NDT - Residual Stress Analysis
Confirm the Quality of Surface Enhancements
October 17, 2012
We find more and more parts in high-tech applications with surfaces modified to enhance some aspect of the part’s performance. This surface enhancement is likely to be performed by a specialist department or sub-contractor, and in many cases the new surface does not observably reveal the degree of modification. In other cases the surface may have received a further treatment that removes visual evidence of the enhancement. How are we to demonstrate that the required surface modifications have actually been carried out as the designer intended?
Engineered surfaces are often produced in a batch process. For example, in case hardening a batch of prepared parts may be treated together in a single carburizing furnace load. The parts in the load may not even be of the same shape. As such, often the batch is defined by the nature of the treatment with the temperature and the composition of the furnace atmosphere monitored and controlled. Representative test coupons or sacrificial parts can be included in the batch for later testing to demonstrate uniformity through the batch, and from batch to batch.
Processes in which components are treated individually demand a different approach to quality control. Individual part processes may include mechanical treatments such as machining, burnishing and peening, and thermal ones, such as induction hardening. Not only are parts treated individually in these examples, the processes are carried out progressively over the surface of each part, and the details of the process are specific to an individual design. Often, the significant properties of the treated, finished surface can only be measured destructively, so it is not possible to test the part itself. The significant properties of the surface cannot be compared directly to the designer’s intention. If the functional surface itself cannot be tested, it is necessary to reproduce the surface exactly, reliably and repeatedly on every part. This requires excellent process control.
Early Automotive Example
The introduction of an engineered surface, in a production part, is often in response to a manufacturing or service issue. The service life of valve springs in the automotive industry of the 1930s and early 1940s provides an excellent example. Steel coil springs were (and still are) used to close inlet and exhaust valves in four-cycle automobile engines. The repeated cycles cause metal fatigue, leading to cracking and eventual fracture of the spring. The problem was initially attributed to inadequate cleaning following heat treatment, and various cleaning methods were evaluated. Chemical treatments to remove scale gave no improvement. Sand blasting reduced the scatter in spring life, but shot blasting eliminated early failures as well as dramatically extending the average life of the springs. Shot blasting continued to be used to clean valve springs until it was recognized that shot blasting was one of the techniques used in the new field of “pre-stressing” critical engine parts such as connecting rods and crankshafts to prevent fatigue fracture.
In 1944, J. O. Almen invented a method to quantify the intensity of shot blasting. Almen took flat strips of spring steel similar in composition and hardness to the valve springs and shot peened one side of the strips. The strips curved so that the peened surface was on the outside of the curve. Almen measured the curvature as arc height in a gage. He found that heavier shot and higher velocity increased the arc height, and that increased peening time initially increased arc height rapidly, then much more slowly beyond a “knee” in the curve of arc height as a function of peening time. This curve is now known as a saturation curve, and it is used to set up individual peening machines to deliver a specific intensity in the stream of shot.
With the measurement method in place, and specifications governing the size, uniformity, composition and hardness of the shot, the process could be controlled and monitored. Cleaning the valve springs by shot blasting could be replaced by shot peening the parts to introduce a reproducible compressive pre-stress. Almen arc height provides part of the control needed in practical shot peening processes.
Coverage of the peened surface is also an important parameter. It needs to be adequate, but not excessive. Coverage is particularly influenced by the hardness and other mechanical properties of the workpiece. It is generally accepted that close to 100% coverage is necessary. It is also recognized that the last 1% or so of coverage requires as much time as the first 99%. During this time a significant proportion of the surface is receiving additional shot impacts that are unnecessary and may even be harmful. Coverage is estimated visually by examining the part. A fluorescent dye may be applied to the surface before processing, to be removed by the shot particles as they strike the surface. Remaining dye identifies areas not contacted by the shot. Excessive peening introduces roughness, laps and folds, which may act as origin sites for future fatigue cracks if the residual stress field is reduced or eliminated during service or by mechanical damage to the part.
Recently developed processes, including laser shock and ultrasonic peening, can produce deeper compressively stressed layers, with less local plastic deformation than conventional shot peening. These processes are aimed at specific points on the surface, with little overlap of the treated areas, while shot peening generates multiple overlaps. Nevertheless, all of them can use Almen strips and arc height measurements to control the intensity of peening. Almen’s invention allows shot peening to be carried out consistently and reliably, and the strips themselves are a permanent record of the process that was performed. However, Almen strips do not provide insight to the details of the residual stresses introduced in the surface of the treated part from the shot peening process. For that information we must turn to a different test involving X-ray diffraction.
X-rays were discovered by Wilhelm Conrad Röntgen in 1895 and by 1912 X-rays were widely used in medical imaging. Their nature was still unknown, hence the name. Suspecting that X-rays were waves, Max von Laue and his assistants shone a beam of X-rays at a crystal of copper sulphate and recorded a diffraction pattern on a photographic plate placed behind the crystal. This confirmed the wave property of X-rays and later the same year William Henry and William Laurence Bragg (father and son) confirmed his findings. Laue was awarded the Nobel Prize for Physics in 1914, and the Braggs received the 1915 Prize. They described the condition for diffraction of X-rays by crystals in the Bragg Equation:
nλ = 2dsinθ,
Where: n is an integer, allowing multiple orders of reflection, λ is the wavelength of the X-rays, d is the distance between parallel crystal planes, and θ is the diffraction angle. This equation provides a very powerful way to probe matter with an X-ray beam. It basically says that an X-ray beam will reflect off of a crystalline surface (such as found in metals) at very specific angles that directly relate to the distance between the atoms in the material. As such, X-ray diffraction can be used as a type of strain gage, and with knowledge of the elastic mechanical properties of the material, the measured X-ray strains can be converted to elastic stresses. This important tool can therefore be used to measure the residual stress in a component resulting from a surface enhancement process such as shot peening.
Routine preparation of Almen strips, evaluations of coverage, and X-ray diffraction residual stress measurements all play a role in process control of parts with surface enhancements. With these measurements there is adequate information to confirm the quality of surface enhancements on existing and new parts, and to evaluate serviced parts, repairs and competitive processes.