Objective
• To gain an understanding of the fatigue properties of a pure base material sample from S420G2+M steel.
• To approximate a SN-curve based on the results.
Experimental procedure
In order to determine the appropriate stress range for the main welded specimens, some fatigue tests were conducted on pure base material. The preparation was performed accordance to the requirements in ASTM E466-15 [45]. The standard covers the procedural steps in order to design a specimen in the fatigue stress range where the strains are predominately elastic. It is limited to axially loaded specimens subjected to constant amplitude loading at ambient temperature. The specimen can be either notched or unnotched. It is designed for measuring the effects of variations of material, geometry, surface condition and residual stress of metallic specimens at a rather large amount of cycles.
Careful considerations of the procedural steps are vital in order to verify that the results are viable and reproducible. To achieve this, a tight control of variables is necessary; such as hardness, cleanliness, grain size, composition, directionality, surface residual stress, surface finish, etc. All obtained data should be logged and stored for future evaluation [45].
The design of the specimen dimensions should be such that the eventual failure occurs in the reduced area in the test section. It is therefore vital when reducing the area of the test section, that the radius introduced from the machining doesn’t cause any detrimental stress concentrations. In addition, using a square or rectangular cross section might reduce the fatigue life of the specimen, due to reduced resistance to plane slippage at the edges. In a circular cross section, the surrounding grains confines the material and inhibits the process and might increase fatigue life [45].
The specimen dimensions were designed as a rectangular cross section with tangentially blended fillets between uniform test section at the ends. The radius of the specimen is supposed to be eight times the specimen test section width to minimize the stress concentration. The specimen test section width should be 2-6 times the thickness, whereas the resulting area should lie in between 19.4-645 mm2. The test section length should be 2-3 times the test section width.
The width of the grip should be 1.5 times the test section width. The length of the whole specimen was 300 mm, resulting in a grip length of 40 mm on each side.
This test dimension is designed for pure base material specimens, but due to the following tests had to fit two welds, some adjustments had to be made of the original design. In order to fit both welds in the test specimen section the length had to be increased. The finished design of the specimen was as shown in Figure 9-6.
Figure 9-6 - Specimen dimensions according to ASTM E466.
The following procedural steps was taken to machine and prepare the said specimen. according to ASTM E466-15 Appendix X1. Example of the machining procedure:
1. X1.2.1 – Machining was done gradually where the second-last step was 0.4 mm and last 0.2 mm.
2. X1.2.2 - Removed the next 0.1-0.2 mm on the front and backside of the specimens with cylindrical grinding (plane Surfaces Grinding Machine). It was not possible to use the plane surface grinding machine on the sides of the samples, so this was done later with abrasive paper.
3. X1.2.2 - The final grinding was done manually with abrasive paper. Sanding was done with Hermes WS Flex Waterproof P180, Struers FEPA P # 500, Hermes WS Flex Waterproof P1000, Struers FEPA P # 1200 and Silicon Carbide 1200/4000
4. X1.2.4 - Requirements after grinding were that all slip marks would be along the test direction of the test specimens. Finally, a visual check was made with a magnifying glass, where no transverse grinding marks were accepted. A visual log was made where all errors were written down.
5. A roughness check was carried out. Requirements for Maximum surface roughness were 0.2um in the longitudinal direction
6. After surface treatment and control, specimens were lubricated into grease while waiting for testing. Specimens were stored in towels.
Figure 9-7 - Finished test specimen sample.
Results and discussion
The results from the tests on the pure base material fatigue specimens are as shown in Table 9-4.
Table 9-4 - Results from fatigue test of unwelded base material plate specimens.
Fatigue test R=0.1
Unwelded base material plate specimens
Specimen ID BM1 BM2 BM3 BM4 BM5
Run out/Fracture Fracture Run out Fracture Run out Fracture
Specimen ID Description References
BM1 Fracture at 430 after 15 643 cycles Plastic fracture. Low-cycle fatigue range.
BM2 Run out at 350 MPa after 2 760 746 cycles.
BM3 Fracture at 400 MPa after 394 701 cycles.
Crack initiated in the side face.
BM4 Run out at 375 MPa after 6 770 000 cycles.
BM5 Fracture at 400 MPa after 426 615 cycles.
Crack initiation at surface face.
For a crack to initiate, the stress range had to be above yield point level. At a stress range of 375 MPa, which equals to 417 MPa in maximum load due to stress ratio R=0.1, the test ran out at 6 770 000 cycles. The yield point of the metal was approximately 430 MPa which means that the specimen was subjected to dynamic loads equaling 97 % without fracturing.
Three tests were loaded until fracture; one at 430 MPa and two at 400 MPa. With stress ratio R=0.1 the resulting maximum load was 478 MPa and 444 MPa. The specimen at stress range 430 MPa failed prematurely in the low-cycle fatigue stress range.
The final crack initiation in the high-cycle fatigue specimens may have nucleated in intrusions due to plane dislocation slippage. The fatigue life of the high-cycle fatigue specimens consisted mainly in the crack initiation stage. When a crack finally was initiated the subsequent crack
propagation growth rate was extremely fast. It wasn’t possible to detect any visible crack before fracture, despite regular check-up during testing.
Figure 9-8 - Unwelded base material fatigue specimens after testing.
Summary
The main fatigue life of the unwelded base material test specimens consisted of the crack initiation stage. This was increased even further by following the procedural steps in ASTM E466-15. The surface finish greatly improved the fatigue life.
These tests stands in direct contrast to the previous fatigue tests on the welded “prior to fabrication specimens” having reduced fatigue strength due to discontinuities and defects in the weld.
To create an approximated SN-curve based on the testing results was not possible due to the limited sample size. It was also not advisable to conduct further fatigue tests on the used specimens at an increased stress range due to uncertainties from the influence of strain hardening.