Pintle Engine PoC
LASC Lander Challenge · engine development article

A throttleable rocket engine built on a lathe, flown from 100 m away

Gaseous oxygen and ethanol meet at a brass pintle inside a graphite liner. Every valve is driven by the pad computer, every step is gated by a measurement, and the whole stand carries over to the lander.

Same engine, tank and valves in both phases. Only the oxygen regulator and the pintle tip change.

240NThrust at sea level
9.4barChamber pressure
128g/sTotal propellant flow
1.3O/FFuel-rich mixture ratio
40%Throttle floor, same as the lander
10sBurn length per firing
Live simulator

Fire it from the control table

Arm the stand, fire, and drag the throttle during the burn. Flow animates in each line, closed valves turn solid, and the sequencer aborts on its own when a check fails. Click any component in the schematic for a picture and what it does, or inject a fault to watch the safety logic work. The model is simplified, with the numbers from the design point.

Safe T − Click any component to see what it is. While safe or armed, click a valve a second time to open or close it by hand (cold-flow mode).
Oxygen Ethanol Nitrogen Propane ■ solid valve = closed yellow actuator = energized
100%
1×
Chamber pressure, bar Thrust, N ÷ 20 Sleeve, °C ÷ 10
Inject a fault
Event log
Hardware

Eleven parts, all lathe work

Drag to orbit, scroll or pinch to zoom. Pick a part on the model or in the list and the camera flies to it. The exploded view follows the assembly order, so every part leaves the way it goes in; Play assembly puts it back together step by step. Dimensions follow the 400 N design, and the SolidWorks models can replace these shapes later.

0%
    Design choices

    Why it is built this way

    Every material here is something the team can buy and machine today: 6063 aluminium, brass, stainless, graphite and copper refrigeration tube. Each choice below follows from that.

    Chamber pressure is set by the regulator

    At 15 bar the 6063 sleeve sees about 10 MPa of hoop stress, under a tenth of its yield. The real limit is the oxygen regulator: the venturi needs about 25% more pressure upstream than the injector, and the valves take 2 to 3 bar more. Phase one runs a 15 bar regulator at 9.5 bar chamber pressure; phase two adds a 24 bar regulator for the full 15 bar.

    10 MPasleeve hoop stress at 15 bar, against 110 MPa yield

    A graphite liner takes the flame

    The gas is near 2,900 K. Aluminium melts at 660 °C and burns in oxygen, so a bare 6063 chamber would fail in the first second. One piece of graphite forms the chamber, throat and nozzle, and it slides out for inspection after every firing.

    Ø16.0 mmthroat, checked after each firing for erosion

    Ceramic paper keeps the sleeve cool

    Graphite heats through in about 2 s. Three millimetres of ceramic paper slows the heat into the sleeve, centres the brittle liner and absorbs the difference in expansion: aluminium grows about six times more per degree than graphite.

    ~20 kW/m²heat into the sleeve through the paper

    The sonic venturi is the flowmeter

    While the gas in its throat is at the speed of sound, oxygen flow depends only on the inlet pressure and temperature. One transducer and one thermocouple make it a flowmeter, and the choked throat stops chamber pressure swings from travelling back up the line.

    Ø5.4 mmventuri throat, choked below 85% of inlet pressure

    Two throttle actuators, on purpose

    Oxygen is throttled by a 1/2 in ball valve on a 150 kg·cm servo, because no small needle valve passes 290 Nm³/h. Ethanol is throttled by a needle valve on a NEMA 17 stepper, which gives fine steps on the liquid side. Flying one of each shows which actuator the lander should use.

    3–6 N·mball-valve breakaway torque at 24 bar

    Tie rods with disc springs

    Four M8 A2-70 rods clamp the head, sleeve and retainer plate. Heat cycles loosen nuts more than vibration does, so disc springs keep the clamp load steady as parts grow, all-metal lock nuts sit at the hot end, and paint stripes show any movement.

    2.2×rod safety factor at the 30 bar worst case
    Safety

    It fails safe without software

    Cutting power is always the safe move. The E-stop loop drops solenoid power through a relay: main valves close and nitrogen purges open, with no code involved. High-pressure oxygen is only ever opened by hand.

    DeviceWith power lostWhy
    SV-1, SV-2 main valvesCloseNormally-closed solenoids stop both propellants
    SV-3, SV-4 igniterCloseNo torch gas without a command
    PV-1, PV-2 purgesOpenNormally-open solenoids flush both manifolds with nitrogen
    RV-1, RV-2 reliefsOpen at 30 barSpring-loaded, no power needed
    CV-1 to CV-4Block backflowFuel cannot reach the oxygen side, or the reverse
    Automatic abort whenCheck
    No torch flame 0.3 s after the sparkTC-5
    Chamber below a third of full pressure 0.3 s after fuel opensPT-5
    Chamber pressure off its setpoint by more than 20%PT-5
    Sleeve above 120 °CTC-1
    Venturi no longer chokedPT-1, PT-2
    Fuel tank pressure dropping toward the chamberPT-3
    Telemetry link lost for 200 ms, or E-stopHardware
    Test campaign

    Twelve gates before the showcase firing

    Each step has a written procedure and a go/no-go, and the next starts only after the one before it passed. Nitrogen stands in for oxygen in every cold flow. Hot steps need the mentor present.

    1. Clean and leak-checkOxygen-clean the GOX side, leak-check everything with nitrogen at 25 bar
    2. Hydrostatic proofFuel tank and engine with the throat plugged, water at 45 bar
    3. Electronics bench testEvery abort path, E-stop, power loss and link loss, with no gas
    4. Fuel cold flowWater through the injector: flow against pressure drop, spray on camera
    5. Oxygen-side cold flowNitrogen through the venturi and annulus, then both sides together
    6. Throttle mappingMap both throttle valves and tune the control loops on cold flows
    7. Igniter aloneAt least 20 lights, timing command to flame
    8. Ignition pops0.5 s, then 1 s at 60% flow
    9. Short firings3 s and 5 s at full flow, throat measured after each
    10. Steady 10 s firingsc* efficiency from chamber pressure and measured flows
    11. Throttle steps100% to 60% and back, open-loop, then closed-loop on chamber pressure
    12. Showcase firingFilmed for sponsors and the milestone report
    Road to the lander

    Same chamber pressure, half the thrust

    The stand, sequencer, control loops, pintle fuel side, liner construction and procedures carry over unchanged. The lander engine swaps gaseous oxygen for liquid N2O and doubles the flow.

    Proof of conceptLander engine
    PropellantsGOX and ethanolLiquid N2O and ethanol
    Thrust240 N → 400 N900 N
    Chamber pressure9.5 → 15 bar15 bar
    Mixture ratio1.33.5
    Total flow128 → 204 g/s496 g/s
    Throat, chamberØ16, Ø40 mmØ25, Ø56 mm
    Burn10 s20 s
    ThrottleTo 60%, 40% stretchTo 40%, held within ±5%
    Oxidizer feedRegulated gasSelf-pressurized liquid, 50–63 bar

    Phase-one budget

    About R$13,100 in total, R$11,600 with borrowed safety gear. Estimates for late 2026, not quotes.