Nonlinear dynamic modeling of axial force and torque in a friction stir welding tool
- Chamorro R. ,
- Serpa A.L. and
- Ramirez A.
- Chamorro R. ,
- Serpa A.L. and
- Ramirez A.
Links
Repository
Description
A nonlinear dynamic model to study the axial force and torque in the tool of a friction stir welding (FSW) process was developed by training an artificial neural network (ANN) multi-perceptron layer structure with a nonlinear autoregressive exogenous (NARX) model. An experiment was designed to determine the dynamic variation of axial force and torque during FSW of AA 5052-H34 aluminium alloy (Al, 2.21%Mg, 0.71%Si) plates. The nonlinear model was formulated empirically by using tool rotational speed, traverse speed and penetration as inputs and axial force and torque as outputs. The NARX model employed input and output delays to predict output signals. The methodology was validated by comparative analysis between experimental and model data.
11th International Symposium, 17-19 May 2016, Session 2B: Modelling I, Paper 02
Related Records
