Evaluation of a digitial displacement pump in a load haul dump application

(1)

LIU-IEI-TEK-A--19/03563—SE Spring 2019

Linköping University SE-581 83 Linköping, Sweden +46 013 28 10 00, www.liu.se

EVALUATION OF A

DIGITAL DISPLACEMENT

PUMP IN A LOAD HAUL

DUMP APPLICATION

JAYASURYA MADHUSUDANAN

Supervisor: Samuel Kärnell

Examiner: Liselott Ericson

(2)

2

COPYRIGHT

The publishers will keep this document online on the Internet – or its possible replacement – for a period of 25 years starting from the date of publication barring exceptional circumstances. The online availability of the document implies permanent permission for anyone to read, to download, or to print out single copies for his/hers own use and to use it unchanged for non-commercial research and educational purpose. Subsequent transfers of copyright cannot revoke this permission. All other uses of the document are conditional upon 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 procedures for publication and for assurance of document integrity, please refer to its www home page: http://www.ep.liu.se/.

(3)

3

ABSTRACT

Hydraulics has always been the first choice of actuation in off-road, construction and mining vehicles due to its power density, low cost, built in cooling and lubrication. However, the current state of our environment along with stricter regulations has brought light to newer technologies within hydraulics to improve the existing system. This urge to enhance efficiency and reduce energy consumption has led to a point where new technologies must be evaluated. One such technology is the programmable hydraulic pump called the digital displacement pump (DDP). This new pump may have the potential to revolutionize mobile hydraulics as it can be used to improve part load efficiencies, response and make it easier to control from a system perspective.

The DDP is a radial piston pump that has been fit with solenoid on/off valves at the inlet of each cylinder to control the flow of the working fluid. The displacement setting of the pump depends on the displacement of each cylinder controlled digitally by the 'active' inlet valve. The pump can act as a single unit to supply one circuit or it can dedicate pistons for supplying several circuits in parallel using different pump outlet configurations. They can be setup to run in pressure controlled or flow controlled systems to achieve the above mentioned flow sharing capability.

An energy study based on two fixed drive cycles (short and intermediate) are conducted on the existing system of a loader used for mining called the ST14 Battery. A breakdown of the energy consumption in the machine is created to look at the impact of the three main actuators (boom, bucket and steering), pump losses and throttling losses have. The losses due to simultaneous load handling and the energy that can be saved by swapping the pumps with a digital displacement pump are also found out and analysed.

A model of the existing hydraulic system is made using Simulink and Hopsan using the data and results from the energy study. It will be used to simulate and evaluate future system architectures. This model is then used to simulate a system architecture where the existing pumps are swapped with digital displacement pumps. This architecture is more energy efficient due to the higher energy efficiency of the pump. The findings from the energy study and simulations are compared and results are obtained regarding power losses, energy consumption and overall usability of the models.

The addition of the two DDP’s instead of the existing inline pumps has resulted in energy savings resulting in 4% more running time in the intermediate cycle and 5.6% in the short cycle while keeping the functionality of the machine.

(4)

4

ACKNOWLEDGEMENT

I would first like to thank my thesis supervisor, Samuel Kärnell of the Flumes department at Linköping University. The door to his office was always open whenever I ran into trouble or had a question about my research or writing. He consistently allowed this paper to be my own work, but steered me in the right the direction whenever he thought I needed it.

I would also like to thank my examiner, Liselott Ericson for providing valuable insights into the research objectives, her patience and motivation.

I would also like to acknowledge Unnikrishna Menon Ravikumar and Sandesh Tumkur at Linköping Univeristy as the second readers of this thesis, and I am gratefully indebted to them for their very valuable comments on this thesis.

My sincere thanks goes to Erik Alden, Simon Magnusson, Robert Lejonberg and everybody else on the Hydraulics team at Material handling, Epiroc. Your valuable comments and help along the way helped shape this thesis.

I would like to thank the guys over at Danfoss and Artemis especially Chris and Ove for the support and suggestions.

I also take this opportunity to thank all my friends and family in Linköping and Örebro for always being there for me and providing the much needed ‘leisure time’ during the time I spent for the thesis. Paru Chechi, Ramu Chettan, ADITI, Vishnu, Ranjith Bhai, Prabhu Chettan, Sajan Chettan, Bijin Bhai, Unni, Sumsu, Shannika, Mervin, Renma, Midba, Param, Sam, Dom, Dany, Asku, Natalia, Nazanin, Sanju Sajeev, Sanju M, Sen, Akhil, Aswin, AK. Thank you for being a part of my life. I appreciate everything you do for me.

Finally, I must express my very profound gratitude to my parents and my brother for providing me with unfailing support and continuous encouragement throughout my years of study and through the process of researching and writing this thesis. This accomplishment would not have been possible without them. Thank you.

(5)

5

L

IST OF

F

IGURES

FIGURE 1: THE DIGITAL DISPLACEMENT PUMP [2] ... 9

FIGURE 2: BATTERY AND ACTUATORS IN THE ST14 BATTERY [3] ... 11

FIGURE 3: A SIMPLE BLOCK DIAGRAM SHOWING THE HYDRAULICS OF THE AUXILIARY SYSTEM OF ST14 BATTERY . 12 FIGURE 4: LOAD SENSING SYSTEM WITH A VARIABLE PUMP [4] ... 13

FIGURE 5 PRESSURE VS FLOW DIAGRAM OF A LS SYSTEM WITH TWO LOADS [4] ... 13

FIGURE 6: PUMP EFFICIENCY AND POWER LOSS AS A FUNCTION OF DISPLACEMENT AT 200 BAR AND 1500RPM [7] . 15 FIGURE 7: EXPLODED VIEW OF THE DDP WITH PARTS LABELLED [9]... 16

FIGURE 8: STATIC ALLOCATION OF A DDP [9]... 17

FIGURE 9: DYNAMIC ALLOCATION OF A DDP [9] ... 17

FIGURE 10: FLOWCHART HIGHLIGHT ING THE METHODOLOGY OF THE THESIS... 18

FIGURE 11: SHORT LOAD CYCLE ... 20

FIGURE 12: INTERMEDIATE LOAD CYCLE ... 21

FIGURE 13: OVERVIEW OF THE VECTOR CANALIZER SOFTWARE ... 21

FIGURE 14: BLOCK DIAGRAM REPRESENTING THE PUMP SWAP OF THE EXISTING SYSTEM... 26

FIGURE 15: TIME VS TOTAL POWER FOR AN INTERMEDIATE CYCLE ... 28

FIGURE 16: TIME VS TOTAL POWER FOR A SHORT CYCLE ... 29

FIGURE 17: ENERGY BREAKDOWN FROM THE BATTERY FOR AN INTERMEDIATE CYCLE [13] ... 30

FIGURE 18: ENERGY SPLIT FROM T HE BATTERY FOR A SHORT CYCLE ... 31

FIGURE 19: ENERGY CONSUMPTION OF THE AUXILIARY SYSTEM DURING AN INTERMEDIATE CYCLE ... 31

FIGURE 20: ENERGY CONSUMPTION OF THE AUXILIARY SYSTEM DURING A SHORT CYCLE ... 32

FIGURE 21: DISPLACEMENT SETTING, POWER LOSS AND ENERGY LOSS VS TIME FOR INTERMEDIATE CYCLE ... 33

FIGURE 22: ENERGY CONSUMPTION OF THE AUXILIARY SYSTEM DURING AN INTERMEDIATE CYCLE WITH DDP ... 34

FIGURE 23: ENERGY CONSUMPTION OF THE AUXILIARY SYSTEM DURING A SHORT CYCLE WITH DDP ... 34

FIGURE 24: COMPARISON OF POWER LOSS AND ENERGY LOSS VS TIME BETWEEN THE TWO PUMPS... 35

FIGURE 25: HYDRAULIC SYSTEM DESIGN IN HOPSAN WHERE THE VARIOUS END EFFECTORS ARE IN SUBSYSTEMS OF THEIR OWN ... 39

FIGURE 26: MODEL OF THE HOIST SYSTEM OF THE ST14 BATTERY... 40

FIGURE 27: MODEL OF THE BUCKET SYSTEM IN THE ST14 BATTERY... 40

FIGURE 28: THE OVERVIEW OF THE STEERING APPLICATION DESIGNED TO MIMIC THE VARIOUS CHARACTERISTICS OF THE MACHINE ... 41

FIGURE 29: MODEL OF THE DRIVE CYCLE FOR FORWARD SIMULATION... 42

FIGURE 30: OVERVIEW OF THE REGULATOR... 42

FIGURE 31: REGULATOR SYSTEM... 43

FIGURE 32: THE PUMP MODEL OF T HE EXISTING PUMP CREATED IN SIMULINK [14] ... 44

FIGURE 33: MASK OF THE PUMP MODEL OF THE DDP [12]... 44

FIGURE 34: OVERVIEW OF THE ENT IRE SYSTEM DESIGN OF THE EXISTING SYSTEM IN SIMULINK ... 45

FIGURE 35: OVERVIEW OF THE ENT IRE SYSTEM DESIGN OF THE PUMP SWAP IN SIMULINK ... 45

FIGURE 36: ACTUATOR DISPLACEMENT VS TIME FOR THE BOOM, BUCKET AND STEER FOR INTERMEDIATE CYCLE FOR THE EXISTING SYSTEM ... 46

FIGURE 37: HOIST CYLINDER PRESSURE VS TIME FOR THE INTERMEDIATE CYCLE ... 47

FIGURE 38: BUCKET CYLINDER PRESSURE VS TIME FOR THE INTERMEDIATE CYCLE... 47

FIGURE 39: STEER CYLINDER PRESSURE VS TIME FOR THE INTERMEDIATE CYCLE... 48

FIGURE 40: ACTUATOR DISPLACEMENT VS TIME FOR THE BOOM, BUCKET AND STEER FOR THE INTERMEDIATE CYCLE AFTER PUMP SWAP... 49

FIGURE 41: COMPARISON BETWEEN THE ENERGY LOSS AND POWER LOSS BETWEEN T HE EXISTING SYSTEM AND THE DDP FOR INTERMEDIATE CYCLE... 50

FIGURE 42: BLOCK DIAGRAM OF THE SYSTEM ARCHITECTURE WHERE MULTIPLE OUT LETS ARE UTILISED ... 52

(7)

7

L

IST OF TABLES

TABLE 1:ENERGY LO SS O F T HE SY ST EMS DURIN G T HE TWO CYCLE S ... 35

TABLE 2:COMPARISON BET WEEN T HE EXIST ING PUMP AND T HE DDP IN T ERMS OF RUNNING T IME AND PRODUCT IVIT Y ... 36

TABLE 3:INT ERFERENCE LOSSE S O F T HE EXIST ING SY ST EM DURIN G T HE T WO CYCLES... 37

TABLE 4:CLA SSIFICAT ION OF INT ERFERENCE LOSSE S ... 37

TABLE 5:ENERGY COMPARI SON BET WEEN T HE FORWARD AND BACKWA RD FACIN G AN ALY SI S ... 51

Abbreviations

LHD Load haul dump

DDP Digital displacement pump

RCS Rig control system

ST14 Scooptram 14

LS Load sensing

EOD Eject’o dump

(8)

8

1. Introduction

1.1. Background

In this day and age at the forefront of technology, the number one problem humans as a whole face is the adverse effects of Global warming. The exploitation of fossil fuels and negligence towards newer greener fuels has led to this path. The obvious solution to this problem is to use more sustainable sources of energy and reduce our existing carbon footprint. However the process required to achieve this goal would require every country and organisation to participate in this deal but we are nowhere near the point in time when we can completely depend on renewable sources of energy; technology wise nor socially.

The next best solution is to use a ‘transition technology’ which can help reduce the usage of fossil fuels, but at the same time start our journey towards a more renewable and sustainable future. The answer is to improve battery technology and electrify our existing systems. While there has been significant strides in this path, we have to adapt our existing systems to use a battery and to run on renewable electricity rather than fossil fuels. This can most likely future proof the system architecture for a few years down the road.

For mobile applications, the use of battery packs and electric motors might seem to be a step down from the current innovations in IC engines especially for off-road applications. Like mentioned earlier, the changes to be made to on the existing system to run on electric power can be an opportunity to look into technology that can improve the efficiency of the machine as a whole and increase the range or workable hours of the vehicle. In the past, the size of the fuel tank could be increased to get a greater range/working hours, however it’s not as simple in an electric vehicle.

Regulations that are currently in place and the ones that might be placed in the near future all point to an emission free mine and battery technology is the way forward to meet this goal. Not only does it help match the regulations, it helps save energy used by the machine which will translate to energy saved on cooling and also helps decrease the ventilation costs required in a mine significantly. It also makes the mining environment safer and healthier. These are major factors as to why a more energy efficient system is being pursued by all major mining equipment manufacturers.

Epiroc Rock Drills AB is a Swedish company in the mining sector created in 2018 after the split from Atlas Copco. The company produces a variety of machines for rock excavation such as for drilling, material handling, infrastructure etc. Epiroc Rock Drills AB strives to increase productivity while reducing the energy consumption in all mining activities.

1.1. Digital displacement technology

A technology that has the capability to drastically improve productivity and efficiency of hydraulic machinery is the digital displacement technology found in the new programmable hydraulic pump called the digital displacement pump (DDP). This new pump can revolution ise

(9)

9

mobile hydraulics as it can be used to improve part load efficiencies, improve the response and make it easier to control in a system perspective. The DDP is a radial piston pump that has been fit with solenoid on/off valves at the inlet to control the flow of the working fluid. The total displacement of the pump depends on the displacement of each cylinder controlled digitally by an active inlet valve. The pump can act as a single unit to supply one circuit or it can dedicate pistons for supplying several circuits in parallel. They can be setup using several algorithms to achieve the above mentioned flow sharing capability with different pump outlet configurations . [1]

Figure 1: The Digital Displacem ent Pum p [2]

Even though the concept of digital displacement has been around for a while, this might be the first working model ready for manufacturing that satisfies all the claims of digital displacement. Hence, a study needs to be done to check if this is indeed the case and if the DDP when adopted in to the existing circuit has the potential to improve the efficiency as claimed by utilising the higher part load efficient and flow sharing feature.

The DDP in question here shown in Figure 1 is designed and developed by Artemis Intelligent Power located in Loanhead, Scotland. The manufacturing of the pump is being done in conjunction with Danfoss Power Solutions who acquired Artemis in late 2018.

1.2. Aim

The main aim of the thesis is to analyse the current hydraulic applications excluding the powertrain in an electric load haul dump truck (LHD) called ‘ST14 Battery’ and introduce a new hydraulic pump (DDP) into the circuit to see if the part load efficiency of the machine has been increased. The reason why the part load needs to be improved is that the ideal goal of the battery machine is to finish an entire shift (8hours) on a single battery. To achieve this, the battery needs to be bigger, an improvement in the efficiency of the powertrain or an improvement in the auxiliary hydraulics need to be made. The DDP improves the latter. In order to get the required result, the new DDP needs to be utilised to the fullest extent. The flow sharing capability along with a proper control strategy needs to be looked at for the new pump in a simulation environment.

(10)

10 1.3. Research Question

The mission of the thesis work is to answer the following questions:

1. How much energy is consumed by the current system on the electric Scooptram called ST14 Battery and by what? It will be based on two different drive cycles and calculated using backward facing simulation (Static).*

2. Will the addition of two DDP’s instead of the inline piston pumps improve the energy efficiency of the machine while keeping its functionality? It will be based on the same two drive cycles but calculated using a forward facing simulation (Dynamic).

1.4. Delimitations

The delimitations of the thesis are:

1. The energy study performed in the thesis only consider the three main service outlets of the pump which are bucket actuation, boom actuation and steering actuation. Other hydraulic applications like the braking system, lubrication and cooling are not considered.

2. The energy consumption of the powertrain is not considered in the analysis.

3. Some of the sampling frequency for various sensors and signals are different and mathematical interpolation is used to obtain an acceptable result.

4. Dynamics of the hoist, bucket and steer during motion might affect the flow through the actuators as different amounts of flow are required at various stages of actuator movement and this has not been considered for the forward analysis.

5. Two different models of the ‘existing pump’ are used for the backward and forward facing simulations and this will contribute to the error between the forward and backward facing simulations.

(11)

11

2. Theory

2.1. System design and architecture

The scooptram shown in Figure 2 is basically a wheel loader that has been squished down to make it suitable to use in mines. Therefore just like a wheel loader, the scooptram has the same functions, pick and drop the material using the bucket, raising and lowering using the boom and traverse using the steering. The battery and actuators are shown in Figure 2 for better clarity. The ST14 Battery as the name suggests has a load capacity of 14 tonnes and is run using battery power. The battery pack is placed in the back of the machine and can be replaced after one battery pack is depleted.

Figure 2: Battery and Actuators in the ST14 Battery [3]

The Battery powers two electric motors one dedicated to the powertrain and one to the auxiliary hydraulics. The latter, manufactured by ABB is a synchronous motor driven by a variable frequency drive and is rated for 200 kW peak power. This electric motor is connected to a dual path gearbox which provides power to two 63cc hydraulic pumps. Each pump has a maximum flow capacity of 135 litres per minute. The pumps have slightly different margin pressures so that they don’t work against each other. The block diagram in Figure 3 represents this system architecture.

(12)

12

Figure 3: A sim ple block diagram showing the hydraulics of the auxiliary system of ST14 Battery

Other fixed pumps like the gear pumps supplying oil to the transmission, cooling and the brakes also run on the same shaft as the main pump. This means that there is a chance that the electric motors run at times only to accommodate flow to the transmission which leads to energy loss. The flow from the pumps are guided to the actuators through two valve blocks and it consists of a Parker K220 and a L90 block which are brought together using a custom made manifold. These are valve blocks used in mobile hydraulics application. The K220 is a set of four valves while the L90 has two valves. Hoist and Bucket has been allotted two valves each to control the output actuators and steering has one. The boom and bucket utilise the K220 while the steering uses the L90 valve blocks. The two other valve slots are utilised for EOD option and Hydraulic Emergency steering.

The ST14 Battery has a battery pack with four sub packs and each sub pack is 92 kWh which translates to a battery size of 368 kWh. The battery is manufactured by NorthVolt and uses Li-ion NMC technology. It uses a Load sensing system to provide an energy efficient and responsive system and this gives a brief overview of the scooptram ST14 Battery.

(13)

13 2.2. Load Sensing

Figure 4: Load sensing system with a variable pum p [4]

The concept of load sensing (LS) was used to design the hydraulic system of this machine. An example of a LS system with two actuators and a variable pump is shown in Figure 4.This concept has many advantages over ordinary constant pressure systems. The energy loss due to the pump can be minimised because LS systems continuously adapts the pumps pressure to a slightly higher pressure than the largest load pressure. A load sensing valve usually is accompanied by a compensator over the valve. It helps with the controllability to reduce load dependency and load interference. [5]

(14)

14

In Figure 5, the pressure and flow are plotted when two actuators are simultaneously used. Pp

is the pump pressure while PL1 and PL2 are the load pressures. qL1 and qL2 are the flows due the

loads L1 and L2. C1 and C2 represent the losses in this case. C1 is caused due to the margin pressure of the pump while C2 is a combination of the margin pressure and difference between load pressures. This loss is lower than that of equivalent constant pressure and constant flow systems. However if there is a stark difference between the two loads, the power losses can be quite high.

2.3. Hydraulic pumps

The transfer of power in a hydraulic system is done through the fluid acting as a medium between two systems which convert one source of energy into another. The pump is considered to be the heart of the hydraulic system as it provides the power. [4] They are classified into different groups based on various parameters like size, shape, cost, operating range etc. One major type of classification is based on how the working fluid is used, which means hydraulic machines can either be rotational machines or piston machines. The sealing gaps might be the main difference between them as the piston machines have smaller gaps while rotating machines have larger sealing gaps. This means that piston machines can work with faster speeds and more pressure. [4]

Rotational machines include gear machines, vane machines and vane machines. Piston machines are further classified as:

a) In-line piston machines

These type of machines consist of a number of pistons arranged in a circular shape in a barrel. This barrel presses against a fixed valve plate and is connected to the drive shaft. The pistons rotate along with the barrel. A swash plate controls how much the barrel is angled to the pistons. The angle of the swash plate decides the displacement of the machine which makes these machines variable machine. The pump used in the existing system is an inline piston machine. The torque is transferred by the pistons here. It has high efficiency and high operating pressures. [4]

b) Bent axis machines

The barrel in which the pistons are placed are kept at an angle to the drive shaft. The pistons do not transfer any torque in this case. The angle of the barrel is changed to vary the displacement of the machine. Higher the angle, longer the piston can travel and more flow. This machine has high efficiency and high operating pressures too.

c) Radial machines

The pistons are arranged radially and the piston stroke makes a linear radial movement. This type of machine can have separate valves for each piston instead of valve plates. [4] The DDP is a radial piston machine and uses active inlet valves to achieve its specialities.

(15)

15 2.4. Digital displacement pump

The digital displacement pump was invented by Artemis Intelligent Power LTD. The company started as a result of research and development of fluid power and renewable energy at the University of Edinburgh. The DDP’s have an active inlet valve that can switch on/off as required. This can allow the machine to have high efficiencies and low idle losses. As they are electronically controlled, they have fast and accurate response and can be tuned electronically with different control modes. [6] They have showcased the effectiveness of the technology by using the DDP for the transmissions of off-shore wind turbines since the beginning of 2013. Another example would be the installation of the DDP in a 16 tonne excavator called ‘Dexter’. The project was able to sum up that the DDP was able to save the amount of fuel used by 16-21% when operated similar to the original excavator by only performing a pump swap. [7] The efficiency of the DDP is comparable with that of inline piston pumps and bent axis pumps in full displacement. However none of the other technologies can match the efficiency of the DDP during part load operation. Figure 6 clearly shows the efficiency of the DDP is much higher than those of swash plate pumps in the low displacement region.

Figure 6: Pum p Efficiency and power loss as a function of displacem ent at 200 bar and 1500rpm [7]

Figure 6 shows the improvement of the DDP over swash plate pumps like the one used in the

ST14 battery. The DDP has an almost uniform efficiency curve from 20% -100% displacement. Even for low displacement percentages, the DDP shows high efficiency. [7]

Both axial piston pumps and DDP’s can adjust the flow rate of the fluid during a revolution. The axial piston pump uses a swash plate to vary the stroke length of the piston while the DDP uses a couple of valves to provide rapid change in flow demand. The low pressure valve is a solenoid on/off valve that is used to control fluid intake of the pump. The high pressure valve is a passive valve which can be a poppet valve connecting the working chamber to the load.

(16)

16

The control strategies that can be developed for the DDP uses the solenoid on/off valves to control the amount of flow. [6]

Hysteresis is an error caused by friction or drag of various components in the control loop. For swashplate pumps, the friction at the swashplate is responsible for this error. 5% hysteresis error is a typical level found on most pumps today. [8] Response times for normal pumps can vary from 20-100 ms and is usually around 50 ms. The DDP is resistant to fluid contamination all the while having a 20 ms response and zero hysteresis. The pump can have digital CAN communication and software configurable control modes. [9]

Figure 7: Exploded view of the DDP with parts labelled [9]

Figure 7 gives an exploded view of the pump, valve assembly and crankshaft. The crankshaft

of the pump requires so little torque that a human can easily rotate the shaft with ease.

The highlight of these pumps is that they have the capability to have multiple outlets i.e. a group of pistons are allocated to a single outlet. A single DDP can have a maximum of four outlets each with three pistons. These can be called as ‘pumplets’. [1] This is the benefit of independent piston control with which the DDP works. The pump controller can be programmed to recognize multiple “services” with the ability to control flow, pressure, torque, power, etc. per service. Due to the mechanical constraints of the 96 cc pump design, the smallest possible grouping is 3 pistons per outlet. [9]

A tandem 96*2 cc pump will be used for the ST 14 battery as that is the only size available at the moment. Although the existing pumps are only 63 cc each, due to the very low idle losses and high efficiency at low displacements, the 96cc DDP will be a viable option.

(17)

17

A special feature of the DDP is that the piston allocation can be either static or dynamic. Static means that the pistons are always connected to the same services. Dynamic allocation can allow the change in pistons assigned to services. An external valve might be needed for dynamic allocation and multi service operation may or may not require external plumbing.

The following figures show the difference between the two allocations.

Figure 8: Static allocation of a DDP [9]

Figure 8 shows that the two DDP’s are split as 9+15 piston pumps and this is called static

allocation. The pistons are allocated and the same pistons supply the same outlet always. This is the principle behind SA2. This type of allocation reduces load interference and dependency.

Figure 9: Dynam ic allocation of a DDP [9]

Figure 9 shows the dynamic allocation of a DDP. It also shows the six piston pumplet along

with the three piston pumplet at the bottom are allocated as static and the middle pumplet with three pistons is dynamically allocated. An external manifold connects the middle pumplet to either propel or work function through switching valves. This type of allocation optimises flow as per demand. [9] This is the principle used for SA3. The dynamically allocated pump can be switched off when they are not required hence saving energy.

6

3

(18)

18

3. Methodology

The following flowchart represents the procedure and logic used in the thesis.

Figure 10: Flowchart highlighting the m ethodology of the thesis

(19)

19

The energy study on the existing system began by first going through how the hydraulic circuit has been laid out, what components are used and why. The system is designed around load sensing technique. Upon learning that, more details about the schematic of the electric scooptram (ST14 Battery) was looked at. For example, how the boom of the machine works and how exactly it is being lifted, lowered and monitored. That led to an idea on how to calculate the energy or power consumed for that particular function with the sensors already on board the machine.

After finding the largest contributor to the losses in the hydraulic system, an ideal solution would be to swap the existing pump with another that is more efficient at smaller flow rates and lower loads. To make sure the simulation model closely represents the real world counterpart, the existing system needs to be simulated first. If the flows and pressures matched, this would validate the hydraulic system. The next step would be to simulate the hydraulic system with the new DDP to obtain the results. Figure 10 shows in more detail how the thesis has been conducted.

3.1. Backward and forward facing simulation

Backward simulation basically is forward simulation of the inverse model. The model used for backward simulation is the same model used for forward simulation. The difference is in the definitions of inputs and outputs. Backward simulation can be used to compute the required input for a given output. [10]

In this context, a backward facing model is developed where the data from the end effectors can be used to find the data of another component down the line. The output is used to find the input so to say. Since the prototype already exists, some of the data can be used to find what we want. For example, since the position and pressures of the bucket actuator is logged using sensors, the flow through this actuator can be calculated. Similarly these parameters can be found for all of the actuators. The backward facing analysis can be used to find the losses and the energy consumption of the hydraulic system.

The energy analysis is especially useful because it is being done on a battery operated machine. The range of the vehicle or the maximum operating hours on a single charge can be improved by first understanding where the losses occur and then finding solutions to decrease these losses. Therefore the end goal of the analysis is to understand how much energy is being consumed or wasted, where this happens and to find these with good accuracy.

Since the study is being carried on a LHD vehicle and the energy study is done to get an idea how much energy can be saved by using a DDP, the three main sources of power consumption are the boom actuator, bucket actuator and the steering actuator. Hence, only these applications are considered while doing the analysis. The DDP does not have an effect on the powertrain as they are independent of each other.

The forward facing simulation here outputs the position, pressure and flow of the actuators which are the inputs to the backward facing simulation. The forward simulation is also dynamic in nature. [11] In this context, the hydraulic system is designed in Simulink and Hopsan and

(20)

20

these software are used to obtain the results. The pump model is given inputs like speed, load sensing pressure and margin pressure to provide flow and torque as outputs. The model of the hydraulic system uses flow from the pumps as inputs and provides flow, pressure and position of the actuators as outputs.

Therefore the backward facing simulation is used to find the energy consumption and distribution while this data is being used in forward facing simulation and the hydraulic system is being simulated dynamically.

3.2. Drive cycle

Two drive cycles which best represent the working conditions in real life are chosen to be the reference drive cycles to test the machines.

They are:

 Short cycle

 Intermediate cycle

The short cycle is a cycle that is similar to the ‘Y’ cycle used for wheel loaders as can be seen in Figure 11. It has been improvised to accommodate the single pile of material. The material is picked up and dropped in the same pile. The short cycle starts when the machine is used to load the material into the bucket, move a distance of around 50 m and then dump the material, travel with an empty bucket for the next 50 m and this is being repeated thrice. The boom is lifted to the highest point when the material is being dumped. It takes approximately 60 s to finish one cycle.

Figure 11: Short load cycle

The second one as shown in Figure 12 is an intermediate cycle where the impact of the vehicle travelling a greater distance can be looked at. It starts off by the machine collecting the material, travelling for around 500 m, dumping the material and then travelling for another 500 m with

(21)

21

an empty bucket. This is repeated thrice as well. The distance travelled can highlight the effect of steering on the power consumption of the machine. This cycle takes approximately 180 s.

Figure 12: Interm ediate load cycle

After the drive cycle has been agreed upon, the machine can be made to perform as pe r the drive cycles.

The drive cycles performed by the machine must now be translated to a drive cycle that can be used to perform the energy study. This is done by the on board computer which saves and records all the CAN messages by using the Rig control System (RCS). In order to calculate the energy consumption and losses, the required signals need to be mapped, calibrated and logged. It is done using a software called Vector CANalizer. However they wouldn’t make any sense if there isn’t a reference to compare the data obtained. Figure 13 shows the overview of Vector CANalizer.

(22)

22

This data is has to be converted to a readable format for the Vector Canalizer by another software called the Kvaser Memorator. It is being converted to .mf4 from .kmf format. This format can store more Meta data, hence it’s more advisable to use the .mf4.

This data is opened in Vector CANalizer and all the signals added into the configuration is logged. The logged file is then converted into a .mat format so as to open in Matlab. Each individual signal is stored as a separate vector. The size of the vector depends on the total time of the logged file and the sample time of the sensor. After importing the data into Matlab, we can see that the drive cycle takes more time to complete than what has been stated theoretically. We use the drive cycle information from the vehicle. This shows us that the short cycle takes approximately 100s and the intermediate cycle approximately 300s to complete. This might be because the operator who drove the vehicle might not have pushed the vehicle to its limit like seasoned mine workers. This could affect the proportion of losses and can also create deviations propulsion and hydraulic energy consumption.

The forces involved in the actuators are also a part of the drive cycle. This is found out using information about the vehicle such as weight, lifting capacity etc. The force at the actuator ends decide the pressure within the cylinder. These together made it possible to calculate the estimated forces the actuators have to handle. More information about the forces in the drive cycle cannot be highlighted in the report due to the confidential nature of the data.

3.3. Energy consumption and distribution

The vectors obtained from the logged data from Vector CANalizer are manipulated using various functions in Matlab. The constant parameters are added first and then the .mat file containing the logged vectors are loaded. The size of the vectors might be different and this is fixed using the interpolation function in Matlab. They are all made the same size and now operations can be performed on these vectors according to the various equations discussed in the theoretical study. Initially the cylinder pressures and actuator positions are read. The actuator positions are converted into linear positions and this is used to find the flow through these actuators. The compressibility of the fluid is accounted in this stage of the calculation. The total flow from the pump is calculated after which the flow through each actuators and cylinder pressure is multiplied appropriately to obtain the required power for each actuator. The direction of flow determines whether the power can be regenerated or not. The hydraulic power of the pump is calculated after which efficiency maps for the pumps are utilised along with interpolation functions to obtain the respective efficiency reading at a given speed, pressure and displacement setting.

The following set of equations are used to find the various results like flow, energy loss etc. The metric system is used throughout the report. After obtaining the linear displacement of the end actuators over the time of operation, it is differentiated to obtain the velocity. This velocity is used to find the flow of the hydraulic oil into the cylinder.

(23)

23 𝒒 = 𝒗 ∗ 𝑨 + 𝑽

𝜷𝒆∗ 𝒅𝒑

𝒅𝒕 (1)

where q is the flow through the actuator, v is the velocity of the actuator and A is the area of the actuator, V is the instantaneous volume of the actuator, βe is the bulk modulus of the fluid and 𝒅𝒑

𝒅𝒕 is the derivative of the load pressure.

The pressures readings on the base end and rod end of the actuators are measured for all three end effectors. These readings along with the flow through the actuators gives the power requirements of the actuators.

𝑷 = 𝒑 ∗ 𝒒 (2)

where P is the power required for the actuator, p is the pressure at the actuator ends and q is the flow through the actuator. Therefore the power consumption by each actuator is calculated. The direction of flow is used to identify if the power is used for useful work or if it is throttled away to lower the actuator.

The difference between the pump pressure and the output pressures of the individual actuators multiplied with the respective flow will give the losses over the valve block.

Pvalve loss = ( ppump – pload ) * q (3)

The sum of these losses can be calculated along with the actual flow from the pumps. This can be done using the following equation.

qpump = D * ℰ * n (4)

where D is the displacement of the pump, ℰ is the Displacement setting of the pump and n is the speed of the pump.

This will help calculate the required hydraulic power of the pumps used.

Now considering the battery as the input and to find how much energy is being used by the powertrain, auxiliary motor and other electrical components, an analysis is done. This has to take into account the regenerative capability of the powertrain. The inverter provides values of power usage of the traction motor and also the auxiliary motor. This is double checked by finding the power consumption using the following equation.

Pmotor = n * T (5)

where T is the torque of the motor and n is the speed at which the pump rotates.

This gives both the input and output power of the system. This can be utilised to find the efficiency of the system. [4] [7]

(24)

24

The energy distribution is visualised using Sankey diagrams where the input power is being divided based on the power consumptions of various actuators. There are different Sankey diagrams portraying breakdown of energy from the battery to the end effectors. The various divisions in the Sankey diagram showing battery consumption are:

A. Traction system

The power used by the traction motor is integrated over time to obtain the energy used by the traction system

B. Regeneration into battery

From the energy regenerated by the traction system, a fraction of it is sent back into the battery. This is not the most energy efficient method and is not preferred. It can be calculated by looking at the battery power in the positive direction.

C. Regeneration into the auxiliary system

The other fraction of the regenerated energy from the traction system is sent to the auxiliary system directly and this is the best way to use regenerated energy.

D. Energy used by other electrical use

The energy taken by other electrical sources like the light, cabin heating, cooling system etc are part of this.

E. Auxiliary system

The power consumed by the auxiliary motor is integrated over time to obtain the energy used the auxiliary system. The total energy used by the electric motor is the sum of the regenerated energy and the auxiliary system.

The various divisions in the Sankey diagram of auxiliary hydraulics are: A. Pump loss

The pump loss is where the improvement in the energy efficiency should be seen significantly while using the DDP as it claims lower idle losses and higher efficiency overall. This is modelled by taking the input energy going into the pump and subtracting the hydraulic output of the pump.

B. Energy loss due to the fixed pumps

The fixed pumps that are on the same shaft as the main are looked into more detail and tests are conducted on it to see how much power these pumps utilise in different speed ranges. This data and the curve fitting tool of Matlab is used to create an equation to find the power usage of the fixed pumps as a function of speed. Hence, the chunk of energy the fixed pumps use can be removed from the input energy of the auxiliary system.

(25)

25

Throttling losses is the energy loss that occurs in the valve blocks due to the difference in pressure across it. It is calculated using equation 3. Throttling losses can be divided into two: upstream throttling and downstream throttling. Upstream throttling is when the actuators are used to perform work. Downstream throttling is the throttling to the tank and happens for example when the boom and bucket are lowered.

Downstream throttling is not considered in the above energy distribution as it is assumed that the pump does not have to push flow through to the input of the actuator. It is also assumed that all of the fluid being throttled out of the valve is used to fill the input of the actuator. This is done because automatic lowering/gravity lowering is used for the boom and bucket actuators.

The pressure difference in the valves can be due to the usage of one or multiple loads simultaneously. Interference losses are the losses that occur due to simultaneous motion of actuators. When the control of one load will affect the control of the other actuators, it is called load interference. These are separately calculated in a Matlab script by utilising the joystick signals to make sure the actuators are activated by the user and not by the motion of the vehicle. The maximum pressure levels are calculated and the difference between this and the other loads are multiplied with the flow through the actuator which has maximum pressure. This is done using ‘for loops’ and ‘if conditions’ to make sure all states of the machine are accounted for.

D. Required energy for hoist, bucket and steering

The required energy for the actuators are calculated using equations 1 and 2. The impact of these actuators will depend on the drive cycle used. This is the actual work done by the auxiliary hydraulics of the machine.

E. Energy used by emergency steering accumulators and other losses

The energy that is left behind is the combination of the energy utilised by the emergency steering accumulators and other losses like gearbox losses and hose loses.

The energy study also takes a look at how the energy is spread throughout the drive cycle. The interference losses of the system along with interference losses of each pair of ac tuators are also found out. These are found out using the same way as the valve losses as in equation 3. The pressure difference in the valves multiplied with the flow through the valve will provide the interference losses. It will be part of the throttling losses that are calculated in the earlier sankey diagrams.

(26)

26

3.4. Simulation of the hydraulic system with the existing pump

As seen in Figure 10, after the energy study has been completed, the next step is create a model of the hydraulic system to replicate the working of the original machine. The simulation of the existing system is done in two parts. The hydraulic system except the pumps are being modelled in Hopsan while the pump model is done in Simulink. The Hopsan model utilises interface components to communicate with the Simulink blocks.

The Hopsan models need to be converted into ‘s’ functions for them to work in Simulink and this can be done using the ‘export model into S function’ tool within Hopsan.

The parameters of the main hydraulic valves, end effectors, compensators, relief valves etc. have been input as per the data sheet of the real world machine. This makes sure that the hydraulic system made in the simulation environment closely matches the real vehicle. The speed, displacement settings and position of the actuators are directly taken from the logged data from the backward facing simulation making the forward simulation as close to the real world details as possible.

The entire model is brought together in Simulink finally where they all interact with each other and the two drive cycles are run and simulated using Simulink and the ‘S’ functions from HOSPAN. The results of the simulation are provided in the results section.

3.5. Simulation of the hydraulic system with the DDP

(27)

27

The existing system is simulated in Simulink, the flow, pressure and position of the end effectors should be comparable to that of its real world counterpart. This makes it clear that the hydraulic system without the pump model is accurate. The losses and the required energy are validated using the results from the backward facing analysis. Therefore the next step of the forward facing analysis to simulate the hydraulic system with the new pump.

Figure 14 shows the architecture for the pump swap. Since the only change in the hydraulic

system is the pump, the Simulink component for the pump is swapped with the pump model for the DDP. The details of the pump model cannot be looked at in more detail as it is confidential. The pump model pf the DDP provides the flow, torque and power loss for the appropriate speed, output pressure and load sensing pressure. [12]

The new pump is connected to the same input port for flow in the ‘S’ function and Hopsan along with Simulink runs the simulation similar to the earlier simulation to obtain results. The same speed and drive cycle are used here to make sure the model closely follows the earlier simulation. The simulation provides the pressures, flow and position of the end effectors along with the power loss in the pumps. These results are showcased in the results section.

After the simulation with the new pump is run, a Matlab script is written to calculate the power loss and energy loss by the pumps. The valve losses along with other losses in the hydraulic system are validated against the backward facing analysis. The Matlab script also compares the power loss and energy loss calculated from the forward facing simulation to the backward facing simulation to find out how different the results are.

The input mechanical energy and the output hydraulic energy are calculated here too.

*The Matlab script written to find all the result is the property of Epiroc Rock Drills AB and these are not shown due to the sensitive nature of the information.

(28)

28

4. Backward facing simulation

4.1. Results

A Matlab script was written and used the logged data from the test runs for the two drive cycles. After the data was filtered, the appropriate equations were used to find the necessary results as shown below.

Figure 15: Tim e vs Total Power for an interm ediate cycle

Figure 15 shows the total required power of the three actuators over time for the intermediate

cycle. We can observe that the average power requirement is around 20KW while the maximum power requirement is 110 kW when the hoist is lifted. Hence the electric motors and inverters will have to be designed to deal with these peaks. This also highlights the reason why such large pumps are required for the auxiliary hydraulics.

The blue graph shows the power requirement of the bucket while red and orange represents hoist and steer respectively. The power required is added on top of each other and this is shows the effects of ‘interference’ in this cycle. The various stages of the drive cycle are also explained at the bottom of the graph.

Figure 16 shows the total required power of the actuators over time for the short cycle. Similar

to Figure 15, the average power is quite low but peaks of over 120KW are obtained when the hoist is lifted. The various colours represent the different actuators and their impact can be seen on the graph. Interference can also be seen in the graph.

(29)

29

Figure 16: Tim e vs Total Power for a short cycle

The various stages of the drive cycle are explained at the bottom of the graph. The vehicle is initially moved to the pile after which the material is loaded into the bucket and is then moved to the turning point and back into the pile. The material is dumped by lifting the pile and bucket rolled out and then this is done without the material in the bucket to duplicate the journey back to the pile.

These graphs showing the power requirement are interesting to look at because they highlig ht the fact that there are times when there is no power requirement or very little requirement. This can hint that a system with very little idle losses can be an improvement to the current system.

(30)

30

4.1.1. Energy breakdown from battery to electric motors

Figure 17: Energy breakdown from the battery for an interm ediate cycle [13]

Figure 17 shows the breakdown of energy from the battery into how much energy is used by

the traction motor, auxiliary motor and splits the amount of recovered energy into two. For the intermediate cycle, 69.45% of the energy is used by the traction motor. This is understandable as the vehicle travels over a longer distance. 19.23% energy is used by the auxiliary motor. This will further be split in the coming sections.

The amount of energy recovered is very low as the test run was done on fairly level ground. It can be seen that out of the 2.7% energy that is recovered, 2.6% goes directly to the auxiliary system. This is why the input for auxiliary hydraulics in Figure 17 is larger than energy used by auxiliary system in Figure 19. For this particular section, regeneration deals with only the recovered energy by the traction motor. This is great for the efficiency of the vehicle as the energy does not have to go back to the battery where there will be further losses. The average power consumption for the intermediate cycle is 75.96 kW and the total energy consumption is 6.33 kWh.

For Figure 18, the traction motor consumes 56.02% of the energy and the auxiliary motor consumes 32.62% of the energy in the short cycle. This is obvious because the short cycle is fairly short and the vehicle only travels around 100m totally. Therefore a higher impact will be seen on the short cycle when the pump swap is done. Similarly only 1.7% energy is recovered during the cycle. The average power consumption of the short cycle is 86.4 kW and the total energy consumption is 2.4 kWh.

It can also be seen that other electrical uses in the machine also take up to 8-10% energy. Further studies showing the classification of these would be useful in the future.

(31)

31

This shows us that the traction motor takes the most amount of energy from the battery and future improvements to the powertrain will have a great effect on the energy efficiency of the entire vehicle.

Figure 18: Energy split from the battery for a short cycle

4.1.2. Energy breakdown from auxiliary motor to end effectors with existing pump

Figure 19: Energy consum ption of the Auxiliary System during an interm ediate cycle Average Power = 16.56 kW

(32)

32

Figure 19 shows the energy breakdown of the end effectors while also showing other consumers

of the energy from the auxiliary motor.

Since there are fixed pumps mounted on the same shaft as that of the main hydraulic pumps, some amount of energy is wasted to run those and maintain the demands of those pumps. This takes around 11.08% of the energy.

The major loss happens at the pump side where 32.86% of the energy is lost. Therefore this should be the area that most effort is to be put to decrease the energy consumption of the vehicle. Losses due to throttling comes next at 11.69%. These are a combination of losses due to one actuator being used or multiple actuators being used i.e. interference losses are included in this portion.

The energy requirement of the end effectors come next and this is the real work that is done by the machine. The rest of the energy is used by the emergency accumulator and by other losses such as gearbox losses and hose losses.

Figure 20: Energy consum ption of the Auxiliary System during a short cycle

The same can be seen for the energy breakdown of the short cycle as seen in Figure 20, although with different shares of the losses. The major factor for losses is still the pump with 30.87% but then the hoist energy requirement comes next as this is quite a short drive cycle where the impact of the hoist is quite larger than the other end effectors. Throttling losses come next at 10.17%

The energy due to the fixed pumps is 9.09% here and the effect of the steering is lower than that of the intermediate cycle.

The average power required for the intermediate cycle at the auxiliary motor is 16.56 kW while that for the short cycle is 29.88 kW. This shows us that more power and energy is required for the short cycle and the running time would be lesser here.

(33)

33

Figure 21: Displacem ent setting, Power loss and Energy loss vs tim e for interm ediate cycle

One of the reasons for the very low efficiency of the existing pumps is that the displacement setting is very low at certain times. This can mean that the pumps are much bigger than they need to be for the vehicles. However there are conditions where the size of the existing pumps are justified as the flow requirements is quite big when the hoist and bucket are lifted.

The average displacement settings of the two pumps are 0.3517 and 0.5490.

From Figure 21, the power loss and energy loss over time is indicative of the fact that even when the displacement settings are low, there is a steady power loss and a steady increase in energy loss. Therefore it can be seen that a pump with better energy efficiency at no load and low loads can be an improvement to the existing system.

4.1.3. Energy breakdown from auxiliary motor to end effectors with the DDP

In Figure 22 and Figure 23, it can be seen that there is a large decrease in the pump losses and this is seen as an improvement as the input energy for the auxiliary motor is decreased hence increasing energy efficiency.

It can be seen that there is a decrease in the energy consumed as the average power consumed is reduced by quite a bit for a single cycle.

The pump losses has decreased drastically while the functionality of the vehicle isn’t prone to change. The energy breakdown for the existing system and the DDP is obtained by using their respective efficiency maps.

(34)

34

Figure 22: Energy consum ption of the Auxiliary System during an interm ediate cycle with DDP

All other functionalities and values remain constant except for the efficiency maps of the pumps.

Figure 23: Energy consum ption of the Auxiliary System during a short cycle with DDP

Therefore even though the fraction of the various factors might change, the absolute values of these factors remain the same.

Average Power = 14.04 kW

(35)

35

Figure 24: Com parison of Power loss and energy loss vs tim e between the two pum ps

Figure 24 clearly shows how much of an improvement the DDP might have after the pump

swap. The orange line shows the power loss of the DDP and it can be seen that it’s lower than that of the existing system which is shown in blue. The dashed line represents the ene rgy loss and is lower than the bold black line. The power loss and energy loss is lower during part load conditions and this is the main reason for the improvement.

Table 1: Energy loss of the system s during the two cycles

DRIVE

CYCLE

ENERGY LOSS

OF THE

EXISTING PUMP

ENERGY

LOSS OF

THE DDP

COMMENTS

INTERMEDIATE 0.4632 kWh 0.2399 kWh

The pump loss is reduced by 48.2 % and saves auxiliary energy by 15.2%

SHORT 0.2599 kWh 0.1351 kWh

The pump loss is reduced by 48 % and saves auxiliary energy by 14.4% P o w e r lo ss o f th e p u m p s (k W ) E n e rg y lo ss o f t h e p u m p s ( k W h )

(36)

36

Table 1 gives the numerical representation of Figure 24 and shows a reduction of pump losses by 48.2% for the intermediate cycle and by 48% for the short cycle. Therefore it can be seen that the pump swap is a good idea to reduce the part load losses. This supports the theory and necessitates further simulation to validate this information using forward analysis techniques

Table 2: Com parison between the existing pum p and the DDP in term s of running tim e and productivity

EXISTING PUMP DDP

Available Running time for the intermediate cycle: 3 hrs 42 minutes

Available Running time for the intermediate cycle: 3 hrs 50 minutes

Number of cycles : 44.5 Number of cycles : 46 Load hauled : 623 tonnes Load hauled : 644 tonnes

Available Running time for the short cycle: 3 hrs 14 minutes

Available Running time for the short cycle: 3 hrs 25 minutes

Number of cycles : 116.5 Number of cycles : 123 Load hauled : 1631 tonnes Load hauled : 1722 tonnes

For further information and an easier grasp on how much impact the pump swap will have on the vehicle, table 2 can be used as a direct comparison between the two variations.

The productivity is a term used here to represent the amount of the material that can be handled by the machines during the running time. It is calculated by multiplying the number of cycles the machines that perform with the load capacity of the vehicle.

An improvement of 8 minutes (4%) running time and 21 tonne (4%) increase in productivity for the intermediate cycle can be seen after the pump swap.

An improvement of 11 minutes (5.6%) running time and 91 tonne (6%) increase in productivity for the short cycle can be seen after the pump swap.

(37)

37

4.1.4 Interference losses

The losses due to simultaneous usage of the end effectors is called interference losses. They can be quite high and can impact the energy efficiency of the entire system. There are various solutions to reduce the interference losses and they are discussed in the future scope section of the thesis.

Table 3: Interference losses of the existing system during the two cycles

Drive Cycle Interference Losses (KWh) Average Interference Losses (KW)

Short 0.0558 kWh 2.0088 kW

Intermediate 0.0745 kWh 0.8940 kW

From table 3, it is clear that the effect of interference losses for short cycle is more than the intermediate cycle. The interference losses is around 5.4% of the total input energy for intermediate cycle and 6.75% of the total energy for the short cycle.

Average power losses are shown to highlight how much of an impact they have as values in kWh might not be a good representative of the amount of energy wasted due to interference losses.

The interference losses between the three actuators are given in table 4:

Table 4: Classification of interference losses

Cycle

Interference due to

Value

Short Boom 0.0078 kWh Bucket 0.0385 kWh Steer 0.0094 kWh Intermediate Boom 0.0089 kWh Bucket 0.0499 kWh Steer 0.0157 kWh

(38)

38

From the above table it can be seen that the interference due to the bucket is higher than the other actuators. Hence, if the bucket, boom and steering functionality are independent of each other, throttling losses can be reduced or even eliminated fully.

This principle is used for the next architecture of the pump called system architecture 2 and system architecture 3 (SA2 and SA3). Certain pumplets of the DDP can be used for these functions independent of each other to obtain the required result.

(39)

39

5. Forward facing simulation

5.1 Modelling of the hydraulic system

5.1.1 Modelling in Hopsan

The hydraulic model of the auxiliary system is modelled using co-simulation between Hopsan and Simulink. The pump model is created in Simulink while the rest of the hydraulic system is created in Hopsan.

Figure 25: Hydraulic system design in Hopsan where the various end effectors are in subsystem s of their own

Figure 25 represents the hydraulic system that was designed in Hopsan. As it can be seen from

the figure, there are interface components that are used to send information back and forth between Simulink and Hopsan. The flow is sent from the Simulink model of the pump directly into the Hopsan model where it continues to the actuator subsystem through a pressure relief valve. The actuators are placed in subsystem to make it easier to understand and also to make the model seem less complex with a lot of connection going back and forth.

(40)

40 The subsystems are namely:

a. Hoist subsystem

Figure 26: Model of the hoist system of the ST14 Battery b. Bucket subsystem

(41)

41 c. Steer subsystem

Figure 28: The overview of the steering application designed to m im ic the various characteristics of the m achine

Figure 26, Figure 27 and Figure 28 show the three actuator systems that are modelled as closely

to the real world counterpart as possible. It follows load sensing like the original machine. The flow from the pump enters each of the actuator and then goes to their respective valves through compensators. The compensators keep the pressure difference between the valves at a constant level and also helps to reduce load dependency and interference. The flow then goes to the actuator. The valves are controlled by a position feedback PI controller.

There are transducers for pressure, flow and position used wherever required. Required data are sent back to the Simulink model.

There are other system that are used to simulate the whole system which are also made in Hopsan. The trajectory or the drive cycle for the model is also modelled in Hopsan like shown in Figure 29. The drive cycle has to provide the reference signal for the actuators to follow. The force that acts on the actuators are also input through the trajectory. The reference signals are the exact same as the backward facing signal to keep the comparison fair. The forces on the actuators are estimated using various parameters like the weight of the vehicle, weight of the bucket and boom, weight of load etc.

(42)

42

Figure 29: Model of the drive cycle for forward sim ulation

Figure 30: Overview of the regulator

Another system modelled in Hopsan is the regulator that is used to simulate the existing system. The regulator is used to output two parameters that are sent to the Simulink model of the pump. The speed is directly taken from the backward facing simulation and using a look up table is sent to the model. The displacement of the pump is a function of the load sensing and output pressure of the system. A regulator piston is used and the two pressure values are provided at the two ends of it. A force equal to the margin pressure of the pump is applied to the piston.

(43)

43

This is then converted to the displacement settings of the two pumps. The displacement setting is kept between zero and one. This can be seen from Figure 30 and Figure 31.

(44)

44

5.1.2 Modelling in Simulink

Simulink was used to model the existing pump and the DDP. The forward simulation model of the existing pump shown in Figure 32 takes in values for speed, load sensing pressure and displacement setting and outputs a value for flow and torque. The model also shows the efficiency of the pump system.

Figure 32: The pum p m odel of the existing pum p created in Sim ulink [14]

The other pump model is that of the DDP. The pump model of the DDP is pressure controlled. The model needs the size and margin pressure of the pump as inputs. It can be scaled in displacement, hence making it possible to use the same model for one of the pumplets. This model takes in speed, load sensing pressure and output pressure and outputs the flow and torque values corresponding to the input values. The mask of the DDP is shown in Figure 33.

Figure 33: Mask of the pum p m odel of the DDP [12]

The pump models created in Simulink is the property of Epiroc and Danfoss, therefore cannot be showcased in the report in more detail.