Let's make a helicopter blade

[Dave_Jackson]

Possibly, the largest constraint in the development of kit, or homebuilt helicopters, is lack of a source of blades.

As a start to bugdevheli's "Lets make a helicopter", what about a team effort toward the development and production of blades, only? The equipment could be given the capability to produce a variety of blade sizes. For example, the largest allowable blade might be for a 2-blade helicopter at the maximum weight of the new Very-Light categories. This equipment would also be given the capability to produce smaller blades for lighter helicopters, and/or for helicopters with more blades.


Alternativly;

IMHO, the single greatest improvement in rotorcraft will be the development of Active Blade Twist. Specifically, an ABT that is capable of large twist at a high twist rate. There is no reason why a group of motivated people on the Internet cannot compete with the universities and companies that are currently looking into ABT.

There would certainly be a market for very-light helicopters, which due to an improved L/D ratio, had a quantum leap on existing helicopters.

Dave

[bugdevheli]

How many and what size and degree of twist would you like? Bug

[Dave_Jackson]

bugdevheli,

How many and what size and degree of twist would you like?

This would be open to a consensus.

An idea that was considered a few year ago was to produce a female mold (two if the blades are to be asymmetrical) to produce the outer surface of the blade. This mold would have a taper and probably a twist. The mold would have a span that was 3 or 4 feet longer than the longest blade. The root chord of the mold would suit the root of the largest blade and the tip chord of the mold would suit the tip of the smallest blade.

Blades could then be produced at any location within the extra long span of the mold. The different blades that were produced would have to have the same taper and twist but their length and chord could differ.

Dave

[Vfrpilotpb]

If this concept for a new rotor blade with Active Pitch Change could be brought about by how much in degrees would this APC need to change the "twist" on that blade, and would the static or non moving blade( not rotating) have or need any structural twist in it or could it be manufactured totally flat to begin with?(albeit with aerodynamic profiles in its length)

PeterR-B
Vfr

[bugdevheli]

Just out of interest. I have made a number of blades that can twist, or more correctly, at the moment the outer skin can twist. I am currently looking to find a suitable method to accomodate the lenghtwise movement of the material both at the trailing edge and the feathering axis. Ten to twelve degrees of active twist is quite easy to achieve from a flat blade. The blades skins are moulded in one shot, no joint at the leading edge. Bug.

[Dave_Jackson]

Re: Active Blade Twist

Vfrpilotpb

...by how much in degrees would this APC need to change the "twist" on that blade...,

The amount of Active Blade Twist will depend upon what the manufacture wants the rotor to be capable of doing. The tip pitch angles will probably not change much from those of today's rotors. However, the root pitch angles will be quite different.

A simple ABT may be required to provide; high negative twist (tip angle < root angle) during hover, low negative twist during forward flight and very low positive twist during autorotation.

An advanced ABT will be required to provide the above three plus high positive twist in the retreating blade during fast forward flight. This high positive twist is required for reverse velocity utilization.

Rough theoretical values for the actual twist under different flight conditions can be found mid way down the web page UniCopter ~ Control - Flight - Independent Root & Tip (IRAT)

.. would the static or non moving blade( not rotating) have or need any structural twist in it ...

I have never though much about the 'stopped rotor' concept.


bugdevheli,

I am currently looking to find a suitable method to accomodate the lenghtwise movement of the material both at the trailing edge and the feathering axis.

What material(s) have you use for the skin?

Can you elaborate a little on the "lenghtwise movement"

Thanks


Dave

[IFMU]

Twist-n-shout

Dave Jackson wrote:
IMHO, the single greatest improvement in rotorcraft will be the development of Active Blade Twist. Specifically, an ABT that is capable of large twist at a high twist rate. There is no reason why a group of motivated people on the Internet cannot compete with the universities and companies that are currently looking into ABT.

Dave,

I'd buy that blade if you can make it. Not an easy thing, though, and I think the statement is akin to "the single greatest improvement in automobiles will be a car that runs on grass clippings and produces only water vapor." So, you have a blade flexible enough to produce large changes in twist at a rapid rate, how stable is that blade going to be? And, now a rotating frame control system is required, with its inherent complexity and criticality. I think the active twist individual blade control solution, though neat, may be pretty difficult to achieve no matter if you are a university geek, internet geek, or a genuine helicopter engineer geek. But, like I said, I'd buy it if you could build it. I think so would everybody else.

-- IFMU

[Dave_Jackson]

IFMU

The concerns that you raise are legitimate, welcomed, and perhaps based on personal knowledge. It is these very concerns that represent the current challenge and hopefully the future pleasure.
No pain ~ no gain
______________________

The problem:

Low efficiency is, arguably, the greatest handicap of helicopters. This inefficiency was described by Prouty, in Rotor & Wing, Jan 99

"The most efficient angle of attack for an airfoil is where its lift-to-drag ratio is the highest. For most rotor airfoils, this angle is in the neighborhood of 8º. ...... An airplane designer can arrange things so that the entire wing is flying near the optimum angle of attack at cruise speed ......The overall airplane lift-to-drag ratio can be 10 to 30, depending on the configuration, whereas the maximum a helicopter can do is 4 to 6."

The solution:

Active Blade Twist is probably the most viable means of improving the lift-to-drag ratio of helicopters. It is a way to optimizing the blade's angle of attack at all polar coordinates (radius x azimuth). As you mention, a stronger rotor will be necessary. In addition, a slowed rotor, larger chord and pusher propellers will be beneficial.

The means:

The anisotropic nature of composite construction is resulting in components that have high strength, light weight and ordered flexure characteristics. A pultruded or filament-wound spar with the unidirectional threads running spanwise should meet the structural concerns.

Mechanical independent root and tip pitch controls should be possible today.

A near term solution should be the combining of a single mechanical (or electrical) pitch control with the rapidly developing piezoelectric active fiber composites. A +45/-45º bias wrap over the full span of a unidirectional spar should provide the necessary displacement, force and frequency for a near optimal blade.

Further into the future it should be possible to sectionalize the bi-directional piezoelectric tape wraps so that the elements along the blade can be individual twisted.

I'd buy that blade if you can make it.

I'm trying.
UniCopter ~ Rotor
UniCopter ~ Control - Flight


Dave

[Vfrpilotpb]

Dave,

What about making the entire length of the blade from two parts, an iner core(that was fixed to the Head) this would ressemble the rotating part of an ordinary electric motor with sophisticated windings along its length, and the outer part(the blade face's) which would be the opposing part, by working out the resistance to certain parts of the length of the blade a differential reaction could be obtained that would be translated into a form or warping (twist) to desired areas of that outer skin or blade surface. The twist or warp being created by torque produced from the opposing parts of the "Linear windings in the blade, when electricity was introduced

By reversing polarity as can be done with ordinary electric motors to make them run in reverse this would then allow a reverse warp to be created on the outer surface of the construction.

To do this would need sophisticated computer controlled switchgearing to translate from the pilots hand controls to electronically produced twist inputs at the blade.

What dost you think, can I leave my day job now!

PeterR-B

[slowrotor]

Dave,
A group might share molds or other tooling for a conventional blade. Not so realistic for an experimental active twist blade.

Rotor blades are probably the most torsionally stiff structures ever designed to prevent flutter. I dont want any flexable blades unless you can prove it to be simple, cheap and safer than stiff blades. High performance is not needed. Any performance would sell if the cost was reasonable. Just my opinion.

Slower rotor is a good idea.

[IFMU]

Dave Jackson said:
Mechanical independent root and tip pitch controls should be possible today.

I see challenges with this. That torque tube has a lot of potential windup over the span of a blade. And, even those rigid rotors generally aren't all that rigid. So, it's not like the torque tube is going to have a nice straight tunnel to ride in as the blade whirls around. This will induce bending, chafing, and honest to God fatigue cycles.

Dave's website says:
The skin of the blade is firmly attached to the spar at all locations along the span. The skin and the spar twist in unison. The root of the blade is controlled by the root pitch horn. The tip of blade is controlled by the tip pitch horn. The tip and the tip pitch horn are connected by a torque tube, which is located in the middle of the spar.

You already have a strong piece of structure absolutely required that can serve duty as a torque tube. It's called the spar. Why wouldn't you use that as your torque tube, secure the airfoil to the tip, and have seperate pitch horns inboard, one for the spar and one for the airfoil. That reduces the parts count by one. Of course, the same problems with chafing and bending will still be present.

Vfrpilotpb said:
To do this would need sophisticated computer controlled switchgearing to translate from the pilots hand controls to electronically produced twist inputs at the blade.

The sophisticated computer controlled switchgearing is the easy part. You can find somebody to write all the equations (or do it yourself if that's your thing) to define mixing to the active twist. The tricky part is getting that fly by wire stuff from the pilot's hand in the fixed frame to the rotorhead in the rotating frame, and with enough redundancy and reliability that you are reasonably sure you'll go home everyday after flying it. I think the other tricky part is basic blade stability with this scheme.

slowrotor said:
Rotor blades are probably the most torsionally stiff structures ever designed to prevent flutter.

That's part of what I was trying to say in my original post.

I may come across as too negative in this little open source rotor design community. But, I think that what we are talking about is tommorow's technology, not todays. In the immortal words of Igor Sikorsky, "To invent a flying machine is nothing. To build it is little. To make it fly is everything."

-- IFMU

[bugdevheli]

Bloody Hell Dave, Futuristic or what. My twisting skin concept is very basic compared to yours. Imagine a blade where the trailing edge can slide accross itself. Fix the tip to the spar and twist the root by conventional means. This produces a constant active linear twist. Build the skin with variation of laminates and twisting of different sections of the blade is possible. Materials I am using at the moment are prepreg carbons both unidirectional and various weaves.

[Lu Zuckerman]

Active blade twist?

It is my understanding that helicopter blades are negatively twisted starting at the root with the greatest static pitch to the blade tip which is usually at zero or near zero AOA relative to the basic low collective static position. The purpose of the twist is to attempt to equalize the lift across the blade span to minimize blade bending during flight.

With active blade twist this fixed relationship between the root and the tip would be constantly changing developing a very strong traveling wave from the tip to the root requiring some very heavy vibration damping gear on the head. It might also lead to serious fatigue problems due to the traveling wave both on the blade and at the root attachment to the rotorhead.

But then again I’m not a graduate engineer just a fat kid from Cleveland.

[Dave_Jackson]

Vfrpilotpb,

Your thought of using an electro-mechanical means to twist the blade is probably what we will see in the not too distant future. They are coming up with neat products such as piezoelectric actuators etc. If interested, this provides some information on piezoceramics.


slowrotor,

I'm open to participating in the 'simple', the 'complex', or both. However, the latter one really "turns my crank".


IFMU,

"That torque tube has a lot of potential windup over the span of a blade. And, even those rigid rotors generally aren't all that rigid. So, it's not like the torque tube is going to have a nice straight tunnel to ride in as the blade whirls around. This will induce bending, chafing, and honest to God fatigue cycles."

Thanks for the very valid concerns. The following is the 'theoretical' attempt to overcome them and any comments will be appreciated.

The torque tube will only have layers of biased filament wound carbon tow. The odd numbered layers will have a +45º bias and the even numbered layers will have a -45º bias. The innermost layers will extend out to the tip of the spar. The overwrapping layers will have progressively shorter lengths, in the spanwise direction. The final result will be a torque tube that has an extremely strong resistance to twisting and very little resistance to bending. Ref. Picture

The spar used in the Sikorsky XH-59A ABC was fairly rigid , however, it was only made of titanium because back then they were not willing to trust composites and pultrusion was probably not yet known of. It appears that even the rigid Bo105 spar Ref. Pictures was not pultruded. My belief (hope) is that the final spar will be very rigid inplane and out-of-plane, but very soft in torsion.

The torque tube is to fit inside the spar. It will be anchored to the spar at the tip end and rotate within the spar at the root end.

A long elastomeric bearing is intended to bridge the gap between the OD of the torque tube and the ID of the spar, all the way along the span. It is intended that the elastomeric bearing will off-load an ever-increasing portion of the twist from the torque tube to the spar on its way from the root to the tip. The tip is the final and firm transfer of twist. It is also intended that this elastomeric bearing will negate the possibility of chafing.

The skin of the blade will be made from fiberglass cloth, for flexibility. The leading and trailing edges will contain pultruded carbon, for rigidity.

"You already have a strong piece of structure absolutely required that can serve duty as a torque tube. It's called the spar. Why wouldn't you use that as your torque tube, secure the airfoil to the tip, and have seperate pitch horns inboard, one for the spar and one for the airfoil. That reduces the parts count by one. Of course, the same problems with chafing and bending will still be present."

The Independent Root & Tip Control by Floating Root Method will certainly be easier to manufacture. However, it will result in a smaller diameter spar for a comparable airfoil size since the 'bearings' must fit between the spar and the skin. It will also mean that the start of the twist must be a fair way inboard from the tip of the blade. In addition, it places the majority of the blade's forces on the tip control and tip bearing.

The desire is to use the Independent Root & Tip Control by Torque Tube Method since the major blade forces will be handled by the spar and therefor by the root control and root bearing. The root pitch angle will be set primarily by the collective and the airspeed. For a Very-Light helicopter, this might allow the tip control to be connected to the cyclic stick without servo assist.


Bug

Thanks.


Lu,

"I'm ...... just a fat kid from Cleveland."

Was the hospital food really that good?


Scathing or polite rebuttals accepted; begrudgingly.

Dave

[Graviman]

Hmmm - interesting thread, but i can see a lot of unnecessary complication creeping in here. I don't see any reason to directly control the AOA (rel TPP) of the tip, unless it is in reverse air flow (ie very low RRPM vs airspeed).

The ideal blade tip shape for fans and propellers is elliptical, which is generally created by just a circular tip in cooling fans. This is slightly complicated by the need to have an additional taper of 1/radius for optimum disk velocity profile. This results in a good downwash pattern, but with sensibly low tip votices. The efficiency should be able to approach a fixed wing.

If the tip is always operating in forward airflow, the easiest way to achieve this is by fitting a powerfull trim tab to the trailing edge of the tip. In this way the swash plate controls the root, while blade torsional characteristics provide the best washout. This also provides good end damping for Lu's blade eigenmodes.

One comment not adressed, which also favours single swash plate control (other than simplicity), is control forces. If this is for a private helicopter powerful hydraulics are not an option. This means that blade needs to be designed for very low torsional rigidity, while being relatively stiff in bending, and having very good flutter stability.

Not trying be awkward , but the message is: keep it simple...

Mart

[Dave_Jackson]

Mart,

As just mentioned on another thread, I believe that that a top-down approach should also be taken toward the development of an optimal blade.[list=1][*]The lift/drag ratio of an airplane wing is far better than that of a helicopter rotor.[*]The primary reason, by far, is that the segments of an airplane's lift surfaces are operating at close to the optimum angle of attack.[*]What, when, where, why, who can be done to significantly improve the helicopter rotor's angle of attack at all blade segments, and velocities?[/list=1] IMHO, Active Blade Twist is, BY FAR, the best way.

Now for the considerations at the next lower level in this top-down approach.


Dave

[Graviman]

"...top-down approach should also be taken toward the development of an optimal blade."

Agreed (i take this approach with most things). An important part of this is also to ensure that the design goes down avenues most likely to result in practical design. In this way you get the perfomance, but keep good reliability without introducing cost.

"IMHO, Active Blade Twist is, BY FAR, the best way."

Not disputing this at all. Just trying to say that there are easier ways of implementing it. Consider a "fixed" wing - IMHO the tip should ideally offer 0 degrees AOA to the airflow, with the rest of the blade increasing AOA elliptically towards root. This produces the weakest tip vortices (ie induced drag) for any given lift. The wing root would ideally have AOA altered for manouvering and velocity trim change, but the elliptical lift distribution is a given. On banking the elliptical lift would effectively develope a slight inflexion. Once the aircraft is established in a turn, the tip will twist to keep it's AOA 0' rel airflow, so the span pressure distribustion becomes elliptical once again. This is easist achieved by a wing that is stiff in bending, but allows torsional flex - ie the wing/blade warping you have proposed. And the easiet way to achive this? - a powerful tip tab, set to 0 degrees.

In a rotor wing the situation is complicated by the wing rotating around an axis, hence inflow velocity being proportional to radius. This means the blade must now have it's form multiplied by an inverse taper (ie 1/radius) for constant downwash velocity, but the dynamics still hold. The stiffness is chosen to allow an elliptical response to any torsion along the blade. A powerful tip trim tab ensures that the tip is always at 0 degrees relative to the airflow. The blade thus only need be controlled from the root, allowing a simple actuation method for Active Blade Twist.

Naturally this only works as long as the air is flowing in the forward direction across the tip. The reverse flow region of the root presents it's own difficulties, and is why i suggest feathering the retreating blade so as to produce no lift. In this way only the advancing blade actually produces downwash, allowing a near perfect downwash pattern in all regimes of flight. Since Prouty's example helicopter achieves a "glide" slope of 6:1, i suggest you aim for at least 10:1 with such a system.

Mart

[Dave_Jackson]

Mart,

This is also one of the reasons for the Advancing Blade Concept, and its requirement for twin main rotors.

Time to get back to a real job.

Dave

[Graviman]

Must admit i'm not to familiar with the ABC control system. I gather it was fly-by-wire though - presumably had a lot of stuff that would end up on Comanche (Wonder what's next now that proj is canned).

I can see actively trying to control tip and root introducing a lot of (at this stage) unecessary modes of failure, not to mention development costs. For a start you have to design blade for structural integrity, flutter resistance AND you need a method to actuate the tip (beit electric servo or torque tube). You then need (say) to package the twin swash plates or servos, and control system. The system itself will definately need some level of computation to figure out required twist, since the objective is to reduce pilot workload - especially in autorotation where the biggest twist benefit is found. When fitted to intermesher you double up on all this!

It just strikes me that a purely structural and aerodynamic method, like powerful tip tabs and torsional compliance, gets you everything you want without added complexity. The control system would be conventional, accept that the tips sort themselves out for minimum induced drag. The retreating blade feathering would best be accomplished by connection to collective only, requiring correction for speed/climb

The downsides i can see are tip speed and control forces. Clearly the RRPM must always remain high enough for positive airflow over tip, setting a maximum theoretical u = 0.5. I don't really see how an intermesher could stop rotation anyway, since rotors are always assymetrical to avoid clash. Control forces would require a VERY torsionally soft blade, which would need some damping to avoid eigenmodes (clearly mass distribution for flutter avoidance is a given).

Perhaps think of this as a halfway design towards full IRAT, in the same way that the CVT+HS is a a step towards rigid. Full blown IRAT would be capable of u>0.5, but becomes more complex. You need the twin swash plates and a blade layup that causes tip to flex in opposition to blade bend for the forward quadrant, where aero-divergence could be a big problem. Again better to do this in steps - a ground test rig ideally providing hard data for each stage of development...

"Time to get back to a real job"

Yeah, i'm becoming an obsessive-compulsive forum surfer...

Mart

[IFMU]

If the ABC was fly-by-wire, those wires would have been the thick, hollow kind of wires that are mounted to belcranks. I don't think FBW was in vogue during that era.