# Control surfaces actuators design/preferences

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

#### geosnooker2000

##### Well-Known Member
What are the specific reasons torque tubes are preferred over tension wires? I know one reason would be, you never have to worry about checking the tension on a torque tube, and torque tubes can't "fray". What else?

#### Jay Kempf

##### Curmudgeon in Training (CIT)
They don't tighten and loosen as the airframe flexes or changes dimensions with temperature changes. But they are super light for crossing long distances.

#### geosnooker2000

##### Well-Known Member
They don't tighten and loosen as the airframe flexes or changes dimensions with temperature changes. But they are super light for crossing long distances.
You are saying the tension wires are super light for crossing long distances, right?

#### TFF

##### Well-Known Member
Wires are lighter weight and have more precise control because the control is always pulling to move it. Runs can be long without problems. Requires lots of planning to get the runs perfect.
Control tube is simple and almost as precise but bearing wear looses some precision over time. Just about anyone can make the tube with the right parts cheaply. Good cable tools are expensive, especially for one time. Pulleys have gotten expensive too for cables. There usually is one that will fit an application better, than the other.

#### Dana

##### Super Moderator
Staff member
Pushrods typically have a lot less friction than cables and pulleys.

Yes.

#### geosnooker2000

##### Well-Known Member
What about flutter? I seem to remember a discussion about the Cri-Cri having a problem with an undersized torque tube for the ailerons? Would cables have eliminated that problem or exacerbated it?

#### TFF

##### Well-Known Member
Torque tubes on ailerons can be an issue on some planes. There is always twist and you don’t want any. Simplicity like a Tailwind or a Grumman it makes sense. Make the tube big enough and most pilots don’t ever notice. Planes that put a lot of loads on the ailerons like the Stephan’s Acro pushed torque tubes to the limits and skilled pilots could feel them wind up. Cri Cri are built two ways. The original complex bonded and the rivet aluminum style. The aluminum version was under designed in a few ways from the space shuttle complex original. The aluminum was an adaptation not from the original designer.

#### BJC

##### Well-Known Member
HBA Supporter
Aerobatic monoplanes evolved from aileron torque tubes to push-pull rods.

BJC

#### Jay Kempf

##### Curmudgeon in Training (CIT)
Flutter is not related to cables or pushrods. One or the other can be a root cause but neither is apparently flutter proof or worthy. Slop can cause flutter in a badly maintained control system.

#### Map

##### Well-Known Member
As it was already said in other posts, torque tubes will twist under load. The longer a torque tube is the worse it gets. So it is ok to use them if they are short, made from a stiff material (carbon, steel..) and have a large diameter.
It is easier to make a stiff control system (requirement for avoiding flutter and getting control surface deflections that match the stick input) by using push-pull tubes or cables. But the design of the system also has a big influence on stiffness. Attaching a support to a soft piece of structure can have the same undesirable effect as a long flexible torque tube.

#### wsimpso1

##### Super Moderator
Staff member
Log Member
What are the specific reasons torque tubes are preferred over tension wires? I know one reason would be, you never have to worry about checking the tension on a torque tube, and torque tubes can't "fray". What else?
First off, you are attempting to get us to discuss cables vs torque tubes. What about push-pull tubes? Torque tubes are appropriate in few places and are used sparingly by airplane designers for a variety of reasons.

I did a quick analysis of all three as we might apply them to ailerons, using a wing span of 24 feet, chord of 5 feet, ailerons 5 feet by 1.25 feet, airspeed of 180 knots. The ailerons were all hinged at their leading edges to keep this comparison straight up.

Push-pull tubes - They are usually sized by buckling limits. Mine turned out to need 1-3/4 x 0.035 aluminum tube, lost motion due to control circuit deflection is about 0.5 degrees on a 15 degree throw. This will be low friction, and relatively bug free to design and implement, with ratio changes made at idlers and bellcranks if you need it. System will have a number of rod ends in it, so the system should be designed so that air loads take the slack out in nominal flight;

Torque tubes -They are usually sized to get reasonable circuit deflections. I stopped at 3 x 035 aluminum tube (172% of push-pull tube weight), lost motion is 3.3 degrees. These can be modest friction if wing deflections are small and executed properly. Steve Wittman used them. Wing deflections must be low and system must be really excellently executed, or they can be sticky. Ratio changes can be made inside the fuselage if you need them. Cautions about loading rod ends apply here as well;

Cables - Are sized for adequate strength. I needed a 3/16 cable for this job with arms sticking outside the airfoil profile to keep weights low. These usually have numerous pulleys needed and with excellent alignment required of cables to pulleys just to get down to decent friction. Muddy feeling controls can result with cables if much in the way of misalignment or fairleads are needed. I got 4 degrees of lost motion on my 15 degree control throw. The cables themselves are lighter than push-pull tubes (69% of push-pull rods), but the devil is in the details. Many pulleys are sometimes needed, and this can make system weight approach push-pull tubes on weight.

Cables require preload to keep slop down and make the system feel responsive. Ratio changes can be fussy because the net change in cable length over the control throw needs to be kept very small or the system will bind or go slack as the stick moves. Design it to be perfect, then change one lever length, and now it is not so good. Additionally, steel cables change length less with temperature than aluminum and less again than wood or fiberglass. Get into carbon fiber, and well, maybe the cables should be the same material and schedule as the structure. Story is that Paul Allen's heavy lifter airplane - correction - White Knight 2 has carbon fiber rudder cables for just this reason.

Billski

Last edited:

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
Attaching a support to a soft piece of structure can have the same undesirable effect as a long flexible torque tube.
This highlights one of the misconsecptions of good design practice. There is no one single best method. There are no universally specific reasons one chooses a particular method over another. You have to think in terms of systems, not isolated parts. Add in parameters that are external to the actual engineering, as bean counters like to do, and you may end up with a completely different solution.

#### Marc Zeitlin

##### Exalted Grand Poobah
Get into carbon fiber, and well, maybe the cables should be the same material and schedule as the structure. Story is that Paul Allen's heavy lifter airplane has carbon fiber rudder cables for just this reason.
I believe you're conflating WK2 with Stratolaunch.

WK2 was an unboosted aircraft with a 140 ft. wingspan and two fuselages, so all the control surface cable loops were incredibly long. It has carbon fiber "cables" fabricated specially so that the CTE is the same as the structure, and all cables were run as close to the neutral bending axis of the wings and fuselage as possible.

Stratolaunch's control system is basically a cannibalization of two 747-400's (along with 6 of the 8 engines from two 747's). So it's a boosted plane, working pretty much as does a 747.

#### geosnooker2000

##### Well-Known Member
First off, you are attempting to get us to discuss cables vs torque tubes. What about push-pull tubes? Torque tubes are appropriate in few places and are used sparingly by airplane designers for a variety of reasons.

I did a quick analysis of all three as we might apply them to ailerons, using a wing span of 24 feet, chord of 5 feet, ailerons 5 feet by 1.25 feet, airspeed of 180 knots. The ailerons were all hinged at their leading edges to keep this comparison straight up.

Push-pull tubes - They are usually sized by buckling limits. Mine turned out to need 1-3/4 x 0.035 aluminum tube, lost motion due to control circuit deflection is about 0.5 degrees on a 15 degree throw. This will be low friction, and relatively bug free to design and implement, with ratio changes made at idlers and bellcranks if you need it. System will have a number of rod ends in it, so the system should be designed so that air loads take the slack out in nominal flight;
Billski
When I asked the original question, I said "torque tube" when what I really meant was push-pull tubes. I am sorry for my misunderstanding the terminology.

Okay, so for kicks, I checked out where to buy aluminum tube, and how much cabbage..... OMG.
With the caveat that this is according to June 5th 2021, so reading this in the future, keep that in mind:
Aircraft Spruce has short of what you are talking about in 2024-T3 - 10' length, but 1 1/4"dia, (not 1 3/4"dia) = $147 (a 10' stick of their 1 3/4" smallest wall thickness would be over$500)
MetalsDepot has 1 3/4"dia with .065 wall thickness in 6061-T6 - 12' length = $126 Onlinemetals has something I'd like to know more about... "Aluminum Pipe 6061 Extruded T6 Schedule 40", outer diameter 1.9", wall thickness .15", 20'-0" piece =$126, so a 10' section (apples to apples) = $63. I'm sure the "schedule 40" indicates it is for liquid or gas, not structural load, but 6061 is 6061... #### gtae07 ##### Well-Known Member You're comparing 2024 to 6061 and the 2024 is almost always going to be more expensive. If you look at Spruce's 6061-T6 they have a 10ft stick of 1.75x.035W for$125.

I'm not a structures guy but it's my understanding that, for the same geometry, the material property of interest in column buckling is stiffness (not yield strength). Thus, there's no need to use 2024 since it has about the same stiffness as 6061, both being aluminum and all. Bill, am I on track with this?

#### Geraldc

##### Well-Known Member
I'm sure the "schedule 40" indicates it is for liquid or gas, not structural load, but 6061 is 6061...
Schedule 40 relates to wall thickness to diameter ratio.
For pipe work you only need to specify nominal bore and schedule to get the pipe needed. Schedule 80 is thicker wall than 40.

#### Dana

##### Super Moderator
Staff member
When they say "schedule" (usually 40 or 80) it's considered pipe, not tubing, and intended primarily for liquid flow, not structural. Not that it can't necessarily be used for structural purposes, but the dimensions are usually inconvenient, and it's usually too heavy for aircraft work.

#### Heliano

##### Well-Known Member
Billski is right: the devil is in the details. There is no universally better solution. Friction, elasticity and weight depend a lot on how the system is designed.

PUSH PULL RODS: The Embraer Tucano dos not have cables - only push pull rods with ball bearings. Very low friction, you fly it with the tip of your fingers. However push pull rod length is limited by column buckling. Axial forces on a push pull rod are pilot force multiplied by a geometric factor. Can easily reach 600-800 pounds. And push pull rods use up more space. Additionally, the larger the aircraft the more complicated and heavier a push pull rod system becomes.
TORQUE TUBES: Torque tubes must be thoroughly evaluated regarding their natural torsional frequencies when combined with control surface inertia/aerodynamic loads. It is no fun having aileron flutter because of that.
CONTROL CABLES: Cable strength is not an issue. A 1/8" steel cable - used for primary controls in most small aircraft - withstands as much as 2100lb tensile load. As for the friction, adding or not pulleys to change cable direction make all the difference and should be avoided if the change in direction is more than, say, 70 degrees. If that happens, a combination of push pull rod and cable may be the way to go. As for temperature effect on control cables there are fairly simple mechanical devices that can be used to keep them with constant tension. One example that comes to my mind is the Embraer ERJ145, used by American Eagle, United Express, etc. : it has mechanical elevator control, cable-based, with such a device. If a small aircraft whose operational ceiling is no more than say 12000ft and temperature range goes from ISA+25 to ISA -20, normally cable tension is not much of an issue (there are a few cases where it IS an issue, though - again, depending on the design)

#### Map

##### Well-Known Member
When you say 6061 you also have to specify which temper condition. In condition -0 the material is too soft to use as control tubes, but can easily be bent for piping. -T6 is what you should get for structural applications.