|| Tutorials home | Decreasing risk exposure | Safety tour | Emergencies | Meteorology | Flight Theory | Navigation | Communications ||
Builders guide to safe aircraft materials
Hardware fittings in aircraft structuresContributed by the late Tony Hayes
Rev. 3 — some material added by the tutorials author 24 July 2011
Perhaps more than anything they can determine ease of use of the aircraft, both in servicing and dismantling the machine for transport and storage. They also have a great impact on a 'finished' look to the completed build that will impact greatly on downstream resale value; as well as on-going perception of build quality!
Non-welded aluminium tube structure aircraft (commonly termed 'rag and tube') represent the most significant usage of hardware fittings in terms of variety. The various aluminium alloy tubes that comprise the airframe, normally in a triangulated brace layout, have to be connected together to enable the overall structure to obtain a unity of strength. This is normally achieved by connections that may be generically termed 'brackets', 'hinges' or simply 'fittings'. They come in all shapes and sizes, according to the function they serve. Some are quite complex, needing to be manufactured using computer numerical control [CNC] machines and may serve several purposes, others may be very simple.
There are additional hardware components that provide other functions. These will include joints for trike control frame down-tubes and bars, cable terminal connections for stainless steel flying, ground and bracing cables; mounting points for pulleys, rudder pedals and control columns; actuating arms, pivot points, and others. And, of course, wooden aircraft also require these same fittings.
Most commonly, fittings are machined from stainless steel but sometimes are fabricated from chrome-molybdenum steel, light alloy sheeting, and various proprietary extruded alloy mouldings (a channel cross-section for example) that may be sawn and drilled to the shape required. A lot depends upon the design force the fitting is to undergo and how much movement (therefore friction wear) the fitting is likely to experience in operation.
The following text is chiefly associated with aluminium tube structures, but it is equally relevant to the other types of metal airframe structures, where the hardware (in quite large quantity) is still required — it's the variety that may differ. And, as mentioned above, wooden aircraft use many of the same metal fittings.
It may sound over-simplified, but to put your aircraft together you are going to require connectors, and this is a bit more than just rivets and bolts. In fact, you will be facing several hundred fittings for just a small aircraft type and many of them will be in totally different designs. Just making the fittings from scratch could take equivalent time to building the rest of the aircraft!
As much thought is required for the hardware, its design, materials, load-bearing capability and longevity, as any other of the more obvious main structure.
Some historical pointers
All of the original ultralight types were scratch built, in some cases in the remarkably short space of a few weeks between design inception and first flight. This was enabled because those early builders relied very heavily upon the marine industry for small hardware items, so leaving only the specialist design main fittings to make themselves, as well as the main structure. That is a relatively quick task when dealing with tubular alloy.
A great deal of those original fittings were standard proprietary 'off the shelf' components for yachts. There are many parallel functions in terms of attaching shroud lines and guy wires to masts etc. This in turn led to a lot of marine nomenclature being used for ultralight components.
Wing battens are a form of wing ribbing (once again from marine sail technology) to provide aerofoil wing sections without the weight and complexity of using fixed, constructed wing ribs. These are generally made from light alloy tube but require plastic protectors at each end to avoid wear with the fabric coverings.
The trailing edge plugs are simple and fast enough to turn up on a lathe using plastic or even wood. But the leading edge "duck bills" are a complex shape and, while only worth a few dollars for several, would be an arduous task making 40 or 50 identical examples.
Bushes are important and will typically act as buffer sleeves between main metallic component connections. They may be equally used as a simple bearing on large moving parts. For example, capturing both ends of the control column torque tube or supporting a connecting drive between two elevator halves through a main fuselage boom.
These are simple and cheap enough to make on a lathe but ensure you use inert plastics. Nylon, for example, has water retention properties that can lead to component swelling and potentially corrosion where you cannot see it without a dismantling job. Other plastics are intolerant to exposed positions and may become brittle and start cracking.
Some hardware may be so simple that it is simply overlooked. Open tube ends should be sealed with removable capping. Failure to do this not only detracts from a finished appearance, it attracts unwelcome visitors! Examples are:
(a) A thriving colony of rats living in a 1.2 metre long nest inside a 2.5 inch diameter leading edge wing spar;
(b) Two metres of puzzled brown snake arriving beside a pilot's right ear, in flight, from the root end of a leading edge tube spar;
(c) The constant battle with mud wasp nests in any accessible area, where you can accumulate many generations of nests over a period of time; that eventually does nothing for the aircraft's trim!
Sleeving is normally the same material specification and wall thickness as the main component and internal sleeving will have an outside diameter (o.d.) equal to the inside diameter (i.d.) of the main component. They are mostly internal fittings and are fixed in place usually with light alloy rivets, whose only function is to keep the sleeve in place rather than taking any particular force. The Thruster wing spars, for example, each have four internal sleeves (one is 600 mm long) but all are virtually invisible to outside inspection. External sleeves can be quite significant components.
The main challenge here is there may be no metal specifications available and some of the manufacturing may not suit aeronautical practice. For example, bend radii too tight, poor line-up of load bearing force lines through the fitting and its attachments.
A more practical alternative is to use the now existing recreational aviation manufacturing industry. The design of your aircraft could be made to use approved aviation fittings that are readily available as spares for production types that have some form of certification. This would be the most practical way for the scratch builder.
Bear in mind also that you should consider later in-service use and the potential need for replacement parts from loss, damage or just wear. A very simple small fitting could represent several hours work to make in a well equipped workshop and maybe a long time to source materials!
Good sources of supply for simple structure, fabric and tube aircraft would be the Thruster and Drifter types that are now well time-proven, simple and light. An hour or so with a camera on an existing aircraft would be an excellent investment before you go too deeply into detailed design of your own machine.
The first Thruster Geminis had all aluminium alloy framework tail units. This required some reasonable diameter tubing (for overall strength) some of which had to be bent around curves for appearance shaping, but then it all had to be connected together.
This was achieved by a number of shaped brackets in tee and elbow format — all of which had to be made in exactly the right shape. But this became a game of dominoes! To securely affix the brackets to the tubes, so that the tubes could be solidly connected, numerous rivet holes had to be drilled in primary structure at an airframe area that took repeated pounding from the taildragger configuration of the aircraft.
The outcome was extreme and rapid wear, rivet holes becoming oval, fittings becoming loose and fatigue cracking happening under these fittings, in primary structure that then required a mandatory inspection every 50 hours to keep in check!
The next model Thruster (the TST) switched to a welded chrome-molybdenum steel space frame layout. This was not much heavier (if any heavier), far stronger, and had less profile drag from the smaller tube diameter, and instantly removed the need to make all the hardware fittings!
Getting into welded chrome-molybdenum primary structure may appear daunting — but is it? You only have to source the material, cut out the design, and have a competent airframe welder put it together for you in an hour! That is a long way ahead of making all those brackets!
The main areas are as follows. The front and rear spars of the wings should be dismountable from the fuselage, the lift struts (if fitted) should be a simple connection at each end, and the entire tail unit assembly should be equally easy to remove in component sections. On some types the fin may be a fixed item.
Thought also has to be applied to the control circuits (especially if cable driven). Disconnection of the ailerons and flaps (if fitted) to enable removal of the wings, needs to be simple, accessible, positive and foolproof! Fitting design must be such that it is impossible to make connection and give reversed controls!
Rudder circuits may usually be left entire if aircraft support equipment includes a 'fin bow', a shaped piece of wood that slips over the fin and rudder, preventing rudder movement. These normally carry warning 'flags' so their presence is not overlooked after reassembly; as with any other control chocking device.
Elevator circuits are very easy if a pushrod system is used. A neat solution for a cable driven system occurs on the Thruster models where the entire circuit remains complete and fully tensioned. This requires a substantial hardware fitting turned out of stainless steel that crosses through the rear end of the boom, has a male connector at each end, has a welded set of elevator actuating horns the drive cables remain connected to and is mounted in plastic bushes to enable ease of rotation. The two halves of the elevator have female connectors.
While this seems to be an 'over-engineered' fitting for a simple ultralight, in fact it concentrates function into one fitting and off-loads the requirement for several further hinge fittings for the elevator to the tailplane. The elevators are mounted in a couple of minutes simply by sliding their outer ends onto fixed pins (part of the main structure) and connecting the male/female inboard ends with a nut and bolt each!
Special consideration must be given to aircraft with wing mounted fuel tanks. Removal of the wings will require opening the fuel system in perhaps two places. Additional fuel taps may be then required to prevent drainage of now open tanks; a risk of those taps then inadvertently being left 'off' at reassembly; and the possibility of fuel leakage at fuel connection joints, particularly after repeated connections.
Although this section is not about threaded fasteners, in general terms you should plan your hardware fittings so, while the majority will be firmly riveted or bolted in place, those main components listed above should be easily accessible and affixed by clevis pins and safety pins, or, via bolts with castellated nuts and safety pins. Either of these systems enable rapid dismantling with a few very simple tools.
Primary structure connections
Bracing/flying wire connections and tensioners
On most simple ultralights these are usually tangs (flat or shaped small plates of stainless steel) to provide interconnection between fixed structure and correctly fashioned swaged loops at the ends of cables. Normally small 'D' shackles with clevis pins make the fastening.
Cables may be manufactured to length to automatically achieve required tensions when connected or alternatively a simple tensioning device may be introduced. This is normally a U shaped metal fitting with a square bottom to the U. A threaded eye bolt, with lock nut, goes through the bottom of the U, enables cable connection and effects the tension required. The open ends of the U are affixed to main structure via a transverse clevis pin.
Engine, exhaust and cooling system mounts
The main engine mounts may be regarded as part of the aircraft primary structure. However, when two stroke motors are employed it is essential that flexible mountings are also introduced to absorb vibration from the motors.
Common parts are sourced from the automotive industry that are used to insulate the engine from the main frame and are made to suspend exhaust systems from on vehicles.
Similar, lighter, fittings are used to insulate mufflers and liquid cooling systems (radiators especially) from main rigid structure.
Fuel tank restraint
The main support for the fuel tank(s) will normally be main structure. Often fuel tanks are bolted directly to this structure and become a rigid part of the overall airframe. With ultralights however 'barrel tanks' are popular. These will again usually be supported by main structure underneath and forward of the tank, but also they have to be restrained in position. So we are looking at yet more hardware requirement!
This is normally achieved by metal restraining flat straps or suitably padded cable (to avoid cutting into the tank over time). Some tensioning method is further required to make the tank totally located in position. It is worth thinking about that in terms of the actual design strength of hardware (and support structure) you require!
A full, 75 litre barrel tank contains 50 kg of fuel, so it is heavy to start with! Your aircraft may be designed to withstand (say) 10g deceleration forces in a heavy landing/crash scenario. That means you are restraining half a metric tonne in the fuel tank and contents! Maybe worth thinking about that in terms of strength of supports, restraints and material specifications — then do a few calculations!
Instrument panels and instrument related hardware
Some hardware components are required for the instrument system. Most importantly the instrument panel(s) should be mounted on flexible supports to absorb engine vibration and airframe shocks when moving on the ground. Modern instruments are very delicate devices and will rapidly wear internally if not given a suitable environment.
You will additionally require pitot and static heads (that are both hardware items) and if you are contemplating the increasingly popular fuel monitoring and economy devices they you may need space and additional hardware mount requirements for some of their sub-assemblies.
In addition to the actuating fittings there are various forms of interconnection fittings plus the airframe mounting structure to provide secure pivot points for the mechanism inclusive of any required bearings or bushes. In situations where a bolt shank acts as a pivot it is standard practice to insert a replaceable metal or plastic bushing between the bolt shank and the structure to avoid wear in the structure.
Cables or wire ropes are made by laying-up (twisting or winding) a number of single wires into strands, then laying a number of strands to form a cable. If the wires are relatively thick a single strand is often used as a cable. The strength and the flexibility of aircraft cables are dependent on the following.
Cables are typically made up as:
Pivots, pins, guides and cable keepers
The canny builder will make provision for pivot and pin fixing as part of the main structure as much as possible. This is normally only practical in space frame welded or composite (e.g. GRP – glass-fibre reinforced polymer) structures. Metal framework monocoque, wooden and alloy tube will require a doubling up of hardware items. In this case you have to make a fitting that you can attach to the main structure, and that fitting will hold the pin or attachment point where the actual connection will be made. This area alone is worth a lot of thought for time, labour, complexity, maintenance and aesthetic reasons.
Guides (commonly referred to as fairleads) ensure that long control runs are kept in position and do not foul other structure (causing unwanted wear) or most importantly other control connections. These are usually light and simple plastic fittings that can be attached to main structure where appropriate. They are also of a less resilient material than the main control so any friction wear is absorbed by the fitting and not the control.
Any pulleys used on the aircraft require keepers. This will be effected by either a box type design of the pulley mounting fitting or by using (normally) light stainless steel rods that go through the mounting fitting. The objective is to ensure that there is insufficient physical space for a cable to exit the pulley groove and that the cable (even if slack for some reason) will stay around the pulley.
Control column torque tubes and bushing/bearings
This is borderline between main structure and hardware. The torque tube is actually hardware as it is a moving part, with a number of hardware sub-components, and can be completely removed without influencing main structure.
It is often a tube with mounts for the control column to enable it to be rocked side to side for aileron control via attached horns or arms on one end of the tube. Inside the tube is the provision for fore and aft control column movement actuating the elevator circuit. The tube is held in place by fore and aft mounted bushes or bearings to enable easy sideways rotation.
Sounds complicated (and is in terms of what it actually does) but they are reasonably simple devices.
1. A bad idea #1! The front tailplane mounting connection on the early Thrusters is a spring loaded pip pin. Easy and quick to use, totally secure etc, they have an obvious pull ring on the end that invites a tug to see how it works! A curious bystander may have a go and leave your unattended aircraft with a disconnected tailplane that you may potentially miss on a pre-flight inspection! So these then have to be wire locked on these types to prevent that happening.
2. A bad idea #2. Again the Thruster types. Most have a very small multi purpose fitting on top of the fin. This provides the top rudder hinge pin and two tail plane bracing wire connections. It is stainless steel, very small, and if you drop it in long grass you will lose it. You then will have to make another (when you have the materials and welding facility) but until then your aircraft is useless.
3. A good idea. A very simple example of brilliant engineering was on the Glasflugel sailplanes with T tails. This required the elevator control circuit to transit 90 degrees from the horizontal fuselage and up the vertical fin to the elevator. This was effected by a simple stainless steel ring. pivoted at a diagonal, so push and pull forces from the main control push rod were translated into vertical up and down forces. Lovely piece of engineering!
4. A fatal idea. An American kit built metal sailplane relied on flap interconnection at assembly on a sort of socket that captured the root end of the flap for actuation. This was strong enough, if fitted right. A less than desirable 'build' resulted in full union not being achieved and the flap root came out of the socket, on one side only, under flight loads of deployment of full landing flap. With asymmetric flap then deployed the aircraft rolled straight onto its back and dived vertically into the ground from 200 feet, killing the test pilot. The builder is still traumatised by the situation over a quarter of century later!
Put care, thought and sound design into your hardware fittings. They are as important as your main spars and inherent stability etc! Too often very nice designs have been 'finished off' too quickly with almost bodged up connections that obviously look so.
These are not simply potentially dangerous — they could totally undermine hundreds of hours of careful main structure build and then maybe an expensive paint job. The careful pilot and airman will look beyond aesthetics to functional practicality. All those hardware fittings actually make the aircraft function!
Some photographs of fittings in the Thruster and Vision series of aircraft follow.
Copyright © 2005 Tony Hayes
Main undercarriage legs on a Vision V600T showing separate suspension strut and shock absorber. Wheel brake and mountings are out of sight. Contrasts in connection here! The heavy load bearing undercarriage uses entirely AN bolts while the access panel to the pod is attached by light alloy rivets. An often removed panel (e.g. an engine cowling) would use Dzus type spring locked twist fasteners for easy pre-flight inspection and access.
V600T three piece front spar mount assembly and the main longitudinal boom collar fitting to capture the aircraft's front A frame. Note that the front and back spar mount plates have been angle drilled to give the correct design angle of incidence for the wings. To the right is the large and complex engine mount to take the Jabiru 2200 engine. This important load bearing structure is affixed by AN bolts and heavy stainless steel blind rivets are used where access is not possible to attach nuts. The engine mount is also affixed by stainless steel blind rivets although commonly one or two of these may be replaced by nuts and bolts at the open boom end for use in electrical earthing connections.
Example of a multipurpose fitting. This captures the front of the fin, front of both tailplanes, both the boom rear support struts, fairleads for the rudder drive cables, and also provides fixing for the shoulder harness restraint cables. Most of the tail unit leading edge points are located by fixed male/female fittings that require no connectors. Note the double bolted support spars emerging from the bottom of the picture. The main fitting sleeve is connected to the boom by heavy stainless steel rivets – generally all rivets used with tubular structures must be the blind type rather than solid rivets because there is no access for a bucking bar.
Typical light tailwheel assembly of spring; tailwheel yoke with steering arms; steering arms on the rudder, steering connections and springs. Also lower tailplane bracing wires with turnbuckles and connection fitting to fuselage. Tailwheel assemblies are subject to considerable forces and potential wear. No room for rivets here! Most of this assembly is well designed 4130 chromoly welded framing and AN bolts used for the remaining few connections required.
A lot in the V600T cockpit to be made or acquired. Two piece rudder pedal assembly, drive arms and mounts; rudder circuit tensioning bungees; front of rudder cables; pulley fairleads; reverse pulley for elevator circuit; control torque tube including control column mount; control column; hydraulic brake master cylinder and mount; brake actuating lever; two brake hydraulic lines; stick grip. Lower centre of photo is the top of a bracket that gives support to the pod undersurface.
The cockpit area contains just about the whole spectrum of connector types. These range from AN bolts on the primary control system and many clevis pins or bolts on cable connections. Rivets used extend from heavy duty stainless steel employed to secure the rudder pedal mounts, while much lighter alloy rivets are used for guides that are not under load.
The Jabiru 2200 is a fair lump of motor to dangle on the end of an alloy tube and requires a well designed and built mount (see earlier photo). Note the simple square U stainless steel mounting bracket for the engine port support strut.
Note here that the front A frame on the right of the picture is secured by two AN bolts as this needs to be kept rigid. In comparison the engine support struts in the centre of the picture are captured by single bolts. This allows some movement in the structure to counter for torque and landing load flexing of the boom.
A simple two piece wing rear spar mount. A plate riveted to the boom and a U shaped mount that is centrally bolted through the boom. This allows the mount to swivel and give perfect alignment to the wing angle of incidence. The rivets used to secure the mounting plate to the boom are heavy duty stainless steel as this is very definitely primary structure.
Fuel tank restraining cables, D shackles and tensioners. A cable thimble end can be clearly seen. Again note the simple stainless steel square U bracket for the front capture of the boom rear support strut. Some more fixing contrasts: Pod components are connected with light alloy rivets as there is little load and maintaining shape is the requirement. On the other hand you can see that the pod is very firmly bolted to the black space frame bulkhead to control loads on the pod as an entity.
Wing tip former. Complex to build but then serves as a permanent jig to periodically check profile of the battens this type of wing uses. The mounting brackets are pre-formed alloy channel mouldings cut to size and shape required. Not a huge amount of loading out here but it is primary structure. So the mounting brackets are bolted through the entire spars and the wing tip former is also attached with bolts.
Multipurpose wing bracket securing a compression strut and the outboard end of the (presently folded) drag spar, plus one end of an internal bracing wire. Note the plastic tube on the drag spar to prevent abrasion when the wing is dismantled and folded. Also to be seen is the top of the sleeve fitting that picks up the front of the jury strut and the front flying wire of the wing.
Contrasts once more. The main components are single AN bolted to allow some structural flexing; the main fitting is bolted through the spar; the rivets visible are light alloy as their purpose is just to locate internal sleeves and they bear no loading.
A small but neat and robust fitting for the lower end mounting of the lift strut assembly. Note also to the right an alternative to internal sleeving; a spacer tube has been installed in the root of the rear lift strut to prevent tube crushing from over-tightening. Top of the photo is the inboard end of the top plate connecting the rear and front lift struts.
The designer has spared no effort here as this is a critical primary structure area. The lift strut cores are both riveted (stainless steel) and firmly bolted to the aerofoil section struts. Additionally a bracing fish plate has been added that is also bolted to the assembly; including at the other end which is out of view.
Two alternate (but similar) methods of attaching lift struts to wings. The photo above is from earlier Thrusters and is a cup and flange bracket. The later T600 and Vision 600 brackets (below) are full collars but of generally similar concept.
The next module in this metals and hardware group is 'AN, MS hardware — rivets, bolts and locking devices'
Builders guide to aircraft materials – metals and hardware modules
| Guide contents | Properties of metals | Metal corrosion | [Hardware fittings in aircraft structures] |
| AN, MS hardware ? rivets, bolts & locking devices | Safetying |