs at. Ranked by magnitude within the lift sign the run is about, so a downforce surface reads correctly. |  |
| CM | Pitching-moment coefficient of the wing at the starred point, nose-up positive. What a stabiliser has to balance; Trim does that sum with the CG where you put it. |  |
| Span eff. e | Oswald span efficiency: 1 for an elliptical loading, less for anything else, fitted to the spanwise loading with a Glauert series. Induced drag scales with 1/e. |  |
| Stall margin | Headroom before the first station stalls at the cruise point: the smallest Clmax(y) − Cl(y) along the span, as ΔCl. A run saved before this existed shows Δα to the 3D polar peak instead, and says so. |  |
| First stall | Where the wing stalls first, as a fraction of the half-span from the root, with a verdict. Root-first is wanted; tip-first means washout or a higher-lift tip section; mid is inboard of 70% but not at the root. |  |
| S_ref · b · AR · c_ref | What the coefficients are referenced to: planform area, span, aspect ratio and reference chord, all projected (the wing seen from above) so they agree with the Design tab's derived figures. |  |
| Checks | A requirements re-check on the run, pass / warn / fail / n/a: where it stalls first, the tip Reynolds number against the library's floor, the cruise pitching moment. n/a means not measured here, not failed. |  |
| Polars | Four charts of the run. α sweep: CL, L/D and CM against α, and the drag polar. Constant lift: trim α, CD, L/D and the required CL against speed, in the selected unit; an × marks a speed the wing cannot carry the load at. |  |
| Spanwise loading | Local lift coefficient along the half-span at the reference angle, from a vortex-lattice solve on either method. Dashed is the elliptical ideal; dotted is each station's own Clmax at its local Re. |  |
| Download CSV | Downloads the run as wing_polar.csv, columns alpha, CL, CD, CM, L_over_D, one row per point. A constant-lift run writes the trimmed α of each speed but no speed column; read the speeds off the α-vs-V chart. |  |
| Save analysis | Saves the run to your Dashboard under Analyses, and with the box ticked the wing it was run on as a design, so you can reopen the geometry and keep iterating. A connected agent can read both. |  |
