Determining the proper preload is critical when designing a reliable joint. There are many methods for measuring preload. However one of the least-expensive techniques that provides a reasonable level of accuracy versus cost is by measuring torque. The fundamental characteristics required is to know the relationship between torque and tension for any particular bolted joint. First, the desired design preload must be identified and specified, then the torque required to induce that preload is determined.
Within the elastic range, before permanent stretch is induced, the relationship between torque and tension is essentially linear. Some studies have found up to 75 variables have an effect on this relationship such as materials, temperature, rate of installation, thread helix angle, and coefficients of friction. One way that has been developed to reduce the complexity is to depend on the empirical test results. That is, to perform experiments under the application conditions by measuring the induced torque and recording the resulting tension. This can be done with relatively simple, calibrated hydraulic pressure sensors, electric strain gages, or piezoelectric load cells. Once this data is gathered and plotted on a chart, the slope of the curve can be used to calculate a correlation factor. This technique has created an accepted formula for relating torque to tension.
T = K × D × P
T = torque, lb-in.
D = fastener nominal diameter, in.
P = preload, lb
K = nut factor, tightening factor, or K-value
If the preload and fastener diameter are selected in the design process, and the K-value for the application conditions is known, then the necessary torque can be calculated. It is noted that even with a specified torque, actual conditions at the time of installation can result in variations in the actual preload achieved.
One of the most critical criteria is the selection of the K-value. Accepted nominal values for many industrial applications are:
K = 0.20 for as received steel bolts into steel holes.
K = 0.15 steel bolts with cadmium plating, which is like a lubricant.
K = 0.28 steel bolts with zinc plating.
It is readily apparent that if the torque intended for a zinc-plated fastener is used for a cadmium- plated fastener, the preload will be almost two times that intended. It may actually cause the bolt to brake.
Friction is also a criteria to consider. For steel bolts, approximately 50% of the installation torque is consumed by friction under the head, 35% by thread friction, and only the remaining 15% inducing preload tension. Therefore, if lubricant is applied just on the fastener underhead, there will not be full friction reduction. Similarly, if the material against which the fastener is bearing, for example aluminum, is different than the internal thread material, such as cast iron, the effective friction may be difficult to predict. These examples show that it important to identify the torque-tension relationship. Typically, the lubricant manufacturer has K-value information.
The recommended seating torques for headed socket screws from Unbrako Engineered Fasteners, Cleveland, are based on inducing preloads reasonably expected in practice for each type. Sometimes a design may require preloads higher than those that are customary, so knowledge of torsion-tension yield and tension capability after torquing is helpful. First, once a headed fastener has been seated against a bearing surface, the inducement of torque will be translated into both torsion and tension stresses. These stresses combine to induce twist. If torque continues to be induced, the stress along the angle of twist will be the largest stress while the bolt is being torqued. Consequently, the stress along the bolt axis, axial tension, will be less. This is why a bolt can fail at a lower tensile stress during installation than when it is pulled in straight tension alone. Research has indicated the axial tension can range from 135,000 to 145,000 psi for industry socket screws at torsion tension yield, depending on diameter. Including the preload variation that can occur with various installation techniques, such as up to 25%, it can be understood why some recommended torque induce preload reasonably lower than the yield point.
When straight tension is applied immediately after installation, there will be some relaxation, and the torsion component will drop toward zero. This leaves only axial tension, which keeps the joint clamped together. Once the torsion is relieved, the axial tension yield value and ultimate value for the fastener will be appropriate.
Ball lock pins |