Aluminum Extrusion Fsi
Aluminum Extrusion Fsi
Aluminum Extrusion Fsi
COPYRIGHT 2008. All right reserved. No part of this documentation may be photocopied or reproduced in any form without prior written consent from COMSOL AB. COMSOL, COMSOL Multiphysics, COMSOL Reaction Engineering Lab, and FEMLAB are registered trademarks of COMSOL AB. Other product or brand names are trademarks or registered trademarks of their respective holders.
Note: This model requires the Chemical Engineering Module, the Heat Transfer
Module, and the Structural Mechanics Module. In addition, the model uses the Material Library.
Model Definition
The model considers steady-state conditions, assuming a billet of infinite length flowing through the die. In the actual process, the billet is pushed by the ram through the die and its volume is continuously reducing. Figure 1 shows the original complete geometry with four different profiles. To have a model with reasonable dimensions, consider only a quarter of the original geometry.
The simplification involved in neglecting the differences between the four profiles does not affect the numerical scheme proposed. Figure 2 shows the resulting model geometry.
The documentation associated with the benchmark (Ref. 1) serves as the data source for properties of the two main materials: AISI steel for the die and the container (the ram here is not considered) and aluminum for the billet.
Structural Analysis
Because only the steel part is active in the structural analysis, consider a simple linear elastic behavior where the elastic properties are those of the material H11 mod (AISI 610) that can be found in the COMSOL Multiphysics Material Library.
Kal al Cpal
STEEL
Kfe fe Cpfe
Non-Newtonian Flow
The properties of the aluminum were experimentally determined and then checked using literature data for the same alloy and surface state. However the benchmark proposes an experimental constitutive law, suited for the structural mechanics codes usually used to simulate such processes, in the form of the flow stress data. For this model this requires a recalculation of the constitutive law to derive a general expression for the viscosity. The equivalent von Mises stress, eqv, can be defined in terms of the total contraction of the deviatoric stress tensor as eqv = 3 -- : 2
or, using = 2 where is the strain rate and is the viscosity, as eqv = Introducing the equivalent strain rate 2 eqv -- : 3 Equation 1 can be expressed as eqv = 3eqv The strain rate tensor is defined as (Ref. 2)
2 6 :
(1)
1 u + ( u ) = ------------------------------ = -- 2 2 Equation 5-26 on page 142 in the Chemical Engineering Module Users Guide states that the shear rate is defined as = = so that 1 eqv = ------ 3 The flow rule eqv = f states that plastic yielding occurs if the equivalent stress, eqv, reaches the flow stress, f. The viscosity is defined as (see Ref. 2 for further details) f = ------------ 3 eqv The organizers of the benchmark propose specific flow-stress data expressed in terms of a generalized Zener-Hollomon function
-- Z n asinh --- A = --------------------------------- 3 1
1 -- : 2
where A = 2.39108 s1, n = 2.976, = 0.052 MPa1, and 1- RT Z = ------ e 3 with Q = 153 kJ/mol and R = 8.314 J/(Kmol).
Q --------
Structural Analysis
Because the model geometry is a quarter of the actual geometry, use symmetric boundary conditions for the two orthogonal planes. On the external surfaces of the die, apply roller boundary conditions because in reality other dies, not considered here, are present to increase the systems stiffness. The main loads are the thermal loads from the heat transfer analysis and pressures from the fluid dynamics analysis.
Non-Newtonian Flow
At the inlet, the ram moves with a constant velocity of 0.5 mm/s. Impose this boundary condition by simply applying a constant inlet velocity. At the outlet, a normal stress condition with zero external pressure applies. On the surfaces placed on the two symmetry planes, use symmetric conditions. Finally, apply slip boundary conditions on the boundaries placed outside the profile.
On the basis of the results from the simulation, the engineer can improve the preliminary die design by adjusting relevant physical parameters and operating conditions. For this purpose, the slice plot in Figure 3 showing the temperature field inside the profile gives important information. Furthermore, the combined streamline and slice plot in Figure 4 reveals any imbalances in the velocity field that could result in a crooked profile. A proper design should also ensure that different parts of the profile travel at the same speed. In Figure 5 you can see von Mises equivalent strain in the steel part considering the thermal load and also the pressure due to the presence of the fluid.
References
1. M. Schikorra, L. Donati, L. Tomesani, and A.E. Tekkaya, The Extrusion Benchmark 2007, Proceedings of the Extrusion Workshop 2007 and 2nd Extrusion Benchmark Conference, Bologna, Italy, http://diemtech.ing.unibo.it/extrusion07. 2. E.D. Schmitter, Modelling massive forming processes with thermally coupled fluid dynamics, Proceedings of the COMSOL Multiphysics User's Conference 2005 Frankfurt, Frankfurt, Germany.