M2794.
002700 Introduction to Robotics
Problem Set #3 solution
Problem 1 Modern robotics, Chapter 12, Exercise 12.15.
f2
f1
f3
f5
f4
(a) (b)
Figure 1
(a) The conditions for force closure of Figure 1(a) can be arranged into a linear equation of
the form Ax = b:
−1 1 1 1 −1
f1 = c1 , f2 = c2 , f3 = c3 , f4 = c4 , f5 = c5
−1 −1 1 1 1
c1
−1 1 1 1 −1
c2
fx
−1 −1 1 1 1 c3 = fy
√
0 0 − 2 0 0 c4 Mz
c5
Matrix A can be reduced to the following row echelon form via Gauss-Jordan elimination:
1 0 0 −1 0
[I S] = 0 1 0 0 −1
0 0 1 0 0
There does not exist any w ≥ 0 satisfying Sw < 0. Thus, this is not force closure.
(b) In order to make force closure by adding one frictionless point contact, new point contact
should be able to exert positive moment to planar square. Figure 1(b) represents all
possible locations of this contact as red line.
1
Problem 2 Modern robotics, Chapter 12, Exercise 12.17.
(a) From Nguyen’s theorem, this is a force closure grasp.
(b) From Nguyen’s theorem, this is NOT a force closure grasp.
(c) From Nguyen’s theorem, all positions x that ensure that the grasp is force closure are of
the form 0 < x < L.
Problem 3 Modern robotics, Chapter 12, Exercise 12.18. The conditions for force closure of
Figure 2 can be arranged into a linear equation of the form Ax = b:
−µ µ 0 −1
f1a = c1 , f1b = c2 , f2 = c3 , f3 = c4
1 1 −1 0
c1
−µ µ 0 −1 f1
1 1 −1 c2
0 = f2
1 1
c3
0 0 −4 2 f3
c4
Matrix A can be reduced to the following row-echelon form via Gauss-Jordan elimination:
1
1 0 0 −1 + 2µ
−µ µ 0 −1
1 1 −1 1
0 → 0 1 0 −1 − 2µ
1 1
0 0 −4 2 0 0 1 −2
1
1 0 0 −1 + 2µ
1
[I S] = 0 1 0 −1 − 2µ
0 0 1 −2
1
In order to be force closure, all elements of S should be negative. In other words, −1 + 2µ <0
1
and −1 − 2µ < 0.
1
∴µ>
2
Problem 4 Modern robotics, Chapter 12, Exercise 12.19.
(a) The conditions for force closure of Figure 3(a) can be arranged into a linear equation of
the form Ax = b:
0 1 0 −1 0
f1 = c1 , f2 = c2 , f3 = c3 , f4 = c4 , f5 = c5
−1 0 −1 0 1
2
0.5 c
f2
f3
1
h
ŷ
f1a α f1b
x̂
0.5
f1
Figure 2
c1
0 1 0 −1 0 c2
fx
−1 0 −1 0 1 c3 = fy
1 −1 −1 1 0 c4 Mz
c5
Matrix A can be reduced to the following row-echelon form via Gauss-Jordan elimination:
1 0 0 0 −1
[I S] = 0 1 0 −2 0
0 0 1 −1 0
Since there exists w satisfying Sw < 0, this grasp is force closure.
(b) The conditions for force closure of Figure 3(b) can be arranged into a linear equation of
the form Ax = b:
1 −1 −µ µ
f1 = c1 , f2 = c2 , f3 = c3 , f4 = c4
−1 −1 1 1
c1
1 −1 −µ µ f1
−1 −1 c2
1 1
c3 =
f2
−1 1 0 0 f3
c4
Matrix A can be reduced to the following row-echelon form via Gauss-Jordan elimination:
1 −1 −µ µ 1 0 0 −1
−1 −1 1 1 → 0 1 µ2 − 12 − µ2 − 12
−1 1 0 0 0 0 1 −1
1 0 0 −1
→ 0 1 0 −1
0 0 1 −1
3
1 0 0 −1
[I S] = 0 1 0 −1
0 0 1 −1
Regardless of µ, this grasp is force closure except when µ = 0. When µ = 0, there are
only three columns in the matrix, so it is impossible to construct a fourth column with
all elements negative.
π
∴0<β≤
2
A B
A Y [ B
1 X Y
1 \
Z
x
X 1 Z x [
y
y
C 1
C
(a) (b)
Figure 3
Problem 5 Modern robotics, Chapter 12, Exercise 12.20.
(a) µ > tan−1 (π/4n).
(b) µ > 0.