Citation
Measurement and performance analysis of pneumatic braking systems on heavy vehicles

Material Information

Title:
Measurement and performance analysis of pneumatic braking systems on heavy vehicles
Creator:
Martonovich, Mathew
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Civil Engineering, CU Denver
Degree Disciplines:
Civil engineering
Committee Chair:
Janson, Bruce
Committee Members:
Marshall, Wes
Haynes, Roxanne

Notes

Abstract:
Heavy commercial vehicles account for a large percentage of vehicles operating on public roadways, and an even larger percentage of vehicle miles traveled. The pneumatic braking system which heavy vehicles are equipped with, is often found to be deficient by law enforcement and investigators during roadside inspections or following a collision. Deficient pneumatic braking systems increase the distance necessary for a fully laden heavy vehicle to slow or stop. The stopping distance of a heavy vehicle, even when the braking system is fully functional, is significantly longer than that of a passenger vehicle. Slow deceleration rates and long stopping distance can, and do, result in collisions. The licensed commercial driver of a heavy vehicle is federally mandated to complete pre-trip inspections. These pre-trip inspections, and inspections completed by maintenance personal during the servicing of the vehicle, are often inadequate and may fail to identify burgeoning problems within the vehicle’s braking system. Current technology in heavy vehicle braking systems lessens the burden on the driver to constantly maintain the braking system, but does not eliminate the need for regular inspections, and maintenance if needed. Even with current technology, it remains common to find the adjustment of the braking system components on over-the-road heavy vehicles beyond safe limits during roadside inspections. This project presents an overview of the mechanical function of the modern S-cam pneumatic braking system commonly found on heavy vehicles in the United States. The areas of failures and potential deficiencies within S-cam pneumatic brakes is explored and discussed. Additionally, several different pneumatic braking performance analysis methods are presented, which can be used to determine the maximum deceleration rate and speed of a heavy vehicle with a fully operational braking system, and when the braking system is compromised. Examples of each analysis method are included in appendices to this paper. This project presents research examining the potential use of an ultrasonic sensor to, in real-time, measure and relay the adjustment of a heavy vehicle’s brakes to the operator of the vehicle. In this way the operator of the vehicle will be alerted when the adjustment limit of a brake on their vehicle or trailer(s) is reached. Additionally, an alternative design for the studied sensor, which would fulfill the same purpose, is presented in detail.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Mathew Martonovich. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

MEASUREMENT AND PERFORMANCE ANALYSIS OF PNEUMATIC BRAKING SYSTEM S ON HEAVY VEHICLES b y MATHEW MARTONOVICH B.S., Met ropolitan State University, 2008 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Civil Engineering Program 2017

PAGE 2

ii This thesis for the Master of Science degree by Mathew Martonovich has been approved for the Civil Engineering Program b y Bruce Janson , Chair Wes Marshall Roxanne Haynes Date: December 16, 2017

PAGE 3

iii Martonovich, Mathew ( M . S . , Civil Engineering Program ) Measurement and Performance Analysis of Pneumatic Braking System s on Heavy Vehicles Thesis directed by Professor Bruce N. Janson ABSTRACT Heavy commercial vehicles account for a large percentage of vehicles operating on public roadways, and an even larger percentage of vehicle miles traveled. The pneumatic braking system which heavy vehicles are equipped with, is often found to be deficient by law enforcement and investigators during roadside inspections or following a collision. Deficient pneumatic braking systems increase the distance necessa ry for a fully laden heavy vehicle to slow or stop. The stopping distance of a heavy vehicle, even when the braking system is fully functional, is significantly longer than that of a passenger vehicle. Slow deceleration rates and long stopping distance can , and do, result in collisions. The licensed commercial driver of a heavy vehicle is federally mandated to complete pre trip inspections. These pre trip inspections, and inspections completed by maintenance personal during the servicing of the vehicle, are often inadequate and may fail to identify burgeoning problems within the burden on the driver to constantly maintain the braking system, but does not eliminate the n eed for regular inspections , and maintenance if needed. Even with current technology, it remains common to find the adjustment of the braking system components on over the road heavy vehicles beyond safe limits during roadside inspections. This project pr esents an overview of the mechanical function of the modern S cam pneumatic braking system commonly found on heavy vehicles in the United States. The areas of failures and potential deficiencies within S cam pneumatic brakes is explored and

PAGE 4

iv discussed. Addi tionally, several different pneumatic braking performance analysis methods are presented, which can be used to determine the maximum deceleration rate and speed of a heavy vehicle with a fully operational braking system, and when the braking system is comp romised. Examples of each analysis method are included in appendices to this paper. This project presents research examining the potential use of an ultrasonic sensor to, in real rator of the vehicle. In this way the operator of the vehicle will be alerted when the adjustment limit of a brake on their vehicle or trailer(s) is reached. Additionally, an alternative design for the studied sensor, which would fulfill the same purpose, is presented in detail. The form and content of this abstract are approved. I recommend its publication. Approved: Bruce N. Janson

PAGE 5

v ACKNOWLEDGMENTS employer Dr. Jerry S. Ogden, PE for encouraging my pursuit of knowledge and teaching me the skills to succeed. Without your encouragement and support this would not have been possible. I would also like to thank Dr. Bruce Jansen, for providing direction an d support throughout my graduate education. Without your guidance and advice this project would not have become a reality.

PAGE 6

vi TABLE OF CONTENTS CHAPTER I. BACKGROUND ................................ ................................ ................................ ......... 1 II. H ISTORY OF HEAVY VEHICLE PNEUMATIC BRAKING SYSTEM ................ 4 III. H EAVY VEHICLE PNEUMATIC BRAKING SYSTEM CONCEPTS ................... 7 Compressor and Dryer ................................ ................................ ............................ 9 Braking System Plumbing ................................ ................................ .................... 10 Brake Chambers ................................ ................................ ................................ .... 12 Slack Adjustor ................................ ................................ ................................ ....... 15 S Cam ................................ ................................ ................................ ................... 18 Brake Shoes and Drums ................................ ................................ ........................ 19 Valves and Modulators ................................ ................................ ......................... 20 Anti Lock Brakes ................................ ................................ ................................ .. 20 IV. B RAKING SYSTEM DEFICIENCIES ................................ ................................ .... 22 Excessive Pushrod Stroke ................................ ................................ ..................... 23 Thermal Expansion and Failures ................................ ................................ .......... 26 Air Pressure and Delivery ................................ ................................ ..................... 28 Flui d Contamination ................................ ................................ ............................. 29 V. B RAKING PERFORMANCE ANALYSIS ................................ ............................. 31 Commercial Vehicle Factor Method and Skid to Stop ................................ ......... 31

PAGE 7

vii Weight Distribution Method ................................ ................................ ................. 33 Heusser Method ................................ ................................ ................................ .... 34 Bartlett/Heusser Method ................................ ................................ ....................... 35 VI. D ATA COLLECTION AND TESTING ................................ ................................ . 38 Test Series #1 ................................ ................................ ................................ ........ 41 Test Series #2 ................................ ................................ ................................ ........ 43 Test Series #3 ................................ ................................ ................................ ........ 45 Test Series #4 ................................ ................................ ................................ ........ 47 Test Series #5 ................................ ................................ ................................ ........ 49 Test Series #6 ................................ ................................ ................................ ........ 50 Test Series #7 ................................ ................................ ................................ ........ 51 VII. D ATA AND DESIGN ANALYSIS ................................ ................................ ........ 54 Test Series #1: 1.5 Inch Stroke Tests ................................ ................................ ... 56 Test Series #2: 1.5 Inch Stroke and Wet Brake Chamber Face Tests .................. 58 Test Series #3: 1.5 Inch Stroke and Angled Insert Face Tests ............................. 59 Test Series #4: 1.5 Inch Stroke and Mud on Brake Chamber Face Tests ............ 60 Test Series #5: 1 Inch Stroke Tests ................................ ................................ ...... 60 Test Series #6: 1 Inch Stroke and Wet Sensor Tests ................................ ............ 61 Test Series #7: 1 Inch Stroke and Mud on Sensor Face Tests .............................. 63 Analysis of 1.5 Inch Stroke Tests ................................ ................................ ......... 64

PAGE 8

viii Analysis of 1 Inch Stroke Tests ................................ ................................ ............ 65 Observations of Testing ................................ ................................ ........................ 67 VIII. A LTERNATIVE DESIGN ................................ ................................ ..................... 70 IX. C ONCLUSIONS ................................ ................................ ................................ .... 73 REFERENCES ................................ ................................ ................................ ................ 75 APPENDIX A. Commercial Vehicle Factor/Skid to Stop Analysis Example Problem ............ 79 B. Weight Distribution Analysis Example Problem ................................ ............. 80 C. Heusser Method Example Problem ................................ ................................ .. 82 D. Bartlett/Heusser Method Example Problem ................................ .................... 85 E. Data Collection Summary ................................ ................................ ................ 94

PAGE 9

ix LIST OF TABLES TABLE 7.1 Test S eries #1 data ................................ ................................ ................................ . 57 7.2 Test S eries #1 data analysis ................................ ................................ ................... 57 7.3 Test S eries #1 data analysis excluding tests 1 and 2 ................................ .............. 57 7.4 Test S eries #2 data ................................ ................................ ................................ . 58 7.5 Test S eries #2 data analysis ................................ ................................ ................... 58 7.6 Test S eries #3 data ................................ ................................ ................................ . 59 7.7 Test S eries #2 data analysis ................................ ................................ ................... 59 7.8 Test S eries #4 data ................................ ................................ ................................ . 60 7.9 Test S eries #4 data analysis ................................ ................................ ................... 60 7.10 Test S eries #5 data ................................ ................................ ............................... 61 7.11 Test S eries #5 data analysis ................................ ................................ ................. 61 7.12 Test S eries #6 data ................................ ................................ ............................... 62 7.13 Test S eries #6 data analysis ................................ ................................ ................. 62 7.14 Test Se ries #7 data ................................ ................................ ............................... 63 7.15 Test S eries #7 data analysis ................................ ................................ ................. 63 7.16 1.5 inch stroke data analysis ................................ ................................ ................ 64 7.1 7 2 Test for Goodness of Fit ................................ ................... 65 7.1 8 1 inch stroke data analysis ................................ ................................ ................... 66 7.1 9 2 Test for Goodness of Fit ................................ ...................... 66

PAGE 10

x LIST OF FIGURES FIGURE 2.1 Diagram of Lane Air Brake System ................................ ................................ ........ 5 3.1 Pneumatic S cam brake ................................ ................................ ........................... 8 3. 2 Brake system circuit diagram ................................ ................................ ................. 11 3. 3 Pneumatic brake chamber ................................ ................................ ...................... 1 3 3. 4 Pneumatic brake chamber with spring parking brake ................................ ............ 1 4 3. 5 Stroke sensing slack adjustor diagram ................................ ................................ ... 1 5 3. 6 Clearance sensing automatic slack adjustor ................................ ........................... 1 8 3. 7 S cam brakes not applied, S cam brakes applied ................................ ................... 21 4.1 Type 30 brake chamber force curve ................................ ................................ ....... 24 4. 2 Cracked brake drum ................................ ................................ ............................... 27 4. 3 Martensite spotted brake drum ................................ ................................ ............... 28 4. 4 Failed relay valve due to debris ................................ ................................ ............. 29 4. 5 Fluid impregnated brake shoe and drum ................................ ................................ 30 6.1 Vericom VC4000 DAQ data acquisition system ................................ ................... 38 ................................ ................................ ...... 39 6.3 Sensor testing apparatus ................................ ................................ ......................... 40 6.4 Testing setup for Test S eries #1 ................................ ................................ ............. 42 6.5 Test S eries #1 data ................................ ................................ ................................ . 43 6.6 Testing fixture setup for T est S eries #2 ................................ ................................ . 44 6.7 Test S eries #2 data with wet brake face ................................ ................................ . 45 6.8 Angled brake chamber mounting bracket ................................ .............................. 46

PAGE 11

xi 6.9 Test S eries #3 data with angled support ................................ ................................ 47 6.10 Mud applied to brake chamber face ................................ ................................ ..... 48 6.11 Test S eries # 4 data with mud applied to brake face ................................ ............. 4 8 6.12 Testing fixture with st ro k e decreased to 1 inch ................................ ................... 49 6.13 Test S eries #5 data with 1 inch stroke ................................ ................................ . 50 6.14 Test S eries #6 with wet sensor ................................ ................................ ............. 51 6.15 Senso r face with thin layer of mud applied ................................ ......................... 5 2 6.16 Test S eries #7 data with mud applied to sensor face ................................ ........... 53 7.1 Test Series #5 Data ................................ ................................ ................................ 55 7.2 Mean of pushrod retracted and extended ................................ ............................... 55 7.3 Gradual increase in sensor distance from brake face ................................ ............. 63 7.4 Increase in sensor distance due to tests ................................ ................................ .. 66 7. 5 Digital caliper placement ................................ ................................ ....................... 69

PAGE 12

1 CHAPTER I BACKGROUND Large multi axle commercial vehicles, or heavy vehicles, travel on the same highway system as passenger vehicles, necessitating braki ng systems up to the task of rapidly slowing large and heavy vehicles at acceptable rates when compared to much smaller and lighter passenger vehicles. Heavy vehicles, including transit vehicles, account for approximately 10% of the vehicle miles traveled (VMT) on roadways in the United States in 2011 [1] . A fully laden tractor semi trailer can weigh 80,000 pounds or more, whereas, a typical passenger car weighs approximately 4,000 pounds with pickups and SUVs weighing up to 8,000 pounds [2] . Generally, heavy vehicles travel at, or about the speed of passenger vehicles, but possess a much greater amount of kinetic energy which must be dissipated to slow or stop due to their large masses. Generally , brakes function by converting kinetic energy into thermal energy, which is then dissipated into the atmosphere. Maintaining a commercial vehicle braking system in safe working order is of the utmost importance due to the massive amount of kinetic energy which must be dissipated for heavy vehicle to slow and brake, especially when operating alongside passenger vehicles. 1.1 The kinetic energy of a 4000 pound passenger car traveli ng at a highway speed of 75 mph is approximately 7.5 x 10 5 foot pounds, which must be dissipated through the braking system for the passenger vehicle to stop. A fully loaded heavy vehicle (80,000 pounds)

PAGE 13

2 traveling at the same speed must dissipate approximat ely 150 x 10 5 foot pounds, or 20 times system differs vastly from that of a p assenger vehicle. The massive energy of fully laden heavy vehicle at speed, if not properly dissipated, can cause massive damage to any object the heavy vehicle strikes during a collision. The large mass difference between a heavy vehicle and a passenger vehicle will , when they collide generally result in a relatively minor collision for the occupant(s) of the heavy vehicle, but a potentially severe or catastrophic collision for the occupant(s) of the passenger vehicle. Realizing that even a minor collisio n for a heavy vehicle has the potential for producing serious injuries or death to the occupant(s) of a passenger vehicle, it is clear that contribute to a collision tha t would otherwise be avoided. Collisions and single vehicle crashes involving heavy vehicles have been studied extensively. Starting in 2001, the National Highway Traffic Safety Administration (NHTSA) and the Federal Motor Carrier Safety Administration (FM CSA) estimated during its Large Truck Crash Causation Study (LTCCS) that deficient braking systems played a part in 26% of all heavy vehicle crashes [3] . A further study completed by the Michigan State Police Motor Carrier Enfor cement Division called the Fatal Accident Complaint Team (FACT), found that 32.7% of all heavy vehicles involved in the study had braking system deficiencies [4] . The FACT study encompassed inspections of 407 heavy vehicles fol lowing crashes. This paper examines potential tools which could alert drivers when a brake deficiency arises, and hopefully decrease the instance of braking system related failures on heavy

PAGE 14

3 vehicles. The feasibility of employing an ultrasonic distance measuring sensor with real time alert an operator when a brake reaches its adjustment limit, is examined. Brake adjustment information relayed to the heavy vehicle drive r in real time, would potentially reduce the frequency of heavy vehicles operating on public roadways with deficient brakes.

PAGE 15

4 CHAPTER II HISTORY OF HEAVY VEHICLE PNEUMATIC BRAKING SYSTEM Modern heavy vehicles (defined as a vehicle or vehic le combination weighing 10,001 pounds or more) are generally equipped with pneumatic braking systems. Pneumatic braking systems were originally developed for use in locomotives prior to their adaptation to heavy vehicles [5] . G eorge Lane, was the first to develop and deploy pneumatic brakes for on road heavy vehicles. Lane worked as a logging truck driver in the northwest United States and saw the need for better, more reliable braking systems on the logging trucks he operated. In 1919, ssion stroke to pass through a one way check valve and into a holding reservoir where the compressed gas rear axle of a heavy vehicle and was applied using a hand val ve [5] . Figure 2.1, depicts the

PAGE 16

5 Figure 2.1 : Diagram of Lane Air Brake System [5] on heavy vehicles, as follows: the compression stroke. Introduction of contaminants from the engine into the braking system. By 1924 Westinghouse had developed a pneumatic airbrake system which did not rely upon an accumulator valve, but instead relied on an engine driven compressor to provide compressed air for the system. The engine driven compressor heralded the coming of the modern pneumatic braking system. Although refined in the following century, the engine driven compressor remains basically the same and completes the same fun ction. Around the advent of the engine driven compressor, foot operated brake valves (treadle valves) were

PAGE 17

6 developed, along with pressure regulators to ensure that the braking system operated within normalized pressure s . Shortly after the development s by W estinghouse , the United States Government began to regulate braking systems on heavy vehicles. The first government mandated regulations for a required stopping distance for heavy vehicles were issued in 1933 . These regulations requi red a heavy vehicle to stop from 20 mph within 50 feet. The major legislative effort to ensure the safety of heavy vehicle braking systems was the Federal Motor Vehicle Safety Standard (FMVSS) 121 , which came into effect in 1975. FMVSS 121 required newly manufactured commercial vehicles to be equipped with many of the safety features found on modern vehicles , including anti lock brakes. In 1978, due to legal action by the trucking industry and vehicle manu facturers, the requirement for an anti lock brak ing system was eliminated from the law along with the relaxation of the stopping distance requirement. It was not until 1997 that anti lock brakes were again required by FMVSS 121 [5] . During t he century following the invent ion of t he heavy vehicle pneumatic braking system , pneumatic brakes have been refined, with greater efficien cy and reliab ility . Even with modern refinements, many challenges remain in maintaining braking efficie ncy on modern heavy vehicles. The following chapters will outline the function of the major components of a pneumatic braking system and the areas where common failures occur.

PAGE 18

7 CHAPTER III HEAVY VEHICLE PNEUMATIC BRAKING SYSTEM CONCEPTS Pneumatic braking systems have several commonalities with the hydraulic braking system employed in passenger vehicles . However , instead of using a n (ideally) incompressible fluid in a hydraulic system , a pneumatic system uses a comp ressible fluid (air) . Pneumatic braking systems in general are more complex when compared to hydraulic braking systems. The majority of pneumatic braking systems on heavy vehicles in the United States employ a type of brake system called S cam drum br akes. Air disc brakes for heavy vehicles are prevalent in Europe , and while they are available and have been making inroads into the United States market in recent years , they comprise a small percentage of the braking system s found on heavy vehicles in the Uni ted States [6] . Due to dominance of large fleets warry of costs, S cam drum brakes will likely maintain their prevalence in the heavy vehicle market in the United States for many years . P neumatic braking system s use compressed air to activate a series of mechanical linkages , which in turn press friction material (brake shoe/pad) into a heat sink (brake drum/rotor) . Brakes, whether a p assenger vehicle equipped with a hydraulic braking system or a heavy vehicl e equippe d with pneumatic brakes , complete the same function convert ing kinetic energy into thermal energy to slow the vehicle. The thermal energy is then dissipated heat sinks can accept more energy .

PAGE 19

8 Figu re 3.1 : Pneumatic S cam brake [7] An engine driven mechanical compressor compress air from the atmosphere to approximately 120psi. The compressed air is dried to remove contaminates , and stored in a series of air reservoirs . The air reservoirs ensure that sufficient volume of compressed air is available to operate the brakes and other pneumatic systems during normal or emergency operations. Additionally, the air reservoirs are designed with redundancies to ensure that if a f ailure occurs within the pneumatic system, the vehicle can still brake to a stop. Modern Pneumatic braking systems have two braking system ; service brakes and parking brakes . Service brakes are simply the braking system designed for use while the vehicle is in service. Service brakes are applied by pressing on the brake treadle valve with , or by operating a hand valve , if equipped. The p arking , or spring br akes , are used to ensure that a vehicle does not move while parked, and to ensure that a vehicle has sufficient air pressure to operate the service brakes before the spring brakes release to allow

PAGE 20

9 the vehicle to move . S pring brakes are operated by evacuati ng air from the spring brake chamber, which is usually done through a push/pull button on the dash position in the cab . Generally, w hen the operator of a pneumatically braked vehicle applies the service brakes , a valve is opened allowing co mpressed air to flow into the braking circuit. Compressed air released from the h olding tanks flows through a series of check valves and relay valves into the brake chamber s at each axle . The brake chambers are energized by compressed air , and through mech anical linkage they press the friction material against the surface of the heat sink. A more in depth description of the most common braking components utilized by heavy vehicles in North America follows : Compressor and Dryer The m odern compressor fitted on heavy vehicles is much like the engine of these vehicles. The compressor consists of a one to four cylinder piston compressor driven by the reci procating pistons acting as an air pump. Unlike a conventional engine, fuel is not added to the compress ed air, and the there is no ignition source. A fter being compressed , the air is pumped through a one way check valve that allows air to flow into holdin g reservoirs , but not back towards the compressor . Most modern pneumatic braking systems employ an air dryer consisting of a desiccant system that absorbs wa ter and contaminants from the air compressor . When the desiccant material in the air dryer become s saturated , compressed air is pushed through the

PAGE 21

10 top of the air dryer flushing the moisture out of the desiccant material through a drain valve. Ensuring that only dry air fills the air tanks is important as wat er and contaminants can build up in the air t anks and decrease the avail able storage volume for compressed air. Water can also rust components of the braking system, and contaminants can clog air passages . Braking System Plumbing The pneumatic braking system on modern heavy vehicles consist s of two braking circuits ; the primary and the secondary . The primary circuit supplies air to the brakes on the drive axle ( s ) of the vehicle and to the trailer brakes, if attached . The secondary circuit provides air to the brakes on the front axle , and can also supply air to the brake s of an attached trailer. If either one of the air circuits fail, the remaining functional circuit supplies air to the brakes of an attached trailer. The two circuits in a modern pneumatically braked heavy vehicle are separated , so that if one of the circuits has a catastrophic failure, the other circuit can still provide enough air to the brake s to safely, although not as efficiently, stop the vehicle or vehicle combination . Each circuit contains its own dedicated air reservoir (s) used to store compressed air exclusively for that circuit . Modern semi trailers have two separate air circuits as well , known as the emergency circuit and the service circuit. The emergency circuit provides air to the air holding reservoir s , suspension , and springs brakes. The service circuit provides air to the service brake relay valves , which when activated , allow air from the holding reservoir s filled by the emergency circuit, to energize the service brakes during brake application.

PAGE 22

11 secondary circuit by means of a double check valve that allows the circuit with the higher air pressure to feed the emergency circuit . The air in the service circuit is fed much the s ame way . W hen a bra king application is requested using either the treadle valve (foot brake) or a hand valve (Johnson bar) , air flows into the service circuit from whichever vehicle ci rcuit with the highest pressure . This system and the double check valve en sures that if there is a catastrophic failure of one of the air circuit s , the trailer will remain unaffected . Figure 3.2 heavy vehicle tractor. Figure 3.2 : Brak e sys tem circuit diagram [8]

PAGE 23

12 Brake Chamber s S Cam b rake chamber s convert pneumatic pressure to a linear force. A brake chamber is comprised of a steel chamber with a rubber diaphragm sandwiched between the two halves of the chamber. One half of the chamber is equipped with an air inlet where compressed air enters when the service brakes are applied. The compressed air then flexes the rubber diaphragm outward. T he diaphragm is attached to a pushrod , which extends out of the brake cham ber and moves away from the brake chamber a s the diaphragm flexes due to applied air pressure during brake application. On the other side of the brake chamber is a return spring, which acts on the pushrod base to return the pushrod and diaphragm to their original position when the service brake is released , evacuat ing air from the brake chamber. The rated size of brake chambers is measured to the normalized square inch area of the brake chamber diaphragm. Therefore, a type 3 0 brake chamber has a nominal 30 square inch diaphragm, a type 20 has a nominal 20 square inch diagram and so on. The larger the chamber size , the more force can be generated by the brake chamber at the same given air pressure, and the more distance the pushrod and diaphragm can move d uring brake application . F igure 3.3 depicts the components of a service brake chamber.

PAGE 24

13 Figure 3. 3 : Pneumatic brake chamber [9] Another variation in brake chamber design is the long stroke chamber. The long stroke brake chamber has a deeper body, allowing the rubber diaphragm to move further before the diaphragm makes contact with inside of the chamber , thus providing a long er stro ke of the pushrod . An example of this is a type 30 long stroke brake chamber , which can displace the pushrod 2.5 inches prior to reaching its adjustment limit . A regular stroke type 30 brake chamber can only displace the pushrod 2 inch es prior to reaching the adjustment limit [10] . A longer maximum stroke has maintenance advantages , and it has been argued that it may provide braking system efficiency advantages. Additionally, some pneumatic brake chambers are also equipped with a supplemental parking, or spring brake. The spring brake operates the opposite of the service brake chamber, in that in place of air pressure being utilized to displace the pushrod, a large spring is used instead in the absence of air pressure in the cham ber . A spring brake chamber uses air pressure on the opposite side of the rubber diaphragm than a service brake chamber . When air is su pplied to a spring brake chamber, the air pressure pushes the rubber diaphragm and compresses the spring, allowing the br ake pushrod to retract and release contact between the

PAGE 25

14 friction material (shoes) and the heat sink (drum) . The spring brake is configured in such a way that when the braking system does n o t have sufficient air pressure , the brakes at the corresponding whee l s remain applied. This is both a safety feature and a means to ensure a vehicle or trailer does not move when parked. In order to release the spring brakes , a heavy vehicle typically must have approximately 55 psi of air pressure built up in the braking system [11] . If a catastrophic failure of both braking circuits occur s , and the pressure within the braking system falls below typical ly 20 to 40 psi , the spring brake will automatically apply [12] , and apply braking to the corresponding wheel s . Figure 3.4 illustrates the inside of an S Cam brake chamber with an attached spring brake chamber . Figure 3. 4 : P neumatic brake chamber with s pring parking brake [13] Upon brake application and air pressuriz ing the service brake chamber , the rubber diaphragm pushe s outward, extending the pushrod . The opposite end of the pushrod is attached to the slack adjustor through a pin connection . As the pushrod extend s , the pushrod rotates the slack adjuster which applies torque to rotate the S cam and push the brake shoes into contact with the brake drum, thus applying the brakes.

PAGE 26

15 Slack Adjustor Slack adjustors are a mechanical linkage that conv ert translational force into rotational force (torque) and to provide leverage through the geometry of the system to multiply the brake force generated by the brake chamber . The slack adjustor is essentially a lever attached at one end to the pushrod throu gh a pin connection, and at the other end through a splined connection to the S cam torsion bar function is to convert the linear motion of the pushrod into the torque necessary to activate the brakes. When the slack adjust or rotates due to the linear pushrod force, the splined connection to the S cam is torqued and rotate s . As the S cam rotate s, the brake shoes ramp up the sloped face of the S cam , are pushed outward and contact the brake drum. The friction which is then dissipated through the surface area of the brake drum. The slack adjustor also functions to limit the travel needed by the pushrod to activate the brakes. Slack adjustors maintain the clearance distance between the brake shoes and the inner surface of the respective brake drums , ensuring that the stroke of the pushrod is sufficient to fully activate the brakes prior to the rubber diaphragm in the brake chamber reaching its limits . The stroke of the pushrod must be sufficient to fully activate the brakes even when the brake drum has expanded due to heat buildup , such as resulting from braking on a n extended downgrade . There are two types of slack adjustors ; manual slack adjustors and automatic slack adjustors. Manual s lack adjustors require the operator of the vehicle to manually adjust (by turning a hex head at tached to a gearing system) the slack adjustor to maintain proper brake stroke and clearance between the brake shoe and inner surface of the brake drum . Manual

PAGE 27

16 slack adjustors were prevalent on vehicle prior to f ederal regulations requir ing all pneumatical ly braked on road heavy commercial vehicles to be equipped with automatic slack adjustors for vehicles manufactured after October 20, 1994 [14] . Automatic slack adjustor s were [5] in an effort to automatically attempt to maintain the proper brake stroke . Automatic slack adjustors eliminate the need for constant adjustment by the driver or maintenance person ne l. There are two different design methodo logies utiliz ed in automatic slack adjustors ; t he stroke sensing slack adjuster and the clearance sensing slack adjuster . Automatic slack adjustors employing the stroke sensing method use a rod which is actuated during braking application to measure the brake stroke . A s soon as the pushrod stroke measured by the actuated rod reach es a specified limit, a ratcheting mechanism engage s, causing the slack adjustor w o rm gear to rotate, thus rotating the main slack gear . Because the slack gear is att ached via a spline to the S cam, the pushrod stroke is decreased by increasing the initial rotation of the S cam , much the same way brake adjusto rs work in a passenger vehicle equipped with drum brakes. The following diagram is an example of a Meritor stroke sensing automatic slack adjuster.

PAGE 28

17 Figure 3 . 5 : S troke sensing automatic slack adjustor [15] The other method employed in automatic slack adjusters is the clearance sensing method. The clearance sensing method uses a clutch that engages when the brake shoe lining contact s the brake drum. When the clutch engages , the adjustment stops. If there is enough distance between when the brakes are applied or released , and when the brake linin gs engage the drum , a one way clutch engages the next tooth on the clearance notch . This advances the initial rotation of the S cam, thereby decreasing the pushrod stroke . The following diagram is an example of a Haldex clearance sensing automatic slack adjuster, which determines if adjustm ent is needed during the application stroke and then adjusts as the brakes are being released. Some designs determine if adjustment is need ed upon the release of the brake, instead of on application.

PAGE 29

18 Figure 3 . 6 : Clearance sensing automatic slack adjuster [16] S Cam The S cam is a camshaft , or torsion bar, S at one end that contacts the brake shoes , and a splined end on the other that rides in side the slack adjuster. As the slack adjuster rotates due to br ake application, the S cam is torque d S the head forces the brakes shoes to roll up the S cam face and outward , pressing the shoes o nto the brake drum inner surface . Different S cam designs have different geometry of the head, allowing for faster or slower ramping of the brakes, and thus faster or slower brake application. Because the S cam head is fixed in its geometry, all adjustment for braking wear and thermal expansion must be performed by the slack adjuster . If the brake shoes and/or drums are worn beyond in service specifications , or the brake drum has cracked allowing for expansion, the S cam can roll over. S cam roll over occurs when the S S

PAGE 30

19 head , and the brake shoes fall off the l eading face of the S cam. Once an S cam rolls over , th e brake can no lon ger generate any braking force, since the shoe cannot be pushed out to make contact with the brake drum. Figure 3.7, depict s the S application. Figure 3. 7 : S cam brakes not applied, S cam brakes applied Brake Shoes and Drums Brake shoes consist of a semi circular metal backing plate with a fiction lining riveted onto its outer surface . Brake shoe are anchored as a revolute joint at one end by pins , and t he other end consists of a roller which rides on the head of the S cam. The friction lining of the brake shoe is pressed into the brake drum upon brake application , creating friction between the lining and the drum inner surface, converting kinet ic energy into thermal energy. The brake shoes ride within a brake drum, which is a large cast iron covering . The brake drum is the heat sink of the braking systems. The friction generated upon brak e application when the braking lining press es against the brake dru m is the method for generating brake force. Even though brake drums are a large single piece of cast iron, they too can fail. Brake drums can fail by cracking, allowing the drum to expand upon brak e application and limiting the friction which can be genera ted. Brake drums can also wear out,

PAGE 31

20 or become oblong due to wear and excessive overheating. Brake drums are the single heavi est part of the braking system. Valves and Modulators A pneumatic braking system contains several valves that act as relays to speed up the delivery of air pressure to the desired destination. One way check valves keep the air pressure in the desired location , and separate different braking circuits. Quick re lease valves are located near brake chambers and act to release the air pressure build up in a brake chamber during brake application , without the air having to travel all the way to a central location . These quick release valves prevent brakes from remain ing applied after release by the driver . Modulators are necessary on vehicles equipped with anti lock braking systems (ABS). The modulators regulate the air pressure delivered to individual, or groups of brake chambers during full ABS brak e application to prevent wheel from remaining fully lock ed, and the vehicle loosing directional stability. Anti Lock Brakes Since March of 1997 , newly manufactured over t he road heavy vehicles we re mandated to be equipped with an anti lock braking system, or ABS system. In March of 1998 , all newly built semi trailer s we re required to be equipped with an ABS system [17] . Pneumatic ABS, in basic terms, uses a seri e s of valves, modulators and sensors to keep the ceasing rotation, or fully locking and remaining locked during braking application . This is done to generat e the maximum friction between the tire road interface, and to maintain vehicle stability during full braking ABS app lication . Full wheel lockup is prevented by modulating the air pressure at each brake chamber to keep the braking force

PAGE 32

21 below that required to stop wheel rotation . The cyclic rate of a pneumatic ABS braking system (rate at which pressure is modulated to pr event wheel lockup) is much less than that of a modern hydraulically braked passenger vehicle equipped with ABS. The low cyclic rate of pneumatic ABS is due to the large rotating inertia of the heavy system, as opposed to the much smaller rotating inertia wheel system . T he large rotational ine tire takes significantly longer to slow, and regain its rotational velocity than for a passenger ve hicle . The ABS system is generally controlled by a central Electronic Control Module (ABS ECU), which processes the data from various sensors and controls the ABS modulators in response. In a combination vehicle, e.g. truck and semi trailer, both tractor a nd trailer(s) have independent ABS ECUs which do not communicate with each other . I nformation regarding wheel speed is sent to the ABS ECU from H all E ffect sensors mounted at the wheel. These sensors detect a si ne wave generated by a tone ring which rotate s on the wheel hub. The sensors are a simple magnetic pickup that reports the rate of the sin e wave generate d by the rises and depressions of the tone ring. The ABS ECU calculates the wheel speed based upon the frequency of the sin e wave, and when a wheel s speed begins to deviate from the other monitored wheels, the ABS ECU commands the modulator at that wheel to reduce air pressure to the subject brake chamber . The corresponding decrease in brake force at that wheel allows that wheel to spin back up to th e speed of the other wheels. An ABS system does not function during normal braking , but only begins to function when it senses a wheel or wheels approach ing lockup.

PAGE 33

22 CHAPTER IV BRAKING SYSTEM DEFICIENCIES Braking system deficiencies on heavy vehicles are found during inspections on a far too regular basis . I n many parts of North America , it is not uncommon to find a heavy vehicle combination while descending a ste e p downgrade or stopped near the base of the downgrade , with its brakes smoking from excessive heat , or even on fire. The factors that may have preceded such a visible and dangerous event are usually not readily evident without a technical inspection of the braking system. Although this smok y sidesh ow is often traced to driver error when descending a grade , it is often caused or compounded by deficiencies in one or more the brakes on the heavy vehicle . Ineffective pre trip inspections and a lack of proper maintenance lead to many braking deficiencies being overlooked prior to a potentially catastrophic event. In general, pneumatic braking system deficiencies occur when the interface between brake shoes and the drum cannot generate enough force to effectively slow the vehicle . One brake with a braking deficiency will increase s braking distance , and stress the remaining braking system s at the other wheels . The opposite can also occur, where a brake is applied, locks up and then will not release. This can cause a fire at that brake position, or instability of the vehicle. A locked brake can also overheat a tire and cause the tire to explode, which can result in potentially dangerous event s for n earby pe destrians or vehicle traffic . The inability of a braking system to generate adequate force at the brake shoe/drum interface can be due to many different issues within the braking system . The following

PAGE 34

23 section outlines some of the more common and potentially catastrophic braking system deficiencies, but is by no means intended as a comprehensive guide. Excessive Pushrod Stroke The most common ly cited brake system deficiency found during roadside inspection of heavy vehicles is excessive pushrod tra vel during brak e application , otherwise known as out of adjustment brakes [7] . To understand why excessive pushrod travel is such a safety issue on the roadway, one must first understand what pushrod travel refers to , and how it interacts with the maximum force generated at a brake d wheel . Pushrod travel is simply the change in distance between when the pushrod is fully retracted (no braking) and when the pushrod is fully exte nded (full braking air pressure applied ). The difference in the travel of the pushrod as air pressure of 90 to 100 psi fills the brake chamber is commonly referred to as the pushrod stroke . Measuring pushrod stroke is accomplished by selecting an arbitra ry point on the pushrod (usually the clevis pin connection to the slack adjuster ) and measuring the distance to the brake chamber face without brake appli cation (D 0 ) . Following brake application with 90 to 100 psi of pressure , the distance from the brake chamber face to the same arbitrary point on the pushrod is again measured (D 1 ) . The difference between these two meas urements is the pushrod stroke. 4.1 Excessive pushrod stroke decrea se s the force that a brake chamber can apply to activate the brakes. The reason that e xcessive pushrod stroke is detrimental to braking force generation , is that the brake chamber diaphragm can only flex so much before it starts binding on the interior of the brake chamber , during which time the force generated starts to

PAGE 35

24 decrease rapidly. At the extreme end of the excessive pushrod travel, the pushrod strokes out , meaning the diaphragm can no longer move the pushrod any further to apply torque to the S cam . When the diaphragm bottoms out, no additional force can be generated at the braking chamber, no matter what air pressure is applied. Testing published by different brake chamber manufacturers and independent researchers plot the pushrod force versus pushrod stroke at various applied air pressures. Figure 4.1 is a force curve graph is taken from a brake chamber manufacture s testing and shows the buildup of the force measured at the pushrod as the stroke increase s until it hi ts a yield point where the diaphragm begins to contact the brake chamber. Once the diagram begins to contact the brake chamber the force knees over and drops off. Figure 4.1: Type 30 brake chamber force curve [18] If the brake shoe has not moved sufficiently to engage the brake drum as a brake strokes out, the brake will cease to develop any braking force . When one brake fails to

PAGE 36

25 develop force , the amount of work required of the other brakes to slow the vehicle will increase . Increased work by fully functioning brakes, even in non emergency slowing situations, can lead to excessive heat build up , which in turn can produce additional brake failures . Additionally, one non functioning brake on a vehicle can lead to an imbalance , and possibly linear stability while braking . In order to combat excessive pushrod stroke , the Federal Motor Carrier Safety Regulations (FMCSR) were modified to mandate automatic slack adjustors . Although automatic slac k adjustors were mandated in 1994 and have greatly decreased the incidence of excessive pushrod travel, they are not a panacea. Even with automatic slack adjustors, excessive pushrod stroke remains a common deficiency found during roadside or post collision inspections . On September 7, 2017, the Commercial Vehicle Safety Alliance conducted its annual Brake Safety Day , in which 7,698 commercial motor vehicles were inspected . As a result, 14 % of [19] due to the of indicat es that at minimum, 20% of the of adjustment , or the braking system had other significant safety issue (s) . T here are many reasons why e xcessive pushrod stroke can still occur on vehicles equipped with automatic slack adjustors . The following is a list of the main reasons for excessive pushrod stroke: F ailing automatic slack adjustor which no longer is properly a djusting . W orn bushings, brake shoes or other components.

PAGE 37

26 I mproper brake geometry set during replacement of braking components can also result in automatic slack adjustors which do not properly adjust for brake wear. Thermal Expansion and Failures One of the defining features of drums brakes, pneumatic or hydraulic , is that as the brake drum s transfer kinetic energy into thermal energy, the drum i tself heats and expands. T hermal expansion of the drum during brake application increases the travel distance between the brake shoe friction material to where it can make contact with the interior surface of the brake drum. In a pneumatic braking system, this thermal expansion results in an increase in the necessary distance the pushrod trave l s to generate the same force on a heated drum versus a cool drum. This increase in stroke length during thermal expansion of the brake drum becomes critical for a brake already having a stroke length near its adjustment limit , up to the point where the brake may no longer generat e any braking force during application . In order to account for thermal expansion of the drum, the Federal Motor Carrier Safety Regulations (FMCSR) maximum legal brake pushrod adjustment limits have be en established fairly conservative ly , a nd long before the brake chamber diaphragm bottoms out . National Highway Traffic Safety Administration (NHTSA) engineers at the Vehicle Research and Test Center (VRTC) developed an equation to relate the increase in b rake stroke distance as a function of time. The following equation is based upon tests that were conducted at 60 mph and account for thermal expansion and stretch. [20] 4.1 Where I=increase of pushrod stroke in inches, T=time in seconds

PAGE 38

27 Large repeated thermal cycles on brake drums can lead to many issues detrimental to proper brake function. Brake drums can crack if overheated repeatedly or if severely worn. When a brake drum cracks , the brake drum can no longer resist s the brake sho e s outward force during brake application, a nd the brake drum will simply expand in response. Extreme temperatures in brake drums can also lead to a thermal change of the brake drum material composition . The structure of the brake drum can change to Martensite , which does not generate the same frictional resistance to the brake sho e . Brake shoe glazing is another issue with large thermal load , and occurs when the brake shoe friction material heats enough to boil the epoxy resi n in the friction material . The epoxy resin then forms a slick glaze on the outside of the friction material and decrease s the coefficient of friction between the brake sho e and drum surface . Figures 4.2 and 4.3 depict a cracked brake drum, and a brake drum which has experienced heat to the extent that Martensite has form ed . Figure 4.2: Cracked brake drum

PAGE 39

28 Figure 4.3: Martensite spotted brake drum Air Pressure and Delivery Significant a ir pressure is required to activate pneumatic brakes. A failure in air pressure buildup and/ or delivery can lead to catastrophic brake system failure. Over time, a compressor can wear out, and become incapable of maintaining the air pressure demands of a vehicle under continuous or repeated braking. If the brake pressure drops sufficiently low, the spring brakes will automatically apply to slow the vehicle , although they do not generate the same braking for ce as a full , service brake application. A common braking system deficiency occurs when air leaks within the braking are present . An air leak can be as simple as a hole or crack in an air line, or a sticky valve. Another common source of air leaks originates at the brake chamber . If the brake chamber diaphragm develops a hole, air will leak when the brakes are applied. A hole in the brake chamber diaphragm can be difficult to detect, as it will only leak during brake

PAGE 40

29 application. A hole in the brake chamber diaphragm has the addi tional effect of decreasing the force that is developed by the brake chamber with a given air pressure . A faulty relay valve can develop f r om debris entering the braking system, or through wear. A faulty relay valve can receive a demand to supply air, bu t fail to open leaving the brakes supplied by the relay valve inoperable. A failed relay valve will generally disable the brakes on one axle of a vehicle, but if it is the relay valve that supplies air to a trailer that fails , it can render the totality of The photographs of Figure 4.4 depict a relay valve which has failed , preventing the trailer from braking due to debris buildup within the valve . Figure 4.4: Failed r elay valve due to debris Fluid Contamination A failed, or failing gasket at an axle end can allow fluid from the differential and axle differential when leaked onto the braking system, impregnates the friction material of the brake shoes and lubricates the brake drum surface, resulting in the brake shoes lo sing most of their ability to generate friction when pressed against a lubricated brake drum surface . Grease from the axle hub can also leak into the braking system and cause the same decrease in the

PAGE 41

30 ability of a brake to generate friction. Figure 4.5 depicts a brake which has been fluid impregnated. Figure 4.5: F luid impregnated brake shoe and drum

PAGE 42

31 CHAPTER V BRAKING PERFORMANCE ANALYSIS A basic u nderstanding of braking performance analysis is presented to assist the reader in understanding the interplay between the factors previously discussed and braking efficiency . Several methods are available to analyze a pneumatic braking system, from the ver y simple to complex modeling. The methods presented are models that have a basis in both physical constraints and empirical modeling. The select ion of a brake analysis methodology depends upon the information available and the level of precision necessary to assess performance . The models presented here are generally accepted as authoritative, and commonly used to determine the rate of deceleration of a heavy vehicle , and the speed of the vehicle at the beginning of observable brake application . Commercial Vehicle Factor Method and Skid to Stop The Commercial Vehicle Factor (CV F) method uses an empirical slowing ac celeration factor to estimate the ac celeration of a vehicle under full locked wheel brak e application. The Commercial Vehicle Factor method requi res that the coefficient of tire roadway friction is measured or estimate d . Once the coefficient of tire roadway friction for a passenger vehicle on the roadway surface has been determined, a commercial vehicle factor (CVF) is applied to approximate an eff ective drag factor for the heavy vehicle braking on the roadway surface . The commercial vehicle factor is based upon empirical testing of heavy vehicles on surface s with known tire roadway friction. The c ommercial vehicle factor is often ranged anywhere between 65% to 85 % [21] , depend ing upon the condition and tread of the tires, as well as multiple other factors related to tire design . Multiplying the coefficient of friction determined for a passenger vehicle by the commercia l vehicle factor , results in the

PAGE 43

32 effective drag factor for a heavy vehicle. Using the adjusted drag factor , kinematic principles observable brak e application , which is commonly referred to as the skid to stop . The following set of equations are used to complete this analysis. An example of this type of analysis is contained within Appendix A to this report. Work Energy Principal : 5. 1 Where , F=force, D=distance, m=mass, v= velocity at beginning of skid nd law : 5. 2 Where , a=acceleration Ac celeration rate of a heavy vehicle during braking : 5. 3 Where , coefficient of passenger vehicle tire roadway friction , CV F=commercial vehicle factor, g= gravitational constant ( ft/sec 2 ) Ac celeration of a heavy vehicle substituted into E quation 5.2 : 5.4 Second Law statement for f orce substituted in to E quation 5.1 5. 5

PAGE 44

33 Equation 5.5 s olved for velocity : 5. 6 Where, D= length of skid ( ft ) Weight Distribution Method The weight distribution method is the simplest analysis which can account for non functioning brakes on a multi axle vehicle . As with the c ommercial v ehicle f actor method, this method cannot account for brakes with partial function ality . This method expands up on the c ommercial v ehicle f actor method with added considerations . The weight distribution method requires knowing or estimating the weight at each axle end. This is accomplished by either measuring the weight at each axle, or by using general models of w eight distribution based upon loading and configuration. In this analysis , if a brake is non function al the CVF n is equal to 0% at that position, which results no braking force at that brake . The equations to calculate the slowing ac celeration observable brake application using the weight distribution method follow . An example analysis conducted using the weight distribution method is contained in Appendix B to this report. Braking force at each b rake position /axle end : 5. 7 Where , F n =braking force at brake n, CVF n =commercial vehicle factor at brake n, w n =weight at n axle end ( lbs ) Effective braking ac celeration rate of heavy vehicle : 5. 8 Where , W=total weight ( lbs )

PAGE 45

34 Velocity at start of full brake application : 5. 9 Where, D=distance of skid ( ft ) H e usser Method [20] In 1991 , Ronald He usser published the first practical pneumatic braking analysis method considering the air pressure at the brake chamber , and the measured brake stroke. H eusser obtained data from brake dynamometer test s performed by the National Highway Traffic Safety Administra tion, and developed his analysis method based upon applying a regression analysis to the data , as well as data obtained from manufacturers . The Heusser analysis uses a brake force design calculation that was modified to fit empirical data. The Heusser bra ke force equitation calculates the force applied by the brakes at the tire/road interface at each n axle end using the following equation . Braking force from each n brake : 5.10 Where, Pforce=force of pushrod (lbs), SL=slack adjustor length (in), DRad=brake drum radius (in), CamRad= S cam radius (in), TRad=loaded radius of tire (in) All variables except for the pushrod force are directly measured on a vehicle. T he ideal pushrod force can be calculated by multiplying the air pressure at the brake chamber by the square inches of the brake chamber diaphragm. A s an example, a type 20 brake chamber has a 20 square inch diaphragm and when subjected to 100 psi of air pr essure it will, ideally , produce 2,000 pounds of force (20 in 2 ·100psi = 2,000 lbs force) . Direct m easurement of

PAGE 46

35 pushrod force reveals losses in the system that cannot be accounted for in th e idealized equation presented . Pushrod force tables have been gen erated from testing done by brake manufacturers , have two variables ; air pressure and pushrod stroke . Once air pressure at the brake chamber and stroke are determined , pushrod force is entered into E quatio n 5.10, solving for brake force. When using this formula, it is important to check the calculated brake force against the maximum brake force for each braking position/axle end , which is calculated using E quation 5.7 , and then use the smaller of the value s . This step is necessary because a brake cannot generate more force than when it is fully locked. To calculate the speed at the beginning of observable braking , the lesser of the braking forces are summarized into E quation 5.8 and then entered into E quati on 5.9. An example using the Heusser analysis method is presented in Appendix C to this report. Bartlett/He usser Method [22] [23] In 2004, the Heusser m ethod was modified by Wade Bartlett to include the time from brake application to full air pressure at the brake chamber, or what is called air pressure rise time . On vehicles traveling at high speed , the air pressure rise time is almost negligible in a braking analysis, but at low speed rise time can be significant. One issue with including air pressure rise time is that it is hypothetical, meaning that generally in a braking analysis, the point when the vehicle starts braking is correlated to the production of a tire mark. During air pressur e rise time, the vehicle will not be leaving tire marks on the roadway surface . This method is important for calculating, hypothetically , how far a heavy vehicle travel s while

PAGE 47

36 coming to a stop from a low speed. Since this is not the focus of this paper, th e equations will not be presented , but can be found in the original paper referenced. In 2007, the Heusser method was again modified by Wade Bartlett. This time , Bartlett included anti lock braking systems into the analysis. In his paper , Bartlett reports that at a conference and through personal communications, Heusser recommended additional procedure s w hen the attempted brake force ( E quation 5.10) is higher than the force required to lo ck the wheel ( E quation 5.8) , and the vehicle is equipped with anti lo ck brakes . T he additional procedures include adjusting the pushrod force by obtaining pushrod force from a brake force table at 20 psi to 8psi below the pressure necessary to lock the wheel, depending on the vintage of the ABS system. Bartlett found that calculating a linear regression of the tabulated pushrod force data originally published by Heusser produced a close fit. This linear regression must be calculated for each brake chamber, but in general the equation s are as follows: 5.11 Where, F L =pushrod force, m L =the slope of the linear regression, P=the air pressure at the brake chamber, b L =the Y intercept of the linear regression Bartlett rewrites Heusser brake force equation to solve for the pushrod force to lock, by setting the attempted brake force to equal the available brake force at each wheel end . This equation follows. 5.12 Where, P L =pushrod force to lock wheel end (lbs ) , W=weight at wheel end (lbs) , f r =µ*CVF, CamRad= S cam radius (in) , TRad=loaded radius of tire (in) , SL=slack adjustor length (in) ,

PAGE 48

37 DRad=brake drum radius (in) , m L =the slope of the linear regression, b L =the Y intercept of the linear regression If the heavy vehicle is generating more attempted brake force than the maximum available brake force (force to lock the wheel) , then the pushrod force to lock is calculated. Using the pushrod force to lock, the brake application pressure to lock is determi ne d. The adjustment for ABS lock (8psi to 20psi) is subtracted from t he brake application pressure to lock, and using this resul t ant brake application pressure , the brake force is determined. This method has been shown to be an accurate and reliable means to analyze the braking capabilities of a pneumatically braked vehicle equipped with anti lock brakes. A complete work through example analysis using the Bartlett/Heusser method is presented in A ppendix D . The example in Appendix D covers both full activati on of the anti lock braking system on a tractor and trailer , and the situation where only two lightly loaded axles lock up, and a limiting brake application pressure calculated.

PAGE 49

38 CHAPTER VI DATA COLLECTION AND TESTING The data from the testing completed fo r this paper was collected using a distance measuring ultrasonic sensor and a data acquisition system. The ultrasonic sensor tested was the Vericom VC18 measures distance based upon the time of flight of the ultrasonic emission and converts it to an analog signal ranging between 0 to 5 volts. The a na log si gnal can then be interpreted by the Vericom VC4000DAQ data acquisition system (Figure 6.1), and record s sensor outputs at up to 100 hertz . The sensor itself is mo unted to a magnetic base which has a curved recessed bottom. The recess allows for a more secure fit when mounted to the pushrod of a pneumatic brake. Figure 6.1 : Vericom VC 4000DAQ data acquisition s ystem

PAGE 50

39 roke sensor was mounted on the pushrod with its sensor face towards the brake chamber. The sensor emits an ultrasonic pulse which reflect s off and back to the sensor. The time between when the pulse is emitted and received back at the sensor is used to determine the distance the sensor is from the brake chamber face. The sensor operates a t 100 H z and has a range of 1.5 inches to 9.5 inches , and is s hown in Figure 6.2 [24] . Figure 6.2 : u ltrasonic sensor The ultrasonic sensor provides a real time distance measurement to determine the pushrod stroke and the timing of braking application and rise time . The t estin g which follows was completed to determine if this ultrasonic sensor is appropriate for permanent deployment on heavy vehicle s , and if the sensors can provide accurate real time brake measurements alert ing the operator to any brakes which fall out o f adjustment . The sensor is tested and

PAGE 51

40 evaluated based upon its ability to provide accurate, repeatable measurements in conditions commonly found on heavy vehicles. Testing of the was complete d in a controlled laboratory environment , but w as constructed to recreate conditions which the sensor would encounter when deployed on a vehicle . A testing cage was constructed by welding angle steel together and mounting a type 16 brake chamber with a n adjustable pushrod and metal pushrod stop . The pu shrod stop prevent s the brake chamber diaphragm from bottoming out during testing , and to control the stroke length . The sensor was mounted to the brake pushrod by the provi ded magnetic mounting base. Air was provided to the brake chamber from an air compressor regulated to between 90psi and 100psi, the pressure mandated for brakes are out of adjustment . Air was suppl i ed from the compressor and to the brake chamber through a hand operated gate valve. Compressed air enteri ng the brake chamber displaces the diaphragm and pushrod outward until it contacts the metal stop. To retract the pushrod, the compressed air source was disconnected from the brake chamber , allowing air to escape from the brake chamber. Figure 6.3 shows th e testing apparatus. Figure 6.3 : Sensor t esting a pparatus

PAGE 52

41 A digital caliper was used to verify and set the distance between the pushrod and the metal stop. Due to the pushrod end in the testing apparatus not being connect ed to a slack adjustor, or other device to restrain the motion of the pushrod , the pushrod could angle This movement of the push rod added to the variability of the pushrod s troke measure ments . It is suggested that future studies account for this movement, by constraining the pushrod end so that it could only move outward from the brake chamber orthogonally . Seven series of tests were completed, with different parameters and conditions. These seven series of tests were designed to replicate some of the condition s potentially encountered when the sensor is mounted on an over the road vehicle. Test Series #1 The first series consisted of 13 individual tests. These tests were de signed to replicate the ideal conditions in which a permanently mounted brake stroke sensor would operate. The setup and test conditions for this series were as follow s and is shown in Figure 6.4 : Air pressure of between 90 psi to 100 psi Clean brake chamber face Pushrod stroke of approximately 1 ½ inches orthogonally (below FMCSR maximum stroke of 1 ¾ inches)

PAGE 53

42 Figure 6.4 : Testing s et up for T est S eries #1 During the first two tests of this series , sensor was located too close to the brake chamber at just under 1.5 inches, and the sensor did report any change in distance until the pushrod had already initiated mov ement . This can be seen in the chart in Figure 6.5 . On the se first two tests , there wa s no sensor noise when the pushrod was in the retracted position . To correct this, the sensor was moved away from the brake chamber, until it was initially approximately 2 inches from the face of the brake chamber. Figure 6.5 depicts 13 tests of the displac ement of the sensor mounted on the brake pushrod on the vertical axis measured in inches. The blips in the displacement graph when the pushrod is retracted w ere due to adjusting the pushrod in between tests, in an attempt to maintain the pushrod travel ort hogonally from the brake chamber face.

PAGE 54

43 Figure 6.5 : T est S eries #1 d ata Test Series #2 The second series of tests were designed to replicate wet conditions commonly encountered during real world operation , as shown in Figure 6.6 . S oapy water was applied to the brake chamber face between measurements. Soapy water was used to facilitate visual confirmation of the condition of the brake chamber face. Thirteen test s were completed with the following conditions and setup seen in Figure 6 .6 : Air pressure between 90 psi to 100 psi Clean brake chamber face Pushrod stroke of approximately 1 ½ inch es orthogonally Face of brake chamber wetted with soapy water between tests

PAGE 55

44 Figure 6.6 : Testing fixture s et up for T est S eries #2 The graph on Figure 6.7 depicts the displacement of the sensor during the thirteen tests complete d in the second series. During the last test in the series, a digital caliper was used to confirm the distance measured by the sensor when it was extended . When the cal iper crossed the face of the sensor, the distance reading by the sensor dropped from approximately 3.95 inches to 3.55 inches. This is apparent in Figure 6.7 during the time period between 260 and 280 seconds.

PAGE 56

45 Figure 6.7 : Test S eries #2 d ata with wet bra ke f ace Test Series #3 The third series of tests consisted of fourteen tests. These tests were completed with a flexible piece of metal tubing placed at an angle to the brake chamber face. This was done to simulate an angled brake chamber mounting bracket which is commonly found on semi trailers as depicted in the F igure 6.8 .

PAGE 57

46 Figure 6.8 : Angled brake chamber mounting b racket The setup and condition, apart from the angled insert , were the same as in T est S eries #1 . Those condition are as follows: Air pressure between 90 psi to 100 psi Clean brake chamber face Pushrod stroke of approximately 1 ½ inch es orthogonally Flexible metal piping inserted at angle between the brake chamber face and the sensor tests with an angled insert. Significant noise was present both when the pushrod was retracted and when the pushrod was extended. The following graph in Figure 6.9 provides t he displacement data from the third series of tests .

PAGE 58

47 Figure 6.9 : Test S e ries #3 data with angled s upport Test Series #4 The forth series of tests consisted of fifteen tests designed to replicate a dirty brake chamber face . Brake chamber faces are soile d during on road use due to road grime , mud , snow, ice or salt . A layer of mud was applied to the face of the brake chamber prior to testing as shown in Figure 6.10 . The conditions and setup of the fourth series of tests are as follows : Air pressure between 90 psi to 100 psi L ayer of mud applied to brake chamber face Pushrod stroke of approximately 1 ½ inch es orthogonally

PAGE 59

48 Figure 6.10 : Mud applied to brake chamber f ace The following graph in Figure 6.11 depicts the displacement data gathered by the Stroke sensor during testing with mud applied to the brake chamber face. Figure 6.11 : Test S eries #4 data with mud applied to brake face

PAGE 60

49 Test Series # 5 In t he fifth series of tests the adjustment collar fitted to the brake pushrod end was adjusted decreas ing the pu shrod stroke to approximately 1 inch measured orthogonally . The brake chamber face was cleaned prior to the start of this series of testing . Sixteen tests were completed using the following setup and conditions: Air pressure of between 90 psi to 100 psi Clean brake chamber face Pushrod stroke of approximately 1 inch orthogonally The F igure 6.12 depicts the setup of the pushrod adjustment, decreasing the stroke to approximately 1 inch measured orthogonally . Figure 6.12 : Testing fixture wi th stroke decreased to 1 inch sensor by the data acquisitions system are graphed and presented in Figure 6.13 . The sensor

PAGE 61

50 was adjusted between runs 4 & 5 and 15 & 16, in an attempt to maintain the 2 inch initial distance between the sensor face and the brake chamber face. The reason for noise seen when the pushrod was extended in the third test is unknown. Fi gure 6.13 : Test S eries #5 data with 1 inch stroke Test Series #6 The sixth serie s of tests were conducted with the same one inch pushrod stroke , but in this series of tests the sensor face was wetted with soapy water . Soapy water was applied to the face and body of the sensor to replicate traveling in wet conditions. Eighteen tests were conducted with the following setup and conditions. Air pressure of between 90 psi to 100 psi Clean brake chamber face Wetted ultrasonic sensor Pushrod stroke of approximately 1 inch orthogonally

PAGE 62

51 The following graph in Figure 6.14 depicts the sensor displacement data gathered during the sixth series of test . Figure 6.14 : Test S eries #6 data with wet sensor Test Series #7 The seventh series of t est s were conducted with a thin layer of mud applied to the face of the sensor itself, replicating potential conditions encountered during on roadway operation. This was done to replicate road grim e The condition of the sensor with mud applied is shown in F igure 6.15 .

PAGE 63

52 Figure 6.15 : Sensor fac e with thin layer of mud applied Twenty two tests were completed with mud applied to the face of the sensor and an approximate 1 inch stroke distance. The seventh test series were conducted under the following conditions and setup . Air pressure of between 90 psi to 100 psi Clean brake chamber face Thin layer of mud applied to ultrasonic sensor Pushrod stroke of approximately 1 inch orthogonally The following graph in Figure 6.16 depicts the displacement data collected during the twenty two tests conducted du ring test series #7 , which was the final test series completed for this study .

PAGE 64

53 Figure 6.16 : Test S eries #7 data with mud applied to sensor face

PAGE 65

54 CHAPTER VII DATA AND DESIGN ANALYSIS Ninety seven tests were completed, with seven different configurations and conditions. The volume of tests was completed to generate a reasonable statistical sample during al measurements taken using a digital caliper during each test series and found to be reasonably in agreement. The data gathered from the e Vericom Profile 5 software. A statistical mean was taken each time the pushrod was fully retracted and fully extended. Taking the mean over a time period was done to minimize the noise within the sensor. The mean distance of the pushrod when retracted wa s compared to the mean distance when fully extended to determine the pushrod stroke. The following graphs in Figure 7.1 and 7.2 illustrate how the means were determined.

PAGE 66

55 Figure 7.1: Test S eries #5 data Figure 7.2: Mean of pushrod retracted and extended After determining the mean distance of the sensor when the pushrod was fully retracted and fully extended during each test , E quation 4.1 was used to calculate the net

PAGE 67

56 measured brake stroke. After determining the bra ke stroke from each test, descriptive statistic s w ere used to determine the mean, standard deviation, variance, range of the percentage difference from the mean, and the absolute percentage difference from the mean for the data sets . The following sections analyze the results from each test series, and then all tests combined are analyzed to evaluate the function of the sensor for this proposed application. Test Series #1: 1.5 Inch Stroke Tests The data set from the first test series is presented in T ables 7.1 and 7.2 . It was noted during testing and wh ile examining the data , that on the first two tests completed the sensor was placed closer to the brake chamber than the 1.5 inch minimum distance specified by the . The errant placement caused the sensor to output its minimum reading of 1.5 inches when the pushrod was fully retracted into the brake chamber during the first two tests , as discussed in the previous section. tween the brake stroke sensor and the face of the brake chamber prior to applying air to the system. The cells titled air is applied to the brake chamber and the pushrod fully extends, contacting the pushrod stopper plate.

PAGE 68

57 Table 7.1: Test S eries #1 data Test 1 2 3 4 5 6 7 8 9 10 11 12 13 Retracted 1.50 1.50 2.00 2.01 2.02 2.00 2.00 2.01 2.04 2.02 1.99 2.00 2.03 Extended 2.94 3.06 3.60 3.67 3.61 3.70 3.61 3.73 3.71 3.58 3.72 3.74 3.70 Stroke 1.44 1.56 1.60 1.66 1.58 1.70 1.62 1.72 1.67 1.56 1.73 1.74 1.68 % diff. 13.86 6.41 3.63 0.01 4.69 2.55 2.63 3.43 0.60 5.90 4.01 4.65 0.95 Table 7.2: Test S eries #1 data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.635 Inches 0.084 Inches 0.007 Inches 0.294 Inches 4.65% to 13.86% The stroke measurement range, when considering all the data, has an approximate 0.3 inch spread, due partially to the movement of the non constrained pushrod end discussed in the previous chapter, and partially to the bottoming out of the sensor on tests 1 and 2. Removing the two tests where the sensor bottomed out, the data range decreases t o 0.174 inches as seen in Table 7.3 below. Using t he data from the T est S eries #1 with tests 1 and 2 removed results is a two standard deviation spread of ±0.12 inches which marginally meets the criteria needed for measurement of brake stroke by the Federa l Motor Carrier Safety Regulations (FMCSR) , which is measurement of brake stroke to the nearest 1/8 th of an inch (0.125 inch) . Table 7.3: Test S eries #1 data analysis excluding tests 1 and 2 Mean Standard Deviation Variance Range Range % difference from the mean 1.659 Inches 0.060 Inches 0.004 Inches 0.174 Inches 4.65% to 5.90%

PAGE 69

58 Test Series #2: 1.5 Inch Stroke and Wet Brake Chamber Face Tests Data from the second series of tests is presented in T ables 7.4 and 7.5 . The formatting of the tables follows the guidelines laid out for T est S eries # 1. In this test series the brake chamber face was wetted with soapy water to replicate expected real word condition encountered by the sensor in its proposed use. Table 7.4: Te st S eries #2 data Test 1 2 3 4 5 6 7 8 9 10 11 12 13 Retracted 2.00 2.01 2.05 2.04 2.05 2.06 2.09 2.11 2.14 2.13 2.16 2.18 2.21 Extended 3.66 3.76 3.59 3.76 3.72 3.80 3.85 3.85 3.82 3.88 3.89 3.91 3.95 Stroke 1.67 1.74 1.54 1.72 1.67 1.74 1.76 1.74 1.68 1.75 1.73 1.73 1.74 % diff. 2.52 1.93 10.37 0.55 2.22 1.99 3.07 1.88 1.57 2.39 1.13 1.42 1.59 Table 7.5: Test S eries #2 data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.710 Inches 0.059 Inches 0.004 Inches 0.222 Inches 3.07% to 10.37% The data from the second test series has a two standard deviation spread of ±0.118 inches. This spread again marginal ly meets the FMCSR measurement criterion of 1/8 th inch (0.125 inch) . The noise in the data is noticeable in this test series, but an aggressive smoothing function could minimize the noise without compromising the data. The tests suggest reasonably measure displacement for pushrod stroke measurement when the brake chamber is wet.

PAGE 70

59 Test Ser ies #3: 1.5 Inch Stroke and Angled Insert Face Tests Data from the third series of tests is presented in T ables 7.6 and 7.7 . In these tests an angled metal insert was inserted between the brake chamber face and the sensor to simulate a type of typical sem i trailer brake chamber mount. Table 7.6: Test S eries #3 data Test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Retracted 2.99 2.96 3.05 3.07 3.08 3.10 3.15 3.12 2.21 2.16 2.08 2.18 3.05 3.04 Extended 3.88 4.00 4.54 4.73 4.69 4.04 3.96 4.05 3.66 3.90 3.69 3.78 4.02 4.31 Stroke 0.89 1.04 1.49 1.66 1.61 0.93 0.81 0.93 1.46 1.74 1.61 1.61 0.97 1.27 % diff. 36.2 21.1 14.5 25.1 22.5 31.9 45.7 32.4 12.4 30.1 21.9 21.9 27.7 1.11 Table 7.7: Test S eries #3 data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.287 Inches 0.342 Inches 0.117 Inches 0.936 Inches 30.14% to 45.74% The data from the third test series clearly shows that the sensor cannot reliably and accurately measure the distance to the brake face when the brake face is obscured by an angled mounting bracket. A single standard deviation of the series of tests was de termined to be 0.342 inches, which is much greater than the 1/8 th of an inch (0.125 inch) criterion where the head of the sensor is pointed at a flat surface. The te sensor cannot reliably measure the pushrod stroke when there is an angled brake chamber mounting bracket between the sensor and the face of the brake chamber.

PAGE 71

60 Test Series #4: 1.5 Inch Stroke and Mud on Brake Chamber Face Tests D ata from the fourth series of tests is presented in T ables 7.8 and 7.9 . In these tests , a thick layer of mud was applied to the face of the brake chamber. During this series of tests, the sensor face itself was clear of any mud. Table 7.8: Test S eries #4 d ata Test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Retracted 2.11 1.98 2.01 2.01 2.14 2.06 1.98 2.03 2.03 2.09 2.04 2.07 2.11 2.08 2.11 Extended 3.63 3.61 3.77 3.75 3.71 3.65 3.76 3.74 3.74 3.69 3.73 3.74 3.74 3.85 3.73 Stroke 1.53 1.63 1.77 1.74 1.57 1.59 1.78 1.70 1.72 1.60 1.69 1.68 1.63 1.76 1.62 % diff. 8.93 2.03 5.91 4.14 5.89 4.76 6.75 2.17 2.81 4.07 1.11 0.57 2.15 5.51 2.70 Table 7.9: Test S eries #4 data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.667 Inches 0.078 Inches 0.006 Inches 0.259 Inches 6.75% to 8.93% The two standard deviations of the forth test series data result in a ± 0.156 inch spread, which is greater than 1/8 th o f an inch (0.125 inch) FMCSR measurement criterion and is not acceptable. There appears to be more noise in this data set versus T est S eries # 1 and # 2, but not nearly as much as T est S eries # 3 had. The tests suggest can reliably measure pushrod displacement , but not to the accuracy needed whe n the brake chamber face is covered in mud. Test Series #5: 1 Inch Stroke Tests Data from the fifth series of tests is presented in T ables 7.10 and 7.11 . In the following test series, the stroke length was shortened to approximately 1 inch to determine if

PAGE 72

61 any error was dependent on distance. The first series of tests with an approximate 1 inch stroke distance was completed with a clean and dry brake chamber face and sensor. Table 7.10: Test S eries #5 data Test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Retracted 2.00 2.03 2.09 2.11 2.01 2.08 2.13 2.19 2.21 2.23 2.29 2.32 2.35 2.39 2.45 1.98 Extended 3.19 3.26 3.30 3.35 3.21 3.29 3.35 3.30 3.36 3.41 3.41 3.48 3.54 3.58 3.53 3.18 Stroke 1.19 1.23 1.21 1.24 1.20 1.22 1.22 1.11 1.16 1.19 1.12 1.16 1.18 1.19 1.08 1.20 % diff. 0.74 4.05 2.49 4.54 1.75 2.91 3.24 6.13 2.16 0.74 5.05 2.16 0.24 0.74 8.77 1.75 Table 7.11: Test S eries #5 data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.180 Inches 0.044 Inches 0.002 Inches 0.154 Inches 4.54% to 8.77% In analyzing t he data from the fifth series of tests , the two standard deviation spread was determined to be ±0.088 inches which is acceptable based upon the FMCSR requirements for 1/8 th of an inch (0.125 inch) measurement precision. The variance in the data decreased with the shorter stroke, which would indicate that not all the variance is due to noise inherent in the sensor. The tests suggest displacement with in th e limits needed when the pushrod stroke is approximate ly 1 inch, with a clean and dry brake chamber face and sensor. Test Series #6: 1 Inch Stroke and Wet Sensor Tests Data from the sixth series of tests is presented in T ables 7.12 and 7.13 . During these tests the face of the sensor was sprayed with soapy water. The pushrod stroke was set to approximately 1 inch and the sensor and brake chamber face where clean.

PAGE 73

62 Table 7.12: Test S eries #6 data Test 1 2 3 4 5 6 7 8 9 10 11 12 13 Retracted 2.01 2.06 2.09 2.14 2.18 2.23 2.26 2.29 2.32 2.37 2.39 2.40 2.44 Extended 3.26 3.28 3.30 3.36 3.41 3.44 3.47 3.49 3.52 3.52 3.54 3.57 3.63 Stroke 1.24 1.22 1.20 1.22 1.23 1.22 1.21 1.21 1.20 1.16 1.16 1.17 1.19 % diff. 3.79 1.76 0.36 1.84 2.58 1.60 1.02 0.61 0.52 3.46 3.46 2.17 0.56 Test 14 15 16 17 18 Retracted 2.47 2.50 2.52 2.55 2.58 Extended 3.66 3.68 3.70 3.74 3.76 Stroke 1.19 1.18 1.18 1.20 1.18 % diff. 0.31 1.41 1.15 0.14 1.75 Table 7.13: Test S eries #6 data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.198 Inches 0.023 Inches 0.001 Inches 0.087 Inches 3.79% to 3.46% For an unknown reason the data from the sixth series of tests, which consisted of wetting the head of the sensor, had smaller deviations between the data points and the mean, then the fifth test series, which had the same stroke and a dry sensor. Potential ly, the soapy water cleaned any contaminants, or grime off study in the future. The sixth test series when analyzed, have a ±0.046 inch two standard deviation spread which suggests an reliably measure pushrod displacement to the accuracy required under FMCSR for brake stroke when the sensor itself is wet.

PAGE 74

63 Test Series #7: 1 Inch Stroke and Mud on Sensor Face Tests Data from the seventh series of tests is presented in T ables 7.14 and 7.15 . During these tests the face of the sensor was covered in a thin layer of mud. The brake chamber face was dry and free of mud during this test series. Additional tests were completed in rapid secession during this test series as movement of the sensor on the pushrod was observed and further testing was warranted. Table 7.14: Test S eries #7 data Test 1 2 3 4 5 6 7 8 9 10 11 12 13 Retracted 1.96 2.03 2.11 2.17 2.24 2.27 2.34 2.37 2.42 2.47 2.52 2.55 2.58 Extended 3.18 3.28 3.35 3.40 3.44 3.44 3.52 3.56 3.61 3.65 3.65 3.69 3.72 Stroke 1.22 1.25 1.23 1.23 120 1.17 1.18 1.19 1.19 1.18 1.14 1.14 1.14 % diff. 3.33 5.43 4.22 3.57 1.09 1.28 0.25 0.59 0.76 0.34 3.88 3.97 3.70 Test 14 15 16 17 18 19 20 21 22 Retracted 2.61 2.65 2.68 2.70 2.74 2.78 2.81 2.86 2.87 Extended 3.78 3.82 3.85 3.85 3.93 3.96 4.00 4.03 4.06 Stroke 1.16 1.17 1.17 1.16 1.19 1.19 1.19 1.17 1.19 % diff. 1.79 1.28 1.10 2.31 0.25 0.51 0.25 1.10 0.34 Table 7.15: Test S eries #7 data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.183 Inches 0.030 Inches 0.001 Inches 0.112 Inches 5.43% to 3.97% The data from the seventh series of tests suggest that the sensor is adequate in measuring the pushrod stroke to within the 1/8 th of an inch (0.125 inch) measurement

PAGE 75

64 criterion with a two standard deviation spread of ±0.06 inches . This test series suggests that adequate accuracy when thin layer of mud and the stroke is approximately 1 inch. Analysis of 1.5 Inch Stroke Tests The testing and analysis completed determined that the sensor is not the right tool to measure brake stroke when there is an angled brake chamber mounting bracket. A single factor analysis of variance (ANVOA) was completed on all four tests series with a 1.5 inch stroke. The ANVOA analysis when considering the data from the test with the angled bracket rejected the null hypothesis, showing that at least one of the m ean of the test series was not from the same population. Therefore, the data from those tests are excluded from these findings. Also excluded from this analysis was the data from the first two tests, where the sensor was placed t o o close to the brake chamb er . T ests S eries #1, #2 and #4 were aggregated together for analysis. The results of the analysis are presented in T able s 7.16 below. Table 7.16: 1.5 inch stroke data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.679 Inches 0.069 Inches 0.005 Inches 0.259 Inches 6.05% to 9.62% Single Factor ANVOA: Series # 1, 2, 3 ,4 Single Factor ANVOA: Series # 1, 2 ,4 2 tailed t Test: Series # 1 & 2 2 tailed t Test: Series # 1 & 4 2 tailed t Test: Series # 2 & 4 F=15.4 F crit ical =2.78 p V alve=2.95E 7 F=2.03 F crit ical =3.26 p V alve=0.147 t S tatistic = 2.06 t C ritical = 2.08 t S tatistic = 0.31 t C ritical = 2.06 t S tatistic = 1.62 t C ritical = 2.06 Th e single factor ANVOVA analysis and 2 tailed t Test analysis determine d it is probable that the means of Test Series #1, #2 and # 4 are from the same population . T he data from Test Series #1, #2 and # 4 was additionally tested for normal distribution fit 2

PAGE 76

65 T 2 T est are presented in 2 T est determined that the null hypothesis that the data is normally distributed should be rejected 2 2 Critical Value, and the p Value is smaller than the Confidence Interval. T able 7.1 7 : 2 T est for G oodness of Fit Confidence Interval 95% Degrees of Freedom 2 2 Critical V alue p V alue DOF= 3 2 =10.16 2 C ritical= 7 . 81 p V alue= 0.017 Although t he ANVOVA analysis and t Test analysis determine d it is likely that the mean value of Test Series #1, #2, and # 4 are from the same population , that population is not normally distributed . E ven though the spread of two standard deviations on most of the test series f al l within the accuracy required by FMC SR , the data is not normally distributed. The reason for the lack of normality could be due to error within the sensor, or error within the test setup. Further testing with a restrained brake pushrod could eliminate one source of error that is present duri ng this testing and may provide further clarity . Analysis of 1 Inch Stroke Tests The data from the tests with a 1 inch pushrod stroke, has a much smaller range and variance than the data from the tests done with a 1.5 inch pushrod stroke. All tests series completed with a 1 inch pushrod stroke were aggregated together for analysis. The results of the statistical analysis are presented in T able 7.1 8 .

PAGE 77

66 Table 7.1 8 : 1 inch stroke data analysis Mean Standard Deviation Variance Range Range % difference from the mean 1.187 Inches 0.033 Inches 0.001 Inches 0.168 Inches 5.10% to 9.34% Single Factor ANVOA: 2 tailed t Test: Series 5 & 6 2 tailed t Test: Series 5 & 7 2 tailed t Test: Series 6 & 7 F=1.45 F crit ical =3.17 p V alve=0.244 t Statistic = 1.41 t Critical = 2.07 t S tat istic = 0.22 t C rit ical = 2.06 t Statistic = 1.73 t C rit ical = 2.02 Th e single factor ANVOVA analysis and 2 tailed t Test analysis determine d that it is probable the means of Test Series #5, #6 and #7 are of the same population . The data from testing completed with a 1 inch stroke was also tested for a normal distribution 2 Test for goodness of fit 2 2 Test determined that the null hypothesis that the data was normally distributed was acceptable 2 Statistic is smaller 2 Critical Value, and the p Value is larger than the Confidence Interval. T able 7.1 9 : 2 Test for Goodness of Fit Confidence Interval 95% Degrees of Freedom 2 2 Critical V alue p V alue DOF= 3 2 = 4.10 2 Critical= 7.81 p Value= 0. 25 It was determined that at a 1 sensor meets the repeatability and accuracy requirements of this application required by the FMCSR. It was determined that it is probable that the means between the fifth, sixth and seventh series of tests are from the same population using a single facto r ANVOVA analysis and three 2 tailed t Tests. Additionally, it is probable that the data fits a normal distribution based upon the 2 Test for goodness of fit.

PAGE 78

67 Observations of Testing ent that the magnetic mounting bracket did not function to fasten the sensor securely to the pushrod. The plate can be seen in the data captured by the sensor. The following graph taken from Figure 7.3 of T est S eries #7 , shows that each time the pushrod is retracted following a test, it is further away from the brake chamber face. Figure 7.3: Gradual increase in sensor distance from brake face position to the extended position reveals the sensor moving on the pushrod. The figure below is the first six tests of the seventh test series as graphed above. The increase in distance of the position to the extended position and prior to the distance between the sensor and brake chamber face becoming relatively stable. The F igure 7.4, sh own again below, illustrates where the sensor is moving during the testing procedure.

PAGE 79

68 Figure 7.4: Increase in sensor distance due to tests The mounting method of the sensor does not appear to be suitable for repeatable measurements during this testing. It should be recognized that in these tests, the pushrod struck a solid steel stopping plate. On a brake the pushrod is stopped after rotating the slack adjuster, S cam and pushing the brake shoes into the brake drum, which could decrease the shoc k below a level where sensor mount cannot securely maintain its location. Further investigation is required to determine if this is the case or not. During the testing a digital caliper was also used to measure the movement of the liper was placed between the face of the brake chamber and the face

PAGE 80

69 to the output of the sensor as displayed on the Vericom VC400DAQ. Differences between the calip ere noted and determined to very between each measurement comparison . The measure ments difference between the caliper and the sensor were 0.03 inches to 0.13 inches, which is greater than the 1/8 th inch (0.125 inches) accuracy desired in this application. Figure 7. 5 : Digital caliper placement The testing conducted during this research illuminates the need for another tool to accurately measure the brake stroke , especially if the brake chamber is mounted using an angled bracket. The data obtained from testing with the ultrasonic sensor with an angled insert was significantly flawed and unusable with the requirements need ed for measur ing brake stroke. The following section o utlines an alternative design to address this and other problems noted during testing with the sensor as used.

PAGE 81

70 CHAPTER VIII ALTERNATIVE DESIGN The cost of ultrasonic sensors is prohibitive ly expensive , especially if used in large trucking fleets. The ret ail cost of 10 sensors, enough to equip all five axles of a typical tractor trailer combination, at the time of this paper is $5,170 [25] . For example, Swift Transportation currently operates approximately 20,000 tractors [26] , adding the Vericom ultrasonic sensors to both tractors and semi trailers would amount to over $103,000,000 . The mounting system of the ultrasonic sensor is not ideal for everyday use, as vibration could easily manipulate the alignment between the sensor and the brake face . Additionally, shock of repeated brake applications could disturb the location of the sensor, as occurred during testing. T he sensor s hard plastic housing is also not ideal for real word usage , since it is likely susceptible to damage from flying debris. Flying debris should be expected as the sensor would be subjected to the relatively harsh environment underneath a vehicle or trailer . However, t he mounting issue and survivability of the housi ng could be changed to accommodate on road use. One solution to the problems identified in this paper is to create a much simpler, less expensive and more reliable device than using an ultrasonic sensor. A potential solution that meets the se criteria sensor. Instead of the expense and complexity of a sensor which can provide precise real time measurements, the g o/no go sensor would only need to inform the driver when the brake stroke is about to , or when it reaches the adjustment limit. The go/no go sensor would consist of an enclosed metal housing approximately 3 to 4 inches in length. O ne end would fasten and remain fixed to the brake chamber mounting

PAGE 82

71 studs. The inside of the go/no go sensor housing would consist of a metal rod with a metal stop at the free end. A slider would ride on the metal rod and extend out of the bottom of the housing through an open slot which would run along the length of the housing. The bottom of the slider would extend beyond the open slot of the housing , and would have a latching mechanism similar in design to a hose clamp. The latching mechanism would latch onto the brake pushrod. When the pushrod was extended by brake application, the slider would move with the pushrod, while traveling on the metal rod built into the housing. The slider would slide on the metal rod using a non conductive material such as Delrin®. A switch would be mounted o n the interior of the housing mounted on the top side, and positioned where the slider reaches eith er the brake stroke adjustment limit, or desired position before maximum stroke length cut off is sensed . When the slider travels from its initial position the length of the adjustment limit, the slider would contact the switch and close it under condition s where the stroke length cut off is achieved . Closing the switch s , or individual brake has triggered the adjustment limit. The slider would be able to move past the contact switch without damaging the sensor in the event that the brake stroke goes beyond the adjustment limit. The advantages of the go/no go sensor over an ultrasonic sensor are multi fold as summarized in the following: The cost of a go / no go sensor would be much less than an ultrasonic sensor. A go / no go sensor could easily be ruggedized and better able to cope with the various harsh conditions and environmental changes encountered while in use .

PAGE 83

72 Further design work, and testing of a protot ype go / no go sensor would be necessary to determine all the advantages and disadvantages over an ultrasonic sensor, which is beyond the scope of this research .

PAGE 84

73 CHAPTER IX CONCLUSIONS The data examined during the completion of this research supports the need for a brake stroke measuring device to alert drivers when a pneumatic brake or brakes on their heavy vehicle and trailer reaches the FMCSR adjustment limit . A brake adjustment sensing device has the potential to decrease the frequency and severity of collision s involving heavy vehicle pneumatic braking system deficiencies . The analysis methods and failure modes presented in this paper demonstrate the importan ce of proper brake stroke adjustment to the cap ability of a fully laden heavy v ehicle to a stop in safe and efficient manner under emergency conditions . c an be used to measure the brake st r oke of an in service pneumatic brake , even when the brake chamber is wet or muddy. The measurements taken by the s ensor when the sensor is clean of debris on its sensing face and has a flat surface such as the brake chamber from which to read distances, have been shown to be marginally sufficient for the precession required in this application. It should be cautioned, there were several sources of variability of the measurement due to the sensor mount , the design of the testing apparatus and within the sensor itself. It is beyond the scope of this project to determine what the most significant source of the observed er ror was, but it is suggested for future research . The t esting completed during this project determined that ultrasonic sensor can be accurate to the required degree for brake stroke measurement s of approximately 1 inch during brake inspection s performed in a controlled environment. With

PAGE 85

74 the exception of an obstruction between the sensor and a flat surface, such as an angled brake chamber mounting bracket , commonly found on trailer axles . is not an ideal sensor fo r a permanent brake stroke measuring device in real world applications. An alternative design for a brake stroke measuring device was presented, which potentially has many advantages over an ultrasonic sensor. Further study of this proposed alternative go/ no go device is needed to fully determine the feasibility for a real world application and universal adoption by manufacturers and the trucking industry .

PAGE 86

75 REFERENCES [1] Federal Motor Carrier Safety Administration, "Commercial Motor Vehicle Facts March 2013," US Deoartment of Transportation, Washington D.C., 2013. [2] Federal Highway Administration, Code of Federal Regulations Title 23, Part 658 Truck Size and Weight Route Designations Length, Width and Weight Limitations, Washington D.C. : Goverment Publishing Office, 2013. [3] Office of Research and Analysis, "Large Truck Crash Causation Study FMCSA RI 05 037," Federal Motor Carrier Safety Administration, Washington D.C., 2006. [4] Michigan State Police Motor Carrier Enforcment Division, "Fatal Accident Complaint Team Data 1996 2001," Michigan State Polic, Detroit, 2002. [5] L. C. Buckman, Commercial Vehicle Braking Systems: Air Brakes, ABS and Beyond, Warrendale: Society of Automotive Engineers Inc., 1998. [6] Bendix Spicer Foundation Brake LLC, "The Compelling Case for Air Disc Brakes in Heavy Truck Braking: A WHite Paper," Bendix Spicer Foundation Brake LLC, Elyria, 2011. [7] State of California Department of Motor Vehicles, "Commercial Driver Handbook, Section 5: Air Br akes," 2017. [Online]. Available: https://www.dmv.ca.gov/portal/dmv/detail/pubs/cdl_htm/sec5/!ut/p/a1/nZHLboMwEE W_JQuWaIZXMEurSSHQNJWoEuMNomADVTCkcVHbry9E2UZ9zGKkK81czT0D HBhwVYxtXei2V8Vx1nyZJ5vV2opCTELvgSJNomi3X -

PAGE 87

76 _tMHDgABx4qfSgG8iqbszLXmmhdC6UgZM2cHh_ORtYVs e80Z2BZ1F6885QthVkT uVIX7iWKd2iMl1. [Accessed 20 October 2017]. [8] Bendix Commerical Vehicle Systems LLC, Bendix Product School Training Manual, Knorr Bremse Group, 2004. [9] Timmy, "NewTruckspring.com," 6 September 2016. [Online]. Available: http://w ww.newtruckspring.com/category/wiki parts/air brake chambers/. [Accessed 9 November 2017]. [10] Commercial Vehicle Safety Alliance, North American Standard Out of Service Criteria, Greenbelt: Commercial Vehicle Safety Alliance, 2014. [11] Bendix Commercial Vehicle Systems LLC, "SD 03 4510 Service Data Bendix SR 2 Trailer Spring Brake Valve," Bendix Commercial Vehicle Systems LLC, USA, 2007. [12] Bendix Spicer Foundation Brake LLC, "Technical Bulletin Bendix Dash Valve Trip Pressure/DOT Inspectio ns," Bendix Spicer Foundation Brake LLC, 2008. [13] Ministry of Transportation, "The Offical Air Brake Handbook Spring (Parking and Emergency) Brake Subsystem," Ontario, 28 April 2017. [Online]. Available: https://www.ontario.ca/document/official air brake handbook/spring parking and emergency brake subsystem. [Accessed 1 October 2017]. [14] Federal Motor Carrier Saftey Administration, 49 Code of Federal Regulations part 393.53, Subpart C, Washington D.C.: U.S. Goverment Publishing Office, 1998.

PAGE 88

77 [15] Bennett/Norman, "Heavy Duty Truck Systems, 4e," Thomson Delmar Learning, 2006. [Online]. Available: http://slideplayer.com/slide/4008159/13/images/57/Stroke sensing+Automatic+Slack+Adjuster.jpg. [Accessed 1 October 2017]. [16] Haldex Commercial Vehicle Systems, "Service Manual Transit and Coach Applications Automatic Brake Adjusters," Haldex Commercial Vehicle Systems, Kansas City, 2003. [17] Federal Motor Carrier Safety Administration, 49 Code of Federal Regulations Part 393.55 Subpart C Antilock Brake Systems, Washington D.C.: U.S. Goverment Publishing Office, 2011. [18] MGM Brakes, "MGM Brake Tech Self Study Training Program Spring Brakes/Service Chambers," MGM Brakes, Charlotte, 20 11. [19] N. Leandro, CSVA Releases Results from Brake Safety Day, Greenbelt: Commercial Vehicle Safety Alliance, 2017. [20] R. Heusser, "Heavy Truck Deceleration Rates as a Function of Brake Adjustment, SAE910126," SAE International, Troy, 1991. [21] D. E. Brill, Commercial Motor Vehicle Crash investigation, Jacksonville: Institute of Police Technology and Management, 2000. [22] W. Bartlett, "Calculation of Truck Deceleration Basied on Air Pressure Rise Time and Brake Adjustment, SAE 2004 01 2632, " SAE International, Troy, 2004. [23] W. Bartlett, "Calculation of Deceleration Rates for S Cam Air Braked Heavy Trucks Equipped with Anti Lock Brake Systems, SAE 2007 01 0714," SAE International,

PAGE 89

78 Troy, 2007. [24] Vericom Computer, Vericom Computer, 2013. [25] Vericom Computer, "V'Stroke Packages," Vericom LLC, 2016. [Online]. Available: http://vericomcomputers.com/products/advanced brake testing/v stroke packages/. [Accessed 6 November 2017]. [26] Swift T ransportation Company, "History Over 50 Years of Transportation Solutions," Swift Transportation, 2017. [Online]. Available: http://www.swifttrans.com/who we are/history . [Accessed 7 November 2017].

PAGE 90

79 APPENDIX A . Commercial Vehicle Factor/ Skid to Stop Analysis Example Problem

PAGE 91

80 B. Weight Distribution Analysis Example Problem

PAGE 92

81

PAGE 93

82 C. Heusser Method Example Problem

PAGE 94

83

PAGE 95

84

PAGE 96

85 D. Bartlett/Heusser Method Example Problem

PAGE 97

86

PAGE 98

87

PAGE 99

88

PAGE 100

89

PAGE 101

90

PAGE 102

91

PAGE 103

92

PAGE 104

93

PAGE 105

94 E . Data Collection Summary Test S eries # 1 : dry/clean

PAGE 106

95

PAGE 107

96

PAGE 108

97

PAGE 109

98

PAGE 110

99

PAGE 111

100

PAGE 112

101

PAGE 113

102

PAGE 114

103 Test S eries # 2 :

PAGE 115

104

PAGE 116

105

PAGE 117

106

PAGE 118

107

PAGE 119

108

PAGE 120

109

PAGE 121

110

PAGE 122

111

PAGE 123

112 Test S eries # 3 : stroke angled insert

PAGE 124

113

PAGE 125

114

PAGE 126

115

PAGE 127

116

PAGE 128

117

PAGE 129

118

PAGE 130

119

PAGE 131

120

PAGE 132

121

PAGE 133

122 Test S eries # 4 :

PAGE 134

123

PAGE 135

124

PAGE 136

125

PAGE 137

126

PAGE 138

127

PAGE 139

128

PAGE 140

129

PAGE 141

130

PAGE 142

131

PAGE 143

132 Test S eries # 5 : stroke dry/clean

PAGE 144

133

PAGE 145

134

PAGE 146

135

PAGE 147

136

PAGE 148

137

PAGE 149

138

PAGE 150

139

PAGE 151

140

PAGE 152

141

PAGE 153

142

PAGE 154

143 Test S eries # 6 :

PAGE 155

144

PAGE 156

145

PAGE 157

146

PAGE 158

147

PAGE 159

148

PAGE 160

149

PAGE 161

150

PAGE 162

151

PAGE 163

152

PAGE 164

153

PAGE 165

154

PAGE 166

155 Test S eries # 7 :

PAGE 167

156

PAGE 168

157

PAGE 169

158

PAGE 170

159

PAGE 171

160

PAGE 172

161

PAGE 173

162

PAGE 174

163

PAGE 175

164

PAGE 176

165

PAGE 177

166

PAGE 178

167

PAGE 179

168

PAGE 180

169