Keeping It All Together

A three part article in the BMW Owners News, published in 2000.

Owner maintenance is often a topic of conversation when kicking tires and telling lies with my riding friends. The subject of threaded fasteners often comes up, which prompts questions such as:

Through my work as a mechanical engineer and over 25 years of wrenching on things with wheels and wings I have gained insight to these and other related questions. I thought it was time to write it down and share some of this information. My goal is to provide understanding of how threaded fasteners work so that riders that do their own wrenching will do a better job. I hope to also dispel some old wives' tales that tend to flourish during bench racing sessions. I have broken the subject into three parts, the basics of threaded fasteners, fastener types, and torque wrenches. I hope you like it.

By Joseph C. Dille
BMWMOA #24754

Part 1: How a bolt works

Except for a few taper and press fits, our BMW motorcycles are held together with threaded fasteners. Improperly installed fasteners can fail by loosening or through outright breaking, causing a dangerous situation. In this installment, I will describe the basics of how fasteners work and how they should be installed. I will also point out a few simple things to look for when installing fasteners.

A screw thread is an extension of one of the basic machines, the inclined plane that, has been wrapped around a shaft. When the thread is turned, it moves the mating part or nut up the inclined plane. When more turning force, or torque is applied to the shaft, more force is exerted on the nut. This force creates a tension in the bolt, which clamps the mating parts together. Preload is the technical term for the tension caused by tightening the fastener that holds the assembled parts together. Generating sufficient preload force is the key to strong and reliable bolted joints that will not loosen or break under load. Figure 1a shows the forces that act on a bolted joint.

Force Diagram For Typical Bolted Joint
Figure 1a, Force Diagram For Typical Bolted Joint

It is often helpful to think of the fastener as a spring (see Figure 2a). It may seem odd to think of your bike as being held together by a bunch of springs, but this analogy works to show what happens when a bolt is tightened. Rotating the bolt, which in turn stretches the spring, generates the preload force. The more the bolt is rotated, the more it stretches and generates more preload or tension.

A Bolt Acts Like A Spring To Clamp The Parts
Figure 2a

The clamping force Fc is the difference between the preload force and the tension force Ft on the joint. The clamping force is what holds the parts together. This translates mathematically to: Fc=Fp-Ft.

When there are no tension loads applied to the joint, the clamping force equals the preload force. If the tension load is equal to the preload, there is no clamping force. If the tension load is increased beyond the initial preload force, the joint will separate. Even after the joint separates, it will continue to take increased tensile loads until the ultimate tensile strength of the fastener(s) is reached and the fasteners break. As a practical matter, joint failure occurs well before the fasteners actually break because the parts that are being held together will loosen and not function properly.

Joints are loaded with shear force, tension force or a combination of both. In a joint loaded in tension the joint separating forces are opposed by the preload force on the bolt. A good example of this is a cylinder head. It is important to note that for a joint with stiff mating parts the load on the bolt remains constant (at Fp) until the tensile load is greater than the preload force. A simplistic view is that the ultimate strength of the joint is limited by the strength of the bolt. However, the higher the preload force the better the joint, because it will prevent the assembled parts from moving and the joint from loosening. A highly preloaded joint is also more resistant to cycling loads since less of the cyclic portion of the load is experienced by the fastener. In general, the preload force determines the strength of the joint. Joints are stronger and more fatigue resistant with greater preload force.

It is important that the preload force be maintained in the fastener during operation. Highly loaded or critical fasteners tend to be long, and they need to be stretched a relatively large amount to generate the preload force. This allows them to maintain their preload, even if they expand a little or the mating parts shrink. Examples of this include connecting rod bolts, flywheel bolts, K-bike rear wheel bolts, and airhead cylinder studs.

The other type of joint is loaded by shear force (Fs). In a joint loaded in shear, the friction between the parts keep them from moving when subject to a shear force. The friction between the parts carries the load, not the fastener. An example of this type of joint would be a shock absorber mount or the driveshaft flange on an airhead. The greater the preload force, the greater the clamping force, the greater the friction, and the stronger the joint. With a properly designed and tightened joint, the bolt will not experience a direct shear load.

The mating parts also act like a spring, but a much stiffer spring. In the ideal case, the mating parts are much, much stiffer than the fastener. Engine-mounting studs on an airhead boxer are close to the ideal because the engine case and frame are much stiffer than the studs.

Joints with soft gaskets such as the oil pan or valve covers are exceptions to the More-Preload-is-Better rule. High loads can deform the gasket or mating surface, which causes leaks.

Proper preload is the key to reliable bolted joints (see figure 3a).

Who Determines The Strength of A Bolt?
Figure 3a
(Courtesy of SPS Technologies, Aerospace Fasteners Group)

If preload is so important, how is it set? This is the 102,000dm question. The ideal way is to measure it directly with strain gauges or some other force-indicating instrument. This is impractical and unnecessary for all but the most critical applications such as aircraft turbine engines. Tightening the fastener creates the preload. Ideally, the preload is related to the applied torque by the torque-tension relationship:

$$F_p = {T \over D * K}.$$

where:

Fp = the preload
T = the torque
D = the thread diameter
K = a "tightening factor" or "K-value" specific for the assembly conditions

For many industrial and automotive applications, the torque is used to establish the preload. Angle torquing is a more accurate method in which the joint is tightened to a low torque to take up slack in the joint. The fastener is then tightened a specified amount of rotation with the help of an angle indication tool. Angle torquing is sometimes called the "Turn of Nut" method in industrial publications. Special washers are also available that produce an indication when the desired preload is achieved. Since the fastener is essentially a spring, the preload force can be ascertained by measuring the elongation of the fastener. The most direct method is to mount a strain gage on the fastener or mating part to indicate the preload force while tightening. Table 1 gives the Industrial Fastener Institute's estimate of the effectiveness of various preloading methods.

Preload Measuring Method Error
Operator "Feel" +/- 35%
Torque Wrench +/- 25%
Angle Torquing +/- 15%
Load Indicating Washer +/- 10%
Fastener Elongation +/- 5%
Strain Gauges +/- 1%
Table 1

I suspect the +/- 35% variation listed for "feel" is conservative because it is developed in a production environment in which the same wrench is used every time. Wrench type and handle length and shape change the amount of torque applied. For example, I have a 10mm sockets in 1/4, 3/8" and 1/2" drive sizes. I am sure I apply more torque when using the 1/2" drive than the 1/4" drive. It is interesting to note that using a torque wrench gives a relatively small improvement over operator "feel." However, using a torque wrench to set preload is still better than feel.

BMW and other manufacturers specify the torque of critical fasteners. BMW also publishes the DIN/ISO standard for all fasteners not specifically called out in the assembly instructions. However, torque is a relatively poor way of generating preload because the K-value can be greatly influenced by the friction of the threads of the screw or nut on the parts and the friction of the bearing surface of the head. Friction changes with surface finish, material, hardness, and lubrication. Manufacturers are aware of this and use a safety factor when designing assemblies to account for the inconsistency in the K-value.

Attention to detail when reassembling improves the consistency of the K-value and improves the consistency of the preload generated by tightening. Threaded connections should turn freely with no binding when assembling. Test the bolt with the nut or threaded hole to make sure it turns freely. If it does not, inspect the threads for damage. (Bumps on our inclined planes reduce their efficiency.) Light damage can be cleaned up with the appropriate tap or die. Heavily damaged fasteners should be replaced. Dirt and debris in the threads can also cause them to bind. They can be cleaned with spray cleaner and compressed air. Heavily encrusted threads may have to be cleaned first with a tap, or bolt with a flat filed about 1/3 of the way across on one side.

For those readers who may doubt that there can be a +/-25% variation in preload when using a torque wrench, I have experience that shows this to be true. At work, I did an experiment where I torqued 36 screws to the point where they yielded (stretched). I assumed that the preload required to yield all screws was the same because they were from the same production lot from a reputable manufacturer. For the test, I used the same parts, the same torque wrench, the same lubrication, and the same technique. My results gave a perfectly normal distribution of torques with a standard deviation of 7.9%. This statistic indicates that 99% of the screws would yield at the mean torque value +/-23.7%, which is darn close to +/-25%.

Angle torquing is a more accurate way of preloading fasteners. The fastener is tightened to a seating torque to take up the slack and then tightened by turning it a specified number of degrees regardless of torque using a tool to measure the amount of rotation. This method is only used for certain highly loaded critical fasteners where the application warrants the extra time and expense such as with automotive head bolts sometimes use this technique. BMW uses the angle torquing method for highly stressed fasteners such as K-bike connecting rods and R11/12 cylinder heads.

It is important to heed the manufacturer's specification for thread lubrication if provided. One critical area where this is spelled out is in BMW's documentation for the K-bike rear wheel bolts. The bolts should be assembled dry. It may be tempting to add an anti-seize paste to prevent rust and make them easier to remove later. The tightening torque specification is developed using a K-value for dry threads. Lubricating with a high-pressure lubricant like anti-seize reduces the friction, resulting in a much greater preload than the nominal design value. In the extreme case the bolts stretch when tightened to the recommended torque. If anti-seize is used, the tightening torque should be recalculated to account for the lower thread friction.

The threaded fasteners in our motorcycles work by stretching when tightened to produce a tension that clamps parts together. Tightening to a specified torque is the most common way of tightening/preloading fasteners. However, this produces a preload force with a potential for a +/-25% error. The good news is that there is a conservative amount of safety factored into the joint design. Fasteners have a good margin to allow for this variation.

Glossary:

Continue to Part 2, "The Nuts and Bolts of Bolting"