# Control surfaces actuators design/preferences

### Help Support Homebuilt Aircraft & Kit Plane Forum:

#### User27

##### Well-Known Member
If your heart is set on an FBW aeroplane then also include a full aircraft parachute and make the gear strong enough to withstand a parachute landing without wrecking the airframe as it is very likely the parachute will be used during development. Trade-offs are everywhere in all aircraft, but does the time/cost/risk/reward trade stack up for FBW in a homebuilt aircraft? Doesn't work out for me.

#### User27

##### Well-Known Member
... the A380 and the FBW Gulfstreams (and possibly more, I don't know for sure) use a setup like this as a backup to actuators powered off the main hydraulic system, and I think the F-35 may use it as a primary system for some/all surfaces.
Search for electro-hydrostatic actuator (Moog EHA) or (Parker EHA)
There's not much detail on the F-35 systems in the public domain, this is one of the most detailed (F35)

#### wsimpso1

##### Super Moderator
Staff member
Log Member
Do you think that there has been any significant change in the reliability of electronics since the first adoption of FBW?
Well, maybe in home electronics yes. In airplane control surface management, these high levels of reliability were been demonstrated early. Let's remember that the electrical power and signal processing not only has to be really reliable, but so does the wiring, connections, sensors, actuators, etc. Those system components are not just carrying electricity, they are producing mechanical loads cycled over a distance, vibrations, thermal cycling, and weather effects equivalent to the parked time the airplane gets while accumulating the operational time.

It's just that 1000 hrs of run time seems ,hmmmm, very, very, likely, considering that many devices get thrown out or shelved not for failing, but just by being superseded. Its a question as so many parts of an airplane are becoming electrically dependent that in order to achieve that 10e-3 minimum failure rate by having triply redundant systems could drive weight and complexity up and start to become self defeating in a HBA.
As Marc pointed out, the certified level is decent target for a reason. That is the target for operating the control surfaces. Why? Because if the engine stops, we still have pretty good odds of surviving the forced landing if the controls are still working. If a control surface goes hard over and stays there, even a ballistic parachute is unlikely to save the lives of folks in the seats. Better equip this bird with hot seats.

SO how on earth does a failure rate for any single component get established in the world of HBA? Is it up to builders to make a test circuit ,put it on a wall and have an arduino cycle it for six months continiously? Or to put my interest another way, anybody got a nicopress for sale real cheap?
If someone decides on any level of demonstrated reliability, then they still will need some way to show themselves that the level they will tolerate is met. 10,000 hours is about 15 months of moving the actuator against load while seeing everything else exposed to hot and cold and snow and rain and salt air... Environmental test cells might be skipped if you put it out in the weather running against loads someplace that has rain and snow and sun.

The really big thing that has been proven out long ago is that complex systems are really tough to give really long highly reliable lives when they are a single string. We do know through huge experience in everything from boilers to spacecraft that we achieve high levels of reliability through redundancy. You do have to be willing to design, build, and maintain all that redundancy, or it goes back to the lower level of safety from it becoming a single thread. We can put 1000 hours on a system in 42 days. Or more importantly, we can put 1000 hours on multiple systems in the same 42 days.

Now if that is too high a chinning bar, we can just do a robust mechanical system that is generally WAY reliable and be happy.

Billski

Last edited:

#### dog

##### Well-Known Member
Ok lets see if I understand a basic fly by wire, it is system that converts mechcanical
hand controls into an electrical signal useing strain guages, that is then sent by wires to a hydraulic servo valve that then controls hydraulic flow,provided by an engine or electricaly drivem pump,to hydraulic cylinders that are mechanicaly conected to our control surfaces.
There are presumably sensors on the control surfaces and a black box monitering the triply
redundent electrical portion of the control system and talking to an auto pilot.
I thought it through and think that useing an electrical actuator composed of a jack screw and a motor with limit switches would be iffy
in that a non functional actuator will lock the control at that point, and building truely reliable actuators is going to get pricy.
,with two or more motors on each screw ,and the limit switches and or strain guages,or stepper motors with enough steps so that it feels analog.
I just gota couple things to do first and Ill get right on it.

#### Vigilant1

##### Well-Known Member
Ok lets see if I understand a basic fly by wire, it is system that converts mechcanical
hand controls into an electrical signal useing strain guages, that is then sent by wires
There's another part that is often in there. The hand/foot pressure doesn't get converted directly into control surface movement. Instead, the pilot's control pressires get processed by a fantastically reliable and complex computer to figure out what the pilot wants to do, then signals are sent to the control surfaces to accomplish that result. So, the same left stick pressure might produce the same 30deg/sec roll rate whether the plane is going 200 knots or 600 knots. If the left aileron is inop/shot off, the computer might reflex the right aileron to accomplish the roll, and add some rudder to fix any yaw that results. Or, maybe move the stabilators differentially to produce roll.
Yep, super complicated. But one of the best justifications for true FBW--to get maneuverability and damage tolerance that is impossible to accomplish any other way.
A homebuilt doesn't need it.

#### rtfm

##### Well-Known Member
For me, this is enough of this FBW stuff. I think I've figured out how to run cables up through the single wing masts, across the length of the rib to bellcranks.

#### wsimpso1

##### Super Moderator
Staff member
Log Member
There is usually more than that:

There is frequently three (or more) computers each reading the sensors and figuring out how to give what the pilot is calling for, then the actuator signals compared, one selected, then signals sent to actuators.

At the actuators, they tend to be doubled actuators so if one (cylinder or motor/screw) messes up, the other end of the same actuator can still operate. The other tool in the box is to have more ways to get a response - for wing mounted surfaces each separately actuated and/or elevators operated separately to also give roll response.

One other great thing is that the computer has control laws that will not allow bad stuff to happen. Well, until it does. Airbus' fly-by-wire did not preclude a rudder doublet by a pilot from breaking the vertical fin right off of an airplane or kept confused pilots from inputting full back stick and mushing all the way to the Atlantic... Not fool proof.

Then there is benefits for tactical airplanes. Automatic trim and coordination with automatic limiting that changes as stores come off, enhanced maneuvering, etc. The main reasons an F16 pilot needs to touch the rudder pedals are to land in a crosswind or to fire on a target.

Fly by wire single thread would appear to have less pieces, but has lower reliability per piece and higher consequences (severity times frequency is the figure of merit for how bad) of each means they have to install a bunch redundancy in sensors, processors, and actuators to get to acceptable total reliability.

In big/fast/complicated/tactical airplanes, it can be worth the trip. In helos it can be worth the trip too. But in our little birds?

Billski

#### Heliano

##### Well-Known Member
Billski comments are wise. Failure of a primary control, in particular of roll or pitch control, is regarded as catastrophic by FAA's Ac 23.1309. To prevent a catastrophic failure mode from happening a very, very robust system must be designed. The chances of failure of the system as a whole must be one in 10^-8 or less. To meet that requirement, the FBW systems are fault tolerant and are VERY complex. Even so there are examples of unexpected problems with FBW aircraft. A Sukhoy Superjet was hit by lightning and became almost unflyable, crashing onto the runway; an Air France A330, flying from Rio to Paris had a pitot static failure, went into direct mode and the crew made a mess out of it, partly because of the human factors related to the direct mode/sidestick features. When it comes to safety, simpler is better. I love cables, pulleys and bellcranks.

#### REVAN

##### Well-Known Member
Is it true that electrically actuated and FBW are the same thing? If no computer is involved is it still fly by wire?

This definition from Wikipedia: "The movements of flight controls are converted to electronic signals transmitted by wires, and flight control computers determine how to move the actuators at each control surface to provide the ordered response" implies that controls are driven by output from some sort of flight control computer.

A straight analog input that feeds output system is a simplification of that definition. So is a simple electric trim system FBW? Been noodling on an entire secondary flight control system that is just servos and trim tabs. Basically designed for only a fraction of the pitch, yaw, roll rates that the primary controls can create. Only reason is to be able to drive an autopilot.
A computer can be an analog device. A device doesn't have to use a microchip to compute. A computer is anything that takes an input, applies an operation and delivers an output. An electrical servo-actuator with a built-in analog feedback control loop is a "computer". I would consider any electrically operated control system to be FBW, regardless of analog or digital design complexity. If there is no direct mechanical linkage from the pilot to the control surface, if that linkage is broken with an electrical system somewhere in that path, then it is FBW.

#### maya.ayoub.32

##### Well-Known Member
HBA Supporter
Or to put my interest another way, anybody got a nicopress for sale real cheap?
If you're in the SF Bay Area and still need one we can give you a contact for someone who has one. We're borrowing it too.

#### maya.ayoub.32

HBA Supporter
You can make quite a few of them from a six pack for about $10. Well maybe that would be a little thin. You might get in touch with a local architect and see if they have any metal panel samples to get rid of (we always seem to have them coming out of our ears). #### dog ##### Well-Known Member If you're in the SF Bay Area and still need one we can give you a contact for someone who has one. We're borrowing it too. Thanks for the offer.I am in Canada and am much closer to Ireland than SF. And your pulley guards are an excellent beginer metal working project,no need for expensive metal,clean scrap from anything of the right guage will work. I would give you plenty to practice and build, but maybe someone has some closer like say Hawaii. #### TFF ##### Well-Known Member You might be physically closer to Ireland, but shipping is closer to SFO. #### wsimpso1 ##### Super Moderator Staff member Log Member Does anyone have any recommendations for cable covers? The affordaplane uses them, but they seem home-made: View attachment 112284I was wondering if it's recommended to make them ourselves, or to buy them? We found some off VXB, but they're quite expensive at$10 each.
Gaurds like that can be overridden pretty easily. A common way of doing them is to make them similar to the pictures, but with 1/4" of excess, then drill a cross hole for a cotter key that is very close to the cable and sheave.

Billski

#### maya.ayoub.32

##### Well-Known Member
HBA Supporter
local architect
Fantastic idea thank you!

1/4" of excess, then drill a cross hole for a cotter key that is very close to the cable and sheave.
So does this mean using a drilled bolt instead of the one shown in the picture, or to have the excess material near the cable and use the cotter key to guide it? Thank you!

#### Dan Thomas

##### Well-Known Member
Fantastic idea thank you!

So does this mean using a drilled bolt instead of the one shown in the picture, or to have the excess material near the cable and use the cotter key to guide it? Thank you!
Like this:

Get the pin (in the right diagram) real close to the pulley's edge.

#### proppastie

##### Well-Known Member
Log Member
the idea is to keep the wire rope from jumping out of the groove and getting caught between the side of the pulley and the mounting or guard and jamming up and stopping the rope from moving.....not a good idea to have your control completely immobilized. a drilled hole for a cotter pin might be easier that two bends and a hole for the pulley in just the correct places so the rope can not rise up out of the groove.