Comparison of mechanical properties and weight of the riveted and FS welded panel


6.4.1     FEM simulations

FEM simulation in program Ansys Workbench 10 were performed to determine the differences in mechanical properties of the riveted and FS welded panel. The simulations take in account differences in the joint character of riveted and FS welded joint.

6.4.1.1     FEM model

              6.4.1.2 Loading

6.4.1.3     Material characteristics used

Values of Poisson’s ration and Elastic modulus were taken from [81] and values of material strength were taken from Ansys material library which together with generic set of aluminium alloy properties created the material model used in the simulation.

Material                                     Aluminum 2024-T3

Poisson's ratio                            0.33

Elastic modulus                         73100 MPa

Tensile yield strength                285 MPa

Tensile ultimate strength           42 MPa

Compressive yield strength       285 MPa

6.4.1.4     Result evaluation

Four result sets were obtained for two load sets and two different samples. The resulting equivalent stress and deformation was evaluated. The FEM simulation on this level of detail does not provide sufficiently reliable date accurate stress analysis but provides a good representation of the different character of stress and deformation distribution between the riveted and FS welded construction.

6.4.1.4.1     Evaluation and comparison of results for positive Z axis direction loading

Figure 33 shows the distribution of the equivalent stress and deformation throughout the sample. On this scale, the results are in a good agreement with expected distribution for the bent sample. The stress peaks in the corners of the sample are caused by the effect of applied constraints on the mesh with relatively large size of elements in those areas. We can see light buckling of the panel plate.

Figure 33: FEM simulation result for riveted panel and loading in positive axis Z direction (deformation scale magnitude 29x)

The detailed examination of the riveted area in the cross section plane going through the rivets’ axes (Figure 34) shows stress concentration areas around the rivet holes. The value of stress for the given loading does not appear to be dangerous but for the higher loading levels, it could lead to crack initiation and spread. In addition, cracks can nucleate easily on the edges of drilled holes and this can increase fatigue risk further.

In addition, we can see slight separation of the contact surface between the stringer and the panel plate, which can initiate fretting, and lead to further easier crack initiation.

Also the stress distribution shows that our estimate of placing the neutral plane into the contact surface between the stringer and the panel plate was reasonable.


Figure 34: Detailed equivalent stress distribution in the middle cross section of the rivet joint for loading in positive axis Z direction (deformation scale magnitude 150x)

The FS welded sample (Figure 35) shows similar stress and deformation distribution which is also in agreement with our expectations. The stress peaks are again caused by the effect of applied constraints on the mesh with relatively large size of elements in those areas. As seen previously, loading causes light buckling of the panel plate.


Figure 35: FEM simulation result for FS welded panel and loading in positive axis Z direction (deformation scale magnitude 29x)

The detailed view into the weld area (Figure 36) shows stress distribution without peaks. The stress distribution in the panel plate can carry inaccuracy because only there is only one element used on throughout the thickness of the material. The change in stress distribution on the left side of the sample is caused by the meshing pattern.

It is important to note that the critical part of the weld, its root, is never exposed to the highest stress values, so eventual crack initiation is less likely to happen there.



Figure 36: Detailed equivalent stress distribution in the middle cross section of the FS welded joint for loading in positive axis Z direction (deformation scale magnitude 150x)


6.4.1.4.2     Stress distribution comparison

The detailed view into the joint area shows the difference in the stress distribution character. In the case of the riveted panel, the critical points of the construction are the rivet holes where initiation of crack is likely at higher stress levels. In the case of FS welded sample, the critical place is the outer surface of the panel plate where the highest tensional stress occurs. In addition, eventual weld defects and irregularities could lead to stress concentration and crack initiation.

6.4.1.4.3     Evaluation and comparison of results for negative Z axis direction loading

The stress distribution on this sample (Figure 37) does not show unexpected stress of deformation features. The stress peak close to the fixed support constraint is again cause by relatively large element size in the area. Again, light buckling of the panel plate occurs.



Figure 37: FEM simulation result for riveted panel and loading in negative axis Z direction (deformation scale magnitude 29x)

Detailed image of the cross section (Figure 38) shows again stress concentration peak in the rivet area, however, this is the case of compression stress which does not induce higher risk of crack initiation and growth.


Figure 38: Detailed equivalent stress distribution in the middle cross section of the rivet joint for loading in negative axis Z direction (deformation scale magnitude 150x)

The FS welded sample (Figure 39) shows the expected stress distribution with peaks caused by the effect of applied constraints on the mesh with relatively large size of elements in those areas.


Figure 39: FEM simulation result for FS welded panel and loading in negative axis Z direction (deformation scale magnitude 29x)

The detailed view into the weld area (Figure 40) shows stress distribution with no sudden peaks. As mentioned previously, the stress distribution in the panel plate can carry inaccuracy because only there was only one element used on throughout the thickness of the material. The change in stress distribution on the left side of the sample was caused by the meshing pattern.



Figure 40: Detailed equivalent stress distribution in the middle cross section of the FS welded joint for loading in negative axis Z direction (deformation scale magnitude 150x)

6.4.1.4.4     Stress distribution comparison

In this case of loading, the joint, both riveted and FS welded, are not in the dangerous areas as the predominant stress is compression. The critical zones are the top parts of the stringers.

6.4.2     Comparison of maximum stress and total deformation

The values of the maximal total deformation of the samples (Figure 41 and Figure 42) and their weight, were gained from the results of the FEM simulation in program Ansys.


    Figure 41: Maximal total deformation of the samples under the positive Z-axis loading

 

Figure 42: Maximal total deformation of the samples under the negative Z-axis loading

The results produced are shown in Table 3.

 

Table 3: Comparison of maximal total deformation and weight of the samples

Sample type

Loading

FSW/Riveted ratio

Negative 
Z-axis

Positive 
Z-axis

Negative 
Z-axis

Positive 
Z-axis

Max total deformation

[mm]

Riveted panel

0.15511

0.15874

0.94

0.91

FS welded panel

0.14536

0.14498

Sample weight

[kg]

Riveted panel

0.163

0.93

FS welded panel

0.152

 

We can see that the FS welded sample has lower maximal deformation in the both cases of the loading and at the same time lower weight. This shows that the FS welded construction uses the material more efficiently.



Comments