Anisotropic friction for Abaqus/Explicit

This problem contains basic test cases for one or more Abaqus elements and features.

This page discusses:

ProductsAbaqus/Explicit

Elements tested

C3D8R S4R

Features tested

Anisotropic friction for general contact in Abaqus/Explicit.

Problem description

The models consist of a solid element pressed against a shell element, converted to a rigid element for convenience, and displaced along the shell surface in varying directions. At the beginning of the analysis both contact surfaces have a directional preference oriented with the major axis of the scaling ellipse pointing in the Y-direction (see the vector plot of CORIENT1 in Figure 1). The frictional directional preference factor, ϵ, for both surfaces is equal to 0.5. The nominal or average friction coefficient, μiso, is 0.2.

Figure 1. Vector output CORIENT1.

Model:

Model-1:

Step-1 The block is pressed on the shell element with a force of 25 N.
Step-2 The block is displaced with a boundary condition in the X-direction.
Step-3 Release displacement in the X-direction.
Step-4 The block is displaced with a boundary condition in the Y-direction.
Step-5 Release displacement in Y-direction.
Step-6 The block is displaced with a boundary condition in the XY direction (45°).

Model-2:

Step-1 The block is pressed on the shell element with a force of 25 N.
Step-2 The block is displaced with a boundary condition in the X-direction.
Step-3 Release displacement in X-direction and pressure.
Step-4 Rotate the shell element with a rotation of 90°.
Step-5 Apply the pressure again.
Step-6 The block is displaced again with a boundary condition in the X-direction.

After the shell is rotated, the two surfaces have orthogonal directional preferences and the equivalent scaling ellipse becomes a circle.

Material:

Density 7.8e9
Young's modulus 7.0e4
Poisson's ratio 0.3

Anisotropic behavior:

The scaling ellipses can be mathematically treated as tensors. The form in the principal coordinate system is

S=[1+ϵ001ϵ].

The form in a rotated coordinate system is

S=RSRT
R=[cos(θ)sin(θ)sin(θ)cos(θ)]
S=R[1+ϵ001ϵ]RT
S=[1+ϵcos(2θ)ϵsin(2θ)ϵsin(2θ)1ϵcos(2θ)].

Each ellipse is defined as

Ssurf1=Ssurf2=[1+0.50010.5]=[1.5000.5].

The two surface scaling tensors (ellipses) are averaged to determine the contact scaling tensor (ellipse):

Scont=ω1Ssurf1+ω2Ssurf2.

Initially the major axis of the two ellipses are aligned, and using a balanced weighting method, the contact scaling ellipse becomes

Scont=0.5[1.5000.5]+0.5[1.5000.5]=[1.5000.5].

The limit shear force in the principal direction is computed by multiplying S by the average friction coefficient and the normal force,

FLimit=FNScontμiso=25[1.5000.5]0.2=[7.5002.5].

Results and discussion

The reaction forces are checked and compared with the analytical results.

  • Model-1: since the equivalent scaling ellipse is oriented as the Y-direction and the two surfaces are not rotating elatively, the expected reaction forces are equal to Fx = 2.5 N (Step-2) and Fy = 7.5 N (Step-4) during the sliding in the X- and Y-directions.(see Figure 3). In Step-6 the block is sliding in a diagonal direction in the X–Y plane (45°). By definition, the direction of the slip is orthogonal to the equivalent scaling ellipse (see Figure 2).
    FX=Akxyxel,Fy=Akyyyel,
    where
    kykx=(1+ε1ε)2=(1.50.5)2=9.
    For yxel=yyel (45° sliding),
    FyFx=9,θ=tan1(9)83.6°.
    The values of the forces Fx and Fy computed intersecting the scaling ellipse using the angle θ=83.6° are

    Fx=25*0.033=0.825 and Fy=25*0.284=7.1 (see Figure 3).

  • Model-2: during Step-2 the equivalent scaling ellipse is oriented as the Y-direction, so the expected reaction force is Fx = 2.5 N. After the rotation of the shell, the directional preferences of the two surfaces become orthogonal, generating an equivalent scaling ellipse equal to a circle for the isotropic case (see Figure 4).

Figures

Figure 2. Direction of slip respect to the critical force ellipse.

Figure 3. Reaction forces for Model-1.

Figure 4. Reaction forces for Model-2.