Turbofan Bypass Ratio

What bypass ratio is, in one line

Look at a modern airliner engine head-on. The whole thing is mostly fan. The dark core in the centre, where compressors and combustors and turbines live, is a small dot in the middle of an enormous spinning disc of fan blades. That visual disproportion is the entire idea of a high-bypass turbofan: most of the air the engine touches never enters the core at all. It is accelerated by the fan, channeled around the outside of the core through the bypass duct, and ejected out the rear at modest velocity. Only a fraction goes through the compressors, the combustor, and the turbine — the core.

Bypass ratio (BPR) is the ratio of mass flow through the bypass duct to mass flow through the core:

BPR = ṁ_bypass / ṁ_core

A modern Boeing 777X with the GE9X engine has BPR ≈ 9.9: about ten times as much air goes around the core as through it. A 1970s 747 with the original Pratt & Whitney JT9D had BPR ≈ 5. A 1960s 707 with the original Pratt & Whitney JT3D-3B had BPR ≈ 1.4. A Concorde, with the Olympus 593, had no bypass duct at all — BPR 0, a pure turbojet. An F-22 fighter with the F119 has BPR ≈ 0.3.

That single number captures the entire propulsive design intent of the engine. It is one of those rare engineering quantities where one parameter determines the family — much like 'Reynolds number' for fluid mechanics or 'compression ratio' for piston engines.

Why higher BPR is more efficient at subsonic speeds

Net thrust of a jet engine is the rate of momentum change of the air it processes:

F_net = ṁ (V_jet − V_flight)

where ṁ is total mass flow, V_jet is jet exit velocity, V_flight is flight speed. The kinetic energy added to the airflow is:

KE_added = ½ ṁ (V_jet² − V_flight²)

Propulsive efficiency η_p is the ratio of useful work done on the aircraft (thrust × V_flight) to total kinetic energy added to the flow:

η_p = 2 V_flight / (V_flight + V_jet)

That formula has a crucial implication: η_p is maximised when V_jet equals V_flight. In practice you need V_jet slightly above V_flight to make any thrust at all, but the closer the two are, the less kinetic energy is wasted in the jet wake. For the same total thrust F_net, you can either move a lot of air a little (high ṁ, low ΔV — high-bypass) or a little air a lot (low ṁ, high ΔV — low-bypass). The first is more efficient at subsonic speeds.

At Mach 0.85 (V_flight ≈ 250 m/s), the propulsive efficiency of:

• A high-bypass turbofan (V_jet ≈ 300 m/s): η_p ≈ 91 %.
• A medium-bypass turbofan (V_jet ≈ 400 m/s): η_p ≈ 77 %.
• A turbojet (V_jet ≈ 700 m/s): η_p ≈ 53 %.

That gap is the entire economic basis for high-bypass aviation. For an airliner cruising at Mach 0.85, the high-bypass engine burns roughly 1.5–1.7 times less fuel per unit thrust than a comparable turbojet. Over a 12-hour transatlantic flight that translates to tens of tonnes of fuel saved per departure.

Why low-bypass for supersonic

The same formula explains why fighter engines look completely different. At Mach 2 (V_flight ≈ 600 m/s), propulsive efficiency demands V_jet ≈ 600+ m/s. You can't get there with a BPR-12 fan stage — the fan compressors the air just enough to mix with bypass at low velocity. To produce a high-velocity jet, you need a low-bypass engine where most of the air goes through the core, gets compressed 30:1, heated past 1700 °C in the combustor, and accelerated through the turbine and nozzle.

Two additional constraints push fighters toward low-bypass:

• Packaging. A modern fighter has to fit its engine inside a slim supersonic fuselage. The GE9X is 3.43 m diameter; the F-22's F119 is 1.18 m. A BPR-12 engine cannot be stuffed inside a Raptor — physically, the fan is bigger than the airplane's cross-section.
• Afterburning. Military aircraft routinely use afterburners (raw fuel dumped into the core exhaust before the nozzle, igniting in residual oxygen) to double thrust briefly. Afterburning requires hot, oxygen-rich exhaust — exactly what a low-bypass core produces, and exactly what a high-bypass engine's mostly-cold mixed exhaust does not.

BPR	Example engine	Aircraft	Typical use
0 (turbojet)	Rolls-Royce Olympus 593	Concorde	Pure supersonic transport (extinct in civil)
0.3	Pratt & Whitney F119	F-22 Raptor	Mach 2+ air superiority fighter
0.4	F135	F-35 Lightning II	Multi-role stealth fighter
0.6	F110-GE-129	F-16 Fighting Falcon	Multi-role fighter
1.4	Pratt & Whitney JT3D	Boeing 707, DC-8	Early jet airliner
3 to 4	Rolls-Royce Adour	BAe Hawk, Jaguar	Trainer / light attack
5 to 6	CFM56-5B	A320 family (original)	Single-aisle airliner
8	GE GEnx-1B	Boeing 787	Twin-aisle airliner
9	Rolls-Royce Trent XWB	A350 XWB	Twin-aisle airliner
9.9	GE9X	Boeing 777X	Widebody twin (current production record)
11	CFM LEAP-1A	A320neo	Re-engined single-aisle
12.5	Pratt & Whitney PW1100G	A320neo, A220	Geared turbofan, civil production record
~20 (planned)	Rolls-Royce UltraFan, GE/Safran RISE	Future narrow-body	Open-rotor or variable-pitch ducted fan

Worked example — how much fuel does BPR save?

Take a 1970 Boeing 747-100 with four JT9D engines (BPR 5.2, total thrust 4 × 209 kN = 836 kN). Compare to a 2020 Boeing 777-300ER with two GE90-115B engines (BPR 7.7, total thrust 2 × 512 kN = 1024 kN). Cruise mission: 12,000 km at Mach 0.84.

747-100 JT9D (1970, BPR 5.2):
  TSFC (thrust-specific fuel consumption, cruise) ≈ 17.5 g/(kN·s)
  4 engines × 209 kN × 0.65 cruise throttle  = 543 kN cruise thrust
  Fuel burn   = 543 kN × 17.5 g/(kN·s) = 9.5 kg/s = 34.2 t/hour
  12 hr flight: ~410 t fuel total

777-300ER GE90-115B (2004, BPR 7.7):
  TSFC ≈ 14.0 g/(kN·s) at cruise
  2 engines × 512 kN × 0.30 cruise throttle  = 307 kN cruise thrust
  Fuel burn   = 307 kN × 14.0 g/(kN·s) = 4.3 kg/s = 15.4 t/hour
  12 hr flight: ~185 t fuel total

A320neo PW1100G (2016, BPR 12.5):
  TSFC ≈ 13.0 g/(kN·s) at cruise
  Cruise thrust ~50 kN total (smaller airplane)
  6 hr typical flight: ~12 t fuel total

Saved fuel from BPR 5.2 → BPR 7.7 → BPR 12.5:
  Per available seat km, A320neo burns 35 % less than original A320,
  which burns 25 % less than B707, which burned 60 % more than B747.

Three generations of BPR increase combined with countless other improvements (composite materials, more efficient compressors, higher TIT, blended winglets) cut airliner fuel burn per seat-km by roughly 75 percent between 1960 and 2025. Bypass ratio is the single biggest contributor.

The geared turbofan — and why BPR finally jumped past 10

Through the 1990s and 2000s, civil-aviation BPR climbed steadily from 5 to about 8–9. The next jump — past 10:1 — wasn't done by enlarging the fan further on a conventional architecture, because a problem set in: fan tip speed.

To stay efficient, fan tip speed must remain below about Mach 1.2 (otherwise shock losses kill efficiency and noise spikes). For a 2.5 m diameter fan at 6,000 rpm, tip speed is already at the limit. To grow the fan further while staying under that limit, the fan must spin slower.

But the fan, the low-pressure compressor, and the low-pressure turbine all share a single shaft on a conventional turbofan. The LPT wants to spin fast (10,000+ rpm) to be small and efficient. The fan wants to spin slow (3,300 rpm). The compromise hurts both.

The geared turbofan solves this by putting a planetary gearbox between the fan and the LPT shaft. The Pratt & Whitney PW1000G family (PW1100G on A320neo, PW1500G on A220, PW1700G on E195-E2) uses a single epicyclic stage with a 3:1 ratio. The LPT now runs at its optimal 10,000 rpm; the fan runs at 3,300 rpm; both are happy. The fan grew from 2.0 m on the V2500 it replaced to 2.06 m on the PW1100G — modest in diameter — but the BPR jumped from 4.7 to 12.5, a 2.7× increase in mass flow ratio.

The cost is a ~500 kg gearbox at the heart of the engine, transmitting roughly 30 MW continuous. Designing a gearbox for 20,000 hours of service life at multi-MW power is non-trivial, and the PW1000G has had reliability problems in service (combustor liner cracking on Tier 1 engines, fan disc corrosion in saline environments). These are being progressively solved.

The competing CFM LEAP-1A on A320neo uses a conventional direct-drive architecture with BPR 11 — slightly lower than the GTF, slightly more reliable, slightly different fuel burn. Different design philosophy, similar final efficiency.

Beyond the conventional ducted fan

• Open rotor (no nacelle). Remove the bypass duct entirely. Use a large-diameter unducted variable-pitch fan rotating in free air. Equivalent BPR exceeds 25:1 because the 'fan' is really a propeller, and propellers are propulsively efficient at low speeds. Two flight-tested examples: GE Unducted Fan (UDF, 1986, never produced) and the current GE/Safran RISE program (CFM-derivative, targeted for service entry by 2035). Trade: the noise floor of an unducted fan is harder to suppress than a ducted fan.
• Variable-pitch ducted fan (VPDF). A conventional ducted geometry but with variable-pitch fan blades, similar to a turboprop. Allows the fan to operate efficiently across a wider thrust range. Rolls-Royce UltraFan demonstrator (planned 2026 ground test) targets BPR ~15:1 with variable-pitch and a power gearbox.
• Three-spool architecture (Rolls-Royce). Rolls-Royce's Trent family uses three independent shafts: LP (fan), IP (intermediate compressor), HP (high compressor). Allows the engine to behave 'like a geared turbofan' without a gearbox — each spool optimised independently. Rolls-Royce will combine three-spool with a planetary gearbox on UltraFan, getting both benefits at once.
• Ultra-low-bypass (turbofan/turbojet hybrid). Boom Supersonic's planned Symphony engine on Overture targets BPR ~3.5 for sustained Mach 1.7 cruise — a deliberate compromise between turbojet (efficient supersonic, terrible subsonic) and high-bypass (terrible supersonic, efficient subsonic).
• Distributed propulsion / electric fans. Future aircraft concepts (NASA X-57, multiple eVTOL aircraft) replace one large fan with many small electric fans, distributed along the wing or fuselage. Effective BPR is meaningless because the fans are decoupled from any thermodynamic core. Likely the long-term architecture, but two decades from passenger-aircraft scale.

What changes inside the engine when BPR rises

• Fan grows. A BPR-12 fan is roughly 30 percent larger in diameter than a BPR-6 fan for the same total thrust. Mass flow rate scales with disc area; disc area scales with diameter squared.
• Core shrinks (relative). The high-pressure compressor and combustor get proportionally smaller — they're processing less mass for the same total thrust.
• Bypass duct area grows. At BPR 12 the bypass annular cross-section is roughly an order of magnitude bigger than the core cross-section. The entire nacelle geometry shifts to give that air a clean path.
• Low-pressure turbine grows. The LPT must extract enough work to drive the bigger fan. More stages, more blade diameter, more thrust through the LPT. On a direct-drive ultra-high-BPR engine the LPT becomes a large component in its own right.
• Pylon and nacelle redesign. A 3.5 m diameter engine sitting under a wing changes wing-to-engine geometry, ground clearance, weight distribution, and yaw asymmetry handling on engine-out scenarios. The 737 MAX's MCAS controversy traces directly to the LEAP-1B engine being too large to fit under the unmodified 737 wing without changing the airframe's pitch behavior.
• Inlet design. Larger inlet area to capture mass flow at low speed. Modern high-BPR inlets are aerodynamically tuned to handle takeoff (high mass flow, large angles of attack) and cruise (low mass flow, near-axial) simultaneously without flow separation.

Failure modes specific to high-BPR engines

• Fan-blade-off event. The single largest catastrophic risk. A fan blade weighing 25 kg detaching at 3,300 rpm imparts enormous unbalanced load on the engine; the FBO certification test (FAA / EASA mandatory) requires the engine to contain a blade-off without rupturing the case. Modern fan cases use Kevlar wraps to absorb the energy. The 1989 United 232 DC-10 crash was caused by an uncontained fan-disc failure; the entire industry's containment standards were rewritten afterwards.
• Bird strike (fan ingestion). Large fans swallow large birds — and flocks. Modern fan blade design must survive a 1.8 kg single-bird ingestion and 16-bird flock ingestion without catastrophic damage. Composite fan blades (GE90, GE9X) have made huge strides; titanium fan blades on older engines deform more readily.
• Fan flutter. At certain flight conditions the long thin fan blades can resonate at high amplitude. Modern fans use shrouded designs (partial mid-blade shrouds) or composite blades with intrinsic damping to mitigate.
• Geared turbofan gearbox. Reliability problem unique to GTF architecture. The 500 kg planetary gearbox sees enormous loads and must last 20,000+ hours. Pratt & Whitney's PW1000G family had production rejection issues (2017–2019) and accelerated wear in saline environments. Largely solved in current production but worth knowing about.
• Ice ingestion / ice shedding. Large-area fan can accumulate ice on inlet and fan blades. Sudden shedding can imbalance the rotor. Modern engines have heated inlets and routine ice-shedding modes that release ice before it builds up too thick.

Common pitfalls when reasoning about BPR

• Bigger BPR is not always better. Above ~15:1 the drag of the larger nacelle and weight of the larger engine starts to eat the propulsive-efficiency gain. The current sweet spot for narrow-bodies is 11–13; for ultra-long-range twins it's 9–10.
• BPR alone doesn't determine SFC. Thermal cycle efficiency (depending on pressure ratio and turbine inlet temperature) is equally important. A modern BPR-9 engine can outperform a 1980s BPR-10 because the core is so much more efficient.
• The fan does not provide most of the thrust at all speeds. At takeoff a high-bypass engine gets ~75 % of its thrust from the bypass air. At Mach 0.85 cruise the split is similar. But supersonic flow regimes (low-bypass engines) inverts this — most thrust comes from the core jet.
• BPR is not the same as fan-pressure ratio. BPR is mass-flow ratio; fan-pressure ratio is the pressure rise across the fan. The two often correlate (high-BPR engines have low FPR around 1.4–1.6; low-BPR engines have high FPR around 2.5–3.5) but they are independent design parameters.
• "Geared" doesn't mean "higher BPR" automatically. The GTF architecture enables higher BPR but the BPR is a separate design decision. A direct-drive engine can match the BPR of a GTF (CFM LEAP-1A at BPR 11 matches PW1100G's earlier specification), just with different efficiency and noise trade-offs.