 Home |
|  Home |
966KB
|
MATERIAL AND COMPONENT VALIDATION BY SPECKLE INTERFEROMETRY AND CORRELATION METHODS A.Ettemeyer Ettemeyer Consulting, Ulm, Germany Abstract: Challenging the limits of modern material and component designs requires a deep understanding of the product performance with respect to stress distribution, fatigue and lifetime. With the help of modern simulation tools as FEA this process has significantly been improved and simplified. However, for experimental validation in the test, modern and efficient test methods have to be provided. Here, the speckle technique offers an excellent potential. Electronic speckle pattern interferometry (ESPI) has been proven to be a flexible and efficient tool for characterization of material properties of complex materials. Inversing the measuring principle by moving the illumination instead of the object, the ESPI technique also enables contour measurement functions, which have been integrated into the ESPI technique. This enlarged the application range to the inspection of components under load and for direct comparison of the measured contour and strain data with the FEA simulation. In order to meet the increasing need for the analysis of a larger range of deformations, e.g. in forming processes or for inspection of polymers, the development of a new digital 3D image correlation technique was initiated. Two cameras observe the object under investigation and calculate the 3D surface of the component using photogrammetric algorithm. The correlation of images, taken at different load levels leads to the complete information about the deformations and strains on each point of the surface. In this paper, we compare both techniques and give examples of successful applications. Introduction: Modern materials and components are often loaded far beyond the former strain limits. Especially in automotive and aircraft industries, they have to satisfy the toughest strength requirements and meet stringent safety standards. At the same time, product cycles have dramatically been shortened. To reduce cost and time-to-market, rapid design optimisation has become a key issue. In the last decade, many industries adopted Computer-Aided-Engineering (CAE) tools to reduce the number of prototype builds and to speed up the development cycle. These analytical tools are relatively inexpensive to use and faster to implement than the costly traditional test methods as used in traditional design processes. There are, however, many variables that CAE tools cannot adequately address, such as, boundary conditions, material anisotropy etc. Therefore, efficient physical experimental techniques are still necessary for component design validation. A traditional experimental technique for determining component strain/stress characteristics is the application of strain gauges. Although inexpensive and effective for some applications, strain gauges provide only localized information and cannot capture true peak strain or high-resolution information when applied in regions having large strain gradient. Strain gauges also require time consuming and careful application, which have also limited its utility for design optimisation in an intensive, time-critical development effort. Modern optical measuring techniques based on Speckle Interferometry and on Digital Image Correlation Techniques have opened new opportunities for fast and effective investigations on materials and components behaviour. Electronic Speckle Pattern Interferometry: Electronic Speckle Pattern Interferometry (ESPI) is a full field non-contact experimental measuring technique that allows rapid and highly accurate measurement of the three components of deformation and generation of strain/stress distribution with high resolution, /1/. It is well suited for the determination of the true peak strain and the strain distribution in regions of strain concentration that are usually the most critical in an investigation. Fig. 1 shows a schematic of an ESPI set-up. The laser beam is split into two beams, one as the object illumination beam and the other as a reference beam. The light reflected from the object surface is recombined with the reference beam. This results in an interference pattern that is the so- called speckle pattern and is recorded by the CCD camera. Before the object is deformed, the phase difference between the object and the reference beams is φ. After the object is deformed, the phase difference becomes (φ + ∆) and a phase change ∆ results from the deformation. The subtraction of both phase distributions φ and (φ + ∆) gives the phase change ∆ which is related to an object deformation in the direction of the sensitivity vector S /2/. As an example a clamped circular plate loaded centrally is presented in Fig. 1 Laser Refe- rence beam Illumination Measuring Object Observation direction Sensitivity vector (Φ+∆) S (Φ) Camera non deformed (Reference state) Reference- wave deformed (∆) Fig. 1: Principle of ESPI technique For measurement of the complete 3D displacement of the object, the measurement is repeated with minimum 3 different laser illumination directions enabling the measurement of the same object deformation with 3 different sensitivity vectors S1, S2, S3. For practical reasons, generally 4 symmetric illuminations are used, Fig. 2, and the laser illumination is automatically switched 4 times each at the acquisition of the reference state and the deformed state. The complete image acquisition is performed within typ. 1 second. The measurement resolution for displacements is a fraction of the laser wavelength and typ. 30 ... 100 nm. Principle of Strain Measurement: As described above, 3D-ESPI enables rapid and accurate determination of the three components of deformation u, v and w on an objects surface. Because it is of a full field measurement, the distributions of u, v and w over the entire object surface i.e. the functions u(x,y),v(x,y) and w(x,y) can be obtained, /3/. Obviously, the in-plane strains at each point of the measured area can be determined by differentiating the in-plane components u(x,y) and v(x,y): εx =x u∂∂,εy =y v∂∂, γxy =y u∂∂+φ φ + ∆ Object- wave v∂∂Such differentiation can be carried out, quickly and delivers a complete strain field in the measuring area. In Fig. 3, the direct comparison of an ESPI measurement with classical strain gauges is shown. A composite plate with elliptical hole is loaded with a tensile force. The resulting strains in force direction have been measured with a set of 4 strain gauges along the axis of the hole. On the opposite side of the hole, the displacements were measured with ESPI and from there, the strains were calculated. The deviations of the local strain distribution along the axis of the hole between the ESPI measurement and strain gauge measurement are resulting from the averaging effect of the strain gauges. If the ESPI data are integrated in the same area as the dimension of the strain gauges, the results correlate excellently, as shown in Fig. 3. xFig. 2: 3D-ESPI System Q-300, /4/ Strain
[10 ]
0,7
0,6
0,5
0,4
0,3
0,1
0
ESPI
Strain Gauge
0,2ESPI average at
strain gauge position
0 5 10 15 20 25 30
Position [mm]
Fig. 3 : Comparison of ESPI measurement
with strain gauges High Strength Steel Weldings: The full field information of the ESPI measurement is especially valuable for investigations of non homogenous materials, such as e.g. joints. The optimisation of welding materials and processes requires detailed information about mismatches of the different properties of the base material, the heat effected zone and the welding itself, /5/. Fig. 4 shows a section of a specimen with a classical welding seam. The specimen was loaded in the x-direction up to a maximum and unloaded, afterwards. During the complete loading cycle, the strain field was measured in the observed area. Fig. 5 shows the resulting strain stress curves at different locations of the specimen. While the base material behaves linearly (blue curve), the point of maximum strain at the centre of the welding seam is clearly plastic with a large hysteresis effect (red curve). In comparison, the violet curve represents the integrated strains on the complete welding seam. A classical two-point strain measurement would show the average strain value according to the green curve. Average Basis material Strain Fig. 4: Measuring field of a specimen with
welding seam in tensile test R [N/mm2]
400.0
350.0
300.0
250.0
200.0
150.0
100.0
50.0
0.0
0.000% 0.500% 1.000% 1.500%
Längsdehnung
Fig. 5: Strain - stress curves at different positions
of the specimen yWeld xHot spot LoadStrain measurement on curved surfaces: As described above 3D-ESPI can determine the three components of deformation along with the strains of an object that are both given in Cartesian coordinates and these coordinates are denoted as an Oxyz set of coordinates while on a flat surface. When the object under test has a curved surface, the Cartesian coordinates on the object are denoted as an Ox'y'z' set of coordinates (usually the z axis is normal to object surface) and are usually different from the 3D-ESPI Cartesian coordinates as that from the flat surface, Fig. 6. Therefore, the components of deformation have to be transformed into the Ox'y'z' coordinate system. Assuming that u, v and w are the three components of deformation in the Oxyz set of coordinates directly measured by 3D-ESPI they can directly be transformed into the component corresponding to the local coordinate system Ox'y'z' by applying the transformation matrix T: u'
vvu T w' = 'wIn order to determine the transformation matrix, a shape measurement of the object on the measured area is required. After the shape has been measured, the three components of deformation on the curved surface can be determined. The strains on a curved surface can be calculated according to the equations below. εx' =∂The shape of the curved surface can be measured with the same ESPI techniques as above, but using the inverse measuring principle: While the measuring object does not move, the illumination source of the ESPI system is laterally shifted. The resulting fringes correlate with the contour of the object and the optical set-up, /6/. With known geometry of the optical set-up and defined shift of the illumination the shape can be determined. For practical applications, the shape measuring function has been integrated into a miniaturized 3D ESPI system, Fig. 7. ∂,εx' ='u∂' 'xv∂∂, γxy' =''yu∂∂+'yv∂''xz
Fig. 6: Local coordinate system for 3D -ESPI
measurement
Fig. 7: Miniaturized 3D-ESPI system Q-100 for
strain analysis on components, /4/ |
| © NDT.net |