MASTER'S THESISSurface Drill Rig Simulation

(1)

2009:149 CIV

M A S T E R ' S T H E S I S

Surface Drill Rig Simulation

Johan Theander

Peter Holmström

Luleå University of Technology MSc Programmes in Engineering

Mechanical Engineering

Department of Applied Physics and Mechanical Engineering Division of Functional Product Development

(2)

Preface

This thesis is the final part of the Master of Science program in Mechanical Engineering at Luleå University of Technology. The project was performed at Atlas Copco AB between September 2008 and February 2009.

We would like to thank Atlas Copco for the opportunity to do our master thesis at their R&D department. It has been a very meaningful and instructive time for us. We would like give a special thank to Morgan Norling at Roctech for his help and time and his endless knowledge and understanding of MSC Adams. We would also like to thank our supervisors at Atlas Copco Mikael Johnson, Patrick Kenger and Göran Tuomas for their good advices and directives. We also thank our supervisor Mikael Nybacka and examiner Tobias Larsson at Luleå University of Technology for their shown interest and help with all the administrative work, without you this thesis work wouldn’t have been possible. We are also very grateful for the support and

assistance from all employees at Altas Copco.

On a personal level, we would like to express gratitude to our friends and families who have supported us during our work.

In Örebro February 2009

(3)

2

Abstract

This master thesis work where performed at Atlas Copco AB division SDE techniques in Örebro. SDE techniques designs and manufactures surface drilling equipment for a demanding worldwide market.

The constant ambition to reduce the development cost and time demands new and improved dimensioning tools. ADAMS is a simulation software that can be used to find the dynamic loads on a vehicle early in the design phase. This thesis work is a start for a new method of dimensioning boomsystems and other machine parts for surface drill rigs.

The surface drill rig was built in ADAMS with moment of inertia and mass from CAD models. The boomsystem hydraulics was modeled with a function describing the oil compressibility and the flexibility of the cylinder walls. The complete model was compared with measurements on a surface drill rig during tramming (propulsion of the rig).

(4)

3

Nomenclature

BLC – Boom lift cylinder BSC – Boom swing cylinder FDC – Feed dump cylinder FEC – Feed extension cylinder FSC – Feed swing cylinder

SDE – Surface drilling Excavation

ADAMS – Automatic Dynamic Analysis of Mechanical Systems FEA - Finite Element Analyses

FEM - Finite Element Method MBS - Multi Bodies System DTH - Down The Hole

Tramming - Propulsion of the drill rig CAD - Computed Aided Design Mnf – Modal Neutral File

(5)

4

1. Introduction

This chapter will give the reader an introduction to the thesis work. Background, purpose, objectives and limitations of the thesis project will be discussed. At the end of this chapter follows a presentation of this thesis structure.

This thesis work was performed at the Atlas Copco Surface Drilling Equipment (SDE) division. Surface Drilling Equipment is a division inside the business area construction and mining technique. The main manufacturing plant is located in Örebro and has 340 employees.

Manufacturing is also located in Garland Texas, Yokohama in Japan, Nanjing and Zhangjiakou china and Nasik in India [1].

Increasing the machine reliability demands a better understanding of the forces acting on surface drill rigs. Improving the dimensioning tools can do this. Mines and quarries is a very rough environment and presents a challenge for machinery, the rough surface induces

vibrations and shocks that cause fatigue in machine parts. By developing more precise methods the number of expensive breakdowns and the need for on the location repairs can decrease. Underdimensioned parts also lead to that Atlas Copco must replace the part in the field to great expense. Decreased maintenance will lead to increased competiveness on the demanding worldwide market of mining machinery.

(8)

7

1.1 Background

This master thesis was initiated by Mikael Johnson at Atlas Copco AB division SDE techniques in Örebro, to improve their methods for dimensioning the boom on surface drill rigs from here on called rigs. The methods today are foremost FE-analysis and hand calculations. The input for the loads has been obtained from field tests, rig geometry, weight and assumptions. To improve the precision in dimensioning the rig, and to avoid fatigue and fracture problems, methods for stress and strain calculations need to be continuously improved. The thesis focuses on surface drill rig L8, but methods should be the same on similar machinery. Simulations’ using dynamic software ADAMS has become a good option for evaluating designs before the vehicle is actually built and will be used in the project.

1.2 Purpose

The purpose of this thesis project is to describe how ADAMS can be used to simulate a L8 surface drill rig during tramming. The mission is to find the loads in the boomsystem that causes fatigue. This thesis is a start to a new method to dimensioning booms and other machine parts for surface drill rigs, from here on called rig.

1.3 Objectives

The main objective is to examine the possibility to simulate loads on rigs with ADAMS and possibly Easy5. The model should be the beginning of a dimensioning tool for rigs.

The secondary objective is to reduce the loads by using accumulators or active damping. Limitations:

Since the timeframe of this master thesis is twenty weeks some limitations are introduced:

(9)

8

2. About ADAMS

This chapter explains how ADAMS works and what is needed to make a model. Knowing how ADAMS calculates and what information is needed can be vital to avoid problems in the thesis project.

The ADAMS software simulates the static and dynamic behavior of a Multi Body system (MBS). A Multi Body System consists of a collection of rigid and/or flexible bodies that are connected to each other by kinematic joints and/or force elements [5]. Each body have six degrees of freedom, three translational and three rotational. Rigid bodies cannot deform. Flexible bodies deform due to loads and are prepared in a FEA-program like Marc or Ansys and then imported to ADAMS, with an add-on module to ADAMS, ADAMS/Flex, one can directly transform rigid parts into flexible parts within ADAMS.

A kinematic joint is a “perfect” connection that removes 1 to 6 degrees of freedom between two bodies, e.g. a spherical joint removes 3 translational degrees of freedom [4].

Forces can be used to describe arbitrary functionality (see Figure 1 for an example of a simple MBS). A force can as an example describe a spring and the function value is given mainly by length and velocity. A force can also describe an external force by a function or measured data. A force can also describe a connection like a hinge with gap and friction.

After a simulation results can be investigated by plots and animations. Positions, velocities, accelerations for all bodies and forces and torques in all joints are calculated by default.

Figure 1. Example of a simple MBS.

(10)

9 users can often prepare their own user-written subroutines in languages such as FORTRAN or ANSI ‘C’.

Figure 2. Overview of ADAMS.

For each rigid body in the system it is necessary to define the mass, centre of mass location and mass moments of inertia. Each body will possess a set of coordinates and are considered to move with the part during simulation. These points are used to define centre of mass locations, joint locations and orientations, force location and directions. The relative motion between parts in the system is constrained by using joints, joint primitives, couplers, gears and user-defined constraints. The next step is to define the external and internal forces. These forces can be defined as constant, time dependent, or functionally dependent. They can also be

translational or rotational and be defined to “follow” the part during simulation.

Users can also set up internal forces acting between two parts like springs, dampers, cables or rubber mounts. The elements can also be defined to act in only tension or compression and may be linear or non-linear.

(11)

10

3. Surface drill rig – ROC L8 LF

This chapter gives the reader an understanding of the machinery and nomenclature used in surface drill rigs.

Simulations are made on a ROC L8 long feed, see Figure 3. The L8 is intended for open pit mining. The machine consists of three main parts, carrier, boomsystem and feed. It has a single boomsystem and is equipped with a down-the-hole hammer (DTH) drilling system.

Figure 3. Surface drill rig L8 with C20mk2 feed.

(12)

11

Figure 4. The boomsystem C20mk2

Table 1. The boomsystem and feed nomenclature

4. Part Abbreviation in ADAMS model and text A Wagon frame alt carrier CAR

B Boom attachment BAT

C Boom BOOM

D Boom head BOH

E Feed holder FEH

F Feed FEED

G Rock drill RDR

H Boom swing cylinder BSC I Boom lift cylinder BLC

J Right Feed dump cylinder FDC_RIGHT K Left Feed dump cylinder FDC_LEFT L Feed swing cylinder FSC

(13)

12

5. Mechanical model - ADAMS

This chapter explains how the model was build up and how the surface drill rig is modeled. Flexible elements in ADAMS are discussed and also the limits of the model built.

To analyze the behavior of the boomsystem on a surface drill rig during tramming an ADAMS model was built see Figure 5.

Figure 5. ADAMS base model of surface drill rig.

The boom, boom head, boom attachment, feed and feed holder are built in Pro Engineer and imported to ADAMS/View as parasolid files. This makes it easier to make the model more realistic looking and it is also handy to find the right connection places between the various parts. The parasolid files describe the volume that ADAMS uses to calculate the moment of inertia and centre of gravity, (see appendix G for details translating CAD geometry to ADAMS models). They were also used as inspiration when modeling the appearance in ADAMS. The parts were then connected with joints. The hydraulic cylinders where built in ADAMS/View and added to the model.

5.1 Rigid bodies

(14)

13

5.2 Joints

The parts in the mechanical model are connected with different types of joints see Figure 6. The joints in the model do not have to be the same as in the real machine, but they give the same functionality. One reason for example is having one constrain in one point instead of having two constrains in two different points that’s gives the same function is to shorten calculating time and to avoid lockups and crashes in the model.

(15)

14

5.3 Hydraulic cylinders

The stiffness and the movement of the hydraulic cylinders are modeled with single component forces, stretching from the hydraulic cylinders hook joints see Figure 7. Friction and damping are also modeled using a single force see appendix H and I.

(16)

15

6. Hydraulics

This chapter explains the basics of hydraulic systems and formulas used when calculating forces in the boomsystem. The compressibility of hydraulic fluids will also be investigated.

The L8 surface drill rig boomsystem is steered with hydraulics, for schematics see appendix F. In a hydraulic system of the type hydrostatic system is the energy transfer of displacement type. The fluid is pressurized by a displacement pump lead by valves into a hydraulic cylinder see Figure 8.

Figure 8. Schematic hydraulic cylinder.

The force created by the hydraulic cylinders according to Linköpings Universitet IEI (2005) is

2 2 1 1 A P A P F = ⋅ − ⋅ _. ₍₁₎

Where A1 are the area of the piston and A2 the area of the piston minus the rod. The

boomsystem is steered with six hydraulic cylinders.

6.1 Counterbalance valve

The purpose of the counterbalance valve in the boomsystem is preventing the boomsystem to free-fall when the directional valve is shifting direction or in case of breakage of the hoses. Another aspect is to prevent damage on the hydraulic system when the boomsystem is exposed for overrunning loads.

As the load extends, internal pilot supply gives smooth control with little energy loss [8]. After work contact, as system pressure builds, the external pilot fully opens the counterbalance to relieve all backpressure in the cylinder. Figure 9 shows the schematics of internally/externally piloted counterbalance valve whish mounted on the BLC and FDC cylinders. The

counterbalance valves are mounted directly on the inlet on the cylinders, and have max pressure of 250 bar before valves open backwards and even out the pressure between the chambers in the cylinders.

Due to this it is not necessary to consider the expansion of the hoses and the compress ability of the fluid in the hoses in the hydraulic system under the validation of the model.

2

P

(17)

16

Figure 9. The schematics for the counterbalance valve for BLC.

6.2 Compressibility of hydraulic fluids

Transient pressure variations in hydraulic systems can often be referred to oil compressibility. Pressure changes occur instead of motion. The compressibility κ is defined as the change in volume per unit for a given pressure change. The compression module K is defined as the inverted value of the compressibility. The compression module of a hydraulic fluid is defined as

V V p K / ∆ ∆ = . (2)

The value of the K module is dependent on the compression process. Under constant

temperature the compression is isothermal. With no heat exchange with the surroundings and without friction the compression is isentropic. For fluids the isentropic and isothermal

compression is approximately the same. Only with exact calculations consideration of both isentropic and isothermal compression is required.

6.3 Hydraulic cylinder Stiffness

It is assumed that there is no leakage between the two cylinders. Leakage would cause a damping effect but this is neglected. The stiffness is only dependent on the elasticity of the cylinder walls and the compression module of the hydraulic fluid. The volume for a cylinder is

l b

V = ⋅ 2⋅

π ₍₃₎

where b is the radius and l is length of cylinder and will increase with

(18)

17 The formula for cylinder volume (3) Inserted in volume change (4) gives the rate of change of the volume l l b b V V ∆ + ∆ = ∆ 2 . (5)

The hydraulic cylinders in the ROC L8 have hydraulic locks. The hydraulic cylinders therefore have constant pressure when the boomsystem stands still. The force acting on the cylinders due to feeder movement are shown in Figure 10.

Figure 10. Entities and forces acting on a hydraulic cylinder

According to Roark 2007 [9] the elasticity of the material gives the radius change of the hydraulic cylinder with the Poisson’s ratio =

ν

2 2 2 2 ) 2 1 ( ) 1 ( b a b a E b P b − − + + ⋅ ⋅ ∆ = ∆

ν

(6) And the length with

2 2 2 ) 1 ( b a b E l P L + − ⋅ ⋅ ⋅ ∆ = ∆

ν

. (7)

The change of diameter (6) and length (7) inserted in the formula for change of volume (5) gives the complete formula for the expansion of the hydraulic cylinder due to pressure increase

(

2 (1 ) 3 (1 2 )

)

) ( 2 2 2 2 +ν + − ν − ∆ = ∆ b a b a E P V V . (8)

The hydraulic fluid will also be compressed with

β

P V V = ⋅∆ ∆ . (9)

Where β is the compressibility constant for hydraulic fluid.

(19)

18       − − + + + ⋅ ⋅ ∆ = ∆ ) ( ) 2 1 ( 3 ) 1 ( 2 1 2 2 2 2 b a E b a V P V

ν

β

(10) and the total length change is

      − − + + + ⋅ ⋅ ∆ = ∆ ) ( ) 2 1 ( 3 ) 1 ( 2 1 2 2 2 2 b a E b a l P l

ν

β

. (11)

The force change is

2 2 1 1 P A P A F = ⋅∆ − ⋅∆ ∆ _. ₍₁₂₎

The formula for volume change (10) and length (11) inserted to force change (12) with ∆l1=∆l2= ∆X and after rewrite gives the stiffness of a hydraulic cylinder

1 2 2 1 2 2 2 2 2 1 ) ( ) 2 1 ( 3 ) 1 ( 2 1 l A l A b a E b a l l F X +       − − + + + ⋅ = ∆ ∆ ν ν β (13) which gives the equation

K F X 1 = ∆ ∆

, where K is the total spring constant for the cylinder walls and the fluid.

(20)

19

7. Flexible parts

This chapter describes the approach for creation of flexible parts in ADAMS with original CAD geometry and describes the needs for flexible bodies in the model.

A completely stiff boom model would miss-represent the boomsystem response to applied forces. The difference between a rigid and a flexible body is damping in the flexible part and deflection. The flexible parts react more realistic than rigid. The use of large flexible bodies in ADAMS increases simulation time, which limits the number of bodies used in models.

ADAMS/Flex allows flexible components to be included into ADAMS model to achieve more realistic simulation results. The approach to obtain a flexible part to ADAMS is shown in Figure 11.

Figure 11. Approach to obtain ADAMS flex model.

7.1 Boom

The boom is represented with a FE model to get a flexible structure. The boom is meshed in UGS NX 5 with 3D element tetra 10. ADAMS needs an interface point to create joints between the FEM parts and the rigid parts in ADAMS. The connections are created from one connection surface to one point with rigid elements (RBE2). Figure 12 shows how the rigid element is connected between the interface surfaces and interface point.

The FEA-software MSC Marcis capable of creating a Modal Neutral File (mnf), which can be integrated into the ADAMS model as a flexible body. No more analysis takes place in Marc after generating the mnf. Generating an mnf from Marc is based on performing the most general method of Component Mode Synthesis (CMS) techniques, named the Craig-Bampton method. To simplify and shorten the simulation time in ADAMS solver, ADAMS flex is used to optimize the mnf-file. The optimized mnf-file contains only surface nodes.

(21)

20

Figure 12. Mesh representation of the boom with interface point’s connection elements.

7.2 Boom Head

With just the boom as a flexible body the results will not be satisfying so the rigid boom head is replaced whit a flex body. The flexible body of the boom head is represented in the same way as the boom with RBE2 elements in the interface points. The boom head is locked in the feed holder interface point and the in the Boom interface point, see Figure 13. For verification of the flexible boom and boom head see appendix J

Boom head interface point

(22)

21

Figure 13. Boom head interface points and rigid elements connections.

7.3 Limits in the mechanical model

The model built is only a model of a L8 rig and some factors are not taken into account. In reality the bearings joints has some play and friction. Play in the joints may cause different force impulses. Friction in joints is hard to find and is modeled in this thesis project as a damping in the hydraulic cylinders. The damping was changed until a satisfactory force response was found. Friction in hydraulic cylinders can be calculated from the column between the piston and cylinder wall, but the column size is not known. The moment of inertia of the carrier could not be found. This has a small impact on model validation with the direct applied forces in

cylinders (see chapter 7 and 8 for validation). Knowing the carriers moment of inertia is crucial when using a track model (see chapter 9.1.2 ) due to the movement of the carrier.

In the model, only the boom head and boom is flexible. In reality all parts are more or less flexible which may cause a different response of the machine when tramming. Especially the feed flexes significantly during tramming and even more when driving over bumps.

Feed holder interface point

Feed swing cylinder interface point

Boom interface point

(23)

22

8. Measurements

This chapter explains the measurements that took place before this thesis project. Method and entities measured is described. How and what measurements that are used as input is also described.

All measurement in this thesis was performed on an Atlas Copco L8C20LMK2 rig. Mikael Johnson at Atlas Copco carried out the measures before this thesis work in cooperation with Ødegaard & Danneskiold-Samsøe Sweden in April 2008. The measurement site was located at Atlas Copco facilities in Örebro, Sweden. The objective of the measurement was to measure the resulting loads acting on the structure at different operating conditions. Forces are determined based on pressure signals from the hydraulic cylinders on the drilling rig. Length of the

cylinders was measured with wire transducers. Accelerations were measured with accelerometers at both carrier and feed. For full table of measure points and entities see appendix K. All data is gathered in Matlab files and can be imported into ADAMS. One of the tests was interesting for this master thesis and has been used to validate the model.

8.1 Tests

Test run used for validation see appendix E. The drill rig runs over a small bump on flat ground see Figure 14. The drill rig trams towards a small bump, slowly crawls over it, stops and reverses over the bump and stops. The bump is made of a concrete brick (0.26*0.26m). It was noticed during the test that the edges crumbled. Acceleration on the feed, pressure and length in all cylinders is measured. Before analyzing the test data it was noticed that the rig had problems with stability when tramming over the bump. Because it is risky and hard to drive over the bump in full speed the driver did not maintain constant velocity.

Figure 14 Rig L8 tramming test run.

(24)

23 difficulty of how to measure carrier movement and insert it into the model without using tracks.

8.2 Sensors

To measure the effect on a rig during tramming pressures in cylinders, lengths, accelerations where measured according to Table 2. Accelerations are measured at the feeder on two

positions and at the carrier frame on three positions. Length and pressure on both sides of the hydraulic cylinders are measured. See appendix K for exact positions of sensors.

Table 2 Measured entities during tests.

Name Quantity Unit Info Boom lift Pressure [bar] Boom lift Pressure [bar] Feed dump left Pressure [bar] Feed dump left Pressure [bar] Feed dump right Pressure [bar] Feed dump right Pressure [bar] Boom swing Pressure [bar] Boom swing Pressure [bar] Feed swing Pressure [bar] Feed swing Pressure [bar] Feed extension Pressure [bar] Feed extension Pressure [bar] Boom lift (stroke) Length [mm] Feed dump (stroke) Length [mm] Boom swing (stroke) Length [mm] Feed swing (stroke) Length [mm] Feed extension (stroke) Length [mm] Frame pkt1 x-acceleration Acceleration [m/s2]

Jaw and X acceleration of frame

Frame pkt1 y-acceleration Acceleration [m/s2] Y acceleration of frame Frame pkt1 z-acceleration Acceleration [m/s2]

Roll & pitch acceleration of frame

Frame pkt2 x-acceleration Acceleration [m/s2] Jaw acceleration of frame Frame pkt2 z-acceleration Acceleration [m/s2]

Roll and Z acceleration of frame

Frame pkt3 z-acceleration Acceleration [m/s2] Pitch acceleration of frame Feed pkt1 x-acceleration Acceleration [m/s2] X acceleration of feeder Feed pkt1 y-acceleration Acceleration [m/s2] Y acceleration of feeder Feed pkt1 z-acceleration Acceleration [m/s2] Z acceleration of feeder Band oscillation left Pressure [bar]

Band oscillation left Pressure [bar] Band oscillation right Pressure [bar] Band oscillation right Pressure [bar] Feed pkt2

x-acceleration/Res Acceleration [m/s2] Pitch acceleration of feeder Feed pkt2

(25)

24

8.2.1 Wire transducers

The wire transducers were mounted by steel hose clamps. The calibration of the length and positioning was made by finding the min value of all transducers from all measurement. That position has been considered as the zero position throughout the analysis of data. The wire transducers are used to determine the positioning of the boomsystem. In this case the wire transducer measure the full length of the stroke of cylinders, which is approximately 2 meter on the larger cylinders in boomsystem. This means that the accuracy of the measurement in the future can be more accurate with wire transducers with shorter range.

8.2.2 Accelerometers

Accelerometers were all of ICP type (integrated circuit piezo electric) and were installed with magnets on frame and feed see appendix K. They were supplied with power from the

measurement equipment. The accelerometers, cabling and measurement device were all calibrated before the measurement started with a calibrated hand held vibration shaker providing 10mm/s RMS at 159.2 Hz.

8.2.3 Dynamic pressure

The pressure in the hydraulic cylinders is measured with pressure transducers. The transducers are mounted at its designated attachments and the hydraulic cylinder.

8.2.4 Signal quality

During measurements two major disturbances occurred, one with the data transfer through the telemetry system and one causing delays and lack of measurement of data. The disturbance in the telemetry system was most likely caused by severe vibrations of the transmitter and antenna together with unstable conditions for the electronics located behind the driver seat. Moving the antenna reduced the disturbances.

The disturbances in the recording system were caused by ground loops that disturbed the FireWire connection between the recorders and the computers. Efforts were made to reduce ground loops but at some points they remained. The main cause was the conversion from the 24V system to 230V.

8.3 Measurements used as input to ADAMS model

(26)

25 0 10 20 30 40 50 60 70 60 70 80 90 100 110 120 130 140 Time (sec) P re s s u re ( B a r)

Figure 15. Pressure in BLC cylinder side during test run.

0 10 20 30 40 50 60 70 10 20 30 40 50 60 70 80 90 Time (sec) P re s s u re ( B a r)

Figure 16. Pressure in BLC rod side during test run.

(27)

26 0 10 20 30 40 50 60 70 -0.5 0 0.5 1 1.5 2 2.5 3x 10 5 F o rc e ( N ) Time (sec)

Figure 17. Force in BLC during test run.

Therefore the model will be validated with calculating forces from the measured pressures in the different hydraulic cylinders. Simply inserting a force in one and read the models response in another and compare with measurement.

8.3.1 Positioning the boomsystem

The boomsystem is positioned using the wire transducers and compared with footage. The wire transducers are calibrated using the min and max value from the entire measurement and that value are set to the min or max value. This gives the cylinders lengths according to table 3. This gives some uncertainty when positioning the boomsystem. It is unknown if the driver did move the boomsystem to the mechanical stops. When inserting the values from the wire transducers it was clear that they were not calibrated satisfactory. Therefore footage was used to correct the position of the boomsystem. An incorrect positioned boomsystem in the ADAMS model causes incorrect static and dynamic forces when comparing with the forces from the measurement. The true position of the boomsystem is not known.

Table 3. Cylinder lengths.

(28)

27

8.3.2 Feed dump cylinders – FDC

The pressured side shows similar behavior for both FDC, but the other side has pressure difference between cylinders, see Figure 18. In this example the cylinders are performing a plus stroke, which means that the cylinder side is pressurized. The piston rod side over the center valve is open and there is a flow thru.

The over center valve on the right FDC has a narrower opening than the left side. This cause a much smaller pressure drop, which cause the cylinders to act with different forces. Figure 19 shows the pressure in left and right cylinder. Left cylinder is dragging behind and the right takes much higher forces. This behavior can also be seen during negative stroke. This concludes that there is something wrong with the hydraulics at the non-pressurized side in the left FDC on the test-rig, it does not have the ability to sustain the flow rate it should have. This may depend on an incorrectly mounted over center valve, wrong over center valves, or dirt in the system.

0 10 20 30 40 50 60 70 50 100 150 Time [s] P re s s u re c y lin d e r s id e [ B a r] 0 10 20 30 40 50 60 70 50 100 150 200 Time [s] P re s s u re p is to n r o d s id e [ B a r]

(29)

28 0 10 20 30 40 50 60 70 0 50 100 Time [s] F o rc e [ N ] 0 10 20 30 40 50 60 70 250 300 350 400 Time [s] S tr o k e [ m m ]

Figure 19. Force and stroke in both left (blue) and right (red) FDC.

(30)

29

9. Model validation and ADAMS simulation result

This chapter discusses methods used for validating the ADAMS rig model. Also results are shown and explained.

The data from the measurements where used as input to the ADAMS model. If new

measurements were done it would be possible to validate the model by using a track model. But building a track model was removed due to time requirements.

9.1 Model validation methods

The methods described below are the options to validate the model with the measurements from the previous chapter. Four ways to validate the model was thought of, one using the contraction/detraction measuring wire transducers mounted on the hydraulic cylinders, using frame accelerations to set the boomsystem in motion, the measured forces can also directly be inserted into the model and calculate theoretical extension using the pressure measurements. The four methods are described below.

9.1.1 Method A - Using the wire transducers

In theory one could use the length sensors to induce movement to the force functions in the model. This would make validation of the model simpler. The length measurement moves cylinder and the response force in the hydraulic cylinder can be calculated. The force can then be compared with the measured pressure in the same cylinder. But the length sensors do not have enough sensitivity to be used as an input signal in ADAMS and moving the cylinders. Their sensitivity is about ~0.1mm and with the rapid changes implied by the pressure sensors it is not enough. But they can be used to determine the positioning of the boomsystem during test runs. The data also has a lot of interference, especially at 1, 2 and 3Hz, see fast Fourier diagram - FFT diagram in appendix A. These three pikes are 3 modes of the same unknown interference. It could be the Eigen frequency of the sensor itself that interfere the measurement. Instead hydraulic positioning sensors should have been installed.

9.1.2 Method B - Acceleration measurements in the frame

The accelerations of the carrier are measured in three points on the frame. The acceleration in X direction is measure in point 1and point 2, acceleration in Y direction is measured in point 1 and the acceleration in Z direction is measure in all the points. With these measurements it is possible to determine the movement of the carrier. This means that the movement of the carrier could be used as input to the boomsystem

(31)

30 error of the position in the ADAMS model is the double integration of the acceleration error. The original orientation of the accelerometers used during this thesis work was not known. The orientation of the frame and therefore also the orientation of the accelerometers can easily be changed when using the band oscillation on the drill rig.

9.1.3 Method C - Inserting the measured force directly into model

The forces in the cylinders can be modeled by inserting the pressuresfrom the measurement directly into the model. The models response in another cylinder is then compared with the measurement. For example force is inserted to the BLC and the response force in the FDC in the model is compared with the test run measurement from the FDC.

9.1.4 Method D - Calculating extension from hydraulic pressure sensors

The movement of the boomsystem cylinders introduced by tramming can be calculated from the pressure measurements. The movement is in theory only dependent on the elasticity of the cylinder walls and the compression module of the hydraulic fluid. The measured force is used as input to the stiffness equation (13), which results in cylinder movement during the test run. The cylinder length movement is inserted to the hydraulic cylinder equations in the ADAMS model. The resulting force in the model simulation can then be validated with the force from the measurement.

Therefore the model will be validated with calculating forces from the measured pressures in the different hydraulic cylinders. Simply inserting a force in one and read the models response in another and compare with measure.

9.2 Model validation rigid bodies

Method C and Method D where used to validate the model. Method A using the wire transducers was discarded due to poor signal, see appendix A. Method B acceleration measurement yielded poor results because acceleration under 0.5 m/s^2 was not measured. Because of that the data from the accelerometers is unreliable and cannot be used as input to the ADAMS model. The original orientation of the acceleration measurement must be known, to sync it with the coordinate system in the ADAMS model. If not, the angle differences in the coordinate systems will cause an error. This means also that the error of the position in the ADAMS model is the double integration of the acceleration error. The original orientation of the accelerometers used during this thesis work was not known. The orientation of the frame and therefore also the orientation of the accelerometers can easily be changed when using the band oscillation on the drill rig

9.2.1 Method C - Inserting force into model

(32)

31 BLC had to be evaluated with several design evaluations in ADAMS. When the system finally was balanced the length of the BLC was evaluated to the fourth decimal in millimeters and the length of the BLC was changed from 1501mm to 1430mm. This change has an effect on how the system reacts to impulses. When the drill rig hits the bump the angle of the machine with boomsystem changes and the boomsystem becomes unbalanced and drops. Therefore using this method it is only possible to simulate until T = 18 sec.

Figure 20 shows the result of simulation from T=6 when the rig stands still to T=18 when it hits the bump.

Figure 20. Validation feed dump cylinder right.

The simulation result shows a magnitude error but no phase error. The simulation is 75% match with the measurement, which is a good result. But the model cannot be validated due to the fact that the method does not work when the drill rig hits the bump.

9.2.2 Method D - Calculating extension from hydraulic pressure sensors

(33)

32

Figure 21. Validation boom lift cylinder.

Method D has the same problem as Method C. The drill rig hits the bump and the angle of the entire machine changes, which cannot be simulated with the method. The static force offset between the simulation and measurement is small.

Figure 22 shows the relevant time sequence from stand still to T=18 when the drill rig hits the bump.

Figure 22. Validation Boom lift cylinder.

(34)

33

Figure 23. Validation Boom lift cylinder.

(35)

34

9.3 Model validation flexible bodies

The flexible body analysis is done with the same boom position as in the rigid body analysis using method D. The boom head and boom rigid bodies are replaced with flexible parts. The feeds centre of gravity (see Figure 24) is located to the left of the boomsystem, and viewed in the drill rigs driving direction. The two feed dump cylinders receive different loads because of the centre of gravity location.

Figure 24. Centre of gravity and interface points for FDC.

(36)

35

Figure 25. Force in right (Blue) and left (red) FDC.

The difference between the two cylinders is about 37 500N in the boomsystem position given by table 4. As shown in Figure 24 the centre of gravity of the feed is located at the left side of the left feed dump cylinder. Figure 25 shows that the force in the right FDC is larger than the left FDC. The higher force in the right FDC is because of the bending moment caused by the feed centre of gravity position. The boom bends to left, which cause tension in the right boom plate and pressure tension in the left boom plate. This will give higher bending moment in the right FDC. The right FDC will elongate more due to the bending of the boom to the left, which is the reason for the larger force in the right FDC. This phenomenon is not possible to see with rigid bodies. Therefore it is important to use flexible bodies in a surface drill rig ADAMS model.

Table 4. The position of boomsystem.

Boom lift cylinder (BLC) 1501mm Boom swing cylinder (BSC) 929mm Feed swing cylinder (FSC) 1086mm Feed dump cylinder (FDC) 2001mm Feed extension cylinder (FEC) 1827mm

(37)

36

Figure 26. Comparison flexible (red) and rigid (black) parts force in feed dump cylinders.

The force in BLC is shown in Figure 27 where the red curve is the simulated force and the blue is the results from measure. In Figure 28 the forces are compared between T=7.5 and 16.5 sec.

Figure 27. Simulated (red) and measured (blue) in BLC.

Figure 28. Simulated (red) and measured (blue) in BLC.

(38)

37

Figure 29. Flexible parts (blue) and rigid (red).

(39)

38

9.4 Conclusions

The model cannot be validated with the measurement used in this thesis project. It is therefore not possible to use the model for finding forces in the boomsystem.

The measured data is not good enough to validate the model. It needs to be more accurate and made especially for validating the model. The objective of the measurement campaign used in this thesis project was to measure the resulting loads acting on the structure at different operating conditions. The measured loads were intended for Atlas Copco to make structural design verification, and were therefore not intended for model validation. It is vital that the measurement is designed to validate the specific simulation model. In this case the

accelerometers that measure the movement of the carrier should be able to register all acceleration magnitudes. This was not the case for the measurement campaign and the

accelerometers could not be used to set the boomsystem in motion and validate the model. It is concluded that the model needs to be developed if used for load analysis of rigs.

9.5 Further studies

There are some ways to improve the model, but first it has to be validated. The ADAMS model of the L8 surface drill rig needs to be validated if it should be used for design. There are some ways to do this. Either validate it with accelerometers in the frame or build a track model. To improve the model more, flexible part could be added. The hydraulic functions could also be replaced with simulation software like Easy5 or Matlab/Simulink.

9.5.1 Accelerometers

Accelerometers in the frame could be used to validate the model when driving over bumps. The method of using pressure in the hydraulic cylinders does not take into account the movement of the entire drill rig. A good measure of the accelerations in the frame would solve this problem and the model, except tracks, could be validated. It is important that the model and the measurement have the same original conditions, so that the coordinates system of the measurement is the same as in the model.

9.5.2 Track model

(40)

39

Figure 30. ADAMS track model.

If an accurate working track model exists one can drive over a bump in an ADAMS simulation and the forces in the boomsystem and hydraulic cylinders could be found. The forces are then compared with forces from driving over the same bump where the pressures in the hydraulic cylinders are measured. The pressures can be calculated to force and compared with the forces given by the ADAMS simulation.

9.5.3 Flexible feeder

The feeder sways heavily during tramming, which affects the loads at the boomsystem. A flexible feed is therefore needed to make a more exact model. The Finite element method is a possibility to make a flexible part of the feed. But the complexity with lots of parts and different materials makes this difficult. The feed stiffness can be found with a ping test. By inserting the result from the ping test to a discrete flexible link, which is built by rigid links connected with spring-dampers, a model of a flexible feed can be made. Not as correct as finite element analysis but better than a rigid body. Errors are presumed to arise because of the complex structure. A simple aluminum beam would yield a good result using this method but not the feed.

9.5.4 Easy5 – Hydraulic model

(41)

40

Figure 31. ADAMS and Easy5 co-simulation.

It was decided early in the thesis project that Easy5 was excessive to be used for simulating the hydraulics when tramming. When tramming between holes the boomsystem is not steered and the hydraulic locks on the cylinders are active. Then only the cylinder has to be taken into account when calculating the stiffness. When moving the boomsystem the hydraulic locks are open and the stiffness becomes different because of the dynamics of the hydraulic system. If forces need to be simulated during steering of the boomsystem either Easy5 or

MatLab/Simulink should be used.

Easy5 is also better to model the stiffness of the hydraulic cylinders, because it has a better way of modeling cylinders than the function used in this thesis work. The downside is that Easy5 will increase the simulation time. Easy5 can also be used to achieve the secondary objective of this thesis project, reducing the loads with accumulators. Different accumulator setups and pre-charge pressure can be investigated, see Axelsson and Fredriksson 2007 [2].

9.5.5 Cylinder length measurements

It should be investigated if the length change of the hydraulic cylinders when tramming can be measured with good accuracy. If the length change could be measured with ~0.01mm this could be an option. This could not alone validate the model, but can be used as comparison with measurements when using accelerometers or a track model.

9.6 How to proceed

If the ADAMS model is going to be used as a start for a new dimensioning tool it needs to be improved.

It is recommended that a track model be built to verify the model against measurements. Easy5 should also be used to increase the flexibility of the model. With Easy5, design of accumulators is possible and finding loads when the boomsystem is maneuvered. Easy5’s model of the hydraulic system is also considered to more extensive. Work should also be put into a flexible feeder. It is shown during measurements that the feeder sways heavily and this will affect the force response in the boomsystem.

New measurements should be carried out. Under this thesis project the lesson was learned that the conditions in the ADAMS model must be carefully recreated if verification could be

(42)

41 verifying a ADAMS rig model with tracks, pressure measurements in the hydraulic cylinder can be used. These measurements have good accuracy and are easy to compare with an Easy5 and ADAMS co-simulation result. The measurement used in this thesis project did not measure the boomsystem adequately. It is recommended that angle sensors be used to monitor the

boomsystem position. Angle sensors have much better accuracy than wire transducers. The boomsystem translations should be monitored and correlated towards the ADAMS model. If the boomsystem is incorrectly positioned the force response will be incorrectly. It is also important that the carriers band oscillation is known. Either measure it, or make sure that it is unchanged during the measurements. The surface used in the measurement should be flat so that it can be built in ADAMS. The hinder used should also be much smaller than the used in this thesis project, so that the rig can drive over it in constant speed. Constant speed is easier to work with in ADAMS. The hinder dimension should also be well known and it should be of a material that does not crumble when driven over or change form.

The model that was developed during this master thesis is easily configured so that other types of boomsystems and carriers also can be simulated. It is relatively easy to change different parts in the boomsystem so that new designs can be tried. With easy5 it is also possible to try out changes in the hydraulics, which can be interesting.

(43)

42

10. References

[1]Atlas Copco official hompage www.atlascopco.com

[2] Axelsson, E and Fredriksson, V. Design of passive ride control systems for scooptrams. Master Thesis Linköping: Institute of Technology 2007

[3]Freedman, R and Young,H. University Physics with Modern Physics with Mastering Physics (12th Edition): San Francisco: Pearson Addison Wesley .April 2, 2007. 080538684X

[4] Mechanical Dynamics, Incorporated Basic ADAMS Full Simulation Training Guide VERSION 11.0 PART NUMBER 110VIEWTR-03 Mechanical Dynamics, Incorporated.2001

[5] Blundell, M and Harty, D. (2004). The Multibody Systems Approach to Vehicle Dynamics. Norfolk: Elsevier Butterworth-Heinemann, 0-7506-5112-1.

[6] Kreyszig, E. (2007). Advanced Engineering Mathematics 9 edition. Ottawa: Wiley, 04-714-8885-2.

[7] Jönsson, P. (2006). MATLAB beräkningar inom teknik och naturvetenskap. Poland: Studentlitteratur, 978-91-44-01780-8.

[8]Linköpings Universitet IEI. Formelsamling i hydraulik och pneumatik, Linköping Institute of Technology. May 1995

(44)

43

11. Contacts

Atlas Copco AB

Support and help in related CAD and dimension questions Anders Hellman Mechanical Engineer

anders.hellman@se.atlascopco.com Supervisor in ADAMS software Morgan Norling Senior MBS Analyst morgan.norling@se.atlascopco.com Supervisor Dec-Feb

Patrick Kenger Group Manager Drill Rig Mechanics patrik.kenger@se.atlascopco.com

Supervisor Sep-Nov

Göran Tuomas Group Manager Drill Rig Mechanics Goran.Tuomas@se.atlascopco.com

Supervisor Sep-Dec

Mickel Johnson Mechanical Engineer johnson.micke@gmail.com

University contacts

Luleå University of Technology

Supervisor at Luleå University of Technology Mikael Nybacka doktorand at LTU

mikael.nybacka@ltu.se

Help and advisor with questions regarding the FEA software Marc Anders Lundbäck Research Engineer at LTU

andreas.lundback@ltu.se External contacts Stacke hydralik

(45)

44

12. Appendices

The appendices is containing detail data and detail information that’s has been used in this thesis project. The information in the appendices is not necessary needed to follow the line of thought in the thesis, but it is needed for repeatability of the project. The authors have chosen to move so much as possible of the information to gain a readable thesis.

(46)

45

Appendix A - filtering of data from length sensors

This chapter explains the filtering of measure raw data from the length sensors and some conclusions regarding the length sensors.

The signals from the length sensors has lot of distortion, see Figure 32. For the purpose of using the measurement as input signal to ADAMS, the signal has to be filtered.

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 269.5 270 270.5 271 271.5 272 272.5 273 273.5 274 274.5 B L C l e n g h t [m m ] Data

Figure 32. Length measurement during tests in boom lift cylinder – BLC.

FFT – fast Fourier transformation analysis [6], see Figure 33, shows frequency spectrum of the BLC hydraulic cylinder. Three pikes occur at 1, 2, and 3 Hz. These are probably some kind of interference, possibly the Eigen frequency of the sensor itself. The boomsystem vibrates during operation, which could interfere with the measurement. Also the wind could interfere.

(47)

46 0 5 10 15 0 0.05 0.1 0.15 0.2 0.25 Frequency [Hz] L e n g h t [m m ] Figure 33. FFT analysis of BLC.

Three Bandstop filters are applied in ADAMS to sort out these three frequencies. Engine and other vibrations that are not modeled in ADAMS and therefore not of interest are removed with a low-pass filter of 0.9 Hz and order 3. For result see Figure 34.

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 270.5 271 271.5 272 272.5 273 B L C l e n g h t [m m ] Data

Figure 34. Filtered length measurement from BLC.

(48)

47

Appendix B - Data used in ADAMS model

This appendix contains detailed data regarding the cylinders in the boomsystem.

Boom swing (BSC)

Manufacture Parker Stroke

Piston diameter 125 mm min 855 mm

Rod diameter 63 mm max 1078 mm

Piston area +stroke 12272 mm^2 Piston area -stroke 9155 mm^2

Weight

Max force +stroke 306800 N Piston 35 kg Max force -stroke 625 N Cylinder 45 kg Inertia

Piston 2.0416543744E+006,2.0580137285E+006,1.7108839376E+005 Cylinder 3.778032915E+006,3.7839901354E+006,3.8174776008E+004 TP Distance from Ref (Z) mm

Piston 400 mm

Cylinder 300 mm

Boom lift (BLC)

Manufacturer Stackeryd Stroke

Weight

Max force +stroke 636175 N Piston 112 kg Max force -stroke 477125 N Cylinder 83 kg Inertia

Piston 435 mm

Cylinder 320,55 mm Feed dump (FDC) 2st

Manufacture Stackeryd Stroke

(49)

48 Weight

Max force +stroke 384850 N Piston 56.3 kg Max force -stroke 288625 N Cylinder 65.7 kg Inertia

Piston 629,7 mm

Cylinder 556 mm

Feed swing (FSC)

Weight

Max force +stroke 384850 N Piston 31 kg Max force -stroke 288625 N Cylinder 52.5 kg Inertia

Piston 489,7340201 mm Cylinder 262,63 mm Feed extension (FEC)

Weight

Max force +stroke 196350 N Piston 31.5 kg Max force -stroke 118425 N Cylinder 38.5 kg TP Dis from Ref (Z) mm

Inertia

(50)

49 Complement data

(51)

50

Appendix C - Boom mode shapes

This appendix contains results from the evaluation of the mnf files used to create flexible elements in ADAMS. The figures show the mode shapes of the boom are the same for ADAMS and NX Nastran but in different phase.

The mode shape is the same for ADAMS and NX Nastran but in different phase.

Figure 35. Mode 7 from ADAMS. Figure 36. Mode 8 from ADAMS.

Figure 37. Mode shape 9 from ADAMS. Figure 38. Mode shape 10 from ADAMS.

(52)

51

(53)

52

Appendix D - Boom head mode shapes

This appendix contains results from the evaluation of the mnf files used to create flexible elements in ADAMS. The figures show the mode shapes of the boom head are the same for ADAMS and NX Nastran but in different phase.

Figure 43. Mode shape 7 from ADAMS. Figure 44. Mode shape 8 from ADAMS.

(54)

53

Figure 47. Mode shape 7 from Nx Nastran. Figure 48. Mode shape 8 from Nx Nastran.

(55)

54

Appendix E - Measured forces

This appendix presents measured forces in all hydraulic cylinders during the test run. The forces are calculated from pressure in plus and minus side.

0 10 20 30 40 50 60 70 -12 -10 -8 -6 -4 -2 0 2 4x 10 4 Time (sec) F o rc e ( N )

Figure 51. Force in feed dump cylinder left (FDC).

0 10 20 30 40 50 60 70 -0.5 0 0.5 1 1.5 2 2.5 3x 10 5 F o rc e ( N ) Time (Sec)

Figure 52. Force in boom lift cylinder.

(56)

55 0 10 20 30 40 50 60 70 -1.5 -1 -0.5 0 0.5 1 1.5x 10 5 Time (sec) F o rc e ( N )

Figure 53. Force in boom swing cylinder.

0 10 20 30 40 50 60 70 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2x 10 5 F o rc e ( N ) Time (Sec)

(57)

56 0 10 20 30 40 50 60 70 0 0.5 1 1.5 2 2.5 3 3.5 4x 10 4 Time (Sec) F o rc e ( N )

(58)

57

Appendix F - Hydraulic schematics

(59)

58

Appendix G - Moment of inertia and centre of gravity

This appendix describes the approach to gain the moment of inertia and centre of gravity of the included parts of the drill rig. The appendix also contains command file that can be used to merge the included parts.

Moment of inertia and centre of gravity for the different parts in the boomsystem where found by importing a parasolid from Pro Engineer1_{into ADAMS. The parasolid that usually contains}

lots of small parts is then merged in ADAMS. The centre of gravity, weight and moment of inertia from the merged parasolid are used as input to the rigid body ADAMS parts. The quality of the weight and inertia is very dependent on the quality of the cad-model, but also the size of the polygons in the parasolid file.

The feeder is a complex structure consisting of many parts. This makes it difficult to find the centre of gravity and the correct moment of inertia. The CAD-models given by Atlas Copco where built up by approximately 3000 parts. That many parts make it very time consuming to merge them by hand in ADAMS.

The feeder consists of mainly two materials, steel and aluminium. Therefore it cannot just used be merged cause the densities would also merge. The feeder aluminium beam framework needs to maintain its correct density. This combined with the large number of parts concluded the need for a merge macro. The merge macro command file merge bodies to a single body and gives the body correct mass and correct moment of inertia despite the original bodies has different density. The merge command file where written by Morgan Norling at Atlas Copco division ROCTEC.

var set variable_name = pnames string_value = (eval(DB_FILTER_NAME(DB_CHILDREN( .matare , "part" ),"*PART*")))

for var = ii start_value = 2 end_value = (eval(max(shape(pnames))))

1

(60)

59 var set var= jj integer_value = (eval(ii))

part modify rigid mass_properties & part_name = (eval(pnames[jj])) & mass = (eval(eval(pnames[jj]).mass)) & ixx = (eval(eval(pnames[jj]).ixx)) & iyy = (eval(eval(pnames[jj]).iyy)) & izz = (eval(eval(pnames[jj]).izz)) if cond=(jj > 1)

part merge rigid_body part_name=(eval(pnames[jj])) into_part=.matare.PART2

end end

lis var var=pnames !clean up

var del var = pnames,jj !,nn

Centre of gravity and moment of inertia for BLC and FDC was found in cooperation with Stacke hydraulic cylinders. FEC, FSC and BSC were approximated from Parker product information. Models where built in ADAMS with the correct dimensions and weight and approximated values for the moment of inertia was found.

(61)

60

Appendix H - Mechanical stop

This appendix describes how the mechanical stops are modeled in the cylinders.

The mechanical stops for the different hydraulic cylinders are modeled with BISTOP [4] functions. The BISTOP function models a gap element Figure 56. The gap element consists of a slot that defines the domain of motion for the hydraulic cylinder, its stroke. Part I in our case the rod is free to move without forces acting on it. When the rod tries to move beyond the physical definition of the slot, impact forces representing contact are created.

Figure 56 Illustration of gap function.

The BISTOP impact force has two components, a stiffness component and damping

component that may be used to model energy loss. To prevent a discontinuity in the damping force at zero penetration, the damping coefficient is defined as a cubic step function of the penetration, see Figure 57. The damping coefficient has a maximum at c-max at a user-defined penetration.

Figure 57. Damping coefficient vs penetration.

Stiffness and damping for the BISTOP functions where given by Atlas Copco see Table 4.

Table 4. Input to BISTOP functions.

Stiffness force coefficient 105N /mm

Stiffness force exponent 1,7

Damping Coefficient 1000Nm /s

(62)

61

Appendix I - Friction in joints and damping

This appendix describe how the friction and damping is modeled and which assumptions is made to gain a realistic friction and damping in the joints.

In reality the system itself has a damping effect due to damping in materials. The hydraulic cylinders also have leakage, which gives a damping effect. To be able to cope with this effect a damping force need to be added to the model.

Friction Idealized case

There are three phases that defines friction, stiction, transition and dynamic, see Figure 58. The idealized case will cause discontinuity due to the infinite slope from stiction to transition and therefore simplified case is required. Discontinuities lead to solution convergence difficulties.

Figure 58. Idealized friction case.

(63)

62

Figure 59. STEP function.

(64)

63

Appendix J - Verification of flexible bodies

This appendix contains data and results from the verification of the flexible part of the boomsystem. To verify the ADAMS flex model of the boom a comparison is done in the FEM solver NX Nastran. A simple static FEM analyze is done with single force with a magnitude of 60 000N, first measure in X, second measure Y and third Z direction acting on the Boom head interface point. The boom is constrained in the feed dump cylinders interface point and in boom attachment interface points.

The boom head is verified in same way as the boom but the forces acting on the feed dump cylinders interfaces points and are fully constrained in feed holder interface point and in boom interface point.

The difference between NX Nastran and ADAMS is calculated with equation (14) and the difference between them is shown in percents see table 5 and 6.

1 − = results Adams results Nastran Nx Difference (14)

Table 5. Result comparison of the boom head.

Direction ADAMS (mm) NX Nastran (mm) Difference (%) X Measure X 0.03872 0.03877 0.13 X Measure Mag 0.0413 0.0416 0.73 Y Measure Y 0.1562 0.1562 0 Y Measure Mag 0.1569 0.1569 0 Z Measure Z 0.1787 0.1928 7.89 Z Measure Mag 0.2072 0.2222 7.24

Table 6. Result comparison of the boom.

Direction ADAMS (mm) NX Nastran (mm) Difference (%) X Measure X 0.04674 0.04694 0.43 X Measure Mag 0.1630 0.1644 0.85 Y Measure Y 1.494 1.513 1.27 Y Measure Mag 1.502 1.521 1.26 Z Measure Z 1.523 1.554 2.00 Z Measure Mag 1.523 1.554 2.00

The first six modes are free body modes so it is not of interest and is not compared here. The first real modes 7 to 10 are compared to verify the translation between the different software. The results from modal analyses is show table 7 for the boom and the for boom head is shown in table 8.

Table 7. Boom mode comparison.

Modes ADAMS (Hz) NX Nastran (Hz) Difference (%)

(65)

64

8 326.8 322.5 -1.31

9 345.7 341.6 -1.12

10 391.1 392.2 2.81

Table 8. Boom head mode comparison.

Modes ADAMS (Hz) NX Nastran (Hz) Difference (%)

7 310.7 320.2 3.05

8 696.4 692.4 -0.57

9 796.2 767.0 -3.66

10 973.8 914.8 -6.05

Result from flex body verification

The difference in displacement of the boom between ADAMS and Nx Nastran is between is small. Which means ADAMS gives less displacement of than Nx Nastran with same static load. As long it ADAMS does not gives more displacement than Nastran or other FE-programs it is safe to say the flex model is a better representation than a fully rigid model is shown in table 4. The difference in displacement in the boom head is more than the difference in the boom but still on the safe side.

(66)

65

Appendix K - Sensor functions and positions

This appendix contains tables of the sensors function and figures of its position of the sensor used under the measurement used in this thesis.

Table 9 describes the function of the sensor. The positions and directions of accelerometers are shown in Figure 60.

(67)

66

Figure 60 is showing postions of the sensors.

(68)

67

Figure 61show the position the accelerometers on the carrier frame.

Figure 62 shows the more detailed desciption of the position of the accelerometers.

(69)

68

Appendix L - Geometric properties

This appendix contains Geometric properties of included parts Carrier.

Table 10. Geometric properties of the carrier and boomsystem.

Mass 1.225E+004 kg

Center marker CARRIER.cm name

Location -1518.0, 11.9, 1470.9 mm

Orientation 0.0, 0.0, 0.0 deg

Mass Inertia Tensor IXX 1.0E+010 kg/mm^2 Mass Inertia Tensor IYY 1.0E+010 kg/mm^2 Mass Inertia Tensor IZZ 1.0E+010 kg/mm^2

Boom attachment

Mass 356.7917966549 kg

Center marker BOOM_ATTACHMENT.cm name Location 668.877365986, -516.177203346, 1414.819994230 mm Orientation 229.487294519, 8.628822346, 49.084401503 deg Mass Inertia Tensor IXX 2.8085032855E+007 kg/mm^2 Mass Inertia Tensor IYY 2.4128964617E+007 kg/mm^2 Mass Inertia Tensor IZZ 1.2066481282E+007 kg/mm^2

Boom (rigid)

Mass 642.722404372 kg

Center marker BOOM.cm name

Location 1627.573794061, -499.791219610,1576.402244214 mm Orientation 90.0059003729, 89.4165453362, 88.7749271705 deg Mass Inertia Tensor IXX 3.5598297118E+008 kg/mm^2 Mass Inertia Tensor IYY 5470256338E+008 kg/mm^2 Mass Inertia Tensor IZZ 3.2707315956E+007 kg/mm^2

Boom head (rigid

)

Mass 523.124747375 kg

Center marker BOOM._HEADcm name

Location 2994.992668854, -531.556882114, 2019.38030161 mm Orientation 89.6515894784, 93.9291451926, 191.5854310541 deg Mass Inertia Tensor IXX 4.2276675742E+007 kg/mm^2 Mass Inertia Tensor IYY 3.2460455508E+007 kg/mm^2 Mass Inertia Tensor IZZ 2.5841268739E+007 kg/mm^2

Feed

Mass 4342.2683541502 kg

Center marker FEED.cm name

(70)

69 Orientation 88.9370160517, 153.9154274158, 151.4427844551 deg

Mass Inertia Tensor IXX 4.4746030769E+010 kg/mm^2 Mass Inertia Tensor IYY 4.4704914898E+010 kg/mm^2 Mass Inertia Tensor IZZ 9.31308416E+008 kg/mm^2

Feed holder

Mass 703.7395191876 kg

Center marker FEED_HOLDER.cm name Location 2757.137370455, -478.265551887, 3270.12722722 mm Orientation 87.4346289354, 153.9293086035, 173.800903019 deg Mass Inertia Tensor IXX 9.3311637082E+008 kg/mm^2 Mass Inertia Tensor IYY 9.3081804474E+008 kg/mm^2 Mass Inertia Tensor IZZ 2.1542560209E+007 kg/mm^2

Rock drill

Mass 242.0222576373 kg

Center marker ROCK_DRILL.cm name

Location 1252.941986807, 160.164766414, 6380.540938080 mm Orientation 270.2869014608, 26.0988721115, 179.7666812072 deg Mass Inertia Tensor IXX 1.6149526497E+007 kg/mm^2 Mass Inertia Tensor IYY 1.507922775E+007 kg/mm^2 Mass Inertia Tensor IZZ 3.3514046253E+006 kg/mm^2

Cylinders

Boom lift cylinder

Mass 83.0 kg

Center marker BOOM_LIFT_CYL_CYL.cm 1. name 2. Location 3. 1341.4884264766, -500.0, 1175.8662796589 4. mm 5. Orientation 6. 90.0, 80.8694585162, 90.0 7. deg 8. Mass Inertia Tensor

IXX

9. 6.076756E+006 10. kg/mm^2 11. Mass Inertia Tensor

IYY

12. 6.110637E+006 13. kg/mm^2 14. Mass Inertia Tensor

IZZ

15. 6.31381E+005 16. kg/mm^2

Boom lift piston

Mass 112.0 kg

MASTER'S THESISSurface Drill Rig Simulation

M A S T E R ' S T H E S I S

Surface Drill Rig Simulation

Johan Theander

Peter Holmström

Preface

Abstract

Nomenclature

Table of contents

1. Introduction

1.1 Background

1.2 Purpose

1.3 Objectives

2. About ADAMS

3. Surface drill rig – ROC L8 LF

5. Mechanical model - ADAMS

5.1 Rigid bodies

5.2 Joints

5.3 Hydraulic cylinders

6. Hydraulics

6.1 Counterbalance valve

6.2 Compressibility of hydraulic fluids

6.3 Hydraulic cylinder Stiffness

ν

ν

ν

ν

(

)

β

ν

ν

β

ν

ν

β

7. Flexible parts

7.1 Boom

7.2 Boom Head

7.3 Limits in the mechanical model

8. Measurements

8.1 Tests

8.2 Sensors

8.2.1 Wire transducers

8.2.2 Accelerometers

8.2.3 Dynamic pressure

8.2.4 Signal quality

8.3 Measurements used as input to ADAMS model

8.3.1 Positioning the boomsystem

8.3.2 Feed dump cylinders – FDC

9. Model validation and ADAMS simulation result

9.1 Model validation methods

9.1.1 Method A - Using the wire transducers

9.1.2 Method B - Acceleration measurements in the frame

9.1.3 Method C - Inserting the measured force directly into model

9.1.4 Method D - Calculating extension from hydraulic pressure sensors

9.2 Model validation rigid bodies

9.2.1 Method C - Inserting force into model

9.2.2 Method D - Calculating extension from hydraulic pressure sensors

9.3 Model validation flexible bodies

9.4 Conclusions

9.5 Further studies

9.5.1 Accelerometers

9.5.2 Track model

9.5.3 Flexible feeder

9.5.4 Easy5 – Hydraulic model

9.5.5 Cylinder length measurements

9.6 How to proceed

10. References

11. Contacts

12. Appendices

Appendix A - filtering of data from length sensors

Appendix B - Data used in ADAMS model

Appendix C - Boom mode shapes

Appendix D - Boom head mode shapes

Appendix E - Measured forces

Appendix F - Hydraulic schematics

Appendix G - Moment of inertia and centre of gravity

Appendix H - Mechanical stop

Appendix I - Friction in joints and damping