Fig. 1: IrT setup according to [6]
This infrared radiation is detected by the camera, which is connected to a processing unit that converts the radiation data into thermal images. An optional thermal stimulation unit is needed for active IrT methods. The method of passive IrT measures the temperature variations on the object’s surface. Internal conversion processes for instance, that go along with (local) temperature changes can be detected measuring the attending changes in infrared radiation. Thereby, temperature hot spots resulting from crack formation and crack growth processes for instance can be detected. For more detailed information on IrT see [6].
Active IrT methods on the other hand use external excitation units for the thermal stimulation of the specimen. Optical LT for instance uses an external heat source, denoted as optional thermal stimulation in Fig. 1, introduces a heat flow into the specimen. The energy flow moves through the specimen and is partly reversed at material boundaries, such as voids or similar. The reversed energy flow is detected by the IrT camera system. Local differences in infrared radiation can thereby be visualized. This way, thermal images can be used to find subsurface irregularities.
2.2. Test Description
Two test series are performed, one under quasi-static and the other under fatigue loading. For each test, three different configurations are tested: specimens without patch, specimens with patch monitored from the metallic side and specimens with patch monitored from the patched side. The infrared camera used is of type ImageIR R 8380 S (InfraTec GmbH, Dresden) in combination with a tele zoom lense (\(f=50\) mm). The test equipment used is a tensile and compression machine of type Instron 5567 (Instron, Darmstadt) for the quasi-static tests and a servo-hydraulic test machine of type Schenck POZ 160 with a static nominal load of +/-160 kN. For the quasi-static tests, a test velocity of 9 mm/min and a monitoring frame rate of \(f=200\ \)Hz is used. The fatigue tests are performed with a stress ratio \(R=0.1\), mean load \(F_{m}=4000\) N, load amplitude \(F_{a}=3272\) N and a monitoring frame rate\(f=100\) Hz. To reduce data, recording is done every 10000 load cycles for about 5 seconds.
All specimens are made of aluminium 5083 with a thickness of\(h=3\) mm. The patch material used is a 0/90 epoxy-based carbon fibre reinforced plastic prepreg (SGL Technologies GmbH of type C W410-TW2/2-E323/42%/6k and a ply thickness of \(h=0.6\ \)mm). Specimen geometries and dimensions are given in Fig. 2.