For an assignment, my group was tasked to perform 2 FEA studies on a test specimen. A tensile and compression to measure the materials properties of AL 2024-T4. This test specimen was designed to follow BS EN ISO 6892-1:2016 Fusion 360 is used to model the test specimen and Ansys is used to model the test.
The free meshing is used due to the test specimen having curved geometries, a range of meshing is used between 1 mm and 6 mm to do a convergence study. The End-time in the analysis setting is changed to 0.001 s to minimise the run time. A constraint of fixed support is set on the left grip section of the test specimen and a displacement of 18 mm is placed of the right grip section to simulate a physical test. Under solution, the normal stress and strain and maximum shear stress and strain are chosen to find the stress and strain of the gauge length.
In a tensile test, a Force-displacement curve can be plotted. A load-displacement curve measures the extrinsic properties of the test specimen, from the graph the specimen’s stiffness, point of failure, ultimate load and displacement, and work done.
From the data, a stress-strain curve can be obtained. The curve measures the intrinsic properties of the material being tested; this is due to the data being normalise for the specimens’ dimensions. The materials Young’s modulus, yield stress and strain, ultimate stress and strain and modulus of toughness (energies to yield and failure).
A stress-strain curve is preferred over a force-displacement curve due to a force-displacement curve only being relevant to the specific test specimen. While a stress-strain curve is relevant to all specimens of the material.
The absolute value of the compresive stress and strain were used to compare the curve between tensile and compressive stress. Due to aluminium being a ductile material, pure deformation due to compression is a rarity. Compressive strength is considered to be equal to that of tension.
The fixed support constraints the left grip section in the x, y and z-axis, the displacement is placed on the right grip section to allow the shoulder length of the specimen to be extended, this reflects a tensile test using machinery. A displacement of 18mm is chosen as the material begins its necking behaviour, the software is unable to detect the stress or strain of the breaking point, therefore the Normal Stress and Normal Strain until 0.0006.5s are taken to reduce the error of the software. ANSYS retrieves the minimum, maximum and average results of stress, the maximum of the data point are used to improve the accuracy of the result as it is extrapolated from the area of maximum stress.
Though obtaining the Maximum Normal Stress before breaking for different element sizes between 1mm and 5.5mm. As seen on the graph the results begin to converge at a single value of 633.9 MPa. The maximum Normal Stress value at 1mm element size is 605.79MPa compared to the value 608.03MPa at 2.5mm element size which is 0.37% different from the value at 1mm. The time taken also is relatively small only being 2 minutes and 42 seconds compared to 130 minutes and 38 seconds for the first value.
The FEA study shares a similar constant increase in the elastic region, but in the plastic region, the FEA study doesn’t produce a smooth curve.
Table 2.1 shows that the FEA study had larger results for each material properties. Factors affecting these changes include:
· Data of the material in ANSYS being insufficient or incorrect
· Numerical errors of the solution of the FEA equation
· Modelling errors due to simplification
· Boundary conditions being wrong
Due to aluminium 2024 T4 high strength and fatigue resistance, they’re widely used in aircraft, this could be dangerous do to the factor of safety produced by the FEA study being larger than the actual value. To improve the accuracy of the FEA, the material values and FEA solution would need to be checked.