PSI - Issue 13

W. Teraud / Procedia Structural Integrity 13 (2018) 238–242

241

4

Teraud W. / StructuralIntegrity Procedia 00 (2018) 000–000

The purpose of “Recognizer” is transformation of graphic presentation of a specimen in a frame into internal program objects. Transformation carries automatically or semi-automatically, if automatic recognition of some specimen’s elements resulted in a large error. The recognition process takes place in the following way: “Recognizer” extracts a specimen from the frame to fix its right and left side boundaries, as well as upper and lower edges; then it recognizes the reference lines upon the surface. Using the first frame to contain original geometry of the specimen with known sizes, one calibrated the scale of the pictures. Images obtained upon recognition are stored in internal program objects and transmitted to “Solver”. The “Solver” software automatically transforms the objects into metrical geometry. “Solver” run results in a virtual specimen that is to the maximum identical to the natural one. Figure 2b shows a virtual cylindrical specimen at some point of its deformation. Then using “Post-processor”, graphs are constructed on the basis of metric data (virtual specimen) and data is generated for further analysis. It is also possible to represent dependencies of various parameters upon time t and x specimen axial coordinate. Some parameters can be represented in form of fields within a graphic image of the specimen (See Figure 3). For each type of specimens it is possible to obtain both standard data (changes to length / height  l ( t ) /  h ( t ) and deformation  ( t )), and specific one (distribution of true stress  ( t , x ) along a test portion of the specimen and tangent circle radius R ( t ) at necks of specimens subject to stretch and cylinder side surface swelling distribution B ( t , x ) (specimens subject to pressing). For each experiment it is possible to create a video file. The video file can contain a standard deformation process and/or augmented reality in form of specimen coloring according to deformations or acting stresses (Fig. 3).

Fig. 3. A stretched specimen with augmented reality.

3. Application With a material subject to stretching, at a certain point of time localization of deformation occurs (a “neck” appears). Due to difficulties related to its measurements at elevated temperatures it usually remains not addressed. Upon localization of deformations, a value of a stretched specimen becomes less informative. This point of time in experiments both before and after is usually unknown. Of the utmost interest is therefore to study occurrence and further development of a neck. The method was also used for cylinders under pressure. The test pattern was as follows. First, a specimen fixed in rods was heated to an operating temperature T. Upon rise of the temperature to a specified value, a camera was switched on, set to take pictures at given time intervals or specimen elongation intervals. A load was then quickly applied onto the specimen to reach a specified value of the initial axial stress  0 using a weight P 0 in the lower part of the test facility. Further deformations carried under creep conditions at a constant force P 0 till fracture. Two materials were selected for tests: D16T Al alloy and VT-6 titanium alloy. Flat- and cylinder-shaped specimens were subject to stretch, and cylindrical ones were pressed at temperatures ranging from 400 °C to 600 °C. More details on the test results obtained are given in [14]. Field tests with use of the developed method of measurement were carried out at the Institute of Mechanics of Lomonosov Moscow State University using the test facility IMEX-5. A camera NIKON D300s was also used, as