One of the major problems with the use of bolted joints is the precision, with regard to achieving an accurate preload, of the bolt tightening method selected. Insufficient preload, caused by an inaccurate tightening method, is a frequent cause of bolted joint failure. It is important for the Designer to appreciate the features and characteristics of the main methods employed to tighten bolts. Presented below is a brief summary of the major bolt tightening methods. Note however that whatever method is used to tighten a bolt, a degree of bolt preload scatter is to be expected.
There are six main methods used to control the preload of a threaded fastener. Specifically:
1. Torque control tightening.
2. Angle control tightening.
3. Yield controlled tightening.
4. Bolt stretch method.
5. Heat tightening.
6. Use of tension indicating methods.
Torque Control Tightening
Controlling the torque which a fastener is tightened to is the most popular means of controlling preload. The nominal torque necessary to tighten the bolt to a given preload can be determined either from tables, or, by calculation using a relationship between torque and the resulting bolt tension.
When a bolt is tightened the shank sustains a direct stress, due to the elongation strain, together with a torsional stress, due to the torque acting on the threads. Most tables of bolt tightening torques ignore the torsional stress and assume a direct stress in the threads of some proportion of the bolts yield stress, usually 75%. For high frictional conditions the magnitude of the torsional stress can be such that when combined with the direct stress, an equivalent stress over yield can result, leading to failure. A more consistent approach is to determine the magnitude of the direct stress which, when combined with the torsional, will give an equivalent stress of some proportion of yield. The proportion commonly used with this approach is 90%.
Torque prevailing fasteners (such as Nyloc, Cleveloc nuts etc.) are often used where there exists a risk of vibration loosening. The prevailing torque has the effect of increasing the torsional stress in the bolt shank during tightening. This affects the conversion of the tightening torque into bolt preload and should be allowed for when determining the correct torque value for this type of fastener.
As can been seen by study of the above chart, a fundamental problem with torque tightening is that because the majority of the torque is used to overcome friction (usually between 85% and 95% of the applied torque), slight variations in the frictional conditions can lead to large changes in the bolt preload. This effect can be reduced by the use of so called friction stabilisers. These are substances which are coated onto the fasteners to reduce the frictional scatter. Other ways to improve the accuracy of the method are:
1. Do not use plain washers; their use can result in relative motion to change from the nut to washer, to washer to joint surface, during tightening. This as the effect of changing the friction radius and hence affects the torque-tension relationship. If, because of excessive bearing pressure, a larger bearing face is required, thought should be given to the use of flanged nuts and bolts.
2. Determine the correct tightening torque by the completion of tests. Strain gauges can be attached to the bolt shank and tightening completed on the actual joint. A load cell under the bolt head can be used, however it is not as accurate as strain gauging, since the joint characteristics have been changed.
3. If it is not feasible to establish by testwork the actual tightening torque, determine the tightening torque using the best information available i.e. fastener finish, nut head bearing surface size and prevailing torque characteristics, if applicable. (The computer program TORQUE developed by Bolt Science can allow for all these effects.)
4. Ensure that the tightening torque value is specified on the assembly drawing. Quotation of a plus or minus 5% tolerance is good practice. More unusually, quote that a calibrated torque wrench is to be used to check the torque after installation. The method used to tighten the bolt has a significant influence on the preload scatter (see below).
Angle Controlled Tightening
This method, also known as turn of the nut method, was introduced for manual assembly shortly after the second World War when a certain tightening angle was specified. The method has been applied for use with power wrenches, the bolt being tightened to a predetermined angle beyond the elastic range and results in a small variation in the preload due, in part, to the yield stress tolerance. The main disadvantages of this method lie in the necessity for precise, and, if possible, experimental determination of the angle; also the fastener can only sustain a limited number of re-applications before it fails.
Yield Controlled Tightening
This method, developed by the SPS organisation, is also known under the proprietary name "Joint Control Method". Very accurate preloads can be achieved by this method by minimising the influence of friction and its scatter. The method has its roots in a craftsman's "sense of feel" on the wrench which allowed him to detect the yield point of the fastener with reasonable precision. With the electronic equivalent of this method, a control system is used which is sensitive to the torque gradient of the bolt being tightened. Rapid detection of the change in slope of this gradient indicates the yield point has been reached and stops the tightening process. This is achieved by incorporating sensors to read torque and angle during the tightening process. Since angle of rotation and torque are both measured by the control system, permissible values can be used to detect fasteners which lie outside their specification (having too low a yield for example).
A small degree of preload scatter still results from this method due to the influence of friction. The method detects the yield point of the fastener under the action of combined tension and torsion. The higher the thread friction, the higher the torsional stress, which, for a given yield value, results in a lower preload due to a lower direct stress.
The method has been used in critical applications, such as cylinder head and conn-rod bolts, in order that consistently high preloads can be achieved (which can allow smaller bolts to be used). However, because of the cost of the tools necessary to use this method (a hand wrench incorporating the control circuitry costs many times more than a conventional torque wrench), widespread adoption of this method is unlikely. (Although manufacturers may be able to invest in the equipment, unless service staff have similar equipment, the Designer cannot depend upon high preloads being maintained in the field.)
Bolt Stretch Method
A problem relating to the tightening of large bolts is that very high tightening torques are required. Although this can be partly overcome by the use of hydraulic torque wrenches (the reaction of the torque however can be a problem), the use of hydraulic tensioning devices is commonplace for bolts over 20mm in diameter. The method uses a small hydraulic ram which fits over the nut, the threaded portion of the bolt/stud protrudes well past the nut and a threaded puller is attached. Hydraulic oil from a small pump acts upon the hydraulic ram which in turn acts upon the puller. This is transmitted to the bolt resulting in extension occurring. The nut can then be rotated by hand with the aid of an integral socket aided by a tommy bar.
Control of the hydraulic pressure effectively controls the preload in the bolt. A small amount of preload reduction however does occur when the pressure is removed as the nut elastically deforms under the load. Removal of nuts corroded to the bolts can be a problem with this method.
Heat tightening utilises the thermal expansion characteristics of the bolt. The bolt is heated and expands: the nut is indexed (using the angle of turn method) and the system allowed to cool. As the bolt attempts to contract it is constrained longitudinally by the clamped material and a preload results. Methods of heating include direct flame, sheathed heating coil and carbon resistance elements. The process is slow, especially if the strain in the bolt is to be measured, since the system must return to ambient temperature for each measurement. This is not a widely used method and is generally used only on very large bolts.
Tension Indicating Methods
This category includes the use of special load indicating bolts, load indicating washers and the use of methods which determine the length change of the fastener. There are a wide number of ways bolt tension can be indirectly measured and the discussion presented here is not exhaustive.
Special bolts have been designed which will give an indication of the force in the bolt. One such fastener is the Rotabolt which measures bolt extension by the use of a central gauge pin which passes down a centrally drilled hole in the bolt. Underneath the head of the gauge pin, a rota is retained which is free to spin in a very accurately set gap. The fastener stretches elastically, whereas the gauge pin does not move since it experiences no load. As tightening continues, the bolt will stretch sufficiently to eliminate the gap and prevent the rota from being able to be rotated. This is the indication that the bolt is correctly loaded. Another proprietary fastener uses a similar method. The HiBolt uses a pin located centrally down the bolt as does the Rotabolt except the pin is gripped by the slight contraction of the bolt diameter; the pin being locked when the correct preload is reached.
The use of load indicating washers is widespread in structural engineering. Such washers have small raised pips on their surface which plastically deform under load. The correct preload is achieved when a predetermined gap is present between the washer and the underhead of the bolt. This is measured using feeler gauges. Generally they are not used in mechanical engineering, but are, extensively, in civil engineering.
The extension which a bolt experiences can be measured either using a micrometer or by a more sophisticated means such as using ultrasonics. The extension can be related to preload either directly, by calibration, or indirectly, by calculation. If ultrasonic measurement is used then the end of the bolt shank and the head may require surface grinding to give a good acoustic reflector.