Question 2 of 6: Type-1 system — gain for a ramp-error spec, then step/ramp errors
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
Paper: National Exams — May 2013 · 07-Mec-A3 System Analysis and Controls · 3 hours · closed book (Casio/Sharp calculator + semi-log graph paper only) · any four of six questions constitute a complete paper; all questions of equal value. All six questions are solved here as a study resource.
Reference texts (subject). K. Ogata, Modern Control Engineering, 5th ed. (root locus Ch. 6, frequency response/Bode Ch. 7, steady-state error §5.8, Routh–Hurwitz §5.6); N. S. Nise, Control Systems Engineering, 8th ed.; R. C. Dorf & R. H. Bishop, Modern Control Systems, 13th ed.; G. F. Franklin, J. D. Powell & A. Emami-Naeini, Feedback Control of Dynamic Systems, 7th ed. Laplace-transform pairs are supplied with the exam.
Question 2: Type-1 system — gain for a ramp-error spec, then step/ramp errors (equal value)
Design target: a steady-state ramp error of $5\%$, i.e. velocity error constant $K_v=20$ (the paper labels this constant $K_r$).
Find. the gain $K$; then (a) the steady-state error to a unit step, and the full transient error responses to (b) a unit step and (c) a unit ramp.
Error transients for the designed loop ($K=500$). The unit-step error starts at $1$ and decays to $0$ (type-1 loop rejects a step exactly); the unit-ramp error starts at $0$ and settles to the design value $e_{ss}=0.05$.
Approach. A single free integrator makes this a type-1 loop, so $K_v$ fixes the ramp error and $K$; the transient error $E(s)=R(s)/(1+G)$ is then inverted for each input.
Gain from the ramp-error spec. For a type-1 loop the velocity error constant and ramp error are
$$K_v=\lim_{s\to0}sG(s)=\frac{K}{25},\qquad e_{ss,\text{ramp}}=\frac{1}{K_v}.$$
Setting $e_{ss,\text{ramp}}=0.05$ gives $K_v=20$ and
$$\boxed{\,K=25\,K_v=500\,}.$$
The closed-loop characteristic equation is then $s^{2}+25s+500=0$, with poles $-12.5\pm j18.54$ ($\omega_n=22.36$, $\zeta=0.559$).
(a) Steady-state error to a unit step. Because the loop contains an integrator, the static position constant is infinite:
$$K_p=\lim_{s\to0}G(s)=\infty\;\Longrightarrow\;\boxed{\,e_{ss,\text{step}}=\frac{1}{1+K_p}=0\,}.$$
The integrator drives the step error to zero in the steady state.
(b) Transient error to a unit step. The error transfer function is $\dfrac{E}{R}=\dfrac{1}{1+G}=\dfrac{s(s+25)}{s^{2}+25s+500}$, so for $R=1/s$
$$E(s)=\frac{s+25}{s^{2}+25s+500}=\frac{(s+12.5)+12.5}{(s+12.5)^{2}+18.54^{2}}.$$
Inverting,
$$\boxed{\,e(t)=e^{-12.5t}\!\left[\cos 18.54t+0.674\sin 18.54t\right]\,}.$$
The error jumps to $e(0^{+})=1$ (the plant cannot respond instantly) and rings down to zero with time constant $1/12.5=0.08$ s — consistent with $e_{ss,\text{step}}=0$.
(c) Transient error to a unit ramp. For $R=1/s^{2}$,
$$E(s)=\frac{s+25}{s\,(s^{2}+25s+500)}=\frac{0.05}{s}+\frac{-0.05(s+12.5)+0.375}{(s+12.5)^{2}+18.54^{2}}.$$
Inverting,
$$\boxed{\,e(t)=0.05-e^{-12.5t}\!\left[0.05\cos 18.54t-0.0202\sin 18.54t\right]\,}.$$
The ramp error starts at $e(0)=0$ and climbs to the steady value $e(\infty)=0.05$ — exactly the $5\%$ the gain was designed to deliver.