The drive for further evolution

Abstract

Construction machines, also referred to as engineering vehicles or earth movers, are used in a variety of tasks related to infrastructure development and material handling. While modern construction machines represent a high level of sophistication in several areas, their suspension systems are generally rudimentary or even nonexistent. This leads to unacceptably high vibration levels for the operator, particularly when considering front loaders and dump trucks, which regularly traverse longer distances at reasonably high velocities. To meet future demands on operator comfort and high speed capacity, more refined wheel suspensions will have to be developed. The aim of this thesis is therefore to investigate which factors need to be considered in the fundamental design of suspension systems for wheeled construction machines. The ride dynamics of wheeled construction machines are affected by a number of particular properties specific to this type of vehicle. The pitch inertia is typically high in relation to the mass and wheelbase, which leads to pronounced pitching. The axle loads differ considerably between the loaded and the unloaded condition, necessitating ride height control, and hence the suspension properties may be altered as the vehicle is loaded. Furthermore, the low vertical stiffness of off-road tyres means that changes in the tyre properties will have a large impact on the dynamics of the suspended mass. The impact of these factors has been investigated using analytical models and parameters for a typical wheel loader. Multibody dynamic simulations have also been used to study the effects of suspended axles on the vehicle ride vibrations in more detail. The simulation model has also been compared to measurements performed on a prototype wheel loader with suspended axles. For reasons of manoeuvrability and robustness, many construction machines use articulated frame steering. The dynamic behaviour of articulated vehicles has therefore been examined here, focusing on lateral instabilities in the form of “snaking” and “folding”. A multibody dynamics model has been used to investigate how suspended axles influence the snaking stability of an articulated wheel loader. A remotecontrolled, articulated test vehicle in model-scale has also been developed to enable safe and inexpensive practical experiments. The test vehicle is used to study the influence of several vehicle parameters on snaking stability, including suspension, drive configuration and mass distribution. Comparisons are also made with predictions using a simplified linear model. Off-road tyres represent a further complication of construction machine dynamics, since the tyres’ behaviour is typically highly nonlinear and difficult to evaluate in testing due to the size of the tyres. A rolling test rig for large tyres has here been evaluated, showing that the test rig is capable of producing useful data for validating tyre simulation models of varying complexity. The theoretical and experimental studies presented in this thesis contribute to the deeper understanding of a number of aspects of the dynamic behaviour of construction machines. This work therefore provides a basis for the continued development of wheel suspensions for such vehicles.

Introduction and background 1.1 An introduction to construction machines

Since the origin of the first steam-powered excavating machines in the 19th century, mobile construction machines have been crucial for the development of modern infrastructure. While most early machines were basically modified agricultural tractors, modern construction machines have evolved into purpose-built, integrated vehicles.

The most elementary task for a construction machine is earth moving – excavating and/or moving soil, sand, rock or similar crude material from one location to another. While most machines are designed with this task in mind, construction machines can be found in numerous other roles, such as pallet handling, pipe laying, or any other task that requires a rugged vehicle with high load capacity and off-road capability. Earth moving, as well as many other types of transport tasks performed by construction machines, is usually performed in driving cycles that include loading, transporting and unloading the material. The transport capacity of an earth-moving machine can be described by two basic parameters: the amount of material carried per cycle, and the total cycle time. Increasing the task performance can therefore be accomplished either by increasing the load capacity, or shortening the cycle time. Depending on the specifics of the earth-moving operation, various machines may be selected for the task. The research work presented in this thesis focuses on the vehicle dynamic behaviour of the machine, meaning that the relevant parts of the work cycle are the “haul” and “return”. The machines of interest are those that travel over reasonably long distances at higher speeds, mainly including front loaders and dump trucks. Excavators, scrapers, bulldozers and similar machines are either used at very low speeds or mainly in a stationary position, meaning that vehicle dynamic behaviour is less relevant to these machine types.

The drive for further evolution

The evolution of construction machines has historically been driven primarily by productivity demands. Therefore, the load capacity is the primary figure of merit, followed by secondary objectives such as tractive performance, reliability, durability and ease of maintenance. Vehicle dynamic qualities such as ride comfort and handling quality have been considered less important, which is reflected in the lack of refined wheel suspensions. Usually construction machines feature at best rudimentary suspension systems, and often rely solely on the tyres to provide vibration isolation and road holding. The vehicle layout may also be less optimal with regard to the ride and handling characteristics, instead being optimised for maximum load capacity and ease of operation.

The lack of refined suspension systems, in combination with the rough surfaces where construction machines operate, generally means that vibration levels in earth-moving machines are high. The detrimental health effects of whole-body vibrations are well known, including back pain and various internal organ disorders, and vibration exposure can therefore be considered a major occupational hazard for earth mover operators. Furthermore, the directive states that no operator shall at any time be exposed to vibration levels exceeding 1.15 m/s2.

Whole-body vibrations are only countered to some degree by current technology. Cab and seat suspensions may alleviate vibrations, but they are limited by the short stroke length available and are mainly effective in the vertical direction. It has also been seen that, due to inadequate design, many seats actually amplify vibrations in the critical frequency range, thus aggravating the situation rather than improving it. Increased tyre damping could possibly improve vibration isolation in vehicles without suspension, but is generally undesirable due to increased rolling resistance and thermal stress. Moreover, off-road tyre design is generally governed primarily by demands for durability and traction, which limits the possibility for modifications. A more radical solution is to use remotely operated or autonomous vehicles, which solves the vibration problem completely by removing the operator. Considerable advances in unmanned machinery have been made, especially in the mining industry, and autonomous earth-moving operations in laboratory environments have been successfully demonstrated as well. While the use of unmanned machines is likely to increase, it still seems highly probable that manned construction machines will remain in use for the foreseeable future.

Given the inherent limits of current methods for vibration isolation in construction machines, it seems obvious that the logical progression is further improvement of suspension systems. With an efficient wheel suspension, surface-induced vibrations can be reduced at the source, thus improving the dynamic behaviour of the entire machine rather than merely shielding the operator. This is not only an efficient way of reducing ride vibrations, but will also reduce dynamic wheel loads, thus making higher transport speeds possible. This translates directly into increased productivity, since the total cycle time will be decreased. Hence, improved suspension systems will also lead to increased task performance. This is true for unmanned machines as well, as wheel load fluctuations and shaking of the carried load make higher transport velocities impractical without wheel suspension, even if no operator is affected. Secondary effects also include reduced shock loads on the vehicle frame, which can allow lighter structural design and thereby reduced energy consumption. Decreased wheel load variation could lead to less tyre wear, as slippage is reduced. Hence, although more refined suspensions certainly mean increased procurement cost and complexity, there is much to be gained besides improved ride comfort, both economically and environmentally.

Research approach

As stated above, wheel suspension has the potential to improve ride comfort and productivity for a construction machine that is used to a large extent in a transporting role. However, the introduction of a more complex suspension system will also have a radical effect on the dynamic response of the machine, especially that of machines such as wheel loaders, which are traditionally built without wheel suspension. This leads to the research question considered in this thesis: “What particular vehicle dynamic properties need to be considered in the design of a wheel suspension for a construction machine, and how is the dynamic behaviour of the machine affected by the addition of suspended wheel axles?”.

Vehicle dynamic theory provides a natural starting point for finding the answer to this question. The fundamentals of vehicle dynamic theory were first formalised in early papers on ride dynamics (Olley, 1934) and handling (Segel, 1956). These papers mark a first application of theoretical analysis in ground vehicle design. Even if the methods have evolved over time, it has been observed that many of the rules of thumb established in the early days of vehicle dynamic analysis are still adhered to by modern designers (Barak, 1991). However, the main focus of published research has been on passenger cars, to some degree extending to heavy road vehicles as well, although it has been acknowledged that the increased complexity and variety of heavy vehicles complicates the application of generic, simplified models (Gillespie and Karamihas, 2000). Another observation (Blundell and Harty, 2004) is that theoretical methods are generally applied in the early phases of a design project, while the later design stages tend to draw more from common practice and past experiences. This is a natural progression as vehicle development today takes place in increments, with each new generation differing only slightly from the previous generation. Hence, it is possible to rely on established practices and simplified design criteria.

The design of a construction machine suspension represents a different case. Rather than an incremental change, the introduction of wheel suspension systems represents a major evolutionary step in the overall design of the machine. It is therefore necessary to revisit the foundations of vehicle dynamic theory and apply these using parameter combinations relevant to construction machines. In this way, a thorough understanding of the machine dynamics can be obtained and applied to the design problem at hand. This is the philosophy behind the research presented in this thesis. Applying vehicle dynamic theory to construction machines reveals a number of areas that are critical for the dynamic behaviour of the machine, and hence illuminate certain factors that influence suspension design. This theoretical approach is complemented by simulation models and empirical testing, in order to study more complex dynamic phenomena that are not covered by elementary theoretical models. A number of limitations have also been made. Since the main focus is suspension design, the analyses performed have been limited to ride comfort and handling stability. Longitudinal dynamics such as tractive performance have not been considered, since such properties are generally influenced more by the drive train and tyre properties than by the suspension. This also means that less attention has been paid to the particular properties of off-road terrain. While the behaviour of vehicles on deformable terrain is an important research area, it is generally more relevant to tractive performance than to ride vibrations and handling. Hence, terramechanic considerations have been deemed outside the scope of this thesis.

Methods used

As stated above, the objective of the research work presented here is to provide an increased insight regarding the particular vehicle dynamic behaviour of construction machines, in areas relevant to the design process for suspension systems. To this end, a number of methods on different levels of complexity have been applied. Linearised, low-order analytical models have been utilised to study ride dynamics and lateral stability on an elementary level. This provides a fundamental understanding of dynamic behaviour and may also be used as a base for more refined analysis methods.

To study more complex dynamics, multibody dynamic simulations have been used. This allows more realistic and complex studies of the vehicle dynamic behaviour, including higher order ride dynamics and lateral dynamics such as body roll and load transfer, which are typically not covered by simplified analytical models.

Scale model testing, using a purpose-built remote-controlled vehicle, has also been used to study the lateral stability of articulated vehicles. The use of a scale model represents a compromise between cost-efficiency and accuracy. Moreover, since the vehicle is unmanned, it allows the performance of hazardous tests without the safety considerations that are usually necessary in full vehicle trials.

Full vehicle testing has been utilised to a limited extent. Basic tests with an experimental suspended wheel loader, subjected to a single obstacle excitation, have been performed to gather data for comparison with multibody dynamic simulations. A rolling test rig for large tyres has also been evaluated, using a road vehicle tyre as a test object. The test results show that this test rig produces repeatable and robust tyre data and may therefore be utilised to gather data on more complex tyres as well, thus making it a useful tool for future evaluations of the dynamic behaviour of off-road tyres.

Ride dynamic aspects

Ride vibrations are generally defined as tactile or visual vibrations in the frequency range of 0 to 25 Hz. Higher frequency disturbances are generally referred to as noise, and are less relevant to fundamental suspension design. Thus, the ride dynamics of a ground vehicle are mainly related to the translational and angular oscillations of the suspended mass, as caused by surface irregularities. The response of the vehicle to external excitations is directly related to the suspension properties, and can therefore be seen as the most fundamental consideration in suspension design for optimal ride characteristics.

The typical construction machine considered in this thesis is a wheeled, heavy off-road vehicle. As such, it differs in a number of respects from the road vehicles which are the focus of most published studies. Nevertheless, fundamental methods for ride analysis can provide important information on the dynamic behaviour of construction machines, if the particular differences are included in the analysis.

Roll dynamics

As in the case of the pitch and bounce frequencies, it is desirable to maintain a roll frequency in the range of 1.0 to 1.3 Hz for optimal ride comfort. However, roll comfort generally represents a more difficult case than pitch and bounce comfort, since the effective roll stiffness is also governed by the handling stability. It can be shown that the rollover stability margin decreases with the roll angle, and it has also been observed that the roll angle displacement may be used as a measure of the handling stability. This means that high roll stiffness is desirable from a stability standpoint. Furthermore, the design of construction machines often places the centre of gravity in a high position, especially when loaded. The inertial forces therefore create an overturning moment which needs to be compensated for by sufficient roll stiffness. This is often detrimental to ride comfort, as it leads to high lateral accelerations when the vehicle is subjected to large asymmetric surface disturbances. Given the conflicting demands for ride comfort and rollover stability, especially on uneven surfaces, it seems unlikely that a passive system can fulfil both requirements. Thus, a semi-active or active system may be necessary. Such systems have the capability to reduce large roll displacements when there are transient loads, while maintaining low accelerations in normal driving. Although the majority of research work on active systems has been carried out for road vehicles, some examples exist where the technology has been applied to off-road vehicles. Semi-active systems have been demonstrated for agricultural tractors, as well as for military off-road vehicles, and have been seen to reduce transient roll motion while at the same time minimising lateral accelerations. Fully active roll control in the form of an active anti-roll bar has also been implemented on an off-road passenger vehicle, and was seen to improve the handling stability with preserved ride comfort. One particular concern for construction machines is the large loads involved in active ride control. This may restrict the possibility for active control, possibly limiting the potential for control to semi-active damping.

Higher order dynamics

Besides the elementary oscillations of the suspended and unsuspended mass, other resonances may exist in the frequency range relevant to ride comfort. One cause of such vibrations is individual components attached to the vehicle frame, as the large masses of these components lead to relatively low natural frequencies. An overview of resonance modes in heavy trucks has been presented by Gillespie (1985), who has listed a number of component resonance modes in the 1 to 20 Hz range, for example exhaust stack longitudinal oscillation (6.9 Hz), radiator pitching (10.1 Hz) and cab bouncing (15.2 Hz). This condition is similar to that for construction machines, which feature a number of components of high mass such as the cab, lifting attachment and driveline assembly. As construction machines may vary considerably in layout and size, the exact spectrum of vibrations is highly individual for each machine type and needs to be determined using modal analysis methods.

Elastic deformation of the vehicle frame is also a source of low frequency vibration in heavy vehicles, the most typical example being the bending and torsion of the longitudinal frame members in heavy trucks. Such flexible modes are said to occur at 6- 10 Hz. Similar conditions exist in buses, where structural optimisation of the upper body has been shown to improve ride comfort. Finite element studies of an articulated dump truck frame have shown a first bending mode at 10.2 Hz and a first torsion mode at 17.4 Hz. As these modes are computed on the bare frame, it seems likely that the resulting frequencies are lower on the full vehicle, as attached masses lower the resonance frequency.

From the simulations and studies cited above, it can be concluded that higher order modes exist and most likely influence the ride dynamics, but that such modes generally occur at frequencies above the main ride frequencies which are important in suspension design. Furthermore, many of the higher order modes are consequences of fundamental layout decisions and are therefore difficult to change, unless a major vehicle redesign is undertaken. Thus, the higher order modes are best considered as boundary conditions for suspension design, rather than a part of the design space.