Transcript
Occupational traineeship report
Performance evaluation of a Liebherr 996 hydraulic shovel. DCT 2005.59 Tom Kemps
University of Queensland Faculty of Mechanical Engineering Brisbane, Australia
Supervisors: Dr. Ross McAree Prof. Dr. ir. M. Steinbuch
[email protected] [email protected] [email protected]
Co-workers: Paul Siegrist
[email protected]
Performance evaluation of a Liebherr 996 hydraulic shovel
Abstract This report represents a part of a project that evaluates the performance of a Liebherr 996 hydraulic shovel operating in an open-cut nickel mine. A Liebherr 996 weights 668 tonnes and can move 60 tonnes of material each dig. The evaluation is done by establishing correlations between operator style, the productivity they achieve under different digging conditions and the machine damage the operators cause. Several sensors and a computer are installed on the shovel to measure and log a.o. hydraulic cylinder extensions, oil pressures, joystick signals. A kinematic model of the shovel arm is used to calculate the bucket position and orientation out of the cylinder extension data. This calculation is done using the multi dimensional Newton-Raphson method for non-linear systems. The parameters for the model are estimated experimentally using photos. The results of these estimations are not accurate enough and a calibration is performed by matching the simulation plots with the calibration photos. In the two 2-week field trials at the mine site the following data is collected: sensor data, annotation data (dig, swing dump, return, wait, propel, etc.), video data, diggability data, payload data. The annotation data is manually generated each day for 3 to 5 hours. Each annotation session is video recorded. All data is summarized in a spreadsheet. On average the operators fill a truck in 5.8 passes with 228.5 tonnes dirt and the average bucket load is 40.5 tonnes. The time it takes to load a truck is 3 min. and 23 sec. and the total dig-swing-dump-return cycle time is 35.4 sec. with a mean dig time of 24.3 sec. The mean face height is 14.8 m. Bucket teeth trajectories are visualized in x-y-plots and dig, swing, dump and return times are presented in box plots to study the operator style. To find relations between operator style, diggability and productivity these plots are printed together with the corresponding bucket load, bench height, diggability, dumps per truck and truckload. Clear correlations could manually not be found studying these prints and a more systematic approach is needed. The used approach is verifying hypotheses using analysis of variance (anova) in Matlab. The assumption that operators stand further back while digging a high bench appeared to be untrue. The cycle times digging high faces are not longer either. Cycle times are shorter and the bucket load is higher when the digging conditions are good is proven to be true.
2
Performance evaluation of a Liebherr 996 hydraulic shovel
Samenvatting Dit verslag gaat over een deel van een project waarin gekeken wordt naar de prestatie van een Liebherr 996 hydraulische graafmachine die men gebruikt in een nikkel dagbouwmijn. De Liebherr 996 weegt 668 ton en kan 60 ton materiaal in 1 keer verplaatsen. Het onderzoek wordt gedaan door correlaties te zoeken tussen bestuurdersstijl, de productiviteit bij de verschillende graafcondities en beschadiging die aan de machine wordt toegebracht door de bestuurder. Verscheidene sensoren en een computer zijn geïnstalleerd op de graafmachine om o.a. hydraulische cilinderverlengingen, oliedrukken and joystick signalen te meten en vast te leggen. Om de bucketpositie en -oriëntatie te berekenen uit de cilinderverlengingsdata is een kinematisch model gebruikt. Dit model is opgelost door gebruik te maken van de multidimensionale Newton-Raphson methode voor niet-lineaire systemen. De parameters die nodig zijn voor het model zijn experimenteel afgeschat door gebruik te maken van foto’s. De nauwkeurigheid van de resultaten hiervan zijn onacceptabel en door de simulatieplots te vergelijken met de kalibratiefoto’s zijn de parameters gekalibreerd. Gedurende twee bezoeken aan de mijn die elk 2 weken duurde is de volgende data verzameld: sensordata, annotatiedata (graven, draaien, dumpen, terugkeren, wachten, voortbewegen, etc.), videodata, graafbaarheidsdata en truckladingsdata. De annotatiedata is elke dag handmatig gegenereerd gedurende 3 tot 5 uur. Elke annotatiesessie is gefilmd. Alle data is samengevat in een tabel. Gemiddeld vult de bestuurders een truck in 5.8 keer dumpen met 228.5 ton materiaal en de gemiddelde bucket lading is 40.5 ton. De tijd die nodig is voor het vullen van een truck is 3 min en 23 sec. en de totale graaf-draaidump-terugkeer-ketentijd is 35.4 sec. met een gemiddelde graaftijd van in 24.3 sec. De hoogte van de te graven ‘gevel’ is gemiddeld 14.8 m. De bewegingsbanen van de buckettanden zijn gevisualiseerd in x-y-plots en de graaf-, draai-, dump- en terugkeerperiodetijden in boxplots om de bestuurdersstijl te bestuderen. Om relaties tussen bestuurdersstijl, graafbaarheid en productiviteit te vinden, zijn de plots uitgeprint samen met de bijbehorende bucketlading, gevelhoogte, graafbaarheid, dumps per truck en trucklading. Duidelijke correlaties zijn handmatig niet gevonden door de prints te bestuderen en daarom is er een systematischere methode nodig. De toegepaste methode test hypotheses door middel van analyse van de variantie (anova) in Matlab. De aanname dat bestuurders verder van de gevel af gaan staan wanneer deze hoog is, blijkt onwaar te zijn. Ook zijn de graaf-draai-dump-terugkeer-ketentijden niet langer in het geval van hoge gevels. Het is wel zo dat die ketens korter duren als de graafbaarheid goed is. In die goede graafbaarheidssituatie is de bucketlading beter.
3
Performance evaluation of a Liebherr 996 hydraulic shovel
Table of Contents Abstract ...............................................................................................................2 Samenvatting ......................................................................................................3 Table of Contents................................................................................................4 Glossary ..............................................................................................................6 1 Introduction ......................................................................................................8 2 Kinematic model for the Liebherr996 shovel.................................................9 2.1 Introduction..................................................................................................9 2.2 Kinematic constraints ................................................................................10 2.3 Newton-Raphson.......................................................................................12 3 Parameter estimation for kinematic model..................................................15 3.1 Introduction................................................................................................15 3.2 Photogrametric method for estimating lengths ..........................................15 3.3 Measurement of the hydraulic cylinder lengths. ........................................17 3.4 Kinematic Calibration ................................................................................18 4 Productivity analysis .....................................................................................19 4.1 Introduction................................................................................................19 4.2 Summary of data collected ........................................................................19 4.3 explanation summary table........................................................................19 4.4 Overall .......................................................................................................19 4.5 Summary data tables ................................................................................20 5 Results............................................................................................................24 5.1 Introduction................................................................................................24 5.2 Dig trajectories ..........................................................................................24 5.3 Hypotheses ...............................................................................................25 5.4 Errors ........................................................................................................29 5.5 Summary ...................................................................................................30 6 Conclusions and recommendations ............................................................31 6.1 Conclusions...............................................................................................31 6.2 Recommendations ....................................................................................31
4
Performance evaluation of a Liebherr 996 hydraulic shovel Acknowledgements ..........................................................................................32 References ........................................................................................................33 Appendix 1 Liebherr 996 data [1]....................................................................34 Appendix 2.1 Configuration simulation script..................................................44 Appendix 2.2 Trajectory simulation script .......................................................46 Appendix 2.3 Sub-scripts used in simulation scripts.......................................48 Appendix 3.1 Cable transducer data [5] .........................................................51 Appendix 3.2 Alternative cylinder length measurement ..................................54 Appendix 4.1 Generation of box plots.............................................................55 Appendix 4.2 Box plot interpretation...............................................................59 Appendix 5.1 Trajectories and Boxplots Stage 1 ............................................60 Appendix 5.2 Trajectories and Boxplots Stage 2 ............................................66 Appendix 6 Matlab scripts for testing hypothesis............................................74 Appendix 6.1 Dig reach vs bench height script...............................................74 Appendix 6.2 Cycle time vs bench height script .............................................75 Appendix 6.3 Cycle times vs digging conditions script ...................................76 Appendix 6.4 Bucket load vs digging conditions script ...................................78
5
Performance evaluation of a Liebherr 996 hydraulic shovel
Glossary Annotate:
Create data that contains information describing the shovel actions; Dig, Swing, Dump, Return, Propel, Clean up, Waiting and Idle.
Bench:
The wall of dirt and rock that the shovel digs away.
Boom:
Link of the shovel arm that is connected to the shovel house by a rotary joint. The main function of the boom is to raise and lower the bucket through actuation of the boom cylinder.
Box plot:
Full name is box and whisker plot. Plot showing a graphical interpretation of the statistical distribution of a population.
Bucket:
The part of the arm that actually is going through the dirt and has storing space for the dirt that has been digged.
Cable transducers:
A string potentiometer used to measure extension of the hydraulic cylinders.
Clean up:
Clean up the big rocks to avoid damage to the tires of the trucks.
Constraints:
Adding constraints reduces the number of degrees of freedom of a mechanic system as the shovel arm.
Cycle time:
Duration of the subsequent dig, swing, dump and return times.
Cylinder:
See Hydraulic Cylinder.
Dig:
Filling bucket with dirt.
Diggability:
See digging conditions
Digging conditions:
Property of the dirt. Good digging conditions are soft and well fragmented. Bad digging conditions are hard dirt with big rocks due to bad fragmentation.
Dirt:
The material that the shovel digs. The properties of the dirt is influenced by the original geology and the blast properties like fragmentation.
Dump:
Emptying the bucket into a truck.
Face:
See Bench
6
Performance evaluation of a Liebherr 996 hydraulic shovel Hydraulic Cylinder:
Actuator for moving the shovel arm links powered by oil pressure.
Idle:
The shovel is idle if it is doing nothing, for example when the operator has a conversation over the 2-way radio.
Kinematic Calibration:
Adjustments made to let the model represent the reality within certain allowable error.
Kinematic Model:
Mathematic model that describes the relation between cylinder lengths and vector angles with respect to the horizon.
Model:
See Kinematic Model.
Numeric Method:
Method that uses the computer for solving mathematical problems.
Propel:
Shovel movement by actuation of the hydraulic track motors.
Return:
Movement of the bucket from the dump position to a dig spot.
Stick:
Link of the shovel arm that is connected to the boom by a rotary joint. The main function of the stick is to extend the arm horizontally through actuation of the stick cylinder.
Swing:
Moving the bucket from the dig spot to the dump spot above the truck.
Trajectories:
The paths that the bucket teeth make through the dirt seen from the site of the shovel (x,y plane)
Waiting:
If there is no truck after a dig action, operator has to wait before he can dump.
7
Performance evaluation of a Liebherr 996 hydraulic shovel
1 Introduction This report summarizes my ‘occupational traineeship’ performed at the University of Queensland with the mining research group CRCMining. CRCMining is one of the Australian Commonwealth Governments Cooperative Research Centres, which bring together university academics and students with Australian industry. CRCMining works with the mining industry to create technologies that help reduce costs and worker hazards [6]. CRCMining has a division at the Department of Mechanical Engineering at the University of Queensland [7]. The project described here is about a performance evaluation of a Liebherr 996 hydraulic shovel at Mount Keith Mine. The Leibherr996 is capable of moving 60 tonnes of material at each dig. It weighs 668 tonnes and has 2 V-16 diesel engines generating 2240 kW (3000 HP). More details are in Appendix 1 Liebherr 996 data. Mt Keith is an open-cut nickel mine located in North West of Kalgoolie in Western Australia. This is remote mine is approximately 1500km from the nearest major city (Perth). As part of the project 15 days were spent at Mt Keith. The aim of the project is to establish correlations between the way in which operators use the machine, the damage they cause, the productivity they achieve, and the digging conditions in which the machine operates. This problem has many factors and the interactions are complex. The contribution to the project was in the planning and participation of field trials to collect data to and the analysis of that data. To collect data a computer (data logger) and many sensors were installed on the shovel. The sensors included pressures sensors monitoring hydraulic pressures in the actuating cylinders, cable sensors to measure the extension of these cylinders. Accelerometers and a rate gyroscope mounted to the machine house. The data logger sampled these sensors at 50 Hz. An additional data stream was provided by a ‘cycle annotation’ process in which an on-board observer recorded the activity of the machine. For example dig, swing, dump, return, propel, idle, and so on. This report summarizes some results from the study. The study is on-going, so final conclusions are not drawn. The structure of the report is as follows. Chapter 2 develops a kinematic model of the shovel that obtains the bucket teeth positions and orientations from the actuating cylinder extensions (forward kinematics) based on multi-dimensional Newton-Raphson solver. Chapter 3 investigates the calibration of this model from experimental data, in particular photogrametric methods. In Chapter 4 the productivity is reported and in Chapter 5 the results of the data analyses are discussed. Finally, some conclusions that could be made out of the limited results are drawn in Chapter 6. Recommendations for future investigation can be found in the chapter as well.
8
Performance evaluation of a Liebherr 996 hydraulic shovel
2 Kinematic model for the Liebherr 996 shovel 2.1 Introduction This chapter gives a formulation of the forward kinematics of the shovel. The purpose of this model is to convert measurements of the cylinder positions into the position and orientation of the bucket. The basic geometry of the shovel is shown in Figure 1. This figure also gives the common names of the links, namely the boom, stick, and bucket. The kinematic model assumes that all bodies excluding the hydraulic cylinders are rigid. The shovel arm is a 3 bar (boom-stick-bucket) linkage and is attached to the shovel housing at pivot point b. The stick pivots about the boom at point f and the bucket pivots about the stick at point g. The motion of the arm is controlled by the positions of the three hydraulic cylinders.
Figure 1 Basic geometry of a Liebherr 996
9
Performance evaluation of a Liebherr 996 hydraulic shovel
2.2 Kinematic constraints To get the bucket position and orientation, a set of constraint equations need to be solved. These equations are based on the three constraint loops shown in Figure 2. [3]
Figure 2 Constraint loops for calculation of the bucket position and orientation.
Horizontal set:
Loop 1: l1 ⋅ cos ( t1 ) + l2 ⋅ cos ( t2 ) + l3 ⋅ cos ( t3 ) = 0 Loop 2: l4 ⋅ cos ( t4 ) + l5 ⋅ cos ( t5 ) + l6 ⋅ cos ( t6 ) = 0 Loop 3: l7 cos ( t7 ) + l8 ⋅ cos ( t8 ) + l9 ⋅ cos ( t9 ) + l10 ⋅ cos ( t10 ) = 0
Vertical set:
Loop 1: l1 ⋅ sin ( t1 ) + l2 ⋅ sin ( t2 ) + l3 ⋅ sin ( t3 ) = 0 Loop 2: l4 ⋅ sin ( t4 ) + l5 ⋅ sin ( t5 ) + l6 ⋅ sin ( t6 ) = 0 Loop 3: l7 sin ( t7 ) + l8 ⋅ sin ( t8 ) + l9 ⋅ sin ( t9 ) + l10 ⋅ sin ( t10 ) = 0
10
Q (l,t)
Performance evaluation of a Liebherr 996 hydraulic shovel To solve Q (l,t) the following arrays are introduced:
lv = [l3
l4
lc = [l1 l2 tv = [ t 2
l9 ] ' l5
l6
l7
l8
t3 t4
t6
t8
t9 ] '
tc = [t1 t5
t7
l10 ] '
t10 ] '
Where the subscript v refers to variable lengths and angles and the subscript c refers to constant lengths and angles. The hydraulic cylinder lengths lv are the inputs to the kinematic model, which will be used to solve for the variable angles, tv. If t has an index x (tx) than this angle represents the angle between vector lx and the horizon where counter clockwise is the positive direction. The angles in tc are constant or can be calculated out of angles that already are determined. For instance, t5 can be computed out of t2, which is already a variable. This is because vector l2 and l5 are in the same rigid body and therefore dependent. The same for t10 with t5 and t6 with t7. See Equ. 1, 2 and 3
t5 = t2 + c2,5
(1)
t7 = t6 + c6,7
(2)
t10 = t5 + c5,10
(3)
cx,y represents the constant angle between vector x and y. In Q (l,t) are 6 equations with 6 unknowns and therefore the equations set is solvable. The outcome of this problem will only provide the six angles in tv. The actual bucket tip info is derived from these angles and constant lengths. In Chapter 3 the estimation of the required lengths is discussed.
11
Performance evaluation of a Liebherr 996 hydraulic shovel
2.3 Newton-Raphson To compute the variable angles from the hydraulic cylinder extensions, the equations set Q(l,t) is solved using the Newton-Rahpson method for non-linear systems. [3] This method is used for its simplicity and minimal computational requirements. Taylor series expansion is used to formulate the perturbations (∆l, ∆t) about a certain configuration (l, t) of the shovel arm. Here ∆l, ∆t, l and t are arrays. λ
m 1⎛ m ∂ ∂ ⎞ Q(l + ∆l , t + ∆t ) = ∑ ⎜⎜ ∑ ∆li + ∑ ∆ti ⎟⎟ Q(l , t ) = 0 ∂li i =1 ∂ti ⎠ λ = 0 λ! ⎝ i =1 ∞
(4)
Where m is the number of vectors that are used in the kinematic tracking model, for the kinematic model of the shovel m equals 10. When only the first order terms are considered (λ=0 and λ=1) and the arrays lv, lc, tc and tv are substituted, Equ. 4 can be written in a matrix form as follows. Q(l + ∆l , t + ∆t ) = Q(l , t ) +
∂Q ∂Q ∂Q ∂Q ∆lc + ∆lv + ∆tc + ∆tv = 0 ∂lc ∂lv ∂tc ∂tv
(5)
Where:
⎡ − l2 s 2 ⎢lc ⎢ 2 2 ∂Q ⎢ 0 =⎢ ∂tv ⎢ 0 ⎢ 0 ⎢ ⎣⎢ 0
− l3 s3
0
0
0
l3c3 0
0 − l4 s 4
0 − l6 s6
0 0
0
l4c4
l6c6
0
0
0
0
− l8 s8
0
0
0
l8c8
⎡c1 c2 ⎢s s ⎢1 2 ∂Q ⎢ 0 0 =⎢ ∂lc ⎢ 0 0 ⎢0 0 ⎢ ⎣⎢ 0 0
0
0
0
0
0 c5
0 c6
0 0
0 0
s5
s6
0
0
0
0
c7
c8
0
0
s7
s8
0⎤ 0 ⎥⎥ 0⎥ ⎥ 0⎥ c10 ⎥ ⎥ s10 ⎦⎥
0 ⎤ 0 ⎥⎥ 0 ⎥ ⎥ 0 ⎥ − l9 s9 ⎥ ⎥ l9c9 ⎦⎥
⎡c3 ⎢s ⎢ 3 ∂Q ⎢ 0 =⎢ ∂lv ⎢ 0 ⎢0 ⎢ ⎣⎢ 0
0 ⎡− l1s1 ⎢ lc 0 ⎢ 11 − l5 s5 ∂Q ⎢ 0 =⎢ l5c5 ∂tc ⎢ 0 ⎢ 0 0 ⎢ 0 ⎣⎢ 0
0 0 c4 s4 0 0
0 0 0 0 − l7 s7 l7 c7
0⎤ 0 ⎥⎥ 0⎥ ⎥ 0⎥ c9 ⎥ ⎥ s9 ⎦⎥ ⎤ 0 ⎥⎥ 0 ⎥ ⎥ 0 ⎥ − l10 s10 ⎥ ⎥ l10c10 ⎦⎥ 0
Where si = sin(ti) and ci = cos(ti) The lengths in array lc are constant at all times i.e. ∆lc = 0. In the NewtonRahpson loop tc is assumed constant i.e. ∆tc = 0. These angles are updated after
12
Performance evaluation of a Liebherr 996 hydraulic shovel each iteration using the current solution for tv. Since lv varies in a different loop than the Newton-Rahpson loop itself, it is assumed constant as well i.e. ∆lv = 0. Eq (5) becomes,
Q(l , t ) +
∂Q ∆tv = 0 ∂tv
(6)
Rearranged to determine ∆tv we have, −1
⎡ ∂Q ⎤ ∆tv = − ⎢ ⎥ ⋅ Q(l , t ) ⎣ ∂tv ⎦
(7)
Algorithm 1 shows the implementation for the Newton-Rahpson solver. The for loop generates tv-column from every lv-column. The Newton-Rahpson calculation is done in the while loop. Algorithm 1: For k = 1 till number of data points in lv array lv,k = lv measurement data matrix row k; i = 1; F,i = ones(6, 1) While (maximum F,i > Minimum error) Update tc,i-1 out of tv,i-1; F,i-1 = F(lc, lv, tc,i-1, tv,i-1); Inverse Jacobian = Inverse Jacobian of F,i-1 tv,i = tv,i-1 – Inverse Jacobian * F,i-1 i=i+1 end TV = [TV tv]; end
This algorithm creates a matrix TV where every column represents one arm configuration. An initial guess is needed for all arrays. lc is constant and is loaded before the algorithm starts. The lv guess is the first column of the lv array. This array is generated before from the measurement data using Equ.8 in Chapter 3.3. The initial guess for tc and tv are derived by solving the initial constraint equations using the MATLAB command solve. (Appendix 2.3 Sub-scripts used in simulation scripts) Each constraint loop has 2 solutions. To determine the correct solution a logic rule has to be applied according to the constraints of the arms motion. When the simulation is running, the tv from the previous configuration is used as an initial guess for calculating tv for the current configuration. Because the difference between these configurations is always little, the Newton-Raphson
13
Performance evaluation of a Liebherr 996 hydraulic shovel method does not need many iteration steps to find a solution within the error bound. Algorithm 1 is used in 2 simulations. One is for displaying the configuration of the shovel arm and bucket used for calibration of the model parameters. This calibration is discussed in Chapter 3. The other simulation plots the bucket teeth trajectories and will be used in Chapter 5. These two simulation scripts can be found in Appendix 2.1 Configuration simulation script’ and Appendix 2.2 Trajectory simulation script’
14
Performance evaluation of a Liebherr 996 hydraulic shovel
3 Parameter estimation for kinematic model 3.1 Introduction The purpose of this chapter is to give details of the calibration of the kinematic model of chapter 2. To analyze the movement of the bucket, it is necessary to know the constant lengths and angles. They were not provided by the manufacturer so it was necessary to determine them experimentally. In Section 3.2 the photogrametric method for estimating lengths is discussed. Then in 3.3 the measurement and the conversion method of the cylinder lengths is explained. The calibration of the parameters is shown in 3.4.
3.2 Photogrametric method for estimating lengths Photos are taken of eight configurations of the shovel from as far away as possible to minimize perspective effects. An example of such an image is shown in Figure 3. This figure illustrates that it is sometimes hard to see where the center of the joints are, especially the joint that connects the stick cylinder to the stick. AutoCAD is used to get the lengths and angles. Scaling the AutoCAD lengths are converted to meters using the track as a reference. Since there are eight photos taken for the calibration procedure, an average is taken for the non-cylinder lengths. Moreover, they supposed to be constant in all pictures.
Figure 3 Bucket-In configuration photo used for estimating lengths.
15
Performance evaluation of a Liebherr 996 hydraulic shovel Table 1 gives the values of the shovel model parameters determined by this experimental method. All lengths are in meters. Boom
Stick
Bucket
Photo: Cylinder Cylinder Cylinder BOOMOUT 7.94 5.02 5.31 STICKOUT 5.57 5.26 5.38 STICKIN 5.88 3.19 4.93 BUCKETOUT 5.91 5.2 6.3 BUCKETIN 6.53 3.9 4 BOOMIN 4.84 5.26 6.34 Average
l1
l2
l5
l6
L7
l8
l10
3.16 3.10 3.24 3.12 3.13 3.22 3.16
5.49 5.86 5.84 5.83 5.73 5.95 5.78
4.57 4.64 4.64 4.63 4.63 4.64 4.63
1.60 1.58 1.65 1.54 1.65 1.58 1.60
4.62 4.74 4.69 4.72 4.69 4.76 4.70
1.09 1.11 1.11 1.09 1.10 1.10 1.10
0.97 1.03 1.00 1.00 1.00 1.00 1.00
Table 1 Values of the shovel model parameters.
16
Performance evaluation of a Liebherr 996 hydraulic shovel
3.3 Measurement of the hydraulic cylinder lengths. To find out the total lengths of the cylinders in meters the cylinder extensions are measured by cable transducers. The signal of the cable transducers is logged in Volts. Therefore a conversion equation is needed. The current through the transducers changes linearly when the cylinders extend. The voltage over the transducers must be continuously 24 V. Since the data logger is only able to measure voltages, the current is led trough a constant resistance (R). The voltage over this resistance is quantified and logged in the data logger. To derive the actual total cylinder length the following equation is used (Equ. 8).
Li =
V − Vmax,i R⋅S
+ Lmax,i
(8)
Where = Total length of cyclinder i in meters Li = Current for cylinder i in Ampere Ii Vi = Voltage over R for cylinder i in Volts R = Resistance in Ohm S = Sensitivity in Ampere per meter Lmax,i = Maximum length cylinder i in meters Vmax,i = Maximum voltage for cylinder i in Volts The attachment of the sensor onto the hydraulic cylinder is depicted in Figure 4
Cable extension ~ Vmax Cable Transducer
Lmax Figure 4 Cable transducer attachment.
The cable transducer that is used is a Celesco PT9420 (see Figure 5 and Appendix 3.1). The range is 5.08 m (200 inch) and the current changes 16 mA. The sensitivity is then 0.00315 A/m. The constant resistance that is used was 470 ohm. When the sensitivity and the resistance are unknown an alternative method can be used. This is described in Appendix 3.2 Alternative cylinder length measurement.
17
Figure 5 Celesco PT9420
Performance evaluation of a Liebherr 996 hydraulic shovel
3.4 Kinematic Calibration Since there are still uncertainties in the parameter estimation a calibration of the kinematic model is needed. This is done by comparing configuration photos with configuration plots from the model. The most uncertain parameters are adjusted such that the plot configuration matches the photo configuration for al eight configurations. One of those eight comparisons is displayed in Figure 6. A synchronous video is generated as well and can be found on the CD-rom (SynchroneV2.m2v).
Figure 6 Comparison of model configuration with photo configuration
After this calibration the final model uses the parameters given in Table 2.
l1 l2 l5 l6 l7 l8 l10
= = = = = = =
3.15 5.78 4.9 1.7 5 1.15 1.05
Arm: m c2,5 m c6,7 m c5,10 m m L1 m L2 m L3
= = =
5 deg 6 deg 23 deg
= = =
7.6 m 5.0 m 4.2 m
S R Lmax,Boom Vmax,Boom Lmax,Stick Vmax,Stick Lmax,Bucket Vmax,Bucket
18
= = = = = = = =
Hydraulic Cylinders: 0.00315 470 8.3 3.577 5.6 2.314 6.7 2.871
A/m Ohm m V m V m V
Table 2 Parameters used in the final kinematic model
Performance evaluation of a Liebherr 996 hydraulic shovel
4 Productivity analysis 4.1 Introduction The purpose of this chapter is to report on productivity of the shovel based on 43 annotation sessions. These were logged over two 2-week periods for 3 to 5 hours each day.
4.2 Summary of data collected Table 3 in section 4.5 summarizes each of the logging periods. Stage one is displayed in Table 3.1 and stage two in Table 3.2.
4.3 explanation summary table Every row represents an annotation session and every column session information. The blocks column is representing a unique code for the blast area that is dug in that session and contains information about the dig material and dig location. Numbers that are shown in the Riley’s Diggability index column are Diggability numbers. Tim Riley, a continuous improvement engineer with 30 years experience in mining and quarrying, judged the ease of digging daily and rated them on a scale from 1 to 5, where 1 stands for easy digging and 5 for very difficult. The rilling column is an index of how well the material is rolling down. The average bucket loads are calculated each annotation session using the annotation videos and the recorded truckloads. In the videos is visible how many dumps it takes the operator to fill the truck. The truckload divided by this number gives the average bucket load. In the notes column are Paul’s (CRCMining employee) remarks he made after every session. Interesting factors to compare are, diggability, face height, average bucket load and mean digging time. This will be discussed in Chapter 5. The cycle time is the duration of a dig, swing, dump and return cycle together. To see the difference between dig, swing, dump and return times and their variances graphically box and whisker plots (box plots) are made for each annotation session. These plots are used for comparison in Chapter 5. The method that is used for generating the box plots and for calculating the cycle times can be found in Appendix 4.1 Generation of box plots.
4.4 Overall The operators filled a truck with 228.5 tonnes dirt. They needed 5.8 passes for that and every cycle took 35.4 sec. They loaded a truck in 3 min and 23 sec. The mean face height was 14.8 m. The average bucket load was 40.5 tonnes and filled in 24.3 sec. The maximum average bucket load was 50.0 tonnes in very easy digging conditions and a bench height of 12–14 m. The minimum average bucket load was 23.6 tonnes when the rilling was poor to good and the bench height was 20 m. The shovel was digging in the edge procedure zone i.e. more dangerous zone than normal.
19
Performance evaluation of a Liebherr 996 hydraulic shovel
Date
Tue Nov 9
Graham
Thu Nov 11 Johnny
Fri Nov 12
Paul Paul
Sat Nov 13 Paul
Clive
Sun Nov 14 Paul Paul
Mon Nov 15 Mutt Mutt
Tue Nov 16 Tinky Mutt
Pimmy and Thu Nov 18 Jimmy
Pimmy and Jimmy
Fri Nov 19
hours of annotation
Operator
Pimmy and Jimmy
Afternoon session 1 hr 45 min
Morning Session
shot number
20
blocks
Riley's digability face height index
NON F-394-20-342 18m
rilling
good to 3 very good
number of average average number of trucks payloads passes average bucket loaded recorded per truck truck load load
21
0
6,2 n/a
n/a
mean dig time
11,1
mean dig mean mean time mean swing time mean dump time variance swing time variance dump time variance
14
7,5
4,6
4,1
2,1
mean return time variance
mean return time
8
7,9
mean mean cycle time cycle time variance
30,7
video
16,8 from cab
1 hr 40min
45
NON F-424-45-311 20m
4 poor
20
0
5,6 n/a
n/a
15,9
27,7
7,1
4,8
4,5
2,6
7,4
9,2
34,4
30,4 from cab
Afternoon Session 1 hr 30 min
45
NON F-424-45-311 20m
4 poor
19
0
5,52 n/a
n/a
18,4
34,6
7,6
6,4
4
2,5
8,5
18,3
38,5
71,1 from cab
Morning Session 1 hr 45 min Afternoon Session 1 hr 50 min
Morning Session
1 hr 30 mins
Afternoon Session 1 hr 45 mins
Morning Session 1 hr 45 min Afternoon Session 1 hr 40 min
Morning Session Afternoon Session 1 hr 25 min
Morning Session 1 hr 40 min Afternoon Session 1 hr 5 min
Morning Session
40
NON F-439-40-276 10m
3 good
19
14
6,42
228
35,2
14,8
22,9
6,3
2,7
3,9
2,2
9,2
10,4
34,1
40
NON F-439-40-276 10m
3 good
18
14
6,22
223
36,4
16,2
27,9
6,3
3,7
4,3
1,5
8,1
13,9
35,3
67
67
NON F-469-67-210 10-12m
NON F-469-67-210 12-14m
1 very good
1 very good
23
25
16
22
5,09
4,88
235
242
45,8
50
12,5
16,8
14,6
28,3
7,5
6,8
4,3
6
4,3
4
2
1,2
9,7
6,6
9,6
5,7
33,7
NON F-469-67-210 12-14m
1 good
22
18
5,27
231
46,3
15
15,1
7,3
4,9
4,1
1,5
10
13,3
36
1 good
19
15
6,47
228
35,6
17,8
34,7
5,9
3,2
4
1,3
7,4
5,9
35,9
1 good
16
11
6,25
231
37,2
NON F-469-67-210 12-18m
1 good
20
12
6,2
220
35
15,6
23,6
6,3
4,6
4,2
1,4
6,4
4,1
32
41,2 from pit edge This annotation set was done by Tom in the morning of Monday the 15th of November. Unfortunately tom could not get the traintest software working so annotation will need to be done at a later time with the video tape that was recorded . Digging is in the same bench as yesturday, although Tom noticed changing digging conditions accross the face. To the left the face height was lower and the material much softer to dig. The the middle of the face the hard from cab and rocky with a much higher face (maybe 18m). Operating conditions were much the same as the morning annotation session done eariler that 22,5 from pit edge day. The digging location was the same as the previous day. The material was relitively soft and rilled well. Tinky set himself in good positions and was able to fill trucks within five pass with consistant 50 ton bucket fills. He did not slam the clam shut and hit the tracks or the stops.
NON F-469-67-210 12-18m
1 very good
25
13
4,44
239
49,9
18,1
39,4
7,4
5,1
4
2
8,8
17
38,4
57,8 from cab
67
NON F-469-67-210 12-18m
1 very good
14
12
5,29
246
47,3
14,7
39,5
5,9
5
4
1
6,9
8,5
32,8
41,5 from pit edge The session was finishing of the last of the bench that had been dug the last few days. Pimmy (the trainer) claimed he was operating all day with Jimmy in the hot seat. The shovel was digging in the drop cut that was started on Thursday last week. It was difficult to tell the digging conditions form afar but it did appear that they had improved from the previous week. The face height was variable and the material appeared to rill well. Some bucket fills were very low maybe indicating that the digging is not as soft as appeared. The start of dig height appeared to vary 41 from pit edge significantly. This annotation session was conducted from the pit rim on The afternoon of Thu the 18 of November. Again Pimmy was training Jimmy so we should see a large variation in dig styles through the session. Noticable differences in Bucket fills and truck payloads was noticed during the annotation. Digging Conditions looked hard. Large rocks were observed falling from the 45,1 from pit edge face. The face stood up and did not naturally rill after each each dig.
45
NON F-424-45-311 20m
4 good/poor
22
17
5,36
231
42,5
17,6
29,1
6,2
4,4
3,9
1,6
10,2
25,3
37,2
Afternoon Session 1 hr 30 min
45
NON F-424-45-311 20m
4 good/poor
20
19
5,6
235
42,7
17,1
23,4
5,8
5,9
4
1,5
8,8
21
34,8
1 hr 45 min
34,8 from cab
We were in the same pooftah dirt as yestdurday although it had become more rocky in parts. The face was higher and stood-up for longer. Some mid face digging actions were required to pull material down. Good bucket fill factors were still obtained. Photos IMG_1218 to 1239.jpg are pictures taken from the cab through the morning. Photos IMG_1241-1244.jpg show the approximate digging position and the bench height. Approximate digging location: 31900N, 9550E
67
1 hr 45 min
Morning Session
The digging conditions were excellent. The material was very soft and rilled extremely well from the face. The face height was approximately 10-12m. Very high bucket fill rates were achieved. Digging Location: 32025N, 9575E. Photos IMG_1191-1197.jpg show the digging conditions taken through the morning from the cab. Photos IMG_1198-1200.jpg show the digging location from the pit edge. Digging was done to the left of the diggers position shown in the photos.
As in the morning the dirt was very soft (poofter dirt) and rilled extremely well. The face height crepped a little higher (maybe 12 - 14m). Production rates were extremely high with most trucks 30,3 from pit edge having loads greater than 230 ton. IMG_1211-13.jpg show the digging position.
NON F-469-67-210 12-14m
NON F-469-67-210 12-18m
The muck pile was well fragmented and the face rilled well. The maximum face height was not more than 10m. The beginner operator was not smooth. There were some digs that were very rough on the machine and should show up on the acceleration data. He also occasionally had 30,2 from cab difficulty with filling the bucket. Digging position: 32500N, 9750E. Tom and I annotation this data from the lookout on the pit rim. The digging location and the 42,3 from pit edge operator were the same as those I annotated in the morning
24,6 from cab
67
67
We were digging in hard transitional rock with bench heights of approximately 20m. The digging was very difficult with lots of large rocks and a face that would not rill naturally. The operator was constantly checking the face for rock falls while swinging and dumping. IMG_1152.jpg to IMG_1159.jpg in directory November_11_2004_anno are taken from the cab after the annotation session. IMG_1160.jpg to IMG_1163.jpg were taken at the lunch break that day and show the recess where the shovel had been digging. IMG_1164.jpg to IMG_1171.jpg show the condition of the sensors taken at the same lunch-time. The shovel spent the remainder of that day digging in the same area. After the laptop battery went flat, I continued to annotate using the digital voice recorder (files A_001_007,8,9). The data for the recording is the data directory. Digging Location: 32425N, 9825E. This annotation set was tom's first session in the excavator and was done in the afternoon of Thu the 11 of November 2004 Mt Keith time. The operator and digging conditions are the same as those experienced during my annotation session that morning. Tom is not confident that his annotation data is well representitative of the machines actual state.
34,2
67
67
notes After cleaing up some rill under the one-way section haul road we walked to a oar bench. The bench was approximately 18m high and was made up of Millarite Oar. This is a hard and brittle oar with glass sounding qualities and little fines. Towards the end of the annotation period we moved into Pendlandrite oar. This is considered the best digging oar material at the mine. This oar is softer and heavier then Millarite and is much easier to Photos IMG_1128 to IMG_1131.jpg show the Millarite oar dug. Photos IMG_1132, IMG_1138 to IMG_1141.jpg show the Pendlandrite oar. Digging Positions: 32250N, 9900E followed by 32500N, 10000E.
45/49
NON F-424-45-311 NON F-424-49-319 20m
4 good/poor
16
12
5,88
238
42,3
21,8
42,3
6,7
6,6
4,4
2,6
11
25,7
43
The digging was again very variable, some very compact and hard dirt in the middle of the cut where the face was the highest. Here rilling had to be iniciated by taking digs half way up the face. To the left (closer to the windrow) the face was lower and digging easier. To the right of the hard stuff the digger also became easier. Here the face was still high but the material softer 85 from pit edge and consisting of more fines.
Pimmy
Sat Nov 20 Kazzy
Kazzy STAGE 1 average:
Afternoon Session 2 hrs
Morning Session
45/49
NON F-424-45-311 NON F-424-49-319 20m
4 good/poor
1 hr 40 min
45/49
NON F-424-45-311 NON F-424-49-319 15-20m
4 good/poor
Afternoon Session 1 hr 10 min
45/49
NON F-424-45-311 NON F-424-49-319 15-20m
4 good/poor
8 different operators Table 3.1 Summary data sheet Stage 1
15.9m
25
21
16
18
5,08
242
47,8
15,7
24,4
7
6,9
4
2,5
9,3
26,8
35,5
6,33
226
35,9
12,9
19,7
7,3
4,6
4,2
2,2
11
19,5
35,3
5,62
219
41,5
14,3
24,6
6,4
4,9
4,3
2
10,9
28,4
35,1
5,67
232,13
41,96
15,91
26,99
6,74
4,92
4,12
1,87
8,79
15,03
35,38
41,9 from cab
The digging was again very variable, some very compact and hard dirt in the middle of the cut where the face was the highest. Here rilling had to be iniciated by taking digs half way up the face. To the left (closer to the windrow) the face was lower and digging easier. To the right of the hard stuff the digger also became easier. Here the face was still high but the material softer and consisting of more fines.
The digging conditions were highly variable. To the left of the cut (near the windrow) the face height was at it lowest and digging was moderately easy (IMG_1333.JPG). Further to the right the digging became very hard (IMG_1334.JPG). The face height was very high 15m bench plus 4 to 5 m of heave. At the base of the face the material was very tight (as if the blast had not separated it). Above the face did not rill well and digging half way up the face was needed to bring it down, at which large boulders would drop. Further to the right the digging became easier. The face was ust as height but the digging was much more softer and fragmented. The material would rill more naturaly and when it was brought down by the digger it large boulder would shatter. The colour of this stuff was black with lots of green powerdery aspestos stuff. 40,3 from cab This annotation session was done from the pit wall in the afternoon of Saturday the 20th of November. The digging location and operator was the same as the morning. The session was interupted by a emergency rock fall and then maintenance on the shovel. The sensor data was 49,5 from pit edge lost but we have annotated data from the laptop. 41,46
Performance evaluation of a Liebherr 996 hydraulic shovel
Date
Thu Dec 2
Fri Dec 3
hours of annotation
Operator
Troy
Troy
session 1
2 hrs 10 min
8
session 2
1 hr 50 min
8
session 1
session 2
Sat Dec 4
Sun Dec 5
Troy
Troy
session 1 session 2
session 1
session 2
Mon Dec 6
Troy
session 1
session 2
Tue Dec 7
Troy
shot number
2 hrs 30 min
1 hr 35 min
2 hr 35 min 1 hr 30 min
2 hrs 30 min
1 hr 30 min
1 hr 45 min
45 min
blocks
LGO F-379-08-335 SGO F-379-08-337 SGO F-379-08-337 LGO F-379-08-336
SSG F-379-12-229 TAO F-379-13-338 LSO F-379-13-339 12 and 13 NON F-379-13-340
NON F-379-13-340 LSO F-379-13-339 12 and 13 TAO F-379-13-338
37 37
12
12
44
44
NON F-409-37-320 NON F-409-37-320
LGO F-379-12-334
LSO F-379-12-330
HSO F-424-44-313 SSG F-424-44-314
face height
Riley's digability index
rilling
number of trucks loaded
number of average average payloads passes per average bucket recorded truck truck load load
mean dig time
mean dig mean mean mean mean time mean swing time mean dump time mean return time mean cycle time variance swing time variance dump time variance return time variance cycle time variance
video
15m
3 very good
42
28
5,02
232
46,1
13,6
17,4
6,8
4,6
3,9
2,2
7,6
16,6
31,3
33,6 from cab
15m
3 very good
29
17
5,07
232
46,5
14,4
15,7
6,9
5,9
3,7
1,6
8,1
19
32,3
35,9 from cab
15m
3 very good
15m
n/a n/a
3 very good
5* 5*
15 m
3 very poor
15 m
10 - 13m
good good
3 very good
3*
HSO - very good SSG - poor
25
27 17
26
33
25
28
14
16 11
20
23
15
5,25
5,24
6,46 6,44
6,07
5,63
5,76
233
227
236 219
230
230
228,5
44,3
43,6
36,3 34,7
37,9
41,6
40,1
14,6
14
16,5 16,1
15,7
14,9
18
17,2
13,7
26,9 24,7
22,8
18,2
31,2
7,3
7
7,8 7,8
7,5
7,9
6,8
4,4
3,5
4,4 3,8
5,7
5,4
3,5
4,1
3,7
4,4 4,1
4,1
4
3,8
2
1,5
1,6 2,2
1,9
1,6
1,2
10,7
9,7
10,6 11
10,3
10,2
9,6
23,1
20,5
27,3 26
31,7
27,1
25,9
36,4
34,4
38,9 37,7
37,7
37
38,5
NON F-424-47-316 LGO F-424-47-315
10 - 13m
NON - poor LGO - good
27
22
5,5
225
41,1
18
24,5
6,3
2,8
3,6
2,1
9,4
24,75
37
poor
24
20
6,2
220
35,8
18,2
39,5
6,5
7,5
3,8
2,2
9,2
28,6
38,2
3*
session 1
1 hr 45 min
47
LGO F-424-47-315 NON F-424-47-316 LGO F 424 47 315
10m (peeking to 13m towards the far left of the 10mbench) (peeking to 3*
session 2
2 hrs
47
SSG F-424-47-318
13m towards
3*
NON - poor LGO - good
36
13
9
5,71
230
41,4
16,1
17,5
7,2
4,9
3,7
1
9,2
17,6
35,7
notes The material rilled very well giving the face a triangular shape. Fragmentation for the material varied from very fine green dust to large boulders. The boulders, however, were easily smashed into smaller pieces. Digging became a little harder to the west where fragmentation was a little larger with less fines. Here the bench held up a little more prompting the ocassional dig half way up the face.
40 from cab
Digging was very good. The material was well fragmentated although it did have a range of rock sizes. During the first annotation session the face was made up of larger boulders and sloped at about 30 deg (see photos IMG_1420-26.jpg) This was a result of the trim shot and required a lot of cleaning up and dig preperation.
36,8 from cab
Once this material was dug out the material underneath had more fines mixed in with the fragmented rock. Photos IMG_1427-44.jpg shows the dig conditions once the low rock had been dug out. The face height was not more than 15 meters and the material rilled well. The ocassional large boulder was easily broken up. During the second annotation session the boom was place all the way up (a good reference to calibrate the slightly damaged boom cable transducer)
49,3 from cab 37,3 from cab
56 from cab
Today we were still digging out the one-way section covered by the blast yesterday. The aim of the digging was to clear rill (from the blast) from the one-way section and to widen the haul road to two way. The dig material is transitional rock which is very hard and abrasive. The material was reasonbly fragmented into largish rocks, but was still well compacted making it difficult to penetrate the bucket into the face. Some points were extremely hard and may have been the birm underneath the blast (the haul road sloped down from the blast). Face hieght wasnt realy an issue becase the face was sloping over the haul road. The maximum hight of the face was maybe 10m. Most digs were started low with the ocassional dig high. Although the face did not rill well because the low face height it wasnt really a problem. There were many disruptions; the truck were loaded on the off-side and because of this some truckies had difficulty backing up, trafic was high on the one-way having to pass behind the shovel, and the tight space ment the shovel had to walk a lot.
Free facing trim shot 15, digging LGO F-379-12-334. The dig material was a combination of rocky (large in size) and unblasted rock. Digging the unblasted rock was extremely tough going with each requiring a number of attemps to get a reasonable bucket fill. The vibration on the house during these digs were extreme.
42,6 from cab
Moved further south in E stage to LSO F-379-12-330. This was good digging. The material was a combination of fines and rock and was well fragmented. Troy was happy to get the shovel very close to the face to start a dig. The material rilled naturally and he would move back if the rill looked like it was to cover the tracks. Face height was 15m.
42,9 from cab
Boom cable transducer damaged. The HSO was the green powdery stuff seen in IMG_1510.jpg. Very easy to dig, rilled well and good bucket fill factors The SSG was much harder, rocky and and compact. The digger found it hard to penetrate the muck pile and bucket fill factors were poor. The material didn't rill well and Troy wasn't phase to get the shovel up on its back tracks close to the face bringing material down.
46,1 from cab
The NON much like the SSG above. Very tight digging with poor bucket fills. Close to the floor the digging was particularly tight with the bucket shudering badly while trying to penetrate. The LGO offer better digging (not as easy as the HSO). Good fragmentation and rilling. Fragmentation size was a little larger then in HSO.
55,8 from cab
LGO was good digging, the material was well fragmented and rilled well, reasonable bucket fills could be obtained. diggin NON was difficult going but according to Troy not a difficult as yesturday. Bucket fills were generally poor because of difficultly penetrating the muck pile. Rilling was better then yesturday.
54,9 from LV
poor annotation session, conducted with Dave Medland
Wed Dec 8 Charlie
Thu Dec 9
Fri Dec 10
session 1
1 hr 45 min
47
session 2
1 hr 10 min
47
3*
very good
28
24
5,6
233
43,9
13,7
23,3
8,1
5
3,9
1,4
9,7
21,9
34,5
36,3 from cab
3*
good to poor
17
16
5,3
226
43,8
15,3
41
8,1
8
5
8,2
9,2
18
37,5
65,1 from LV
poor annotation session, conducted with Dave Medland. Digging condition were much the same as above.
No Annotation
Charlie
Sat Dec 11 Charlie
Sun Dec 12 Graham
session 1
2 hrs 20 min
99
NON F-439-99-279
4 - 20m
5 good
19
18
5,6
215
39,1
14,3
13,1
8,8
4,2
4,8
1,5
12
38,9
40,8
61,1 from cab
session 2
3 hrs
99
NON F-439-99-279
4 - 20m
5 good
21
19
6,14
221
36,6
14,6
26,3
8,8
4,2
5,2
2,2
10,9
23,5
39,8
63,1 from cab
Continually lost connection to data logger during this session. The digging at shot 99 was improved from yesturday. The face was higher, increasing from about 4 meters to the left to about 20 meters on the right. The material was mostly fine dirt with occasion large and very large boulders some of which needed to be broken up. Except when breaking rock the digging was smooth. Annotation was a mixed bag today. Two annotation sessions were conducted but during them I repeatedly lost the network connection. I don't believe this was caused by the data logger locking-up because it appeared the data was still collected. On inspection a lunch I found that the cross-over cable in the data logger box was a little loose.
Morning digging today was excellent. The NON material was well fragmented, very easy to penetrate and rilled very well. We dug along the face. to the right the face height was approximately 20m, to the left it dropped down to ground level. The digging here was frequently interupted by clean-up sessions by the wheelie dozer.
session 1
1 hr 30 min
session 2
1 hr 45 min
session 1
session 2
Mon Dec 13 Graham
session 1
session 2
Tue Dec 14 Graham STAGE 2 average:
SSG F-424-47-318 SSG F-424-47-318 LGO F-424-47-315 NON F-424-47-316
10m (reaching to 20 at the far right of the bench) 10m (reaching to 20 at the far right of the
The majority of the both session was spent in the SSG (to the far right) which offered the best digging. The SSG consisted of a conbination of green powder and small to mediam rocks. The face height was highest at the far right (reaching up to 20m). Here large boulders hung and threated to fall onto the shovel. Further to the left the face height reduce (to about 10m). Here rilling was good. The LGO was a thin band in the middle of the length of the face and offered the hardest digging. Only 1-2 were dug from here. The NON was at the far left and offered digging conditions somewhere inbetween the LGO and SSG. Here the face was again higher and rocks larger but fragmentation was good. Digging was done here at the end of the second session. The shovel continued to dig here up until lunch.
session 1 session 2
33 min
1 hr
2 hr 10 min
1 hr 50 min
2hr 1 hr 55min
3 different opreators Table 3.2 Summary data sheet Stage 2
NON F-424-53-322 52 and 53 NON F-424-52-323 NON F-424-53-322 52 and 53 NON F-424-52-323
53
47
53
NON F-424-53-322
4 to 3
very good
22
17
5,85
222
39
11,5
11,9
8,3
3,8
4
1,4
10,7
23
34,5
35,5 from cab
20m
4 to 3
very good
25
19
5,28
212
39,9
11,4
20,4
7,8
3,5
3,7
1,4
10,8
20,6
33,4
28,5 from cab
20m
SGO F-424-47-317 EPZ - 20m
NON F-424-53-322
LSO F-379-16-354 LSO F-379-15-353 SSG F-379-15-352 LSO F-379-15-351 15 and 16 LGO F-379-15-350
15 15
20m
LSO F-379-15-351 SSG F-379-15-352
3 good/poor
3*
good/poor
8
9
1
3
6,19
8,25
220
213
35,5
23,6
11,3
15,2
12,4
36,6
6,7
6,4
2,3
2,2
3,1
3,8
0,7
1,8
9,2
10,3
7,7
17,8
31
36,4
17,5 from cab
In the morning we started at NON F-424-53-322 (shown in the images IMG_1583 to IMG_1586.jpg). This was the same location I dug with charlie yesturday although today the digging was a lot harder. Unfortunately because problems with the data logger I didnt get a full annotation set with truck payloads but I did take 1.5hours of video.
62,2 no video
After smoko we tramed to the EPZ at SGO F-424-47-317. Here we loaded some large bolders into trucks. Digging was also tough here until we made it into the green powdery stuff at the far right of the EPZ here we had some full buckets for a change but only 2 trucks were loaded in this stuff. Unfortunately I did not managed to take any video or truck payload data for this session.
20m
3 poor
27
12
7,78
224
32,1
13,4
25,1
6,2
2,5
3,4
1,1
9,1
15,5
31,9
34 from cab
sloping away from 4 to 15m
3 good
25
15
6,64
225
34,1
11,75
16
6,7
2,7
3,6
1,4
10,8
20,2
32,6
33,3 from cab
15m
13.7m
3 very good 3 very good
28 28
13 14
5,52 5,67
221 223
40,3 40
11,3 11,4
11,6 10,1
7,2 6,4
4,7 3,3
3,3 3,4
1,3 1,3
9,3 9,6
16,3 16,5
30,8 31
5,92
224,90
39,05
14,39
21,55
7,30
4,28
3,92
1,87
9,88
22,00
35,39
16,9 from cab 22,6 from cab 41,97
The first was the same location as yesturday although the digging was much harder today. The face was approximately 18m high and only rilled occasionally, digs were often required mid face height to bring down material. There were some very hard parts probably inbetween blast holes. Submerged boulders were very hard to remove with the weight of the face on top of them. (IMG_1592 to IMG_1596) After smoko we tramed down to the bottom of F stage to dig some oar. During the annotation session. Digging here was an improvement on earlier that morning. We were digging the blast face, so intially the face was low as the morning continued we move into a higher face. Graham continued to dig here into the afternoon and N/S. Much of an improvement on Monday. We were digging again on the 379 level again and now the face had built up sufficiently to approximately 15m. Good bucket fill were being acheived through free rilling and well fragmented material. Large Boulders were easily smashed into smaller pieces.
Performance evaluation of a Liebherr 996 hydraulic shovel
5 Results 5.1 Introduction In this chapter data is analyzed to find relations between operator style, diggability and productivity.
5.2 Dig trajectories A part from the statistical information in Chapter 4 (truck load, bucket load, dig time, etc) dig trajectories can also be used to identify an operator style. For this purpose, bucket teeth trajectories are visualized in the x,y-plane. Examples are shown in figure 7 and 8.
Figure 8 Dig trajectory A
Figure 7 Dig trajectory B
The trajectory line is colored red when the operator was digging according to the annotation data. For comparison a summary is printed. This summary shows the trajectory plot and the box plot for each session. Below these plots the bench height, diggability, bucket fill, buckets per truck and truckload are displayed. See Appendix 5.1 and Appendix 5.2. The examples in Figure 7 and 8 are displayed here, because the difference is evident. Interesting to see is that trajectory A shows a style where the face is entered at one height and B is showing one that is entering at two different heights. Method B has shorter average cycle times than method A, but there was no correlation with the average bucket fill. Another remark is that the shovel is closer to the face in situation A than in situation B. In situation A the return movement is more curved than in situation B. Even though there are a few differences to be seen in the trajectory plots, more and clearer correlations are not easily found by studying the prints. A more systematic approach is needed and this will be discussed in the next section. 24
Performance evaluation of a Liebherr 996 hydraulic shovel
5.3 Hypotheses Verifying hypotheses is a more systematic method and the first hypothesis is formulated as follows, 1: “For high bench height operators stand further away from the bench.” It is plausible that operators do not feel safe close to the face when big rocks could fall down from the high bench. When the shovel is close to the bench it can lift the bucket more easily, because of a shorter lever arm. The Matlab script used to test this hypothesis is displayed in Appendix 6.1 Dig reach vs bench height script. Figure 9 shows the result of an analysis of variance for Hypothesis 1. On the vertical axis, the 1 stands for 20m bench, the 2 stands for 15m bench and 3 stands for 10m bench. The horizontal axis represents the dig reach in meters.
Figure 9 Result analysis of variance Hypothesis 1
Since the p-value is 0.0219 and is close to 0.05, the difference between dig reaches is little. Figure 9 shows that the difference between 20m and 15m bench is only 10 cm. The operators digging 20m benches are closer to the bench than the other ones. This proves that Hypothesis 1 is untrue.
25
Performance evaluation of a Liebherr 996 hydraulic shovel To find out if cycle times are influenced by bench height Hypothesis 2 is: 2: “Cycle times are longer when the bench height is higher.” In general operators start their dig at the same height, but in case of a high bench, the dig can end higher. Than the bucket has moved a longer distance through the dirt. Therefore it is plausible this takes more time. The swing, dump and return states are assumed the same for all bench heights. The Matlab script used to test this hypothesis is displayed in Appendix 6.2 Cycle time vs bench height script.
Figure 10 Result analysis of variance Hypothesis 2
After testing this hypothesis the results are surprising. The p-value of the null hypothesis is 1.058 e-13 and is much smaller than 0.05. Therefore there is significant difference between cycle times for different bench heights. Figure 10 shows that the cycle times are the longest for operators digging small bench heights. On the vertical axis, the 1 stands for 20m bench, the 2 stands for 15m bench and 3 stands for 10m bench. The horizontal axis shows the cycle times in seconds. Hence, Hypothesis 2 is untrue.
26
Performance evaluation of a Liebherr 996 hydraulic shovel A part from the bench height digging conditions are compared with the cycle times as well. 3: “If digging conditions are good cycle times are shorter.” When the digging conditions are good the dirt is soft and consistent. It is plausible this is easier to dig and therefore less time consuming. If the dirt is really hard the bucket teeth have difficulties getting through the dirt. To test this hypothesis all the annotations are divided into 3 groups again, group 1 is very soft and well grinded dirt, 3 is normal and 4 is hard dirt. The difference very soft and normal is bigger than the difference between normal and hard dirt. That is why these numbers are used. They are corresponding with the Riley’s diggability index discussed in chapter 4. The Matlab script used to test this hypothesis is displayed in Appendix 6.3 Cycle times vs digging conditions script. After testing Figure 11 shows that the operators digging the soft dirt have significant shorter cycle times. Normal dirt has the longest cycle times. The numbers on the vertical axis represent the digging conditions: 1 for very soft, 3 for normal and 4 for hard. The horizontal axis shows the mean cycle time in seconds.
Figure 11 Result analysis of variance Hypothesis 3
The p-value is 3.109 e-14 and is much smaller than 0.05. Hypothesis 3 is proven to be true. This is not implying that worst digging conditions have the longest cycle times.
27
Performance evaluation of a Liebherr 996 hydraulic shovel For investigating the correlation between the bucket load and digging conditions the following hypothesis is posed, 4: “If digging conditions are good the average bucket load is higher.” It is for sure that operators can dig soft dirt quicker than the harder dirt, but are able to get the bucket fuller in a shorter time? The Matlab script used to test this hypothesis is displayed in Appendix 6.4 Bucket load vs digging conditions script.
Figure 12 Result analysis of variance Hypothesis 4
With a p-value of 0.0008, which is bigger than p-values in hypotheses 2 and 3, Figure 12 shows that bucket fills are in soft dirt significantly better. The numbers on the vertical axis represent the digging conditions: 1 for very soft, 3 for normal and 4 for hard. Horizontal axis shows the average bucket fill in tonnes. Hence, Hypothesis 4 is true.
28
Performance evaluation of a Liebherr 996 hydraulic shovel
5.4 Errors Even though this is a systematic way of testing few things have to be considered. The method used for testing the hypotheses uses the Matlab command anova1, which assumes that all sample populations are normally distributed and have equal variance. Figure 13 shows the box plot of hypothesis 3 (cycle time vs dig conditions). These distributions are not normal and the variances differ slightly. The distributions of the other 3 hypotheses are not perfectly normal either.
Figure 13 Box plot of Hypothesis 3
Another assumption that anova1 makes is that all observations are mutually independent. Consider the bench height hypotheses, the 20m group has in general harder dirt than the 15m and 10m group. This is an example where the observations are not independent. In some annotation sessions the bench height varied a lot during annotation, if the height was most of the time more than 15m than that annotation data ended up in the 20m group. This introduces errors as well. In Hypotheses 2 and 3 cycle times are considered. It is interesting to look at the dig times as well. Dig times are not reviewed, because the dig times have a bigger variance. This is because of different people doing the annotation. On the other hand, cycle times are not so sensitive to timing of those people. If dig state is too late change into swing state, the cycle time can still be accurate but the dig time not.
29
Performance evaluation of a Liebherr 996 hydraulic shovel
5.5 Summary After plotting the trajectories and the box plots for each session different styles are visible, but a more systematic approach is used to prove correlations. The outcome of this approach is that the dig reach and the cycle time are not directly influenced by the bench height. The cycle times are shorter and the bucket load is higher when the digging conditions are good. Even though the approach is systematic there is still uncertainty. The statistical tools assume that the populations are normally distributed and have equal variance. The analyzed populations do not have these properties. On top of that, the values like bench height and cycle times have their own uncertainties as well. Despite these uncertainties the results are interesting and rather valuable to draw conclusions.
30
Performance evaluation of a Liebherr 996 hydraulic shovel
6 Conclusions and recommendations 6.1 Conclusions • • • •
Using method B (Figure 8) the operators are more time efficient. The bench height does not have to be decreased in the future, because the operators do not appear to be more afraid of high bench than a low one. It is important to optimize the digging conditions. Even though it is not possible to change the dig material, the fragmentation is depending on the quality of the blast. Smaller rocks result in better bucket load. Operators still have to wait on a truck quite often, this is inefficient and relative easy to improve.
6.2 Recommendations • •
• •
Investigate the bucket shape, it seemed there is consistently to little dirt in the back of the bucket (the non-teeth site). Calculate the bucket teeth forces out of the logged hydraulic pressure data to quantify damage caused to the shovel. Look for correlations between dig trajectories and damage caused to find an optimal trajectory for minimizing maintenance. Check if the motivation of the operator is a big influence to the productivity. Color the joints of the shovel arm when the pictures are made for the estimation of the lengths. This way the measure points are better visible on the pictures.
31
Performance evaluation of a Liebherr 996 hydraulic shovel
Acknowledgements I want to thank Paul, Ross and Andy especially for being helpful to me during my time at the UQ and even with extracurricular activities! Rose helped me with the visa application and organized all office facilities right from day one. Steve got me real quick on the Internet and the mech-network and borrowed me all equipment I asked for. John for welding my (pauls) bike two times for free. These people work at the Mechanical Engineering department at the university of Queensland during my occupational traineeship. Mark and Bart helped me in the preparation phase of my traineeship and got me in contact with Ross. They answered all my questions without any exception. Maarten Steinbuch for stimulating to do a traineeship abroad and taking away the doubt about the sufficient academic level of my work.
32
Performance evaluation of a Liebherr 996 hydraulic shovel
References [1] [2] [3] [4] [5] [6] [7]
Liebherr-France SAS, http://www.liebherr.com/downloads/TB_R996-GBUS-TB_pdf.pdf Matlab V6.1.0.450 Realease 12.1, Boxplot (statistics toolbox) Help, 2001 Hall, A.S., Characterizing the Operation of a Large Hydraulic Excavator, Thesis, School of Engineering at University of Queensland, Brisbane 2003 http://web2.concordia.ca/Quality/tools/4boxplots.pdf Celesco Transducer Products, Inc., http://www.celesco.com/_datasheets/pt9420.pdf Cooperative Research Centre Mining, http://www.crcmining.com.au/ Department of Mechanical Engineering at the University of Queensland, http://www.uq.edu.au/mecheng/
33
Technical Description Hydraulic Excavator
R 996 litronic
Operating Weight with Backhoe Attachment 659 t/1,452,800 lb Operating Weight with Shovel Attachment 668 t/1,472,700 lb Engine Output 3000 HP (2240 kW) Bucket Capacity 25,00 – 36,00 m3/32.7 – 47.1 cuyd Shovel Capacity 25,00 – 36,00 m3/32.7 – 47.1 cuyd
`
Technical Data Engine 2 Cummins diesel engines Rating per SAE J 1995 ________________ 3000 HP (2240 kW) at 1800 RPM Model ________________________ K 1800 E Type __________________________ 16 cylinder V-engine, water-cooled, direct injection, turbo-charged, after-cooler Displacement ____________ 50,3 l/3069 cu.in Bore/Stroke ________________ 159/159 mm/6.26/6.26 in Air cleaner ______________________ dry-type air cleaner with pre-cleaner, with automatic dust ejector, primary and safety elements Fuel tank ________________________ 13 000 l/3440 gal Electrical system Voltage ______________________ 24 V Batteries ____________________ 8 x 170 Ah/12 V Alternator __________________ 2 x 24 V/150 Amp Engine idling ____________________ sensor controlled
Swing Drive Hydraulic motor ________________ 4 Liebherr axial piston motors Swing gear ______________________ 4 Liebherr planetary reduction gears Swing ring ______________________ Liebherr, sealed triple roller swing ring, internal teeth Swing speed ____________________ 0 – 3.5 RPM Swing-Holding brake ________ hydraulically released, maintenance-free, multi-disc brakes integrated in each swing gear
Uppercarriage Design
torque resistant designed upper frame in box type construction for superior strength and durability Attachment mounting ________ parallel longitudinal main girders in boxsection construction Catwalks ________________________ on the right side with a hydraulically driven access ladder, additional emergency ladder in front of the cab
Hydraulic System Hydraulic pumps for attachment and travel drive ________________ 8 variable flow axial piston pumps Max. flow __________________ 8 x 840 l/min/8 x 222 gpm Max. hydr. pressure ____ 320 bar/4640 PSI Hydraulic pumps for swing drive ____________ 4 reversible swash plate pumps, closedloop circuit Max. flow __________________ 4 x 413 l/min/4 x 109 gpm Max. hydr. pressure ____ 350 bar/5076 PSI Pump regulation ______________ electro-hydraulic, pressure compensation, flow compensation, automatic oil flow optimizer Hydraulic tank capacity ____ 4600 l/1215 gal Hydraulic system capacity __________________________ 8200 l/2166 gal Hydraulic oil filter ______________ filtration of entire return flow, 1 high pressure filter for each main pump Hydraulic oil cooler __________ 2 separate coolers, 4 temperature controlled fans driven via hydraulic piston motors Electronic engine speed sensing __________________ over the entire engine RPM range Lubrication ______________________ central lubrication system
Hydraulic Controls Servo circuit
independant, electric over hydraulic proportional controls of each function Emergency control ______ via accumulator for all attachment functions with stopped engine Power distribution ____________ via monoblock control valves with integrated primary relief valves and flanged on secondary valves for travel Flow summation ________ to attachment and travel drive Control functions Attachment and swing ________________________ proportional via joystick levers Travel ________________________ proportional via foot pedals or hand levers Bottom dump bucket __ proportional via foot pedals Operation with one engine possible
2
____________________
TD R 996 Litronic
____________________________
Service Flap Design
hydraulically actuated service flap, easily accessible from ground level to allow: – fuel fast refill – hydraulic oil refill – engine oil quick change – splitterbox oil quick change – swing gearbox oil quick change – swing ring gearing grease barrel refilling via grease filter – attachment/swing ring bearing grease barrel refilling via grease filter – windshield washer water refilling Quick coupler upon request ____________________________
Technical Data Operator’s Cab Design
resiliently mounted, sound insulated, large windows for all-around visibility, integrated falling object protection FOPS Operator’s seat ________________ suspended, body-contoured with shock absorber, adjustable to operator’s weight Cabin windows ________________ 20,5 mm/0.8 in tinted armored glass for front window and left hand side windows, all other windows in tinted safety glass, high pressure windshieldwasher-system with 75 l/20 gal watertank, sun louvers on all windows in heavy duty design Heating system/ Air conditioning ________________ heavy duty, high output air conditioner and heater unit Cabin pressurization ________ ventilation unit with filters Controls __________________________ joystick levers integrated into armrest of seat Monitoring ______________________ via LCD-Display, data memory Automatic engine shut off ____________________________ in case of low engine oil pressure or low coolant level Destroking of main pumps ____________________ in case of engine overheating or low hydraulic oil level Safety functions ______________ additional gauges with constant display for: engine speed, hourmeter, engine oil pressure, coolant temperature and hydraulic oil temperature ____________________________
Central Lubrication System Type ________________________________ Lincoln Centromatic lubrication system for the entire attachment and swing ring Grease pumps ________________ 2 Lincoln Powermaster pumps with switch over function, plus 1 separate pump for swing ring teeth Capacity __________________________ 1 x 600 l/158.5 gal bulk container for attachment and swing ring, separated 1 x 80 l/21 gal grease drum for swing ring teeth
Attachment Design
box type structure with large steel castings in all high-stress areas Pivots ____________________________ sealed with double side centering with 1 single floating pin per side, all bearings with wear resistant, steel bushings, bolts hardened and chromium-plated Hydraulic cylinders __________ Liebherr design, all cylinders located in well protected areas Hydraulic connections ______ pipes and hoses equipped with SAE split flange connections Kinematics ______________________ Liebherr parallel face shovel attachment geometry ____________________________
Undercarriage Design
3-piece undercarriage, box type structures for center piece and side frames, stress relieved Hydraulic motor ________________ 2 axial piston motors per side frame Travel gear ______________________ Liebherr reduction gear Travel speed ____________________ 0 – 2,2 km/h/0 – 1.4 mph Parking brake __________________ spring engaged, hydraulically released wet multi-disc brakes for each travel motor, maintenance-free Track components ____________ maintenance-free combined pad-link, heavy duty track shoes Track rollers/ Carrier rollers __________________ 7/3 Automatic track tensioner ________________________ pressurized hydraulic cylinder with accumulator, maintenance free Transport ________________________ undercarriage side frames are removable ____________________________
TD R 996 Litronic
3
Dimensions A2
E
A1
D
A
F
W C
H1 H P
K Q L
N
U
S Z
B
V X
mm/ft-in 7000/22’11” 7430/24’ 4” 8250/27’ 1” 7650/25’ 1” 9070/29’ 9” 7550/24’ 9” 7795/25’ 7” 2780/ 9’ 1” 6275/20’ 7” 8480/27’10” 2845/ 9’ 4”
A A1 A2 B C D E F H H1 K
mm/ft-in 7500/24’ 7” 10000/32’ 9” 2985/ 9’ 9” 1435/ 4’ 8” 6000/19’ 8” 1400/ ’55” 9750/32’ 0” 14350/47’ 1” 22600/74’ 1” 12465/40’10”
L U P Q S N W V X Z
A2 A1
E
A
D
F
C
W1 H1 H P
K Q N
L U
S Z
B
V1 X1
A A1 A2 B C D E F H H1 K
4
TD R 996 Litronic
mm/ft-in 7000/22’11” 7430/24’ 4” 8250/27’ 1” 7650/25’ 1” 9070/29’ 9” 7550/24’ 9” 7795/25’ 7” 2780/ 9’ 1” 6275/20’ 7” 8480/27’10” 2845/ 9’ 4”
L U P Q S N W1 V1 X1 Z
mm/ft-in 7500/24’ 7” 10000/32’ 9” 2985/ 9’ 9” 1435/ 4’ 8” 6000/19’ 8” 1400/ ’55” 8500/27’10” 17800/58’ 4” 23450/76’11” 12465/40’10”
Backhoe Attachment m
ft
18
60
16 50 14 12
40
Digging envelope Max. reach at ground level Max. teeth height Max. dump height Max. digging depth Max. digging force Max. breakout force
20,00 m/65’ 7” 16,60 m/54’ 5” 10,50 m/34’ 5” 8,80 m/28’10” 1500 kN (153,0 t)/337,100 lb 1670 kN (170,2 t)/375,300 lb
10 30 8 6
20
4 10 2 0
0
-2 -10 -4 -6
Operating Weight and Ground Pressure The operation weight includes the basic machine with backhoe attachment and bucket 33,00 m3/43.1 cuyd. Pad width Weight Ground pressure
mm/in 1400/55 kg/lb 659000/1,452,800 kg/cm2/PSI 2,81/39.97
-20
-8
m0 ft 0
2
4 10
6 20
8
10 30
12 40
14
16 50
18 60
20
-10 22
-30
70
Bucket Cutting width SAE Capacity SAE heaped Weight Suitable for material up to a specific weight of Wear kit level 1)
mm/in m3/cuyd kg/lb t/m3/lb/cuyd
4150/163”1) 25,00/32.7 35000/77,160 2,50/4000 III
4800/189”1) 30,00/39.2 39000/85,980 2,20/3700 II
4800/189”1) 33,00/43.1 41200/90,760 1,80/3000 II
4800/189”1) 36,00/47.0 41500/91,500 1,60/2700 II
Bucket with delta cutting edge and tooth system Posilok size S 145.
Level II: For heavy rock, not detoriorated or cracked. Has to be shot to be dug. Level III: For highly-abrasive materials such as rock with a high silica content, sandstone etc.
TD R 996 Litronic
5
Shovel Attachment m
ft 70
20 18
60
Digging Envelope Max. reach at ground level Max. dump height Max. crowd length Bucket opening width T
15,60 m/51’ 2” 14,30 m/46’11” 6,40 m/21’ 0” 2800 mm/110”
16 50 14
T
12
40
Crowd force at ground level Max. crowd force Max. breakout force
1960 kN (199,8 t)/440,450 lb 2340 kN (238,5 t)/525,850 lb 1905 kN (194,2 t)/428,100 lb
10 30 8 6
20
The operation weight includes the basic machine with shovel attachment and bottom dump bucket 34,00 m3/44.4 cuyd.
4 10 2 0
Operating Weight and Ground Pressure
0
Pad width Weight Ground pressure
mm/in 1400/55 kg/lb 668000/1,472,700 kg/cm2/PSI 2,848/40.53
-2 -1 -4 m0 ft 0
2
4 10
6 20
8
10 30
12 40
14
16 50
18 60
Bottom Dump Bucket Cutting width SAE Capacity SAE heaped Weight Suitable for material up to a specific weight of Wear kit level 1)
mm/in m3/cuyd kg/lb t/m3/lb/cuyd
4150/163”1) 25,00/32.7 49050/108,130 2,50/4000 III
4700/185”1) 29,00/37.9 53600/118,160 2,20/3700 II
Bottom dump bucket with delta cutting edge and tooth system Posilok size S 145
Level II: For heavy rock, not deteriorated or cracked. Has to be shot to be dug. Level III: For highly-abrasive materials such as rock with a high silica content, sandstone etc.
6
TD R 996 Litronic
5500/217”1) 34,00/44.4 59400/130,950 1,80/3000 II
5500/217”1) 36,00/47.0 64000/141,100 1,60/2700 II
Component Dimensions and Weights L
Cab L Length H Height Width Weight
H
mm/ft-in mm/ft-in mm/ft-in kg/lb
3215/10’ 6” 2885/ 9’ 6” 1900/ 6’ 3” 2800/6,200
Cab Elevation with Fuel Tank
L
L Length H Height Width Weight
H
L
mm/ft-in mm/ft-in mm/ft-in kg/lb
Powerpack Modules L Length H Height Width Weight
H
4150/13’ 7” 3100/10’ 2” 2700/ 8’10” 8000/17,650
(two)
mm/ft-in mm/ft-in mm/ft-in kg/lb
5280/17’ 4” 3640/11’11” 2070/ 6’ 9” 2 x 22000/2 x 48,500
Rotation Deck (with swing ring, swing gears and control valve bracket)
L
L Length H Height Width Weight
H
L
9750/32’ 0” 4250/13’11” 4270/14’ 0” 83100/183,200
mm/ft-in mm/ft-in mm/ft-in kg/lb
1250/ 4’ 1” 3430/11’ 3” 7360/24’ 2” 60000/132,300
Counterweight L Length H Height Width Weight
H
mm/ft-in mm/ft-in mm/ft-in kg/lb
Hydraulic Oil Cooling L
L Length H Height Width Weight
H
L
H
with hydraulic tank without hydraulic oil mm/ft-in mm/ft-in mm/ft-in kg/lb
Compartment Panel L Length H Height Width Weight
4210/13’10” 3100/10’ 2” 2100/ 6’11” 8000/17,650
(two)
mm/ft-in mm/ft-in mm/ft-in kg/lb
4145/13’ 7” 3100/10’ 2” 950/ 3’ 1” 2 x 1500/2 x 3,300
kg/lb
8000/17,640
Hydraulic Oil Weight
TD R 996 Litronic
7
Component Dimensions and Weights Miscellaneous
L
L Length H Height Width Weight
H
Side Frame
L
L
4500/14’ 9” 2600/ 8’ 6” 2000/ 6’ 7” 5000/11,100
mm/ft-in mm/ft-in mm/ft-in mm/ft-in kg/lb
10000/32’ 9” 2985/ 9’ 9” 2700/ 8’11” 2225/ 7’ 4” 2 x 117000/2 x 258,000
(two)
L Length H Height Width over travel drive Width without travel drive Weight
H
mm/ft-in mm/ft-in mm/ft-in kg/lb
Undercarriage Central Girder L Length H Height Width Weight
H
mm/ft-in mm/ft-in mm/ft-in kg/lb
4000/13’ 1” 2690/ 8’10” 4600/15’ 1” 40000/88,200
mm/ft-in mm/ft-in mm/ft-in kg/lb
8650/28’ 4” 3300/10’10” 3350/11’ 0” 59140/130,400
L
Shovel Boom L Length H Height Width Weight
H
Hoist Cylinder
L
(two)
L Length Ø Diameter Weight
mm/ft-in mm/in kg/lb
5430/17’10” 600/ ’24” 2 x 5910/2 x 13,050
L Length H Height Width Weight
mm/ft-in mm/ft-in mm/ft-in kg/lb
5620/18’ 5” 2300/ 7’ 6” 3350/11’ 0” 27150/59,850
Crowd Cylinder
(two)
L Length Ø Diameter Weight
mm/ft-in mm/in kg/lb
Shovel Stick
L
H
L
3880/12’ 9” 490/ ’19” 2 x 3430/2 x 7,560
Bottom Dump Bucket (including clam cylinders) Cutting width mm/in Capacity L Length H Height Width Weight
H
L
4150/163” 4700/185” 5500/217” 5500/217”
m3/cuyd 25,00/32.7 29,00/37.9 34,00/44.4 mm/ft-in 4650/15’3” 4650/15’3” 4650/15’3” mm/ft-in 4500/14’9” 4500/14’9” 4500/14’9” mm/ft-in 4150/13’7” 4700/15’5” 5500/18’0” kg 49050 53600 59400 lb 108,130 118,160 130,950
36,00/47.0 4650/15’3” 5040/16’6” 5670/18’7” 64000 141,100
TD R 996 Litronic
8
Component Dimensions and Weights L
Bucket Tilt Cylinder L Length Ø Diameter Weight
(two)
mm/ft-in mm/in kg/lb
4690/15’ 5” 490/ ’19” 2 x 3670/2 x 8,090
Gooseneck Boom with Two Stick Cylinders
L
H
L
L Length H Height Width Weight
mm/ft-in mm/ft-in mm/ft-in kg/lb
Hoist Cylinders
(two)
L Length Ø Diameter Weight
mm/ft-in mm/in kg/lb
12500/41’ 0” 4500/14’ 9” 2800/ 9’ 2” 68950/152,000
5430/17’10” 600/ ’24” 2 x 6060/2 x 13,360
Stick with Two Bucket Cylinders
L
L Length H Height Width Weight
H
mm/ft-in mm/ft-in mm/ft-in kg/lb
7500/24’ 7” 3000/ 9’10” 2500/ 8’ 2” 40800/89,950
Backhoe Buckets Cutting width L
H
Capacity L Length H Height Width Weight
mm/in m3/cuyd mm/ft-in mm/ft-in mm/ft-in kg lb
4150/163”
4800/189”
4800/189”
4800/189”
25,00/32.7 30,00/39.2 33,00/43.1 36,00/47.0 4650/15’3” 4650/15’ 3” 4650/15’ 3” 4900/16’ 0” 3150/10’4” 3150/10’ 4” 3300/10’10” 3400/11’ 2” 4200/13’9” 4850/15’11” 4850/15’11” 4850/15’11” 35000 39000 41200 41500 77,150 86,000 90,760 91,500
TD R 996 Litronic
9
Printed in Germany by Wolf RG-BK-RP LFR/SP 8410797-3-04.04 Illustrations and data may differ from standard equipment. Subject to change without notice.
Liebherr-France SAS 2, Avenue Joseph Rey, B.P. 287, F-68005 Colmar Cedex +33 (0)389 21 35 10, Fax +33 (0)389 21 37 93 www.liebherr.com, E-Mail:
[email protected]
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 2.1 Configuration simulation script clc; global start fin t SH SH = 100; % plot Shrink fartor start = 1; fin = floor( (length(signals.CPBoom)/SH) ) ; comp; % Computes the cylinder lengths out of the logged sensor data (voltage) and generates lvVec. plot(ti,bo); extension)
%decide which part of the data set you want (time against Boom cylinder
SH = 10; %trajec shrink factor tstart = input('Enter begin time number: '); tfin = input('Enter end time number: '); start = tstart *50/SH; fin = floor(tfin *50/SH);
% data is logged at 50 Hz on the shovel
comp; % Computes the cylinder lengths out of the logged sensor data (voltage) and generates lvVec. initialiseV2; % Loads the values of lc tc and calculates the initial guess for tv. k=1; I=[]; TVDEG=[]; BX=[]; BY=[]; lc = [l1 l2 l5 l6 l7 l8 l10]'; tc = [t1 t5 t7 t10]'; tv = [t2 t3 t4 t6 t8 t9]'; del_tv = zeros(6,1); L_lvVec = length(lvVec); for k=1:L_lvVec, lv = [lvVec(k,1) lvVec(k,2) lvVec(k,3)]'; minError = 0.0001; i=1; % Run an intitial solution to get the algorithm up and running. % Initially set Gamma large to enter while loop F = ones(6,1); while (max(abs(F)) > minError) update; [F] = Gamma(lc, lv, tc, tv); [J, inv_J] = Jacobian(lc, lv, tc, tv); del_tv = -(inv_J * F); tv = tv + del_tv; i=i+1; end I=[I i]; tvdeg=tv./con; TVDEG=[TVDEG tvdeg]; %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Parameters
44
Performance evaluation of a Liebherr 996 hydraulic shovel alf = 100; % degrees Angle between t8 and line trough the joint - tip. bet = 52; %degrees Angle between line trough the joint - tip and teeth direction. tB = tvdeg(5) + alf + bet; % Last but not least, the position of the bucket (B) % Parameters bx = 0.7; % meters What's the x distance to point b (b is where the boom and the machine house meet) by = 6.3; % meters What's the y distance to point b L1 = 7.6; % meters distance between house-boom joint and boom-stick joint L2 = 4.7; % l7 (not seventeen) meters distance between boom-stick joint and the stick-bucket joint L3 = 4.04; % meters distance between stick-bucket joint and B C1 = 0; % degrees This is the constant angle between L1 and t2 % calculation of the position of B g = (tvdeg(1) + C1)/180*pi; h = (tvdeg(5) + alf)/180*pi; Bx = bx + L1*cos(g) + L2*cos(tc(3)) + L3*cos(h); By = by + L1*sin(g) + L2*sin(tc(3)) + L3*sin(h); % Plot configutation x = [0;... bx;... bx + L1*cos(g);... bx + L1*cos(g) + L2*cos(tc(3));... bx + L1*cos(g) + L2*cos(tc(3)) + 1.74*cos(tv(5));... %1.74 moet beter uitgezocht worden bx + L1*cos(g) + L2*cos(tc(3)) + 2.9*cos(tv(5) + (45/180*pi));... %2.9 unsure Bx]; y = [0;... by;... by + L1*sin(g);... by + L1*sin(g) + L2*sin(tc(3));... by + L1*sin(g) + L2*sin(tc(3)) + 1.74*sin(tv(5)) ;... %idem by + L1*sin(g) + L2*sin(tc(3)) + 2.9*sin(tv(5) + (45/180*pi));... %idem By]; clf plot(x,y,'-ko','lineWidth',1,'MarkerSize',2) %axis equal; axis([0 20 0 20]) pause(0.12) end
45
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 2.2 Trajectory simulation script clc; global start fin t SH SH = 100; % plot SHrink fartor start = 1 fin = floor( (length(signals.CPBoom)/SH) ) ; comp; % Computes the cylinder lengths out of the logged sensor data (voltage) and generates lvVec. plot(ti,bo);
%decide which part of the data set you want
SH = 10; %trajec shrink gactor tstart = input('Enter begin time number: '); tfin = input('Enter end time number: '); start = tstart *50/SH; fin = floor(tfin *50/SH); comp; % Computes the cylinder lengths out of the logged sensor data (voltage) and generates lvVec. initialiseV2; % Loads the values of lc tc and calculates the initial guess for tv. k=1; i=0; I=[]; TVDEG=[]; BX=[]; BY=[]; lc = [l1 l2 l5 l6 l7 l8 l10]'; tc = [t1 t5 t7 t10]'; tv = [t2 t3 t4 t6 t8 t9]'; del_tv = zeros(6,1); L_lvVec = length(lvVec); for k=1:L_lvVec, lv = [lvVec(k,1) lvVec(k,2) lvVec(k,3)]'; minError = 0.00001; % Run an intitial solution to get the algorithm up and running. % Initially set Gamma large to enter while loop F = ones(6,1); while (max(abs(F)) > minError) update; [F] = Gamma(lc, lv, tc, tv); [J, inv_J] = Jacobian(lc, lv, tc, tv); del_tv = -(inv_J * F); tv = tv + del_tv; i=i+1; end I=[I i]; tvdeg=tv./con; TVDEG=[TVDEG tvdeg]; %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Parameters alf = 100; % degrees Angle between t8 and line trough the joint - tip. bet = 52; %degrees Angle between line trough the joint - tip and teeth direction.
46
Performance evaluation of a Liebherr 996 hydraulic shovel tB = tvdeg(5) + alf + bet; % Last but not least, the position of the bucket (B) bx = 0.7; % meters What's the x distance to point b (b is where the boom and the machine house meet) by = 6.3; % meters What's the y distance to point b L1 = 8; % meters distance between house-boom joint and boom-stick joint L2 = l7; % (not seventeen) meters distance between boom-stick joint and the stickbucket joint L3 = 4.4; % meters distance between stick-bucket joint and B C1 = 0; % degrees This is the constant angle between L1 and t2 % calculation of the position of B g = (tvdeg(1) + C1)/180*pi; h = (tvdeg(5) + alf)/180*pi; Bx = bx + L1*cos(g) + L2*cos(tc(3)) + L3*cos(h); By = by + L1*sin(g) + L2*sin(tc(3)) + L3*sin(h); % trajectory BX = [BX; Bx]; BY = [BY; By]; % Plot configutation end plot(BX,BY,'-ko','lineWidth',1,'MarkerSize',2) axis equal;
47
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 2.3 Sub-scripts used in simulation scripts Comp: t = 0:0.02:length(signals.CPBoom)*0.02-0.02; % time array creating (50 Hz) [ti]=short(t,SH); [bo]=short(signals.CPBoom,SH); [st]=short(signals.CPStick,SH); [bu]=short(signals.CPBucket,SH); % bu= (bu - 2.52)./(470 * 0.00315) + 4.26; % st= (st - 2.17)./(470 * 0.00315) + 3.38; % bo= (bo - 2.22)./(470 * 0.00315) + 4.9; % min voltage/ min length calculation bu= (bu - 6.13)./(470 * 0.00315) + 6.45; %6.71 st= (st - 5.3)./(470 * 0.00315) + 5.5; %5.7 bo= (bo - 7.18)./(470 * 0.00315) + 8.1; % max voltage/ max length calculation lvVec=[bo(1,start:fin)' st(1,start:fin)' bu(1,start:fin)'];
InitializeV2: %convert from degrees to radians con=pi/180; % set parameters and link lengths l1 = 3.16; %meter l2 = 5.78; %meter l5 = 4.63; %meter l6 = 1.6; %meter l7 = 4.7; %meter l8 = 1.1; %meter l10 = 1; %meter % define the conversion angles c_2_5 = 5*con; % Constand angle between t2 t5 in degrees c1 c_5_10 = 6*con; % Constand angle between t5 t10 in degrees c2 c_6_7 = 20*con; % Constand angle between t6 t7 in degrees c3 % definer constant angle t1 = 120*con; % define variabel lengths dv l3 = lvVec(1,1); l4 = lvVec(1,2); l9 = lvVec(1,3); [t2,t3] = solve([num2str(l1),'*cos(',num2str(t1),')+',num2str(l2),'*cos(t2)+',num2str(l3),'*cos(t3) '], ... [num2str(l1),'*sin(',num2str(t1),')+',num2str(l2),'*sin(t2)+',num2str(l3),'*sin(t3)']); t2 = double(t2); % acw means anti clock wize and makes sure t3 = double(t3); % all used angles are in between 0 and 2*pi %Which one of the solutions is the one we wont to use? ff1 = find(t2 > -10*con & t2 < 90*con); t3 = t3(ff1);
48
Performance evaluation of a Liebherr 996 hydraulic shovel t2 = t2(ff1); t5 = t2 + c_2_5; t5 = acw(t5); [t4,t6] = solve([num2str(l4),'*cos(t4)+',num2str(l5),'*cos(',num2str(t5),')+',num2str(l6),'*cos(t6) '], ... [num2str(l4),'*sin(t4)+',num2str(l5),'*sin(',num2str(t5),')+',num2str(l6),'*sin(t6)']); t4 = double(t4); t6 = double(t6); % Find the right solution: % Angle [t5 - t6] (between l5 and l6) must be inbetween 20 and 160 deg ff2 = find((t5 - t6) < 160*con & (t5 - t6) > 20*con); t4 = t4(ff2); t6 = t6(ff2); t7 = t6 + c_6_7; t7 = acw(t7); t10 = t5 - c_5_10; t10 = acw(t10); [t8,t9] = solve([num2str(l8),'*cos(t8)+',num2str(l7),'*cos(',num2str(t7),')+',num2str(l10),'*cos(', num2str(t10),')+',num2str(l9),'*cos(t9)'], ... [num2str(l8),'*sin(t8)+',num2str(l7),'*sin(',num2str(t7),')+',num2str(l10),'*sin(',num2st r(t10),')+',num2str(l9),'*sin(t9)']); t8 = double(t8); %t8(1)=acw(t8(1)); t8(2)=acw(t8(2)); t9 = double(t9); %t9(1)=acw(t9(1)); t9(2)=acw(t9(2)); % Find the right solution: % Angle [t7 - t8] (between l7 and l8) must be inbetween 20 and 170 deg ff3 = find( (t7 - t8) > 20*con & (t7 - t8) < 170*con ); t8 = t8(ff3); t9 = t9(ff3); if t8 == [] [t8,t9] = solve([num2str(l8),'*cos(t8)+',num2str(l7),'*cos(',num2str(t7),')+',num2str(l10),'*cos(', num2str(t10),')+',num2str(l9),'*cos(t9)'], ... [num2str(l8),'*sin(t8)+',num2str(l7),'*sin(',num2str(t7),')+',num2str(l10),'*sin(',num2st r(t10),')+',num2str(l9),'*sin(t9)']); t8 = double(t8); %t8(1)=acw(t8(1)); t8(2)=acw(t8(2)); t9 = double(t9); %t9(1)=acw(t9(1)); t9(2)=acw(t9(2)); ff3 = find( (360*con + t7 - t8) > 20*con & (360*con + t7 - t8) < 170*con ); t8 = t8(ff3); t9 = t9(ff3); else t8=t8; t9=t9; end
Update: % converion constants c_2_5 = 5*con; % Constand angle between t2 t5 in degrees c1 c_5_10 = 6*con; % Constand angle between t5 t10 in degrees c2 c_6_7 = 20*con; % Constand angle between t6 t7 in degrees c3 t5 = tv(1,1) + c_2_5; t7 = tv(4,1) + c_6_7;
49
Performance evaluation of a Liebherr 996 hydraulic shovel t10 = tc(2,1) - c_5_10; tc = [tc(1,1) t5 t7 t10]';
Gamma: function [F] = GammaKin(lc, lv, tc, tv) %Develop F matrix (should be 6X1) F=[ lc(1)*cos(tc(1)) + lc(2)*cos(tv(1)) + lv(1)*cos(tv(2)); lc(1)*sin(tc(1)) + lc(2)*sin(tv(1)) + lv(1)*sin(tv(2)); lv(2)*cos(tv(3)) + lc(3)*cos(tc(2)) + lc(4)*cos(tv(4)); lv(2)*sin(tv(3)) + lc(3)*sin(tc(2)) + lc(4)*sin(tv(4)); lc(5)*cos(tc(3)) + lc(6)*cos(tv(5)) + lv(3)*cos(tv(6)) + lc(7)*cos(tc(4)); lc(5)*sin(tc(3)) + lc(6)*sin(tv(5)) + lv(3)*sin(tv(6)) + lc(7)*sin(tc(4))];
Jacobian: % Develop partial derivate matrix (6X6) of del F del Theta function [J, inv_J] = Jacobian(lc, lv, tc, tv); J = [-lc(2)*sin(tv(1)), -lv(1)*sin(tv(2)), 0, 0, 0, 0; ... lc(2)*cos(tv(1)), lv(1)*cos(tv(2)), 0, 0, 0, 0; ... 0, 0, -lv(2)*sin(tv(3)), -lc(4)*sin(tv(4)), 0, 0; ... 0, 0, lv(2)*cos(tv(3)), lc(4)*cos(tv(4)), 0, 0; ... 0, 0, 0, 0, -lc(6)*sin(tv(5)), -lv(3)*sin(tv(6)); ... 0, 0, 0, 0, lc(6)*cos(tv(5)), lv(3)*cos(tv(6))]; inv_J = inv(J);
50
Cable-Extension Position Transducer ▼ Industrial Grade – Medium to Long Range ▼ 4...20 mA Output Signal ▼ Intrinsically Safe
R
Specification Summary:
PT9420
GENERAL Full Stroke Ranges–on this datasheet ..... 0-75 to 0-550 inches, see next page Output Signal ................................ 4 - 20 mA (2-wire), 0 - 20 mA (3-wire), see Sensitivity, 2-wire ....................................................... 16 mA/full stroke, ± 0.25% Sensitivity, 3-wire ....................................................... 20 mA/full stroke, ± 0.25% Accuracy ...................................................................... + 0.12% full stroke, max. Repeatability ................................................................ + 0.05% full stroke, max. Resolution ............................................................................... essentially infinite Measuring Cable ............... stainless steel, nylon-coated or thermoplastic, see Enclosure Material ............... powder-painted aluminum or stainless steel, see Sensor .................................................... plastic-hybrid precision potentiometer Potentiometer Cycle Life ............................................................... 250,000, min. Maximum Retraction Acceleration ............................................................. see Maximum Velocity ..................................................................................... see Weight, Aluminum (Stainless Steel) Enclosure ..................... 8 lbs. (16 lbs.) max.
R
ENVIRONMENTAL Enclosure Design ..................................... NEMA 4/4X/6, IP 67/68, see and Hazardous Area Certification (see ) CSA 22.2: ......................................................... Class 1, Groups A, B, C and D Cenelec: .............................................................................. LCIE EEx ia IIc T4 Operating Temperature ............................................................... -40o F to 200oF Thermal Effects Zero ........................................................................ 0.01% full stroke / oF, max. Span ........................................................................................ 0.01% / oF, max. Vibration ............................................................... up to 10 G's to 2000 Hz max. EMC COMPLIANCE PER DIRECTIVE 89/336/EEC Emission / Immunity ..................................................... EN50081-2 / EN50082-2 IMPORTANT!
Dimensions are only valid for Enclosure/Cable-Tension Options 1 & 3 only refer to Supplement 9-A for increased cable tension.
0.85 [21.6]
6.15 ± 0.06 [152,2 ± 1,5]
"A" DIMENSION MEASURING CABLE DIAMETER 0.034 in. 0.047 in. 0.062 in. 0.22 0.29 0.37 0.29 0.39 0.49 0.44 0.59 0.73 0.58 0.79 0.98 0.73 0.98 1.22 0.88 1.18 1.47 1.02 1.38 1.71 1.17 1.57 1.96 1.31 1.77 N/A 1.46 1.97 N/A 1.61 N/A N/A
RANGE 75 100 150 200 250 300 350 400 450 500 550
4.80 [121,9]
DIMENSIONS ARE IN INCHES (MM)
1.10 ± 0.08 [27,9 ± 2,0] measuring cable at full extension
TOLERANCES ARE ±0.03 IN. (±0.5 MM) UNLESS OTHERWISE NOTED
measuring cable at full retraction
The PT9420 is a great value for demanding long-range applications requiring a 4 - 20 mA linear position feedback signal. Sealed to meet NEMA 4 standards, this Cable-Extension Transducer will perform even under the harshest of environmental conditions. The PT9420 provides a 4 - 20 mA feedback signal that is proportional to the linear movement of a traveling stainless-steel extension cable. Simply mount the body of the transducer to a fixed surface and attach the extension cable to the moving object. Electrical Output Signal:
20 mA
4 mA
OUTPUT SIGNAL (milliamperes)
ELECTRICAL Input Voltage ................................................................................. 12 to 40 VDC Input Current ....................................................................................... 4 - 20 mA Loop Resistance (Load) ........................... (loop supply voltage - 12 ) / 0.02, max Circuit Protection ..................................................................... 38 mA maximum Impedance ................................................................ 100 MW @ 100 VDC, min. Zero and Span Adjustment ............................................................. 2:1 turndown
FULL STROKE RANGE
0.27 [6,9]
A
4.52 ± 0.08 [114,8 ± 2,0]
0.64 [16,3]
5.16 [131,1] max. 4.90 [124,5]
4.75 ± 0.06 [120,7 ± 2,0]
0.25 [6,4] 7.87 [199,9] max.
Celesco Transducer Products, Inc. 5.30 [134,6]
169
7800 Deering Avenue • Canoga Park, CA • 91309 tel: (800) 423-5483 • (818) 884-6860 • fax: (818) 340-1175 www.celesco.com •
[email protected]
PT9420 • Cable-Extension Transducer • 4...20 mA Output Signal
▼ Ordering Information Model Number:
PT9420-
R
order code:
-
A
B
C
- 1D
0 E
F
G
Full Stroke Range: R order code: full stroke range, min:
0075
0100
0150
0200
0250
0300
75 inches
100 inches
150 inches
200 inches
250 inches
300 inches
R order code: full stroke range, min:
0350
0400
0450*
0500*
0550*
350 inches
400 inches
450 inches
500 inches
550 inches
note: *42 oz. cable tension (see
below) for these ranges is strongly recommended!
Enclosure Material and Measuring Cable Tension: 1 2 A order code: enclosure: see drawing (front page) see Supplement 9-A enclosure material: powder-painted aluminum cable tension (+30%): 26 oz. 42 oz. maximum velocity: 60 inches per second 200 inches per second max. retraction acceleration: 1G 5 G's
3
4
see drawing (front page) see Supplement 9-A 303 stainless steel 26 oz. 42 oz. 20 inches per second 80 inches per second 0.33 G 2 G's
Measuring Cable: B order code: cable construction:
1*
2**
3***
.034 nylon-coated stainless steel
.047 stainless steel
.062 thermoplastic
0.190 in. (4.83 mm) dia. thru
0.190 in. (4.83 mm) dia. thru
0.190 in. (4.83 mm) dia. thru
0.047 in. (1.19 mm) dia.
0.034 in. (0.86 mm) dia.
0.170 in. (4.32 mm)
0.170 in. (4.32 mm)
notes:
* available in all ranges
0.062 in. (1.57 mm) dia.
0.170 in. (4.32 mm)
** available in ranges up to 500-inches only
*** available in ranges up to 400-inches
Cable Exit: C order code: direction:
1
2
3
4
front
top
rear
bottom
7800 Deering Avenue • Canoga Park, CA • 91309 • tel: (800) 423-5483 • (818) 884-6860 • fax: (818) 340-1175 www.celesco.com •
[email protected]
170
PT9420 • Cable-Extension Transducer • 4...20 mA Output Signal
Output Signals:
1
2
3
4
5*
6*
4...20 mA [2-wire]
20...4 mA [2-wire]
0...20 mA [3-wire]
20...0 mA [3-wire]
4...20 mA [2-wire]
20...4 mA [2-wire]
E order code: output signal configuration:
20 mA
20 mA
20 mA
4 mA
4 mA
20 mA
20 mA
0 mA
4 mA
0 mA
4 mA
R
hazardous area certification
none
none
none
20 mA
none
CSA Standard 22.2 Class 1 Groups A, B, C and D Cenelec LCIE EEx ia IIc T4
*Important: Intrinsically Safe when powered from a CSA certified zener barrier rated 28 VDC max, 110 mA max as per Installation Drawing 677984.
Electrical Connection: F order code: electrical connection:
1
2
3
4
6-pin plastic connector and mating plug
10 ft. waterproof cable
6-pin metal connector and mating plug
25 ft. instrumentation cable
MS3106E-14S-6S 5/16" (8 mm) max cable dia. 16 AWG max conductor size
MS3106E-14S-6S 5/16" (8 mm) max cable dia. 16 AWG max conductor size
3-CONDUCTOR, 18 GA., TYPE SJOW-A ~0.35 in. (9 mm) dia.
~ 3.2 in. (81 mm)
B
F
C
E D
IP rating: NEMA rating: notes:
MS3102E-14S-6P
connections
A
connections
2-wire 3-wire 12...40 VDC 4...20 mA out common case ground
~0.2 in. (5 mm) dia.
~ 3.2 in. (81 mm)
MS3102E-14S-6P
contact view
MULTICONDUCTOR, 24 GA., SHIELDED
A B n/a D
A C B n/a
contact view
12...40 VDC 4...20 mA out common case ground
67 6, 4X**
WHT BLK n/a GRN
n/a n/a n/a n/a
connections
2-wire 3-wire
B
F
C
E D
67, 68* 6, 4X**
* requires factory submersion test
connections
A
2-wire 3-wire
12...40 VDC 4...20 mA out common case ground
67 4
A B n/a D
A C B n/a
2-wire 3-wire 12...40 VDC 4...20 mA out common case ground
RED BLK n/a GRN
RED BLK WHT n/a
67 6
** applies to stainless steel enclosure, see
▼ Sample Model Number
PT9420- 0200 order code:
Specifications:
R
- 1A 1 B 1 C - 1 D 1 E 1 F 0 G
Full Stroke Range: Enclosure Material: Measuring Cable: Cable Exit: Output Signals: Electrical Connection:
200 inches powder-painted aluminum 0.034-in dia. nylon coated stainless steel cable front 4-20 mA, 2-wire, output increasing with cable extension 6-pin plastic connector
7800 Deering Avenue • Canoga Park, CA • 91309 • tel: (800) 423-5483 • (818) 884-6860 • fax: (818) 340-1175 www.celesco.com •
[email protected]
171
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 3.2 Alternative cylinder length measurement The minimum and maximum cylinder lengths are estimated from the calibration photos. With each minimum or maximum length the value of the transducer signal is noted. Assuming the relation is linear, following equation is used. (Equation 9 and Figure 14)
Li =
Lmax,i − Lmin,i Vmax,i − Vmin,i
⋅ Vi + B
(9)
Where Lmax,i = Maximum length cylinder i in meters Lmin,i = Minimum length cylinder i in meters Vmax,i = Maximum voltage for cylinder i in Volts Vmin,i = Minimum voltage for cylinder I in Volts B = Li when Vi equals 0 or Intersection with Li axis in meters Vmaxi-Vmini Lmaxi-Lmini
L [m] B V [V]
Figure 14 The lineair realtion between Lenght and Voltage
The value of B can be calculated when Lmax,i, Lmin,i, Vmax,i and Vmin,i are known by filling in a known data point, for instance Lmax,i with Vmax,i.
54
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 4.1 Generation of box plots. To get the box plots a matlab script was written. (see Script 1 and cyclebreakdown.m on cd-rom) How to interpret a Box plot is in Section 4.2. To observe variation in annotation state times pie charts are generated as well. The pie plots are not analyzed in this report. Each annotation state is logged as a number. This script calculates the time that the annotation array is continuously on one number and puts the time in the corresponding array, for example the dig time-array. When the dig, swing dump and return time-arrays are available the box and pie plot were generated using commands as ‘boxplot’, ‘pie’ and ‘subplot’. The cycle times are calculated every annotation session at the end of the m-file, but not plotted in any figure. Script 1: % % % % % % % % %
Dig = 4 Swing = 5 Dump = 7 Return = 8 Wait = 6 Start = 3 Propel = 10 cleanup = 11 Idle = 12
%load DataSet.mat ts = 0.02; % time step walkcount = 0; cleanupcount = 0; digcount = 0; swingcount = 0; dumpcount = 0; retcount = 0; waitcount = 0; idlecount = 0; startcount = 0; digdis = 0; dumpdis = 0; retdis = 0; swingdis = 0; walktime = []; cleanuptime = []; digtime = []; swingtime = []; dumptime = []; rettime = []; waittime = []; idletime = []; starttime = []; order = []; state = 1; j = 1;
55
Performance evaluation of a Liebherr 996 hydraulic shovel
for k=(1:size(signals.AnnoState')-1) if signals.AnnoState(k) == 10 state = 10; walkcount = walkcount + 1; end if signals.AnnoState(k) == 11 state = 11; cleanupcount = cleanupcount + 1; end if signals.AnnoState(k) == 4 state = 4; digcount = digcount + 1; end if signals.AnnoState(k) == 5 state = 5; swingcount = swingcount + 1; end if signals.AnnoState(k) == 7 state = 7; dumpcount = dumpcount + 1; end if signals.AnnoState(k) == 8 state = 8; retcount = retcount + 1; end if signals.AnnoState(k) == 6 state = 6; waitcount = waitcount + 1; end if signals.AnnoState(k) == 12 state = 12; idlecount = idlecount + 1; end if signals.AnnoState(k) == 3 state = 3; startcount = startcount + 1; end %---------------------------------------------------------------------if signals.AnnoState(k) ~= signals.AnnoState(k+1) if signals.AnnoState(k) == 4 && signals.AnnoState(k) == 11 % if the current annotation is a dig and the next annotation is a cleanup cleanupcount = digcount; digcount = 0; state = 11; else if state == 10 %walk walktime = [walktime walkcount*ts]; order(j,:) = [walkcount*ts state]; j = j + 1; walkcount = 0; end if state == 11 %cleanup cleanuptime = [cleanuptime cleanupcount*ts]; order(j,:) = [cleanupcount*ts state];
56
Performance evaluation of a Liebherr 996 hydraulic shovel j = j + 1; cleanupcount = 0; end if state == 4 %dig if digcount*ts > 3 && digcount*ts < 35 digtime = [digtime digcount*ts]; order(j,:) = [digcount*ts state]; j = j + 1; digcount = 0; else digdis = digdis + 1; digcount = 0; end end if state == 5 %swing if swingcount*ts > 2 && swingcount*ts < 15 swingtime = [swingtime swingcount*ts]; order(j,:) = [swingcount*ts state]; j = j + 1; swingcount = 0; else swingdis = swingdis + 1; swingcount = 0; end end if state == 7 %dump if dumpcount*ts > 1 && dumpcount*ts < 9 dumptime = [dumptime dumpcount*ts]; order(j,:) = [dumpcount*ts state]; j = j + 1; dumpcount = 0; else dumpdis = dumpdis + 1; dumpcount = 0; end end if state == 8 %return if retcount*ts > 2 && retcount*ts < 30 rettime = [rettime retcount*ts]; order(j,:) = [retcount*ts state]; j = j + 1; retcount = 0; else retdis = retdis + 1; retcount = 0; end end if state == 6 %wait waittime = [waittime waitcount*ts]; order(j,:) = [waitcount*ts state]; j = j + 1; waitcount = 0; end if state == 12 %idle idletime = [idletime idlecount*ts]; order(j,:) = [idlecount*ts state]; j = j + 1; idlecount = 0; end if state == 3 %start starttime = [starttime startcount*ts]; order(j,:) = [startcount*ts state]; j = j + 1; startcount = 0; end end end
57
Performance evaluation of a Liebherr 996 hydraulic shovel end dumptot = sum(dumptime); digtot = sum(digtime); swingtot = sum(swingtime); rettot = sum(rettime); starttot = sum(starttime); idletot = sum(idletime); walktot = sum(walktime); cleanuptot = sum(cleanuptime); waittot = sum(waittime); %-------create at pie chart for the total time breakdown-----------h = pie([digtot swingtot dumptot rettot idletot waittot cleanuptot walktot]); textObjs = findobj(h,'Type','text'); oldStr = get(textObjs,{'String'}); val = get(textObjs,{'Extent'}); oldExt = cat(1,val{:}); Names = {'Digging: ';'Swinging: ';'Dumping: ';'Returning: ';'Idle: ';'Waiting: ';'Cleaning up: ';'Walking: '}; newStr = strcat(Names,oldStr); set(textObjs,{'String'},newStr); %-------lets create box plots for the dig, swing, dump and return states------figure subplot(1,4,1) boxplot([digtime'],1); title('Dig'); axis([0.5 1.5 0 30]); grid; subplot(1,4,2) boxplot([swingtime'],1); title('Swing'); axis([0.5 1.5 0 30]); grid; subplot(1,4,3) boxplot([ dumptime'],1); title('Dump'); axis([0.5 1.5 0 30]); grid; subplot(1,4,4) boxplot([rettime'],1); title('Return'); axis([0.5 1.5 0 30]); grid; clear walkcount, cleanupcount ,digcount, swingcount, dumpcount, retcount, waitcount, idlecount, startcount; % % % % %
Dig = 4 Swing = 5 Dump = 7 Return = 8 Wait = 6
m=1; for p=(1:size(order)-3) if order(p,2) == 4 && order(p+1,2) == 5 && order(p+2,2) == 7 && order(p+3,2) == 8 cycles(m,1) = [order(p,1)+order((p+1),1)+order((p+2),1)+order((p+3),1)]; m=m+1; end end
58
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 4.2 Box plot interpretation The box plot is a graphical representation of data that shows statistical information. The box plot is helpful in evaluating small data sets that are not suitable for histograms. Because of the small size of a box plot, it is easy to display and compare several box plots in a small space [4] The graph shows an example of a notched box plot.
This plot has several graphic elements: The lower and upper lines of the "box" (Q1 and Q3) are the 25th and 75th percentiles of the sample. The distance between the top and bottom of the box is the interquartile range and 50% of all data is in that range. The line in the middle of the box (Q1) is the sample median. If the median is not centered in the box, that is an indication of asymmetry of the distribution. The "whiskers" are lines extending above and below the box. They show the extent of the rest of the sample (unless there are outliers). By default, an outlier is a value that is more than 1.5 times the interquartile range away from the top or bottom of the box. The dots or plus signs at the top of the plot are indications of outliers in the data. These points may be the result of a data entry error, a poor measurement or a change in the system that generated the data. The notches in the box are a graphic confidence interval about the median of a sample [2]
59
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 5.1 Trajectories and Boxplots Stage 1
60
Performance evaluation of a Liebherr 996 hydraulic shovel
61
Performance evaluation of a Liebherr 996 hydraulic shovel
62
Performance evaluation of a Liebherr 996 hydraulic shovel
63
Performance evaluation of a Liebherr 996 hydraulic shovel
64
Performance evaluation of a Liebherr 996 hydraulic shovel
65
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 5.2 Trajectories and Boxplots Stage 2
66
Performance evaluation of a Liebherr 996 hydraulic shovel
67
Performance evaluation of a Liebherr 996 hydraulic shovel
68
Performance evaluation of a Liebherr 996 hydraulic shovel
69
Performance evaluation of a Liebherr 996 hydraulic shovel
70
Performance evaluation of a Liebherr 996 hydraulic shovel
71
Performance evaluation of a Liebherr 996 hydraulic shovel
72
Performance evaluation of a Liebherr 996 hydraulic shovel
73
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 6 Matlab scripts for testing hypothesis Appendix 6.1 Dig reach vs bench height script load cycledatastage2; load cycledatastage1; maxBXstage1; maxBXstage2; h20_1_maxBX = [Nov11_anno1_johnny_20.maxBX; Nov11_anno2_johnny_20.maxBX; Nov18_anno1_pimmy_20.maxBX; Nov18_anno2_pimmy_20.maxBX; Nov19_anno1_jimmy_20.maxBX; Nov19_anno2_pimmy_20.maxBX; Nov20_anno1_kazzy_20.maxBX; Nov20_anno2_kazzy_20.maxBX]; h10_1_maxBX = [Nov12_anno1_paul_10.maxBX; Nov12_anno2_paul_10.maxBX; Nov13_anno1_paul_10.maxBX; Nov13_anno2_clive_10.maxBX]; h15_1_maxBX = [Nov14_anno1_paul_15.maxBX; Nov14_anno2_paul_15.maxBX; Nov15_anno2_mutt_15.maxBX; Nov16_anno1_tinky_15.maxBX; Nov16_anno2_mutt_15.maxBX]; h20_2_maxBX = [Dec11_anno1_charlie_20.maxBX; Dec11_anno2_charlie_20.maxBX; Dec12_anno1_graham_20.maxBX; Dec13_anno1_graham_20.maxBX]; h15_2_maxBX = [Dec14_anno1_graham_15.maxBX; Dec14_anno2_graham_15.maxBX; Dec2_anno1_troy_15.maxBX; Dec2_anno2_troy_15.maxBX; Dec3_anno1_troy_15.maxBX; Dec3_anno2_troy_15.maxBX; Dec5_anno2_troy_15.maxBX]; h10_2_maxBX = [Dec6_anno1_troy_10.maxBX; Dec6_anno2_troy_10.maxBX; Dec7_anno1_troy_10.maxBX; Dec7_anno2_troy_10.maxBX; Dec8_anno1_charlie_10.maxBX; Dec8_anno2_charlie_10.maxBX]; h10_maxBX = [h10_1_maxBX; h10_2_maxBX]; h15_maxBX = [h15_1_maxBX; h15_2_maxBX]; h20_maxBX = [h20_1_maxBX; h20_2_maxBX]; group_maxBX = [ones(size(h20_maxBX),1); 2*ones(size(h15_maxBX),1); 3*ones(size(h10_maxBX),1)]; data_maxBX = [h20_maxBX; h15_maxBX; h10_maxBX]; [p_maxBX, tb_maxBX, stats_maxBX] = anova1(data_maxBX',group_maxBX'); figure [c_maxBX m_maxBX] = multcompare(stats_maxBX);
74
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 6.2 Cycle time vs bench height script load cycledatastage2; load cycledatastage1; h20_2_cycles = [Dec11_anno1_charlie_20.cycles; Dec11_anno2_charlie_20.cycles; Dec12_anno1_graham_20.cycles; Dec13_anno1_graham_20.cycles]; h15_2_cycles = [Dec14_anno1_graham_15.cycles; Dec14_anno2_graham_15.cycles; Dec2_anno1_troy_15.cycles; Dec2_anno2_troy_15.cycles; Dec3_anno1_troy_15.cycles; Dec3_anno2_troy_15.cycles; Dec5_anno2_troy_15.cycles]; h10_2_cycles = [Dec6_anno1_troy_10.cycles; Dec6_anno2_troy_10.cycles; Dec7_anno1_troy_10.cycles; Dec7_anno2_troy_10.cycles; Dec8_anno1_charlie_10.cycles; Dec8_anno2_charlie_10.cycles]; h20_1_cycles = [Nov11_anno1_johnny_20.cycles; Nov11_anno2_johnny_20.cycles; Nov18_anno1_pimmy_20.cycles; Nov18_anno2_pimmy_20.cycles; Nov19_anno1_jimmy_20.cycles; Nov19_anno2_pimmy_20.cycles; Nov20_anno1_kazzy_20.cycles; Nov20_anno2_kazzy_20.cycles]; h10_1_cycles = [Nov12_anno1_paul_10.cycles; Nov12_anno2_paul_10.cycles; Nov13_anno1_paul_10.cycles; Nov13_anno2_clive_10.cycles]; h15_1_cycles = [Nov14_anno1_paul_15.cycles; Nov14_anno2_paul_15.cycles Nov15_anno2_mutt_15.cycles Nov16_anno1_tinky_15.cycles Nov16_anno2_mutt_15.cycles]; h10_cycles = [h10_1_cycles; h10_2_cycles]; h15_cycles = [h15_1_cycles; h15_2_cycles]; h20_cycles = [h20_1_cycles; h20_2_cycles]; data_cycles = [h20_cycles; h15_cycles; h10_cycles]; group_cycles = [ones(size(h20_cycles),1); 2*ones(size(h15_cycles),1); 3*ones(size(h10_cycles),1)]; [p_cycles, tb_cycles, stats_cycles] = anova1(data_cycles',group_cycles'); figure [c_cycles m_cycles] = multcompare(stats_cycles);
75
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 6.3 Cycle times vs digging conditions script load cycledatastage2; load cycledatastage1; %-------------------------------------------------------------------------Dec11_anno1_charlie_20.digcond = 4*ones(size(Dec11_anno1_charlie_20.dig)); Dec11_anno2_charlie_20.digcond = 4*ones(size(Dec11_anno2_charlie_20.dig)); Dec12_anno1_graham_20.digcond = 3*ones(size(Dec12_anno1_graham_20.dig)); Dec13_anno1_graham_20.digcond = 3*ones(size(Dec13_anno1_graham_20.dig)); Dec14_anno1_graham_15.digcond = 3*ones(size(Dec14_anno1_graham_15.dig)); Dec14_anno2_graham_15.digcond = 3*ones(size(Dec14_anno2_graham_15.dig)); Dec2_anno1_troy_15.digcond = 3*ones(size(Dec2_anno1_troy_15.dig)); Dec2_anno2_troy_15.digcond = 3*ones(size(Dec2_anno2_troy_15.dig)); Dec3_anno1_troy_15.digcond = 3*ones(size(Dec3_anno1_troy_15.dig)); Dec3_anno2_troy_15.digcond = 3*ones(size(Dec3_anno2_troy_15.dig)); Dec5_anno2_troy_15.digcond = 3*ones(size(Dec5_anno2_troy_15.dig)); Dec6_anno1_troy_10.digcond = 3*ones(size(Dec6_anno1_troy_10.dig)); Dec6_anno2_troy_10.digcond = 3*ones(size(Dec6_anno2_troy_10.dig)); Dec7_anno1_troy_10.digcond = 3*ones(size(Dec7_anno1_troy_10.dig)); Dec7_anno2_troy_10.digcond = 3*ones(size(Dec7_anno2_troy_10.dig)); Dec8_anno1_charlie_10.digcond = 3*ones(size(Dec8_anno1_charlie_10.dig)); Dec8_anno2_charlie_10.digcond = 3*ones(size(Dec8_anno2_charlie_10.dig)); %-------------------------------------------------------------------------Nov11_anno1_johnny_20.digcond = 4*ones(size(Nov11_anno1_johnny_20.dig)); Nov11_anno2_johnny_20.digcond = 4*ones(size(Nov11_anno2_johnny_20.dig)); Nov12_anno1_paul_10.digcond = 3*ones(size(Nov12_anno1_paul_10.dig)); Nov12_anno2_paul_10.digcond = 3*ones(size(Nov12_anno2_paul_10.dig)); Nov13_anno1_paul_10.digcond = ones(size(Nov13_anno1_paul_10.dig)); Nov13_anno2_clive_10.digcond = ones(size(Nov13_anno2_clive_10.dig)); Nov14_anno1_paul_15.digcond = ones(size(Nov13_anno2_clive_10.dig)); Nov14_anno2_paul_15.digcond = ones(size(Nov14_anno2_paul_15.dig)); Nov15_anno2_mutt_15.digcond = ones(size(Nov15_anno2_mutt_15.dig)); Nov16_anno1_tinky_15.digcond = ones(size(Nov16_anno1_tinky_15.dig)); Nov16_anno2_mutt_15.digcond = ones(size(Nov16_anno2_mutt_15.dig)); Nov18_anno1_pimmy_20.digcond = 4*ones(size(Nov18_anno1_pimmy_20.dig)); Nov18_anno2_pimmy_20.digcond = 4*ones(size(Nov18_anno2_pimmy_20.dig)); Nov19_anno1_jimmy_20.digcond = 4*ones(size(Nov19_anno1_jimmy_20.dig)); Nov19_anno2_pimmy_20.digcond = 4*ones(size(Nov19_anno2_pimmy_20.dig)); Nov20_anno1_kazzy_20.digcond = 4*ones(size(Nov20_anno1_kazzy_20.dig)); Nov20_anno2_kazzy_20.digcond = 4*ones(size(Nov20_anno2_kazzy_20.dig)); %-------------------------------------------------------------------------cycles_1 = [Nov11_anno1_johnny_20.cycles; Nov11_anno2_johnny_20.cycles; Nov18_anno1_pimmy_20.cycles; Nov18_anno2_pimmy_20.cycles; Nov19_anno1_jimmy_20.cycles; Nov19_anno2_pimmy_20.cycles; Nov20_anno1_kazzy_20.cycles; Nov20_anno2_kazzy_20.cycles; Nov12_anno1_paul_10.cycles; Nov12_anno2_paul_10.cycles; Nov13_anno1_paul_10.cycles; Nov13_anno2_clive_10.cycles; Nov14_anno1_paul_15.cycles; Nov14_anno2_paul_15.cycles Nov15_anno2_mutt_15.cycles Nov16_anno1_tinky_15.cycles Nov16_anno2_mutt_15.cycles];
76
Performance evaluation of a Liebherr 996 hydraulic shovel
cycles_2 = [Dec11_anno1_charlie_20.cycles; Dec11_anno2_charlie_20.cycles; Dec12_anno1_graham_20.cycles; Dec13_anno1_graham_20.cycles; Dec14_anno1_graham_15.cycles; Dec14_anno2_graham_15.cycles; Dec2_anno1_troy_15.cycles; Dec2_anno2_troy_15.cycles; Dec3_anno1_troy_15.cycles; Dec3_anno2_troy_15.cycles; Dec5_anno2_troy_15.cycles; Dec6_anno1_troy_10.cycles; Dec6_anno2_troy_10.cycles; Dec7_anno1_troy_10.cycles; Dec7_anno2_troy_10.cycles; Dec8_anno1_charlie_10.cycles; Dec8_anno2_charlie_10.cycles]; cycletime = [cycles_1; cycles_2]; digcond_1 = [Nov11_anno1_johnny_20.digcond; Nov11_anno2_johnny_20.digcond; Nov18_anno1_pimmy_20.digcond; Nov18_anno2_pimmy_20.digcond; Nov19_anno1_jimmy_20.digcond; Nov19_anno2_pimmy_20.digcond; Nov20_anno1_kazzy_20.digcond; Nov20_anno2_kazzy_20.digcond; Nov12_anno1_paul_10.digcond; Nov12_anno2_paul_10.digcond; Nov13_anno1_paul_10.digcond; Nov13_anno2_clive_10.digcond; Nov14_anno1_paul_15.digcond; Nov14_anno2_paul_15.digcond; Nov15_anno2_mutt_15.digcond; Nov16_anno1_tinky_15.digcond; Nov16_anno2_mutt_15.digcond]; digcond_2 = [Dec11_anno1_charlie_20.digcond; Dec11_anno2_charlie_20.digcond; Dec12_anno1_graham_20.digcond; Dec13_anno1_graham_20.digcond; Dec14_anno1_graham_15.digcond; Dec14_anno2_graham_15.digcond; Dec2_anno1_troy_15.digcond; Dec2_anno2_troy_15.digcond; Dec3_anno1_troy_15.digcond; Dec3_anno2_troy_15.digcond; Dec5_anno2_troy_15.digcond; Dec6_anno1_troy_10.digcond; Dec6_anno2_troy_10.digcond; Dec7_anno1_troy_10.digcond; Dec7_anno2_troy_10.digcond; Dec8_anno1_charlie_10.digcond; Dec8_anno2_charlie_10.digcond]; digcond = [digcond_1; digcond_2]; %-------------------------------------------------------------------------[p_cycles, tb_cycles, stats_cycles] = anova1(cycletime',digcond'); figure [c_cycles m_cycles] = multcompare(stats_cycles); %--------------------------------------------------------------------------
77
Performance evaluation of a Liebherr 996 hydraulic shovel
Appendix 6.4 Bucket load vs digging conditions script load cycledatastage2; load cycledatastage1; payloadstage1; payloadsstage2; payload_1 = [Nov11_anno1_johnny_20.payload; Nov11_anno2_johnny_20.payload; Nov18_anno1_pimmy_20.payload; Nov18_anno2_pimmy_20.payload; Nov19_anno1_jimmy_20.payload; Nov19_anno2_pimmy_20.payload; Nov20_anno1_kazzy_20.payload; Nov20_anno2_kazzy_20.payload; Nov12_anno1_paul_10.payload; Nov12_anno2_paul_10.payload; Nov13_anno1_paul_10.payload; Nov13_anno2_clive_10.payload; Nov14_anno1_paul_15.payload; Nov14_anno2_paul_15.payload; Nov15_anno2_mutt_15.payload; Nov16_anno1_tinky_15.payload; Nov16_anno2_mutt_15.payload]; payload_2 = [Dec11_anno1_charlie_20.payload; Dec11_anno2_charlie_20.payload; Dec12_anno1_graham_20.payload; Dec13_anno1_graham_20.payload; Dec14_anno1_graham_15.payload; Dec14_anno2_graham_15.payload; Dec2_anno1_troy_15.payload; Dec2_anno2_troy_15.payload; Dec3_anno1_troy_15.payload; Dec3_anno2_troy_15.payload; Dec5_anno2_troy_15.payload; Dec6_anno1_troy_10.payload; Dec6_anno2_troy_10.payload; Dec7_anno1_troy_10.payload; Dec7_anno2_troy_10.payload; Dec8_anno1_charlie_10.payload; Dec8_anno2_charlie_10.payload]; payload = [payload_1; payload_2]; pay_digcond_1 = [4*ones(size(Nov11_anno1_johnny_20.payload)); 4*ones(size(Nov11_anno2_johnny_20.payload)); 4*ones(size(Nov18_anno1_pimmy_20.payload)); 4*ones(size(Nov18_anno2_pimmy_20.payload)); 4*ones(size(Nov19_anno1_jimmy_20.payload)); 4*ones(size(Nov19_anno2_pimmy_20.payload)); 4*ones(size(Nov20_anno1_kazzy_20.payload)); 4*ones(size(Nov20_anno2_kazzy_20.payload)); 3*ones(size(Nov12_anno1_paul_10.payload)); 3*ones(size(Nov12_anno2_paul_10.payload)); ones(size(Nov13_anno1_paul_10.payload)); ones(size(Nov13_anno2_clive_10.payload)); ones(size(Nov14_anno1_paul_15.payload)); ones(size(Nov14_anno2_paul_15.payload)); ones(size(Nov15_anno2_mutt_15.payload)); ones(size(Nov16_anno1_tinky_15.payload)); ones(size(Nov16_anno2_mutt_15.payload))];
78
Performance evaluation of a Liebherr 996 hydraulic shovel
pay_digcond_2 = [4*ones(size(Dec11_anno1_charlie_20.payload)); 4*ones(size(Dec11_anno2_charlie_20.payload)); 3*ones(size(Dec12_anno1_graham_20.payload)); 3*ones(size(Dec13_anno1_graham_20.payload)); 3*ones(size(Dec14_anno1_graham_15.payload)); 3*ones(size(Dec14_anno2_graham_15.payload)); 3*ones(size(Dec2_anno1_troy_15.payload)); 3*ones(size(Dec2_anno2_troy_15.payload)); 3*ones(size(Dec3_anno1_troy_15.payload)); 3*ones(size(Dec3_anno2_troy_15.payload)); 3*ones(size(Dec5_anno2_troy_15.payload)); 3*ones(size(Dec6_anno1_troy_10.payload)); 3*ones(size(Dec6_anno2_troy_10.payload)); 3*ones(size(Dec7_anno1_troy_10.payload)); 3*ones(size(Dec7_anno2_troy_10.payload)); 3*ones(size(Dec8_anno1_charlie_10.payload)); 3*ones(size(Dec8_anno2_charlie_10.payload))]; pay_digcond = [pay_digcond_1; pay_digcond_2]; %-------------------------------------------------------------------------[p_payload, tb_payload, stats_payload] = anova1(payload',pay_digcond'); figure [c_payload m_payload] = multcompare(stats_payload); %--------------------------------------------------------------------------
79
