THE CHALLENGE

The bolted joint presents users and designers with many problems. In part this is because it is

‘‘alive’’—it keeps changing state in response to service and environmental conditions, as we’ll see. A more common source of problems, however, is the fact that the assembly process and the in-service behavior are affected by literally hundreds of variables, many of which are difficult or impossible to control or to predict with accuracy. As a result, when we deal with bolted joints we must inevitably deal with a lot of uncertainty. What follows is a quick review of some of the sources of this uncertainty. We’ll take a closer look at most of these things in later chapters.

FIGURE 1.2 Bolts and joint members deform elastically when the bolts are tightened. In effect, they act like stiff springs as suggested by this sketch. This fact that they act like springs greatly influences the behavior of the joint.

1.3.1 ASSEMBLY PROCESS

Bolts and joint members in both tension and shear joints respond in the same way to the act of tightening the bolts. There are differences in the accuracy with which we must tighten them, but most of the discussion which follows applies to all joints.

As far as all tension and most shear joints are concerned, the goal of the assembly process is to establish an initial clamping force between joint members, to introduce the first energy into bolt and joint springs. And, in tension joints, we’re usually interested not just in tensioning the bolts but in tensioning them by a desired amount, because the life and behavior of such joints are so dependent on the right amount of clamping force. We want enough clamping force to prevent a variety of failure modes, but we must also make sure that the bolt tension and clamping force do not exceed an upper limit set by the yield strengths of the materials, the anticipated loads to be placed on the joint in service, and other factors.

Unfortunately, as already mentioned, hundreds of variables affect the results when we tighten a group of bolts, so predicting or achieving a given clamping force is extremely difficult.

We attempt to control the buildup of clamping force by controlling the buildup of tension or preload in the bolt. In most cases, we do that by controlling the amount of torque applied to the nut or head.

The work we do on the fastener while tightening it is equal to one half the applied torque times the angle (measured in radians) through which the nut turns. Typically, about 10% of this input work ends up as potential energy stored in the joint and bolt springs. The rest is lost in a variety of ways.

Most of the kinetic energy is lost as heat, thanks to friction restraints between the nut and joint surface and between male and female threads. Some energy is used to twist and, often, to bend the bolt a little. Some energy may be lost simply in pulling heavy or misaligned joint members together or dragging a bolt through a misaligned or interference fit hole. More is lost by spreading the bottom of the nut, a process called nut dilation.

A major problem for the designer and assembler is that it is virtually impossible to predict how much of the input work will be lost due to factors such as these. The amount lost can and usually will vary a lot from one bolt to another, even in the same joint.

In spite of these uncertainties and losses, some potential energy is developed in each bolt as it is tightened, and it starts to create some clamping force in the joint. But then the bolt relaxes—loses some energy—for a couple of reasons.

A process called embedment occurs as high spots on thread and joint contact surfaces creep out from under initial contact pressure and the parts settle into each other. More drastically, a previously tightened bolt will relax somewhat when its neighbors in the joint are subsequently tightened. We call this process elastic interaction, and it can eliminate most or even all of the tension and energy created in the first bolts tightened in the joint. We’ll examine this phenomenon in detail in a later chapter.

The amount of relaxation a bolt will experience is even more difficult to predict than the amount of initial tension it acquires when first tightened, increasing the challenge of the assembly process.

Anything that reduces the amount of energy stored in a bolt reduces the force it exerts on the joint. Too little torque, too much friction, rough surface finish, twisting, bending, hole interference, and relaxation all can result in less stored energy, less preload, and less clamping force.

Anything that increases the energy stored will increase the force. There are a couple of things which can do this during assembly: too much torque or too little friction, thanks perhaps to a better-than-anticipated lubricant.

Again, all of these factors are difficult to predict or control, making it very difficult to achieve a particular amount of preload or clamp force at assembly. Because many factors can

give us less preload than desired and only a couple can give us more, we often—perhaps usually—end up with less than expected at assembly.

Bolts in shear joints are subjected to the same assembly problems and variables as are bolts in joints loaded in tension. There’s a difference, however.

In tension joints we always care about the amount of preload, tension, clamping force, and potential energy developed during assembly because of the way such joints respond to service loads. We’re not so concerned about this when dealing with shear joints. We’ll see why when we examine the in-service behavior of such joints.

1.3.2 IN-SERVICEBEHAVIOR

The in-service behavior of tensile joints differs substantially from that of shear joints, and this is reflected in the different ways we design and assemble the two types. Here’s a preliminary look at the differences.

1.3.2.1 Joints Loaded in Tension

We encountered many uncertainties when we assembled a tensile joint. Further uncertainties are introduced when we put such a joint to work—when we load it, expose it to vibration or shock, subject it to change in temperature, anoint it with corrosive fluids, etc. Being alive, it responds to such things; and as it responds, the tension in the bolt and the clamping force between joint members change.

First, and most important, the tensile load on the joint will almost always increase the tension in the bolts and simultaneously decrease the clumping force between joint members.

This is undesirable and unavoidable, and it is the major reason why we care so much about the exact amount of bolt tension and clamping force developed at assembly.

If the assembly preloads are too high, the bolts may yield or break when they encounter the service loads. On the other hand, if assembly preloads are too small, the clamping force on which the joint depends may all but disappear when service loads decrease it.

Other service factors can also change bolt tension and clamping force and will affect our choice of assembly preload. For example, relaxation processes like embedment or gasket creep are increased by loads and by elevated temperatures. Vibration, shock, or thermal cycles can cause the bolt to self-loosen. Differential expansion between bolts and joint members can increase bolt tension and clamping force simultaneously, or can reduce both. In this case, heat energy is being used to increase or redistribute the energy stored in the parts.

Chemical energy, exhibited as corrosion, can increase the clamping force as corrosion products build up under the face or the nut or head of the bolt.

These factors present an additional challenge to the designer. They increase the difficulty of predicting joint behavior, because the designer can rarely predict the exact service loads or conditions the joints will face. The joint’s response, furthermore, will be influenced by such hard-to-pin-down factors as the condition of the parts or the exact dimensions and material properties of the parts. Behavior will also be influenced by the hard-to-predict amount of preload in the bolts, which the designer must somehow specify.

Once again, however, the factors that lead to less clamping force are more common than the ones that can lead to more clamping force. Since this is also true, as we’ve seen, of the assembly process, we are forced to recognize Bickford’s little-known First Law of Bolting:

Most bolted joints in this world are providing less clamping force than we think they are.

1.3.2.2 Shear Joints

Shear loads do not affect the tension in the bolts or the clamping force between joint members, at least until such loads become so high that the joint is about to fail. Predicting behavior and

avoiding failure are therefore easier when we’re dealing with shear joints than when we are dealing with tensile joints. This, in part, explains why people who design airframe, bridge, or building structures rely so heavily on shear joints and avoid using tension joints whenever possible.

I don’t mean to imply that shear joints won’t respond to service loads and conditions; they will. Bolt tension and clamping force will change if temperatures change. Vibration or shock can loosen the bolts, parts can rust, and corrosion products can build up and alter bolt and joint stresses. If the loads on the joint are cyclical, the stresses in bolts and joint members will fluctuate. But the in-service uncertainties the designer faces, and their consequences, are usually less than those he’ll face when dealing with tensile joints.