# Why do jet engine fan blades have a twisted shape?

I'm thinking: as the blades rotate, the tip moves faster than the hub. So to make the blades efficient, the blades are usually twisted. The angle of attack of the blades at the tip is lower than at the hub because it is moving at a higher velocity than the hub. The amount of thrust generated is higher near the tips than at the root. Engineers want the thrust to be equivalent on every part of the blade. Is that right?

Source: Rolls-Royce plc. flickr.com.

• Yes that's about right, but it does not fit the format of this site to put the answer imside the question. Ask the question, then answer your own question in an answer. Jul 25 '17 at 0:30
• The fan is not to generate thrust. It compresses air. The nozzle generates thrust. Jul 25 '17 at 2:41
• The fan most definitely produces thrust. That's a high bypass turbofan we're looking at. Jul 25 '17 at 6:11
• Related to the shape: aviation.stackexchange.com/questions/34798/… and tangentially aviation.stackexchange.com/a/26947/4108 Jul 25 '17 at 12:12
• You aerodynamics engineers crack me up! I see at least four people commenting on the question and one commenting on the -1 answer who sound like they know enough to provide an actual answer. Yet the person who actually answered got down voted and told he was wrong. Why not answer the question? Jul 26 '17 at 3:50

There's been patent wars regarding this invention.

What you show is not just a twist, but also a varying sweep. According to Rolls-Royce's patent from 1996, which was the subject of one of the lawsuits, a basic summary is as follows:

This feature produces a distinct mid-height bias to the airflow distribution across the span of a blade, with the result that airflow through the mid-height regions of the flow passage is increased and the airflow through the tip regions is reduced. This has an advantageous effect on the overall efficiency of the blade.

and

Forward sweep is employed near the hub 4 to counteract the rearward sweep of the outboard sections of blade 30 in order to make the design mechanically feasible.

If you look at the image in the middle, it shows how the blade angle changes, the higher the $$S_n$$ number, the farther from the hub. And with the two velocity vectors from rotation and airflow, it's clear the mid-sections will produce the majority of the thrust, which helps reduce the tip losses.

Another object of the invention, is "increased resistance to foreign object damage", "in particular bird strike"; with the majority of the blades's top surface not facing directly ahead, less impact energy is imparted on the blades.

I don't think it's accurate to say that the desired thrust is the same along the entire blade. For a large turbofan blade like the one in your image, the outer part is designed to act more like a propeller (efficient for the bypass air) while the inner part is designed to act more like a compressor (efficient for the core). You're not incorrect in stating that the twist does help optimize the angle of attack with the speed of the section, but that's more about not breaking the speed of sound at the tip than keeping thrust constant along the blade.

For any specific engine, though, the blades are highly optimized based on an incredible amount of complex CFD and wind-tunnel analysis for which there is no simple or intuitive summary. The design of a turbofan blade is driven by several factors, mainly

• aerodynamic efficiency,
• structural efficiency,
• noise,
• temperature,
• vibrations, and
• bird-strike resistance.

All of these go into the resultant shape, but you're unlikely to find specifics that are non-proprietary to the manufacturer and available for anyone to analyze. The twist specifically is just another parameter in the extensive, complicated optimization.

• Peter, I disagree extremerly with the statement that "the outer part is designed to act like a propeller". Do you have a reference for this? I have read many books on gas turbine performance, and never ever have read anythng that suggests this. e.g Mattingly's Elements of Propulsion, or Oates Aerothermodynamics of Gas Turbine and Rocket Propulsion, or Gas Turbine Theory by Saravanamuttoo. Performance modeling of the fan is identical to the LPC or HPC - using a compressor map defining pressure ratios, mass flow, and efficiency. But the rest of your comments are correct. Regards Aug 2 '17 at 11:43
• @Richard: The concept I'm trying to convey is that the most efficient thing for bypass air is to move a lot of it with as small a change in velocity as possible (the goal of a propeller) and the most efficient thing for the core is to have the air compressed as much as possible. The turbofan blades need to do both, so some efficiency is lost; the twist is one way of mitigating these losses. Compare a propfan, where the functions are decoupled and higher efficiency can be achieved. Aug 2 '17 at 22:21

Why do jet engine fan blades have a twisted shape?

Your question is about fan blade twist, but actually the answer is valid for nearly all rotating blades. If a blade weren't twisted, the angle of attack would be very different from its root to its tip.

Twist equalizes the angle of attack along the length. I say equalize for simplification, the angle may not be designed to be equal, other aspects may requires the blade to include some variation of the angle of attack, e.g. to prevent the blade to oscillate (flutter), still the general goal is a to equalize the angle.

Angle of attack

The angle of attack is the angle between airflow and blade chord.

For the blade to be efficient this angle must be close to the maximum lift angle, which is roughly around 15° for usual airfoils:

However we're talking about the direction air actually comes from, as seen by the blade. When the aircraft is moving straight and level, for the pilot air comes from ahead, but this is not true for the fan blade, due to blade rotation.

Velocity created by rotation (tangential velocity) depends whether the point is close to the center of rotation (smaller velocity), or close to the tip (larger velocity).

For a rotating blade, air direction is the (vector) sum of aircraft translation, which is aircraft airspeed, and blade rotation. The sum has a fixed component, the translation, and a variable component, the tangential velocity at the point under consideration:

Let's imagine we have tweaked the blade angle of attack at location A. What happens at locations B and C if the blade is not twisted?

As we move closer to the blade tip, the tangential speed increases relatively to forward speed, so the sum (purple) gets more vertical. Angle of attack is now too small for locations B, and in C it is even negative, that is the blade pushes air in the wrong direction.

To offset air direction rotation as we get closer to the tip, the blade must be twisted by a value shown in gray in the picture above. Near the root where forward motion is the largest, the chord is close to the direction of travel; near the tip where tangential speed is the largest, the chord is nearly in the plane of rotation of the fan:

Source: The excellent Bjorn’s corner

Effect of twist: Varying incidence

The angle of attack of the blades at the tip is lower than at the hub because it is moving at a higher velocity than the hub

From what was discussed we are able to distinguish the actual angle of attack from the apparent angle, which is generally called pitch or incidence, that is the angle between the engine longitudinal axis and the chord.

On a twisted blade, incidence varies, and is larger at blade tip. However the angle of attack, which actually depends on rotational velocity and cannot be evaluated visually, should be more or less constant.

In turbofans, the fan acts on two separate flows: Primary flow, the smaller in mass which is used to produce gas to spin the turbine; and secondary flow, the largest which bypasses the engine core and propels the aircraft.

Accordingly, fan blades in such engines are designed in two sections. Near the hub, blades are part of the low pressure compressor, their role is to increase pressure. Elsewhere they accelerate air a bit to create thrust, like a propeller.

These two designs must deal with their own constraints and optimizations, this gives a particular shape. Modern fan blades are formed hollow sandwiches titanium/aluminum and composites, and are comparatively lighter and more complex. When rotating they are less stressed, in particular where velocity is higher, than heavier blades, and they can be wider without being subject to deformation. This allows to give them a more optimized shape, including twist, taper, dihedral, sweep, etc. The shape evolution over time is visible in the last picture of this answer.

• Is it the same reason why compressors are also twisted? The value of relative velocity is different for a hub, tip, and mean radius. Sep 24 '21 at 23:44
• @Auberron: You're right. For compressor blades (and fan blade inner section), there is also a need to make the axial velocity constant from hub to tip to control flow turbulence and so the exit velocity is constant. The purple vector modulus (length) changes from hub to tip, it must be corrected. Velocity is converted into static pressure by next stator, this is the principle of axial compressors. Therefore stationary vanes are twisted for velocity equalization. The blade shape must match the next vane for smooth flow. This plays also a role in blade twist...
– mins
Sep 25 '21 at 12:04
• ... To become familiar with axial velocity and velocity triangle, you may have a look at this video. Someone may suggest a better one.
– mins
Sep 25 '21 at 12:10

(Hot section) turbine blades are twisted (in the chord wise direction) because they are partially utilising reactive aero mechanical design, and impulse aero mechanical design. (The blade is also twisted in the radial direction due to vortex flow. See the last paragraph of the development section here. This is also mentioned in Frank Whittle's book, Gasturbine Aerothermodynamics. Rolls Royce's book, "The Jet engine" says on page 50: "The reason for the twist is to make the gas flow from the combustion system do equal work at all positions along the length of the blade and to ensure that the flow enters the exhaust system with a uniform axial velocity.")

With fan blades, I am not so sure - I think your answer is along the right lines, (though it's better to say even pressure rise than thrust, although ultimately that pressure translates to thrust), but if that was the only factor at play, I think the blade shape would be a smooth arc from the hub to the tip, and that's not the case. Usually in a gas turbine, the outer fan case diameter is relatively straight, but the hub expands as the air flows though the fan, then into the compressor, the air travels through a bend. So, the air is moving in an axial and radial direction. This maybe an added factor that causes fan blades to be twisted. Fan and compressor blades have the difficult job of making the air flow in the direction of increasing pressure. Flow seperation will cause stall, so care needs to be taken to avoid any local flow disturbances. The air can also choke within a compressor, which is why sometimes air is bleed of the rear stages for stability control. To avoid choking in one region, the air needs to be distributed evenly. (This is more an issue in the booster or HPC than the fan however). Reducing fan noise is another requirement.

Note: Bits in () in 1st para added in subsequent edit, to include feedback in comments.

• Sorry, my first sentence is misleading, and I think that has caused confusion. The first sentence refers only to blades in the turbine (hot) section, as a contextual remark. The rest of the comment is about blades in the cold section - fan, or booster/IPC or HPC. I agree that some fan blades are swept to avoid sonic issues, that results in lower noise and improved performance, as discussed in the related question you highlight. Jul 26 '17 at 11:48

Because airflow will meet the leading edges at different velocities ( at the hub is slower and needs more AOA and at the external edge is faster and fine pitch is needed ). To benefit from the all available airflow and not get aerodynamic( for compressor) stalled air the twisting must be added to the blades. So to not have Stall of the blades is the answer.

• That way the blades will have a more surface for the diameter and double action/ reaction response because if you redirect the airflow that much. So first must take care of the meeting airflow with various points of the diameter of both sides blade and after that to make sure that reaction force is moving airflow at a desired angle (this is part of the 50 degrees you pointed out) . But the twist is because of the different speed.The fan from the initial post is from the engine that bye a good amount of airflow. Sep 25 '21 at 19:31
• You should read this page, this is about propellers, but this is the same principle than for fans. AoA is constant: "For a propeller to operate efficiently, it is desirable to have the majority of the blade at the same angle of attack relative to the local incident airflow [...] Ideally, the whole blade is operating at the lift coefficient at which the blade airfoil achieves its maximum lift-to-drag ratio. We achieve this uniform angle of attack distribution by twisting the propeller blades properly"
– mins
Sep 25 '21 at 20:48