Delta canard jet
Mach 4 at α 2° · 0 m · 18 November 2026
One compressible CFD solution with SU2 8.5.0: 0.99 million points, wing or aircraft, exploratory
- Mach number
- 4.000
- Angle of attack
- 2.00°
- Speed
- 1,361 m/s
- Dynamic pressure
- 1,135.02 kPa
- Air
- 101.3 kPa · 288 K
- Stagnation temperature
- 1,146 K
- Reynolds number
- 56.38 × 10⁶
- Reference area
- 0.554 m²
1Summary
- L/D 1.01. At Mach 4 and α 2°, Delta canard jet makes CL 0.0356 against CD 0.03514. 12% of the drag is skin friction; the rest is pressure drag, the shocks' wave drag with any base drag.
- Converged by the loads. The density residual fell 4.1 orders in 5,000 iterations of the last pass; over its last 500 the axial force swung 0.042%, the normal force 0.027% and the pitching moment 0.047% of their means.
- Exploratory. No validation case covers this geometry at this Mach number: the numbers are the solver's, with no case like this one checked against a measurement.
- Skin up to 1094 K. In steady flight a skin of emissivity 0.8 reaches 821 °C at its hottest, where the heat the air brings equals what it radiates. By area, 2% needs nickel alloy (to 650 °C), 98% needs a superalloy, ceramic or thermal protection: typical long-exposure limits; a material's own data decides.
- Boom 82.8–84.4 dB PL. Flown at 15.0 km, its boom reaches the ground under the track at 7.1–8.2 Pa (0.15–0.17 psf), 24–26 ms long. The run was not meshed for the boom, so its cells far out smear the waves.
2The case and the method
Delta canard jet was solved from delta-canard-jet.step as one steady, compressible RANS case with SU2 8.5.0: the free stream at Mach 4 and α 2°, the standard atmosphere at 0 m (p∞ 101.33 kPa, T∞ 288.2 K, ρ∞ 1.2250 kg/m³, 1,361 m/s), thermally perfect air (its γ falls as it heats), the walls held at 288 K (a cold wall, whose heat flux gives the skin's temperature below).
It runs in two passes: the first from the free stream on a coarser mesh, the second on a surface remeshed where the first found shocks, from the first's solution. The final mesh has 0.99 million points: 1.31 million prisms in the wall's layers and 2.04 million tetrahedra, 25,114 points on the walls.
Coefficients use the reference area 0.554 m² and length 0.605 m on q∞ = 1,135.02 kPa; moments are taken about (0.000, 0.000, 0.000) m in SU2's axes (x downstream, y to the right, z up). Lift is across the free stream, drag along it. The Reynolds number on the reference length is 56.38 million.
3What it can be trusted for
- ExploratoryNo validation case covers this geometry at this Mach number. Thermally perfect air (NASA Glenn polynomials), verified against exact thermally perfect shocks (step 4.6). Its numbers are the solver's, with no case like it checked against a measurement.
- Converged by the loadsThe residual fell 4.1 orders; the loads swing at most 0.047% over the last 500 iterations.
4The surface
The walls' own values from the solution: the pressure, the temperature, the friction and the size of the first cells, on the same body the solver meshed.
5The shock waves
The nose shock is not checked against cone theory. The nose is not round: its half-angle runs from 11.2° to 21.1° round it, so cone theory does not apply.
6The flow on planes
Planes cut through the solver's own cells, each filled from the values at its cells' corners. The schlieren on the first page is the first of them.
7Sonic boom on the ground
The run's near field (the pressure over the free stream's on lines under the flight path, beside the body) carried down through the standard atmosphere from 15.0 km at Mach 4.00, taken to hold there as it did in the run at 0.0 km (same Mach and attitude, scaled by the air pressure up there): still air, the workshops' standard humidity, ground reflection 1.9. It is carried by the augmented Burgers equation along each ray (nonlinear steepening, absorption, oxygen and nitrogen relaxation) and rated by Stevens' Perceived Level, as NASA's sBOOM does; on the AIAA workshop's axisymmetric body this matches sBOOM within 0.3 dB.
The lines differ by 1.6 dB: the solution's cells grow away from the body and smear the waves, so take the range as the answer's uncertainty. A run meshed for the sonic boom would narrow it. The front and rear shocks arrive 18–20 ms apart, closer than the ear's 70 ms: one bang, so its loudness keeps both shocks' energy.
8Loads
| Total | Pressure | Friction | |
|---|---|---|---|
| Lift, CL | 0.03557 | 0.03564 | -0.00008 |
| Drag, CD | 0.03514 | 0.03098 | 0.00416 |
| Pitching moment, Cm | 0.01887 | 0.01876 | 0.00011 |
| Side force, CY | 0.00008 | 0.00008 | 0.00000 |
| Rolling moment, Cl | -0.00001 | -0.00001 | 0.00000 |
| Yawing moment, Cn | 0.00001 | 0.00001 | 0.00000 |
Force (21,304.5, 52.8, 23,122.3) N and moment (-4.6, 7,180.2, 4.6) N·m in SU2's axes.
9How far it converged
The rule: the density residual five orders below its peak, or each body-axis load within 0.1% (its range over its mean) over the last 500 iterations. The residual fell 4.1 orders in 5,000 iterations.
| Load | Last | Mean | Swing | Drift | Cycle |
|---|---|---|---|---|---|
| CA | 0.03388 | 0.03388 | 0.042% | -0.004% | 110 it. |
| CN | 0.03677 | 0.03677 | 0.027% | 0.004% | 60 it. |
| Cm | 0.01887 | 0.01887 | 0.047% | -0.008% | 116 it. |
AComputation
| Pass | Points | Iterations | Wall time | Peak memory |
|---|---|---|---|---|
| 1 | 0.74 M | 3,000 | 57 min | 6.9 GB |
| 2 | 0.99 M | 5,000 | 2 h 20 min | 7.8 GB |
delta-canard-jet · Mach 4 · α 2°: 3 h 19 min in all on 5 cores, with the surface, the meshes and the result files. Generated by Deneb Aero from the run's own result files; nothing in it is solved again.