Image Acquisition Setup

The DIC equipment setup shown in Figure 34 was a 5M non-contact optical 3D deformation measuring system manufactured paired with the ARAMIS deformation analysis software, and sensor controller board, all manufactured by GOM Optical Measuring Techniques, and a Dell Precision M6700 laptop. The digital camera system consisted of two 5 megapixel Baumer TGX15 monochrome cameras with a maximum acquisition rate of 7 Hz at full (5 MP) resolution and 29 Hz with binning (2.5 MP). Each camera has a 2/3” Charged Coupled Device (CCD) sensor, and an image resolution of 2448 x 2050 pixels (width x height). In this thesis the images acquired have the opposite dimensions since the cameras were rotated 90° to acquire more pixels along the height of the coupon as explained previously. All images are stored in the main memory (RAM) of the computer attached. Attached to the camera fixture bar are two LED lights with polarization filters. Finally MTS data is supplied to the DIC system through the controller

board as a voltage signal in order to synchronize DIC and MTS data.

Figure 34. DIC system: GOM 5M [84]

Before recording, the camera system has to be positioned and set-up depending on the dimensions of the object to be monitored. To achieve this the manufacturer provided tables with different parameter settings, such as distance from the object and camera position along the fixture bar, depending on the surface view and camera system used. In this investigation the 50mm lens/5M system configuration was used for a 65 x 55 mm2

and 50 x 42 mm2 _{field of view for monotonic and fatigue tests, respectively, with the}

corresponding parameters shown in Table 5.

Table 5. DIC camera setup for tests

Parameter Monotonic Fatigue

Measuring distance (mm) 485 395

Slider distance (mm) 176 136

Camera angle (°) 25 25

The reason why the monotonic setup was not used for fatigue tests was because with the

camera system

sensor controller board laptop

MTS load (Volts)

longer tabs the field of view was reduced, plus the grips obstructed some of the view. By choosing a closer measuring distance this compensated for some of the field of view lost. The cameras were also aligned vertically such that one camera was above the other to increase the pixel density in the acquisition. Next the lighting and focus had to be adjusted using a 55x44 calibration panel provided by the manufacturer. The appropriate focus was achieved by making sure that small size letters of a business card were read clearly. Achieving the correct light exposure involved changing the iris size and polarizer position in the camera and the LED lights, respectively, to achieve violet tones in a false color image of the calibration cube in real-time. Once the lighting and focus were satisfactory, a calibration procedure in the ARAMIS software was carried out, using a 55 x 44 mm2

calibration cube based on the object volume corresponding to the setting in Table 5. This process is performed by taking a sequence of images of the calibration cube in different positions and orientations relative to the camera, following the calibration procedure in ARAMIS. The general calibration results for both monotonic and fatigue tests were the following:

Table 6. Calibration results

Parameter Value

Calibration deviation (pixel) 0.036 – 0.040 Scale Deviation (mm) 0.001 – 0.003

Camera Angle (°) 24.4 – 25.0

Avg. Intersection Deviation (pixel) 0.022

system can acquired is constrained by the PC RAM, size of the images, and the frame rate chosen. For full 5 MP resolution the maximum amount of images allowed was approximately 400 and for 2.5 MP (using binning) was approximately 1,800. Therefore to optimize the system to ensure that the complete duration of the test could be captured, for monotonic tests a 2 Hz frame rate at 5 MP resolution was chosen whereas for fatigue tests, since they last longer, a python script [43] was used to specify different frame rates at different load levels taking 2.5 MP resolution images with binning. This adjustment was done in fatigue tests to reduce the amount of data gathered from a continuous frame rate such that images were only taken at specific loads levels of interest specified in the script as voltages, based on a conversion factor of 8896.404 N/V. The script initiated when mean load was reached the first time, taking 13 images at 27 Hz frame rate to capture approximately one full cycle, then 10 images were taken with a delay in between of 10- 15 s depending on the estimated duration of the test, and then the sequence was repeated after a delay of 10-15 s starting when the mean load was reached again. The exposure time was approximately between 28-30 ms in both loading cases.

Once the images were acquired, the image correlation was performed in ARAMIS. The facet field characteristics used for the auto-correlation algorithm in the post-processing of monotonic and fatigue tests are shown in Table 7. Facet size, step and overlap were always the same, whereas the facet field varied by a few pixels due to the speckle pattern, lighting, etc. Moreover a smaller facet size was used in the fatigue test results given the reduced gage length to achieve a slightly higher sensitivity.

Table 7. Facet Field Characteristics

Parameter Monotonic Fatigue

Facet Size (pixels) 30 20

Facet Step (pixels) 15 10

Facet Overlap (%) 50 50

Facet Field Size ( X x Y pixels) 46 x 152 40 x 44

For the strain calculations, the minimum computation size was used which is 3. This means that a 3x3 field of 3D points is used to calculate the strain value of the center point. This setting is particularly suitable for the assessment of local strain means.