Supersonic Engines

(This is not meant to be an exhaustive text on gas turbine theory and application, plus I’m trying to refrain from being my usual aero-nerd self. Some things are not mentioned like combustion chamber design and the varieties of turbine one could opt for. I know! This is meant to focus on the really critical parts of cobbling together a supercruise engine.)

 

An adequate powerplant for supersonic flight is a massive sticking point for anyone hoping to develop a civilian transport. Efficiency at high speed without excessive fuel consumption is a balancing act that becomes more difficult the less one is able to spend on exotic materials. For a civilian designer faced with a budget that resonates more with FBO than DOD, we should explore what is required with regards to creating a supersonic powerplant for non-military aircraft.

The most important thing for any aircraft flying faster than the Mach is to have an exhaust speed which is equal to (in actuality, slightly faster) than the true airspeed. The drawback is very high noise levels from exhaust shearing (like I care) and very high fuel consumption for a given thrust setting (this I really do care about). The first solution most people would consider is to increase the bypass ratio. The complicated reality is that there is not a single bypass ratio that is optimal for all phases of flight, especially for an aircraft with a 750+ knot speed range. What may help at Mach 0.60 will probably hurt at Mach 1.15. Furthermore, by increasing the fan size or fan rotational speed, the core of the engine has to work harder and either give up thrust or increase turbine temperatures to compensate. Any increase in fan diameter has to be weighed against these factors.

We can rule out large bypass ratios of greater than 5, such as those used for modern airliners. The fan diameter is too large and creates too much frontal area drag. In addition, the massive volume of cool exhaust flow is not able to move fast enough to push the aircraft to supersonic speeds. Theoretically one could take a large diameter fan and spin it faster (many large civilian turbofan engines spin their fans at speeds below 3000RPM) but this would require a lot of extra energy from the core. On the other hand, a military style bypass ratio of 0.2 to 0.8 is not enough to produce the required TSFC for a civilian aircraft that cannot refuel in-flight or carry external tanks. Thus somewhere between a bypass ratio of 1.0 and 5.0 is the optimal choice for our engine.

Way too wide for our purposes, but perfect for the C-17.

On the front end, the fan pressure ratio affects the specific thrust and indirectly, speed of the air through the engine. Specific thrust is the thrust divided by the inlet airflow in pounds per second. Low bypass engines tend to have very high specific thrust values while large high bypass turbofans have a very low specific thrust. Civilian turbofans usually use one large fan whereas high performance military turbofans typically use 3 or more fan stages for this exact reason (as a reference, the F100-PW-229 has a specific thrust of 71.77, the F101-GE-102 ranks at 48.84 and the GE90-B4 comes in at only 28.78). The now “hardened” bypass air is able to move through the engine with enough energy (combined with the core flow) to provide a choked nozzle condition at the exhaust orifice (Mach 1 flow at the narrowest point). The flow can then be exploited by a variable system aft of the choke point to accelerate the air to the required supersonic speed (I have simplified so much that it annoys me, but otherwise, this article would go on for days).

We also must mention overall pressure ratio, which is the amount of total compression achieved by the engine/inlet combination. A lower pressure ratio means fewer exotic materials required in the hot section, a lower engine weight, less extravagant methods of cooling and as a result, a lower manufacturing cost. On the downside, the thrust level is lower for a given engine as compared to the same engine with a higher pressure ratio. Another negative effect is that a lower overall pressure ratio raises TSFC for a given thrust setting. Depending on the aircraft budget and expected operating environment, trading extra fuel burn for lower initial cost may be acceptable. A word of note is that as Mach number increases, TSFC increases as well. This may be offset by ram pressure recovery (mentioned later) but it is important to know that the TSFC at Mach 0.80 is not going to be the same at Mach 1.3.

The intake setup is important even though strictly speaking it is a part of the airplane, not the engine. As an aircraft with the proper fan pressure ratio moves faster, it is able to delay thrust decay and in some cases, reverse the process thanks to ram recovery. But this is if and only if the intake is designed properly. At low supersonic velocities, a simple normal shock inlet is sufficient to allow adequate pressure recovery at the fan/compressor face. As velocities increase past roughly Mach 1.5, the losses mount exponentially and thrust will degrade accordingly. A multiple shock inlet can reduce these losses significantly, however there may be issues with shockwave placement at off-design speeds. In the quest for low cost and low weight, a fixed position normal shock inlet is probably the best choice for a civilian supersonic jet. If one wishes to engage in what the military terms “carefree handling”, then intake design must be given far more attention to ensure that certain flight conditions do not lead to disturbed flow. Turbulent air at the fan/compressor face can lead to surges and stalls of varying severity.

If it’s good enough for the Viper, it should be good enough for us…fixed normal shock inlets are lightweight, inexpensive and have no moving parts.

Finally, the most important piece of the puzzle is the exhaust nozzle. Ideally, a jet engine exhausts air at ambient pressure to produce a stable column of thrust. A given engine can force air out at a higher pressure than ambient, but this flow will simply overexpand, collapse in upon its now low pressure core and possibly re-expand. This is very inefficient and can be hazardous to the aircraft’s operation. To allow the higher than ambient pressure flow to expand under control so that it’s energy is translated aft rather than radially, a divergent section of nozzle is required. Every angle made with respect to the convergent and divergent sections has a particular Mach number and pressure ratio associated with it. Knowing this, for an aircraft to have maximum efficiency across a wide range of Mach numbers, a variable convergent-divergent nozzle would be required. However, a variable exhaust nozzle is extremely complex to build and requires a system to activate it (oil or bleed air in most cases). A fixed nozzle will have far less efficiency but an attendant lower cost.

Controlled by bleed air, these F100-PW-220 nozzles are very complex.

To recap, we are in need of a low bypass turbofan probably between 1.0 and 5.0 with a high fan pressure ratio, moderate overall pressure ratio, adjustable exhaust and fixed inlet. Starting at the front of the engine, we can assume a bypass ratio of 2.5 for no other reason than it’s halfway (and back-of-the-napkin wit). With this we are assured of an acceptable frontal drag penalty, while still having a fan small enough to stage if required. To keep core requirements within a reasonable range, a specific thrust range of 50-70 allows us to move air fast enough without taxing the core too much. An added bonus is that multiple fan stages can eliminate the need for a low-pressure compressor altogether. At the rear of the engine, the exhaust nozzle should be adjustable to a certain extent. A dual-position nozzle may be the best alternative to a fully articulated iris nozzle when cost and complexity is considered.

As of now, there are several civilian engines that qualify with minimal modifications and many that could fit the bill with more extensive changes. In the interest of rapid development, low cost and minimal risk to the aircraft, a minimal-change option is the best choice for a civilian budget. The Williams FJ44-2 and Pratt & Whitney Canada JT15D are both contenders for small aircraft. The medium to large designs could be well served by the Rolls Royce Tay 611-8, Spey 511-8 or Pratt & Whitney JT8D with very few changes (certainly for less effort and expense than a cleansheet design). While some of these engines are no longer in production, there are enough examples to support a test program and even limited run manufacturing of aircraft.

This is a classic chicken vs egg issue. Engines will not be produced unless there is an airframe that requires them and airframes will not be built unless there is a reliable engine available to power it. Somebody has to blink first.