Journal

of
Journal of Terramechanics 42 (2005) 15–33
Terramechanics
www.elsevier.com/locate/jterra

Mechanics of cable shovel-formation interactions

in surface mining excavations
Samuel Frimpong *, Yafei Hu, Kwame Awuah-Offei
School of Mining and Petroleum Engineering, Chemical-Mineral Building, Room 606,
University of Alberta, Edmonton, Canada ABT6G 2G7

Accepted 7 June 2004

Available online 20 August 2004

Abstract

The cable shovel excavator is used for primary production in many surface mining opera-
tions. A major problem in excavation is the variability of material diggability, resulting in var-
ying mechanical energy input and stress loading of shovel dipper-and-tooth assembly across
the working bench. This variability impacts the shovel dipper and tooth assembly in hard for-
mations. In addition, the geometrical constraints within the working environment impose pro-
duction limitations resulting in low production efficiency and high operating costs. A potential
solution to the above problems is the deployment of an intelligent shovel excavation (ISE)
technology, with real-time formation identification, recording and knowledge transmission ca-
pabilities. This paper advances the ISE technology by developing dynamic models of the cable
shovel using the Newton–Euler techniques. The models include the main factors that influence
shovel performance including the effect of both linear and angular motions of dipper handle
and dipper. A path trajectory is modeled to demonstrate the dynamic velocity and acceleration
profiles. Numerical examples show that the critical performance variables include geometrical
and physical properties of the dipper and dipper handle, digging strategies and formation
properties. The kinematic results show that the critical phase occurs between 1.5 and 2.0 s
of a 3-s excavation cycle with occurrence of maximum kinematic effects. The dynamic results
also show a similar trend with maximum dynamic effects between 1.5 and 2.0 s. The results

*
Corresponding author. Present address: Centre for Advanced Minerals and Energy Research, Civil–
Electrical Building, Room 220, University of Alberta, Edmonton, Alta., Canada T6G 2G7. Tel.: +1-780-
492-4053; fax: +1-780-492-0249.
E-mail address: [email protected] (S. Frimpong).

0022-4898/$20.00 Ó 2004 ISTVS. Published by Elsevier Ltd. All rights reserved.

doi:10.1016/j.jterra.2004.06.002
16 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33

also show that the maximum resistive force occurs at 1.625 s within the excavation cycle. At
this point the maximum breakout force of the equipment is reached and any increase in the
resistive load will require further fragmentation. The results provide appropriate information
for excavation planning and execution. These models form the basis for developing dynamic
shovel simulators for the ISE technology.
Ó 2004 ISTVS. Published by Elsevier Ltd. All rights reserved.

Keywords: Intelligent excavation; Surface mining; Dynamic simulation; Numerical modeling

Nomenclature

The following symbols are used in developing the cable shovel kinematics and
dynamic equations
CðH; HÞ _ generalized Coriolisand centripetaltorque
D(H) generalized inertiamatrix
G(H) generalized gravity torque
(H) vector of generalized variables
Fn normal reaction force
Ft tangential reaction force
F cable shovelÕs breakout force (design parameter)
Fload(Ft, Fn) = Fr resistive force due to the formation being excavated by the
shovel dipper
Cij _
(i, j) element of Coriolis and centripetal torque CðH; HÞ
Dij (i, j) element of inertia matrix D(H)
Gi ithcomponent of gravity torque G(H)
Cij (i, j) element of control matrix C(h)
ci, si coshi and sinhi, respectively
cij, sij cos(hi + hj) and sin(hi + hj), respectively
l1 length of of crowd arm from pivotal point to connection point between
arm and dipper
l2 length between dipper tip and connect point of arm and dipper
mi mass of linki
m1 mass of crowd arm
m2 mass of dipper
0
x0 angular velocity of the shovel base
0_
x0 angular acceleration of the shovel base
1
x1 angular velocity at the center of the crowd arm or dipper handle
1_
x1 angular acceleration at the center of the crowd arm
1_
v c1 linear acceleration at the center of the crowd arm
s1 torque at the center of the crowd arm
1
N1 torque at the center of the crowd arm
1
F1 inertial force at the center of the crowd arm
1
f1 inward iteration of the crowd arm
 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 17

1
n1 torque balance of the crowd arm
X^ i ; Y^ i ; Z^ i ði ¼ 1; 2Þ displacement vectors along the ith 3D Cartesian coordinate
systems
1
0R transformation matrix from base frame to the crowd arm based coor-
dinate
2
1R transformation matrix from crowd arm based coordinate to the dipper
based coordinate
3
2R transformation matrix from dipper based coordinates to the dipper tip
based coordinate
hi ; h_ i ; €
hi angular displacement, velocity and acceleration of joint i [i = 1 (crowd
arm), 2 (dipper)]
d i ; d_ i ; d€i linear displacement, linear velocity and acceleration of joint i [i = 1
(crowd arm), 2 (dipper)]
0
P1 location vector for the rotation point of the crowd arm
1
Pc1 location vector for the center of the mass of the crowd arm
2
x2 angular velocity at the center of the dipper
2_
x2 angular acceleration at the center of the dipper
2_
vc 2 linear acceleration at the center of the dipper
2
N2 torque at the center of the dipper
2
F2 inertial force at the center of the dipper
2
f2 inward iteration of the dipper
2
n2 torque balance of the dipper
3
f3 inward iteration of the dippertip
3
n3 torque balance of the dippertip
f31, f32 respective xand y components of the interaction force between soil and
dipper tip
Izz1 moment of inertia of crowd arm about centroidal axis parallel to z1-
axis
Izz2 moment of inertia of dipper about centroidal axis parallel to z2-axis
~
k zi unit vector on the zi-axis
di offset distance of the gravity center in link i
d1 displacement between O1 and C1 in Fig.10
d2 displacement between O2 and C2 in Fig.10
Ii moment of inertia of link I about centroidal axis parallel to zi-axis
~
k zi unit vector on the zi-axis
hb angle between dipper bottom and O2 D in Fig.11
hc 2 angle between O2 D and O2 C2 in Fig.11
C0 compactness and cutting resistance index
s cutting edge index
ez tool plate thickness
kz index for the type of cutting
d cutting plate thickness
w length of horizontal chip
b angle of cutting
18 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33

1. Introduction

Shovel excavators are widely used as primary production equipment in surface

mining operations for removing overburden and ore materials. The shovel-truck
mining method in surface mining is flexible, efficient and it can be easily relocated
to different operating environments. The efficiency and costs of mining operations
greatly depend on the efficient use of these capital-intensive shovel excavators. How-
ever, the shovel-truck mining method can be rendered inefficient from operating, en-
vironmental, topography, and operator constraints. The physical and mechanical
properties of the materials being excavated also have severe impact on the efficiency
of the shovel excavator. Complex and unstructured mining environments have a
huge impact on the shovel dipper and tooth assembly. They usually result in varying
mechanical energy input and stress loading of shovel dipper-and-tooth assembly
across the working bench when the shovel dipper is in contact with mining surfaces
during excavation. Stress development and fatigue failure in equipment components
such as booms cause unplanned downtimes, reduced efficiency and high production
costs. A potential solution to the above problems is the application of an intelligent
shovel excavation (ISE) system that is responsive to interactive forces during excava-
tion as illustrated in Fig. 1 [1].
Fig. 1 consists of a shovel excavator, vision systems, sensors and receivers that
work together to capture mechanical energy, stress and resistance information from
a mining face, and process and display, in real time, on an on-board screen in the
operatorÕs cabin. An operator can therefore assess local variation in the bench envi-
ronment based on the information displayed on the screen and make real-time
decisions. These decisions may lead to fragmentation of extremely hard formations,
in-pit blending and stockpiling of materials as per the plant requirements. The devel-

Fig. 1. Sensor–vision system in the ISE technology [1].

 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 19

opment of the ISE system requires kinematic and dynamic models of the shovel op-
erating modes to describe the evolution of the excavator motion with time. Hend-
ricks et al. [2] developed a dynamic model for cable shovels using Lagrange
formulations without including the resistive forces from the mined environments.
Daneshmend et al. [3] dealt with the same problem using an iterative Newton–Euler
formulation. That work did not include the dipper handle crowd, which is very im-
portant for the complete description of the dynamic behavior of a cable shovel. Also
no model predictions were presented in these papers.
In this paper, the dynamic models of shovel excavators are derived using the
Newton–Euler formulation in the local coordinate frame. These models are based
on the main functional components as free bodies. The components include the dip-
per handle and the dipper for electrical cable shovels as shown in Fig. 2. The kinemat-
ics and dynamic responses from given digging trajectory are predicted. These models
form the basis for developing comprehensive simulator models for efficient shovel op-
erations in constrained mining environments. Section 2 contains the working mecha-
nism of a typical cable shovel, which forms the basis of the kinematics and dynamic
models. The models are presented in Section 3 with detailed derivations in Appendix
A. Sections 4 and 5 contain numerical example and associated results for a scaled ca-
ble shovel using a given digging trajectory, with corresponding kinematics and dy-
namics characteristics. Finally, appropriate conclusions are drawn in Section 6.

2. Cable shovel dynamics in surface mining

Fig. 2 illustrates the cable shovel excavator for bulk surface mine production
operations. The shovel consists of three major assemblies: the lower, upper and

Fig. 2. Nomenclature of a cable shovel [4].

20 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33

the attachment. The lower provides a stable base for the machine and includes the
propel drive and crawler system. The upper provides a platform for the hoist and
swing machinery, boom attachment, electronic control cabinets, operatorÕs cab
and supporting equipment. The attachment consists of the boom, crowd machinery,
dipper handle and dipper. The primary motions of a cable shovel include propel,
swing, hoist, and crowd/retract. The shovel uses the propel function to move (tram)
the machine from one digging site to another. The shovelÕs upper assembly is roller
and center pin mounted on the lower. Electric motors drive the swing transmissions
that in turn drive swing pinions engaged in the swing gear. Multiple swing transmis-
sions are typically used on electric mining shovels to swing the machine from the dig-
ging face to the haulage unit and back to the formation.
The hoist machinery on the cable shovel consists of a rope drum that is rotated by
the electric motor-driven hoist transmissions. Dipper lowering is accomplished when
the hoist drum is rotated in the opposite direction and a controlled amount of rope is
unspooled. The crowd motion refers to the motion of the dipper moving away from
the centerline of the machine and toward the digging face. Depending on the shovel
manufacturer, this motion and the associated retract motion are accomplished with
either a pinion driven dipper handle or a rope controlled dipper handle. In either
case, digging is accomplished when the dipper is crowded and hoisted in unison
up through the digging face. A complete digging phase involves the digging tool (dip-
per) to the bank, filling it and then raising it clear to the bank as illustrated in Fig. 3.
All the primary motions will influence the productivity of a cable shovel and/or its
operating environments. For example, proper positioning of the cable shovel by pro-
pel motion will result in efficient bucket digging. With the increase in bucket capacity
and limited space for dump truck parking, optimized swing strategies may shorten
the time necessary to spot dump trucks and, therefore, shorten the duty cycle and
improve productivity. Also, hoisting and crowd/retract motions are critical to the

Fig. 3. The digging phase of a cable shovel [4].

 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 21

efficient digging of excavated materials. Thus, a proper understanding of all motions

involved in the digging cycle is essential for efficient and productive digging.

3. Dynamic modeling of the cable shovel

As discussed previously, the dynamic equations can be obtained by applying

either Lagrange method or iterative Newton–Euler method. Here iterative New-
ton–Euler method is used because the iterative formulation lends itself to efficient
software implementation, which is of importance for real-time simulation and pa-
rameter estimation [2,3]. Also, it provides detail information on all links and joints,
which will be useful in possible stress and/or strength analysis of the components
during excavation. The Newton–Euler dynamic algorithm for computing the crowd
force and the hoist torque comprises two parts. First, the velocities and accelerations
are iteratively computed from dipper handle to dipper and the Newton–Euler equa-
tions are applied to each of them. Second, the interactive forces and torques and
joint actuator torques are computed recursively from dipper back to dipper handle
[5,6].
The attention here is restricted to the boom, dipper handle and dipper assembly of
the cable shovels because this research is only concerned with the digging section of a
duty cycle. Therefore, the rotation of the upper structure is not modeled and the dip-
per handle and dipper assembly only moves on the vertical plane. The dynamic equa-
tion for cable shovel excavator is given as Eq. (1) and the procedure is summarized in
the equations developed under Appendix A.
€ þ CðH; HÞ
DðHÞH _ H_ þ GðHÞ ¼ F
F load ðF t ; F n Þ; ð1Þ
where D(H), CðH; HÞ _ and G(H) are inertial matrix, Coriolis and centripetal effects,
and gravity, respectively. H is a vector of generalized variables. F is related to the
crowd and hoist force effects, and Fload(Ft, Fn) is a resistive force due to soil-dipper
interaction. For a cable shovel, the variables in equation can be summarized in
Eqs. (2)–(5) as developed in Appendix A.


m1 þ m2
m2 d 2 s2c2
DðHÞ ¼ ; ð2Þ

m2 d 2 s2c2 I zz1 þ I zz2 þ m1 d 21 þ m2 ðl21 þ 2l1 d 1 c2c2 þ d 21 Þ
" #
_ ¼ 0
ðm1 d 1 þ m2 ðl1 þ d 2 c2c2 ÞÞh_ 1
CðH; HÞ ;
2ðm1 d 1 þ m2 ðl1 þ d 2 c2c2 ÞÞh_ 1 0
ð3Þ


ðm1 þ m2 Þgs1
GðHÞ ¼ ; ð4Þ
ðm1 d 1 c1 þ m2 ðl1 c1 þ d 2 c12c2 ÞÞg


F t c2hb
F n s2hb
F load ðF t ; F n Þ ¼ : ð5Þ
F t ðl1 þ l2 Þs2hb þ F n ðl1 þ l2 Þc2hb
22 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33

It may be observed from Eq. (1) that F, which is the force provided by crowd
and hoist motors, is mainly used to overcome two parts of external effects.
One is the dynamic part, including the inertial effect, Coriolis and centripetal ef-
fects, and gravity. The other is the interaction force between the bucket and the
surrounding media. The inertial part, D(H), and gravity part, G(H), summarize
the effect of geometrical and material properties of the dipper handle and dipper,
while CðH; HÞ_ is an indicator of the kinematic and dynamic effect of the digging
strategies including the digging profile and time distribution during the digging
process. Fload(Ft, Fn) is mainly related to the physical and mechanical properties
of the materials being excavated and the geometrical characteristics of the dipper
although other effects such as digging strategies may not be negligible. Other ef-
fects, which are not considered here, may include friction and losses in the trans-
mission process. Therefore, from the dynamic model, it can be seen that the
material properties to be excavated, the geometrical and physical properties of
the dipper handle and dipper, as well as the digging strategies are the main fac-
tors in determining the required crowd and hoist forces, and thus, the perform-
ance of a cable shovel. The formulation of the dynamic model presented above
depends on precise knowledge of the various kinematic and dynamic parameters,
which characterize a particular cable shovel. These parameters can be estimated
from data supplied by shovel manufacturers.

4. Numerical examples of cable shovel dynamics

A scaled cable shovel model is used here for the demonstration of the kinemat-
ics and dynamic behavior of the cable shovel during surface mining excavation.
The main geometry and physical data for the shovel simulated is listed in Table
1. The digging trajectory shown in Fig. 4 is first generated based on the reference
coordinate (xw, yw, zw) in Fig. 10. Fig. 4 shows a generalized excavator bucket tip
trajectory for completing the digging section of a working cycle using this refer-
ence coordinate. The bucket begins the digging cycle at point A and completes
the cycle at point C. The excavator sits on the elevation A–B. The amount of
digging below and/or above A–B elevation depends on the type of excavator.
The bucket penetrates into formation at point A until it reaches a maximum

Table 1
Main shovel structural data for a scaled cable shovel
Length (m) Mass (kg) Inertia momentsa (kg m2)
b
Handle 11.888 65800 7.749E + 5
Dipper 5.800c 24500 6.458E + 4
Bucket capacity 30.6 m3
a
Moment of inertia about the gravitational center.
b
Length between the end of the handle and the joint for handle and bucket.
c
Length between the joint for handle and bucket and bucket tip.
 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 23

Fig. 4. A trajectory for a shovel dipper tip.

cut depth of 0.186 m. Then it starts to rise to be clear of the formation. It takes
3.5 s to finish the whole trajectory.

5. Discussions and analysis of results

For the assumed dynamic trajectory applied to the dipper tip of the cable shovel
as shown in Figs. 4–7 show the change of ðd 1 ; h1 Þ; ðd_ 1 ; h_ 1 Þ and ðd€1 ; €h1 Þ with time t,
respectively. Note that in all these figures, the vertical axes have been normalized
for a comparative purpose. In Fig. 5, the dipper handle,d1, advances at first, experi-
ences a maximum displacement around the middle of the path profile (t = 1.75–2.0 s)
and then retracts following the given digging profile, while h1 always increases with
time. Although h1 is increasing constantly, its rate of change, h_ 1 , is not constant as
illustrated in Fig. 6. It has a maximum around the path mid-point (t = 1.75–2.0 s).
For the dipper handle crowd rate, d_ 1 , it is decreasing constantly as shown in Fig.
6. The dipper handle crowd acceleration, d€1 , and the angular acceleration, €h1 , are
more complex for the given digging path profile and are illustrated in Fig. 7. The an-
gular acceleration first increases to a maximum at t = 1.25 s, and then decreases to a
minimum at t = 2.25 s and gradually increases through the rest of the excavation cy-
cle. The dipper handle crowd acceleration decreases at first, experiences a minimum
acceleration at t = 1.625 s and then increases throughout the remaining excavation
cycle. It is noted from these figures that the dipper handle crowd and angular rota-
tion have different trends. The dipper handle crowd for the given digging trajectory
has a much smaller change than that of the angular rotation, indicating a
24 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33

Fig. 5. d1 and h1 with time t.

1.0
Normalized Gen.Vel.

0.5

0.0 d 1 /|d1 | max

θ 1 /| θ1 |max

-0.5

0.0 0.5 1.0 1.5 2.0 2.5 3.0

t (s)

Fig. 6. d_ 1 and h_ 1 with time t.

predominant rotation effect in the inertial part. When d 1 ; d_ 1 ; d€1 , and h1 ; h_ 1 ; €h1 are de-
termined for each time point, the left-hand side of Eqs. (A18) and (A19) can be cal-
culated and their normalized changes with time are shown in Fig. 8. This figure
shows that the dipper handle crowd inertial force and torque have opposite trends
with time. The inertial force reaches a local maximum early in the cycle, gradually
decreases to a local minimum, increases to a global maximum at t = 1.875 s and fi-
nally decreases sharply to a global minimum at the end of the cycle. The crowd arm
torque decreases to a minimum at t = 2.0 s and gradually increases to the end of the
cycle. It is also noted that there are some jaggednesses for d€1 at the very early stage,
which may be attributed to the time differential of the d1 and h1, not from numerical
inaccuracies.
The cable shovel is designed with a breakout force, which is the maximum force
available to overcome the resistive force from the material being excavated by the
 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 25

1.0
.. ..
d 1/|d 1|max
.. ..
Normalized Gen.Acce.
0.5 θ1/| θ1|max

0.0

-0.5

-1.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0

t (s)

Fig. 7. d€1 and €h1 with time t.

-0.6

F1/|F 1|max
τ 1/|τ1|max
-0.7
Normalized Gen. Force

-0.8

-0.9

-1.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0

t (s)

Fig. 8. Inertial force and torque with time t.

shovel. The difference between the breakout force, F, and the resistive force, Fr, de-
termines whether or not the formation can be excavated by the shovel. There are dif-
ferent models for predicting the resistive force, Fr. Zelenin et al. [7] provide a basis
for estimating the resistive force Fr as shown in the following equation:
F r ¼ 10C 0 d 1:35 ð1 þ 2:6wÞð1 þ 0:0075bÞð1 þ 0:03sÞez k z ; ð6Þ
where C0 is compactness and cutting resistance index, d is depth of cutting, w is
length of horizontal chip, b is the angle of cutting, s is the cutting edge index, ez is
the tool plate thickness and kz is the index for the type of cutting. Based on the Ze-
lenin model, the resistive force according to the current trajectory is calculated and
shown in Fig. 9. The resistive force increases sharply from the beginning of the
26 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33

Fr: Resistance
0.0 Force

Resistive Force Fr/(Fr)max

-0.4

-0.8

-1.2
0.0 0.5 1.0 1.5 2.0 2.5 3.0
t (s)

Fig. 9. Normalized resistive force with time.

excavation cycle to a maximum at t = 1.625 s and decreases from there at the same
rate to the end of the excavation cycle. At t = 1.625 s, the breakout force of a given
shovel is reached and further increase in the resistive force will require a shovel with
greater breakout force or increased fragmentation. Alternatively, Eq. (1) may be
solved for the breakout force, F, to select an appropriate shovel that can excavate
the formation. Then the relationship between the crowd force and hoist torque from
crowd and hoist motors and the resistive force Fr may be established. Therefore, by
measuring the the crowd force and hoist torque, one can know the current resistive
force from the mined materials and, therefore, other information such as diggability.

6. Conclusions and contributions

Cable shovel dynamic models have been developed by analyzing the main func-
tional components (the dipper handle and the dipper) as free body diagrams using
an iterative Newton–Euler method. The models include the main factors that influ-
ence the performance of a cable shovel including the effect of both linear and angular
motions of dipper handle and dipper. Based on the dynamic model, it is identified
that the material properties to be excavated, the geometrical and physical properties
of the dipper handle and dipper, as well as the digging strategies are the main factors
that determine the required crowd and hoist forces and, thus, the performance of a
cable shovel. The responses from a given digging trajectory are predicted as an illus-
tration of the dynamic velocity and acceleration profiles. The kinematic results show
that the critical phase occurs between 1.5 and 2.0 s of a 3-s excavation cycle with oc-
currence of maximum kinematic effects. The dynamic results also show a similar
trend with maximum dynamic effects between 1.5 and 2.0 s. The results also show
that the maximum resistive force occurs at 1.625 s within the excavation cycle. At
this point the maximum breakout force of the equipment is reached and any increase
 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 27

in the resistive load will require further fragmentation. The results provide appropri-
ate information for excavation planning and execution. The next step in this research
is the characterizing, modeling and simulation of the interaction between the shovel
and the mining environments. This will eventually enable complete simulation of the
shovel operation, as well as assessment of ore bench digging conditions based on on-
line measurements of the information from the main functional components of the
shovel excavators.

Appendix A. Detailed mathematical modeling of cable shovel dynamics

The digging path of the dipper is produced by the extension/retraction of the han-
dle (crowd) and by the cable hosting action as in Fig. 3. Hoisting of the dipper is ac-
complished by means of cables attached to the dipper, which pass over sheaves at the
boom point and spool on a deck mounted powered drum. The crowd action is pro-
duced either by cables or a direct rack and pinion gear drive. In Fig. 10, Xw, Yw, Zw
are the coordinates at the lower and upper structures of the shovel, O0, X0, Y0, Z0
(base frame), O1, X1, Y1, Z1 and O2, X2, Y2, Z2 are the respective coordinates of the
boom, crowd arm (dipper handle) and dipper. It is considered that the upper struc-
tures of the cable shovel are fixed, and thus, the kinematics and dynamic models of a
cable shovel excavator are mainly related to dipper handle and dipper.

Fig. 10. Elements of shovel dynamics.

28 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33

The Newton–Euler method is utilized to analyze the kinematics and dynamic be-
havior of the cable shovel. The advantage of this method is that it follows an iterative
procedure and it is easy to implement in a simulation environment. Also, it provides
detail information on all links and joints, which will be useful in possible stress and/
or strength analysis of the components during excavation. The Newton–Euler dy-
namic algorithm for computing the crowd force and the hoist torque comprises
two parts. First, the velocities and accelerations are iteratively computed from dipper
handle to dipper and the Newton–Euler equations are applied to each of them. Sec-
ond, the interactive forces and torques and joint actuator torques are computed re-
cursively from dipper back to dipper handle [5]. The procedure is summarized in the
equations below. The upper structure of the shovel is fixed, and hence it does not ro-
tate during operation, resulting in the following equation:
2 3 2 3
0 0
0
0 6 7 d x 0 6 7
x0 ¼ 4 0 5; ¼ 4 0 5; ðA1Þ
dt
0 0

where 0x0 and 0 x_ 0 ¼ ðd0 x0 =dtÞ are the respective angular velocity and angular accel-
eration of the shovel boom. The first step of the Newton–Euler method is an out-
ward iteration for the crowd arm. This procedure includes the calculation of the
angular velocity, 1x1, angular acceleration, 1 x_ 1 , the linear acceleration at the gravity
center of the crowd arm or dipper handle, 1 v_ c1 , the torque at the gravity center of the
crowd arm, 1N1, and the inertial force at the gravity center of the crowd arm, 1F1.
The angular velocity is given by the equation
0 10 1 0 1 0 1
c 1 s1 0 0 0 0
1 1 0 _ 1^ B CB C B C B C
x1 ¼ 0 R x0 þ h1 Z 1 ¼ @
s1 c1 0 A@ 0 A þ @ 0 A ¼ @ 0 A; ðA2Þ
0 0 1 0 h_ 1 h_ 1

where 10 R is the transformation matrix from base frame to the crowd arm based co-
ordinate and h_ 1 is the rotation angular velocity of the crowd arm based coordinate
relative to the base frame. 1x1 is necessary for estimating the Coriolis and centripetal
effects of the crowd arm. The angular acceleration of the crowd arm is given by
1
x_ 1 ¼ 10 R0 x0  h_ 1 1 Z^ 1 þ 10 R0 x_ 0 þ €
h1 1 Z^ 1
0 10 1 0 1 0 10 1 0 1
c 1 s1 0 0 0 c1 s1 0 0 0
B CB C B C B CB C B C
¼ @
s1 c1 0 A@ 0 A  @ 0 A þ @
s1 c1 0 A@ 0 A þ @ 0 A
0 0 1 0 h_ 1 0 0 1 0 €h1
0 1
0
B C
¼ @ 0 A; ðA3Þ
€
h1

where €h1 is the rotation angular acceleration of the crowd arm based coordinate rel-
ative to the base frame. 1 x_ 1 is used to calculate the angular moment of the crowd
 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 29

arm. The linear acceleration, 1 v_ 1 , at the connection point of the crowd arm is given
by the equation
   
1
v_ 1 ¼ 10 R 0 x_ 0  0 P 1 þ 0 x0  0 x0  0 P 1 þ 0 v_ 0 þ 21 x1  d_ 1 1 X^ 1 þ d€1 1 X^ 1
0 1
gs1 þ d€1
B C
¼ @ gc1 þ 2d_ 1 h_ 1 A; ðA4Þ
0
where P1 is a position vector, which locates the rotation point of the crowd arm. d_ 1
0

and d€1 are the linear velocity and linear acceleration of the crowd arm, respectively.
The gravity effect 0 v_ 0 is calculated as 0 v_ 0 ¼ g^y 0 . The linear acceleration at the gravity
center of the crowd arm is given by
0 2 1

d 1 h_ 1 þ d€1 þ gs1
1
  B C
v_ c1 ¼ 1 x_ 1  1 P c1 þ 1 x1  1 x1  1 P c1 þ 1 v_ 1 ¼ @ d 1 €h1 þ 2d_ 1 h_ 1 þ gc1 A; ðA5Þ
0
1
where Pc1 is the position vector, which locates the center of the gravity of the crowd
arm. 1 v_ c1 is used to calculate the inertial force of the crowd arm. EulerÕs equation is used
to calculate the torque, 1N1, at the gravity center of the crowd arm in the equation
1
N 1 ¼ c1 I 1 1 x_ 1 þ 1 x1  c1 I 1 1 x1
0 10 1 0 1 00 10 11
0 0 0 0 0 0 0 0 0
B CB C B C BB CB CC
¼ @ 0 0 0 A@ A @ A @@ 0 0 0 þ 0  0 A@ 0 AA
0 0 I zz1 €
h1 h_ 1 0 0 I zz1 h_ 1
0 1
0
B C
¼ @ 0 A: ðA6Þ
I zz1 €
h1
NewtonÕs equation is applied to calculate the inertial force, 1F1, at the gravity center
of the crowd arm in the equation
0 2 1

m1 d 1 h_ 1 þ m1 d€1 þ m1 gs2
1 B C
F 1 ¼ m1 1 v_ c1 ¼ @ m1 d 1 €
h1 þ 2m1 d_ 1 h_ 1 þ m1 gc1 A: ðA7Þ
0
The same procedure can be applied to the dipper to calculate its angular velocity,
2
x2; angular acceleration, 2 x_ 2 ; the linear acceleration at the gravity center of the dip-
per, 2 v_ c2 ; the torque at the gravity center of the dipper, 2N2; and the inertial force at
the gravity center of the dipper, 2F2. They are
0 10 1 0 1 0 1
c2 s2 0 0 0 0
2 2 1 _ 2^ B CB C B C B C
x2 ¼ 1 R x1 þ h2 Z 2 ¼ @
s2 c2 0 A @ 0 A þ @ 0 A ¼ @ 0 A; ðA8Þ
0 0 1 h1 0 h_ 1
30 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33

2
x_ 2 ¼ 21 R1 x1  h_ 2 2 Z^ 2 þ 21 R1 x_ 1 þ €
h2 2 Z^ 2
0 10 1 0 1 0 10 1
0 1
c 2 s2 0 0 0 c2 s2 0 0 0
B CB C B C B CB C B C
¼B CB C B C B
@
s2 c2 0 A@ 0 A  @ 0 A þ @
s2 c2 0C B C B C
A@ 0 A þ @ 0 A
0 0 1 h_ 1 0 0 0 1 h€1 0
0 1
0
B C
¼B C
@ 0 A; ðA9Þ
€
h1
2
v_ 2 ¼ 21 Rð1 x_ 1  1 P 2 þ 1 x1  ð1 x1  1 P 2 Þ þ 1 v_ 1 Þ þ 22 x2  d_ 2 2 X^ 2 þ d€2 2 X^ 2
0 2
1

l1 h_ 1 c2 þ d€1 c2 þ l1 €
h1 s2 þ 2d_ 1 h_ 1 s2 þ gs12
B C
¼ @ l h_ 2 s
d€ s þ l € h c þ 2d_ h_ c þ gc A; ðA10Þ
1 1 2 1 2 1 1 2 1 1 2 12
0
2
v_ c2 ¼ 2 x_ 2  2 P c2 þ 2 x2  ð2 x2  2 P c2 Þ þ 2 v_ 2
0 2
1
h1 þ 2d_ 1 h_ 1 s2 þ d€1 c2 þ gs12

ðl1 c2 þ d 2 cc2 Þh_ 1 þ ðl1 s2
d 2 sc2 Þ€
B C
¼ @ ðl s
d s Þh_ 2 þ ðl c þ d c Þ€ _ _ € A; ðA11Þ
1 2 2 c2 1 1 2 2 c2 h1 þ 2d 1 h1 c2
d 1 s2 þ gc12
0
2
N 2 ¼ c2 I 2 2 x_ 2 þ 2 x2  c2 I 2 2 x2
0 1 0 1 0 1 00 10 11
0 0 0 0 0 0 0 0 0
¼ @ 0 0 0 A @ 0 A þ @ 0 A  @@ 0 0 0 A@ 0 AA
0 0 I zz2 €
h1 h_ 1 0 0 I zz4 h_ 1
0 1
0
¼ @ 0 A; ðA12Þ
I zz2 €
h1
2
F 2 ¼ m2 2 v_ c2
0 2
1
h1 þ 2m2 d_ 1 h_ 1 s2 þ m2 d€1 c2 þ m2 gs12

m2 ðl1 c2 þ d 2 cc2 Þh_ 1 þ m2 ðl1 s2
d 2 sc2 Þ€
B C
¼ @ m ðl s
d s Þh_ 2 þ m ðl c þ d c Þ€ _ _ € A:
2 1 2 2 c2 1 2 1 2 2 c2 h1 þ 2m2 d 1 h1 c2
m2 d 1 s2 þ m2 gc12
0
ðA13Þ

Having computed the forces and torques acting on the crowd arm and the dipper,
it now remains to calculate the joint torques which will result in these net forces
and torques being applied to the crowd arm and the dipper. This is achieved by
writing force and moment balance equations based on free body diagrams of the
crowd arm and the dipper. The inward iteration for the dipper is first calculated
as follows:
 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 31

Fig. 11. Interaction between dipper and environment.

2
f2 ¼ 32 R3 f3 þ 2 F 2
0 2
1

m2 ðl1 c2 þ d 2 cc2 Þh_ 1 þ m2 ðl1 s2
d 2 sc2 Þ€h1 þ 2m2 d_ 1 h_ 1 s2 þ m2 d€1 c2 þ F t chb
F n shb þ m2 gs12
B C
¼ @ m ðl s
d s Þh_ 2 þ m ðl c þ d c Þ€h þ 2m d_ h_ c
m d€ s þ F s þ F c þ m gc A;
2 1 2 2 c2 1 2 1 2 2 c2 1 2 1 1 2 2 1 2 t hb n hb 2 12
0
ðA14Þ

where Ft and Fn are the tangential and normal components of the interaction force,
respectively, between the dipper tip and the media (Fig. 11). Eq. (A14) is the force
balance equation for the dipper. The torque balance equation for the dipper is given
by the equation
2
n2 ¼ 2 N 2 þ 23 R3 n3 þ 2 P c2  2 F 2 þ 2 P 3  23 R3 f3
0 1
0
¼@ 2
0 A
I zz2 €
h1 þ m2 l1 d 2 s2c2 h_ 1 þ m2 ðl1 c2c2 þ d 2 Þd 2 €h1 þ 2m2 d 2 d_ 1 h_ 1 c2c2
m2 d 2 d€1 s2c2
0 1
0
þ@ 0 A:
m2 d 2 gc12c2 þ F t l2 s2hb þ F n l2 c2hb
ðA15Þ

Following the same procedure, the inward iteration for the crowd arm is also given
by

1
f1 ¼ 21 R2 f2 þ 1 F 1
0 2
1

ðm1 d 1 þ m2 ðl1 þ d 2 c2c2 ÞÞh_ 1
m2 d 2 s2c2 €h1 þ ðm1 þ m2 Þd€1 þ F t c2hb
F n s2hb þ ðm1 þ m2 Þgs1
B C
¼ @
m d s h_ 2 þ ðm d þ m ðl þ d c ÞÞ€h þ 2ðm þ m Þd_ h_ þ F s þ F c þ ðm þ m Þgc A;
2 2 2c2 1 1 1 2 1 2 2c2 1 1 2 1 1 t 2hb n 2hb 1 2 1
0
ðA16Þ
32 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33
0 1
0
1 B C
n1 ¼ 1 N 1 þ 12 R2 n2 þ 1 P c1  1 F 1 þ 1 P 2  12 R2 f2 ¼ @ 0 A
ðI zz1 þ I zz2 Þ€
h1
0 1
0
B C
þ@ 0 A
ðm1 d 21 þ m2 ðl21 þ 2l1 d 1 c2c2 h1 þ 2ðm1 d 1 þ m2 ðl1 þ d 2 c2c2 ÞÞd_ 1 h_ 1
m2 d 2 d€1 s2c2
þ d 21 ÞÞ€
0 1 0 1
0 0
B C B C
þ@ 0 Aþ@ 0 A: ðA17Þ
ðm1 d 1 c1 þ m2 ðl1 c1 þ d 2 c12c2 ÞÞg F t ðl1 þ l2 Þs2hb þ F n ðl1 þ l2 Þc2hb

Eqs. (A16) and (A17) are the respective force and moment balance equations for the
crowd arm. Finally, the equations for the crowd inertial force and torque are ob-
tained in the following equations:
2
F 1 ¼ ðm1 þ m2 Þd€1
ðm1 d 1 þ m2 ðl1 þ d 2 c2c2 ÞÞh_ 1

m2 d 2 s2c2 €
h1 þ ðm1 þ m2 Þgs1 þ F t c2hb
F n s2hb ; ðA18Þ

s1 ¼ ðI zz1 þ I zz2 Þ€
h1 þ ðm1 d 21 þ m2 ðl21 þ 2l1 d 1 c2c2 þ d 21 ÞÞ€h1
þ 2ðm1 d 1 þ m2 ðl1 þ d 2 c2c ÞÞd_ 1 h_ 1
m2 d 2 d€1 s2c
2 2

þ ðm1 d 1 c1 þ m2 ðl1 c1 þ d 2 c12c2 ÞÞg þ F t ðl1 þ l2 Þs2hb þ F n ðl1 þ l2 Þc2hb :

ðA19Þ
Eqs. (A18) and (A19) can be rearranged concisely as follows:
 
 
F1 m1 þm2
m2 d 2 s2c2 d€1
¼
s1
m2 d 2 s2c2 I zz1 þI zz2 þm1 d 21 þm2 ðl21 þ2l1 d 1 c2c2 þd 21 Þ €h1

 
0
ðm1 d 1 þm2 ðl1 þd 2 c2c2 ÞÞh_ 1 d_ 1
þ
2ðm1 d 1 þm2 ðl1 þd 2 c2c2 ÞÞh_ 1 0 h_ 1

 
ðm1 þm2 Þgs1 d1
þ
ðm1 d 1 c1 þm2 ðl1 c1 þd 2 c12c2 ÞÞg h1


F t c2hb
F n s2hb
þ : ðA20Þ
F t ðl1 þl2 Þs2hb þF n ðl1 þl2 Þc2hb

If H is assumed to be {d1 h1}T for cable shovels, D(H), CðH; HÞ, _ G(H) and
Fload(Ft, Fn) in Eqs. (2)–(5) can be determined from Eqs. (A20) and (A20) becomes
the dynamic model for the motion of cable shovels illustrated by Eq. (1).

References

[1] Frimpong S, Szymanski J, Pedrycz W, Gao Y. Intelligent shovel excavation in varying oil sands
formation and bitumen content. A proposal submitted to COURSE, University of Alberta, Edmonton,
Canada; 2001.
 S. Frimpong et al. / Journal of Terramechanics 42 (2005) 15–33 33

[2] Hendricks C, Daneshmend L, Wu S, Scoble M. Design of a simulator for productivity analysis of

electric mining shovels. In: Proceedings of 2nd international symposium on mine mechanization and
Automation, Lulea, Sweden; 1993.
[3] Daneshmend L, Hendricks C, Wu S, Scoble M. Design of a mining shovel simulator. Innovative mine
design for the 21st century, Kingston, Ont., Canada; 1993.
[4] Peak performance: excavator selection. P & H MinePro Services; 2003.
[5] Craig JJ. Introduction to robotics: mechanics and control. Reading (MA): Addison-Wesley; 1986.
[6] Vaha PK, Skibniewski MJ. Dynamic model of excavator. J Aerospace Eng 1993;6:148–66.
[7] Zelenin AN, Balovnev VI, Kerov IP. Machines for moving the earth. New Delhi: Amerind
Publishing; 1985.