Evaluation of a Programmable Hydraulic Valve for Drill Rig Applications

(1)

Evaluation of a Programmable Hydraulic Valve for

Drill Rig Applications

Jonathan de Brun Mangs and Mikael Tillquist

Division of Fluid and Mechatronic Systems

Master thesis

Department of Management and Engineering LIU-IEI-TEK-A–18/03026–SE

(2)

(3)

Linköping University | Division of Fluid and Mechatronic Systems Master Thesis | Mechanical Engineering | Electrical Engineering Spring 2018 | LIU-IEI-TEK-A–18/03026–SE

Evaluation of a Programmable

Hydraulic Valve for Drill Rig

Applications

Jonathan de Brun Mangs and Mikael Tillquist

Supervisor: Magnus Sethson Examiner: Liselott Ericson Linköping University SE-581 83 Linköping 013-28 10 00, www.liu.se

(4)

(5)

Avdelning, Institution

Division, Department

Institutionen för ekonomisk och industriell utveckling Fluida och mekatroniska system

Department of Management and Engineering Fluid and Mechatronic Systems

Datum 2018-06-15 Date Språk Language Svenska/Swedish Engelska/English Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats övrig rapport

URL för elektronisk version

http://www.ep.liu.se

ISBN

—

ISRN

LIU-IEI-TEK-A–18/03026–SE

Serietitel och serienummer

Title of series, numbering

ISSN

—

Titel

Title Evaluation of a Programmable Hydraulic Valve for Drill Rig Applications

Författare

Author Jonathan de Brun Mangs and Mikael Tillquist

Sammanfattning

Abstract

The increase of intelligent systems can be seen in every industry. Integrated sensors and processors are used with internal control systems to create better performance for mobile hydraulic applications.

The report describes how an evaluation was made to see if the productivity of a drill rig could be increased. This was done by implementing a programmable hydraulic valve to control the hydraulic drilling functions. The productivity would be increased by reducing the downtime due to jamming in the drill hole. Jamming occur when the system does not compensate for changes in rock conditions. By conducting a series of tests in a controlled environment with simulated loads, the response time of the CMA system and original system could be determined and compared. The CMA system had a response time that was 60-64% faster than the original system.

Two different implementations of a controller was tested. Ziegler-Nichols method was used to get the initial value of the PI parameters. The controller that was implemented onboard the valve’s CPU was considered more successfull to reduce jamming.

A drill test was conducted to ensure that the programmable valve could handle a drilling procedure with the controller that was implemented onboard the valve’s CPU. The valve handled the drilling procedure well.

Nyckelord

Keywords Hydraulic valve, CAN, Drill rig, Rock drill, Blast hole drilling, Time delays, Ziegler-Nichols, PI-control

(6)

(7)

Upphovsrätt

Detta dokument hålls tillgängligt på Internet — eller dess framtida ersättare — under 25 år från publiceringsdatum under förutsättning att inga extraordinära omständigheter uppstår.

Tillgång till dokumentet innebär tillstånd för var och en att läsa, ladda ner, skriva ut enstaka kopior för enskilt bruk och att använda det oförändrat för ickekommersiell forskning och för undervisning. Överföring av upphovsrätten vid en senare tidpunkt kan inte upphäva detta tillstånd. All annan användning av dokumentet kräver upphovsmannens medgivande. För att garantera äktheten, säkerheten och tillgängligheten finns det lösningar av teknisk och administrativ art.

Upphovsmannens ideella rätt innefattar rätt att bli nämnd som upphovsman i den omfatt-ning som god sed kräver vid användomfatt-ning av dokumentet på ovan beskrivna sätt samt skydd mot att dokumentet ändras eller presenteras i sådan form eller i sådant sammanhang som är kränkande för upphovsmannens litterära eller konstnärliga anseende eller egenart.

För ytterligare information om Linköping University Electronic Press se förlagets hemsida http://www.ep.liu.se/

Copyright

The publishers will keep this document online on the Internet — or its possible replacement — for a period of 25 years from the date of publication barring exceptional circumstances.

The online availability of the document implies a permanent permission for anyone to read, to download, to print out single copies for his/her own use and to use it unchanged for any non-commercial research and educational purpose. Subsequent transfers of copyright cannot revoke this permission. All other uses of the document are conditional on the consent of the copyright owner. The publisher has taken technical and administrative measures to assure authenticity, security and accessibility.

According to intellectual property law the author has the right to be mentioned when his/her work is accessed as described above and to be protected against infringement.

For additional information about the Linköping University Electronic Press and its proce-dures for publication and for assurance of document integrity, please refer to its www home page: http://www.ep.liu.se/

c

(8)

(9)

Abstract

The increase of intelligent systems can be seen in every industry. Integrated sensors and processors are used with internal control systems to create better performance for mobile hydraulic applications.

The report describes how an evaluation was made to see if the productivity of a drill rig could be increased. This was done by implementing a programmable hydraulic valve to control the hydraulic drilling functions. The productivity would be increased by reducing the downtime due to jamming in the drill hole. Jamming occur when the system does not compensate for changes in rock conditions. By conducting a series of tests in a controlled environment with simulated loads, the response time of the CMA system and original system could be determined and compared. The CMA system had a response time that was 60-64% faster than the original system.

Two different implementations of a controller was tested. Ziegler-Nichols method was used to get the initial value of the PI parameters. The controller that was imple-mented onboard the valve’s CPU was considered more successfull to reduce jamming.

A drill test was conducted to ensure that the programmable valve could handle a drilling procedure with the controller that was implemented onboard the valve’s CPU. The valve handled the drilling procedure well.

(10)

(11)

Acknowledgments

We would like to give a special thanks to the supervisor Magnus Sethson for valuable inputs along the way. We would also like to thank our examiner Liselott Ericson. The thesis was conducted at Epiroc Rock Drills AB. The resources and support received from the company have been increadibly valuable. A special thanks to our industrial su-pervisors Fredrik Öhman, Dan Storås and Simon Magnusson for helping us understand the drill rig. We would also like to express our gratitude to the valve manufacturer, Eaton, for all the support regarding the CMA valve.

(12)

(13)

Nomenclature

∆ L Change in response time s

ˆ

G Modelled transfer function

-ωcut,of f Cut off frequency for the low pass filter in the controller Hz

b Steepest gradient of the step bar/s

e Control error

-F Controller

-G Transfer function

-Gc Closed loop transfer function

-iv Control current mA

K The pressure to current ratio of the feed pressure mA/bar

Kp Proportional gain

-M Current offset for the electro-proportional valve mA

nrot Rotation motor speed rpm

pdamp,pump Damping pressure reference for pump 1 bar

pdeadband Minimum control error bar

pf eed,drill Desired feed pressure for drilling bar

pf eed,max Maximium feed pressure bar

pf eed,min Lower pressure limit for the feed bar

pf eed,ms Mode switch feed pressure bar

pf eed,ref Reference feed pressure bar

pf eed,ss Steady state feed pressure bar

pf eed Feed pressure bar

(20)

xviii Contents

pjamming Rotational pressure limit to enter anti-jamming mode bar

pperc,collaring Desired percussion pressure for collaring bar

pperc,drill Desired percussion pressure for drilling bar

prot,drill Desired rotational pressure for drilling bar

prot,ms Mode switch rotational pressure bar

prot,pump Rotational pressure reference for pump 2 bar

prot,ss Steady state rotational pressure bar

Td Derivative parameter -Ti Integrational parameter -TS Sample time s T1,2 Time s u Control signal -L Time delay s

(21)

Acronyms

CAN Controller Area Network 17 CMA Controls Mobile Advanced 3 CPU Central Processing Unit 17

CSV Comma Separated Value files 38, 41, 43, 45 EDS Electronic Data Sheet 19, 24

IFC Intelligent Flow Control 18 IIR Infinite Impulse Response 39 PDO Process Data Object 17

PID Proportional Integral Derivative 9 PLC Programmable Logic Controller 19 PWM Pulse Width Modulation 18

RPCF Rotation Pressure Controlled Feed 3 SDO Service Data Object 17

UFC Universal Flow Control 18 VSM Valve System Module 17

(22)

(23)

Chapter 1

Introduction

Drill rigs are used to drill holes in rock. The drilled holes are normally between 64 to 127 mm in diameter and nearly 30 m in depth during blast hole drilling, for the drill rig type used in this project. A top hammer rock drill is crushing the rock in the drill hole while rotating. A feed force is applied to make sure that the drill tool always have contact with the rock. Blast hole drilling is a mining and construction concept that drills a hole on the rock surface or in an underground environment. The holes are filled with explosives that are detonated to make the rock crack and break up into smaller fragments. The holes are strategically planned beforehand to make the rock crack efficiently for easy excavation. Ore can then be extracted from the smaller fragments and be used for other applications.

When quarrying, the task could be to process limestone into aggregate or cement that are used for civil infrastructure and urban development. If aggregate and cement are combined into a mixture, the compound will be concrete.[1] The demand for pro-cessed limestone is high, since it is also used in products like paper, paint and plastic.[1] During mining, metals such as gold, silver, platinum, copper and zinc are extracted to be used in everyday products, infrastructure or vehicles.

During drilling the drill tool can get stuck in the hole. This creates downtime for the drill rig which is undesirable. The length of the downtime depends. It can take 30 seconds to one hour to get the drill tool loose from the rock. In some cases, it is not possible to get the drill tool out in one piece which will require that a new hole have to be made. When taking this into account, there can be a lot of downtime that prevents the operators to keep the time frame of their work. The time it takes for drilling must be minimized. A faster system could prevent the rock drill from getting stuck which reduce the downtime. A reduced downtime would result in greater drilling economy. The environmental impact from the drill rig would also be reduced, since the drill rig will only use fuel to conduct the drilling instead of using it to get lose from the rock. The rock drill can get stuck caused by a change in friction or other varying rock conditions. At first, the rotational speed of the rock drill will slow down and then the drill string jams in the rock. This problem can be prevented by reducing the feed force when the rotational speed slow down. This project is about evaluating this controlling process with a new valve. The method for the evaluation will be to reconstruct the drill rig’s hardware, on its hydraulic systems, with simulated loads instead of actuators. The feed

(24)

2 Introduction

cylinder and hydraulic rotational motor are disconnected and an electro-proportional valve simulates the load on the rotational motor meanwhile a variable orifice simulates the load on the feed cylinder. This approach will give a controlled test environment and repeatable results while the drill rig will be drilling in air instead of rock. Tests and experiments will be conducted on the drill rig until the hardware and software is ready for a drill test. Then the actuators will be connected to the programmable valve which will then be evaluated while drilling in rock.

1.1 Background

More sensors and electronics are integrated with mechanics which generates high-intelligent and smarter products. There is a continuous demand for short development time and new technologies that increase the productivity and efficiency. This requires constant evaluation of new components. There are new programmable hydraulic valves available on the market which can be set up and be adjusted by software algorithms. These valves show great potential to improve performance and lead time for hydraulic applications.

Figure 1.1: A SmartROC T40 drill rig that is used for blast hole drilling.

Epiroc is a global company that is within the mining and rock excavation technique business area. Surface and Exploration Drilling, SED, is a division within the com-pany. They develop, manufacture, and market rock and exploration drilling equipment for various applications. These applications are in the areas of civil and geotechnical engineering, quarrying and both surface and underground mining. Epiroc wanted to

(25)

1.1 Background 3

evaluate a programmable hydraulic valve to see its benefits and potential when imple-menting it on a drill rig, which can be seen in figure 1.1. The programmable valve used in this project is called Controls Mobile Advanced (CMA) valve, which can be seen in figure 1.3.

Different functions can be used to have good drilling conditions. One function that is used during drilling is called Rotation Pressure Controlled Feed (RPCF). This function has a constant rotational speed and it also aims to keep the rotational torque within given limits. The torque depends on the friction in the hole. The friction is affected by the force pressing down the rock drill and this force can be changed with the feed pressure. Thus, the RPCF function uses the rotational pressure, which is proportional to the rotational torque, to calculate the required feed pressure in order to keep rotational torque at the desired level. The friction will also depend on the rock conditions and the type and state of the drilling tool that is in contact with the rock. The rock conditions are unknown and changing for different depths in the drill hole. If these rock conditions change quickly, the RPCF might not be fast enough to compensate for the change. Then the rotational pressure can reach a limit value which put the drill

pf eed

prot

prot,drill pjamming

pf eed,drill

pf eed,min

Figure 1.2: The figure show how the RPCF function operates. The desired rotational pressure is prot,drill. If prot increases, pf eed decreases to compensates the change in

pressure. This is done between the dotted lines. The rock drill goes into anti-jamming mode if p_rot is higher than p_jamming for a specific amount of time.

rig into a safe mode. The safe mode ensures that the drilling equipment does not get damaged during operation. The safe mode starts if the drill string is jammed and can take away valuable time from the drilling operation if it is not detected at the right time.[2] With the CMA valve, the system response time might become faster compared to the original system. The CMA system should detect anomalies on the rotational pressure faster and compensate the rotational pressure with the feed pressure before the drill rig enters the safe mode.

(26)

4 Introduction

1.2 Purpose and Objectives

The purpose of this project was to evaluate the use of a CMA valve, made by Eaton, in drill rig applications. This was done to see if the functionality of the drilling procedure could be improved with the CMA valve. The evaluation was made by comparing the CMA valve with the original valve system. The objectives were the following questions: • How fast is the response time, the time that it takes to detect and act on differen-tiates in rotational pressure during varying rock conditions, of the CMA system compared to the original system?

• Can the CMA valve be used in drill rig applications?

• Are the original control modes of the CMA valve enough for good drilling perfor-mance?

• Can the CMA system handle a drilling operation well (the drilling is smooth and continuous)?

The objectives that was studied should contribute in making the RPCF work more efficient, prevent unwanted downtime and make the drilling operation smooth and continuous to increase the drill rig’s productivity.

Figure 1.3: The programmable CMA valve from Eaton. [3]

1.3 Limitation

• The custom control algorithm was implemented on the CMA valve by the valve manufacturer.

• The time frame for this project was until June 2018.

• A PLC was used to control the CMA valve. CoDeSys, version 2.3.9.42 was used to program the software on the PLC. The CANopen standard was used for the communication between the PLC and the CMA valve.

(27)

1.3 Limitation 5

• This project only evaluated one programmable valve and compared it to the original valve system on a drill rig from Epiroc. The tests were only conducted on a specific drill rig, the smartROC T40, with a top-hammer rock drill that is used for blast hole drilling and only on the RPCF function.

(28)

(29)

Chapter 2

Theory

2.1 Drilling

The four hydraulic principles required for rock drilling are percussion, rotation, feed and damping. Some of the principles can be seen in figure 2.1.

Figure 2.1: Illustration of the principles required for rock drilling.[4]

Percussion uses a piston. The force that accelerates the piston is the pressure multiplied by the working area. The magnitude of the impact energy depends on the mass and velocity of the piston. The velocity depends on the working pressure and stroke length of the piston. The piston is released in a defined frequency and generates an impact force that is crushing the rock. The rotation is made by a hydraulic motor that is connected through a gearbox in the rock drill. The rotation makes sure that new uncrushed rock is crushed with each impact stroke. If the rotation speed is too high or low, it results in poor drilling economy and quick ware of the tool that is penetrating the rock. The rotational speed is set depending on the rock type and its characteristics. Feed force is used to overcome the impulse reflex from the impact and is also working as a normal force to keep the drilling tool in constant contact with the rock. The feed

(30)

8 Theory

force is generated by the feed pressure that is pressing down the feed piston along with the weight of the drill string and rock drill. In good drilling conditions, the feed force is kept constant regardless of the drill depth or rock type. A feed force that is too high causes the rotational speed to decrease and it can also have an impact on the shape of the hole. Too low feed force causes the drill string to vibrate and the penetration rate is decreased. A hydraulic damping system is used to absorb impulse wave reflex from the rock and is also used to see if there is contact between the drill tool and rock. When the actual drilling is about to start, the rock drill is set to collaring mode. This mode decreases the percussion and feed force when the drill string is collaring towards the rock. When the drill string has contact, collaring mode is switched to drill mode and the percussion and feed is increased to forces that are optimal for good penetration rate.

2.1.1 Rotation Pressure Controlled Feed, RPCF and Jamming

The RPCF function is used to protect the drill tool from getting stuck in the hole or loosing contact with the ground. The feed pressure is actively changing depending on the rotational pressure in order to obtain good drilling conditions. For example, when the rotational pressure is increased to a level above normal condition, due to deviation of friction or difficult rock types, the RPCF function reduces the feed pressure to counteract the increase of the rotational pressure. If the RPCF fails to compensate the feed pressure, the drilling tool can be jammed into the rock. The characteristics of the original RPCF can be seen in figure 1.2.

The rotational pressure has a maximum limit when the RPCF fails to compensate for the increasing rotational pressure. When the limit is reached, the drill rig detects jamming and enters anti-jamming mode. In this mode the drill string is retracted by reversing the direction of the feed flow until one of two things happens. Either the rotational pressure is low enough to begin the drilling operation again or the rock drill has been in anti-jamming mode for too long. If the time limit for the anti-jamming mode is reached, the rock drill enters an idle state and waits for a manual measure by the operator. It is not desirable to enter anti-jamming mode unless the drill string has actually jammed. However, when jamming is detected, it is important to enter anti-jamming mode quickly enough to prevent the drill tool from getting stuck in the drill hole.

During drilling it is important that the operator is focused. By listening for different sounds the operator can hear if something is wrong in the hole. A ringing sound occurs if the threads are not tightened enough which entails that the rotational torque is too low. A swooshing sound occurs if the drilling tool has gone loose. The crushed rock that is extracted from the hole is analyzed visually. It is normally smaller than the size of a thumbnail. The feed pressure need to be reduced if the crushed rock is bigger. [2]

2.2 Hydraulic Valve

A hydraulic valve’s function is to control the the flow through it. The flow is regulated by orifices. The flow can have different characteristics depending on the velocity, orifice

(31)

2.3 Control 9

geometry and the viscosity of the fluid. The viscosity of the hydraulic oil is temperature dependent. The temperature varies due to the working temperature of the drill rig and the pressure drop over the valve. A temperature rise of 15◦C can reduce the viscosity to half.[5]

The orifice controls the flow or pressure of the hydraulic oil generated by a hydraulic pump. The actuators are affected by the friction between the drilling tool and the rock which varies with different rock types and cavities. Since the friction is changing and occasional cavities in the rock can occur, the valves need to compensate the flow and pressure for this. If the valves fail to do it quickly enough, the drill rig will enter a safe mode and stop the drilling.

A conventional valve uses electrical current to control the spool position and external sensors to regulate the pressure and flow in the system. The CMA valve has on board electronics and built-in sensors which eliminate the need to integrate them. The sensors measure pressures, spool positions and temperatures which help detect anomalies in the system faster. The demand is controlled by CAN to achieve functionality and high precision on the machine. The CMA valve has a software interface that can be used to tune the valve’s performance electronically instead of having parts manually calibrated, which saves time. Each spool can control its designated work port because of the independent metering. This allows a smooth and effective load control. The sensor feedback allows to use different control modes in the same valve. It is also beneficial for the control algorithms used in mobile applications.

2.3 Control

A controller measures a systems output signal and modifies its input signal in regard of an error value. Commonly, for industrial applications, a Proportional Integral Deriva-tive (PID) controller is used. It can be modified to be more flexible to the system it is used for. It uses a proportional gain, P, integrational gain, I, and derivative gain, D. A common expression of it can be seen in equation 2.1.[6]

u(t) = Kp e(t) + 1 Ti t Z 0 e(τ )dτ + Td de(t) dt (2.1) 2.3.1 Ziegler-Nichols

The Ziegler-Nichols method for tuning a PID controller can use the Ziegler-Nichols model to do it. It will give starting values of the controller. The model is derived from an open loop step response, with a unit step.[7] The structure of the model can be seen in equation 2.2.

G(s) = b se

−sL _(2.2)

Parameters can be determined by using the open loop step response. The parameter

(32)

10 Theory

can be seen in figure 2.2. If the input signal is not a unit step response (from 0 to 1), the coefficient has to be divided with the magnitude of the step.[7]

L b

time output

Figure 2.2: The model is derived from the step response as shown in the graph. L is the time delay and b is the steepest gradient.

Ziegler-Nichols discrete time PID is described in equation 4.1.[8] The tuning pa-rameters are based on the model papa-rameters and can be seen in table 2.1.

Table 2.1: Ziegler-Nichols tuning method for a PID-controller is shown in the table, based on a Ziegler-Nichols model from a step response.[6]

Parameter PI controller PID controller

K 0.9/b L 1.2/b L

Ti 3L 2L

Td 0 L/2

The model used in this method does not specify what is happening during steady state. When using this method, it is more important what happens during changes in the system.

2.3.2 Time Delays

Time delays occur in different parts of a system. In this project, the time delays occur in the communication system, the mechanical system and the control program. For a fixed communication cycle, the messages are sent continuously at a specific time. When a value changes during the cycle, it will not be sent until the next communication cycle has started. A maximum time delay of one communication loop may occur because of this. The mechanical time delay is the time it takes to move the mechanical parts in the system. It takes one program cycle to calculate the new values after receiving the updated values.

A time delay in a system results in a negative phase shift.[9] The phase shift results in a smaller phase margin which limits the maximum stable gain of the controller. The limitation of the gain results in a slower controller. One way to reduce the effect of a

(33)

2.3 Control 11

time delay is to use a Smith predictor. The Smith predictor changes the feedback to the controller. This creates a controller that attempts to eliminate a time delay on the feedback.

Gc=

F Ge−sL

1 + F ˆG(1 − e−sL_{) + Ge}−sL (2.3)

The closed loop transfer function Gc can be calculated with equation 2.3. In the

equation, G is the systems transfer function without time delay and ˆG is the model of

that transfer function. L is the time delay and F is the controller.

Gc=

F Ge−sL

1 + F G (2.4)

If the modelled transfer function ˆG is equal to the system transfer function G, the time

delays in the denominator cancel each other out. This leads to the closed loop transfer function described in equation 2.4. This closed loop transfer function has the same time delay as the open loop transfer function.

The Smith predictor requires an accurate and linear model of the system to be implemented in the controller.[10] There are ways to extend the Smith predictor to nonlinear systems like a hydraulic system.[11] However, this is a complex procedure which requires more computational power. Because of this, it is desirable to minimize the time delay in a system. Especially if the system can not be modelled.

(34)

(35)

Chapter 3

System Overview

3.1 Drill Rig

The system is powered by a diesel engine. The engine powers the hydraulic system through hydraulic pumps. There are three pumps supplying the hydraulic system. The hydraulic oil used in the drill rig had a viscosity of 46 cSt at 40 ◦C.[12] Inside the operator cabin, there was screen with a graphical user interface that was used to enable drilling functions and set drill rig parameters such as rotational speed and different pressures.

3.2 Feed Boom

The feed boom consists of several components such as the rock drill, button bit, drill rods, shank adapter and coupling sleeves. They can be seen in figures 3.1 and 3.2. The button bit is the drill tool that crushes the rock with the impulse wave generated by the percussion. Drill rods are used to connect the button bit with the rock drill. The drill rods are threaded to coupling sleeves to connect drill rods to each other and to the shank adapter. When the drill hole gets deeper, more drill rods are connected to create a longer drill string.

(36)

14 System Overview

Figure 3.1: The feed boom.[4]

The rock drill is located on the feed boom which can be seen in figure 3.1. The rotation, percussion and damping subsystems is located inside the rock drill. The feed subsystem is located on the feed boom. These four hydraulic subsystems each have their purpose for the drilling principles showed in figure 2.1.

Figure 3.2: The rock drill components. 1: Flushing system. 2: Damper. 3: Percussion piston. 4: Shank adapter. 5: Gearbox. 6: Rotational motor.[4]

The purpose of the rotational subsystem is to move the button bit to uncrushed rock. The subsystem is powered by one of the hydraulic pumps, which is called pump 2 in the report. The rotational motion comes from a hydraulic motor that is connected to a gearbox and then to the drill string. This motor receives a constant flow to keep a constant rotational speed during the drilling. In figure 3.2 the rotational motor can be seen on the rock drill.

The feed subsystem makes sure that the drill string always has contact with the rock. The feed subsystem is supplied by a different pump than the rotational subsystem, which is called pump 1 in the report. The actuator of the feed subsystem can be a cylinder. This cylinder is controlled by a pressure controlled valve, which means that

(37)

3.2 Feed Boom 15

the valve outputs a constant pressure corresponding to the control signal. The feed cylinder can be seen in figure 3.1.

The percussion subsystem creates the impact force that generates the impulse wave through the drill string. The percussion subsystem is supplied by the same pump as the feed subsystem, pump 1. It consists of a piston and a control valve. The valve increases the pressure behind the piston to accelerate it. The piston generates an impact force, when hitting the shank adapter, that will travel through the drill string down to the button bit. The percussion piston can be seen on the rock drill in figure 3.2.

The damping subsystem damps out the impulse wave reflex. It is supplied by the same pump as the feed- and percussion subsystem, pump 1. The damper can be seen on the rock drill in figure 3.2.

(38)

16 System Overview Pump 1 Pump 2 WPA WPB Valve section 1 Valve section 2 Valve section 3 Valve section 4 WPB WPB WPB WPA WPA WPA Percussion piston Damper Feed piston Rotational motor

Figure 3.3: The figure show how the CMA valve was connected to the pumps and actuators. The valves shown in the schematic are controlled with pilot valves.

(39)

3.3 Communication 17

3.3 Communication

The information that is sent and received by the valve need to be packed and unpacked according to the CMA valve’s communication protocol. The CMA valve uses a CAN network with the CANopen standard. In this project the drilling functions controlled by the CMA system have a separate communication system from the rest of the drill rig. This communication system have a shorter loop time, which has to be taken into account when evaluating the response times for the different systems.

3.3.1 Controller Area Network, CAN

Controller Area Network (CAN) is used to send and receive information with a Central Processing Unit (CPU). In this application there are two CPUs, one for each node of the CAN network. The first node is the PLC, which is used to initialize the CAN network. This node also has the user interface. The other node is the CMA valve’s VSM, which is the interface from the internal CAN network on the valve. The information is transferred in a specified baud rate by using two wires, CAN-high and CAN-low. Signal reflection is suppressed by having a 120 Ω resistance to terminate each end of the bus line. The information is placed on bits that are either a one or a zero. A CAN message is sent as an array of bytes (each byte is eight bits). The messages can either be sent by the type Process Data Object (PDO) or Service Data Object (SDO). The structure of these messages can be seen in appendix C. The PDOs are sent continuously during operation meanwhile the SDOs can only be sent during the pre-operational mode when the valve in inactive.

3.4 Electro-proportional Valve

An electro-proportional valve was used to simulate a load on rotation and replaced the rotational motor during response time measurements in this project. This valve is controlled with a current controlled PWM signal which will result in a spool position. There is a current feedback on the PWM signal.

3.5 CMA Valve

The CMA valve is the programmable valve that was compared with the original valves that are controlling the rock drills hydraulic subsystems. It is a CAN-enabled electro-hydraulic sectional mobile valve and can be seen in figure 1.3. The valve has predefined control modes that are setup by a controller on a PLC. It also has a custom made control mode for the RPCF drilling function that was built-in to the valve’s VSM. A cross-section of one valve section can be seen in figure 3.4. The valve measures supply pressure and uses it for control. The Valve System Module (VSM) is the interface module for the valve that is acting like a CPU. The VSM is used as a CAN gateway and is the main controller in the valve system. This means that there was one internal CAN bus for the valve and one external for the drill rig. The advantage with the VSM is the internal high speed closed loop control which give the valve a faster response

(40)

18 System Overview

time. By using a separate CAN bus, the load on the drill rig’s CAN bus is reduced. It uses sensors along with software control algorithms to make independent metering of multiple states. There are pressure sensors situated on all load sensing line, pressure line, tank line and workports which enables monitoring and pressure control.

Figure 3.4: Cross section of a work section on the programmable CMA valve from Eaton.[3]

The setup of the valve enables flow sharing, passive and overrunning load control. The Intelligent Flow Control (IFC), is a twin spool controller that enables flow con-trol on one workport and pressure concon-trol on the other. Flow concon-trol is used on the inlet workport and pressure control is used on the output workport when the load is passive. Flow control is used on the output workport and pressure control on the in-let workport when the load is overrunning.[13] The Universal Flow Control (UFC), is a twin spool mode during which both spools operate in a pressure compensated flow control mode.[13] The valve has the possibility to read the load demand of the rotation pressure difference from one section to control feed pressure in another section. This is only possible if the motor load is passive during the entire operation.

The programmable valve has two spools that can be working together to control double acting functions in each work section. They can also be working separately and be controlled from any of the work sections. An independent pilot spool is controlling the main stage spools and closed loop control can be made for each work section locally by sensors.

Six different control modes (not including the custom RPCF mode) can be chosen depending on how applicable they are to the RPCF function. The six modes are flow, pressure, spool position, PWM, float and idle. In flow control mode, the valve delivers a constant flow out of the valve without regard to the pressure. The pressure control mode keeps a constant load pressure for the feed pressure. It adjusts the flow to make sure that the pressure do not drop below the required level. The spool position control mode allows the user to control the system by changing the desired spool position. Pulse Width Modulation (PWM) lets the user specify how much of the maximum

(41)

3.6 PLC 19

control current that should be used to control the spool (a value between 0 and 100%). Float control mode opens both of the spools to get a specific flow until the pressure in both workports is dropped below a set pressure. When the pressure is reached, both spools is opened completely. Idle is the natural state of the valve. If the valve stop receiving command messages, it enters this state after 200 ms.[14] In this mode the valve is closed.

The CMA valve uses the supply pressure in the control algorithms. It has a built-in sensor that measures this pressure. The system is designed to be supplied by the same pressure for all sections. In this drilling application, the rotational valve section is supplied by a different pump. Because of this, the rotational work section will be positioned nearby pump 2 that is supplying the rotational subsystem. The supply pressure to this work section has to be measured and sent via the VSM with a special SDO that can be sent during operational mode. The other work sections, along with the VSM, that are controlling the feed, percussion and damper were positioned on the feed beam. A hydraulic schematic of the system can be seen in 3.3.

3.6 PLC

The Programmable Logic Controller (PLC) used in this project was a IFM CR7132. It was used to control components for the thesis. The PLC has 4 MB of flash memory, which was enough for the thesis since only a few functions was programmed on to the PLC. The PLC was programmed using CoDeSys V2.3.9.42. The electro-proportional valve used to simulate the load in the rotational subsystem was controlled with a current. This current could be controlled with a resolution of 1-2 mA.[15]

An Electronic Data Sheet (EDS) file was loaded onto the PLC for the CAN com-munication with the programmable valve. This file contained predefined PDOs that was used to write control programs in CoDeSys.[16]

(42)

(43)

Chapter 4

Control

This project worked with an actual drill rig. This required that safety functions were coded into the controller. The controller had an idle state. When the idle state was entered, all the hydraulic systems were shut down. The feed system was load holding while the others subsystems pressures were set to zero.

Software for the CMA system was developed for the tests in this project. The controller made for drilling automatically controlled the feed pressure to keep a constant rotational pressure. Another controller worked similar as the original RPCF function, see figure 1.2. This controller was used to compare the response time of the CMA valve system with the original system. Software for control of the damping and percussion was also developed.

4.1 Control Algorithm

The control algorithm was designed to be stable and fast. Thus, the derivative of the control signal, du_dt, was not limited more than the standard limitation in the CMA valve. The standard limitation in the CMA system contained a maximum of 300 bar/s in pressure change or a taper off filter with a filter frequency of 3 Hz. The signal was processed through both filters and the method with the smallest difference between the old value and the new value was used for that sample. Oscillations in the system were undesirable, but if the oscillations had a high frequency and a limited amplitude, it might have been damped out by the hydraulic system. The controller was tuned so that the oscillations would not induce movement in the mechanical parts. The energy of the control signal was not important for this controller. The reason for this was that the feed pressure were supplied by the same pump as the percussion that had a much higher pressure. Thus, the feed pressure was not affect by the pressure settings on the pump.

4.1.1 Rotational Pressure Controlled Feed

The rotational reference pressure was calculated for a specific drill rig, drilling condition and hardware setup. This value usually does not change during the drilling operation. Thus, the RPCF function could be treated as a regulator problem instead of a servo

(44)

22 Control

problem. This means that the controller should keep the controlled signal on a constant level and eliminate the effects of disturbances on the system. For a regulator problem, a controller that uses the control error is suitable. In this case, the control error was the difference between the actual rotational pressure and the reference rotational pressure. Other possible control features was a feedback from the disturbance. For a drilling application, the disturbance can be the friction in the hole and the flow out of the feed cylinder. The friction in the hole can be estimated with measurements of the feed and rotational pressure, but it was changing in an unpredictable way that make it unsuitable for this application. Measuring the flow is considered easier since it does not change as much. However, because of the low flows to the feed cylinder, the flow measuring method was not accurate enough for this application. Thus, there was no feedback from disturbances in the controller used in this project.

4.1.2 Controller on the PLC

Frot Ff eed Gf eed Grot

erot

prot,ref

LP-filter

Made by valve manufacturer

pf eed,ref pf eed prot

Figure 4.1: A schematic over the transfer function that was implemented for the CMA valve. Frot is the outer controller that outputs the reference feed pressure. Ff eed is

the inner controller, made by the valve manufacture, which outputs the feed pressure. LP-filter is a low pass filter.

The PLC controller could only use the six predefined valve modes. For the RPCF function, the pressure control mode and the flow control mode were used during normal operation. The flow control mode was used for the rotational motor and the reference flow was set to 40 l/min. This control mode was chosen to keep a constant rotational speed for the rock drill. The rotational pressure was controlled using a cascade loop. The outer controller (F_rot) was a discrete time PID controller, which can be seen in equation 4.1.[8] uk= Kp ek+ TS Ti k X n=0 en+ Td ek− ek−1 TS (4.1) By using the valve manufacturer’s pressure control mode as the inner control loop, the outer controller only had to provide a reference feed pressure. The changed feed

(45)

4.1 Control Algorithm 23

pressure would then affect rotational pressure through the change of friction in the hole. By saturating the feed pressure control signal at the maximum and minimum feed pressure limits, the feed pressure was contained. This can be seen in equation 4.2. Anti windup was added to the integrational part of the PID if the control signal would be saturated. pf eed,ref =       

pf eed,max, if u(t) > pf eed,max

u(t), if pf eed,min< u(t) < pf eed,max

pf eed,min, if u(t) < pf eed,min

(4.2)

To avoid instability, the outer control loop was made slower than the inner control loop. The inner control loop, in this case, was the pressure control mode. A schematic of the controller can be seen in figure 4.1.

A deadband for the acceptable error can also be used which can be seen in equation 4.3. This was used since the demanded pressure of the outer controller was discrete. The precision of the control signal was 0.125 bar.

ek= (

0, if |erot| < pdeadband

erot, if |erot| ≥ pdeadband

(4.3)

4.1.3 Controller on the VSM

A new control mode was created for the VSM controller. This mode used the internal communication loop of the CMA valve for the outer controller (Frot). The same control

strategy was applied here as in the PLC controller. However, the tuning parameters differed since the program loop time was different. A schematic over this controller can be seen in figure 4.1. The valve manufacturer implemented the derived code for this controller based on the specification that was worked out for this project.

4.1.4 Anti-jamming Mode

The drill rig enters the anti-jamming mode if the filtered rotational pressure reach the limit for jamming. This means that the regular RPCF controller is overridden by the anti-jamming controller. This can be seen in figure 4.2. The anti-jamming controller used the pressure control mode of the CMA valve. It sent a constant reference feed pressure to the valve, with a reversed flow direction. This controller was only programmed on to the PLC.

(46)

24 Control

START

RPCF

prot> pjamming Anti-jamming

prot< pexit,jamming

TRUE

FALSE

TRUE

Figure 4.2: The flowchart shows how the controller starts and how it switches between modes during operation.

4.1.5 PID Tuning

The friction in the system was unpredictable and also expected to change a lot, which means that the system would not stay in steady state for a long time. Thus, the Ziegler-Nichols method was deemed to work well with this application. The method was used with the Ziegler-Nichols model generated from a step response. Another method that could have been used was an oscillation test, since the system handle high pressure the step response method were considered to be more safe.

4.2 Set up of PLC

ProFX (software from the valve manufacturer) was used to reconfigure CAN-addresses to be able to receive the rotational pressure to programs in the PLC. An EDS was provided by the valve manufacturer and implemented in the PLC program. This file helped with the initialization of the CAN network. The programming of the PLC was done with function block diagrams and structured text. Programs were created for most cases, but the packaging of CAN messages was created by function blocks. The function blocks could be used in more than one program.

The CMA valve used the supply pressure for the control modes. If the supply pressure differed from the measured value, the valves control modes might not work

(47)

4.3 Control Required for Drilling 25

properly. Since this project used two different pumps with different pressure settings to supply the CMA valve, the PLC had to send a measurement of the supply pressure for the rotation to the VSM. The supply pressure from the other pump was measured by a sensor inside the CMA valve.

4.3 Control Required for Drilling

During actual drilling, feed and rotation were not the only drilling principles that were required to be controlled. There was also a need to control the percussion and the damping, which is stated in section 2.1. The percussion was controlled to a constant pressure and the damping to a constant flow. The drill- and the anti-jamming modes were the only modes used during the hardware simulation. However, for the drill test, a few more modes had to be developed. A manual mode was developed to be able to turn the percussion on and off and to change the rotational flow and feed pressure. In order to start the drilling mode, the rock drill would first enter collaring mode. In collaring mode, the percussion pressure and the feed pressure were lower than during drilling. The RPCF was disabled since the low feed pressure would not give the desired rotational pressure. The feed pressure was instead kept at a constant level. The rotational flow and damping flow were the same in the collaring mode as in the drilling mode. An idle mode was created to tie all these modes together. The idle mode was also the starting mode that the drill rig entered after initialization. From the idle mode it was possible to update parameters that were sent to the CMA valve as SDOs. A flow chart over these modes and how to move between them can be seen in figure 4.3.

Since the percussion used a much higher pressure than the rotation, the two pumps supplied two different pressures levels. This required that a rotation supply pressure measurement was sent to the CMA valve. It was sent using an SDO, which would hold the value until the previous value got overwritten by the new value.

4.3.1 Implementation of Manual Control

The joysticks were configured to output a zero value in center position. The value was proportionally increased until it reached the end position which outputted the specified maximum value. The two different end positions corresponded to the two different flow directions. This enabled the drill rig to be able to switch the feed pressure and rotational flow from positive to negative direction in a smooth, controlled operation. The switch enabled the percussion to be on or off and the percussion pressure parameter could be set in the graphical user interface if different pressures were required.

(48)

26 Control Start Initialize Idle Manual Activate Activate Activate Collaring Drilling 10 seconds Anti-jamming prot > pjamming prot< pexit,jamming Runtime True True True True True True True False False False False False False Manual Collaring

Drilling in collaring exceeded

Figure 4.3: The flowchart how to switch between the different drill modes during an actual drilling. If the Anti-jamming runtime was exceeded, the operator had to enter the manual control mode to reset the warning. Otherwise the drill rig could not enter collaring mode. The manual mode was used by the operator to get the drill bit loose in event of jamming.

(49)

Chapter 5

Testing

Conducting tests on the drill rig required several safety aspects to ensure that nothing was damaged or affected the surroundings or the people operating it. Proper safety equipment was used when the drill rig was running such as safety glasses, proper cloth-ing, gloves and shield walls to prevent any hydraulic oil injection accident.

In this chapter the different hardware setups are described along with the different tests. It summarizes why the tests were done and how they were executed. The re-sponse time test evaluated, with a repeatable result, how fast the CMA valve system was compared to the original valve system. The tuning test evaluated if the controller worked effectively. The anti-jamming test evaluated if the CMA system worked cor-rectly and safely during jamming. The drill test was performed last to see if the valve could handle a drilling procedure using the result of the tuning- and anti-jamming test.

5.1 Hardware Setups

There were two different hardware setups for the testing equipment. The first setup was used to perform hardware simulations on both the original system and the CMA system. The second was used to perform tests in a real drilling environment with the CMA system.

5.1.1 Wiring

Shielded wires were used for CAN connections to prevent unwanted disturbances on the measurements.

The VSM connection cable was split into two parts. The first part was a powercable that was connected to a lamp outlet (24 V). The second part was the CAN connection that could be connected to either the PLC or the CAN-card. The valve section for the rotation was separate from the other valve sections. This required a long CAN-cable to connect it to the internal CAN bus. The PLC was powered from a 24 V cigarette outlet on the drill rig. The wiring diagram for the cables can be seen in appendix B.

(50)

28 Testing

5.1.2 Hardware Simulation

The hardware simulation setup was used to have a controlled environment. The drilling procedure was modeled using hardware and software. The tests performed with this setup had to be repeatable and have the possibility to compare the results for the different valve systems. A schematic of how to setup the hardware is shown in figure 5.1.

iv = K · pf eed+ M (5.1)

The electro-proportional valve was controlled with a current calculated from equa-tion 5.1. The loads were simulated with orifices. Both the feed piston and the rotaequa-tional motor were disconnected. The feed piston was replaced with a variable orifice and the rotational motor with an electro-proportional valve. The feed beam was tilted to a horizontal position, showed in figure 5.2, that enabled easy access to the valve blocks and hydraulic hoses.

Test Equipment

• A portable data logger. It was used to read and record flow, pressure and tem-perature using the existing channels.

• An electro-proportional valve. It was used to be able to simulate the counteracting torque when drilling with different friction interference by changing the rotational pressure.

• Variable orifice controlled by levers. It was used to simulate a normal force from the rock.

• Adapters. They were used to connect the equipment and hoses. • Hoses. Used to connect the equipment.

• A flow sensor. It was used to measure the flow rate.

• Pressure sensors. They were used to measure feed and rotational pressure. • Temperature sensor. It was used to read the working temperature of the hydraulic

oil.

• Display. It was used to read and change the parameter for the current reference. • PLC hardware. It was used for controlling the electro-proportional valve.

• M12 wires. They were used to connect the sensors to the portable data logger’s channels.

• CAN-card. It was used to enable the software to communicate with the CMA valve.

(51)

5.1 Hardware Setups 29 Connect p q p p T Connect Connect Connect Control valve Rotation Control valve Feed PLC

Figure 5.1: The figure shows a schematic of how the hardware was set up for the hard-ware simulation. There was one flow sensor and one pressure sensor for measurements on the rotational subsystem. Then there was two pressure sensors for measurement of the feed. One for the PLC and one for the data logger. The connect blocks were quick couplings which makes it easier to switch between the original and the CMA control valves.

Assembly of Hardware

The hoses that connected the outgoing ports of the feed valve block to the feed piston were removed and thereby disabled the feed piston. Instead, a feed subsystem was made that consisted of a variable orifice with a control lever, see figure A.4 in appendix, connected in series with a pressure and temperature sensor which can be seen in figure A.5 in appendix. The schematic of the feed subsystem can be seen in figure 5.3. The sensors were connected to the data logger using M12 wires. They can be seen in figures

(52)

30 Testing

Figure 5.2: The feed beam was put in a horizontal position to enable easy access to the equipment.

A.1 and A.6 in appendix. The temperature was recorded to be able to ensure that the hydraulic oil had reached working temperature before all tests. Also, another pressure sensor was assembled in series with the other sensors that was connected to the PLC. The PLC then uses this to control the electro-proportional valve. The assembled feed subsystem can be seen in figure 5.4. The hoses that were connected to the rotational motor ports were also disconnected. This disabled the rotational motor and prevented unwanted disturbances when the shank adapter rotates. Instead, a rotational subsystem was made that consisted of the flow and pressure sensor which was connected in series with the electro-proportional valve. The rotational subsystem can be seen in figure 5.5. Both of the sensors were connected to the data logger. The function of the electro-proportional valve was to change the rotational pressure. It was only work port A, tank port and pump port that were used on the electro-proportional valve and the rest of the working ports were plugged. The assembled rotational subsystem can be seen in figure 5.6. A display was connected to the PLC. This display (if enabled) set the current value parameter which the PLC used to control the electro-proportional valve. The display changed the parameter with five milliampere increments. It was programmed using CoDeSys and initial values were set near the desired working area.

(53)

5.1 Hardware Setups 31 p p Quick coupling female T To PLC To data logger Quick coupling female

Figure 5.3: Schematic of the feed subsystem containing pressure and temperature sen-sors with a variable orifice. An image of the quick coupling can be seen in figure A.7 in appendix.

Figure 5.4: The assembled feed subsystem containing pressure and temperature sensors with a variable orifice controlled by a lever.

the rotational pressure changed. This required that the rotational pressure was higher than prot,drill and lower than pjamming (see figure 1.2). The rotational pressure was

controlled with a control current for the electro-proportional valve. By staying inside this pressure range there were no internal delays and the anti-jamming function did not start. The drill rig had its rotational flow calibrated to maintain a rotational pressure within the range of the RPCF. The feed pressure were also calibrated to a level within the range of the RPCF.

(54)

32 Testing p q Quick coupling female Quick coupling female T o data logger From PLC

Figure 5.5: Schematic of the rotational subsystem. The quick coupling can be seen in figure A.7 in appendix.

Figure 5.6: The assembled rotational subsystem.

(55)

5.1 Hardware Setups 33

to connect them in parallel with each other. This was done by adding a T-coupling from the pump to the supply and tank hoses. From this, T-coupling hoses were connected to both control valves. See figure A.8 in appendix. The feed piston subsystem was modified to have two quick couplings that could be pulled off one valve and put on the other. The quick couplings can be seen in figure A.7 in appendix. The valve, that was not used, was thereby plugged by using the quick couplings. The supply- and tank hoses from the rotational pump were also T-connected and branched to both work sections that control the rotation, to the original- and the CMA valve. The rotational subsystem was also given quick couplings.

The percussion and damping were disconnected to make the hydraulic system simple and to make sure that the tests were not interfered with external factors. All hydraulic hoses and work ports were converted by adapters to 3/4" in diameter to have the same dimension for the response time tests. The length of the hoses that were T-connected had a similar length to not make any time delays occur during tests.

5.1.3 Hardware Drilling Environment

This setup was used for an actual drilling environment. The tests performed with this setup generated a perception on how the CMA valve works during a drilling operation.

Test Equipment

• A portable data logger. It was used to read and record flow, pressure and tem-perature using the existing channels.

• Adapters. They were used to connect the equipment and hoses. • Hoses. They were used to connect the equipment.

• Pressure sensors. They were used to measure pump, feed and rotational pressure. • Temperature sensor. It was used to read the working temperature of the hydraulic

oil.

• PLC hardware. It was used for controlling the drill functions.

• CAN-card. It was used to enable the software to communicate with the CMA valve.

• M12 wires. They were used to connect the sensors to the portable data logger’s channels.

• Joysticks. They were used to manually operate the rotation and feed. • Switch. It was used to manually operate the percussion.

• Valve attachment. It was used to mount the valve in a safe position on the feed beam.

(56)

34 Testing Supply Pressure for Rotation

In this test the supply pressure for the rotation was measured and sent to the CMA valve. This allowed the two pumps to produce different supply pressures.

Assembly of Hardware

In order to make a drill test with the CMA system, the drill rig made several hardware changes. The feed piston, rotational motor, percussion and damping were connected to the CMA valve. The CMA valve was mounted to the feed beam with a custom made attachment. The attachment can be seen in figure A.11 in appendix. This was done to secure the valve when it was positioned vertically on the feed beam during drilling. The assembled hardware can be seen in figure 5.7 and 5.9. The pressure sensor for rotation was assembled on the inlet of the rotational motor which can be seen in figure 5.8.

Figure 5.7: The CMA valve assembled on the attachment with percussion, damping and feed actuators connected to it.

The hydraulic subsystems needed the ability to be controlled manually in case that jamming would occur or if the drill string needed to get up from the hole during the test. Two joysticks and a switch were provided to be able to control the feed, rotation and percussion, which can be seen in figures A.9 and A.10 in appendix. Flat pins was assembled to them in order to connect them to the PLC.

(57)

5.1 Hardware Setups 35

Figure 5.8: Down in the left hand corner of the figure, it can be seen where the rotational pressure sensor was assembled on the rock drill.

Figure 5.9: The CMA valve rotational work section assembled on the drill rig with the rotational actuator connected to it. Two pressure sensor are located on the inlet work port that is connected to the PLC and data logger for monitoring of the different pressures.

Evaluation of a Programmable Hydraulic Valve for Drill Rig Applications

Evaluation of a Programmable Hydraulic Valve for

Drill Rig Applications

Evaluation of a Programmable

Hydraulic Valve for Drill Rig

Applications

Jonathan de Brun Mangs and Mikael Tillquist

Abstract

Acknowledgments

Contents

List of Figures

List of Tables

Nomenclature

Acronyms

Chapter 1

Introduction

1.1

Background

1.2

Purpose and Objectives

1.3

Limitation

Chapter 2

Theory

2.1

Drilling

2.2

Hydraulic Valve

2.3

Control

Chapter 3

System Overview

3.1

Drill Rig

3.2

Feed Boom

3.3

Communication

3.4

Electro-proportional Valve

3.5

CMA Valve

3.6

PLC

Chapter 4

Control

4.1

Control Algorithm

4.2

Set up of PLC

4.3

Control Required for Drilling

Chapter 5

Testing

5.1

Hardware Setups