Locations of strain gauges for fatigue analysis of welded joints (1)

Click to increase image sizeClick to decrease image size

Notes

^1. Fatigue assessment is assured either by damage-tolerant design or safe-life design. γ_M is a partial safety coefficient of fatigue resistance related to stress, used in safe-life design. In this advice, there are no general recommendations but in safe-life design, it is 1.15 when γ_M causes damage to a secondary member, an effect that causes damage for some part (or structure), 1.30 when the damage is to the entire structure and 1.40 when the damage is to human life. When a structural member is produced at usual quality, and it is inspected regularly during use, γ_M = 1 is appropriate¹.

The maximum value for Δσ_s,d in the 1995 edition is 26 MPa for steel and 10 MPa for aluminium alloy, in each case lower than the 1996 edition.

^2. The value m used in the formula N = K/Δσ^m of the S − N curve for nominal stress is m ₁ = 3.0 when number of iterations N < 5 ⊆ 10⁶ and m ₂ = 5.0 when number of iterations N≦5 ⊆ 10⁶, with N = 10⁸ being the cut-off. When this is used in cumulative damage calculation, recent research indicates that assessments that use this formula are rather unstable and, in order to avoid this problem as far as possible, the sum of the minor side should not be 1 m, as is usual, but 0.5, as shown below:

^3. FAT is the stress range (N/mm²) at N = 2 ⊆ 10⁶ iterations of the design S − N curve. For an explanation of the S − N curve, etc. for aluminium alloys, see Reference 6.

^4. For example, in the IIW recommendations¹, there are allowable values for some transverse butt joints and load transmitting cruciform joints; for example, the allowable quantity for misalignment of the latter is less than 15% of sheet thickness.

^5. For the stress concentration factors due to these misalignments, see, for example, Refs 7 and 8.

^6. The materials used, according to the IIW recommendations¹, are structural steel with ferrite/pearlite microstructure or bainite microstructure and aluminium alloys used in welding structures, up to yield strength σ_y≦700 MPa. The EC 3 standards² are for structural stainless steel and uncoated weather-resistant steel. The EC 9 standards³ are for 7020 alloy and 5000 and 6000 aluminium alloys.

^7. Usually, electrical resistance wire strain gauges are used for stress concentration measurement in experimental stress analysis. The gauge length should preferably be short and 0.4 mm is recommended. A two-strand or three-strand strain gauge is used according to measure the structure depending on the stress state but details are omitted here.

The parts where they are used are as follows:⁴

	(1) Measurement of nominal stress of a structure at a location distant from the structural discontinuity, that is, other than the region of the geometric stress concentration is by an appropriately located strain gauge with the result that this is separated into axial force, bending moment, shear force and torsion, etc.
	(2) Geometric strain measurement to determine the hotspot stress for local discontinuities in a structural component.
	(3) The same as the above but in this case measurement to establish the hotspot S − N curve for the fatigue specimen.
	(4) Studies of dynamic response of a structure to evaluate the actual dynamic load coefficient used in stress calculation.
	(5) Verification of the results obtained by FEM; this comparison is one of the conditions used in the hotspot stress approach.
	(6) Experimental determination of the geometric stress concentration using (1) and (2) above.
	(7) Recording of the stress history of either (1) or (2) above, to obtain data on the stress generated in structural components subjected to variable amplitude loading.