PSI - Issue 58

Mikkel L. Larsen et al. / Procedia Structural Integrity 58 (2024) 73–79 M.L. Larsen et al. / Structural Integrity Procedia 00 (2024) 000–000

74

2

In this paper, a digital twin of a lifting arm of an agricultural disc mower is developed based on experimental results. The experimental setup is developed to test one major load direction while also reducing the number of complex boundary conditions. A digital twin is then developed, based on the finite element method (FEM). It is well known that discrepancies between FE analyses and experimental results can occur due to incorrect modelling of joints and joint properties, boundary conditions, geometry di ff erences and material properties (Arora (2011)). In the current study, discrepancies between the experimentally obtained results and FE predictions are observed. To reduce the discrepancies, parameter-based finite element updating of the support conditions is performed. The results show very good agreement between the FE predicted results and the experimental results. Finally, the levels of non-proportionality are quantified for four points on the lifting arm based on the updated FE model and a quantification method based on principal component analysis (Larsen et al. (2022a,b)). The results show that the levels of non-proportionality are highly dependent on the location of interest, which means weld optimization can be performed by taking this information into account.

2. Experimental setup of lifting arm in mower structure

The non-proportionality levels of a lifting arm in an agricultural disc mower are of interest in the current study. An overview of the mower structure is shown in Fig. 1. As seen from the figure, several parts and components comprise the agricultural mower. The main assembly consist of the hitch connection that connects the mower structure to the tractor, the lifting arm and the hydraulic actuator capable of lifting the arm to facilitate various work and transport operations. Furthermore, the mower unit is connected to the lifting arm using a special link-system. In this study, only the lifting arm is considered as it has to transfer the loads from the mower unit to the hitch, while also being heavily welded.

Hydraulic actuator

Link-system

Hitch

Lifting arm

Mower unit

Fig. 1. Overview of agricultural mower structure.

The purpose of the tests is to obtain experimental data that will facilitate the development of a digital twin of the lifting arm. The test setup consists of two support structures where the lifting arm can be connected using large diameter pins, see Fig. 2. The support structures remove the necessity of using a hydraulic actuator between the lifting arm and the supports in the setup, which considerably reduces the complexity of the test. Furthermore, a frame system manufactured using I-beams is placed above the lifting arm. Using the frame it is possible to apply two di ff erent sets of loading, namely a horizontal and vertical loading. An overview of the test setup is shown in Fig. 2 (a) and a picture of the real setup is shown in Fig. 2 (b). It should be noted that in this paper, only a horizontal force is applied. The loading is applied using hydraulic actuators and the loads applied are comparable to real loads measured during field work using the mower structure. To capture the response of the structure, several measurement gauges are applied to the structure. Several uniaxial strain gauges and rosette gauges are placed on the structure. The gauge overview is shown in Fig. 3. The gauges are 3 mm 3-wire epoxy coated TML WF waterproof strain gauges. In total 21 uniaxial gauges and 7 rosette gauges have been used in the setup. The gauges placed far from stress raisers such as geometry changes or welds are denoted global gauges (G) and gauges closer to welds and geometry changes are denoted local gauges (L) in Fig. 3. Rosette gauges are denoted by an (R). Wire-potentiometers are attached to the hydraulic actuators that measure the applied displacement during loading. The load in each direction is applied in two steps in the experiments. The first step consists of applying between 5 and 10 % of the total force, which ensures that any slack in the system is removed. The strain gauges and wire-potentiometers are then zeroed and the full loading is applied. All experimental data is captured using the Dewesoft Data Acquisition systems.