Size a chamber,
not a guess.

CSTAR is a Windows desktop suite for preliminary liquid rocket engine design. One solve carries thermochemistry, nozzle contour, conjugate heat transfer and structural margins — and every model states what it was checked against.

Windows 10/11 · offline · no account

CSTAR v1.0 STANDARD
Vacuum thrust 54.06 kN
Specific impulse 326.9 s
Exit Mach 3.67
Peak heat flux 50.1 MW/m²
Nozzle flowfield Gas dynamics Heat flux Structural
Mach number through the example thrust chamber, from subsonic in the chamber to Mach 3.67 at the exit.
Example engine Methalox · O/F 3.0 · 60 bar · ε 16
Real output from the example engine that ships with the application.

What every number was checked against.

Three release gates. A build that fails any one does not ship. The fault is named, the tolerance is published, and the worst operating point is listed — not just the average.

Read the full validation report
Worst deviation across each gate, from the shipped test suite.
Model and reference Worst deviation Points Result
Equilibrium thermochemistryNASA CEA via rocketcea 2.923% c* · 2.686 K Tc 145 Pass
Coolant propertiesCoolProp, NIST-grade EOS 4.214% cp · 2.005% ρ 6,000 Pass
Nozzle contour, planar MoCExact Prandtl–Meyer theory 0.0031% Mach · 0.0284% mass 16 Pass

The CEA sweep spans 10–250 bar and area ratios of 4 to 150. The 6,000 CoolProp states are sampled at cell centres. The method-of-characteristics sweep covers 16 configurations from γ 1.15 to 1.40.

The whole chamber, in one solve.

Every plot below is real output from the example engine that ships with the application. Same run, four views. The numbers do not change between them.

Nozzle cross-section shaded by Mach number, from the chamber through the throat to the exit.

Nozzle flowfield

Mach number through the example engine. Chamber, throat and exit are marked. The supersonic contour is a Rao parabolic bell, not a cartoon nozzle.

Contour
Rao parabolic bell
Planar MoC
Validated, 16 configurations
Axisymmetric
Withheld — did not pass mass conservation
Pressure, temperature, Mach number and velocity along the chamber axis.

Gas dynamics

Pressure, temperature, Mach and velocity along the axis, from a two-branch area–Mach solver with self-consistent variable γ.

Model
1D, variable γ
Propellants
Six presets
Exit Mach
3.67
Heat flux and hot-wall temperature distributions along the chamber axis.

Heat flux and wall temperature

Bartz on real gas properties, Gnielinski with roughness-aware friction, and Sieder–Tate at the coolant-side wall. Channel lands are solved as fins.

Coolants
Five, temperature- and pressure-dependent
Peak flux
50.1 MW/m²
Max wall
482 K
von Mises stress and safety factor along the chamber axis.

Structural margins

von Mises stress and safety factor along the example engine. The liner is a thin wall spanning the cooling-channel lands, loaded by pressure and the through-thickness gradient.

Yield SF
1.28
Melting SF
2.82
Method
Closed-form thin-wall

What CSTAR is not.

You are an engineer. You would find these out anyway, and you would trust the rest less for having had to dig. So here they are.

  • A preliminary design aid. Not certified by any aviation, space or standards authority. Everything it outputs needs independent verification before it touches hardware.
  • Heat transfer inherits Bartz. The correlation carries ±20–30% scatter against hot-fire data. That is a property of the method, not of the implementation.
  • Gas radiation is a grey-gas approximation. Real H₂O/CO₂ emission is band-structured. The magnitude lands in the right range; the model is crude, and the coefficients are exposed.
  • Structural analysis is thin-wall, not FEA. No land fillets, no creep, no bulging mode. The fatigue number is a screening estimate with generic constants.
  • The nozzle contour is a Rao approximation. A verified planar method of characteristics ships alongside it. The axisymmetric branch is not validated and refuses to run.
  • Flow is one-dimensional. No shocks, no separation prediction, no film cooling, no combustion instability.
A rocket engine firing on a test stand at NASA Lewis in 1973, with a glowing exhaust plume.
Rocket Engine Test Facility, NASA Lewis, 1973. Ground test remains mandatory. NASA, public domain, via Wikimedia Commons.

One licence. One machine.

The demo is the same executable, not a crippled build. It runs your engine and shows the flow solution. Heat transfer, cooling and structural analysis are what you buy.

Free Demo edition · evaluation only
$149 one-time Standard edition · one machine · offline
What the demo and the standard edition include
What you get Free Demo edition · evaluation only $149 one-time Standard edition · one machine · offline
Nozzle geometry & flowfieldIncludedIncluded
Gas dynamics — M, P, T, vIncludedIncluded
Thrust, Isp, mass flowIncludedIncluded
Runs on your own engineIncludedIncluded
Heat flux & wall temperaturesNot includedIncluded
Regenerative cooling circuitNot includedIncluded
Structural margins & fatigueNot includedIncluded
CSV / CAD / Excel / PDF exportNot includedIncluded
Watermark-free outputNot includedIncluded
Full documentation & validation reportNot includedIncluded
Buy CSTAR — $149 Download the demo Checkout in crypto · keys issued by hand

After you buy

  1. Download and run CSTAR. It opens in DEMO; press Activate in the toolbar.
  2. Send your Hardware ID. Open the activation page and send the ID it shows. It is derived from your Windows installation and processor, and contains no personal data.
  3. Get your key by email. Paste it in and STANDARD unlocks offline. Keys are issued by hand, so allow a few hours.

Questions worth asking.

If a question is not answered here, write to cstar.propulsion@gmail.com. A person replies, usually the same one who built it.

Can I run my own engine in the demo, or only a fixed sample?

Your own. Every input is editable in DEMO and the solver runs your geometry, propellant and chamber conditions. What DEMO withholds is the thermal, cooling and structural results — an unlicensed run does not compute them for delivery at all.

Is it really validated against NASA CEA, or is that marketing?

Really. The comparison runs through the official CEA code via the rocketcea wrapper, across 145 operating points spanning 10–250 bar chamber pressure and area ratios of 4 to 150. The test suite ships with the application and prints the worst-case deviation for every quantity against a documented tolerance.

Does it need an internet connection?

No — not to run, not to activate, not ever. There is no activation server, no telemetry and no phone-home. The licence key is verified locally against a public key compiled into the application. Air-gapped machines work fine.

What happens if I change my hardware?

The Hardware ID is derived from your Windows installation GUID, processor and architecture. It deliberately excludes network adapter addresses, so VPNs, docking stations and virtual machines do not disturb it. Reinstalling Windows or moving to a different machine does change it — contact us and we re-issue.

Which propellants and coolants are supported?

Propellants: methalox, kerolox, hydrolox, ethanol, methanol and propane, all with LOX. Coolants with full temperature- and pressure-dependent properties: methane, hydrogen, oxygen, RP-1 and ethanol. Any other fluid can be entered as constant properties.

Can I use it commercially?

Yes. A STANDARD licence covers commercial use on one machine. Need several seats? Get in touch and we will sort it out.

Run it on your engine.

The demo takes a minute and runs your real geometry. If the flow solution does not convince you, nothing on this page should.

Licence
$149 one-time, one machine
Platform
Windows 10/11, 64-bit
Activation
Offline, by email