Hole Drilling Techniques for Measuring Residual Stress
As referred to in a recent Technical Tip on ‘perceptions’ and easy observation of features which assist in Failure Analysis, one of the phenomena encountered is that related to Residual Stress. For current purposes, residual stress may be regarded as that stress that remains in a component or structure when it is in the unloaded condition. It may also be regarded as that stress that is intrinsic in the component, due to its fabrication or loading history, and is essentially in self-equilibrium. Familiar examples in everyday life include distortion of components after welding, toughened glass (as used in vehicle windscreens and windows), pre-stressed concrete/reinforcing bars in civil structures, and carburised and nitrided gear surfaces, which have an induced compressive stress surface layer to resist fatigue crack initiation. Although residual stress induced by carburising and shot peening are typically beneficial, generally residual stresses arising from fabrication are deleterious to component performance, mostly through premature fatigue initiation or stress corrosion cracking (SCC).
There are various techniques for measuring residual stress in components, the most important of which are dependent on either i) diffraction, or ii) stress relaxation (and more specifically centre-hole drilling). In a metal, diffraction techniques i) rely on the physical deformation of the array of atoms in the crystal micro-structure being deformed under local stresses, and hence the reflected diffraction pattern obtained using X-rays (or a neutron beam) are altered. These changes in reflected angles can be used to determine the relative extent of strain as well as direction of distortion, using Bragg’s law, but needs to be scanned from 0 to 90⁰ to the surface, as well as over the full 180⁰ rotationally for each location point and for reliability. Such diffraction techniques are complex and delicate to perform and provide only a relative measure of deformation from the norm, so hence also require a diffraction test for unstressed material as a calibration standard. With X-rays one can only detect stresses in the first few microns of the surface, making specimen preparation crucial to ensure one is truly measuring the intrinsic inherent residual stress and not the effect of specimen preparation through grinding or polishing.
A preferable diffraction technique relies on neutron diffraction, rather than X-rays, where neutrons can easily penetrate several millimetres in steels, and so is less susceptible to surface preparation, and can provide a better averaging measure of the residual stress. However this technique, in turn, requires a neutron source (or nuclear reactor) and normally one has to bring the component to the neutron source/reactor, which is not always convenient or possible. The diffraction equipment is expensive and sophisticated but with care yields excellent measurements of residual stress.
The best of the stress relaxation techniques, are the ‘hole drilling’ techniques in which a small hole is drilled (using a high-speed turbine or air abrasive techniques) into the surface of the component and the changes in strain caused during the hole formation are recorded using strain gauges. Even though the hole drilled into the components is only typically a millimetre across and about as deep, the techniques should be viewed as semi-destructive. The strains relieved during the formation of the hole are measured using a strain gauge rosette (with individual gauges typically aligned at 0, 45 and 90 degrees) where the strain gauge portion is seldom more than a couple of millimetres in length and the whole gauge rosette, including gauges and solder tabs, is typically in the region of 6 by 16 mm (see Figure of a typical residual stress hole drilling strain rosette). The strain gauges themselves are arranged to point toward a common origin, or focus, and it is there that the hole is drilled. A crucial part of the technique is that the positioning, both of the gauge rosette, and particularly the drilled hole introduced, needs to be precisely at this focus point, within an accuracy of a few microns. Furthermore, the hole needs to be perpendicular to the surface, straight sided and must not introduce additional stresses due to its formation (hence the requirement for using a high-speed turbine and air abrasive techniques). As the hole is drilled, the hole (being a hole) can carry no stress, and so the strain in its vicinity redistributes and the changes are detected and recorded as a change in strain by the strain gauges. By interpreting the change in strain (often as a function of depth) recorded by each of the three strain gauges, one can readily determine the principal residual stress that was there before the hole was introduced, both in terms of magnitude and also direction. The location of the drilled hole, relative to the focus of the gauges, is of vital importance, as an ‘off centre’ hole will result in the gauges detecting disproportionate strain changes, governed more by hole proximity and less by the intrinsic residual stress, and additionally, the mathematical interpretation becomes progressively incorrect. Although setting up to a few microns accuracy is challenging, and usually achieved using an optical microscope, the technique is robust and extremely convincing, and is one of the most reliable and accurate residual stress measuring techniques. Undertaking an array of measurements, in judiciously chosen locations, can be used to map the distribution of the residual stress field.
In special symmetric applications, such as drill pipe tubes, where the residual stress is predominantly circumferential, realistic average strain measurements can be obtained using a ring splitting technique, together with an accurate measure of the diameter, before and after ring splitting. Similar destructive techniques can be used with trepanning methods and machining away sections, but care is needed in interpretation.
As residual stresses are often difficult to quantify using analytical techniques, they are often simply ignored even though they can reach yield stress magnitudes. Residual stress often causes/contributes to failure and, in critical situations, cognisance of their potentially detrimental consequences needs to be included in design. Owing to the complexity of accurate analysis, such analyses should be backed up by measurement of the true residual stress developed on the actual component (or models that simulate manufacture techniques). As illustrated in this Technical Tip accurate quantification of the residual stresses is indeed possible by relatively simple techniques.
Published in Technical Tips by Origen Engineering Solutions on 1 May 2018