Skip to content
Tutorials Center
My learning

Topics

Lesson 5 of 37 ยท Modelling and Transfer Functions

Modelling Physical Systems

Derive differential equations for mass-spring-damper, RC circuit, DC motor and tank systems, see their analogies, and linearize nonlinear models in Python.

Before you can control a system you need equations that describe it. This lesson shows how to derive models from basic physical laws for mechanical, electrical, electromechanical and fluid systems. The result is always a differential equation relating an input to an output. The remaining lessons of this chapter turn those equations into transfer functions and block diagrams.

The strategy is the same everywhere: pick the variables, write one balance law per energy-storing element (Newton's law, Kirchhoff's law, mass or energy conservation), and eliminate the extra variables.

Mechanical translation: mass, spring, damper

A mass m is pushed by a force F(t). A spring pulls back with k x and a viscous damper resists with c x'. Newton's second law (sum of forces equals mass times acceleration) gives

Text
m x'' + c x' + k x = F(t)

This is a second-order equation because it has two energy stores: the mass (kinetic energy) and the spring (potential energy). The damper dissipates energy. Car suspensions, door closers and many sensors follow this model.

Electrical: the RC and RLC circuits

For a resistor R in series with a capacitor C driven by voltage v_in, Kirchhoff's voltage law with i = C dv_out/dt gives

Text
R C v_out' + v_out = v_in

This is first order with time constant R C. Adding an inductor L in series gives the RLC circuit:

Text
L C v_out'' + R C v_out' + v_out = v_in

Compare it with the mass-spring-damper. They are the same equation with different names.

Mechanical Electrical (series RLC) Role
Force F Voltage v Effort or source
Velocity x' Current i Flow
Mass m Inductance L Stores kinetic-type energy
Spring 1/k Capacitance C Stores potential energy
Damper c Resistance R Dissipates energy

This analogy is why one theory serves many fields: learn the second-order equation once and you can handle suspensions, circuits and more.

One equation, many systems

Mass-spring-damper m x'' + c x' + k x = F
Series RLC circuit L q'' + R q' + q/C = v
Thermal mass First-order lag with R and C.

Same maths, different symbols.

1 / 2

Electromechanical: the DC motor

A DC motor is the workhorse of control examples. Its electrical side is an inductor and resistor driven by the armature voltage V, and the motor produces a back-EMF proportional to speed. Its mechanical side has inertia J and viscous friction b.

  • Electrical: L di/dt + R i = V - Ke w (back-EMF Ke w)
  • Mechanical: J dw/dt + b w = Kt i (torque Kt i)

Here w is angular speed. When the inductance is small, set L = 0 to get i = (V - Ke w)/R and substitute:

Text
J w' + (b + Kt Ke / R) w = (Kt / R) V

That is a first-order system in speed: the electrical coupling acts as extra damping. With J = 0.01, b = 0.1, Kt = Ke = 0.01 and R = 1, the equation is 0.01 w' + 0.1001 w = 0.01 V, with time constant 0.01 / 0.1001 = 0.0999 s and DC gain 0.01 / 0.1001 = 0.0999 (rad/s)/V.

Fluid and thermal systems

A tank with inlet flow q_in, cross-section area A, and an outlet whose flow is proportional to level h (q_out = h / Rh) satisfies mass conservation:

Text
A h' = q_in - h / Rh      =>      (A Rh) h' + h = Rh q_in

A thermal mass works identically: heat stored C_th T' equals heat in minus heat lost (T - T_amb)/R_th. Many industrial processes are first-order lags like this, sometimes with a transport delay.

Linearization

Most real systems are nonlinear. A pendulum of length l with damping obeys

Text
theta'' = -(g/l) sin(theta) - b theta'

The sin makes it nonlinear, and the whole machinery of transfer functions needs linear equations. The fix is to linearize around an operating point. Near theta = 0, sin(theta) is approximately theta, giving theta'' + b theta' + (g/l) theta = 0. In general, replace f(x) near x0 by f(x0) + f'(x0)(x - x0): the tangent line.

The approximation is excellent for small deviations and poor for large ones. Check it by simulation:

Python Nonlinear pendulum versus its linearization

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The relative error grows with the starting angle. For 0.1 rad it is small; for 1.5 rad (86 degrees) the linear model is clearly wrong. A controller designed on the linear model is trustworthy only while the system stays near the operating point.

Simulating the motor

Use the motor equation from above to see the first-order behaviour.

Python DC motor speed response

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The speed approaches gain * V = 1.199 rad/s, reaching 63% after one time constant of about 0.1 s.

Check your understanding

Quick check ยท 5 questions

of 5 answered You scored / 5 ยท perfect!

  1. 1. Which pair is analogous in the mechanical and series-RLC electrical systems?

    The damper dissipates energy like a resistor. Mass corresponds to inductance, the spring to capacitance, and force to voltage.
  2. 2. Why does a mass-spring-damper give a second-order equation?

    The mass and the spring each store energy, so the system has two independent state variables and a second-order differential equation.
  3. 3. What is the linearization of sin(theta) near theta = 0?

    The tangent line of sin at 0 has slope cos(0) = 1 and passes through 0, so sin(theta) is approximately theta for small angles.
  4. 4. A tank has area A and outlet resistance Rh. What is its time constant?

    From (A Rh) h' + h = Rh q_in, the time constant is A Rh.
  5. 5. When is a linearized model reliable?

    The tangent-line approximation is accurate only for small deviations from the point where it was computed.

Summary

  • Models come from balance laws: Newton for mechanics, Kirchhoff for circuits, conservation for tanks and heat.
  • Mass-spring-damper and series RLC are the same second-order equation, so analogies transfer results between fields.
  • A DC motor reduces to a first-order speed model when armature inductance is neglected.
  • Linearize nonlinear models with a tangent line around an operating point, and check the range of validity by simulation.

Next: The Laplace Transform for Control