 |
 |
Root Cause Analysis
Bridgestone/Firestone, Inc.
|
|||
| Summary Root Cause Analysis Bridgestone/Firestone, Inc. |
|
Outline: |
|
|
Based upon detailed, thorough and scientific analysis, Bridgestone/Firestone, Inc. has concluded that the increased rate of tread separation claims related to the P235/75R15 Radial ATX and ATXII tires and Wilderness AT tires of the same size manufactured at the company's Decatur, Illinois facility is not due to one over-riding factor. Instead, the company has concluded that a combination of design factors, external factors such as low inflation pressure and certain manufacturing factors at the company's Decatur plant - in extreme cases and working together - have contributed to the phenomenon. This summary report does not address vehicle performance or driver response issues after a tread separation occurs. That interaction, which is certainly an important factor in determining the cause of the serious accidents related to tires and the vehicles upon which they are mounted, is beyond the scope of the company's review and this summary. |
|
| II. The Analytical Process | TOP |
|
The scientific process to determine the root cause of the increased accident rate associated with P235/75R15 Radial ATX tires and Wilderness AT tires of the same size manufactured at Bridgestone/Firestone, Inc.'s Decatur, Illinois plant involved field testing, laboratory work, data analysis, computer modeling and extensive dissection of hundreds of returned recalled tires. The analysis was performed by a team of engineers, statisticians and technical experts from both the United States and Japan. In addition, the company analyzed reported claims information, tire performance information and manufacturing information from all of its plants in the United States. Outside laboratories were used to provide independent analysis. |
|
|
In addition to laboratory work, computer modeling and outdoor testing, inspections were performed on more than 2,500 tires. Cut tire inspections were performed on more than 850 tires to examine potential internal variances as well as any service effect on internal components. The review included as many as 40 measurement points per tire and included analysis of the following tire components:
For reference, the internal components of a tire are shown in Attachment 1, indicating the relative positions of certain of the key components. |
|
| III. Tire Design and Its Effect on Stress | TOP |
| The shape of the tread, the location of the various components in relation to the tread patterns, and the material properties of the components govern how each component is stressed while a tire is in use. Through each rotation of a tire, tremendous forces are transferred and absorbed by the tire components. The relative shape of the tread and the relative location of each component to impact points on the tread affect how the road loads are transferred to the tire components. Aspects of the component geometry in the recalled Radial ATX and Wilderness AT tires have been analyzed using finite element modeling to determine comparative stress levels at the tire surface and at the interbelt area. | |
| In the design of the P235/75R15 ATX and Wilderness AT tires, there is a pattern of hollowed-out spaces or cavities between the lugs around the edges of the tread on the tire's shoulder. These spaces are called "shoulder pockets." The shoulder pocket design of the recalled Radial ATXII tires and the P235/75R15 Wilderness AT tires are significantly different (see Attachment 2). It was observed that deep surface cracking at the bottom of the shoulder pocket was more likely to occur in the Radial ATX in the recalled size than in other tires, including the recalled Wilderness AT tires. | |
| To determine how this cracking occurred, the team examined the strain at the surface of the pocket through finite element modeling. Finite element modeling (FEM) is performed by creating a grid of elements that represents specific points of the tire and a computer is used to keep track of how each of the points responds to loads on the tread as if the tire were actually rolling on the road. As shown by the FEM, the shoulder pocket design of the recalled ATX permits more stresses to occur inside the shoulder pocket and could lead to cracking at the shoulder pocket bottom (see Attachments 3, 4 and 5). | |
| In addition to the FEM, the company conducted a "painted tire" test, painting the tread and shoulder pocket areas of recalled tires and rolling them on a large glass plate. The stresses on the pocket area were captured with a digital camera and the resulting photographs analyzed (samples shown in Attachment 6 and 7). The photographs show a clear stress pattern at the bottom of the shoulder pocket in the recalled Radial ATX as compared to the P235/75R15 Wilderness AT. | |
| The review of the shoulder pocket area was important to the team because cracks in that area (when combined with other factors, including design differences between the recalled Radial ATX and the Wilderness AT, manufacturing process issues and external factors) could, in extremely rare circumstances, become the starting point of a tread separation. | |
| In addition to the shoulder pocket, the team reviewed the area between the tire shoulder and the edge of the top steel belt. More specifically, the measurement, called the "W7 gauge", is the distance from the top of the #2 steel belt (the steel belt closest to the surface of the tire) and the inside of the shoulder pocket (the deepest part of the surface of the tire). In reviewing quality assurance data from the Decatur plant, it appears that some Decatur Radial ATX tires may have had a shallower W7 gauge. This thinner gauge, combined with the shoulder pocket design, appears to have contributed to permitting cracks to reach the belt edge, therefore contributing to tread belt separations, especially when combined with the different skim stock produced at Decatur. After 1996, there is no evidence of W7 gauge variances. | |
| IV. Effect of Material Strength on Tire Strength | TOP |
| The basic elements of the strength between the steel belts of a tire are the adhesion strength of the rubber coating the steel cords (the "skim stock") and the configuration of that material. Note that the two belts are not "glued" together; rather, the rubber skim stock covering the steel cords are vulcanized together during the curing process, as are the other rubber components of the tire. When the dissection analysis of several hundred of the returned recalled tires was completed, there was no apparent difference among the various manufacturing plants in the amount of the material covering the cords. However, it was clear from the claims analysis that the rate of tread separations from the Decatur plant was higher. Therefore, a more in-depth analysis of the strength of the skim stock was necessary to determine if there was a difference among the plants in the strength of the material covering the steel belts. A test was established to measure the force required to pull apart a one inch strip of the steel belt package to compare the adhesion strength between the two belts. | |
| This test determined that the skim stock from the Decatur manufacturing process was different from the skim stock used in making these tires at other BFS plants and required less force to separate the two belts. Although the pull strength of the Decatur skim stock used in making these tires is within the range of tire design requirements and Firestone's specifications, its characteristics differed, including an initial relative strength that provided less margin than that found in tires from other BFS plants. | |
| The Decatur system used to handle the rubber skim stock coming out of the rubber mixing machine resulted in the use of more processing lubricants, and therefore, more inorganic material was found in the Decatur skim stock rubber, as compared to skim stock rubber from other BFS plants. Based on their experiments, the review team concluded that while the skim stock was within tolerances, there was a correlation between the amount of residual materials in the skim stock and the force required to pull apart two strips of rubber cured together. | |
| V. How Tire Use Affects Stress and Strength | TOP |
|
|
| The essential question in a particular tire's resistance to separation is a comparison of stress versus strength. The stresses on a tire depend on the vehicle load, speed, road conditions and air pressure in the tire. However, the strength of a tire at any point in time is governed by the tire's past use, temperature, inflation pressure and original design and construction. The role of inflation pressure as an aspect of use has both an accumulative effect in building up a life history of overloading and hotter running temperatures and an immediate effect in decreasing the strength of the components in a tire. A tire exposed to high heat can be less resistant to stress because heat deteriorates rubber, which can reduce adhesion. | |
| Several scientific analysis methods were used to assess the role of inflation pressure in the performance of tires. The results of field-testing performed at the BFS test track located in Fort Stockton, Texas, are indicated in Attachment 8. The test was performed by running a set of Wilderness AT tires in size P235/75R15 at various inflation pressures for approximately 45 minutes at constant speed until the tire temperature stabilized, then stopping to immediately take the temperature of the internal materials at the belt edge with a needle pyrometer. Attachment 8 shows a clear increase in temperature at lower inflation pressure. | |
| The importance of the higher temperatures in the strength of the interbelt materials is shown by the experiment expressed in Attachment 9, which shows the relative decrease in adhesion strength of the steel belt rubber compound at different temperatures. This experiment was conducted using sample strips of tires. Although this test does not account for all the factors in the strength of the belts, the results do show a definite decrease in adhesion strength at higher temperatures and highlights the difference between properly inflated and improperly inflated tires discussed in the prior experiment. | |
| The role of inflation pressure in a tire's performance is also evident in the tire industry standards for safe tire design, referred to as the Tire and Rim Association Standards, which are incorporated into federal government standards under the Federal Motor Vehicle Safety Standards. These standards list maximum loads for tires, which loads decrease as recommended inflation pressure decreases. All modern tires are designed to comply with these standards and the recalled tires were within the safe range set by these standards. However, based on testing at the company's Fort Stockton proving grounds, it appears that the vehicle load levels, when coupled with the tire pressure initially specified for the Ford Explorer with the P235/75R15 size tire, resulted in a tire that was approaching the limits of its load carrying capacity (see Attachments 10 and 11). Since tires on average lose about one psi of pressure per month, in as little as four months, an unchecked P235/75R15 tire mounted on the left rear of an Explorer would be overloaded, according to the Tire and Rim Association Standards and the Federal Motor Vehicle Safety Standards. Attachment 12 shows a sample of cold tire pressure measurements that are excerpted from surveys conducted over a period of five years as part of the company's continuous tire monitoring program covering a variety of passenger and light truck tires. | |
|
|
| Tire industry experience with all populations of steel belted radial tires has demonstrated that the substantial majority of tread belt separations are caused by service-related tire damage. These causes include such conditions as cuts, punctures, improper repairs, severe impacts and improper tire-vehicle applications. The root causes discussed in this summary were not the cause of all tread/belt separation incidents: many were attributable entirely to unrelated damage conditions. As an example, Firestone examined both recalled tires and those that were subject to claims to determine if repairs might have an impact on the durability of a tire. In its survey, only about 16% of returned recalled tires (tires which were still being used at the time of their return) had one or more puncture repairs. However, about 35% of tires which were the subject of tread separation claims (returned to the company because the tire had failed and allegedly had resulted in damage of some type) had puncture repairs. Improper repairs that did not effectively stop air leaks made up about two-thirds of these repairs. It appears that improper repairs are a factor in tread separations. | |
| VI. Claims Data Correlation and Support | TOP |
| In addition to the analysis of the physical properties and design characteristics of the recalled tires and their manufacturing history, the team analyzed the claims database for the recalled ATX and Wilderness AT tires as well as the other tire lines included in the NHTSA Preliminary Evaluation dated May 2, 2000 (the "PE"). Particular attention was paid to the claims rates of other Decatur produced tires included in the PE. An independent team of auditors verified the accuracy of information contained in the database (based on a review of the actual claims files), and confirmed that all relevant information, such as DOT numbers, were included in the database. Once all of the data had been verified, a team of independent statisticians conducted an extensive analysis of the data. The results of this statistical analysis were then compared with the findings of the technical team to determine how the various identified factors compared to the claim reports. | |
| From the outset, the statistical analysis confirmed several of the company's important early findings: that the number of tires that were more likely to experience belt separation was extremely limited, and that the initial safety recall was more than adequate. Wilson and Joliette produced P235/75R15 Wilderness AT tires experienced a parts per million claims rate of 7. This is substantially lower than the recalled Decatur-produced Wilderness AT, which experienced a parts per million claims rate of 49. These two numbers, however, are still significantly lower than the parts per million claims rate of the ATX line of tires: Decatur-produced ATX tires showed a claims rate of 436 ppm, while ATX tires produced at other Firestone plants had a 94 ppm claims rate. Additionally, the analysis showed that all 16 inch Wilderness tires, regardless of manufacturing plant, exhibited extremely low claims rates. In fact, there are no claims recorded against the Decatur produced 255/70R16 Wilderness AT tires and the Wilson and Joliette Wilderness AT tires of the same size experienced a claims rate of 5 ppm. (Attachment 13 and 14 provide summaries of claims rate experience for the tires addressed in this paragraph.) | |
| The final piece of the statistical analysis related specifically to Decatur-produced tires. At issue was whether the manufacturing differences at that plant affected other tire lines. As noted above, the technical review found that these variances in manufacturing alone did not contribute to the higher claims rate. Instead, it was found, and confirmed by the claims data, that the manufacturing variances must work in combination with other factors such as tire design differences and external factors such as inflation specifications. The statistical analysis clearly supports the technical findings. If the conditions at the Decatur plant were a sole contributor to the increased claims rate, all tires manufactured in Decatur would show a similar claims rate history. To the contrary, and based on the detailed analysis of the tires included in the PE, more than 55% of tires identified in the PE and produced in Decatur have never experienced a single claim in over ten years. Overall and on a ppm basis, other Decatur produced tires that are part of the PE show a claims rate of just 35 (see attachment 15). | |
| VII. Conclusion | TOP |
| The statistical and technical analyses conducted by Firestone in its root cause review demonstrate that: the August 9, 2000 recall was more than adequate to protect public safety; the higher percentage of tread separations were isolated to a very small percentage of its tires; and a series of factors, acting in combination and in extreme conditions, were responsible for the increased rate of tread separation claims experienced by the recalled tires, especially the Radial ATX tires manufactured in Decatur. | |
| The technical analysis involved a review of external conditions, tire design changes between the Radial ATX and the Wilderness AT (both in size P235/75R15), and an examination of manufacturing variances among Firestone's passenger and light truck tire plants. The statistical review included verification of the reports contained in the company's claims database and an analysis of the population of recalled tires as compared to other all tires contained in the PE and produced both at Decatur and at other Firestone plants. | |
|
Based upon its findings, the company is implementing a number of changes to further enhance public safety and regain the confidence of the driving public. This includes changes in how the company collects, analyzes and manages data and the implementation of recommendations from a special Comprehensive Quality Review Team. This Comprehensive Quality Review Team, which was formed following the announcement of the recall, is focusing it attention on production operations, tire design enhancements and further quality assurance standards.
|
|
| Click on any image below for enlarged view. | |
| #1 | TOP |
 |
|
| #2 | TOP |
 |
|
| #3 | TOP |
 |
|
| #4 | TOP |
 |
|
| #5 | TOP |
 |
|
| #6 | TOP |
 |
|
| #7 | TOP |
 |
|
| #8 | TOP |
 |
|
| #9 | TOP |
 |
|
| #10 | TOP |
 |
|
| #11 | TOP |
 |
|
| #12 | TOP |
 |
|
| #13 | TOP |
 |
|
| #14 | TOP |
 |
|
| #15 | TOP |
 |
|
| TOP | |