Semi-active hydro-pneumatic spring and damper system

Semi-active hydro-pneumatic spring and damper system

ABSTRACT The basic concept of land vehicle transportation has not changed much in the last few decades,although much progress was made in improving and optimising vehicle design and technology. Spring and damper characteristics determine to a large extent the ride quality and handling of a vehicle. Since the requirements for good handling and good ride are conflicting, adjustable suspension elements are developed. In this study a twostate semi-active hydro-pneumatic spring, in conjuction with with a two-state semi-active semi-active hydraulic damper is investigated. Two types of tests were performed on a prototype spring-damper unit, namely characterisation tests and single degree of freedom tests. The characterisation tests included characterising the hydro-pneumatic spring, while for the single degree of freedom tests, the step response were determined. Good correlation was obtained between measured and simulated data for the characterisation, as well as the single degree of freedom tests. The spring –damper model can be incorporated into a full 3-D vehicle model in order to predict the ride and handling of a vehicle fitted with such a system.

1


INDEX

SR. NO. 1

1.1 2

CONTENTS NAME

PAGE NO.

Introduction

Preamble

8

Literature and historical overview

2.1

Preamble

11

2.2

Need of semi-active hydro-pneumatic suspension

11

systems 2.3 3

4

Historical overview

11

Experimental work

3.1

Preamble

13

3.2

Experimental spring and damper unit

13

3.3

Experimental test setup

14

3.4

Hydro-pneumatic spring characterization

17

3.5

Single degree of freedom testing

18

3.6

Ride height adjustment

19

Heat transfer effects on hydro-pneumatic

2


suspension systems

4.1

Preamble

21

4.2

Determination of spring characteristics

21

4.3

Experimental verification

24

4.4

Modes of heat transfer

27

4.5

Effect of damper heat build-up

27

Conclusion

30

References

31

3


LIST OF FIGURES

2. 1

Swiss Mowag Piranha

3.1 3.1

Exper Experim iment ental al semi semi-a -act ctiv ivee spr sprin ing g and and dampe damperr unit unit

3. 2

Char haracterization te test se setup

3.3 3.3

Sing Single le degr degree ee of free freedo dom m tes testt set setup up

3. 4

Flo Floating pi piston ac accumulator

3.5 3.5

SemiSemi-ac acti tive ve spri spring ng char charac acte teri rist stic icss (0.0 (0.01m 1m/s /s))

3.6 3.6

Two Two st state ate dam dampe perr cha chara ract cter eriisti stics

3.7 3.7

Step Step res respon ponse se for for dif diffe fere rent nt com combi bina nati tions ons of spr sprin ing g and and damp damper er

3. 8

Ride height adjustment

4. 1

Exper perimenta ntal spring unit nit

4.2 4.2

Meas Measur ured ed and and pre predi dict cted ed spr sprin ing g char charac acte teri rist stic icss

4.3 4.3

Eff Effect ect of exci excittati ation freq freque uenc ncy y

4.4

Compar Comparisi ision on betwe between en adiab adiabati atic, c, isoth isotherm ermal al and and predi predicte cted d spring spring charact characteri eristi stics cs

4.5 4.5

Eff Effects ects of chan change ge in gas gas tem tempe pera ratture ure

4


1.

INTRODUCTION

1.1 Preamble

The basic concept of land vehicle transportation has not changed much in the last few decades,although much progress was made in improving and optimising vehicle design and technology. The quest to always go faster,further and more comfortably,has lead in recent years to the development of advanced suspension system. An improved suspension system allows a vehicle to achieve higher speeds over rougher terrain,and results in better handling,as well as improved ride comfort. Passive suspension systems(suspension without controllable elements),always represent a compramise between ride comfort and handling,since a stiff suspension is required for good handling,while a more compliant suspension is needed for good ride comfort. Implimenting a controllable suspension (adaptive,slow-active,fully-active) is therefore an attempt to narrow the gap between the opposing requirements for optimal ride comfort and handling. In armoured fighting vehicles and other heavy off-road vehicles it is important to have a soft suspension (low spring rate) allowing big wheel travel when negotiating rough terrain. A soft suspension however compromises good h andling and straight line stability at high speed on good roads. It is known that semi-active dampers can dramatically enhance ride cofort and handling over rough terrain, still making the ch oice of spring rate a compromise between ride comfort and handling. It is now possible to choose between two spring rates- one that favours good ride comfort and another favouring good handling. These suspension systems are mainly developed for heavy off-road vehicles like trucks, military vehicles or passenger vehicles. By introducing a semi-active hyd ro pneumatic spring and damper system two very important suspension elements can be

5


changed to suit specific driving conditions. When driving on good roads the hard spring and damper settings can be used to to improve vehicle stability stability and handling. By being able to switch between a soft and a hard suspension good off-road capabilities as well as good handling and stability can be achieved. This study focuses on a semi-active suspension system,consisting of a two-state switchable hydraulic damper,as well as a two-state switchable hydro-pneumatic spring. The different elements of the spring/damper system are characterised to predict the system performance.

6


7


Fig 1.1 Armoured Personnel Carrier Fitted With Semi-active Damper

2. LITERATURE AND HISTORICAL OVERVIEW 2.1 Preamble

In this chapter, an overview is is given of semi-active dampers and hydro pneumatic springs.It also focusses on literature literature concerned with large off-road vehicles, but in cases where the applicable technology has not yet been demonstrated on heavy vehicles, reference is made to commercial and passenger vehicles.

2.2 Need Of Semi-active Hydro-pneumatic Suspension Systems

The suspension systems which are used for normal cars d o not suit the large offroad vehicles, since the weight of such vehicles is much more than the normal n ormal cars. Also even if we use such suspension systems, a large stress results on them eventually making them of no use. So to overcome this difficulty semi-active hydro-pneumatic suspension systems using spring and damper were invented. Semi-active dampers were conceptualized in the 1970’s and numerous configurations and control strategies were simulated and tested since then. Semi-active suspension systems greatly influence the vehicle dynamics (ride comfort and handling). This is also the reason for developing semi-active suspension systems, namely to improve ride comfort without compromising handling and stability, by switching between hard and soft spring/damper characteristics.

2.3 Historical Overview

8


Hydro-pneumatic suspensions have been introduced on battle tanks in the 1950’s. The first hydro-pneumatic struts were fitted to a prototype tracked vehicle, as result of research done by two German companies, Frieseke and Hopfner from Erlangen and Borgwald from Bermen into the use of compressible co mpressible fluids in suspension systems (Hilmes 1982). Since then, several other military vehicles were fitted with hyd ro pneumatic suspensions, but most of them did not go into production due to reliability problems and short life span of the mechanical components. Initially, confidence in this type of suspension was low, due to sealing and design problems.

But since the introduction of more reliable sealing techniques, hydro-pneumatic springs and dampers have become more popular and are occasionally used in passenger cars and large off-road vehicles. This type o f suspension system is popular due to its nonlinear characteristics and versatility.

9


Hydro-pneumatic suspensions are not commonly used on commercial vehicles due to high capital cost involved. The first production tracked vehicle fitted with a hy dro-pneumatic suspension was the Swiss Strv-103 Main Battle Tank (MBT). This vehicle was fitted with a rigidly mounted weapon and the height adjustable hydro-pneumatic suspension was used to tilt the vehicle upward or downward (Hilmes 1982).

Fig 2.1 Swiss Mowag Piranha

10


3. EXPERIMENTAL WORK 3.1 Preamble

In this chapter, the experimental work is presented. The experimental work can be divided into two stages, namely characterization tests and single degree of freedom tests. The set up, test equipment and characterization procedures for both these stages are discussed in this chapter. Where deemed necessary, some background information is supplied, in order to elucidate the characterization c haracterization process. The test results are presented in graphical and tabular format. In the following paragraphs, firstly the test set up is discussed, with reference to the hardware, software and test equipment. Sec ondly, the characterization of the spring and damper are discussed. After that, the single degree of freedom tests are discussed and finally some closing remarks are made.

3.2 Experimental Spring and Damper Unit

The high and low characteristics for both spring and damper are made possible by channeling hydraulic fluid with solenoid valves (Fig. 3.1). The spring and damper units consists of hydraulic strut (1), two gas filled accumulators (2 and 3), a hydraulic damper (4) and two solenoid valves (5 and 6). The low spring rate is achieved by compressing a large volume of gas in two separate chambers (2 and 3). By sealing off one of the chambers (2), a smaller gas volume (3) is compressed and a higher spring rate is achieved. Spring rates can be

11


individually tailored by changing the two gas volumes. For low damping the hydraulic damper (4) is short circuited by opening a by-pass valve (5). For high damping this valve is closed and the hydraulic fluid is forced through the damper resulting in a high damping force.

Fig. 3.1 Experimental semi-active spring and damper un it 3.3 Experimental Test Setup

Two experimental setups were used, one for the component characterisations (springs, dampers and valves) and another for the single degree of freedom tests. In both the cases, a 160 kN Schenck hydraulic actuator was used to supply the desired input. Fig. 3.2 shows schematically the characterization setup. The top of the strut was fixed to the rigid test frame with a locating pin, while the bottom mounting was fixed to the hydraulic actuator. Spherical rod ends were used, in order to eliminate any bending moments on the strut. In this setup, the required relative strut displacement is generated by vertical actuator motion.

12


Fig.3.2 Characterization Test Setup For the single degree of freedom tests, a separate test frame was build. Fig. 3.3 shows test setup. A lead mass of approximately 3 tons was used to simulate the sprung mass of a vehicle, since a static wheel load of between 2.5 and 3 tons are common for military offroad vehicles. The test frame was equipped with a set of linear bearings, guiding the sprung mass, which was supported by lower mounting of Schenck actuator. Nitrogen cylinder is used to fill the accumulators. The accumulators, dampers and valves were secured on top of the lead mass. The test frame was securely fixed to the test floor, to ensure that the SDOF setup does not fall over.

13


Fig. 3.3 Single Degree Of Freedom Test Setup

3.3 Hydro-pneumatic Spring Characterization 3.3.1 Physical Attributes

There are many different types of hydro-pneumatic springs, but the basic difference lies in the way the gas and oil is separated. Some hydro-pneumatic springs have a rubber bladder separating the gas and the working fluid, while others have a floating piston. Both the hydro-pneumatic springs considered here are of floating piston type.

14


Fig. 3.4 Floating Piston Accumulator The static volumes of the two accumulators are 0.3l and 0.7l respectively. When the valve is open the combined volume of the two accumulators is 1.0l and when the valve is closed only the 0.3l 0 .3l accumulator is connected with the single acting cylinder. Nitrogen gas was used as springing medium. 3.3.2 Characterization Procedure

For the hydro-pneumatic spring characterization, two stages of semi-active hydro pneumatic spring were characterized by subjecting the strut to a sinusoidal displacement, of varying frequency. The excitation speed is defined as the piston speed when moving through the static position. The excitation speed is therefore a function of excitation frequency and excitation amplitude. The amplitude of signal was approximately 100mm. Fig 4.5 two spring characteristics for an excitation speed of 0.01m/s. It can be seen that two very different spring characteristics were achieved.

15


Fig. 3.5 Semi-active Spring Characteristics (0.01m/s) 3.4 Hydraulic Damper Characterization 3.4.1 Physical Attributes

The damper pack used has non-linear damping characteristic, which is achieved by a system of orifices, a sealing washer and Belville springs. This This damper was mounted statically between the strut and the hydro-pneumatic springs (fig 4.1). The damping force is therefore supplied by resistance to fluid flow through damper pack. 3.4.2 Characterization Procedure

The damper characteristics of the two-state damper were characterized by subjecting the strut to a sinusoidal displacement input. The actuator force was recorded when the strut travels through the static position, where the velocity is almost constant and a maximum. Fig. 4.6 shows the damper characteristics for both “on” a nd “off” states. It can be seen that “off” “o ff” characteristic for velocities above 0.25m/s is approximately half that of the “on” state, in the compression direction (negative velocity). The damper force, in the rebound direction, shows an almost constant damping force at high velocities. This is because during rebound motion the driving force behind the hydraulic fluid is the accumulator’s pressure and in the compression direction, the simple acting strut cylinder.

16


Rebound damping can therefore not be made too high, since cavitation may occur at high velocities.

Fig 3.6 Two State Damper Characteristics 3.5 Single Degree Of Freedom Testing

Several single degree of freedom tests were p erformed to evaluate the characteristics and performance of the semi-active spring/damper system. Step response and ride height adjustment feature were evaluated. ev aluated.

Step Response Input

The tests were done by subjecting the system to a step displacement input of 43mm. The sprung mass displacement, for different spring and damper combinations, are shown below. It can be seen that for spring “on” condition an effective natural frequency of approximately 1.5 Hz is achieved. For the spring “off” state, the natural frequency is in the region of 1 Hz. The response for the two cases where the damper da mper is in the “off” state indicates that the system is under damped. For the damper “on” state, the motion is damped out within one cycle.

17


Fig 3.7 Step Response for Different Combinations of Spring and Damper 3.6 Ride Height Adjustment

The ride height of vehicles fitted with hydro-pneumatic suspensions can be easily adjusted by adding or removing hydraulic fluid from the system. A lower ride height results in reduced body roll, lower center of gravity, a more stable firing platform and a lower silhouette. The ride height can also be increased when a greater ground clearance is needed. Ride height adjustments are usually achieved by making use of an external power source, such as an engine driven hydraulic pump. A control system then regulates the amount of hydraulic fluid in the system with a network of pipes and valves. The Swiss Mowag Piranah III is an example of a vehicle fitted with such a system. The semi-active hydro-pneumatic spring/damper system investigated in this study has the added advantage of being able to adjust the ride height without using an external pump. The ride height adjustment works as follows: When valve (6) is closed, the hydraulic fluid in the accumulator (2) is effectively removed from the system. By opening and closing the valve at the right moment an

18


amount of hydraulic fluid can be stored in accumulator (2), thus varying the ride height. To decrease ride height, valve (6) is kept closed and only opened when the pressure in accumulator (3) is higher than in accumulator (2). This is done until the pressure in accumulator (2) reaches a predetermined value corresponding to a specific reduction in ride height. To increase the ride height, h eight, valve (6) is only opened when the pressure in accumulator (2) is higher than in accumulator (3).

Fig. 3.8 Ride Height Adjustment

19


4. HEAT HEA T TRANSFER TRAN SFER EFFECT E FFECTS S ON HYDROPNEUMATIC SUSPENSION SYSTEMS 4.1 Preamble

Literature on hydro-pneumatic suspension systems invariably describes the spring force of the hydro-pneumatic spring by polytropic processes, assuming ideal gas behaviour. This assumption is a simplification simplification of the real situation as heat transfer effects between the gas and its surroundings cannot be ignored. The nitrogen gas used as spring medium cannot be treated as an ideal gas under the pressures and temperatures found in hydro-pneumatic suspension systems. Most hydro-pneumatic suspension systems incorporate the spring and damper into one unit for reasons of cost and packaging. This results in undesirable temperature effects, i.e. variations in ride height and spring rate.

4.2 Determination of the spring characteristic

4.2.1 Classical approach for determining the spring characteristic

The classical approach for determining the spring ch aracteristic is to assume ideal gas behaviour and polytropic gas compression processes. Isothermal and adiabatic characteristics are then calculated. In typical hydro-pneumatic suspension units, the average gas temperature can vary between -20 and +200 degree Celsius, while gas pressure varies between 2 and 110 MPa. Under these conditions, especially at pressures higher than 30 MPa, gas ga s compressibility can result in large errors rendering the ideal gas assumption invalid.

4.2.2. New approach for determining the spring characteristics using real gas behaviour and heat transfer models

To circumvent the errors introduced by the ideal gas assumption, the BenedictWebb-Rubin equation of state, is used to give a more accurate relationship between gas

20


pressure, volume and temperature. Other similar equations of state can be used depending on the operating range of the suspension unit. If a gas in a closed container is compressed, the volume decreases while pressure and temperature increase, storing energy in the process. As the gas is allowed to expand, expand , this stored energy is released. Energy is however lost during this process because of heat transfer between the gas and its surroundings. When the gas is compressed, the temperature and pressure increase and heat is transferred from the g as to the surroundings. When the gas is expanded, the gas temperature drops and heat is transferred from the surroundings to the gas. At very low compression speeds, enough time is available for heat transfer and the gas temperature stays constant, giving the isothermal characteristic. At high speeds, very little time is available for heat transfer and the gas temperature varies, resulting in the adiabatic characteristic. At speeds between these two extremes, the spring characteristic forms a hysteresis loop representing the energy loss in the cycle. This effect has been noted in many publications but is usually attributed to friction and fluid losses in the spring system. Many attempts were made to describe d escribe this effect by adopting speed or frequency dependant polytropic exponents which can be as high as 1.7. This approach is often used in hydraulic accumulator calculations.

Given below is a differential equation which describes the heat transfer between the gas and the environment. e nvironment. Eq. (1) is solved using a fourth order Runge-Kutta method.

21


It can be seen from the definition of the thermal time constant that it varies during the cycle as the wall area (Aw) varies due to piston motion. The convection coefficient (h) also varies due to the change in speed of the gas over the inside surface of the cylinder. However, a constant value of the time constant fits experimental data very well and the analysis is fairly insensitive to its value.

4.2.3. Determination of the thermal time constant

The thermal time constant can be determined experimentally, or calculated from heat transfer models based on empirical data. The average thermal time constant is determined experimentally by subjecting the spring to a step displacement input while monitoring gas temperature, gas pressure or spring force. The thermal time constant is defined as the time needed for the pressure, temperature or force to decrease by 63% of the difference between the peak and final values. Temperature data is more appropriate to use but difficult to measure. In the region whe re the ideal gas assumption is valid, the same result will be obtained from pressure, force or temperature measurements. Because the analysis is fairly insensitive to the value of the time co nstant, pressure measurements

22


are used to determine the thermal thermal time constant. Good correlation is achieved between measured and predicted values for the time constant as indicated in Table 1.

4.3 Experimental verification 4.3.1 Laboratory tests on experimental spring unit

An experimental hydro-pneumatic spring unit was designed and manufactured to verify predicted results. The unit was designed for a static wheel load of 3000 kg and a stroke of 250 mm. The basic b asic design of the experimental spring unit is shown in Fig. 1. The unit was fitted with thermocouples to measure gas and oil temperatures, and pressure transducers to measure gas and oil pressure. Schenck hydropuls equipment was used to excite the spring with sinusoidal displacement inputs of various amplitudes and frequencies. Spring force, displacement, gas and oil pressures and temperatures were recorded.

Fig. 4.1 Experimental Spring Unit

23


4.3.2. Comparison of experimental and predicted results:

The isothermal spring characteristic was determined by compressing the spring slowly in discrete increments, leaving enough time for the temperature and pressure to stabilize at each displacement interval. Correlations between measured an d predicted values are excellent. Fig. 2 shows the measured and predicted spring characteristics for a sinusoidal excitation with a frequency of 0.1 Hz and amplitude of 80 mm. It can be seen that the use of Eq. (1) to predict the dynamic d ynamic spring force is in close correlation with measured data. Fig. 3 shows the effect of different excitation frequencies o n the spring characteristic. It can be seen that the lower the excitation frequency, the smaller the hysteresis loop becomes. This means that the energy loss and thus the damping in the cycle becomes less. Isothermal compression is approximated as there is more time available for heat transfer between the gas and its surroundings. At higher excitation frequencies, the hysteresis loop also becomes smaller as adiabatic compression is approximated and less time for heat transfer is available. The hysteresis loop means that hydro-pneumatic suspensions have an amount of inherent damping which is dependent on the excitation frequency (or speed) just like hydraulic dampers. The amount of damping however decreases as the excitation frequency increases and adiabatic conditions are reached. This is another advantage of hydro-pneumatic suspensions which may be worthwhile to utilize. E.g. trying to manipulate the thermal time constant to get the maximum damping at the resonant frequency of the suspension system as this damping generates no heat because the net heat flux is zero. Fig. 4 shows the adiabatic and isothermal spring characteristics compared to the predicted dynamic spring characteristic. It is clear that assuming assuming isothermal or adiabatic behaviour can result in large errors. The presence of inherent damping may seem insignificant, but becomes an important factor when pe rforming accurate dynamic simulation of vehicles to improve ride comfort or noise, vibration and harshness levels. It must also be taken into account accoun t when developing semi-active damper systems because

24


the inherent damping will limit the minimum value of the off characteristic which usually needs to be as low as possible.

Fig. 4.2 Measured and Predicted Spring Characteristics

Fig.4.3 Effect of excitation frequency

25


4.4. Heat transfer modes

The aim up to now was to predict and verify the nature of the spring characteristic which only involved heat transfer between the gas and the environment. If a complete hydro-pneumatic spring system is analyzed, the effects of heat generation in the integral damper should be included in the analysis. Heat transfer takes place by means of conduction through the floating piston and cylinder walls and convection between the gas, cylinder walls and the environment. 4.5 Effects of damper heat build-up

The energy dissipated in the damper can be calculated by integrating the damper force multiplied by the relative velocity across the damper. The damper characteristic was measured for static gas pressures of 6, 8 and 10MPa. Changing the gas pressure affects the maximum rebound damper force which can be achieved because the damper starts to cavitate when the pressure difference across

26


the damper is higher than the gas pressure. This will cause the oil to aerate which will severely affect damper performance. Fig. 4.6 shows the ambient, oil, gas and wall temperature histories for an excitation frequency of 0.5 Hz at amplitude of 60 mm. It can be seen that due to heat generation in the damper, the oil temperature rises until termination of the tests at 130 degree Celsius. The wall temperature follows the oil temperature trend but the gas temperature rises to only 70 degree Celsius. After the test is terminated, the gas temperature continues to rise indicating that heat transfer between the oil and gas is very slow. Conduction through the steel cylinder wall is much greater than the convection coefficient between the wall and the surroundings, therefore the lumped capacitance method is used to analyze the heat transfer coefficients. Fitting curves to determine the thermal time constant between the oil and the surroundings after termination of tests (when heat input from the damper is zero) z ero) yields time constants of between 3000 s at a temperature difference of 100 degree Celsius and 6500 s at a temperature difference of less than 20 degree Celsius. The effect of the rise in gas temperature (from 23 to 6 8 degree Celsius) on the combined spring and damper force can be seen in Fig. 4.5. The spring rate increases by 58% while the static ride height would increase by 15% (18 mm) after 75 min of testing.

Vehicle tests were performed over typical terrains at representative speeds. After 45 min of continuous driving, the damper oil temperature stabilized at 85 degree deg ree Celsius. The gas temperature showed no increase at all and the ride height variation was zero. The test track included sections of very smooth concrete road over which the heat generation in the damper can be ignored. Thermal time constants of between 150 and 450 s were measured on the smooth parts of the track. These are an order of magnitude lower than that measured in the laboratory, indicating that heat generation in the damper is closely related to terrain roughness and that heat transfer between the spring unit and its surroundings is very much dependant on airflow over the suspension units. The specific design of the test unit with go od thermal separation between the gas and damper oil, together with air flow across the unit, results in a hydro-pneumatic spring

27


configuration which does not suffer from severe temperature effects as often encountered on other systems.

Fig. 4.5 Effect of change in gas temperature

Fig. 4.6 Temperature Histories

28


CONCLUSION Following conclusions can be made from the above report: •

It is possible to have a suspension with a high and a low spring rate, optimized for both ride comfort and handling.

•

Non-linear characteristics are very useful, eliminating eliminating the necessity for a bump stop when the hard spring setting is selected.

•

A method for easily altering the vehicle ride height without any additional costs or further complicating the system was illustrated.

•

This study in short introduced a suspension element which can be altered continuously to suit the terrain and driver input demands.

29


REFERENCES 6.1 Journal Papers •

Semi-active hydro-pneumatic spring and damper system : Els. P .S. & Giliomee C L , 1998 Journal of Terramechanics 35, 109-117.

•

Heat transfer effects on hydro-pneumatic suspension systems : Els. P.S. & B. Grobbelar , 1999 Journal of Terramechanics 36, 197-205.

•

The ride comfort vs. handling compromise for off-road vehicles : Els. P.S., N.J.Theron , M.J.Thoresson & P.E.Uys , 2007 Journal of Terramechanics 44, 303317.

6.2 Books •

Analysis of a four state switchable hydro-pneumatic spring and da mper system.

-

Christiaan Lambert Giliomee Department of mechanical and aeronautical engineering University of Pretoria (2003).

6.3 Web links •

www.sciencedirect.com

30


•

www.interscience.com

•

www.car_suspension_bible/1-4.com

31

Semi-active hydro-pneumatic spring and damper system

Recommend Documents