Purdue News

March 30, 2006

Diagnostic method drives better tire testing for industry

WEST LAFAYETTE, Ind. — Mechanical engineers at Purdue University have developed a system that uses sensors and mathematical models to detect defects in newly manufactured tires better than conventional inspections, promising to help industry meet more stringent federal tire-durability requirements.

Douglas E. Adams and
Timothy J. Johnson

Download photo
caption below

The diagnostic technique works by analyzing vibration waves passing through a tire to detect damage that leads to cracks in the bead area, where the tire connects to the steel rim of the wheel. A crack will sometimes form in the bead area and spread entirely around the tire, causing the tire to lose air or otherwise fail, said Douglas E. Adams, an associate professor of mechanical engineering who developed the system with doctoral student Timothy J. Johnson.

The National Highway Traffic Safety Administration issued the new tire-durability standards covering aspects including the detection of bead-area damage after Congress in 2000 passed the TREAD Act, for Transportation Recall Enhancement, Accountability and Documentation. The law was prompted by a series of SUV accidents in the 1990s, some fatal, which resulted in the recall of 6.5 million tires.

"The fatigue endurance testing needed to ensure that all automotive tires meet the new durability requirements is time consuming and costly," Adams said. "And because the testing is carried out by technicians conducting manual inspections, the results can vary based on a technician's skill and other factors.

"The bottom line is that there hasn't been any objective way to determine when the tire has bead-area damage."

Johnson will present a research paper about the diagnostic technique on April 6 during the Society of Automotive Engineers' 2006 World Congress in Detroit. Adams will give a talk on April 7 focusing on developing better mathematical models and simulations.

"When you build a mathematical model of a tire, part of the purpose is to predict how a crack will spread once it starts," said Adams, whose research is based at Purdue's Ray W. Herrick Laboratories. "But another big part of the model is to determine how the crack gets started in the first place. If industry can better understand how cracks initiate, manufacturers can build a better tire."

Such models promise to help industry improve the tire-testing process. Manufacturers now randomly test new tires using rollers to create punishing forces and conditions designed to push tires to their limits and speed up the wear caused by actual driving.

"Periodically, the technicians will stop the test and manually inspect a tire by passing their hands over the tire bead area to feel for little air bubbles," Adams said.

The bubbles form when rubber in the bead area separates from underlying layers.

"These tests will go on for several days on a single tire to determine the tire's durability, and once the bubbles appear, the test is stopped," Adams said. "You have to have somebody in the plant 24 hours a day doing this testing. Manufacturers would rather automate the testing process, which would enable them to detect bead-area damage at its very earliest stages, saving time and money."

The Purdue researchers compared their diagnostic system's performance to a technician's manual inspection at a tire-manufacturing plant. The technician discovered bubbles in a tire after two and a half days of mechanical endurance testing, and Purdue's system detected damage in the same tire about 12 hours into the test, or one-fifth the time it took for the manual inspection.

"Detecting damage a day and a half earlier than the technician found it means this kind of system should save a lot of time and money," Adams said. "We are able to do this because we can detect damage before the bubble actually forms."

The Purdue engineers developed their diagnostic system using an SUV parked over a testing apparatus so that the left front tire sat on rollers spun by an electric motor at 40 mph, while a hydraulic device exerted downward force on the tire to simulate the weight of passengers. The researchers also heated the tire with a portable furnace to mimic worst-case driving conditions and compared the difference between damaged and undamaged tires.

"Damage causes a decrease in tire stiffness because a tire that has a crack in the bead area doesn't resist as much when you push down on it," Adams said. "That's significant because the more the tire squishes, the less able it is to resist the growth of the crack."

As the test tire rotated, the engineers used sensors called "tri-axial accelerometers" attached to the axle to pick up vibration waves passing through the tire. The data were fed to a computer, where software containing algorithms interpreted the information.

Damage to the bead area causes the tire to rotate unevenly, and data from vibration waves reveal this slight wobble. The data were displayed on a computer monitor in a "wavelet map" that enabled the engineers to not only detect damage but also possibly pinpoint its location on the tire.

"The goal is to collect data and identify when the bead area damage has initiated, and the whole purpose of doing that is so that we can build better models and know the tires are as durable as we want them to be," Adams said.

The engineers plan to extend their research to test what happens to a tire when drivers hit curbs while parallel parking. The engineers will simulate the curb-impact effects with a tower-like apparatus that drops a weight from a height of about 6 feet onto an object being tested.

The research has been industry funded.

The technique also might lead to diagnostic systems in cars and military vehicles that would constantly analyze vibration passing through tires and alert a driver when the bead area has been damaged. The same general method also offers promise in various applications unrelated to tires, including quality assurance in engine manufacturing and detecting damage in jet engines and electrical-generating turbines in hydroelectric and nuclear power plants.

"The method is generic enough that it could be used to detect damage in any mechanical system that rotates," Adams said.

Writer: Emil Venere, (765) 494-4709, venere@purdue.edu

Sources: Douglas Adams, (765) 496-6033, deadams@purdue.edu

Timothy J. Johnson, (765) 496-8438, johnso90@purdue.edu

Purdue News Service: (765) 494-2096; purduenews@purdue.edu

 

Note to Journalists: Timothy J. Johnson will present the research paper at 9:30 a.m. on April 6 in Room W1-53 at the COBO Center in Detroit during the Society of Automotive Engineers' 2006 World Congress.

 

PHOTO CAPTION:
Douglas E. Adams, left, an associate professor of mechanical engineering, and doctoral student Timothy J. Johnson use their diagnostic system to detect tire damage as sensor data are displayed in a "wavelet map" on a computer monitor, enabling the engineers to not only detect damage but also possibly pinpoint its location on the tire. The tire-testing research, at Purdue University's Ray W. Herrick Laboratories, uses sensors and mathematical models to detect defects in newly manufactured tires. (Purdue News Service photo/David Umberger)

A publication-quality photo is available https://www.purdue.edu/uns/images/+2006/adams-tires.jpg

 


ABSTRACT

Rolling Tire Diagnostic Experiments for Identifying Incipient Bead Damage Using Time, Frequency,
and Phase Plane Analysis

Timothy J. Johnson and Douglas E. Adams, Ph.D.,
Ray W. Herrick Laboratories, Purdue University

Automotive tires must meet the U.S. DOT/NHTSA durability regulations set forth in FMVSS139 for vehicles of less than 10,000 lbs GVW. Offline inspection during fatigue endurance testing is time consuming, costly, and involves variability in durability ratings; therefore, a means for determining the health of rolling tires online during testing is desired. This paper investigates the use of passive time, frequency, and phase-plane methods for detecting structural changes in rolling tires. Operational data were taken on a chassis roller where the left front tire of a vehicle was tested under various loads. Acceleration measurements were recorded on the spindle arm and the data were synchronized with the rotations of the wheel and roller to analyze cycle-to-cycle variability. Modal measurements were also made at the end of each operational test; this data provided some indication of structural changes in the tire. Restoring forces generated for the low frequency content of the operating data were inconclusive. Transient wavelets applied to the operational data were more indicative of small-scale tire damage than were harmonic wavelets.


 

To the News Service home page

Newsroom Search Newsroom home Newsroom Archive