Last month, we took a look as some issues that can arise from poor engineering practice in kit and homebuilt airplanes. This month, we turn our attention to one specific area where poor design practice can lead to dangerously-flawed structures.
The structure of an airplane and its systems are only as strong and reliable as the joints holding its components together. The strength of a joint is dependent not only on the properties of the fasteners, but on the design of the joint itself. The way the fasteners are loaded, the strength of the parts the fastener is holding together, and the stability of the joint itself are all important.
There are several common mistakes made in joint design. Most of these result either from a lack of knowledge of the behavior of the materials near the joint or the desire to keep things simple.
A few of the more common joint-design mistakes are:
Rivets through composites:
When a rivet is set into a hole, the rivet body expands to fill the hole tightly. The expanding rivet body puts pressure and stress on the sides of the hole. In metal this is not a problem. The metal is ductile and yields a little around the rivet, giving a tight rivet fit without significantly weakening the metal.
The rivets that hold the piano hinge to this fiberglass cowling have probably expanded within the composite itself. Though lightly loaded, the hinge material serves as a backup to prevent pull-out of the rivets.
Unlike metals, composites are brittle rather than ductile. In spite of their high tensile and compressive strength, they have relatively low bearing strength and cannot withstand a lot of direct pressure. Often the yield stress in bearing of a composite material is less than half of its yield stress in pure tension or compression. When a rivet is set into a hole in composite, the expansion of the rivet body will often cause cracking in laminate around the walls of the hole, weakening the material. Such a joint is a failure waiting to happen. Rivets through composites are used in the “big airplane” industry, but even when installed under carefully controlled conditions, with precision machinery, they tend to be a problem.
Where rivets through composites are useful is in adding a measure of protection against peel failures of bonded joints.
If a fastener, particularly if it is load-bearing, must be put through a composite part, a bolt or other close-fitting, non-expanding fastener is appropriate. Even so, the designer must be very aware of the fact that the bearing strength of the composite material against the fastener is dramatically lower than the basic tensile or compressive strength of the composite material. To safely take much load on a bolt through a composite part, a large metal bushing should be molded into the composite part to properly distribute the bolt loads into the composite material.
Single Bolts in Single Shear:
There are some joints in airplanes which must be held together with a single bolt. Some examples of these are strut-end attach points and bracing-wire attach points.
There are two types of bolted or pinned joints. The first is a single-shear joint, where the bolt goes through each of the parts to be joined only once and is loaded in shear at one point along its length. The bolt is tightened to clamp up against the two parts being joined.
This aft spar attach point on the Van’s RV-1 is a good example of a bolt loaded in double shear. The single wingspar fits into a yoke welded to the fuselage, and a bolt runs through all three layers.
The second type of bolted joint is a double-shear joint, where one part forks around the other and the bolt goes completely through both the fork and the part within the fork. In this joint, the bolt is loaded in shear at two points. The bolt in a double-shear joint may be clamped up, but it is not necessary for the stability of the joint.
While both types of joint can be used safely in some applications, there are some significant problems with single-shear joints; they should be avoided in highly-loaded and flight-critical areas wherever possible.
The first problem with single-shear bolt installations is that the bolt carries all of the load at the single-shear plane. If the same bolt were installed in double shear, it would be able to take twice the load because the load is shared between the two shear planes in the double-shear joint. This, in and of itself, is not dangerous as long as the bolt is properly sized to take the loads it must carry.
The second problem is more severe, and is the primary reason single-shear joints should be avoided in critical areas. A double-shear joint is stable. If the nut loosens, and the bolt moves in the hole, a double-shear joint remains in place, and the internal loads in the joint do not change. The joined parts do not move relative to each other because one is captured by the fork in the other. As long as the bolt still goes all the way through the fork, the strength of the joint is not compromised and nothing shifts undesirably.
This attach bracket for a linear flap actuator is a good example of a bolt loaded in double shear. The actuator attach bolt is much more stable running through two fingers of the mounting bracket, considering the twisting load from the actuator.
A single-shear joint, on the other hand, is dependent on the clamp-up of the nut and bolt head for stability. If the nut loosens, the two bolted-together parts can move apart. If a gap forms between the two parts, the load carried by the bolt is no longer carried in pure shear. The gap provides a lever arm for the forces on the bolt, which then exert bending moments on the bolt and the bolted parts. These moments can bend both the bolt and the bolted parts. The bending causes the bolt to upset and tend to align itself with the forces on the joint. This can cause the gap to open further, and thus increase the moments. This situation is unstable, and will often lead to large distortion, or failure of the joint. Both are highly undesirable, and potentially dangerous.
Single-shear attachment of bracing wires seems to be a relatively common error, particularly among the designers of ultralight and low-performance light airplanes. It is very simple to attach a wire to a tang and simply bolt it to the hard structure. The airplane I mentioned last month as having multiple potential single-point failures in its tail had the tail bracing wires attached this way. Even more frightening was the fact that the tail was not only braced by these wires, but held onto the airplane by tensioning them. A small loosening of the bolt in the single-shear tail-wire attach joint on this airplane could cause the bracing wires to loosen and lead to the departure of the tail.
The proper way to attach bracing wires and control cables is to put a fork on the end of the wire and attach the fork to a tang on the structure or the control horn with a properly-safetied bolt or clevis pin in double shear. Strut-attach and wing-attach fittings should always be designed so they bolt together with the bolt in double shear.
Tear-out and Pull-out:
A joint is only as strong as its weakest component. If a strong fastener is placed in a weak hole, the joint will fail by either tearing the fastener and a piece of material out of the part or by pulling the fastener through the hole.
The strength of a riveted joint in sheet metal is often determined, not by the shear strength of the rivet, but by the shear strength of the material of the riveted-together parts. It does no good to use a rivet with a shear strength of 300 pounds if the material the rivet is holding tears at a load of 100 pounds. When designing riveted joints, it is vital to consider both the shear strength of the rivet and the tear-out strength of the parts being fastened together.
Some of the most critical bolts on an RV are the ones that hold the horizontal tail to the longerons. A strict observance of edge distance requirements is a must to assure that nothing pulls out under load.
The distance between rivets (rivet pitch) and the distance between a rivet and the edge of the material being riveted (edge distance) both affect the strength of the joint. If the rivet holes are too close together or the distance between a hole and an edge is too small, the tear-out load of the rivet hole will be reduced and the joint weakened. The minimum desirable rivet pitch and edge distance varies somewhat, depending on the details of the materials being used, material thickness, and rivet characteristics. Information on proper rivet spacing is available from a variety of sources, including manuals provided by EAA and any one of a number of engineering or aircraft maintenance and repair handbooks.
Another weak-hole problem can arise if a fastener is loaded in tension, rather than shear. The weakest point of a joint with a tension-loaded fastener is often the area where the fastener head bears on the material of the joint. If the material is thin, or the fastener head is small, the fastener head will pull through the hole at a load much lower than the tensile strength of the bolt itself. Tension bolts should go through large washers or backing plates to properly distribute the loads into the bolted-up material. This is particularly important for composite structures. As we have already discussed, composites have low bearing strength. If proper care is not taken to distribute bolt loads, the bolts will pull through at unexpectedly low loads.
Homebuilding airplanes is inherently experimental, particularly for those of us who build original designs. Despite this, it is neither necessary nor desirable to experiment in safety-critical areas where well-known, safe practice has already been defined. The time spent ensuring that an airplane is properly engineered is important and may well save a life.
Barnaby Wainfan is a principal aerodynamics engineer for Northrop Grumman’s Advanced Design organization. A private pilot with single engine and glider ratings, Barnaby has been involved in the design of unconventional airplanes including canards, joined wings, flying wings and some too strange to fall into any known category.