Deep-throttling liquid rocket engine

A rocket engine that can land its own booster has to do something ordinary engines never attempt: run stably at a small fraction of full power. Deep throttling — dialling thrust down and back up on demand — is what lets a first stage slow itself, hover, and set down for reuse. Only six nations operate cryogenic rocket engines at all: the United States, Russia, Japan, France, China, and India.

The difficulty is not raw power but control at the low end. Throttling deep means minimising the pressure drop across the injector while avoiding coupling between the propellant feed system and the thrust chamber — the coupling that produces combustion instability. Heat flux across the chamber swings dramatically as thrust falls, so the cooling jacket must stay effective across the whole range. Landing further demands that the engine restart in flight, more than once, for the deorbit, descent, and touchdown burns.

India has the foundational pieces and is now assembling them for reuse. The CE-20 cryogenic engine operates across 180–220 kN, throttling between nominal and uprated thrust; a 670-second hot test ran at 22 tonnes. In March 2026 ISRO demonstrated boot-strap start-up on the CE-20 for in-flight re-start. The harder target is the LOX-methane path. On 27 January 2026, ISRO hot-tested its first LOX-methane sub-scale thrust chamber — a single-element injector, 3D-printed, reaching 56 bar chamber pressure. The LM110 methane engine is being designed for a 60–110% throttling range and 20 reusable cycles. VSSC has developed a triethylaluminium-triethylboron start fuel ampule for the SE-2000 semi-cryogenic engine and is running injector-level ignition tests for staged-combustion characterisation.

The gap persists because the enabling capabilities sit at different levels of maturity. Cryogenic propellant handling is competitive and regenerative cooling is in production, but the injector design for throttleable engines, the feed-system regulation, and the restart and ignition hardware remain emerging — and these are precisely the critical items that gate a landing engine. Sub-scale methane testing only began in January 2026.

What it would take is a full-scale climb from those sub-scale tests to a flight engine on a fixed schedule. The NGLV programme, approved in September 2024 at ₹8,240 crore over 96 months, specifies LME-1100 engines of 1100 kN thrust with deep throttling for booster recovery, and targets first flight from 2030. Supporting it, ISRO is building an 8-tonne-per-day liquid methane plant at Mahendragiri and a spark torch igniter for reliable multi-restart. The pieces exist; the task is to make them one landing engine.