PSI - Issue 25

Tiago Bento et al. / Procedia Structural Integrity 25 (2020) 234–245 Tiago Bento / Structural Integrity Procedia 00 (2019) 000 – 000

241

8

Note that due to the Bragg fibers operating range limitations (an absolute value 125 0 µε ), two portions of the data had to be extrapolated by a 5th and a 3rd degree polynomials. Considering that in the parameters confi gured into the welding machine it was established a 6 seconds dwell time and a welding speed of 200 mm/min as mentioned in a previous section, the scheme in Fig. 8 could be drawn to explain the evolution of the curve in Fig. 7. Moment "1" represents the plunging and dwell time. During this period no strain alterations were measured in the fiber. As the tool progresses towards the location of the sensor the strain increases until a maximum value is achieved about 66 mm away from the Bragg measuring location, moment "2". Up until this point the stress direction on the fiber has been i n the direction of the tool, fiber in tension, but its intensity starts decreasing until it reaches a minimum when the pin is parallel to the fiber position, moment "3". Finally , th e tool moves away from the fiber conferring an inverse stress direction, compression of the fiber, in the direction of the tool until the welding process is finished, "4". These phenomena are a result of the constant material flow towards the tool to fill the empty space left by the tool pin and explain the origin of residual stresses in the joint. As observed in Fig. 7, an absolute value of 1250 µε the sensors’ operating range was exceeded. After welded and cured, the "smart joint" was mounted in the tensile test machine and submitted to a tensile test to study its elastic behavior. The end of test parameter was stipulated at 40 kN. However, as demonstrated in Fig. 8 the residual stresses on the joint that caused the sensor to be operating outside of its operating range in the last part of the welding had to be neutralized with force applied in the opposite direction. That is why the sensors only started to record new values after a load of around 17 kN was applied. The load witch the real and modelled specimen were subjected to is presented in function of Bragg fibers’ readings in the Fig. 9. As observed in Fig. 9, as the force increased the local strain in the joint increased as well. Due to the residual stresses left by the welding process it is normal that the Bragg fibers in the tensile test start recording values in the negative side of the microstrain axis, the fiber was initially in compression as it was seen in moment “4” in Fig. 7. Also, the variation in the inclination of the strain curve measured by the real FBG sensors should be due to a localized yielding effect that was only felt in the experimental test and not considered in the FEM model.

45

FBG Microstrain

40

35

FBG Microstrain - FEM

30

123

25

20

Load [kN]

15

10

5

0

-3500

-2500

-1500

-500

500

1500

2500

3500

Mircrostrain ( με)

Fig. 9. Lap joint Bragg fibers’ reading during tensile test .

A FEM study replicating the experimental test was performed for the entire specimen under smaller loads and the model had accurate results in relation to the what was experimentally observed.