.Modelica.Mechanics.Rotational.UsersGuide.ModelingOfFriction

Information

Several elements of this library model Coulomb friction with the method proposed in:

Otter M., Elmqvist H., and Mattsson S.E. (1999):

Hybrid Modeling in Modelica based on the Synchronous Data Flow Principle. CACSD'99, Aug. 22.-26, Hawaii.

The friction equations are defined in base model Interfaces.PartialFriction. In the following, there are given explanations.

Assume first the most simplest friction problem: A block sliding on a surface. The friction force f acts between the block surface and the environment surface and shall be a linear function of the relative velocity v between the two surfaces. When the relative velocity becomes zero, the two surfaces are stuck to each other and the friction force is no longer a function of v. The element starts sliding again if the friction force becomes larger than the maximum static friction force f₀ (indicated by f0 in equations below). This element could be defined with a parameterized curve description leading to the following equations:

forward  = s >  1;
backward = s < -1;
v = if forward  then s-1 elseif backward then s+1 else 0;
f = if forward  then  f0 + f1*(s-1) elseif
       backward then -f0 + f1*(s+1) else f0*s;

This model completely describes the simplified friction element in a declarative way. Unfortunately, currently it is not known how to transform such an element description automatically in a form which can be simulated:

The block is described by the following equation:

m*der(v) = u - f

Note, that m is the mass of the block and u(t) is the given driving force. If the element is in its "forward sliding" mode, that is s ≥ 1, this model is described by:

m*der(v) = u - f
       v = s - 1
       f = f_0 + f_1*(s-1)

which can be easily transformed into state space form with v as the state. If the block becomes stuck, that is -1 ≤ s ≤ 1, the equation v = 0 becomes active and therefore v can no longer be a state, that is an index change takes place. Besides the difficulty to handle the variable state change, there is a more serious problem:

Assume that the block is stuck and that s becomes greater than one. Before the event occurs, s ≤ 1 and v = 0; at the event instant s > 1 because this relation is the event triggering condition. The element switches into the forward sliding mode where v is a state which is initialized with its last value v = 0. Since v is a state, s is computed from v via s := v + 1, resulting in s = 1, that is the relation s > 1 becomes false and the element switches back into the stuck mode. In other words, it is never possible to switch into the forward sliding mode. Taking numerical errors into account, the situation is even worse.

The key to the solution is the observation that v = 0 in the stuck mode and when forward sliding starts, but der(v) > 0 when sliding starts and der(v) = 0 in the stuck mode. Since the friction characteristic at zero velocity is no functional relationship, again a parameterized curve description with a new curve parameter s_a (denoted also sa below) has to be used leading to the following equations (note: at zero velocity):

startFor  = sa >  1;
startBack = sa < -1;
        a = der(v);
        a = if startFor then sa-1 elseif startBack then sa+1 else 0;
        f = if startFor then  f0  elseif startBack then  -f0 else f0*sa;

At zero velocity, these equations and the equation of the block form a mixed continuous/discrete set of equations which has to be solved at event instants (e.g. by a fix point iteration), When switching from sliding to stuck mode, the velocity is small or zero. Since the derivative of the constraint equation der(v) = 0 is fulfilled in the stuck mode, the velocity remains small even if v = 0 is not explicitly taken into account. The approach to use the acceleration der(v) = 0 as "constraint" instead of v = 0, is often used in multi-body software. The benefit is that the velocity v remains a state in all switching configurations (there is a small, linear drift, but the friction element would have to stay stuck several days before the drift becomes too large). Consequently, v is small but may have any sign when switching from stuck to sliding mode; if the friction element starts to slide, say in the forward direction, one has to wait until the velocity is really positive, before switching to forward mode (note, that even for exact calculation without numerical errors a "waiting" phase is necessary, because v = 0 when sliding starts). Since der(v) > 0, this will occur after a small time period. This "waiting" procedure can be described by a state machine. Collecting all the pieces together, finally results in the following equations of a simple friction element:

// part of mixed system of equations
startFor  = pre(mode) == Stuck and sa >  1;
startBack = pre(mode) == Stuck and sa < -1;
        a = der(v);
        a = if pre(mode) == Forward  or startFor  then  sa - 1    elseif
               pre(mode) == Backward or startBack then  sa + 1    else 0;
        f = if pre(mode) == Forward or startFor   then  f0 + f1*v elseif
               pre(mode) == Backward or startBack then -f0 + f1*v else f0*sa;

// state machine to determine configuration
mode = if (pre(mode) == Forward  or startFor)  and v>0 then Forward  elseif
          (pre(mode) == Backward or startBack) and v<0 then Backward else Stuck;

The above approach to model a simplified friction element is slightly generalized in model Interfaces.PartialFriction:

• The sliding friction force has a nonlinear characteristic instead a linear one, by interpolation in a table of f(v) values.
• There may be a jump in the friction force when going from stuck to sliding mode (described with parameter peak).