A Rocket Engine Simulator That Shows Its Working

Meco is a rocket engine design simulator. You set the geometry, you fire the engine, and you read what comes back on live graphs. Every number you change answers back. Every node is inspectable. Nothing important happens in a box you cannot open.

I build Meco on my own, and the engine model behind it is written up in a preprint, DOI and all, so you can check my work: 10.5281/zenodo.21864769.

Get Meco on SteamEarly Access · Windows and macOS · Linux coming soon

As an engineering student I really enjoyed it, it covers some complex topics that really make me think and learn. I'd definitely recommend it!

heimdall, on Steam

Five numbers and you have a thrust chamber

Give the Thrust Chamber its throat radius, expansion ratio, contraction ratio, chamber length and section count. It builds the hardware from there: a Rao thrust-optimized bell, the regenerative cooling channel layout, and the chamber solids in 3D. Change one headline number and the hardware rebuilds while you watch.

Push into geometry that would not work and the editor tells you where you left the Rao chart, before you burn an afternoon on it.

Cooling channels you never had to draw

Set the channel count, the land width and the channel height. The assembly tiles them around the chamber wall, four hundred of them if that is what you asked for, and the regenerative cooling solves with the rest of the engine.

Thrust comes from the flow field

Meco integrates the method-of-characteristics flow field through your bell, at the mass flow your simulation actually delivered. There is no thrust coefficient looked up from a table.

Push the expansion ratio too far at sea level and the flow lets go of the wall. The field view grays out past the separation point, and thrust is counted only where the flow is still pushing.

A turbopump that sizes itself

Pick one of four shaft layouts, from a single shaft to a geared direct drive, and the shafts and gears come out sized for the layout you chose. Shafts and gear meshes then cost you real torque, and pumps report the available NPSH beside their performance curves.

Then you wire the whole thing yourself

Assemblies are parts of an engine, not the engine. On the canvas you drop in components, wire them together, and run. Domes, injectors, valves, cooling joints, gas generator, turbine. It is your plumbing, and it fails in your handwriting.

The start is a schedule too. One master ramp runs 0 to 1, and each valve fires as it crosses its own threshold, with its own level, width and delay. Real engines go in order: spin up, crack the valves, light the gas generator, trim.

Fourteen missions, and a senior engineer who has seen worse

Mission 0 drops you at Miku Aoyama's workstation with her Vulcain 1 model open and a clear instruction not to touch it. Thirteen missions follow, across five chapters, from your first model to a complete gas-generator cycle.

Your first designs probably will not hit their targets. That is the point. In a mission, when a build misses, a "Need a hint?" diagnostic names the specific problem and walks you to the fix one item at a time: run the simulation at least once, your mixture ratio is still stoichiometric so dial the gas generator valve richer, that component name looks like a typo. In free build there is no tutor. There it is you, the graphs and the hardware rebuilding as you type.

Chapter V: recreate the Merlin 1D

Chapter V opens with the engine that flies Falcon 9. You place the Turbopump and link its slots, give the pre-placed Thrust Chamber its headline geometry, author the startup master and two of its gates, then tune the gas generator until the whole loop holds at steady state. The debrief puts your numbers next to the published Merlin figures.

SpaceX has never published Merlin's internal geometry, so every dimension in there is derived or back-solved. Take it as inspiration rather than a digital twin.

Once you have flown it, the Merlin 1D is a starting template in free build, yours to keep tinkering with.

Checked against a real engine

Meco ships with the Vulcain 1, the engine that flew Ariane 5, built from its published figures. Here is what the simulator produces against them, including the row that looks wrong.

That last row is the gas generator, and it is the reason the distinction matters. Nozzle-flow Isp counts only what leaves the bell. Engine Isp counts the propellant the gas generator burns as well, which is what a real datasheet quotes. Meco reports both, labelled, and its engine Isp on total flow lands within 1% of the published 431 seconds.

The model behind it is written up as a preprint. Read what it claims carefully, because it is careful: the pump, feed-pressure, cooling-channel and thrust-chamber models are anchored to Vulcain 1's published constants and the RS-25 throat, and the Merlin 1D–class case study measures its own residuals rather than fitting itself to flight data. Mass conservation through the nozzle field closes to 0.92%. The paper calls that verification, not validation against flight data, and it is right to.

What it is, honestly

Meco is in Early Access and playable today. It is not a casual space game. It is an engineering sim with a real learning curve, and it is at its best when you want to know why a cycle works.

Playable now: Chapters I to IV complete, and Chapter V open with the Merlin 1D. Thrust Chamber and Turbopump assemblies in live 3D. The startup schedule. Engine Isp, the number real datasheets quote, with the gas generator's propellant counted against it. The built-in Vulcain 1 sits on its published numbers: 1140 kN of vacuum thrust, and engine Isp within 1% of 431 seconds. Undo and redo across the editor. Display Units, so you can read pressure in bar instead of pascals.

Building next: the rest of Chapter V, the A-7 Redstone and the Vulcain 1. The Injector Head assembly in coaxial, pintle and impinging configurations. Expander-cycle components. The whole engine joined into one machine in 3D. Component mass, for thrust-to-weight numbers. Thermal and load stress limits.

I intend to finish the roadmap in 2027. I build Meco solo, so every purchase funds the work directly.

Why I build an accessible rocket engine simulator

To be a spacefaring civilization we need a lot of people who can build and maintain rocket engines, in every colony. That is a teaching problem before it is a hardware problem.

That is also why I set up the Open Space Launch Foundation. It is a non-profit, and its charity registration is under way. The intention is that it eventually holds LorenLib, the open-source core Meco is built on, so the simulation engine outlives any one product.

Neither Meco nor LorenLib has moved across yet, and neither will until the commercial side is steady enough to carry it. That order is deliberate. The goal has not changed: keep Meco commercially viable enough to sustain me, and eventually a small team.

• Meco is made by Loren Aerospace Ltd, which is me, Dannie Sim.
• Open Space Launch Foundation

Come hang out

Come in and join the folks who are just as into rocket engines as you are.