After a mechanical engineering apprenticeship and a period in an aircraft design technical office, Bob Piggott joined TWI's Fatigue Laboratory in 1958. He transferred to the Design Advisory Service when it was established about a year later.
He was instrumental in establishing the on-site stress analysis facility, providing a strain gauging and data acquisition and analysis service. This has involved him in projects ranging from measurements on offshore structures to instrumentation of a bell in a Cambridge church tower.
Bob, who has achieved Chartered Engineer status through part-time study, is now a member of the Structural Design Group within the Engineering Department, with a particular interest in instrumentation and data collection and analysis. More recently he has been involved in a variety of fatigue testing exercises including expansion of the Department's torsion/tension fatigue testing facility.
In an experimental study of fatigue strength of bolted flange joints in television masts, it became necessary to develop a special instrument - a boltmeter. This device, described here by Bob Piggott, was used to measure the pre-tension of bolts around flanged joints in IBA masts.
To ensure that the integrity of the bolted flange joint under study was not impaired it was essential that each bolt in turn could be slackened and retightened to its original condition. This operation had to be carried out at any desired level on the mast using the usual tools available to the technicians for mast maintenance. Any additional piece of equipment therefore had to be light, compact, robust and simple to use.
Measurement of bolt extension by a micrometer, as used by boilermakers in the past, would have been quite impracticable for this application - as would use of extensometers based on levers and dial gauges. The linear voltage displacement transducer could possibly have been adapted but a simple clip-gauge type transducer was chosen for its many distinct advantages. It has the desirable features listed above and is self-supporting because the cantilever arms act as spring clips.
Having adopted the clip-gauge design it became obvious that its shape would resemble a C to fit over the fixed bolt, with sufficient throat to clear the flanges and provide space to engage spanners to release the bolt.
An L-shaped steel frame forms the stiff body of the transducer and the upper arm of the C is a strain-gauged aluminium sensing element as shown in the photograph. The 350 Ω foil-type gauges form a four-arm bridge to sense bending. Adjusting screws with conical tips enable the unit to be mounted on reference indentations (pop marks) on the ends of the bolt and tightened to achieve the desired clamping force.
As with any strain-gauge type transducer, the sensing bridge must be fed with an excitation voltage and the output amplified so that it can be measured and displayed. Today the task of signal conditioning and amplification is made easier because high-grade instrumentation amplifiers designed for such applications are available. Some manufacturers have gone a step further and produced a single encapsulated module which provides bridge excitation, filtering, and simple gain and balance adjustments.
For simplicity, one of the latter devices was used at the heart of the boltmeter indicator. The boltmeter output is presented digitally on a liquid crystal display in engineering units (kN).
All transducers need calibration, and in the case of the boltmeter it was necessary to mount it on a bolt of the same dimensions as those to be measured on the mast. The bolt was then subjected to an axial tensile load and the meter adjusted accordingly. Load was applied using a Morthrust ring nut, which incorporates its own hydraulically-actuated jacking system supplied with oil under pressure from a simple hand pump.
To make measurements on a bolt in service, pop marks are first made in the head and threaded end of the bolt using a centralising guide. The cone-shaped adjusting screws are located in the pop marks and tightened to provide the desired clamping force. The displayed output is adjusted to zero and then the nut is slackened. The new reading displayed is a measure of the bolt pre-tension. Finally it is simple to re-tighten the nut until the indicator display is again zero.
The usefulness of this type of transducer has already been extended to the study of bolt-load fluctuations under service loading. Again the physical dimensions have been tailored to suit the job and in this application the signals are continuously monitored and recorded on magnetic tape for subsequent analysis.
Conclusion
This short article describes a simple and cheap device which can be used to collect valuable information about bolted joints without impairing their integrity. There are likely to be other applications for such a device, not necessarily related to bolted connections.
Further refinements could include a totally-integrated system where the indicator is housed within the handle, and a go-no go gauge with indicator lights which illuminate within pre-set limits for repetitive checking.
