Characterization of tire wear EPA-4eo/3-8i-o36
November 1981
>IRVIN/CXILSRXiN
CHARACTERIZA TION OF TIRE WEAR PARTICULA TES
by
Leonard Bogdan
Thomas M. Albrechcinski
Advanced Technology Center
Calspan Corporation
Buffalo, New York 14225
November 1981
Contract No. 68-03-2781
Project Officer
Craig Harvey
EMISSION CONTROL TECHNOLOGY DIVISION
U.S. ENVIRONMENTAL PROTECTION AGENCY
ANN ARBOR, MICHIGAN 48105
APPLIED
•Bl-i I i_l lJ^Jl-il Qtay CENTER TECHNOLOGY
•=*•» ' o-r'"^'*-!*.!-!^ T wsai-o i »-« qnoup
P.Q BOX 400, BUFFALO, NEW YORK 14225 TB_. (716) 632-7500
Xffi
-------
FOREWORD
This report is submitted in fulfillment of EPA Contract No. 68-03-2781.
This report has been reviewed by the Emission Control Technology Division,
U. S. Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and policies
of the U. S. Environmental Protection Agency, nor does the mention of trade
names or commercial products constitute endorsement or 'recommendation for use.
-------
ABSTRACT
The principal objectives of this research program were to collect and
characterize tire wear particulates as produced by the operation of passenger
car tires under the controlled test conditions of the laboratory. Additionally,
sufficient quantities of both the airborne particulates (<30 vim), and the non-
airborne particulates were to be collected to permit bioassay tests to be
performed by the EPA on the individual samples.
Tests were performed on a single radial tire (DR78-14) and a single bias
tire (D78-14) using the capabilities of the Calspan Tire Research Facility
(TIRF) that features a flat test surface consisting of a coated steel belt that
simulates the friction properties of actual road surfaces. Tires were tested
under typical load and inflation pressure conditions using a TIRF schedule of
applied wheel torque and wheel steer (slip) angle that results in normal rates
of tire wear and a uniform cross-tread wear. Tires were operated within a
special enclosure fabricated for the purpose of facilitating a gross separation
of the nonairborne particulates from the airborne particulates that were
collected by high volume samplers and a cascade impactor. Particulate charac-
terization was accomplished using the techniques of the optical and the scanning
electron microscope.
Airborne rubber particulates were produced in an extremely small quantity
irrespective of the tire construction used or the relative severity or the
type of wear condition that was imposed on the tire. The bulk of the collected
airborne particulates was identified as contaminant debris, principally stain-
less steel. Tests on the radial tire produced a particle size distribution
with a median particle diameter about one-half of that produced by identical
tests on the bias tire. Consequently, the latter is responsible for the
generation of smaller quantities of airborne particulates than the former.
Samples of the collected particles were sent to EPA for bioassay (Ames)
tests but difficulties (for now) have been encountered in dissolving the
particulates.
This report was submitted in fulfillment of Contract No. 68-03-2781 by
the Advanced Technology Center of Calspan Corporation under the sponsorship
of the U. S. Environmental Protection Agency. This report covers the period
February 5, 1979 to April 4, 1980, and work was completed as of May 9, 1980.
This*report is identified as contractor's Report No. 6500-T-l.
111
-------
METRIC CONVERSION FACTORS
Approximate Conversions to Metric Measures
Symbol Whin You Know Multiply by To Find
LENGTH
Symbol
Symbol
in
ft
yd
mi
in2
ft2
yd2
mi2
oz
Ib
tsp
Tbsp
fl 02
c
pi
qi
gal
ft3
yd3
inches
feet
yards
miles
square inches
square feel
square yards
square miles
acres
ounces
pounds
short tons
(2000 Ib)
teaspoons
tablespoons
fluid ounces
cups
pints
quarts
gallons
cubic feet
cubic yards
•2.5
30
0.9
1.6
AREA
6.5
0.09
O.B
2.6
0.4
MASS (weight)
28
0.45
0.9
VOLUME
5
15
30
0.24
0.47
0.95
3.8
0.03
0.76
centimeters
centimeters
meters
kilometers
square centimeters
square meters
square meters
square kilometers
hectares
grams
kilograms
tonnes
ntilliliters
milliliters
milliliters
liters
liters
liters
liters
cubic meters
cubic meters
cm
cm
m
km
cm2
m2
m2
km2
ha
g
kg
ml
ml
ml
1
1
1
1
m3
m1
TEMPERATURE (exact)
°F
Fahrenheit
temperature
5/9 {after
subtracting
32)
Celsius
temperature
°C
en,2
m2
lun2
ha
9
kg
t
I
I
I
m3
Approximate Conversions from Metric Measures
Whin You Know Multiply by To Find
LENGTH
millimeters
centimeters
meters
meters
kilometers
0.04
0.4
3.3
1.1
0.6
AREA
square centimeters 0.16
square meters 1.2
square kilometers 0.4
hectares 110.000 m2) 2.5
MASS (weight)
grams
kilograms
tonnes (1000 kg)
0.035
2.2
1.1
VOLUME
milliliters
liters
liters
liters
cubic meters
cubic meters
0.03
2.1
1.06
0.26
35
1.3
TEMPERATURE (exact)
inches
inches
feet
yards
miles
square inches
square yards
square mi les
acres
ounces
pounds
short tons
fluid ounces
pints
quarts
gallons
cubic feet
cubic yards
Synibol
fl
yd
in*
v-2
fl oz
Pi
qt
$
yd3
*1 in = 2.54 (exactly). For other exact conversions and more detailed tables,
see NBS Misc. Publ. 286, Units of Weights and Measures, Price $2.25, SD
Catalog No. C13.10:286.
°c Celsius
9/5 (then
temperature add 32)
Fahrenheit °F
temperature
°F 32 98.6
-40 0
1 ' |' ' 1 ' ' 1 '
-40 -20
°C
40 00 1 120
1 1 ' 1 ' 1 ' fl 1
3 20 140
37
160
60
«F
212
200
I 1 » 1 ^
80 IOD
°
-------
CONTENTS
Page
Foreword ii
Abstract iii
Figures vi
Tables viii
Acknowledgment ix
I. Introduction 1
II. Conclusions 3
III. Recommendations 4
IV. Discussion 5
A. Preliminary Investigations 5
1. Selection of a Tire Sample and Test Conditions .... 5
2. The Particle Collection Chamber 7
B. Tire Wear Tests, Phase A 12
1. Test Details and Results 12
2. Discussion of Results 18
C. Tire Wear Tests, Phase B 19
1. Test Plan Details and Rationale 19
2. Test Results 21
a. Analysis of the Settled Particulates 21
b. Analysis of the Airborne Particulates 25
V. References 44
Appendices
A. An Estimate of Truck Tire Wear Particulate Characteristics . . 46
B. Validation of the SEM Technique for Identifying Rubber
Particulates 51
C. The Calspan Tire Research Facility . . 54
D. Test Tire Data and Tread Patterns 60
-------
FIGURES
Number Pa§e
1. Schematic Diagram of An Existing Shroud Assembly Modified for
Preliminary Flow Field Investigations on TIRF ................... 8
2. Photographs of the Modified Shroud Configuration as Installed
on the TIRF [[[ 10
3. Schematic Diagram of the Particle Collection Chamber and
Sampling Devices as Installed on TIRF ........................... 11
4. Particulate Collection Chamber Installed on TIRF ................ 13
5. Internal Modification to Particle Collection Chamber as Used for
Runs No. 4, 5 and 6 (Phase A Tests) . . ........................... - 16
6. Particle Size Distribution of Samples from the Floor of the
Collection Chamber .............................................. 23
7. SEM Photograph of a Rubber Particle from the Floor Sample of Run
No. 8, D78-14 Bias Tire and the Annotated X-ray Spectrum of the
Particle [[[ 27
8. SEM Photographs of Two Spherically-Shaped Rubber Particles from
the Floor Sample of Run No. 8, D78-14 Bias Tire ................. 28
9. SEM Field-View Photograph of Hi-Vol Filter Sample and Annotated,
Composite X-ray Spectrum; Radial Tire ........................... 29
10. SEM Photographs of Hi-Vol Filter Sample at High Magnifications;
Radial Tire [[[ 30
11. SEM Field-View Photograph of Hi-Vol Filter Sample and Annotated,
Composite X-ray Spectrum; Bias Tire ................. ............ 33
12., SEM Photographs of Hi-Vol Filter Sample at High Magnifications;
Bias Tire [[[ 34
13. SEM Field View Photograph of Cascade Impactor Filter Sample
(Stage 4) and Composite X-ray Spectrum; Radial Tire (DR78-14) ,
-------
FIGURES (Cont'd.)
Number Page
15. SEM Field View Photograph of Cascade Impactor Filter Sample
(Stage 6) and Annotated X-ray Spectrum for the Three Identified
Particles 41
16. SEM High-Magnification Photographs of Three Particles from
Stage 6 of the Cascade Impactor; Bias Tire Particulates 42
B-l SEM Photograph of a New Sample of Safety Walk and its
Annotated X-ray Spectrum 52
B-2 SEM Photograph of a Large Rubber Particle and its Annotated
X-ray Spectrum 53
C-l Calspan Tire Research Facility (TIRF) 55
C-2 Tire Research Machine r 56
D-l Pictorial View of the Test Tires Showing the Tread Patterns 61
VII
-------
TABLES
Number Page
1. Test Tire Sample and Test Conditions for Phase A 6
2. Size and Elemental Composition of 50 Tire Wear Particulates;
Floor Sample, Run No. 7, Radial Tire (DR78-14) 24
3. Size and Elemental Composition of 50 Tire Wear Particulates;
Floor Sample, Run No. 8, Bias Tire (D78-14) 26
4. Elemental Composition of Particles from the Hi-Vol Sample,
Radial Tire (DR78-14) 31
5. Elemental Composition of Particles from the Hi-Vol Sample,
Bias Tire (D78-14) 35
6. Elemental Composition of Particles from Stage 4 of the Cascade
Impactor; Radial Tire (DR78-14), Run No. 9 37
7. Elemental Composition of Particles from Stage 5 of the Cascade
Impactor; Bias Tire (D78-14) , Run No. 8 39
8. Weights of Tire Wear Particulates Collected During the Phase
B Tests 43
A-l Comparison of Selected Specifications and Ratings for a Sample
of Typical Automobile and Truck Tires 47
A-2 Average Daily Truck Usage in 11 Urban Areas 48
C-l TIRF Capabilities 57
C-2 Balance System Capability 57
Vlll
-------
ACKNOWLEDGMENT
The cooperation, active interest and technical support of this project
rendered by Messrs. I. Gusakov, E. J. Mack and R. J. Pilie is gratefully
acknowledged. Assistance in test operations was ably provided by Mr. A. J.
LaPres,.in charge of TIRF operations, and by Mr. G. A. Zigrossi who was
responsible for the particulate sampling instrumentation and who also performed
the particulate size analysis using optical microscopy. All of the analyses
conducted with the scanning electron microscope were performed by Mr. E. A.
Gasiecki.
IX
-------
SECTION I
INTRODUCTION
Estimates made of the quantity of tread rubber annually worn from vehicle
tires in the United States place the weight in the order of 5 x 10 kg (1, 2)*.
This debris represents a potential source of pollution especially with regard
to that fraction which remains airborne for an appreciable length of time.
Typically, this airborne category includes particles which are less than»20 ym
in diameter. Within this airborne fraction, the potential health hazard would
be greatest with particles less than about 15 ym in diameter since these are
able to penetrate and be retained in the human respiratory system. Thus two
considerations are involved: (a) the weight fraction of this tire debris that
is found in particulates of a respirable size and (b) the mutagenicity of these
particulates.
A large number of studies aimed at answering question (a) above have been
reported in the literature within the last decade. These efforts have included
passive-type of outdoor studies in which airborne particulates were sampled in
the proximity of busy thoroughfares and within the confines of tunnels (see
for example reference 2) as well as active-type of tests wherein particulates
were collected in the wake of tires on vehicles operated over various types
of roadways (see for example reference 3). In addition, indoor laboratory
studies have been performed to take advantage of the controlled test conditions
which can thereby be achieved in producing tire wear particulates (see for
example references 1 and 4). An excellent survey report (5) summarizes procedural
details and the test results obtained from various 'outdoor and indoor studies
related to the collection, identification and sizing of tire wear particulates.
In contrast to the extensive literature on the physical characteristics of
the wear particulates, there is a dearth of information on the possible mutagenic
properties of these same particulates.
A major criticism directed at laboratory-type tire wear programs has been
the inability to simulate real world conditions as they exist on the roadways.
To a large extent, the test apparatus and the experience in tread wear studies
available at Calspan tend to nullify many of these objections. The TIRF flat-
bed machine (6)** provides a flat, textured test surface similar to a type of
a real world roadway. Steady state and dynamic forces, torques, speeds and
steer angles can be applied to the test tire in a programmed manner. In addition,
tire tread wear schedules have been developed (7, 8) that can simulate a variety
of tread wear rates spanning the range from low to normal to severe with tire
tread surface textures representative of road worn tires.
* Numbers in parenthesis designate references listed at the end of the report.
** See also Appendix C.
-------
The test program described in this report was undertaken to collect and
characterize tire wear particles under simulated roadway conditions including
surface texture, skid number, vertical load, wheel torque, slip angle and speed.
Tests were performed on a low wear rate (radial ply) tire and a high wear rate
(bias ply) passenger car tire. The test effort was divided into two phases.
Phase A comprised steady-state tests of (a) straight-rolling tire operation
at 96 km/h (60 mph) and (b) cornering (constant slip angle) operation at 40
km/h (25 mph) for the purpose of determining the effect of test conditions on
types of wear particulates produced. Phase B tests were designed to simulate
normal wear conditions by operating the test tires at a constant speed of 88
km/h (55 mph) but with time-varying, simultaneous inputs of both slip angle
and wheel torque.
The tires were operated within a multichambered enclosure equipped with
two high-volume samplers ("hi-vols") and a single, six-stage cascade impactor.
Use of the enclosure served two important functions: (a) to provide effective
particulate separation, in a gross sense, by size through the incorporation of
a settling chamber and (b) to reduce the loss of wear particulates.
This report discusses the details of the instrumentation, the test procedures
and the results that were obtained. Particulate analysis was restricted to size
and material characterization. Bioassay testing of the particulates was per-
formed by the EPA, and the results thereof are outside of the scope of the
program described in this report.
Included within the scope of the program was a requirement to estimate the
characteristics of truck tire emissions as compared with emissions from passenger
car tires considering the differences in size, construction, tread rubber
compounding and vehicular use patterns. Results of this study, based on the
application of engineering judgment and the extrapolation of experimental data
for passenger car tires, are presented in Appendix A.
-------
SECTION II
CONCLUSIONS
The itemized conclusions that follow apply to the specific studies completed
during the course of this program. Experimental data were obtained for one
specific radial-ply tire (DR78-.14) and one specific bias-ply tire (D78-14) with
all the testing performed on the Calspan Tire Research Facility.
The more significant findings follow:
• Airborne rubber particulates (± 30 ym) were produced in extremely small
quantities irrespective of tire construction or the type and severity of the
imposed wear condition.
• Airborne particulates were mainly contaminant material with stainless
steel the chief constituent. These particles are generated by the operation
of the stainless steel belt used on TIRF.
• Tests on the bias tire produced a significantly smaller mass of airborne
particulates than did similar tests on the radial tire.
•• Based on the analysis of the floor samples in the collection chamber,
the bias tire produced a median particle size approximately twice that of the
radial tire with both tires operating under identical test conditions.
• There appears to be an absence of rubber particles in the approximate
size range between 1 urn and 10 ym.
• Rubber particle size distributions produced during TIRF tests (no
simulated road dust used) are like those reported for on-vehicle tire tests
performed on highways.
• The introduction of a roadway "dust" (cornstarch) decreased the mean
size of the very large, nonairbome rubber particles.
• An engineering assessment of expected truck tire wear emissions indicates
that, compared with passenger car tires, the wear rate will be comparable, with
the particle size distribution favoring the smaller particles.
-------
SECTION III
RECOMMENDATIONS
Reflecting insight and experience acquired in performing the experimental
studies described in this report, the following itemized recommendations are
offered as worthy of consideration in extending the scope of the present study.
Principally, the purpose of these suggestions is to confirm the relative scarcity
of the rubber particles in the very small size, airborne range by broadening
the range of test variables.
• Satisfactory simulated roadway surfaces, combining micro- and macro-
texture characteristics representative of actual road surfaces, should be
developed for TIRF.
• "Dusts", other than cornstarch, should be investigated which will be
compatible with TIRF operations and facilitate a workable method for separating
these dusts from the wear particulates.
• A larger sample of passenger car tires should be tested to explore the
effects, if any, of different tread compounds on the size distribution of wear
particulates. For similar reasons, different tread patterns should also be
tested including tires with the bulk of the tread rubber worn away.
• Tests of much longer duration should be planned to permit the collection
of larger samples in the particle size range below about 20 ym.
• The loss of very small particulates during a test is a potentially major
problem whose magnitude is not known. The use of tracer techniques to quantify
this problem should be explored.
• Effective methods for in situ separation of the stainless steel debris
from the rubber particulates should be investigated to facilitate size analyses
of the airborne rubber particles.
• The characterization of wear particulates produced by truck tires should
be attacked experimentally.
-------
SECTION IV
DISCUSSION
A. Preliminary Investigations
1. Selection of the Tire Sample and the Test Conditions
The specific tire sample selected for these purposes was one judged to be
representative of in-use automobile tires in terms of size, construction and
tread rubber compounding. Two tire constructions were chosen, the radial-ply
tire and the bias-ply tire. Because of construction differences that affect
cornering stiffness and tread "squirm", the radial tire is characterized by a
significantly lower tread wear rate and rolling resistance than an equivalent
size bias tire operated under the same test conditions. Rubber compounding is
also different between these two constructions with the radial tire using
considerably more natural rubber.
A DR78-14 tire (General Dual Steel II)* was selected in the radial con-
struction since it represents a popular size that is used extensively on current
model compacts. With the present down-sizing trends of cars, this tire size
should see continued, extensive .use. A D78-14 bias tire (General Jumbo 780}*
was selected to provide a high wear rate tire to contrast with the lower wear
rate radial. In the current passenger car tire market, the bias tire is sold
only as a replacement. A total of two radial tires and two bias tires was
purchased. One set was used for preliminary check tests, and one set was used
for the final test runs.
General test conditions were explicitly defined in the contractual statement
of work. Two types of tire wear conditions were stipulated: (a) steady-state,
highway-type of operation (no cornering) at a speed of 96 km/h (60 mph) and
(b) steady-state cornering at 40 km/h (25 mph) with the tires subjected to
vertical, longitudinal and lateral forces consistent with those experienced
in vehicular applications. In addition, it was required that for each of the
above two test conditions, a sufficient quantity of particulate debris, airborne
and nonairborne, be obtained to perform five Ames tests on each sample. An
individual sample of approximately one-half gram was estimated as adequate for
this purpose.
Table 1 summarizes, in a quantative manner, the details of the test program
that was designed to fulfill the requirements of the contractual work statement.
Tire test conditions were based on design data, conventional practices and test
data. Design load rating for the D78-/DR78-14 tires (load range B) is 4982N
(1120 Ib) at an inflation pressure of 165 kPa (24 psi). Since it is common
industry practice to operate tires at about 80% of design load, a test load of
3985N (896 Ib) was selected. Wheel torque figures were obtained by interpolating
* The choice of the manufacturer and tire brand was arbitrary.
-------
TABLE 1
Test Tire Sample and Test Conditions for Phase A
Test Tires
Item
Straight Rolling
Cornering
Bias Ply, D78-14*
Radial Ply, DR78-14**
Maker: General
Tread Design: Ribbed
Conventional Highway
Rim Size: 5.00 x 14
Inflation Pressure, Cold, kPa (psi)
Vertical Load, N (Ib)
Roadway Speed, km/h (mph)
Slip Angle, deg.
Camber, deg.
Wheel Torque***, N-m (ft-lb)
Test Duration (est.), min.
165 (24), Capped
3985 (896)
96 (60)
0
0
108 (80), D78-14
99 (73), DR78-14
120
165 (24), Capped
3985 (896)
40 (25)
3.0 (D78-14), 2.7 (DR78-14)
0
47 (35), D78-14
41 (30), DR78-14
90
*General Tire Jumbo 780
**General Tire Dual Steel II
***Wheel Torque Produced Driving Traction
-------
available test data on actual measured torques taken with vehicles operating
at steady state conditions on a test track (9). The lesser wheel torque
indicated for the radial tire reflects its lower rolling resistance relative
to the bias tire.
For the cornering tests, wheel slip-angle settings were calculated assuming
a 15940-N (3600 Ib = 4 x 896 Ib) vehicle operating in a 30.5-m (100 ft) radius
circle at a speed of 40 km/hr. A further approximation was made that each
vehicle tire contributed equally to the total side force required to maintain
the circular path. Typical tire cornering coefficients for the D78-/DR78-14
tires operating at specified load and pressure conditions (10) were used to
calculate the necessary slip angles. Since radial tires are characterized by
larger cornering stiffness coefficients, a smaller slip angle is required to
achieve a given lateral force for the radial tire than for the bias tire.
Test duration was estimated from the time required to produce one-half
gram of airborne (<30 ym) particulates for bioassay studies (Ames tests). Tire
wear rates were based on previous Calspan experience (7) and available data on
the relative fraction of the total mass of rubber particulates that constitute
the airborne material (1). Each test was designed to be performed on a dry,
flat roadway consisting of a medium-grit, Safety-Walk* surface stoned to a
wet skid number of 30. Each new tire was preconditioned by two hours of
operation at 80 km/h (50 mph) on a 1.2 m (4 ft) diameter drum at an equivalent
load.**
2. The Particulate Collection Chamber
The design and fabrication of an enclosure for the tire under test was
essential to the collection of the airborne and nonairborne wear particulates.
This enclosure had to fulfill three important functions: (a) reduce to minor
proportions the loss of the wear particulates, (b) provide effective separation
between particulates less than«30 ym and those larger than this size, and
(c) act as a mounting base for the particulate sampling instrumentation. To
provide a rational design basis for such a collection chamber, some simple tests
were performed to obtain data on the flow field in the proximity of a shrouded,
rotating tire operating on the moving TIRF roadway.
Preliminary flow survey studies were made using an available spare DR78-14
tire mounted on the TIRF machine and fully enclosed within an existing aluminum
shroud modified with an added splitter plate. This configuration is shown
schematically in Figure 1. To determine the details of the flow field, tests
were performed on the straight-rolling tire operating at 96 km/h (60 mph) and
an applied driving torque of 95N-m (70 ft-lb). Airflow velocities were measured
* 3M Company.
** The load required to produce the same tire deflection on a drum as that
experienced on a flat surface.
-------
TIRF
STEERABLETIRE
POSITIONING AND
LOADING SYSTEM
TIRE SHROUD
SPLITTER
PLATE
PARTIAL SHROUD
FLOOR
DENOTES LOCATIONS AT
WHICH FLOW VELOCITY
MEASUREMENTS WERE
MADE
Figure 1 SCHEMATIC DESIGN OF AN EXISTING SHROUD ASSEMBLY MODIFIED
FOR PRELIMINARY FLOW FIELD INVESTIGATIONS ON TIRF
-------
using a hot-wire anemometer that was inserted at selected locations within the
enclosed shroud and across the flow exit (see ® symbols in Figure 1). From
these measurements, approximate flow rates within the shroud annulus and
through the exit aperature were calculated. Small segments of a sticky-surfaced
paper were strategically placed on the inside surfaces of the shroud perimeter
and on the shroud floor at the exit to define visually the general particle
deposition patterns in the turbulent flow.* Arrows drawn in Figure 1 indicate
the direction of the main flow. Figure 2 shows photographs of the shroud/
splitter plate assembly as it was installed on the TIRF facility. Note (Fig. 2)
that there is no floor in the shroud forward of the tire contact patch.
From the flow field data and observations based on a cursory microscopic
analysis of the particulate debris on the adhesive papers, the induced flow
velocities were judged to be quite high, especially in the proximity of the
roadway and the tire surfaces. The corresponding induced volumetric flow
rates were found to be comparable to the anticipated sampling rates of the
particle collection devices to be used ultimately in the collection chamber
of the final enclosure design. Thus, it was possible that some finite fraction
of the particulate material could be entrained in the induced flow field and
lost either to the chamber walls by turbulent deposition or due to entrainment
in the boundary layer of the road surface beneath the floor of the chamber.
Based on the data of the flow field survey taken within the shroud, design
criteria were formulated and incorporated into the development of a particulate
collection chamber that is shown schematically in Figure 3. The chamber is a
single integral structure consisting of a tire shroud, the main collection
chamber and a secondary collection chamber. A splitter plate extends the width
of the shroud and functions to inhibit the entrainment of particulates within
the shroud by directing the induced flow into the main chamber. A floor panel
extends beneath the entire shroud and main chamber except for the aperture
necessary to permit the tire to contact the roadway.: A subfloor beneath the
main chamber floor provides access to the secondary chamber. Three-inch thick
aluminum hexcel honeycomb** extends across the entire areas of the main and
secondary chambers. Its purpose is to straighten the airflow and thus create
a uniform laminar flow ahead of the sampling devices. A detachable front panel
completely seals the assembly from the ambient air except at the tire contact-
patch aperture and at the inlet to the secondary chamber. The basic structural
materials are plywood and sheet metal. All nonmetallic inside wall surfaces
are covered with an adhesive^backed, light-gage aluminum foil to achieve smooth
surfaces that are not subject to electrostatic charging. Internal corners at
all critical locations are rounded to inhibit local flow stagnation and vortices,
both of which conditions are conducive to undesirable deposition of airborne
particulates.
* Extremely small particles would not impact, of course, but these were not
of interest in this purely visual assessment
** The individual hexcel measured approximately 9.5 nun (3/8 inch) across the flats.
-------
The Shroud/Splitter Plate
Assembly During a Test
(Note: The side panel is
clear plastic)
Close-Up View of the
Splitter Plate and
Partial Shroud Floor
Close-Up View of the Flow
Exit Area Surveyed with the
Hot-Wire Anemometer
Figure 2 PHOTOGRAPHS OF THE MODIFIED SHROUD CONFIGURATION AS INSTALLED
ON THE TIRF FACILITY AND USED FOR THE STUDY OF TIRE/ROADWAY
INDUCED AIRFLOW
-------
SECONDARY
COLLECTION
CHAMBER
TIRE SHROUD
\ , \ INLET TO
MAIN CHAMBER
BATTELLE6-STAGE .
CASCADEIMPACTOR
iTIRF
STEER ABLE TIRE
POSITIONING AND
LOADING SYSTEM -i
HIGH VOLUME
SAMPLER
NO. 2
VOLUME
SAMPLER NO. 1
'A —L
V \ ^ MAIN COLLI
-^-^-V-*^ CHAMBER
MAIN CHAMBER
FLOOR
INLET TO
SECONDARY
CHAMBER
SHROUD FLOOR
ALUMINUM
HEXCELL
FLOW
STRAIGHTENER
ROAD SURFACE
LOCATIONS FOR
FLOW VELOCITY
MEASUREMENTS
NECESSARY FOR
FINAL FLOW
ADJUSTMENTS
A INTERIOR WALLS
LINED WITH FILTER
PAPER
Figure 3 SCHEMATIC DIAGRAM OF THE PARTICLE COLLECTION CHAMBER
AND SAMPLING DEVICES AS INSTALLED ON TIRF
11
-------
The main chamber is approximately 0.46 m (18 in.) deep and 1.02 m (40 in.)
wide. A high volume sampler is located centrally at the top and accommodates
a 20.3 cm x 25.4 cm (8" x 10") filter upon which the particles are collected.
This particulate collection system is designed to acquire sub-30 ym particulates
using the principle that selective separation can be achieved, in accord with
Stokes' law, by the imposition of a suitable velocity on the particle-laden
flow. To establish the desired ambient flow velocity of approximately 3 cm/sec
in the upper regions of the chamber, it was calculated that a volumetric
pumping rate of 0.8 to 1.0 m3/min would be required. Flow rates were adjusted
in the course of checkout tests to optimize the collection efficiency of
particles with equivalent diameter .±50 ym. A six-stage cascade impactor
(Battelle) was mounted also atop the main chamber. Its function was to permit
a measurement of the particle size distribution for the sub-30 ym class of
particulates.
Preliminary flow field studies with the simple shroud suggested that a
finite fraction of the total partieulate debris from the tire would be lost
beneath the floor of the main chamber because of the relatively high flow
velocities in the tire/roadway boundary layer. Consequently, a secondary
chamber, equipped with a high volume sampler, was added for the purpose of
drawing off this flow (with its entrained particulates) from the road surface
as it expands into the inlet portion of this chamber. The volumetric flow rate
of this second high volume sampler was adjusted to equal the ram component of
the air entering beneath the floor of the main chamber. The addition of the
secondary chamber effectively reduced particulate losses and enhanced the
collection efficiency of small particulates within the main chamber by the
balancing of the flow rates of the high volume samplers.
A photograph of the completed particulate collection chamber and asso<
instrumentation, installed on the TIRF machine, is shown as Figure 4. The
entire side panel has been removed to show the internal configuration. During
test operations a gap of approximately 5 mm (0.2 in.) was maintained between
the roadway surface and the bottom of the chamber. Because the loaded radii
of the radial and the bias tire differed by approximately 7 mm (0.3 in.), the
clearance above the roadway had to be made adjustable to accommodate each tire
construction. Maintaining a minimum safe clearance was important in the
attempt to minimize the loss of the small, entrained particulates.
B. TJ.re-Wear Tests, Phase A
1. Test Details and Results
The basic test program was planned to consist of a total of four test runs
at the conditions shown in Table 1. These four runs were to be comprised of
straight-rolling and cornering tests on the bias and the radial tires. As a
first step, the rather large and bulky collection chamber was fitted to the
TIRF machine to check the mechanical clearances and the enclosure dynamics
related to steer angle motions, mechanical vibrations and resonances associated
with machine operation.
12
-------
HIGH VOLUME SAMPLER
BATTELLE 6-STAGE
CASCADE IMPACTOR
MAIN COLLECTION CHAMBER
SPLITTER PLATE
HIGH VOLUME SAMPLER
"""IB TIRE SHROUD
SECONDARY
COLLECTION
CHAMBER
ROAD SURFACE
_HEXCELL
FLOW
STRAIGHTENER
SECONDARY CHAMBER
INLET
MAIN CHAMBER FLOOR
Figure 4 PARTICULATE COLLECTION CHAMBER INSTALLED ON TIRF (SIDE PANEL REMOVED)
-------
At this time the collection chamber was fitted with a full side panel
(see Figure 4) of plexiglass to be used for checkout purposes only. This
panel was pretapped at various strategic locations (see Figure 3) to accommodate
a hot-wire anemometer probe to gather flow velocity data required to adjust
the high volume sampler flow rates. These measurements were made with the
roadway and the tire operating at the proper speeds but not in physical contact
with one another to avoid the possibility of damage to the fragile probe by
the larger particulates. The principal reason for the use of a transparent
side panel was to permit flow visualization studies to be made using photographic
techniques in conjunction with a "smoke" tracer. By injecting the tracer in
the area of the tire patch, it was hoped to follow the path of the tire wear
debris that was airborne. Despite the use of different types of chemical smoke
generators, the attempts at flow visualization were unsuccessful because of the
large turbulence in the flow field around the tire. Upon leaving the injector,
the smoke was almost instantly dissipated, becoming invisible.
A brief preliminary checkout run (run no. 1) was planned to ascertain
that the equipment was functioning properly and that the sub-30 ym particulates
were being collected. In addition, the mass rate of generation of the airborne
wear particulates was sought so that a reasonable estimate could be made of the
test times heeded to produce the necessary quantity of material for the bioassay
tests. For run no. 1, the high volume sampler flow rates were adjusted to
proper values and preweighed filters were installed in the two hi-vol samplers.
The collection chamber surfaces were lined with filter paper.
A project D78-14 bias tire was installed on TIRF, and a straight-rolling
(a = 0°) run was made for a period of one hour at a load of 3985N, a speed
of 96 km/h and a driving torque fo 95 N-m. Post-test examination of the filters
lining the chamber floors showed appreciable amounts of wear debris with many
pieces taking the form of irregularly-shaped rolls 3 to 9 mm in length. Some
strands of rubber were up to about 50 mm in length. Filters from each of the
samplers had a light grayish coating. As shown by microscopic analysis, the
particulates were in the proper size class (»2 to 30 ym). Most importantly,
the change in weight of the filters was negligibly small as measured on equip-
ment with a resolution of about 10 mg. Based on an estimated tread wear rate
for this condition, it was calculated that about 4 grams of rubber particles
should have been generated in one hour. If 5% of these rubber particles were
in the airborne category (2), a total of 0.2 grams should have been produced.
Eve% with a collection efficiency as low as 5% for each sampler, a change in
the weight of the filters should have been detectable. These results indicated
the impracticality of collecting a 0.5-gm sample of airborne particles using the
test methods of run no, 1.
The tread surface of the test tire had a coarse and tacky feel to the
touch. This condition is common to tires operated on steel drums/rolls and
also to many tires tested on the TIRF Safety-Walk roadway surface. In contrast,
most tires operated on outdoor concrete or asphalt roadways are characterized
by a tread surface texture that feels smooth and nontacky.
14
-------
In the effort to improve collection efficiency of small particulates,
run no. 2 was made with both of the high volume samplers operating at their
maximum flow rate. The Battelle impactor was installed for this run at the
top of the main collection chamber. All units were operated during the test
which was made at the same test conditions as used in run no. 1 except that
the test duration was reduced to one-half hour. The results in all respects
were identical to the first run. Because the cascade impactor uses teflon
filter material (rather than the fibrous silica filters used in the samplers),
it was convenient to analyze the composition of the small airborne particulates
with the SEM. These small particulates were found to be primarily stainless
steel.* Results from run no. 2 appeared to indicate that the samplers were
collecting such small particulates as were there to collect.
A logical concern arose concerning the possible deposition of the small
particles on the extremely large surface areas of the hexcell flow straighteners.
Run no. 3 was made as an exact repeat of run no. 2 except that the aluminum
honeycomb was removed. No discernible difference could be detected between
the quantity and type of particulates collected on runs no. 2 and 3.
As a next step in the quest for the elusive small particulates of rubber,
a test was made to collect as much of the debris as possible as soon as
it left the tire by attempting to achieve a test condition approaching isokinetic
sampling. For this purpose, the splitter plate was removed from the enclosure,
and a solid vertical wall was used to isolate the tire shroud from the main
chamber (Figures 4 and 5). A centrally-placed aperture was made in this
partition to accommodate the filter-tray assembly from a standard high volume
sampler. A rectangular converging nozzle was fitted to the filter tray. The
inlet to the nozzle was shaped .to provide a very close fit to the tire at
roadway/tire interface at the aft end of the tire contact patch. A flexible
hose was routed from the filter tray, through the side panel of the chamber
and to two high-volume blowers from an industrial vacuum system. While the
flow rate of this arrangement was not measured, it was known to be larger than
that of the standard high volume sampler operating at maximum capacity. Since
there was no flow into the main chamber, the only other sampler operating was
that on the secondary chamber. The bottom surface of the nozzle was lined with
filter paper.
Test run no. 4 used the exact same conditions as the two previous runs.
Post-test examination of the filter from the sampler equipped with the nozzle
showed little difference from prior tests except for a quantity of visibly larger
particles, estimated at about 100 ym. The quantity of the smaller particulates
on the filter had not increased appreciably, if at all. A sizable quantity of
wear debris had collected on the floor of the nozzle probably as the result of
the relative confinement of the material compared with the earlier configuration
of the collection chamber. At this time it was apparent that the tests were
not producing airborne particulates in quantities that would permit collection
* The TIRF belt is made of stainless steel.
15
-------
HOSE TO HIGH
VOLUME VACUUM
BLOWER
FULL WIDTH WALL
WITH APERTURE FOR
FILTER TRAY
FILTER TRAY
FROM HIGH VOLUME
SAMPLER
RECTANGULAR NOZZLE
EXTENSION TO FILTER
TRAY
Figure 5 INTERNAL MODIFICATION TO PARTICLE COLLECTION CHAMBER AS USED FOR
RUNS NO. 4, 5 AND 6 (PHASE A TESTS)
-------
of a one-half gram sample within a reasonable test duration. The fundamental
question was whether the absence of appreciable quantities (mass) of these
particulates was a normal "real world" condition or an artifact of the type of
tire tread wear that occurs on TIRF. More specifically, is the tacky tread
surface of the tire the cause of (or the result of) small wear particles
agglomerating to form larger-sized debris?
From experience with wear tests on TIRF (7) and other experimental data
(16), it had been known that small quantities of "dust" introduced into the
tire/roadway interface would result in tread surface textures that were smooth
and nontacky. Cornstarch was found to be an acceptable dust. Cognizant of
the adverse effect of the cornstarch on filter porosity, it was decided,
nevertheless, to repeat run no. 4 using a small quantity of cornstarch.* During
the one-half hour test, an estimated 0.11 kg (0.25 Ib) of cornstarch was used
in intermittent applications throughout the duration of the run. As expected,
the high volume sampler filter paper became uniformly coated with cornstarch
to a depth of about one to two millimeters. The cornstarch caught on the
filters had a grayish, discolored appearance apparently caused by entrapped
debris. Subsequent attempts to dissolve the cornstarch and filter the particu-
lates were not successful. The most striking observation concerned the debris
lying on the floors of the chamber. This debris appeared to be finely divided
into individual small particles and was characterized by the complete absence
of the large agglomerates observed on all previous runs. After only 30 minutes
of operation the tire tread surface had assumed a smooth and nontacky texture.
Despite a convincing demonstration that dust inhibits the formation of
large agglomerates, the question concerning the effect on the airborne fraction
was'not resolved.
As the final test run (no. 6) in this checkout series, wear particulates
were collected with the tire operating in a cornering mode. Using the same
enclosure configuration, instrumentation, load and torque as employed on runs
no. 4 and 5, only the speed and slip angle conditions were changed. Speed was
reduced to 40 km/h, and the slip angle was sinusoidally varied from +3° to -3°
at a frequency of 0.1 Hz. A varying slip angle was employed to avoid severe
asymmetrical wear of the tire since cornering represents a very high wear
rate condition. After one-half hour of test, the filters were examined and
found to have the same appearance as in all previous tests. The filters were
weighed, and again no detectable change in weight could be observed. Very
large quantities of rubber were deposited on the floor of the chamber. The
debris tended to be in the form of ragged strands some of which were up to
15 cm in length. These strands were found to be very elastic and could undergo
as much as 50% elongation before failing in tension.
Mean particle size about 10 ym.
17
-------
2. Discussion of Results
Upon completion of the Phase A checkout tests it was clear that a sufficient
quantity of airborne particulates required for the EPA Ames tests would not be
obtained if the planned test program (Table 1) were to be performed. Two
factors were involved: (i) tests of one-half to one hour failed to generate
a measurable weight of material, and (ii) such material as was collected was
predominantly contaminant and not rubber. The cost of extending the test duration
to the length necessary to collect the required weight of material would be
prohibitive.
A continuing major concern has always involved the ability of indoor tests
to duplicate the real world wear conditions of tires. The "real world" condition
is really a catch phrase with no precise meaning because of the breadth of the
ambient conditions involved. A brief reference to the published literature
will support this contention. Tire wear mechanisms are qualitatively discussed
in a survey article (11) which states that tread rubber characteristics and
roadway characteristics interact to create deformation, adhesion and abrasion
in the contact patch with both the micro- and macrostructure of the roadway
surface playing an important role. Studies have shown that average tire tread
life varies geographically over a range of almost four to one within the conti-
nental United States because of differences in highway surfaces and local
topography (winding, hilly roads contribute to increased tread wear) (12).
Road surface textures are also difficult to quantify.
Three basic mechanisms of tire wear have been identified; these are abrasive
wear, fatigue wear and thermal decomposition of the rubber. Briefly, abrasive
wear results from the cutting of tread rubber by sharp asperities and the
removal of the rubber by sliding action in the contact patch. Fatigue wear is
produced by a degradation of the tread rubber by .mechanical fatigue. This type
of wear frequently occurs under moderate wear conditions on smooth road surfaces.
The tread surface appears to be liquid-like and smeared (13). Thermal mechanisms
of tread wear are principally involved in locked-wheel braking conditions and
spinning-wheel situations in acceleration. Temperatures produced in these
circumstances may far exceed the decomposition temperature of the rubber.
Outdoor tests have been performed on specially-constructed tires to assess
the effects of tread rubber compounds, road surfaces and driving conditions on
the tread surface texture (13). It was determined that for tread rubber com-
pounded of styrene-butadiene rubber (SBR), polybutadiene (BR), and mixtures
thereof, all wear surfaces were similar for tests performed on very coarse
(gravel) roads. On smooth concrete surfaces, the BR tread surface was found to
be smooth with a fine granular texture. In contrast, SBR compounds were
characterized by tread surfaces with "liquid-to-crumbled-appearing surface
layers" thought to be caused by fatigue wear. Such tread surfaces have been
reported to be sticky or tacky (13).
The foregoing discussion illustrates how elusive a simulation of real world
tread wear conditions can be. It also shows that the sticky tread surface
18
-------
condition that is frequently found when tires are operated on steel drums,
rolls or coated steel belts (TIRF) is not unique to the laboratory environment.
The most promising method that could be readily implemented in changing the
type of tread wear produced in TIRF was that of modifying the roadway surface
characteristics. To avoid the tacky tread surfaces associated with roadway
surfaces characterized by a fine microstructure, it is necessary to introduce
a macrostructure as well. To this end, a conformable-type of Safety-Walk*
material was obtained and applied over a very coarse open-coat abrasive comprised
of large, blocky grits. The result of this laminated combination was a surface
that combined a micro- and macrostructure that resembled asphalt.
To test the durability of this laminated roadway covering and to determine
the type of tire tread surface that would be produced by a tire operating upon
it, some simple tests were made on a 1.2 m (4 ft) drum. A bias tire was operated
under load on this drum whose test surface was covered with the laminated coating,
After several hours of operation at 80 km/h, the tire surface appeared slightly
abraded, nontacky .and free of any buildup of rubber. The conformable Safety-
Walk appeared to be intact. Based on this encouraging result, a revised test
schedule was formulated and forwarded to the project officer for approval. The
next section discusses the details of the Phase B test effort.
C. Tire Wear Tests, Phase B
1. Test Plan Details and Rationale
The Phase B test plan differed in two major respects from the initial
schedule as summarized in Table 1; namely, the tire wear schedule and the roadway
surface characteristics.
Previous studies concerned with tread wear of passenger car tires on the
TIRF machine (7) had resulted in the development of tire wear schedules that
produced controlled wear rates combined with a uniform wear across the tire
tread surface. Using cornstarch (not practical in the current program), the
worn tire surfaces were judged to be very representative of "road worn" tires.
The test conditions selected on this basis are listed below.
pressure: 165 kPa (24 psi) cold, capped
vertical load: 3985N (896 Ib)
roadway speed: 88 km/h (55 mph)
slip angle: • 0.3° rms, random (Gaussian) amplitude distribution
• period of randomness: 22.7 minutes
* This 3-M product is made in the same surface texture grades as regular Safety-
Walk but uses an aluminum foil backing treated with contact cement.
19
-------
camber angle: 0°
wheel torque: • ±68 N-m (±50 ft-lb) sinusoidally varying
• frequency of input = 0.1 Hz
test duration: three hours
Both the bias and the radial tire were to be subjected to the same test schedule
which was estimated to produce wear rates of about 0.025 mm/km (16 mils/1000
miles) for the former and 0.016 mm/km (10 mils/1000 miles) for the latter.
Tires previously unused under test conditions, but broken-in, were employed for
these tests.
Despite the promising test results obtained with the use of the laminated
roadway surface covering (a conformable Safety-Walk overlay on a coarse, abrasive
paper) on the drum, the feasibility of its use on the TIRF machine remained
uncertain. In the TIRF application these materials would be subjected to
repetitive flexing and sizable longitudinal/lateral forces that did not exist
during the drum tests. Also, the conformable material uses a contact cement
as the adhesive. If the surface integrity of the Safety-Walk were to be com-
promised during testing, the collected sample of debris would be contaminated
with this adhesive which contains known carcinogens. The purpose of the test
would then be nullified. Nevertheless, the decision was made to conduct tests
on the laminated roadway surface material with the backup option of reverting
to the standard Safety-Walk if necessary.
For these tests the test chamber was returned to its normal configuration
as shown in Figure 4. It was thoroughly cleaned and reinstalled on the TIRF
machine. The TIRF roadway was partially covered* with the experimental, laminated
surface.
The first test run was performed on the radial tire. After three hours of
test, the filters from the high volume samplers were generously covered with
debris. Debris that had settled to the floor of the chamber was in the shape
of small granular particles. Tire tread surfaces appeared slightly abraded
but tactilely smooth and nontacky. A slight accretion of rubber was observed
on one of the outer ribs of the tread.
This run was considered to be a checkout test of the roadway surface. While
the composition of the airborne sample was being analyzed with the SEM, a second
test ru.n was started. A portion of the conformable Safety-Walk surface separated
from the TIRF belt after approximately 15 minutes. In the effort to improve
adhesion, a length of conformable Safety-Walk extending the full width of the
steel belt was applied over the 30 cm strip of abrasive underlayer- The adhesion
problem was not resolved by this expedient. Also, the analysis of the particu-
lates showed that the airborne fraction consisted mainly of stainless steel,
aluminum, silicon and contact cement. Clearly, the roadway surface was being
degraded by the loss of entire grits together with the foil backing and the cement
Samples thus were unacceptably contaminated. Consequently, it was necessary
A continuous, circumferential strip, 30 cm wide, was centrally placed on the
71 cm wide TIRF belt.
20
-------
to return to the standard Safety-Walk material to generate the particulate
samples to be used in the EPA bioassay. Nevertheless, caution must be
exercised in the interpretation of mutagenicity test data due to the possible
contamination of the rubber samples by the adhesive used with standard
Safety-Walk as well.
After the collection chamber had been thoroughly cleaned again and the
TIRF belt resurfaced with standard Safety-Walk, three tests of three hours
duration each were performed; one on the bias tire and two on the radial tire.
2. Test Results
This section summarizes the results of analyses performed on tire wear
particulates collected during the final three test runs made on this program;
runs no. 7, 8, 9. Runs no. 7 and 8 were made on the DR78-14 and D78-14 tires,
respectively. Since the total mass of particulates deposited on the floor of
the enclosure during run no. 7 was judged to be marginal, run no. 9 was made on
the same radial tire as a repeat of run no. 7 to collect additional particulates
for the bioassay study. The presentation of the data is divided into two parts;
the first part being devoted to analysis of the nonairborne particulate samples
which deposited on the flat, horizontal surfaces (floors) of the chamber and
the second part concerned with the airborne particulate samples collected by the
high volume samplers and the cascade impactor. Separate treatment is accorded
to the samples from the radial and the bias tires.
a. Analysis of the Settled Particulates - (Floor Samples)
Measurements were made of the particle sizes and their frequency of occurrence
for the wear debris which had settled on the filter paper* lining the floor of
the main settling chamber. Representative filter sections were physically re-
moved from the sheet of filter paper for analyses. Photographs of the samples
were taken with the aid of an optical microscope at various levels of magnifi-
cation to accommodate particulates in the size range below about 300 ym. The
eyepiece of the microscope was fitted with a graduated reticle to permit measure-
ment of particle sizes. Particulate shapes were observed to be very irregular
and characterized by a structure that appeared to be very porous (sponge-like).
Subsequent analysis identified all these larger-sized particles as consisting
essentially of rubber. Particle sizes were quantitively measured from photo-
micrographs of the samples, with the size being given in terms of the long
dimension of the particle. For each of the three test runs, several hundred
particles were sized in this manner to assure reasonable statistics of counting.
The results were then presented in terms of the percentage of the total particle
count that was found within a given particle size range. The size range to
80 ym was divided into fractions, each with a width of 10 ym. The size range
* Pallflex fiberglass filter paper (No, T60A20), a teflon-coated paper, was also
used in the high volume samplers.
21
-------
from 80 ym to 300 ym encompassed 11 fractions, each with a width of 20 ym.
Using the midpoint-size value of each fraction as the ordinate, the data are
summarized in Figure 6 as plots on probability paper where the abscissa is
cumulative probability.
The two runs on the same radial tire show slightly different distributions
which probably are the result of the modest numbers of particles constituting
these sized samples. An alternative possibility is that rubber buildup on
the tread surface with mileage resulted in larger particulates being produced
during the second run (run no. 9) on the radial tire. A bimodal distribution
is found for both samples with a sharp transition occurring at a particle size
of 50 ym. Median particle sizes for these two distributions lie in the approxi-
mate range of 25 ym to 35 ym. A different distribution was observed for the
wear particulates produced by the bias tire. The distribution is curvilinear,
in contrast to the stright-line segments that characterize the distribution
data for the radial tire. Also the bias tire is found to produce larger
particles with the median size at about 55 ym. Median particle sizes shown
in Figure 6 compare favorably with values of 20 ym to 30 ym reported for test
samples collected behind a tire on a vehicle operating in a normal mode on a
concrete highway (3). When the collection system was redesigned to improve
the collection efficiency for the smaller particles, the median size values
dropped to the 10 ym to 15 ym range (3).
In the TIRF tests, the enclosure is designed to separate the smaller
particles by keeping them airborne and hence capable of capture by the high
volume samplers and the cascade impactor. Thus, the floor sample will be
deficient in these smaller particles, and the data of Figure 6 must be viewed
as applicable only to the floor sample of wear particulates and the specific
circumstances and constraints attendant to their collection. Nevertheless,
the agreement between TIRF data and the data from reference 3 (with the
inefficient collection of small particles) strongly indicates that wear par-
ticulate size distributions produced on TIRF are not unlike those produced
during normal vehicle operation on highways. It also follows that the bimodal
distribution observed for the radial tire may be an artifact of the effect of
the test environment on the collection of the sample.
Using the same sections of filter paper employed for the particle-size
measurements discussed above, an analysis of a representative sample of the
particles was made for composition using scanning electron microscope (SEM)
techniques. A description and validation of these techniques is presented in
Appendix B. Basically, the elements comprising the particle(s) are identified
by their characteristic x-ray spectra. Since hydrocarbons cannot be identified
by this method, rubber is detected by the simultaneous appearance of energy
peaks corresponding to sulfur (S) and zinc (Zn) which are constituents of tread
rubber compounds. Table 2 lists the size (length by width or diameter) of 50
randomly selected particles together with the elemental composition of each "
particle from the floor sample generated by the radial tire. Of the 50 particles
examined, 92% were rubber with 65% of these also containing other debris. Of
the contaminated particulates 41% contained iron (probably stainless steel).
22
-------
cc
LU
111
O
H
OC
300
200
100
90
80
70
60
50
40
30
20
f
}
1
j
f /
U
/
/
/
/
//
f
?
i
£
tf
i
•'/
0
jl
6
{
r
ME[
,tf
/
i
o
HAIS
r./
Ef
A
/
>
i
/
X
^
^
r~
0X
S
fs''
,
i \
J
s
rf
s(
Y
$
/
P
/
-d
O— — O RUN NO. 7, DR78-14 TIRE,
TOTAL PARTICLE COUNT =
CD D RUN NO. 8, D78-14 TIRE,
TOTAL PARTICLE COUNT =
£ A RUN NO. 9, DR78-14 TIRE,
TOTAL PARTICLE COUNT =
= 233
- 378
= 261
i ii
10
9
8
7
6
0.5 1
5 10 20 30 40 50 60 70 80 90
CUMULATIVE PROBABILITY, %
95
98 99
99.8 99.9
Figure 6
PARTICLE SIZE DISTRIBUTION OF SAMPLES FROM THE FLOOR OF THE COLLECTION CHAMBER
(Approximately 90% of these Particles Contained Rubber)
-------
TABLE 2
*
Size and Elemental Composition of 50 Tire Wear Particulates;
Floor Sample Run No. 7, Radial Tire (DR78-14)
Particle
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Particle
Size
L x W
(ym)
30 x 40
30 x 100
15 x 40
40 x 80
30 x 40
20 x 80
40 x 80
30 x 50
15 x 30
30 x 150
15 x 20
25 x 50
8 x 30
8 x 110
40 x 120
100 x 130
100 x 150
30 x 70
50 x 60
100 x 150
50 Dia.
120 x 180
15 x 20
220 x 250
100 x 150
Particle Identification:
[Elemental Composition)
Rubber
x
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Fe
X
X
X
X
X
X
Ca
X
X
X
X
X
X
X
X
X
X
X
K
X
X
Cl
X
X
X
Al
X
X
X
X
MR
x
Na'
x
Other
x
Particle
No.
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Particle
Size
L x W
(ym)
50 x 60
150 x 180
50 x 50
15 x 35
25 x 28
25 x 55
30 x 38
160 x 100
70 x 80
120 x 200
50 x 60
60 x 80
40 x 100
90 x 120
20 x 80
25 x 30
50 x 70
40 x 50
15 x 30
50 x 80
20 x 80
80 x 100
50 x 80
30 x 80
50 x 100
Particle Identification:
(Elemental Composition)
Rubber'
x
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Fe
X
X
X
X
X
X
X
X
X
X
X
X
X
Ca
x
x
x
X
X
X
X
X
X
X
X
X
X
K
X
X
X
X
X
X
X
X
Cl
X
X
X
X
X
X
X
X
Al
X
X
Mg
X
X
X
Na
x
Other
x
X
X
K)
-fi.
Fe = Iron,
Ca = Calcium
K = Potassium
Cl = Chlorine
Al = Aluminum
Mg = Magnesium
Na - Sodium
-------
Table 3 shows similar data for 50 particles examined from the floor sample
generated by the bias tire. Here 90% of the particles were rubber with 24%
of these containing contaminants. Iron appears less frequently as a rubber
contaminant (~9%) than in the case of the radial tire sample. In both samples,
tire particulates appear to be blocky or chunky in shape rather than elongated.
For example,~ 75% of the particles generated by the radial tire and«*60% of
those by the bias tire, have a length-to-width ratio of <2.5. A few spherical
particles were also detected. Differences in particle morphology are illustrated
in Figures 7 and 8.
Figure 7a is a photograph of an irregularly shaped rubber particle,
approximately 95 ym in length,* that appears to be an agglomerate. Figure 7b
is an annotated x-ray spectrum of the particle showing energy peaks correspond-
ing to S, Zn, silicon (Si), copper (Cu), gold (Au), and palladium (Pd). The
last two elements originate from the conductive coating that must be applied
to electrically nonconducting specimens (see Appendix B). Figures 8a and 8b
illustrate smooth-surfaced, spherically-shaped rubber particles with diameters
of «15 pm and «50 ym respectively.
b. Analysis of the Airborne Particulates
Airborne particulates, in the context of this section, denote particulates
that were collected by the high volume samplers as well as by the cascade
impactor. In analyzing the particulates from the hi-vol units, the choice was
made to use the samples from the hi-vol located in the secondary chamber since
it collected a larger range of particle sizes, by virtue of chamber design,
than the unit operating in the main chamber. A gross field (30X) SEM photograph
is shown in Figure 9a to illustrate the nature of the particulates collected
by the hi-vol samplers during the tests on the radial tire. The x-ray spectrum
obtained from a scan of the field of Figure 9a is shown in Figure 9b where the
various identifiable elements present are noted. Portions of this field are
shown at magnifications of 300X and 3000X (Figures lOa and lOb, respectively)
to better define the nature of the particles in the«l pn size range. The fibrous
structure of the fiberglass filter paper is clearly a limiting factor in locating
and analyzing particles in the submicron size range. Results of an analysis
of the individual particulates are summarized in Table 4. The larger particu-
lates, individually identified by a number that locates each on either Figure 9a
or lOa, are listed in Table 4A and are seen to be principally rubber particles.
Table 4B lists the elemental composition of the smaller particles that are
identified by size only. These particles are seen to consist chiefly of
contaminants with only a few rubber particles identified.
Rubber particles, in the main, are seen to be coarse, irregular and
sponge-like. They are neither smooth nor cigar-shaped and thus appear to be
abraded from the tire by a scraping action.
* All SEM photographs in this report have a size scale (a horizontal line
terminated by two short vertical bars) at the bottom margin. To the left
of this scale is the numerical value, in micrometers, of the line length.
25
-------
TABLE 3
Size and Elemental Composition of 50 Tire Wear Participates;
Floor Sample, Run No. 8, Bias Tire (D78-14)
Particle
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Particle
Size
L x W
Cym)
120 x 120
40 x 90
10 x 20
150 x 250
20 x 100
50 x 120
150 x 300
10 x 12
100 x 150
200 x 400
20 x 100
50 x 150
30 x 55
25 x 40
150 x 300
30 x 80
20 x 50
30 x 50
80 x 200
40 x 50
30 x 45
12 x 20 „
200 x 300
20 x 100
100 x 500
Particle Identification:
(Elemental Composition)
Rubber
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Zn
X
Fe
x
X
X
X
Ca
x
X
K
x
X
Cl
x
Al
x
X
X
X
Na
x
Other
x
Particle
No.
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Particle
Size
L x W
(ym)
100 x 100
80 x 100
50 x 100
70 x 120
100 x 250
50 x 150
20 x 50
70 x 130
70 x 250
12 x 12
20 x 100
40 x 50
150 x 500
150 x 250
90 x 125
5x8
110 x 500
40 x 80
100 x 200
30 x 250
10 x 30
16 ym Dia.
20 x 18
30 x 40
14 ym Dia.
Particle Identification:
(Elemental Composition)
Rubber
x
x
x
x
x
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Zn
X
Fe
x
Ca
. x
X
X
K
X
x
X
X
Cl
X
Al
X
X
X
X
Na
Other
Fe = Iron
Ca = Calcium
K = Potassium
Cl = Chlorine
Al = Aluminum
Zn - Zinc
Na - Sodium
-------
(a)
MAGNIFICATION = 900X
Figure 7
Cu
SEM PHOTOGRAPH OF A RUBBER PARTICLE FROM THE FLOOR SAMPLE OF
RUN NO. 8, D78-14 BIAS TIRE AND THE ANNOTATED X-RAY SPECTRUM OF
THE PARTICLE
27
-------
MAGNIFICATION = 1000X
(a)
MAGNIFICATION = 1000X
MAGNIFICATION = 2500X
(b)
MAGNIFICATION = 2500X
Figure 8 SEM PHOTOGRAPHS OF TWO SPHERICALLY-SHAPED RUBBER PARTICLES FROM
THE FLOOR SAMPLE OF RUN NO. 8, D78-14 BIAS TIRE
"
-------
I
1
6-
(a)
If 9
-2,3
-4
~ 5
1
Na
Mg J
Al —
Si
P
S
d
II II
Ca Ba pe Cu
K
L
Au
Zn
(b)
MAGNIFICATION = 30X
Figure 9 SEM FIELD-VIEW PHOTOGRAPH OF HI-VOL FILTER SAMPLE AND ANNOTATED,
COMPOSITE X-RAY SPECTRUM; RADIAL TIRE (DR78-14)
-------
8,9 11
7 10
.
:
010(0 u
03-? i?0,fl H 000 0/5
MAGNIFICATION = 300X
-10
(a)
MAGNIFICATION = 3000X
Figure 10 SEM PHOTOGRAPHS OF HI-VOL FILTER SAMPLE AT HIGH MAGNIFICATIONS;
RADIAL TIRE (DR78-14)
-------
Particle
No.
1
2
3
4
5
6
7
8
9
10
11
Rubber
X
X
X
X
X
X
X
X
X
Si
X
Organic;
No Spect-
rum"
X
SEM
Photo
Magnif .
3ox:
CFig.9a)
300X
(Fig.lOa)
(A)
Particle
I.D. by
Size, urn
IS x 20
3x8
'v/dia.
4x5
20 x 30
25 x 30
5 x 30
4x6
5x7
2x3
2x5
20 x 30
10 x 12
2x3
4x5
4x5
Rubber
x
X
X
Zn
x
Ca
x
Fe
x
X
X
X
X
Cl
X
Si
x
Al
Y
Mg
x
Stain-
less
x
No
Spect-
rum
x
X
SEM
Photo
Magnif.
3000X
(Fig.lOb)
TABLE 4
Elemental Composition of Particles From the Hi-Vol Filter
Sample, Radial:. Tire (DR78-14) (Ref. Figures 9a, lOa and lOb)
31
-------
An analysis identical to that described above for the radial tire, was
performed for the hi-vol sample obtained for test run no. 8-performed on the
bias tire. The data are shown in Figures 11, 12 and Table 5 which are counter-
parts to Figures 9, 10 and Table 4, respectively. In general, the same obser*
vations and comments apply to both sets of data. Table 5 shows, as did Table 4,
that the larger particles are primarily rubber (°r rubber plus contaminants)
while the smaller particulates are mostly contaminants. In fact, none of the
small particles examined in Figure 12b contained any rubber.
A six-stage Battelle cascade impactor sampled particulates from within
the main collection chamber while operating at a volumetric flow rate of
12.5 liters per minute. Instrument specifications are tabulated below:
Particle Mean
Stage No. Cut-off Diameter, ym
1 16
2 8
3 4
4 2
5 1
6 0.5
with the mean cut-off diameter defined for spherical particles of unit density*
wherein 50% will impact a given stage and 50% will continue to a succeeding
stage. Millipore Fluoropore filters (teflon) were used to collect the particulates
on each of the six impactor stages.
A representative collection of particulates collected on the 4th stage
filter during run no. 9 on the radial tire is shown in Figure 13a. The
particles are in the approximate 1 ym to 7 ym range, and, as indicated by the
composite x-ray spectrum of Figure 13b, consist chiefly of contaminants.
Table 6 summarizes the composition of 18 specific particles that are identified
(by number) on Figure 13a. The overwhelming majority of the particles contain
stainless steel.**
Figure 14a presents a collection of particles from the 5th stage of the
impactor in run no. 8 (the bias tire). Figure 14b is the composite x-ray
spectrum of the field shown in Figure 14a. Particle sizes are in the 1 ym to
3 ym*range and again contain no identifiable rubber particles. Of 17 particles
analyzed for elemental content (see Figure 14a and Table 7) it is seen that all
but one contained stainless steel.
* Tread rubber density is about 1.18 (3).
** Particles whose x-ray spectra showed both iron and chromium peaks were
classified as stainless steel.
32
-------
2 —
6
7
— 5
O
(a)
Na
<
1
•*
p
h
d
:
c
B
a
a C
r F
e C
j
Z
Si
(b)
MAGNIFICATION = SOX
Figure 11 SEM FIELD-VIEW PHOTOGRAPH OF HI-VOL FILTER SAMPLE AND ANNOTATED,
COMPOSITE X-RAY SPECTRUM; BIAS TIRE (D78-14)
-------
12
11
14
-15
,
OIOiO u M
03-? ao.o n ooo o^o
I
15
14 16
MAGNIFICATION = 300X
I
17
16
17
(a)
MAGNIFICATION = 3000X
(b)
Figure 12 SEM PHOTOGRAPHS OF HI-VOL FILTER SAMPLE AT HIGH MAGNIFICATIONS;
BIAS TIRE (D78-14)
-------
Particle
No.
1
2
3
4
h 5 -\
6
I — -
7
H- §
H
9
10
11
12
13
14
IS
16
17
lubber
X
X
X
X
X
X
X
X
X
X
x
X
X
Ca
X
X
X
Fe
X
X
X
X
Ti
X
Si
X
X
Al
X
Organic,
No
Spectrum
X
X
. ._.
Other
.,
.._ , f
1
'
,
'
'
1
-
SEM
Photo
Magnif .
30 X
(Fig.Ha)!
j
j
|
j
, 300X
(Fig.l2a)|
i
(A)
''article
[.D. by
5izej. _nm
2x3
3x5
3x5
4x8
4x5
2x3
3x5
8 x 15
4x5
Rubber
Fe
x
Zn
x
x
X
Al/Si
x
Gyp-
sum
x
Stain-
less
x
x
x
SEM
3hoto
^lagnif .
^
j
i
3000 X
(Fig.l2b)j
!
\
i
i
TABLE 5
Elemental Composition of Particles From the Hi-Vol
Filter Sample, Bias Tire (D78-14), (Ref. Figures
lla, 12a and 12b)
(B)
35
-------
11,14
(a)
MAGNIFICATION = 2000X
Ci
Fe Ni
(b)
Figure 13 SEM FIELD VIEW PHOTOGRAPH OF CASCADE IMPACTOR FILTER SAMPLE
(STAGE 4) AND COMPOSITE ANNOTATED X-RAY SPECTRUM;
RADIAL TIRE (DR78-14), RUN NO. 9
36
-------
Darticle
Number
1
2
3
4
5
6
•j
8
9
10
11
12
13
14
15
16
17
18
Rubber
Fe
X
X
X
Ca
X
X
X
Cl
X
X
55
X
X
X
X
Si
X
X
x
X
X
X
X
X
X
X
Al
X
X
X
X
X
X
X
X
MS
X
X
X
Na
X
Stain-
less
Steel
X
X
X
X
X
X
X
X
X
X
X
X
Other
X
X
TABLE 6
Elemental Composition of Particles From Stage 4
of the Cascade Impactor;
Radial Tire (DR78-14), Run No. 9 (Ref. Fig. 13)
37
-------
8,10
MAGNIFICATION = 3000X
(b)
Figure 14 SEM FIELD VIEW PHOTOGRAPH OF CASCADE IMPACTOR FILTER SAMPLE
(STAGE 5) AND COMPOSITE ANNOTATED X-RAY SPECTRUM;
BIAS TIRE (D78-14), RUN NO. 8
-------
'article
Number
!
2
•3
4
5
6
7
8
9
10
11
12
13
14
15
16
'17i
lubber
Zn
X
Cu
X
Fe
X
C,a
X
Cl
X
X
S
X
X
Si
X
X
X
X
X
X
Al
X
X
X
X
X
Na ,
X
stain-
less |
5teel
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
x
TABLE 7
Elemental Composition of Particles from Stage 5
of the Cascade Impactor;
Bias Tire (D78-14), Run No. 8 (Ref. Fig. 14)
39
-------
A careful examination of the high magnification SEM photographs of the
impactor samples, see Figure 13a and 14a for example, shows a considerable
number of tiny, regularly-shaped particles. To ascertain the nature of such
particles, the 6th stage sample was analyzed from a run on the bias tire.
Figure 15a depicts a sampling of particles in a size range £ 1 ym. The x-ray
spectrum (Figure 15b) of three identified particles shows the presence of
sulfur and zinc peaks which classify these particles as rubber. Figure 16
consists of high-magnification (20,OOOX) photographs of the three particulates
which show smooth, regular shapes in contrast to the morphology of the larger
particulates. These results indicate an absence of rubber particles in the
approximate size range from 1 ym to 10 ym but their presence on either side
of this size interval. This finding is in agreement with results reported by
General Motors on particle emissions from tires (1) but conflicts with the
work of Pierson at Ford (2).
Filters installed in the two hi-vol samplers and the Battelle impactor
were weighed before and after a test run in the attempt to determine the
weight of the collected particulates. A summary of the results obtained is
presented in Table 8. Approximately four times the mass of particulates was
collected on the hi-vol filters for the radial tire as for the bias tire.
Thus, a lesser portion of the wear particulates generated by the bias tire
was airborne than for the radial tire.
The weight data for the filters from the Battelle impactor are considered
to be within the noise level of the measurement capability. For example, one
of the filters was measured as having decreased on weight by 0.4 mg after a
test run. It is clear that considerably longer duration test times would be
required to collect accurately measurable weights of particulates on the impactor
stages. Due to an oversight created by illness to key technical personnel
during these final tests, measurements of the total weight of rubber removed
for each run were not made, and hence data on the wear rates applicable to these
runs are not available.
40
-------
(a)
MAGNIFICATION = 2000X
(b)
Zn
Cu
Figure 15 SEM FIELD VIEW PHOTOGRAPH OF CASCADE IMPACTOR FILTER SAMPLE (STAGE 6)
AND ANNOTATED X-RAY SPECTRUM FOR THE THREE IDENTIFIED PARTICLES
! !
-------
PARTICLE 1
PARTICLE 2
PARTICLE 3
MAGNIFICATION = 20,OOOX
MAGNIFICATION = 20,OOOX
MAGNIFICATION = 20.000X
Figure 16 SEM HIGH-MAGNIFICATION PHOTOGRAPHS OF THREE PARTICLES FROM STAGE 6 OF
THE CASCADE IMPACTOR; BIAS TIRE PARTICULATES
-------
TABLE 8
Weights of Tire Wear Particulates Collected During the Phase B Tests
Run No .
7
8
9
Tire
DR78-14
(Tire No. 3)
D78-14
(Tire No. 2)
DR78-14
(Tire No. 3)
Test Details
Velocity: 88 km/h
Load: 3985N
Slip Angle: *
Camber: 0°
Torque : **
Pressure: 165 kPa (cap)
Duration: 3 hours
Same as Run 7
Same as Run 7
Hi -Vol.. No. 1
Am , g
0.32
0.04
0.26
Hi-Vo^. No. 2
Am , g
0.14
0.08
0.22
Battelle Impactor
Stage No. Am"*", g
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
0.0010
0.0006
0.0009
0.0005
0.0011
0.0009
0.0008
0.0009
0
0.0002
0.0002
0.0006
-0.0004
0.0003
0
0.0005
0.0006
0.0001
Note: Weights of particulates shown above do not represent total wear. Wear rates were not determined for
these tests. *0.3° RMS With a Gaussian Amplitude Distribution
**+68N-m Sinusoidally Varying, Frequency = 0.1 Hz
t Am = Difference in Post Test and Pretest Weight of Filter
-------
REFERENCES
1. Cadle, S. H. and R. L. Williams. Gas and Particle Emissions From
Automobile Tires in Laboratory and Field Studies. GM Research
Publication GMR-2542R, May 1978.
2.* Pierson, W. R. and W. W. Brachaczek. Airborne Particulate Debris From
Rubber Tires. Rubber Chemistry and Technology, 47 (5): 1275-1299,
December 1974.
3. Dannis, M. L. Rubber Dust From the Normal Wear of Tires. Rubber
Chemistry and Technology, 47 (4): 1011-1037, September 1974.
4. Evans, J. V. and R. D. Hites. Emissions of Pollutants From Tire Wear
Phase I Preliminary Report M71-166 Project 6132, October 1971.
5. Evaluation of Particulate Emission Factors for Vehicle Tire Wear. PEDCo
Environmental, Inc., Draft Report Copy. EPA Contract 68-02-2585, January
1979.
6. Bird, K. D. and J. F. Martin. The Calspan Tire Research Facility:
Design Development and Initial Test Results. SAE Paper 730582, 1973.
7. Bogdan, L. and I. Gusakov. An Evaluation of a Laboratory Technique for
Measuring Tread Wear on Passenger Car Tires for the RMA. Final Report,
Calspan Report No. ZM-5974-T-1, May 1977.
8. Gusakov, I., G. A. Tapia and L. Bogdan. Development of Short Duration
Laboratory Test Techniques and Instrumentation of Evaluation of Tread
Wear of Military Truck Tires. Final Report, TARADCOM Technical Report
No. 12308, January 1978.
9. Bogdan, L. Comparison Between a Double-Roll and a Flat-Bed Dynamometer
Configuration in Regards to Passenger Car Tire Power Consumption,
Fuel Economy and Exhaust Emissions. Report No. EPA460/3-79010, July 1979.
10. Schuring, D. J. Tire Parameter Determination. Report No. DOT-HS-4-00923,
December 1975 (Nine Volumes).
* Related reference material:
a) W. R. Pierson and W. W. Brachaczek. In-Traffic Measurements of
Airborne Tire-Wear Particulate Debris, JAPCA 25^ 404-405 (1975).
b) Proceedings of the EPA Conference on Environmental Aspects of
Chemical Use in Rubber Processing Operations, Univ. of Akron,
12-14 March, 1975. (EPA-501-75-002, PB244172, July 1975, pp.
217-273.)
44
-------
11. How Tires Wear. Automotive Engineering, 81 (5): 28-30, May 1973.
12. Snyder,. R. H. Environmental Effects on Tire Treadwear. Tire Science
and Technology, 1 (2): 202-209, May 1973.
13. Martin, F. R. and P. H. .Biddison. Effect of Polymer, Road Surface and
Driving Conditions on Wear Surface Characteristics of Tire Treads.
Tire Science and Technology, 1 (4): 354-362, November 1973.
14. Bogdan, L., A. Burke and H. Reif. Technical Evaluation of Emission
Control Approaches and Economics of Emission Reduction Requirements
for Vehicles Between 6,000 and 14,000 Pounds GVW. Report No. EPA-
460/3-73-005, November 1973.
15. Motor Trucks in the Metropolis. Wilbur Smith and Associates (Commissioned
by the Automobile Manufacturers Association), August 1969.
16. Thelin, J. H. Laboratory Measurement of Abrasion Resistance Using
a Free Flowing Abrasive. Rubber Chemistry and Technology, 43:
1503-1514 (1970).
45
-------
APPENDIX A
AN ESTIMATE OF TRUCK TIRE WEAR
PARTICULATE CHARACTERISTICS
The purpose of this appendix is to attempt to determine the expected
characteristics of tire wear particulates from truck tires. Extrapolation
from measurements made on passenger car tires is to be made on the basis of
engineering judgment taking into account differences in tire sizes, construction
and tread formulations as well as the differences in usage patterns of auto-
mobiles and trucks. Since this effort must include the entire spectrum of
trucks, the results of this analysis must perforce be broadly drawn and
exceedingly general.
Truck tires are characterized by a broad range both in physical size and
construction as well as in operation. This stands in direct contrast with
passenger car tires which are designed to operate over very limited ranges of
both inflation pressure and load. Tires on passenger cars operate at loads which
range over narrow limits while truck tires are subject to very wide ranges of
loads corresponding to that of an unloaded vehicle to one that is loaded to its
maximum gross vehicle weight rating. Table A-l illustrates the physical
differences in size and ratings for a selected sampling of passenger car and
truck tires. Note that the passenger car tires are limited to a maximum
inflation pressure of 32 psi while truck tires are only limited by the rating
of the rims on which they are used.
Data on the usage of trucks is relatively limited and often atypical of
national averages. The following information has been obtained from references
14 and 15. The majority of motor trucks are used in the immediate proximity of
the home base of operations. Typically only 2% of all trucks regularly travel
more than 200 miles from the point of origin of the trip. Nearly 73% of all
light truck movements are local as contrasted with 50% for heavy units. In
this study, light trucks were defined to include 2-axle, 4-tire vehicles with
payload capacities of one-half to one and one-half tons and a GVW less than
10,000 Ibs. The principal findings are that the usage of motor trucks, like
usage of passenger cars, tends to concentrate in urban areas. There is a clear
difference, however, in usage between light trucks and heavy trucks (GWV > 10,000
Ibs). Light trucks are predominantly urban oriented. Based on statistics
obtained by pooling survey data from 11 urban areas, the light truck accounted
for 72% of all truck trips (on an average daily usage basis), 65% of all truck
miles and 96% of all trucks employed for personal use. Of the total daily one-
way trips made by light trucks, 27% were without load. The light truck was
found to average 28 miles and 5.9 trips per day. The medium heavy trucks
(GVW > 10,000 Ibs) account for 35% of the daily truck miles with a per vehicle
mileage of 36.5 miles and an average trip length of 3.8 miles. Table A-2 is a
summary of the numerical data.
46
-------
TABLE A-l
Comparison of Selected Specifications and
Ratings for a Sample of Typical Automobile
and Truck Tires
Application
Passeng
i
Light
\
Heavy
;er car
truck
truck
t
Size
AR78-13
ER78-13
AR78-14
JR78-14
AR78-15
HR78-15
6.00R16LT
12.00R16.5LT
9.00R20
11.00R24.5
O.D.,
in.
23.36
25.16
23.98
27.94
24.70
29.92
28.80
32.20
40.10
43.48
Load
Range
E
!
5*
r
C/D/E
D/E/F
E/F/G
F/G/H
Ply
Rating
t
\
I
r
6/8/10
8/10/12
10/12/14
12/14/16
Design
Press., psi
2
\
6
'
50/65/80
50/65/80
85/100/115
90/105/120
Max
Press. ,psi
3
!
*
\
2
r
*
r
Max Load at
Design press., psi
940
1240
940
1650
940
1970
1430/1690/1920
3000/3550/4045
4610/5150/5670
5780/6430/7030
*While C and D ranges exist, their usage is uncommon.
**Pressure limit is determined by the rim rating.
Data Source: 1979 Yearbook, The Tire and Rim Association, Inc.
-------
Table A-2
AVERAGE DAILY TRUCK USAGE IN 11 URBAN AREAS"
TRUCK
CLASS
LIGHT i 10.000 LB
MEDIUM-HEAVY > 10,000 LB
TOTAL
TRUCKS
MAKING TRIPS
NUMBER _&
72.989 71.8
28.691 28.2
101.680 100.0
DAILY TRIPS
NUMBER %
608.606 67.7
289.810 32.3
898.410 100.0
DAILY
TRUCK-MILES
NUMBER _%_
2,075,660 65.3
1,104.742 34.7
3,180,402 100.0
DAILY MILEAGE
PER TRUCK PER TRIP
28.4 3.4
36.5 3.8
31.3 3.5
DAILY
TRIPS
PER TRUCK1
8.3
10.1
8.8
00
*REF. 15
These values are for trucks making trips on a typical weekday. When related to all trucks registered in the urban area, the average is 5.9 trips per day,
since a proportion of the registered trucks are idle on any given day.
NOTE: The values are summations of trip values for the 11 areas shown in source.
SOURCE: Comprehensive transportation studies by Wilbur Smith and Associates in Albuquerque, New Mexico; Baltimore, Maryland; Baton Rouge,
Louisiana; Columbia, South Carolina; Lewiston, Maine; Little Rock, Arkansas; Manchester, New Hampshire; Monroe, Louisiana; Richmond, Virginia;
Sioux Falls, South Dakota; and Winston-Salem, North Carolina.
-------
From these figures it can be seen that truck usage, by and large, is very
similar to that for passenger cars. These trucks blend into the flow of the
urban traffic and experience similar accels/decels and cruise speeds. As a
consequence it appears reasonable to assume that, except for the difference in
the magnitudes of the forces and torques involved, tire wear schedules for truck
and passenger car tires are not unlike.
Tread rubber in passenger car tires is principally a blend of synthetic
rubbers with little to no natural rubber (NR) used. Typically, the blend ratio
is 60% styrene butadiene copolymer (SBR) and 40% polybutadiene (BR) together
with carbon black filler. Truck tires use SBR and RR together with NR in the
tread rubber. NR is used here because it has very low hysteretic losses
resulting in less heat buildup which can be destructive in truck applications.
Tread rubber hardness, as measured by the Shore number, typically varies in the
range from 55 to 65 for both passenger car and truck tires (hardness increases
with the numerical value of the Shore number).
One of the important characteristics of the tire rubber that affects wear
is the coefficient of friction. Of the three rubbers used in tread compounding,
SBR has the highest coefficient of friction when compared to BR and NR.
Reference 13 concludes, on the basis of experimental data, that materials with
a higher coefficient of friction (like SBR) tend to be snagged by roadway
asperities and form roll patterns in cornering, leading to fatigue type wear
on concrete highways. Low friction materials, like BR and NR, tend to slide
more readily over the asperities, to break cleanly and form a fine granular
texture when highly stressed. On coarse-texture surfaces (gravel), all materials
snag on the asperities, and the wear surfaces of each are similar in appearance.
Another area of difference between passenger car and truck tires is the
contact pressures in the footprint. Pressure distributions in the footprint
are very uneven and depend on load, inflation pressure, tire construction, etc.
As a rule of thumb, the minimum contact pressure can be taken as equal to the
inflation pressure. Thus while higher contact pressures will lead to higher
cyclic stresses and heat buildup, thermal degradation of tread rubber has not
been shown to be a factor in tread wear* (2, 3). Additionally, localized
stresses in tread rubber in contract with road asperities are usually below
the destructive limit so that the tread rubber does not suffer irreversible
damage (11). There are no known data on the effects of contact pressure on
tire wear except for the known fact that regions of high contact pressure in
the tire patch usually correspond to areas of higher wear.
To summarize, tires designed for truck use are characterized by different
constructions, sizes, loads and pressures when compared to passenger car tires.
Truck tires use different tread rubber formultions which result in a coefficient
of friction lower than that for automobile tires. On fine to medium textured
roads, such as found in urban areas, the lower friction material will tend to
abrade cleanly forming smaller particulates. On the other hand, the higher
* Under normal operating conditions.
49
-------
friction material, more typical of automobile tires, will tend to form larger
particles due to tearing and roll formation. Usage patterns among, trucks and
passenger cars in urban areas are similar. Consequently^ the particulates
generated by truck tires may be expected to be similar to those produced by
passenger car tires except that the size distribution may favor the smaller
size at the expense of the large. In terms of wear rates, it has been estimated
that wear rates for truck tires, on a grams per mile basis should be slightly
less than for bias-ply automobile tires (2). Supporting data are not available.
50
-------
APPENDIX B
VALIDATION OF THE SEM TECHNIQUE FOR
IDENTIFYING RUBBER PARTICULATES
The scanning electron microscope was used in two modes of operation in
the analysis of the tire wear particulates. Using secondary electron emission
detection, photographs were taken at magnifications up to 20,OOOX of the smaller
particulates to show their size, shape and structure at a high level of detail
and with excellent depth of field. In addition, by utilizing the x-rays emitted
by the specimens when bombarded by high-energy electrons in an energy dispersive
x-ray analyzer, an x-ray spectrum is obtained which permits the identification
of the elemental composition. All elements with an atomic number, Z, equal
to 11 and larger, can be identified in this manner.
With an element identification capability restricted to Z — 11, it is
clear that the SEM cannot be used to identify such organic materials as rubber
in a direct fashion. In the formululation of tread rubber compounds, various
additives are included to achieve desired characteristics. Among these
additives are sulfur, S, («1%) and zinc oxide, ZnO, (?»1.5%). Tests were there-
fore performed to ascertain the 'efficacy of identifying rubber particles, in
the presence of contaminants, by their sulfur and zinc "signatures." Antici-
paing that the debris from the Safety-Walk surface would represent the principal
source of contaminant in the tire wear samples, a piece of new (unused)
Safety-Walk was used as a control.
SEM specimens must be electrically conducting. Specimens that are normally
nonconducting are coated with gold (Au)/palladium (Pd) using vapor deposition
techniques. A photograph of a portion of new Safety-Walk surface is shown at
a magnification of SOX in Figure B-la. This material is in an "as received"
condition. The sharp asperities shown are normally blunted by a stoning operation
that is performed on the Safety-Walk surface after it is installed on the TIRF
belt. The lower photograph (Figure B-lb) reproduces the x-ray spectrum corre-
sponding to an integrated scan of the Safety-Walk surface. Silicon (Si) and
magnesium (Mg) are seen to be the principal identifiable* constituents of the
Safety-Walk surface; the gold and palladium peaks are due to the conductive
film applied to the specimen.
A very large rubber particle, worn from a tire during a test, is shown in
Figure B-2a at a magnification of 100X. The lacy, sponge-like appearance is
very characteristic of the large rubber particulates. The x-ray spectrum of
the rubber particle is depicted in Figure B-2b and exhibits the very prominent
peaks of sulfur and zinc. This result demonstrated that adequate sensitivity
was available for the identification of rubber particles in a sample of unknown
composition by the simultaneous presence of the elements sulfur and zinc. Note
the presence of silicon in the x-ray spectrum. Apparently, a fragment of a
silicon dioxide grit was imbedded in this particle of rubber.
* The Safety-Walk is composed mainly of organic materials except for the silicon
dioxide grits which are bonded with a polyurethane composed of aromatic
polyisocynides and polyamines.
51
-------
(a)
MAGNIFICATION = 30X
(b)
Figure B-1 SEM PHOTOGRAPH OF A NEW SHAPE OF SAFETY WALK AND ITS
ANNOTATED X-RAY SPECTRUM
-------
MAGNIFICATION = 100X
(b)
Zn
Figure B-2 SEM PHOTOGRAPH OF A LARGE RUBBER PARTICLE AND ITS
ANNOTATED X-RAY SPECTRUM
-------
APPENDIX C
THE CALSPAN TIRE RESEARCH FACILITY (TIRF)
A photograph of the TIRF facility is shown as Figure C-l; a dimensional
view of the facility is shown in Figure C-2. The primary features of the
machine are:*
Tire Positioning System
The tire, wheel, force sensing balance and hydraulic motor to drive or
brake the tire are mounted in the movable upper head. The head provides steer,
camber and vertical motions to the tire. These notions (as well as vertical
loading) are servocontrolled and programmable for maximizing test efficiency.
The ranges of the position.variables, the rates at which they may be adjusted
and other information are shown in Table C-l.
Roadway
The 28-inch wide roadway is made up of a stainless steel belt covered with
material that simulates the frictional properties of actual road surfaces.
The belt is maintained flat to within 1 to 2 mils under the tire patch by the
restraint provided by an air bearing pad which is beneath the belt in the tire
patch region. The roadway is driven by one of the two 67-inch diameter drums
over which it runs:. The road speed is servocontrolled; it may be programmed
to be constant or varied.
The surfaces usually used are "Safety-Walk".** These surfaces have
excellent microtexture giving a wet skid number*** of about 60 in the untreated
condition. The surfaces are honed to reduce the wet skid number to lower values
(typically surfaces of skid number 50 and 30 are used).
A unique feature of TIRF is the ability to carry out tests under wet road
conditions. A two-dimensional water nozzle spans the roadway. This nozzle has
an adjustable throat which can be set to the desired water depth. The flow
through the nozzle is then varied by controlling the water pressure. At each
test condition the water film is laid on tangential to the belt at belt
velocity. The film thickness may be varied from as low as 0.005 inches up to
0. 4* inches.
* A more complete description of this facility will be found in Reference 6.
** Manufactured by the 3M Company.
*** At 40 mph and 0.020-inch water depth using the ASTME-501 Standard Pavement
Traction Tire.
54
-------
&V •>». *'
CALSPAN TIRE RESEARCH FACILITY (TIRF)
55
-------
«-«•'-•••——>
Figure C-2 TIRE RESEARCH MACHINE
-------
TABLE C-l
TIRF CAPABILITIES
CHARACTERISTIC
RANGE
TIRE SLIP ANGLE (a)
TIRE, INCLINATION ANGLE (y)
TIRE SLIP ANGLE RATE (a)
TIRE INCLINATION ANGLE RATE CY)
TIRE LOAD RATE (TYPICAL)
TIRE VERTICAL POSITIONING RATE
ROAD SPEED (V)
TIRE OUTSIDE DIAMETER
TIRE TREAD WIDTH
BELT WIDTH
±30°*
±30°**
10°/sec
7°/sec
2000 Ib/sec
2"/sec
0-170 mph
Up to 46"
24" MAX.
28"
COMPONENT
TABLE C-2
BALANCE SYSTEM CAPABILITY
PASSENGER CAR
TIRE BALANCE
TIRE LOAD
TIRE TRACTIVE FORCE
TIRE SIDE FORCE
TIRE SELF ALIGNING TORQUE
TIRE OVERTURNING MOMENT
4,000 Ib
±4,000 Ib
±4,000 Ib
±500 Ib ft
±1,000 Ib ft
TIRE ROLLING RESISTANCE MOMENT ±200 Ib ft
TRUCK TIRE
BALANCE
12,000 Ib
±9,000 Ib
±8,000 Ib
±1,000 Ib ft
±2,000 Ib ft
±400 Ib ft
*Can be Increased to 90° with Special Setup,
**Can be Increased to 60° w.ith Special Setup
57
-------
Tire-Wheel Drive
A drive system which is independent of the roadway drive is attached to
the tire-wheel shaft. This separate drive allows full variation of tire slip
both in the braking and driving modes. The tire slip ratio, referenced to
road speed, is under servocontrol.
Balance System*
A six-component strain gage balance surrounds the wheel drive shaft.
Three orthogonal forces and three corresponding moments are measured through
this system. A fourth moment, torque, is sensed by a torque link in the wheel
drive shaft. The load ranges of the basic passenger car and truck tire
balances are shown in Table C-2. Transfer of forces and moments from the
balance axis-system to the conventional SAE location at the tire-roadway
interface is in the data reduction computer program.
System Operation
Data Acquisition Program (DAP) Control
The data acquisition program (DAP) is a software system which controls
machine operation and logs data during tests. DAP controls test operations by
means of discrete setpoints which are generated in the computer by the program.
These setpoints are sent to the machine servos which respond and establish
tire test conditions. After the setpoints are sent to the servos, a delay
time is provided which starts after the machine variables have reached a
steady state value within predetermined tolerances. This allows the system to
stabilize before data are taken. After data are taken, the next set of test
conditions is established and testing continues.
One or two variables can be changed during DAP testing. The other test
parameters are kept fixed throughout the test. Up to twenty data points can
be used for each variable in a run.
A data reduction program is used to operate on the raw data collected
during testing. These new data are reduced to forces and moments in the proper
axis system, arid all variables are scaled to produce quantities with engineering
units. Raw and .reduced data are temporarily stored in a disc file. Both
reduced and raw data can be transferred to magnetic tape and maintained as a
permanent record.
Reduced data points can be listed and plotted, and curves can be fitted to
the points. All of the standard Calspan plots can be generated from DAP test
data.
More detailed information on the balance systems and their calibration may
be found in Reference 6.
58
-------
Data lists and plots are displayed on the screen of a CRT console. Hard
copies of this information can be made from this .display.
Continuous Sampling Program (CSP) Control
The continuous sampling program (CSP) is a software system which controls
machine operation and continuously logs data during tests. Test variables can
be constant or changed at rapid rates. One or all variables can be changed
during a test. Data can be sampled at rates up to 100 samples per second.
Pauses are used so that data can be logged during desired intervals of the
test.
CSP testing can be conducted quickly which in turn reduces tire wear during
severe tests. The high rate of data sampling also permits limited dynamic
measurements to be made.
Two-parameter plots of data can be made. Carpet and family plots of test
data cannot be made with this program at the present time. CSP data will also
reflect time effects if tire characteristics are a function of the rate of
change of testing variables.
Data reduction is accomplished in a manner similar to that employed in
DAP testing.
59
-------
APPENDIX D
TEST TIRE DATA AND TREAD PATTERNS
The two test tires used on this program were a DR78-14 radial-ply tire
and a D78-14 bias-ply tire, both tubeless and with a "B" load range rating.
The radial tire was a Dual Steel II which is made with polyester radial plies
and a steel-belted construction. The bias tire is a Bias Jumbo 780 and
utilizes polyester cords in a cross bias construction. Both tires are products
of the General Tire and Rubber Company.
A pictorial representation of these tires is presented in Figure D-l.
Both tires use a seven-rib tread design which incorporates many sipes. As is
typical of that construction, the radial tire features a relatively open,
aggressive tread pattern as compared with that of the bias tire.
60
-------
GENERAL
DR78-14
DUAL STEEL II
RADIAL
GENERAL
D78-14
JUMBO 780
BIAS
Figure D-1 PICTORIAL VIEWS OF THE TEST TIRES SHOWING THE TREAD PATTERNS
61
------- |