Essay details:

  • Subject area(s): Engineering
  • Price: Free download
  • Published on: 7th September 2019
  • File format: Text
  • Number of pages: 2

Text preview of this essay:

This page is a preview - download the full version of this essay above.


Automatic Self Inflating Tyre System for optimized passenger ride comfort















We, SP GAJANAYAKE, KOV KOTTAHACHCHY, PSH PALLEMULLA and UJP POLWATTE declare that this thesis titled 'Automatic Self Inflating Tyre System for optimized passenger ride comfort' and the presented in it is our own. We conform that:

This work was done wholly or mainly in candidature for a degree at this University.

Where we have consulted the published work of others, which has been always clearly attributed.

Where we have quoted from the work of the others, the source is always given. With the exception of such quotations, this thesis is entirely our own work.

We have acknowledged all main sources of help.

Reg. No. Name of the Member Signature

ENG/13/204 SP GAJANAYKE '''.



ENG/13/195 ULP POLWATTE '''.

Date: ''''..


I certify that the above statement made by the authors is true and that this thesis is suitable for submission to the university for the purpose of evaluation.




Senior Lecturer,

Department of Mechanical Engineering,

Faculty of Engineering,

General Sir John Kotelawala Defence University.


It was a great opportunity in our undergraduate period to involve in project works and enhance the valuable experience of it. This opportunity would have not been a success without some special personnel. So it is our duty to pay our gratitude for those who helped us during this period.

First we would like to extend our gratitude to Mr. WSP Fernando for guiding us through the project as well as providing necessary technical details and necessary advices.

We are also grateful to Mrs. PPSS Pussapitiya and Mr. RMPS Bandara  for provision of expertise advices and guiding us when necessary.

Our special thanks goes to all academic staff and all other lecturers of the Department of Mechanical Engineering for their dedication and hard work which helped us immensely in succeeding in this endeavor.


Vehicle vibrations are a major source of fatigue in the day-to-day activities of many people, especially if they take long trips or travel on rough roads. The product of this research is an intelligent system that tunes the pressure of the tyre so that the ride comfort is optimized. This system automatically identifies the terrain of the road that the vehicle travels on based on preloaded data. It then inflates or deflates the tyres through a pneumatic system. The pressure that the tyres are to be brought up to is governed by a Fuzzy-based controller which comprises the controlling algorithm of the system. After setting the tyres to the required pressure for a certain type of road, the system will continuously monitor the road for changes in the terrain and recalculate the pressures accordingly. A test platform and a quarter car model were built for the purpose of demonstration. The design process of these is presented in this paper as well. While improving ride comfort is the main objective, regulation of the tyre pressures will also prevent the tyres from becoming underinflated thus preventing reduced tread life, increased stopping distance and reduced fuel economy apart from lowering passenger comfort (Tire pressure monitoring final rule, 2000).


Almost everyone is exposed to vehicle vibrations daily. These vibrations affect the whole body and bring discomfort in the short term and may adversely affect the health of an individual in the long term (Nahvi, Fouladi, and Nor, 2009). People suffering from ailments such as lower back pain have their problems exacerbated due to constant vibration from long rides (Rufa'i et al., 2015). All human organs are affected by vibrations up to 12Hz. Those above this threshold have local effects. Vibrations in the lower frequencies (4-8 Hz) resonate the body causing muscle fatigue and make the passengers more vulnerable to back injury (Nahvi, Fouladi, and Nor, 2009). The transient vibrations we experience in vehicles are generated by the roughness of the road, rotating components, the driveline, the engine and the air flow around the bodywork of the car (de Rafael, 2007). Multiple studies have been conducted in the past that deal with minimizing vehicle vibrations by means of modifying the seat (Rebelle and Griffin, 2004), (Liang and Chiang, 2006) and the suspension (Tung et al., 2011). However, this research focuses on the vibrations that stem from the contours and roughness of the road and the transmission of the said vibrations to the vehicle via the tyres.

The British Standard, BS 6841 and the International Standard, ISO 2631 offers descriptive evaluations of human responses to whole'body vibrations. The techniques stated in the ISO 2631-1 are utilized in this research to measure comfort and categorize vibration data into several comfort zones. As per the standard, the value considered in the classification of comfort zones was the frequency-weighted root mean square acceleration. The latter is calculated by subjecting accelerometer data to a series of mathematical manipulations stated in the ISO 2631-1.

The development of the complete system comprised of three parts: the development of the control algorithm and design of the controller; mechanical design and construction; pneumatic design and construction. Fuzzy Logic was implemented in the controller to regulate the tyre pressures at the optimal point. The mechanical design included designing of the rotary joint that interfaces the axle and the wheel along with the test platform and the quarter car model. The pneumatic design involved modifications to the reservoir tank and solenoids for control of the air flow.


There are several self tyre inflating systems already being used. One such system is the Meritor Tyre Inflation System (MTIS) by the company, P.S.I. The MTIS is only used in trucks and trailers to compensate for the pressure loss by slow leaks and tyre punctures. The reason of its selective use is that it uses air from the trailer's air system to reinflate tyres to a preset pressure when needed. This makes the MTIS unfit for passenger vehicles due to the absence of a source of compressed air. The CTIS system developed by Dana Corp and Eaton Corp is electronically-controlled and warns the driver when the speed exceeds tyre pressure settings. This is done to ensure traction control in an array of terrain and load conditions. In case of a tyre puncture, it can keep the vehicle in drivable condition for up to 45 minutes. The CTIS, much like the MTIS, is used in trucks and military vehicles only. The AIRGO T3 is similar in operation and utilization.

The Automatic Tyre Inflation System (ATIS) and the Self Inflating Tyre (SIT) provide an energy-efficient tyre inflation method for passenger vehicles. Both systems comprise a valve mounted on the wheel rim that activates when the tyre pressure falls below a preset. The signal for the pressure drop is obtained from the Tyre Pressure Monitoring System (TPMS). Air is sucked into the valve from the atmosphere by a compression chamber in ATIS and peristaltic tube chamber in SIT. Thus, they eliminate the need for a separate source of compressed air. These systems, however, do not cope with punctures efficiently.

The electromagnetically activated on-wheel air-pump (Patent 6691754), owned by Chrysler corp., characterizes a pump activated by electromagnetic means to automatically inflate tyres when their pressure falls below a certain threshold. The source of air for the system is the atmosphere itself. The air is pumped into the tyre using a plunger which also uses an electromagnet to switch between its open and closed positions.

Yet another tyre pressure maintenance system (Patent 846354) introduces two innovative methods of automatic tyre inflation. One of its preferred embodiments is an air chamber which is sealed once air flows into it. This chamber is then heated, which causes pressurization of the gas. The alternative is a flexible membrane between a pumping chamber and an air transfer chamber, the compression of which provides the pumping force. In both scenarios, micromechanical pumps are used to assist pumping the air into the tyres.

The vehicle wheel including self-inflating tyre pump (Patent 5558730) by Hughes Aircraft incorporates a pump to the wheel. A piston in the pump draws air into a compression chamber by moving radially outwards in a cylinder by making use of centrifugal force. A biasing element include in the pump, can cause it to move radially inwards resulting in the compression of the air in the compression chamber and pumping it into the tyre. The radially outward motion is prevented from occurring until a minimum speed threshold is achieved in order to stop contaminants entering the compression chamber when driving slow over unfavourable terrain. A stopper mechanism halts the entire process when the tyre is at a nominal pressure.

The vehicle wheel including self-inflating mechanism (Patent 5355924) by the same company integrates the reservoir of the pressurized air within the wheel. A regulating valve allows air to flow from the reservoir to the tyre when its pressure drops below a preset threshold. The air in the reservoir is replenished by atmospheric air after a compression process. A radial bore, a piston movable inside the bore and variable volume compression chamber in the bore are the main features of the compression system. When the vehicle is stationary, a spring moves the piston inwards opening a check valve allowing air to enter the variable volume compression chamber. When the vehicle accelerates above a set speed, centrifugal force compels the piston to move outwards against a spring and compress the air in the chamber. When the air is compressed, another check valve opens to enable the compressed air to flow into the reservoir inside the wheel.

The pneumatic tyre wheel having a deformable bladder for adjusting the inflation pressure of said pneumatic tyre wheel (Patent 5119856) presents a different approach to tyre inflation. Inside a firmly sealed chamber between the rim and the tyre, a two phased system exists: a liquid and its own vapour. The saturated vapour is housed in a deformable toric bladder mounted on the rim. The desired tyre pressure can be achieved by changing the temperature of the two-phase system thus shifting its equilibrium point.

The technique patented as inflation and deflation of a tyre in rotation (Patent 4922984) integrates a hose placed on the periphery of the rim of the wheel. One end of the hose is open to the atmosphere while the other end opens into the tyre. A non-rotatable, solid pressure roller causes a local reduction in the sectional area of the hose when the tyre rotates. This motion causes the air to move along the hose and undergo compression during its flow along the length of it before reaching the tyre.

These systems put a stop to underinflation of the tyres and hence improve tread life and risks on the road. In contrast to these systems that maintain a preset pressure, the system developed for this research dynamically changes the tyre pressure to achieve optimum comfort for any terrain. Consequently, it achieves all the objectives of the existing self inflating systems while adding the feature of enhanced comfort.

Project Motivation

The persistent demand for greater ride comfort by the consumer and the increased competitiveness among suppliers to deliver improved ride quality with newer and better vehicle designs were two main reasons for selecting this project. Ride comfort is a major consideration of an individual before selecting a new vehicle to buy. To cater to this requirement, over the years vehicle companies have strived for the perfectly smooth ride. When it comes to the design and manufacture of passenger vehicles, enhancing comfort (other than improving fuel economy), has been a forerunner of the research and development work carried out by major vehicle companies. The innovations they have come up with range from simple tyre pressure monitoring systems to technologically-superior suspension redesigns. So, ride quality being a major selling point in the vehicle manufacturing industry is an aspect to be considered.

The consumer demand for a smoother ride is prevalent in developing countries such as ours. The absence of a well-developed road system leads to drivers having to endure a bumpy ride. This leads to a greater demand for vehicles offering better ride quality among the vehicle users in these countries. As the current state of roads in these countries seems to stagnate in this condition for years to come, the said demand will probably remain.

The array of health risks outlined in section 2.1.2 of this report is yet another reason for being motivated to initiate this project. Travelling on rough roads causes vibrations in a multitude of frequencies to affect the body. Suspension systems of most vehicles dampen only a fraction of these vibrations. Without any other means of attenuating the vibration magnitude, the vehicle users are at risk of developing symptoms of certain disorders when subjected to such vibrations in the long term or even exacerbating existing health issues. This called for a need of an additional means of reducing vibration magnitude transmitted to the passengers. This system could be made part of an integrated whole to bring the vehicle users to a perfectly unaffected level.

The emphasis of maintaining correct tyre pressure was inexorably taken into account as one of the objectives of initiating this project. The effects that under-inflated or over-inflated tyres that can have on the vehicle and the passengers is clearly described in section 2.2 of this report. In summary, under-inflated tyres result in increased stopping distance, reduced tread life, reduced fuel efficiency and reduced ride comfort while over-inflated tyres are prone to blowouts. As a whole maintaining good tyre pressure will lead to less accidents and greater tyre life.

The advent of laws regarding maintenance of correct tyre pressure in a vehicle is already being carried out in developed countries. One example for such legislation is the Transportation Recall Enhancement, Accountability and Documentation (TREAD) Act of 2000 mandated by the United States of America. The first part of the final rule established a new Federal Motor Vehicle Safety Standard that requires the installation of tyre pressure monitoring systems (TPMSs) to warn drivers of under-inflated tyres. The realization of the many threats that under-inflated tyres can potentially pose has been the reason behind introducing this law. Hopefully, other countries, such as ours may follow suit in the future making the benefits of this invention global.


Research Hypothesis ' ' An optimized tyre inflation pressure exists for each type of road condition when the roads are classified according vibration RMS data'

The main goal of this project is to develop an intelligent system for tyre-pressure regulation in vehicles focused towards optimizing ride comfort. The system will incorporate automatic choice of the required tyre pressure for any type of terrain and automatic inflation or deflation of the tyres to achieve it. The system will be fully automated and will require no user involvement at all.


Determining the tyre pressures which optimize ride comfort in a variety of road conditions.

Developing an algorithm to interpret the terrain quality of the road that the vehicle travels on and set the tyre pressures to the value that offers maximum ride comfort.

Designing a mechanism to couple the stationary pneumatic circuit to the rotating wheel hub so that the pneumatic tubing has access to the tyres without risk of the tubes getting tangled.

Developing a tyre pressure monitoring sensor with a rechargeable power source so as to eliminate the need for replacement when the battery dies.

Fabricating a quarter car model to model the vehicle's sprung mass and unsprung mass.

Fabricating a test platform on which the said quarter car model can be placed and simulated to mimic the vehicle's motion on a number of actual road conditions.


This product benefits the passengers and the drivers of vehicles which have to withstand rough road conditions or long trips. The reduction of vehicle vibration is the method for improving comfort used in the system. Therefore, rides on rough roads will be less bumpy. Long trips, even on smooth and even roads tend to tyre the muscles and give a feeling of discomfort to the drivers and passengers. The control algorithm will strive to minimize these vibrations and keep them at a less intrusive level. It will especially be beneficial to people suffering from health problems such as lumbar pain or to safely transport victims of bone fracture.

Literature Review


Transmission of vibrations to passengers in a vehicle may occur in three paths: tactile, visual or aural. Vibrations of the frequency range 0-25 Hz are classified as ride vibrations and those in the range 25-25000 Hz are classified as noise. 25000 Hz is deemed as the limit of vibration frequency for a typical motor vehicle (Luque, Alvarez, and Vera, 2004). Tactile and visual vibrations fall in the ride vibration range and are the main determinants of passenger ride comfort.

The subjective response from whole-body vibrations are claimed to be dependent on not only the frequency, but also the intensity and duration of the vibrations acting on an individual (J''nsson, 2005). In many cases, the magnitudes of the whole-body vibrations overloads the human sensory system i.e. receptors, visual perception and sense of equilibrium. The gush of information that floods the brain requiring to be managed and interpreted tyres out the body. Prevalent, repetitive, low-frequency vibrations were shown to be tiring and transient vibrations were known to cause stress. The combined effect of both types of vibrations results in both physiological and psychological effect. These effects are categorized as comfort and perception factors, health risks and motion sickness (ISO 2631-1 1997)

Comfort and Perception

The passenger's perception of the excitation sources' effects on his/her body is shown in Figure. Excitation sources can be direct or indirect. Direct excitation sources are the engine, the driveline and the tyre/wheel assembly. Indirect excitation sources comprise surface irregularities of the road (bumps, potholes), road roughness (texture) and ground conditions (dry, wet). The aerodynamics of the vehicle may be categorized under either of the two.

Nahvi, Fouladi and Nor (2009) present a number of parameters that may be used in analysis of perception and comfort of passengers. Assessment of the vibrations transmitted were done using Vibration Dose Values (VDVs), kurtosis, frequency-response functions (FRFs) and power spectral densities (PSDs). Seat Effective Amplitude Transmissibility (SEAT) values were calculated from the VDVs to qualify the seat suspension as a vibration isolator. A notable experimental result is that at frequencies lower than 30Hz, vibrations are amplified (upto 5 times in adverse conditions) by the backrest in the fore-aft direction. However, the time taken to reach severe discomfort was long (3 hours). The VDV and FRF were mentioned as the best methods of vibration analysis.

The parts of the human body possessing specific resonance frequencies is an important fact to note (Table). Vibrations received by different parts of the body at their resonance frequencies can have their intensity amplified. This is injurious to the body as a whole. The lumped parameter model of the human body's biodynamic characteristics (Figure) is a noteworthy tool for evaluation in this regard (Sezgin & Arslan, 2012). It consists of masses representing the different parts of the human body interconnected by spring and damper pairs. In order to analyze the vertical vibration effects on ride comfort of the driver, Sezgin & Arslan (2012) used a 11 degree-of-freedom biodynamic model proposed by Qassem et al. (1994). The paper concluded that if a person travels in a vehicle with a passive suspension at a rate of 72 km/h for 5-6 hours on a smooth road, signs of discomfort will start to show. In the range of 8-10 Hz, the vehicle occupant will feel adverse physiological effects such as muscle spasms and abdominal pain. It was recommended that such vibrations should not be experienced for periods longer than 5 hours. However, an active suspension minimized these effects significantly.

Yet another tool used to measure human sensitivity to vibrations is a specialized dummy. Sensitivity to vibration is measured in the 12 axes presented in the ISO 2631-1 standard. Three three-axis accelerometers are connected to the chest, the bottom and the feet along with the pitch, roll and yaw angular accelerations of the seat cushion resulting in the 12 axes. Apart from this, dynamic finite-element models of the human body for vibration simulations are also used at times.


Luque, Alvarez and Vera (2004) put forward that ride comfort is not only dependent on the perception and transmission of vibrations to the passenger (Figure), though it is one of the major causes. Vibrations, much like noise, are an objective factor which means they can be measured regardless of the individual's unique comfort reaction. The subjective factors, on the other hand, are experienced differently by different people. As the sensation perceived by passengers relating to these subjective factors are never perfectly universal, they are difficult to evaluate and standardize. However, the fact that the objective factors have a much larger impact on ride comfort and passenger perception has led to promising research in improving the ride quality, this research being such an example.

A multitude of researches have been conducted to define limits for ride comfort over the years. Among the various tests carried out to assess subjective ride quality are shake table tests, ride measurements in vehicles and ride simulator experiments. These studies attempt to correlate test subject responses in qualitative terms (e.g. fairly uncomfortable, not uncomfortable) by considering parameters that correspond to vibration such as displacement, velocity, acceleration and jerk.

The ISO 2631-1, however, summarizes and presents these testing methods conveniently. Regarding comfort and perception, exposure to periodic, random and transient whole-body vibration estimation methods are given. Although the standard contains a broad range of estimation methods for vibrations other than those caused by vehicles, the data required for this research can be found in the form of the evaluation of comfort reaction of seated persons. The considered frequency is 0.5 Hz to 80 Hz. The ISO 2631-1 has been used in a large number of researches on vibration-related comfort in vehicles. A few of such literature referred to in this research are de Rafael (2007), Varghese (2013), J''nsson (2005), Nahvi, Fouladi and Nor (2009), Sezgin and Arslan (2012) and Gunston, Rebelle & Griffin (2003).

Health Risks

J''nsson (2005) classifies the adverse physiological and psychological effects of vibration on health as acute and chronic. Acute effects are those experienced to a severe and intense degree within a short time while chronic effects last longer or constantly recur. The physiological effects that were listed include:

Cardiovascular (deviations of heart rate, blood pressure)

Respiratory (increased ventilation and oxygen consumption)

Endocrine and metabolic (elevated body temperatures)

Relating to motor processes (fatigue of muscles and tendons)

Sensory (disorders in the sense of balance, illusions of moment)

Relating to the central nervous system (rhythmic changes in EEG)

Skeletal (chronic effects like degeneration and acute effects like fracture)

The study continued to claim that transient vibrations, rather than periodic ones, were especially harmful to the spine. It also noted that musculoskeletal disorders in the neck and arms were due to long term exposure to shock-type horizontal whole-body vibrations when seated.

The ISO 2631-1 (1997) also had a separate section and an annex discussing the health issues of whole-body vibrations. Even though the standard attempted to cover the effects on persons who are standing, seated or recumbent throughout the research, it only presents an evaluation method of seated persons when it came to health. The selected frequency range as 0.5-80 Hz and the vibrations considered were those that are transmitted to the body via the seat pan. Lumbar pain is the main concern of people experiencing such vibrations in a seated posture. The standard estimates the reason to be the biodynamic behavior of the spine and connected nervous system: horizontal and torsional displacement of the vertebrae, greater mechanical stress and hindrance to the nutrition and diffusion processes of the disc tissue. Degenerative processes in the lumbar segments like spondylosis deformans, osteochondrosis intervertebralis and arthrosis deformans are said to be common. Moreover endogenous pathological disturbances of the spine might be exacerbated. The digestive system, the genital/urinary system and the female reproductive organs are also affected but with a much lower probability. A caution zone for vibration exposure from 4 hours to 8 hours is presented in the ISO 2631-1 with the parameter characterizing the vibration magnitude being weighted acceleration.

Rufa'i et al. (2013) has investigated the prevalence, risk factors and impact of lumbar pain among professional drivers in Nigeria. Among the two hundred test subjects, the prevalence of lower back pain was estimated to be 73.5%. It was estimated to have affected 74% of the drivers' driving performance.

Motion Sickness

According to the ISO 2631-1, vibrations at the lowest frequencies (0.1 Hz ' 0.5 Hz) tend to cause motion sickness or kinetosis in individuals. At such frequencies, all parts of the body are said to move in the same oscillatory motion due to there being no resonance. The motion of the head, whether voluntary or involuntary, is mentioned as the major cause of motion sickness. The standard evaluates such vibrations in the seated and standing postures along the z-axis of the surface that supports the person. Due to absence of data, any motion sickness induced due to the roll and pitch motions are ignored.

Motion sickness is associated with prolonged exposure to lower frequency vibrations. The period at which the symptoms would peak are subjective and may differ from individual to individual. It has been found that the human body has the capability to adapt to this vibration range over several days and retain the resistance for some time. A quantity called the Motion Sickness Dose Value (MSDV), calculated from the frequency-weighted r.m.s. acceleration has been defined to evaluate this issue. However, this research will not be considering motion sickness as a major issue as lower frequencies are less frequent in land passenger vehicles.

The Importance of Correct Inflation Pressure

Tyre pressure is an important concern in vehicles because correct tyre pressure leads to better tyre performance, fuel economy and ride comfort. Additionally, demands for increased competitiveness in vehicle ride quality and emerging legislation make maintaining the correct tyre pressure a crucial task.

In 2000, The National Highway Traffic Safety Administration (NHTSA) of the Department of Transportation in USA passed the final rule for the TREAD Act. The new law established a new Federal Motor Vehicle Safety Standard that requires a Tyre Pressure Monitoring System (TPMS) to be implemented in every vehicle to warn the driver of underinflated tyres. The need for such a law in other less developed countries has yet to be realized.

The survey conducted prior to passing the new law revealed that ignorance as the main cause of vehicle owners being oblivious to maintaining correct pressure in the tyres. Only a mere 29% of the participants stated that they check the tyre pressure monthly. Another 29% stated they check their tyre pressure only if one appeared to be underinflated; 19% checked only when their vehicle was serviced; 5% checked before a long trip and 17% carried out tyre pressure checks in other occasions. Hence, 71% of the respondents had stated that they check if their tyres are at correct inflation pressures less than once a month. It is noteworthy that under-inflated or over-inflated tyres can hardly be distinguished by mere observation (Figure)

85% of tyre pressure losses occur as a result of slow leaks and 15% are rapid losses due to road hazards that cause punctures. Tyre manufacturers claim that slow leaks are caused by natural leakage and permeation (approximately 1 psi per month) and changes in ambient temperature owing to seasonal climatic change (about 1 psi per 10''F drop). Slow leaks may also stem from minor punctures.

Under-inflated tyres bring about many complications on the road and sometimes may be the cause of vehicle accidents. When under-inflated, a tyre's sidewalls flex and the temperature of the air inside increase. Consequently, the stress and the risk of failure is increased as well. A loss in lateral traction also occurs leading to difficult handling. These factors along with the increased stopping distance, skidding in lane change maneouvres or hydroplaning on wet surfaces may result in flat tyres and blowouts which are known culprits of deadly crashes.

Reduced tread life is also an issue in under-inflated tyres. As seen in Figure an under-inflated tyre has more pressure placed on its shoulders. This causes incorrect and faster tread wear. Goodyear estimates that a tyre's average tread life will drop by 68% if the pressure dropped by 18 psi. Assuming the process is linear, there would be a 1.78% tread life reduction for 1 psi drop. Moreover, underinflated tyres become flatter than intended while in contact with the road causing the tyre to deflect while it rolls, which builds up internal heat leading to increased rolling resistance and reduction in fuel economy.

Over-inflated tyres, on the other hand, are stiff and cause the tyre to maintain less contact with the road as seen in Figure. Such tyres are easily damaged by bumps and potholes in the road making them vulnerable to blowouts.

Researches such as Mathai & Ranjan (2015) and Ajas et al. (2014) have proposed systems that could monitor tyre pressure and alert the user when it is not within a desired range. Mathai & Ranjan (2015) has used the standard TPMS to sense tyre pressure and temperature. The NuMicro controller would interpret the readings and determine whether the tyre is at optimum inflation. A Zigbee module and the CAN protocol were used to wirelessly communicate the sensor data to the controller. A visual display was also built to alert the user of the current pressures in all four tyres. Ajas et al. (2014) has taken a step further and developed the system into a self-inflation system as well. The sensor used for pressure and vibration readings was a SMART transducer

Even though there is a manufacturer recommended tyre pressure for a vehicle's tyres, research shows that it is better if the tyre pressures could adapt to the terrain the vehicle travels on. This would result in improvement of fuel efficiency, ride quality and better overall performance of the tyres.

Research on Tyre Self-Inflation

Two notable papers that were analyzed to select the best method for automatic inflation were Alexander et al. (2006) and Ajas et al. (2014). The latter has used a compressor-driven pneumatic circuit for tyre inflation to a preset pressure. Alexander et al. (2006) comprised of a detailed concept generation section listing five different concepts along with their advantages and disadvantages.

The first concept they introduced was a thermoelectric heating device to control internal tyre pressure. When alerted of insufficient inflation pressure, a DC-operated motor pump would increase the temperature of the wheel rim. Consequently, the temperature of air inside the tyre would also increase by convection. This leads to an increase in tyre pressure according to the relationship, p'T. Positive traits of the system were identified to be effective control of pressure owing to the thermocouples, full automation, absence of moving components meaning less noise and vibration. The negative aspects included the absence of a method to set the cold tyre pressure, high energy consumption to supply the required temperature and the longer time taken for the components to reach the desired temperature.

The second idea was to incorporate an electromagnet in the wheel hub and a metal strip embedded inside the circumference of the tyre. A sensor detecting excess rolling resistance will send a signal to pass a current to the electromagnet, thus magnetizing it. The magnetic field hence created will repel the metal strip and expand the tyre profile. This effectively improves the stiffness of the tyre. The lack of mechanical components, minimal space and weight requirements are pros while inability to adjust cold tyre pressure, deformation of the tyre profile, additional maintenance and interference caused by the magnetic field are major drawbacks.

Another concept was to integrate miniature pumps into all four wheels. The wheel rotation would generate a suction force proportional to the rpm of the tyres, pushing air into the pump. This air is then used to inflate tyres when necessary. The design is completely automatic, low-pwer and provides an adequate flowrate. Nonetheless, ineffective cold tyre pressure setting, lack of proper control, increased vehicle drag due to constant air intake, extraneous strain on the drivetrain and unattractive appearance were the downsides.

A wheel rim with hollowed-out spokes containing ejector pins was the next design. At insufficient pressure, a DC motor in the wheel hub will cause the ejector pins to radially extend, reducing the volume occupied by the air and hence increasing pressure according to the equation, p'1/V. The advantages are complete automation, minimum space and weight requirements of the rim with the hollowed-out spokes and no change to the outward appearance. Lack of control of the cold tyre pressure, the ejector pins' volume being constrained by the spoke volume, potential noise and vibration and complete redesign of the wheel rim (which would be expensive) were drawbacks.

The final concept is a high pressure reservoir placed on the wheel rim that will supply air to the tyres via passageways when needed. An actuator-controlled valve will control this procedure while a pressure relief valve in the tyre will accommodate pressure reduction. Cold tyre pressure setting, automatic system and minimal power requirement were advantages. The rotating unbalance of the reservoir leading to vehicle instability, high maintenance and unattractive appearance were reasons they discarded the design.

Based on the analyses of aforementioned literature, the centralized compressor system for automatic tyre inflation was decided to be the best means of implementing our system. Possible challenges would be the limited space available for installing the pneumatic circuit in the vehicle, maintaining sufficient flowrate and quick response of the system to the controller's commands.

Concept Generation

 Mathematical Modelling of the Quarter Car Model

Initially, the standard quarter-car model was used to approximate the road condition, shown in Figure 1. The road height under the tyre is taken as y_road (t). m_us is the unsprung mass and includes the mass of one tyre and components attached to it. m_s, the sprung mass, is a quarter of the vehicle body's total mass. v_us and v_s are the vertical velocities of the unsprung and sprung masses respectively. Between the road and the unsprung mass, the rubber tyre is modelled as a spring of stiffness K_t and damper of constant B_t. The vehicle's suspension is modelled as yet another spring and damper with stiffness K_s and damping constant B_s respectively. If the rate of change of road height is y ''_road, a state-space model of the quarter car motion is obtained.

['(v ''[email protected]'(v ''[email protected]'(D ''[email protected] ''_u )))] = ['((-B_s)/m_s &B_s/m_s &K_s/m_s @B_s/m_us &-(B_( s)+ B_t)/m_us &(-K_s)/m_us @'([email protected])&'([email protected])&'([email protected]))    '([email protected]_t/m_us @'([email protected]))]['([email protected]'([email protected]'([email protected]_u )))] + ['([email protected]'(B_t/m_us @'([email protected])))] y ''_road

From the state-space model, a transfer function relating the vertical velocity of the unsprung mass, v ''_s and the rate of change of road height, y ''_road can be obtained.

(v_u (s))/(y ''_road (s) )=(m_s B_t s^3+(m_s K_t+B_s B_t ) s^2+(B_t K_s+B_s K_t )s+K_s K_t)/'(m_s m_us s^4+(m_s B_s+m_s B_t+m_us B_s ) s^[email protected]+(m_s K_s+m_s K_t+B_s B_t+m_us K_s+B_s^2-(B_s^2 m_us)/m_s ) s^[email protected](B_s K_s+B_s K_t+B_t K_s-(B_s K_s m_us)/m_s )s+K_s K_t )

The desired transfer function, however, is one that relates the road profile, y_road to the vertical acceleration of the wheel, v ''_u which can be measured by an accelerometer. This can be accomplished as follows,

(v ''_u (s))/(y_road (s))=v ''_u (s)(s/(y ''_road (s) ))=s^2 ((v_u (s))/(y ''_road (s) ))

The final transfer function to be analyzed is the inverse of the one above i.e. (y_road (s))/(v ''_u (s)),

(v_u (s))/(y ''_road (s) )='(m_s m_us s^4+(m_s B_s+m_s B_t+m_us B_s ) s^[email protected]+(m_s K_s+m_s K_t+B_s B_t+m_us K_s+B_s^2-(B_s^2 m_us)/m_s ) s^[email protected](B_s K_s+B_s K_t+B_t K_s-(B_s K_s m_us)/m_s )s+K_s K_t )/(m_s B_t s^5+(m_s K_t+B_s B_t ) s^4+(B_t K_s+B_s K_t ) s^3+K_s K_t s^2 )

A bode plot of this transfer function will portray the variation of the road profile at different readings of the accelerometer. However, there is no relation to the tyre pressure apparent in this equation. The tyre pressure is brought into the equation by a semi-empirical relation of the tyre pressure to the vertical tyre stiffness developed by TNO, Netherlands, given below,

K_ti=(1+p_fzi dp_i ) K_t0

where dp_i=(p_i-p_0)/p_0 , p_i being the tyre pressure to which the tyre is to be inflated and p_0 being the nominal tyre pressure. K_ti and K_t0 are the tyre stiffnesses at tyre pressures p_i and p_0 respectively. p_fzi is an empirical constant which was determined by plotting sample tyre stiffnesses (K_ti) against the dp_i values, linearizing the graph obtained, determining the gradient and dividing it by K_t0.

Assumptions (Bs=0?) and values of constants (Ks?) substituted to get the tf used to generate the bode plots. Nominal pressure setting?

The bode plots obtained by graphing the final transfer function (number?) are shown below.


As evident by the bode plots shown above, changes in the road profile are not distinctly observable in the transfer function at different pressures. So it was concluded that this method did not provide a basis for a workable terrain identification system.

Estimation of Comfort by the Frequency-Weighted r.m.s. Acceleration

The second method of estimating ride comfort chosen was based on the guidance and information provided in the International Standard ISO 2631-1. Here too, the method used to obtain vibration relevant data was the accelerometer. According to the standard, the method of calculating the weighted r.m.s. acceleration (aw) depended on a quantity called the crest factor. The crest factor was defined as 'the modulus of the ratio of the instantaneous peak value of the frequency-weighted acceleration signal to its r.m.s. value'. The peak value should be determined from the range of frequency-weighted accelerations over the duration of the measurement (T). When the crest factor of the vibrations is below or equal to 9, the usage of the following equation was advised.

a_w=[1/T '_0^T''a_w^2 (t)dt ']^0.5


a_w (t) is the weighted acceleration (translational or rotational) as a function of time (time history), in metres per second squared (m/s^2) or radians per second squared (rad/s^2) , respectively;

T is the duration of the measurement in seconds

The crest factor of vibrations in vehicles were not as high as 9 according to a previously done research (de Rafael, 2007). Therefore, the above equation was sufficient to decide the degree of comfort of the ride based on the comfort regions presented in the ISO 2631-1 (Table 1)

Frequency-weighted r.m.s. acceleration (aw) range Degree of comfort

Less than 0.315 ms-2 not uncomfortable

0.315 ms-2 to 0.63 ms-2 a little uncomfortable

0.5 ms-2 to 1ms-2 fairly uncomfortable

0.8 ms-2 to 1.6 ms-2 uncomfortable

1.25 ms-2 to 2.5 ms-2 very uncomfortable

Greater than 2 ms-2 extremely uncomfortable

The nature of the comfort regions is such that a Fuzzy logic model can easily be implemented.

Sensors and Data Acquisition

Measuring Vibration Intensity

As suggested in the ISO 2631-1, wheel vertical acceleration was the parameter used to measure vibration intensity. The sensor used to capture the accelerations was a three-axis accelerometer. The ADXL345 three-axis accelerometer (Figure 2) was chosen for this project. The main reason for using it was because of the easy interfacing with the Arduino Mega we had selected as the controller. The ADXL345 requires 2.0 ' 3.6 V DC power and has a very high resolution (4 mg/LSB) making it very sensitive to even the relatively small lateral vibrations of the vehicle. It has a dynamic range up to ''16g and supports both SPI and I2C communication. I2C communication was selected because of its need for fewer connections and more flexibility.

The accelerometer was configured to measure the acceleration in g's up to two decimal points and connected to the Arduino Mega for SPI communication (Figure 3). The coding procedure was greatly facilitated by the publicly available Adafruit ADXL345 library.  

Importing Arduino Serial Monitor Data Into MATLAB

In the earlier phases of the design, the calculation of the frequency-weighted r.m.s. acceleration (aw) was accomplished through a code in MATLAB. The equations of the ISO 2631-1 (Section 2.2) were used. On the other hand, the accelerometer readings required for the calculation were output only to the Arduino Serial Monitor. This called for some way to import the Arduino Serial Monitor Data into MATLAB during real-time operation.

Through extensive searching, a third-party software named PLX-DAQ was discovered. The PLX-DAQ served as an add-in to MS Excel (Figure 4) and was capable of displaying the Arduino Serial Monitor data in an Excel sheet. Similar to the Serial Monitor, the PLX-DAQ will also carry out the importing process in real-time. By simultaneously running the Arduino code for the accelerometer and the PLX-DAQ, it was possible to have the g values displayed directly on an Excel sheet.

Afterwards, the MATLAB code was modified to enable it to accept a .xlsx file for input. Consequently, the Excel sheet containing the accelerometer readings can be given as input to MATLAB.

Measuring Tyre Pressure

Using a Tyre Pressure Sensor

Initially, it was decided that a tyre pressure sensor was the most convenient choice for measuring tyre pressure. Tyre pressure sensors that are meant as replacements to damaged ones included in Tyre Pressure Monitoring Systems (TPMSs) are commonly available in the market. The Autel MX sensor (Figure 5) was chosen for our purposes. The Autel MX transmitted at 315 MHz and 433 MHz. As 433 MHz receiver and transmitter modules can be bought inexpensively, it was our favoured choice. The MX sensor vehicle coverage included Suzuki vehicles which made it suitable for the Alto used for trial runs.

The sensor battery and internal circuit is graded to meet severe weather conditions and consists of a high-precision air valve that prevents air leakage (Figure 6). The aluminium valve stem is corrosion-resistant and the clamp-in stem design further broadens its vehicle coverage. Other than measuring the tyre pressure, the sensor can also sense and transmit temperature, its Lithium battery state and location to a TPMS. The sensor and its transmitter's specifications are identical to the standard tyre pressure sensors manufactured by Orange Electric (Table 2)

Pressure Monitoring Range 0~76 psi

Temperature Monitoring Range -40''F to 257''F

Pressure Reading Accuracy ''1 psi

Temperature Reading Accuracy ''4''F

Battery 3.6V

Operating Frequency 433.92 MHz

Storage Temperature -40''F to 250''F

Operating Temperature -22''F to 248''F

Operating Humidity 95%

Transmission Power 5dBm max

The sensor is fixed on the tyre replacing the Schrader valve. While functioning as a sensing element, the valve also allows air intake when inflating the tyre.

Being specifically built for the purpose, the tyre pressure sensor was the best means of accurately reading the tyre pressures. However, the signals that the sensor sent can only be decoded by a TPMS computer. Hacking the TPMS signal was a complicated task and required building a separate circuit and manual analysis of the signal. This was both out of the scope of the project and incurred a cost that exceeded the budget. Due to these reasons, the tyre pressure sensor was discarded.

Using a Pressure Transducer

The tyre pressure sensor was to be replaced by a pressure transducer. An inexpensive generic pressure transducer with a working pressure range of 0-1.2 MPa was selected as the sensing device (Figure 6). A pressure of 1.2 MPa corresponds to roughly 174 psi, which is more than adequate for the purposes of this project. The maximum tyre pressure of a car seldom goes beyond 55 psi.

The specifications of the pressure transducer are given in Table 3

Working Voltage 5V DC

Output Voltage 0.5V - 4.5V DC

Sensor Material Carbon Steel Alloy

Working Current '10 mA

Working Pressure Range 0 - 1.2 MPa

Maximum Pressure Rating 2.4 MPa

Working Temperature 0 - 85''C

Storage temperature 0 ' 100''C

Measuring Error ''1.5%

Response Time ' 2.0 ms

Cycle Life 500,000 pcs

Application non-corrosive gas/liquid measurement

The maximum working temperature of the transducer is actually insufficient for practical application as tyre may heat up to around 120''C in hot days when hauling heavy loads. The tyre pressure sensor (Section 3.3.1), on the other hand, was well-suited for the purpose having a maximum working temperature of 248''F~120''C. Pressure transducers that can withstand greater heat are available, but costly. So, when actually implementing the system in the long run, the pressure transducer selected for this project must be replaced with a more suitable one.

The pressure transducer was fixed on the inner side of the rim with its sensing end exposed to the air inside the tyre. In order to fix the transducer in this way, a hole was bored on the rim's inner surface. The hole diameter was set to be the same as the diameter of the hex nut in the transducer. An epoxy adhesive was applied at the hex nut and rim contact surface to bind the transducer to the rim and make the fix airtight.

 Interfacing the Pressure Transducer with the Arduino

The pressure transducer's location on the inner side of the rim made a direct means of communication with the main controller (the Arduino Mega) problematic. The proposed solutions to the problem were wireless communication methods such as Wi-Fi, Zigbee and Radio Frequency Receiver-Transmitter modules. However, Wi-Fi and Zigbee modules were relatively expensive and they require separate Arduino shields (ArduinoWifi Shield and ArduinoXbee Shield respectively) for their wireless communication with the Arduino. These reasons made RF modules the preferred choice for the prototype.

The rather low cost FS1000A wireless transmitter-receiver module pair (Figure) was selected initially as the means for radio frequency communication. These modules use the RF 433MHz frequency with amplitude-shift keying (ASK) modulation. Literature about the applications of the receiver-transmitter pair were commonly available on the internet so making troubleshooting easier in case of malfunction.

As seen in Figure, both modules were manufactured with antennae requiring no extraneous components for implementing them at all. The required voltage for each module was 3-12V and they supposedly had the capability of communicating within a distance of approximately 20-200m, the range being 200m when powered by a 12V supply. The small size of the transmitter made it ideal for installation inside the tyre with minimal space requirement.

However, when testing the communication between these modules, several problems were encountered. The first problem came up regarding the range. As the maximum voltage that can be given to the transmitter by the Arduino was 5V, the transmission range was inadequate. The transmitter didn't even exhibit its minimal range of 20m when operating at this voltage. The second issue was regarding signal drop. Even if the receiver-transmitter pair was placed very close to each other and operated, there appeared to be an occasional signal drop. This made the use of these modules significantly unreliable.

The second choice of an RF transceiver was the NRF24L01+ (Figure). This particular module was chosen because it only required a voltage in the range 1.9V ' 3.6V to transmit at 2.4GHz frequency. It was said to have a range of about 1000m and could go up to a transmission rate of 2Mbit/s. As the voltage requirement of the NRF24L01+ could easily be supplied by the 3.3V output of an Arduino and its more than adequate range, this transceiver was chosen.

The pressure transducer was connected to an Arduino Nano, the small sizes of which made its installation inside the tyre possible. One transceiver module was also connected to the same Nano to serve as the transmitter. The three components formed one circuit inside the tyre (Figure) powered by a 9V battery. From there, the transmission of the said transmitter was received wirelessly by another NRF24L01+ connected to the Arduino Mega, which was the main controller. The pressure reading was configured to be output in psi value.

Mechanical Design

The Rotary Joint

The inflation of the tyres was to be achieved by pumping compressed air into the tyre through pneumatic tubes. However, the air lines cannot directly be sent into the tyre because the spinning wheels will twist the air lines and damage the pneumatic circuit. This posed a serious design challenge as to how the air lines can reach the tyres safely.

Further reading revealed that the Hummer H1 incorporated a self-inflation system with a centralized compressor as well. The Hummer H1's solution to safely routing an air line into the tyre was a completely new hub design. It is fitted with a geared hub and the axle shaft is fitted to the hub above the wheel's centre. This leaves the centre of the wheel with no solid shaft and allows the feeding of the air line directly into the tyre through the wheel centre. (Figure 7).

Name air line and axle shaft.

This project, however, aims to build a system that can be incorporated into any passenger vehicle. So, a total alteration of the existing components of a vehicle was to be avoided. Moreover, as the axle shaft runs directly into the wheel hub in passenger vehicles, there is no unhindered path for an air line.

The proposed solution to this problem was the design and construction of a rotary joint. The rotary joint was built to couple the stationary axle and the rotating wheel of the vehicle. More specifically, part of the rotary joint rotates with the drive axle hub and the rest remains stationary with the spindle. The hub was fixed to the rotating part of the rotary joint by passing it through the air chamber and held in place by the bearings of the rotary joint. When air is supplied to the air chamber inside the joint, it will be passed on to the air chamber of the joint and enter the wheel hub through a grease hole on the wheel hub housing.


The rotary joint was made of mild steel. The main considerations regarding the rotary joint were ensuring stable rotational motion and making it airtight. It consists of two 6905DU radial ball bearings of dimensions 25mmx42mmx9mm (a) on either side to allow for the rotary movement ' these bearings form the rotating part of the rotary joint. Pneumatic seals of dimensions 25mmx35mmx7mm (b) on either side prevent air from leaking out and lowering pressure. Air is routed to the rotary joint by connecting a pneumatic tube to the mouth of a circular groove (c) on the outer surface of its housing (e). Air enters the air chamber (d) via this circular groove cut through the housing to the air chamber. The wheel hub is inserted through the rotary joint's air chamber to which air from the chamber enters via a grease hole on the housing of the wheel hub.

The Quarter Car Model

In order to facilitate presenting the system, a quarter car model (Figure 8) and a vibrating platform was built. The vibrating platform simulates the vibrational input from a road and the quarter car model is a model of one tyre assembly of the vehicle.

A box bar of dimensions 3x1.5x16 inches was used to represent the swingarm. It was fitted so that it could pivot vertically and allow the suspension strut above it to absorb vibrations. The suspension strut used was common among three-wheelers in the country.

One end of the swing arm was fixed with the wheel hub and rotary joint assembly. To fix the hub and joint, a circular hole was drilled in the box bar and the parts were welded in place. A grease hole on the housing of the wheel hub was used to let the air hose in and route the compressed air into the air chamber (Figure 9).

The second box bar near the top of the assembly serves as the chassis or frame of the vehicle. The load of the vehicle is modelled by gym weights inserted into a thread bar. The thread bar is connected to the chassis box bar and the weights will bear down on it simulating the downward force on one strut.

Later, the base of the vibrating platform was welded to the base frame of the vibrating platform. However, this posed several problems which resulted in another change in the design, separating the bases of the two.

The Pneumatic Rotary Joint was welded inside the box bar and grinded

The Metal Bending(pressing) process in making the 'U' shaped vertical mount.

The Vibrating Platform

The vibrating platform (Figure 10) consists of a bed of rollers mounted on four vertical springs. The platform would move up down and simulate the road profile input frequency. It is assumed that the vibrations felt by the tyre are of constant frequency which is the case when it travels on roads with a consistent roughness. The motion of the platform is best implemented with an industrial grade vibration motor. However, such a motor is expensive.

e and were not within budget. The alternative chosen was to modify a regular motor with an eccentric mass to achieve a vibratory movement.

  Rollers before fixing onto the platform


The Motor and Controller

The motor needed was required to have the following capabilities:

The ability to work at different rpm's (at least 3) in order to simulate several road profile input frequencies.

The torque to bear a weight of around 1.5 kg on its shaft meant to be the eccentric mass (calculation of the mass is stated later in this section)

For these purposes, a DC brush-type motor of rated voltage 220 V, 0.5 HP (~0.37 kW) and 2900 rpm was selected. A motor controller was purchased to run the motor at different speeds.

Testing the Permanent Magnet DC Motor and calculating the starting current.

Calculation of the Eccentric Mass

In order to calculate the eccentric mass needed to vibrate the platform at the required amounts, a one degree-of-freedom model of an eccentric rotating mass vibration motor (ERM) was analyzed (Figure 10). It is assumed that the motor only vibrates in the vertical direction, which is true in most cases as the lateral displacement of the motor is comparatively smaller

The model consists of

The motor with mass M.

The eccentric mass, m and distance from motor shaft to centroid of the eccentric mass (eccentricity), e.

The resultant spring constant of the springs, k.

The damping constant, c.

The rpm of the motor, ''.

Vertical displacement of the motor, x.

The force exerted by the springs according to Hooke's Law is


The damping force of the springs is proportional to the velocity. Hence,

F_d=cx ''

The mass of the ERM without the added mass follows Newton's 2nd law,

F_m=(M-m) x ''

The sum of these three forces should be equal to the input.

F_i=me''^2 sin'(''t)

This results in the equation of motion for the system,


kx+cx ''+(M-m) x ''= me''^2 sin'(''t)

If we consider the moment when the eccentric mass has reached its maximum height (amplitude), sin'(''t)=1. As no external damping was applied on the springs, the damping of the springs were assumed to be negligible i.e. cx ''=0. The equation then reduces to

kx+(M-m) x ''= me''^2

To simulate the vibrations of the road, it was decided that an amplitude of about 5 mm was suitable i.e. x=5 mm when the motor ran at 1000 rpm i.e. ''=1000 rpm=104.72 rad/s. The spring constant of a typical suspension strut was found to be approximately k=4*'10'^4 N/m. The platform should hold an unsprung mass of 40 kg to simulate the load of the car i.e. M=40 kg and the vertical acceleration was set to be the grand mean of the z axis accelerations obtained in the trial runs on different road conditions i.e. x ''=1.00185g=9.82813m/s^2. The eccentricity of the mass was decided to be 1cm i.e. e=1cm constrained by the physical dimensions of the motor and the gap between the motor and the platform's rollers.

Substituting these values to the equation (Number), we get,

(4*('10'^4 N)'(m))*(5*'10'^(-3) m)+(40kg-m)*(9.82813ms^(-1) )=m*(1*'10'^(-2) m)*(104.72 rad's)^2

which gives

m=0.88818 g=888.18 g

As a result, an 800g round mass was selected as the eccentric mass. The eccentricity from the motor's centre was 1cm as per the above reasoning.

Construction of the Eccentric mass and mounted onto the motor shaft.

Reservoir Tank.

For the reservoir tank, it was used a disposing refrigerant gas tank. The emptied gas tank was fiited with a 'T union ' for the two pneumatic lines; from air compressor to the reservoir tank and the outgoing line from the reservoir tank to the rotary joint. A pressure transducer was fixed onto the tank in order to measure the internal air pressure. This tank can store compressed air upto around 100 psi.

Welding a sealed nut to the reservoir tank to set-up the pressure transducer.

Controller Design & Power Management

 Controller Design

It was decided that Fuzzy control logic would be the best choice for controlling the developed system. The reasons for this choice were Fuzzy logic's inherent ability to deal with a system not having an accurate quantitative model and its use of linguistic variables which facilitates the incorporation of the comfort zones presented in the ISO 2631-1 (Figure_in_lit_rev).

One requirement of a Fuzzy logic controller was a smoothly varying input-output relationship. It was observed that increasing the overlap of the input membership function increased the smoothness of the input-output relationship. The manipulation of the membership function and observation of the output were done in Matlab.

Figure 1 shows Fuzzy control with zero overlap in the membership function. The output level sharply rises from a constant lower level to a constant higher level. Hence, there is no gradual variation of the output. The sharp rise in output is undesirable in practice because the output of the system developed for this project is tyre pressure regulation. When the input vibration falls in the region of the sharp rise, a slight change in it will set a significantly higher pressure as the output, ignoring a large range of pressure values in between.

This behavior is somewhat rectified by allowing a small overlap of the input membership functions as in Figure 2. However, when the overlap is increased much further, as in Figure 3, an input-output variation that is feasible for the developed system will be obtained. As seen in Figure 3, the variation is very smooth so that small changes in the input (vibration) will cause a reasonable change in the output (tyre pressure) over all values of the input.

Mamdani inference method was used for the Fuzzy logic implemented. The rulebase formed consisted of 12 rules. The total range of the membership functions was considered when forming the rulebase. When configuring the Fuzzy rulebase, the AND and OR functions were used. The minimization method was used for the AND function and maximization for the OR function. The aggregation method was the maximization method. Defuzzification was performed using the centroid method.

There were 2 inputs to the Fuzzy controller ' the frequency-weighted r.m.s. of the vertical acceleration of the tyre and the tyre pressure. The output was the tyre pressure correction (the amount by which the tyre had to be inflated or deflated). The input space of the frequency-weighted r.m.s. acceleration was split into 4 membership functions with a range of 0.315 m/s2 ' 2.5m/s2. They are

a little uncomfortable: 0.315 m/s2 to 0.630 m/s2

fairly uncomfortable: 0.500 m/s2 to 1.000 m/s2

uncomfortable: 0.800 m/s2 to 1.600 m/s2

extremely uncomfortable: 1.250 m/s2 to 2.500 m/s2

The input space of the tyre pressure input, ranging from 15 psi to 40 psi, was as follows

low pressure: 15 psi to 25 psi

medium pressure: 22.5 psi to 32 psi

high pressure: 30 psi to 40 psi

The output range was chosen to be -15 to +15 with two membership functions

decrement: -15 psi to 3 psi

increment: -3 psi to 15 psi

Power Management

The power supply used initially could supply a maximum of 220V. However, the voltage was stepped down to 110 V using a step-down transformer.

The controlling method of motor speed was, initially, a triac-based motor driver (Figure). The driver was compatible with the motor. The driver was, however, rated at 60 W and drivers having a rated wattage similar to the motor's rated wattage (0.37W) were not available. The unmatched power rating resulted in the driver not being able to provide the motor starting current. For the motor to run, the motor terminals had to be first connected to the main supply (220V) for the starting current and later connected to the triac-based driver for speed control. This was undesirable. To provide the starting current via driver itself, a capacitor was connected in shunt to the motor. The remedy did not work as the capacitor shorted and fried the driver.

Subsequent efforts to connect a potentiometer directly to the 220V power supply for motor speed control failed. The potentiometer would get heated up to a point where it malfunctioned. Around this time, the transformer winding of the power supply shorted and became unusable. To replace the single power supply and transformer, nine 12V power supplies, one 24V power supply and a 48V supply were connected in series. The supplies were set to deliver their maximum voltages by tuning the presets. The multi-meter measured a total output of 218V because the power supplies delivered greater voltages than rated.

A carbon potentiometer, which could withstand greater temperatures than the one used earlier, was used with the supplies' output. It worked well under 60V but a continuous supply of 220V for about ten minutes made this potentiometer malfunction as well. The reason was thought to be the unmatched wattage of the potentiometer and the motor. Finally, industrial-grade selector switches of 4kW power rating were bought. One selector switch could switch between two voltages. So, two selector switches were used to switch between four speeds at voltages 220V, 150V, 120V and 60V.


 Potentiometer which was used during preliminary testing

Carbon Potentiometer

220VAC ' 110VAC Step down Transformer with 110V rectifier.

12V portable tyre compressor used in the system.


Trial Runs to Acquire Real-Time Readings

Before implementing the self-inflation system, to make sure that the frequency-weighted r.m.s. acceleration was a reliable parameter to estimate the comfort reactions of passengers, several trial runs were conducted. Firstly, the accelerometer was fixed to the axle above the rear left wheel of a Suzuki Alto 800 (Figure ?). Some characteristic parameters of the car:-

Engine Type: 12-valve, 796 cc, 3 cylinder

Engine Power: 47bhp

Engine Torque: 7.03kgm

Fuel: Petrol

Transmission Type: Front wheel drive

Gearbox: Manual, 5-speed

Ground Clearance: 160mm

Weight of chassis: 720kg

Tyres: 145/80R12 (tubeless)

Front Suspension: Gas-filled McPherson strut, Torsion roll control device

Rear Suspension: Coil spring, gas-filled shock absorbers with three-link rigid axle and isolated trailing arm

Trial Run 1

The accelerometer was fixed in a way that its z-axis pointed upwards. It was connected to the Arduino by a wire and the Arduino was powered by a laptop which doubled as a data acquisition tool. The Arduino serial monitor and MS Excel with the PLX-DAQ add-in were opened on the computer. The acceleration readings were directly obtained on an Excel worksheet by clicking on the 'Connect' button in the PLX-DAQ GUI.

The vehicle was driven maintaining a speed in the range 40 kmph ' 60 kmph over three types of terrain: asphalt (300m), irregular gravel (50m) and concrete (50m). The roads chosen for the trial runs were in the vicinity of the university and had the typical features of roads in their terrain type. Three runs were carried out per road at three tyre pressures, 35psi, 25psi and 15psi resulting in a total of six runs. The data obtained was then given as input to the MATLAB code that calculated the frequency-weighted r.m.s. acceleration. The outputs were analyzed to ensure that the comfort reactions of the passengers agreed with the information provided in the ISO 2631-1.

The following evaluation was done using MATLAB for the data derived from Trial Run -1.

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 35 psi

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 25 psi

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 15 psi

Orientation : z axis directing vertically upwards

Terrain Type : Coarse Gravel(bumpy)

Tyre Inflation Pressure : 35 psi

Orientation : z axis directing vertically upwards

Terrain Type : Coarse Gravel(bumpy)

Tyre Inflation Pressure : 25 psi

Orientation : z axis directing vertically upwards

Terrain Type : Coarse Gravel(bumpy)

Tyre Inflation Pressure : 15 psi

The analyzed RMS data for the Trial Run ' 1 are as follows;

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 35 psi

Frequency Weighted RMS : 0.6388 m/s^2

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 25 psi

Frequency Weighted RMS : 0.6593 m/s^2

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 15 psi

Frequency Weighted RMS : 0.7499 m/s^2

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 35 psi

Frequency Weighted RMS : 0.9528 m/s^2

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 25 psi

Frequency Weighted RMS : 0.7829 m/s^2

Orientation : z axis directing vertically upwards

Terrain Type : Fine Gravel

Tyre Inflation Pressure : 15 psi

Frequency Weighted RMS : 0.8652 m/s^2

According to the results derived from the trial run ' 1, the conclusion was made that, the vibration RMS value was changed in par with the varying tyre inflation pressure. Here when evaluating the acquired data, a time window of 30 seconds was considered and the RMS was calculated accordingly. When analyzing the RMS values and the graphs, a certain correlation couldn't be built according to the randomness of the vibration data. Hence Trial Run ' 2 was decide to conduct.

Trial Run ' 2

Trial Run ' 2 was conducted with the purpose to acquire data to estimate the effect of tyre inflation pressure to the vibration RMS that is sensed by the Sprung weight and Unsprung weight at simultaneous time instances. 2 accelerometers were utilized where one was fixed onto the Rear Axle in order to acquire the vibration data of the unsprung mass and the second accelerometer was mounted onto the floor of the car(in the middle portion of the vehicle), in order to acquire the vibration data of the sprung mass. Both the accelerometers were connected to two separate arduinos and simultaneous real time readings were acquired and recorded and saved to an MS Excel Worksheet using PLX-DAQ software. The readings were acquired under 3 terrain conditions at manufacturer's recommended tyre inflation pressure of 30 psi. The analyzed time series evaluations and the corresponding frequency weighted RMS values are as follows;

Accelerometer Placement : Axle Mounted

Orientation : z axis directing vertically upwards

Terrain Type : Asphalt

Tyre Inflation Pressure : 30 psi

Temperature : 31'C

Acceleration Range : -0.6G< a(t) < 0.8G

Accelerometer Placement : Floor Mounted

Orientation : z axis directing vertically upwards

Terrain Type : Asphalt

Tyre Inflation Pressure : 30 psi

Temperature : 31'C

Acceleration Range : -0.15G < a(t) < 0.2G

Accelerometer Placement : Axle Mounted

Orientation : z axis directing vertically upwards

Terrain Type : Gravel

Tyre Inflation Pressure : 30 psi

Temperature : 31'C

Acceleration Range : -0.8G < a(t) < 0.9G

Accelerometer Placement : Floor Mounted

Orientation : z axis directing vertically upwards

Terrain Type : Gravel

Tyre Inflation Pressure : 30 psi

Temperature : 31'C

Acceleration Range : -0.2G < a(t) < 0.2G

The above time series acceleration data were fed into the MATLAB program and corresponding Frequency Weighted RMS Values were calculated;

Accelerometer Placement : Axle Mounted

Orientation : z axis directing vertically upwards

Terrain Type : Asphalt

Tyre Inflation Pressure : 30 psi

Frequency Weighted RMS value ; 0.6895 m/s^2

Accelerometer Placement : Floor Mounted

Orientation : z axis directing vertically upwards

Terrain Type : Asphalt

Tyre Inflation Pressure : 30 psi

Frequency Weighted RMS value ; 0.1506 m/s^2

Accelerometer Placement : Axle Mounted

Orientation : z axis directing vertically upwards

Terrain Type : Gravel

Tyre Inflation Pressure : 30 psi

Frequency Weighted RMS value ; 0.9866 m/s^2

Accelerometer Placement : Floor Mounted

Orientation : z axis directing vertically upwards

Terrain Type : Gravel

Tyre Inflation Pressure : 30 psi

Frequency Weighted RMS value ; 0.172 m/s^2

By analyzing the time series acceleration data of above conditions it can be identified that the typical vertical acceleration ranges vary in accordance with 2 conditions.

Terrain Quality

Acceleration Mounting Position

When evaluating the RMS acceleration data, it can be identified that, the RMS acceleration of the axle mounted accelerometer lies between  0.6895 m/s^2  and 0.9866 m/s^2 while the floor mounted accelerometer has a range between  0.1506 m/s^2 and 0.172 m/s^2. This shows that the effect of the road profile vibrations to the Unsprung mass is more effective than that of  Sprung mass. This can be happened because of the heavy damping effect applied by the suspension system of the vehicle.

 Troubleshooting the experimental setup

Incorrect or widely varying readings from the pressure transducer ' The reason behind this was found to be a leakage current from the motor propagating to the transducer as they were physically connected via the quarter car model. The leakage current caused inaccurate readings or large fluctuations in the reading inconsistent with the actual pressure inside the tyre. The first pressure transducer suffered permanent damage to its sensing diaphragm because of this and had to be replaced by a new one.

Remedy: Isolate the bases of the vibrating platform and the quarter car model so that there is no conductive path for the current to flow through to the transducer. To do this, the quarter car model weld was grinded off and fixed to a separate platform. Thus, the only contact between the two setups was the tyre, which is an insulator. To remove the malfunctioning transducer for replacement, the epoxy adhesive binding site was soaked in acetone for an hour. Afterwards, a hot air gun was used to unglue the transducer. The new transducer was fixed in the same way as the old one ' using epoxy adhesive.

The vibrations generated by the motor was transmitted to the tyre through the quarter car model ' In practice, the vibration of the tyre was mainly due to the variations in the road profile apart from the engine and other moving parts of the vehicle which were not within the scope of the project. When the bases of the platform and the model were welded together, vibrations passed to the tyre through the quarter car model other than the vibrating platform. This was undesirable as the vibrations should only be transmitted to the tyre by the platform, which simulated the road condition.

Remedy: The separation of the two bases as explained before took eliminated this problem as well.

The vibrations of the platform caused it to slightly move on the ground ' The platform had to be kept still and would otherwise cause unnecessary relative motion between the tyre and the platform.

Remedy: Among the proposed methods of eliminating this motion were a thick rubber sheet under the platform to damp the vibrations, placing it inside a cavity in a concrete slab and modifying the structure by welding two crossbars connecting the platform base and upper part. The preferred remedy was to anchor bolt the platform to the ground which would completely eliminate motion. However, this was not possible because the ground had to be damaged, which was not permitted in the laboratory the experimentation was performed in. Industrial-grade rubber suckers (Figure) used for lifting glass panes were available and several of these were used to hold the platform to the ground and constrain motion. This was only a temporary solution as the ground was uneven, resulting in the eventual loosening of the suckers as they couldn't hold the vacuum created.

Results and Discussion

 Testing the Set-up

After testing the apparatus under controlled conditions, the below mentioned results were determined. The apparatus was run under 9 testing conditions;

i) The voltage of the vibrating motor = 60V

ii) Set Tyre Inflation Pressure = 15 psi.

i) The voltage of the vibrating motor = 60V

ii) Set Tyre Inflation Pressure = 30 psi.

i) The voltage of the vibrating motor = 60V

ii) Set Tyre Inflation Pressure = 42 psi.

i) The voltage of the vibrating motor = 120V

ii) Set Tyre Inflation Pressure = 15 psi

i) The voltage of the vibrating motor = 120V

ii) Set Tyre Inflation Pressure = 30 psi

i) The voltage of the vibrating motor = 120V

ii) Set Tyre Inflation Pressure = 41.3 psi

i) The voltage of the vibrating motor = 220V

ii) Set Tyre Inflation Pressure = 15 psi

i) The voltage of the vibrating motor = 220V

ii) Set Tyre Inflation Pressure = 30 psi

i) The voltage of the vibrating motor = 220V

ii) Set Tyre Inflation Pressure = 40.9 psi

The conditions under which the apparatus was tested were,

The unsprung mass = 15kg

Room temperature = 26''C

The readings of the accelerometer were obtained in the Arduino serial monitor at each of the tyre pressures and voltages listed above. These readings were transferred to MS Excel using the PLX-DAQ. It must be noted that the accelerometer was connected such that the y-axis was pointed vertically down towards the ground. The y-axis accelerometer readings were filtered out from the Excel sheet. Graphical representations of these readings are shown below (figure). They were then fed into the Matlab code which produced the frequency-weighted r.m.s. acceleration as the output.

The frequency-weighted RMS acceleration values returned from the Matlab code are shown below.

In these graphs, the composite weighted level (AW) is the same as the frequency-weighed r.m.s. acceleration mentioned in the ISO 2631-1 and given by the equation Number.

The fourth power vibration dose (VDV) is the vibration dose value mentioned in the standard and given by the equation Number.

  Frequency Weighted Acceleration : 0.5816 m/s^2

  Frequency Weighted Acceleration : 1.935 m/s^2

 Frequency Weighted Acceleration : 1.266 m/s^2

 Frequency Weighted Acceleration : 2.38 m/s^2

 Frequency Weighted Acceleration : 4.035 m/s^2

Frequency Weighted Acceleration : 3.93 m/s^2

 Frequency Weighted Acceleration : 3.537 m/s^2

Frequency Weighted Acceleration : 4.599

  Frequency Weighted Acceleration : 4.981 m/s^2


It can be observed that the frequency-weighted r.m.s. acceleration has shown a distinct variation with the change in tyre pressure. The comfort zone to which each of the above results fall in, according to the Fuzzy controller used in this project, are given below

Condition 1 (60 V, 15 psi): a little uncomfortable

Condition 2 (60 V, 30 psi): extremely uncomfortable

Condition 3 (60 V, 42 psi): uncomfortable

Condition 4 (120 V, 15 psi): extremely uncomfortable

Condition 5 (120 V,30 psi): extremely uncomfortable

Condition 6 (120 V, 41.3 psi): extremely uncomfortable

Condition 7 (220 V, 15 psi): extremely uncomfortable

Condition 8 (220 V, 30 psi): extremely uncomfortable

Condition 9 (220 V, 40.9 psi): extremely uncomfortable

Here through these test results determined from the quarter-car model, we can explicitly claim that the system identifies the road condition as expected and depicted desired results. During 60V vibration condition, the initial tyre pressure was set to manufacturer's recommended value of 30 psi. At 30 psi, the Fuzzy Inference System depicted that the comfort level of the quarter car model was at 'extremely uncomfortable' region. Then the Fuzzy rules were fired and the tyre pressure was set by the Fuzzy Output function to the value of 15 psi. Then as expected, the comfort level was elevated from 'extremely uncomfortable' to 'a little uncomfortable'. So, the tyre pressure was optimized in order to make the system reach the minimum vibration limits under the prevailing conditions.  

The outputs from the controller are shown below,

Frequency Weighted RMS Acceleration : 0.58 m/s^2

Tyre Inflation pressure : 15 psi

Fuzzy Output Pressure Change : + 4.67 psi

Frequency Weighted RMS Acceleration : 1.93 m/s^2

Tyre Inflation pressure : 30 psi

Fuzzy Output Pressure Change : -4.85 psi

Frequency Weighted RMS Acceleration : 1.26 m/s^2

Tyre Inflation pressure : 42 psi

Fuzzy Output Pressure Change : - 5.51 psi

Frequency Weighted RMS Acceleration : 2.38 m/s^2

Tyre Inflation pressure : 15 psi

Fuzzy Output Pressure Change : - +5.72 psi

Frequency Weighted RMS Acceleration : > 2.5 m/s^2

Tyre Inflation pressure : 30 psi

Fuzzy Output Pressure Change : - -4.85 psi

Frequency Weighted RMS Acceleration : >2.5 m/s^2

Tyre Inflation pressure : 42 psi

Fuzzy Output Pressure Change :  -5.72 psi

By evaluating these Fuzzy Outputs, it is clear that the fuzzy controller always attempt to fluctuate the tyre inflation pressure to an point according to respective conditions. These values are sent to the main Arduino Program and inside the 'loop' of the program, this value will be compared with the prevailing tyre pressure and pressure will be changed until the difference between the 'desired tyre pressure' and the 'prevailing tyre pressure' is equal to zero.

(Prevailing Tyre Pressure) + (Fuzzy Output Pressure Value) = (Desired Tyre Pressure) ;

(Desired Tyre Pressure) ' (Prevailing Tyre Pressure) = x;

Until, x = 0 the process will be carried out.

 Tyre Wear Analysis.

An important question arose during the course of carrying out activities for the project. This question was: How will the change of tyre pressure affect the wear of the tyre over time? Prior research was referred to in order to find a satisfactory answer. In one study (Tong and Jin, 2012), the variation of tyre wear with tyre pressure was analyzed by connecting 84 sensory nodes that detected the wear depth. The wear depth was defined as 'the wear amount of tyre pattern groove in the lateral direction of the tyre surface'. The wear depth vs. node distance graph is given below (Figure)

The three pressures used are equivalent to 36 psi, 43.5 psi and 50.7 psi. The pressure range covers the range of tyre pressures that most passenger vehicle tyres are subjected to. The greatest wear occurs in the middle portion of the curve and the magnitude of wear depth ranges from approximately 2.9 mm ' 3.6 mm. in this region, the higher the pressure, the greater the wear depth. This is evident by a manipulation of the equation to accommodate tyre pressure (Number) given in the same research.

R_w= ''^2 pfr_0 (1+''(t_s-t_0 )+cd)


R_w= The rate of wear (wear depth differentiated by time)

'' = slip angle

p = resilience of the wheel

f = wheel stiffness

r_0,'',c,d = material constants

t_s = tyre surface temperature

t_0 = reference temperature

Consider the equation from (Varghese, 2013), which was used before in the concept generation phase,

f=K_ti=(1+p_fzi dp_i ) K_t0

By substituting to f in equation Number, we get

R_w= ''^2 pK_t0 r_0 (1+p_fzi dp_i )(1+''(t_s-t_0 )+cd)

This form of the equation shows us that R_w increases when dp_i=(p_i-p_0)/p_0  , which depends on the tyre pressure at that time, p_i , is increased. This result is consistent with the graph (Figure)

Consider the middle region where the maximum wear occurs. It can be observed that the maximum increase in wear depth when the pressure is increased from 0.25 MPa (36 psi) to 0.35 MPa (57 psi) is about 0.3mm which is a negligible amount of increase in tyre wear.


The hypothesis of this research,  ' An optimized tyre inflation pressure exists for each type of road condition when the roads are classified according to vibration RMS data'  was successfully proved with necessary data evaluation. Proving this hypothesis, the necessity of designing an Automatic Self Inflating Tyre System was emerged. As the main outcome of the project a prototype of  the  automatic self tyre inflating system was designed and tested. The system was tested to be functioning successfully under the respective terrain conditions.

Future Developments and Recommendations.

Solar powered battery charger circuit to power up the TPMS(Tyre Pressure Monitoring Sensor)

Piston Vibrator to replace the vibrating motor.

Installing 12V portable air compressor with a greater flow rate than the prevailing air compressor.

It's recommended to use industrial grade TPMS rather than pressure transducer.

It's recommended to use the method of anchor bolt fixing of the vibrating platform and the mechanical structure instead of using the suction cup fixing method in order to withstand higher degree of vibration.

This system is strictly recommended only for the land vehicles equipped with tubeless pneumatic tyres.

...(download the rest of the essay above)

About this essay:

This essay was submitted to us by a student in order to help you with your studies.

If you use part of this page in your own work, you need to provide a citation, as follows:

Essay Sauce, . Available from:< > [Accessed 17.01.20].