Docstoc

Road grip test in Arjeplog

Document Sample
Road grip test in  Arjeplog Powered By Docstoc
					                                                                  2008:17



TECHNICAL REPORT


Road Grip Test in Arjeplog




                 Niclas Engström
                  Henrik Andrén
                 Mikael Nybacka
                Lennart Fransson
                 Roland Larsson




                Luleå University of Technology

                       Technical Report
                               -
   Department of Applied Physics and Mechanical Engineering
                 Division of Machine Elements

            2008:17 - ISSN: 1402-1536 - ISRN: LTU-TR--08/17--SE
                                  Road grip test in
                                        Arjeplog
                                17th to 19th of March 2008




Niclas Engströma, Henrik Andrénb, Mikael Nybackac, Lennart Franssonb and Roland Larssona
                        a
                      Division of Machine Elements, Luleå University
                    b
                    Division of Structural Engineering, Luleå University
              c
               Division of Functional Product Devolepment, Luleå University




Supported by Kempe foundations, CASTT and the Swedish road administration




Academic partner:
Preface
This report describes the tests for several road grip measuring devices from March 17, 2008,
to March 19, 2008, in Arjeplog. LTU would like to express our gratitude to the Swedish road
administration for equipment support and personnel, along with covering other costs. We
would like to thank Pon-Cat, ViaTech and SRIS for participation with equipment and
personnel. We also extend our gratitude to Coralba, Halliday Technologies INC. and Race
logics/VBox for their assistance. The substantial support from Kempe foundation is greatly
encouraging in this process to support the car test activities in Norrbotten. This project is a
part of the CASTT portfolio.

Luleå August, 2008

Niclas Engström, Henrik Andrén, Mikael Nybacka, Lennart Fransson and Roland Larsson




   Contents
   Executive summary ................................................................................................................ i
     Project partners ................................................................................................................... i
     Results ............................................................................................................................... ii
     Outlook .............................................................................................................................iii
   1 Introduction .................................................................................................................... 1
     1.1     Groups involved in the project ............................................................................... 2
   2 Goals and objectives....................................................................................................... 2
   3 Equipment information .................................................................................................. 2
   4 Methodology .................................................................................................................. 3
     4.1     Driver instructions.................................................................................................. 4
     4.2     Extract distance from speed and time data............................................................. 4
     4.3     Imprints of tire contact area ................................................................................... 4
     4.4     Presentation of data ................................................................................................ 5
     4.5     Surface types .......................................................................................................... 5
        4.5.1       Old System 2000 surface................................................................................ 5
        4.5.2       Polished ice .................................................................................................... 6
        4.5.3       New System 2000 ice..................................................................................... 8
        4.5.4       Drill sample of ice .......................................................................................... 9
        4.5.5       Footprints of continuous road grip measurement devices.............................. 9
        4.5.6       Wet asphalt................................................................................................... 13
     4.6     Time schedule for tests......................................................................................... 14
        4.6.1       Schedule and preparation for Monday 17th .................................................. 14
        4.6.2       Schedule and preparation for Tuesday 18th .................................................. 16
        4.6.3       Schedule and preparation for Wednesday, March 19, 2008 ........................ 17
        4.6.4       Schedule for extra activities Wednesday 19th .............................................. 17
     4.7     Calibration of RT3 ............................................................................................... 18
     4.8     Calibration of Coralba.......................................................................................... 18
     4.9     Calibration of ViaFriction .................................................................................... 19
     4.10 Calibration of TWO ............................................................................................. 19
   5 Conditions during the test ............................................................................................ 19
     5.1     Conditions on Monday March 17, 2008............................................................... 20
 5.2     Conditions on Tuesday, March 18, 2008 ............................................................. 20
 5.3     Conditions on Wednesday the 19th ....................................................................... 21
 5.4     Night conditions ................................................................................................... 22
    5.4.1     Monday to Tuesday...................................................................................... 22
    5.4.2     Tuesday to Wednesday................................................................................. 23
6 Results and discussion.................................................................................................. 23
 6.1     Test runs on two surfaces 17th .............................................................................. 23
 6.2     Test runs over 3 surfaces 18th ............................................................................... 23
    6.2.1     Test runs 70 km/h, G065 .............................................................................. 24
    6.2.2     Test runs 50 km/h, G065 .............................................................................. 26
    6.2.3     Test runs 30 km/h, G065 .............................................................................. 28
    6.2.4     Test runs 70 km/h, G055 .............................................................................. 30
    6.2.5     Test runs 50 km/h, G055 .............................................................................. 32
    6.2.6     Test runs 30 km/h, G055 .............................................................................. 34
    6.2.7     Test runs 70 km/h, VIAT ............................................................................. 35
    6.2.8     Test runs 50 km/h, VIAT ............................................................................. 37
    6.2.9     Test runs 30 km/h, VIAT ............................................................................. 39
    6.2.10 Test runs 70 km/h, CORA............................................................................ 41
    6.2.11 Test runs 50 km/h, CORA............................................................................ 41
    6.2.12 Test runs 30 km/h, CORA............................................................................ 42
    6.2.13 Test runs 70 km/h, G053 .............................................................................. 43
    6.2.14 Test runs 50 km/h, G053 .............................................................................. 45
    6.2.15 Test runs 30 km/h, G053 .............................................................................. 47
    6.2.16 Test runs 70 km/h, G056 .............................................................................. 49
    6.2.17 Test runs 50 km/h, G056 .............................................................................. 50
    6.2.18 Test runs 30 km/h, G056 .............................................................................. 52
    6.2.19 Test runs 70 km/h, TWOV........................................................................... 53
    6.2.20 Test runs 50 km/h, TWOV........................................................................... 56
    6.2.21 Test runs 30 km/h, TWOV........................................................................... 57
    6.2.22 Test runs 70 km/h, TWOT ........................................................................... 58
    6.2.23 Test runs 50 km/h, TWOT ........................................................................... 60
    6.2.24 Test runs 30 km/h, TWOT ........................................................................... 61
 6.3     Road grip development during March 18, 2008 .................................................. 63
    6.3.1     Comparison of RT3 measurement changes on old System 2000 ice ........... 63
    6.3.2     Comparison of RT3 measurement changes on brushed old polished ice..... 64
    6.3.3     Comparison of normalized road grip on brushed old polished ice .............. 65
    6.3.4     Comparison of RT3 changes on new System 2000 ice................................ 65
 6.4     Relative road grip comparison of different devices ............................................. 66
 6.5     Speed dependency ................................................................................................ 68
    6.5.1     Speed dependencies for continuous devices ................................................ 68
    6.5.2     Speed dependencies of braking devices ....................................................... 72
 6.6     Test runs across three surfaces 19th ...................................................................... 75
    6.6.1     Test runs 30 km/h, G055 .............................................................................. 75
    6.6.2     Test runs 30 km/h, G056 .............................................................................. 77
    6.6.3     Test runs 50 km/h, G055 .............................................................................. 78
    6.6.4     Test runs 30 km/h, SRIS .............................................................................. 79
    6.6.5     Test runs 50 km/h, SRIS .............................................................................. 80
    6.6.6     Test runs 70 km/h, SRIS .............................................................................. 80
    6.6.7     Test runs 30 km/h, VBox ............................................................................. 80
    6.6.8     Test runs 50 km/h VBOX............................................................................. 81
     6.6.9     Test runs 70 km/h, VBOX............................................................................ 82
  6.7     Stopping distance brake test on the 19th ............................................................... 83
     6.7.1     Stopping distance on polished ice at 30 km/h and 50 km/h ......................... 84
     6.7.2     Stopping distance on old System 2000 ice at 30 and 50 km/h..................... 86
     6.7.3     Stopping distance on new System 2000 ice at 30 and 50 km/h ................... 89
     6.7.4     Road grip, SRIS Volvo V70, polished ice ................................................... 92
     6.7.5     Summary of brake tests ................................................................................ 93
  6.8     Circle test runs March 18 and 19 ......................................................................... 93
     6.8.1     RT3s G055 and G065 circular test............................................................... 94
     6.8.2     TWOV circular test ...................................................................................... 94
  6.9     Handling course test runs on March 19th.............................................................. 96
  6.10 Rotational ice friction analysis............................................................................. 96
  6.11 Correlation between weather conditions and Road grip ...................................... 99
  6.12 Data and measurement issues............................................................................... 99
     6.12.1 RT3 G055..................................................................................................... 99
     6.12.2 RT3 G053..................................................................................................... 99
     6.12.3 Measurement challenges .............................................................................. 99
7 Conclusions ................................................................................................................ 100
8 Outlook....................................................................................................................... 101
  8.1     Data base content ............................................................................................... 101
  8.2     Features of road grip measuring equipment....................................................... 101
  8.3     RT3 usage and data acquisition system.............................................................. 102
     8.3.1     Start /stop triggering................................................................................... 102
     8.3.2     Faster data acquisition................................................................................ 102
9 References .................................................................................................................. 102
Appendix A .......................................................................................................................... A
  Vehicle data ...................................................................................................................... A
Executive summary

   The Swedish road administration sees a need to improve the road grip/friction estimation
capabilities for the Swedish road system. The challenge is to find simple, reliable methods to
measure and estimate road grip. This information could then be used in an expert system to
predict and prevent traffic hazards.
   A three day test was conducted on IceMakers test site Lake Kakel, Arjeplog. There were
six different types of measuring devices tested:

•   Three types of devices that continuously measured and logged road grip.
•   Two different devices measuring road grip through brake tests, one using an installed
    instrument and the other using data from the safety systems.
•   One device based on GPS data and three axis inertia measurements.

   Amongst the continuous devices one type, the RT3s used regular winter tires of friction
type. The other two types used Trelleborg Unitester 520 and 523, industrial tires with regular
rubber. Test runs were performed and road grip were recorded on four different surface types:

•   New System 2000 ice, a freshly prepared ice surface with grated furrows created with
    System 2000, to create a macro- and micro rough surface.
•   Old System 2000 ice, where the ice has been aged through weather and wear to micro
    smoothness while still being macro rough.
•   Brushed old polished ice, old polished ice that has been snow covered and then brushed
    off, giving a macro smooth and micro rough surface.
•   Polished ice, this ice is prepared through either flooding or with a Zamboni ice machine.
    This surface is both micro- and macro smooth. Only available on March 19th 2008.

   The weather situation during the three days was stable and weather data was collected
throughout the road grip measurements.

Project partners
  The road grip project is a collaboration between the Center for Automotive System
Technologies and Testing – CASTT at Luleå University of Technology and the Swedish Road
Administration. Industrial partners are listed in Table 1.




                                             i
Table 1                        Industrial partners during the tests in Arjeplog.

 Company                             Device/type                         Units    Test days Web page
 Halliday                            RT3                              G053, G055, 17th, 18th, www.hallidaytech.com
 Technologies                        continuous                       G056, G065     19th
 INC
 Pon-Cat                             TWO                                    TWOV,           17th, 18th www.pon-cat.com
                                     continuous                             TWOT
 ViaTech                             ViaFriction                             VIAT           17th, 18th Viatech.no
                                     continuous
 Coralba                             Coralba µ/3                            CORA            17th, 18th www.coralba.se
                                     Braking/deceleration
 SRIS                                Volvo V70                              SRIS                 19th                    www.sris.nu
                                     On board safety system
 Racelogic                           VBOX III                               VBOX            17th, 19th                   www.racelogic.co.uk
                                     Acceleration


Results
   All the measurement devices detect changes in road grip. Road grip measurement results
were consistent for the same device on different measurements, if a device measured a certain
road grip signature on one run then the following run showed a very similar signature.
Correlation between two RT3 units going down the same path is high.
   Correlation between stopping distance measurements and RT3’s road grip is not perfect,
see Figure 1. One reason for the deviation is that the RT3s measure road grip through forces
created with low lateral slip and stopping distance is based on high slip in the longitudinal
direction. However it is clear that when a RT3 indicates high road grip levels stopping
distances are short.


                          80                                                                80

                          70                                                                70
                                                                                                 Road grip value [N/A]




                          60                                                                60
      Stop distance [m]




                                                                                                                          Average stop distance for all
                          50                                                                50                            vehicles at 30 km/h [m]
                                                                                                                          Average stop distance for all
                          40                                                                40
                                                                                                                          vehicles at 50 km/h [m]
                          30                                                                30                            RT3 G055 average road grip at 50
                                                                                                                          km/h [N/A]
                          20                                                                20

                          10                                                                10

                          0                                                                 0
                                   Polished ice   Old system 2000 ice New system 2000 ice
                                                   Type of surface



Figure 1                       Average stopping distance with initial speeds 30 km/h and 50 km/h on three surfaces, average
                               road grip measured with RT3 G055 before each test.

   Average road grip values for all speeds were used to generate total normalized averages for
three surfaces; old System 2000 ice, brushed old polished ice and new System 2000 ice. The
normalization were done to brushed old polished ice, i.e. all devices have one in road grip on


                                                                       ii
brushed old polished ice, see Figure 2 for 50 km/h results. Results are similar for 30 km/h and
70 km/h. RT3s measure lower grip on new System 2000 ice compared to old System 2000
ice.




Figure 2   Relative grip normalized to brushed old polished ice for all systems at 50 km/h.

   For measurement of road grip on winter roads, we recommend use of rubber compounds
adapted for winter conditions, as the range of the road grip measurement is better with this
type of rubber.

Outlook
   The Road grip project has at this stage only scratched the surface in the field of road grip
prediction and estimation. Before further test runs with the RT3 units, the data acquisition
should be reviewed and updated. Halliday Technologies Inc. has made and is in the process of
making updates on the RT3. An automated data management system should be developed to
enable reports with average road grip values, max value, min value, statistical information
such as standard deviation, distribution curves, etc. There are also opportunities to search for
sections in the data where the road grip values were lower than a set value for a certain
distance. Luleå University of Technology recommends that a project is initiated to pursue
these opportunities.
   Tests with tires made of winter rubber and with a specified tread pattern should be made
with TWO, ViaFriction and ROAR.




                                                iii
1   Introduction
    The condition of the road surfaces significantly impact the safety of transportation on the
Scandinavian road systems during the winter months. One of the main challenges is to
monitor and predict the dangerous conditions. Road grip, the friction between the road and the
tires is of great importance for each vehicle when it comes to avoiding hazards. Ability to
decelerate and change direction quickly is affected mainly by speed and road grip, therefore
knowledge of road grip and ways of measuring it are of great importance. There are many
measuring devices, utilizing different methods to measure road grip/friction. The challenge is
to find out how relevant these are to the occurring traffic situations, i.e. how dangerous are the
conditions on a road section? The Swedish road administration has experience of the weather
conditions impact on road grip and how to mitigate hazardous conditions. The car testing
industry in northern Sweden has experience in reproducing hazardous conditions common on
the roads. There is a need to be able to quantify the road grip on these surfaces and thus
estimate the risks involved. Luleå University of Technology has experience with research and
information collection, and finding correlations between those. It is natural that we
collaborate to increase the knowledge of winter road conditions.
    There have been a lot of tests of different road grip measuring devices. What we need now
is more information on how the different devices measures road grip and correlate with each
other. Even more important is to establish the correlation between the test devices and normal
vehicles operating on the Scandinavian road systems. The data previously collected is not
utilized in a system to predict road grip on the Scandinavian road system. The main objective
for these tests was to see how the RT3 units from Halliday technologies Inc. measure the road
grip on winter roads. This was accomplished through a series of different tests, including
other measuring devices such as two TWO units from Pon-Cat, one ViaFriction from
ViaTech, one Coralba µ/3 from Coralba, two VBOX systems from Racelogic and one Volvo
V70 from the SRIS project, see Appendix A for further details about the vehicles.
    We tested the devices on the following prepared surfaces:

•   New System 2000 ice, a freshly prepared ice surface with grated furrows created with
    System 2000, to create a macro- and micro rough surface.
•   Old System 2000 ice, where the ice has been aged through weather and wear to micro
    smoothness while still being macro rough.
•   Polished ice, this ice is prepared through either flooding or with a Zamboni ice machine.
    This surface is both micro- and macro smooth.
•   Brushed old polished ice, old polished ice that has been snow covered and then brushed
    off, giving a macro smooth and micro rough surface.

   The main parts of the tests were conducted on a straight test track; each surface section on
the test track was a little more than 100 meters. This was to compare the different measuring
devices in the conditions they were made for. To explore limitations, some of the devices
were tested in curved driving such as a circular track and a handling course, see Figure 114.
   Stopping distances were also measured on different surfaces with some of the testing
vehicles, in an attempt to see if road grip could be related to stopping distance. All tests were
conducted at a minimum of two speeds. Laboratory equipment was brought into the field, to
perform the same friction tests that are performed with laboratory ice. Casts of the ice and
footprints of the measuring devices were taken and archived. Weather conditions were stable
during the three test days and were monitored and recorded during the daytime.



                                              1
1.1    Groups involved in the project
   The road grip project is a project between the Center for Automotive System Technologies
and Testing – CASTT [1] at Luleå University of Technology and the Swedish Road
Administration. The University partners are the Division of Machine Element and the
Division of Structural Engineering at Luleå University of Technology. Industrial partners
participating in the project are:

•     Halliday Technologies INC. with RT3
•     Pon-Cat with TWO
•     Coralba with Coralba µ/3
•     ViaTech with ViaFriction
•     SRIS with an instrumented Volvo V70
•     Racelogic with VBOX III 100 Hz, VB20SL and RLVBIMU 02.

    For further information about the equipment see Appendix A.

2     Goals and objectives
   The main goal is to find a method to correctly measure, record and predict road
grip/friction values important to transportation safety on the Scandinavian road system.
   The objective was to evaluate the different devices and their capacity to measure road
grip/friction on different surfaces, i.e. the repeatability, detection of different levels of road
grip, correlation to other road grip measuring devices and how measurements compare to
results from simulations of real traffic situations. Tests with the RT3s were longer and more
tests were performed.

3     Equipment information
    There were several types of road grip measurement devices represented in Arjeplog.
Detailed information about identity and data are presented in Appendix A. For an overview
see Table 2. We have included tire type on the pulling vehicles in an effort to explain changes
in track conditions caused by wear.




                                               2
Table 2      Devices with number of devices, type, method of measuring, designation on the devices, type
             of tires on the pulling vehicle (FT= friction tires, ST=Studded tires) and dates participating in
             the tests.

 Device           Units Type/method                        Designations/tire type          Date present
 RT3              4     Continuous/lateral slip           G053/FT, G055/ST, G056/FT        17, 18, 19
                                                                and G065/ST
 TWO              2        Continuous/longitudinal         TWOV/FT and TWOT/FT             17, 18
                           slip
 ViaFriction 1             Continuous/longitudinal                  VIAT/ST                17, 18
                           slip
 Coralba          1        Discrete/retardation                     CORA/ST                17,18 part of
 µ/3                                                                                       19
 SRIS             1        Discrete/retardation                     SRIS/FT                19
 Volvo V70
 VBOX III         1        Discrete/retardation                    VBOX/ST                 17, 19


4    Methodology
   The main part of the test where performed on Lake Kakel, Arjeplog. IceMakers prepared
test tracks with three different surfaces after each other, see Figure 3.


                                      General test track layout

                         Old system        Brushed old polished      New System
                 Start    2000 ice           ice/Polished ice         2000 ice       End

     Left side
    Right side                                                                               Paths
                                                                        Sand                  for
     Driving                                                                                vehicles
     direction             130 m                 130 m                  130 m


Figure 3     Test track layout on Lake Kakel.

   The test runs were driven with all vehicles present on each test day. Before the tests, tire
pressure was checked and adjusted if needed. The basic steps of a test are listed below:
• Line up in a left and right lane, according to what type of tires the vehicle had.
• Target speed distributed to all drivers, the drivers used the speedometer in the vehicle to
    maintain the speed required.
• Each vehicle accelerates up to the target speed and maintains it throughout the run.
• RT3 started logging of data at the “Start” position, this position was marked with a
    “Sergeant”. For the TWO units and the ViaFriction the logging were started 100 meters
    before the “Start” point, see Figure 3.
• Logging of data ended at the “End”, also marked with a “Sergeant”.

    More details about the tests below.



                                                    3
4.1   Driver instructions
    The instruction to the drivers was to keep the speed steady utilizing the vehicle
speedometer. The logged speed from the equipment correlated well with the base speed.
Before each sequence of test runs information about speed were verbally distributed directly
to each driver. The base speeds during the straight road grip tests were 30 km/h, 50 km/h and
70 km/h. For circular test runs and on the handling course the speeds 20 km/h, 30 km/h and
40 km/h were used. No verification of actual speed was performed during the tests for any of
the vehicles.
    The target was to go down the test track in the same path every time, to make the
comparison between the different devices more relevant. As the GPS data had low resolution
it was impossible to see where the different vehicles had placed the measuring wheels.
    There was sand on the right side of the last test section, see Figure 3. This sand made us
change the direction of the right lane; we made the right lane approach the left lane towards
the end.

4.2   Extract distance from speed and time data
   Since all measurements will be reported based on distance travelled, rather than time, the
RT3 distance had to be calculated from the speed and time. Sometimes speeds were not
correctly registered in the RT3’s log, some values of zero or very high are suddenly
registered. Those are easy to locate as outliers, since the other measurements never deviate
more than 7 km/h from the average. More troublesome is the fact that most of the time 10
samples were recorded per second, however some seconds we found as little as 4 samples. We
used a linear approximation to the distance in order to correct this. Since speed was accurately
kept, this solution produces good results. No such problems were noticed with the other
measurement devices, since they logged the distance travelled directly. Halliday Technologies
Inc. has added distance traveled in the latest software version.

4.3   Imprints of tire contact area
   Imprints of the different tires were made; this was done with Fuji Prescale LLLW pressure
sensitive film on a flat surface. The film has two sheets that are put together to be activated.
The film is not sensitive to the cold, but water can dissolve the coloring, therefore the edges
were taped together and the film was carefully stored. The imprint procedure is done by
lowering the tire carefully onto the film and leaving it on for an extended period of time, as
the pressure sensitive film develops slower at low temperatures. This process can be seen in
Figure 4.




                                             4
Figure 4        TWO’s reference wheel on pressure sensitive film.

   After the imprints where made the sheets were separated, this to stop any further coloring.
Photos were then taken of all sheets. They were later scanned with a high resolution scanner,
together with the color key, for analysis of what level of pressure had been applied onto the
film. The total pressure was calculated, a cross section of the maximum values was made and
the centre of pressure was located.

4.4        Presentation of data
   All graphs are accompanied by the following information:

      1.     Average road grip value for each section. We locate the transition between sections
             and subtract a small distance on each side of the transition point. The average is then
             calculated from the remaining data points in each separate section, yielding an
             average value for each surface
      2.     Data is generally presented with one relative standard deviation, 68% confidence
             intervals.

4.5        Surface types
   Several base types of winter roads were created. They can be divided into three categories.
The first is old System 2000 ice surfaces. All ice and snow surfaces will change
characteristics when vehicles are driving on them, especially if high slip rates are used.

4.5.1 Old System 2000 surface
   The general characteristics of this surface are slightly rounded edges of macro asperities,
with loose new snow in the furrows or covering the surface depending on the amount of
snowfall, see Figure 5 for conditions on March 19, 2008.




                                                    5
Figure 5   Old System 2000 ice with new snow.

   Figure 6 shows how the sharp edges have been worn down and a smoother surface is
created.




Figure 6   Old System 2000 ice.


4.5.2 Polished ice
    During our test runs, we encountered three types of polished ice. The first on Tuesday,
March 18, 2008, this brushed old polished ice was several days old and snow was brushed
away with a rotational brush earlier in the morning. The brushed surface had some roughness
to it as ice crystals had been formed on the ice from frost, some snow and ice particles had
frozen stuck to the ice. The location of this surface is seen in Figure 3. The second and
slipperiest polished ice was created for the tests Wednesday the 19th, with a Zamboni ice
machine, see Figure 7. The new polished ice where made on top of the brushed old polished
ice mentioned above.




                                                6
Figure 7   Zamboni made polished ice surface, only available on March 19, 2008.

   The last type were old polished ice, it was covered with a 20 – 40 mm thick layer of snow.
This polished ice was found on the circular tracks inner section. See Figure 8. It is difficult to
determine if the measurements were effected by the old polished ice under the snow.




Figure 8   Circular track with 20 – 40 mm snow covering the old polished ice surfaces

   One effect seen on all measurements during Wednesday was the sudden increase in road
grip just before and after the new polished ice. This since preparation with System 2000 had
been performed perpendicularly to the test track, to create well defined sections between the
different ice surfaces, see Figure 9. This creates a cross pattern that the RT3 has better road
grip on compared to driving parallel with the System 2000 furrows. The cross pattern distance
was about 10 meters on each side of the polished ice.




                                                7
Figure 9    At each end of the Zamboni ice, new System 2000 ice was prepared perpendicularly to the
            test track. A cross pattern was formed from the old System 2000 ice that previously covered
            the ice surface.


4.5.3 New System 2000 ice
   The third created surface type was new System 2000 ice. This ice is created with a blade or
drum equipped with rounded hard metal teeth that plough through the ice, see Figure 10. The
process creates longitudinal furrows that crack randomly, producing large amounts of sharp
ice particles that partially fill up the furrows. Depending on temperature and air humidity
these ice particles will fuse together and the base surface form create a surface that generates
good grip with several tire types, see Figure 11.




Figure 10   System 2000 blade, equipped with rounded hard metal teeth.




                                                 8
Figure 11   New 2000 ice with small amount of loose snow, Lake Kakel 19th March 2008, large amounts of
            sharp ice particles cover most of the surface except where the ridges are reaching through.


4.5.4 Drill sample of ice
   One test sample was drilled out of the ice, showing that the ice has a high amount of air
bubbles enclosed, see Figure 12. This indicates that large amounts of snow were present as
this ice was formed. The most likely reason for this is that water came up on top of the ice that
formed on Lake Kakel. When the temperature dropped the snow was caught in the ice and
thus created bubbles of air in the ice highest up on Lake Kakel.




Figure 12   Ice core sample, top to the left down to the bottom to the right.

   The road grip between this type of ice and tires is generally a bit higher, compared to ice
without air bubbles. Surface roughness is created when air bubbles in the ice break due to
wear, this creates sharp edges and ice particles, increasing the road grip.

4.5.5 Footprints of continuous road grip measurement devices
   The maximum pressure on the pressure sensitive film was higher than what it could
handle, producing oversaturated footprints. The lowest pressure came from the ViaFriction,
balancing a low wheel force over a moderate area. Looking at the cross section in Figure 13
reveals that many of the measurements bottom out at maximum pressure.




                                                   9
Figure 13   Maximum longitudinal pressure values of the footprint from the ViaFriction, showing that
            the pressure rang of the film were excceded.

   This means that the sum of pressure over the footprint will add up to a too small force. It
also means that the centre of pressure might be slightly off. The color intensity has been
adjusted to reveal the small differences that can be seen from maximum to smaller pressures.

4.5.5.1 RT3 G055
   In Figure 14 we can see the tread pattern of the RT3’s winter tire. The centre of pressure is
found in the middle of the footprint. The tire pressure which was 1.9 bar seems adequate since
the footprint is nicely square shaped, with even pressure over the entire footprint.




                                                10
                                                                              1 cm



Figure 14   Pressure distribution of RT3 G055’s footprint, friction tire.


4.5.5.2 RT3 G065
   In Figure 15 it can be seen that the centre of pressure is slightly misaligned. This could be
indication that the base plate was not completely horizontal.




                                                                              1 cm


Figure 15   Pressure distribution of RT3 G065’s footprint, friction tire.




                                                  11
4.5.5.3 RT3 G065 studded
   The footprint of the studs can easily be recognized in Figure 16. It is interesting to note that
the studs at the edge of the footprint are more visible, closer inspection of the film revealed
that the studs in the centre of the footprint had penetrated the film.




                                                                                 1 cm


Figure 16   Pressure distribution of RT3 G065’s footprint, with studded tire.


4.5.5.4 TWO 100106
   In Figure 17 we can see a comparison between the two wheels on the TWO. The force
applied on the measurement/slipping wheel, is half of the force applied on the reference
wheel. Note that both wheels will be slipping during operation as a large portion of the road
grip comes from hysteresis in the rubber as it slides over the surface. Pon-Cat claim that the
reference wheels slip rate is in the order of 1-2%




                                                 12
Figure 17   On the left side is the footprint of the reference wheel of the TWO 100601, on the right is the
            footprint of the measurement wheel of the same. The footprint is much smaller on the
            measurement wheel.


4.5.5.5 ViaFriction
   In Figure 18 we can see an edge bias in the pressure profile. Despite this, the centre of
pressure is positioned in the middle. This is caused by the oversaturation of the pressure
sensitive film as no difference in pressure is detected even though the pressure seems higher
on the edges to the right and thus could have shifted the centre of pressure towards the right
side.




Figure 18   Pressure distribution of Via Tech 1001’s footprint.


4.5.6 Wet asphalt
   During calibration of the RT3s we used a heated asphalt stretch at Daimler test site in
Arjeplog. During our time slot snow was falling and that made the surface wet. Figure 19
shows the asphalt stretch in dry conditions with polished ice on the sides. The asphalt has a
fairly high level of texture and the edges are sharp, see Figure 20. This asphalt had some
cracks but the overall quality is superior to all but new surfaces on any road system.


                                                  13
Figure 19   Overview of the asphalt stretch at the Daimler test site in Arjeplog. This picture was taken
            two weeks prior to the calibration.




Figure 20   Detailed view of the asphalt surface when dry.


4.6   Time schedule for tests
  The main tests in Arjeplog were conducted during three days, beginning on Monday,
March 17 and ending on March 19, 2008. The schedules for each day can be seen below.

4.6.1 Schedule and preparation for Monday 17th
   Table 3 lists start time, end time, speed and what side of the test track each vehicle used
during each run on Monday, March 17, 2008. Each test was driven in a triplicate of runs. The
four letter description represents a specific vehicle, details about each vehicle are found in
Appendix A. Numbers within brackets have been omitted due to data error.




                                                 14
Table 3     Test runs during Monday, March 17, 2008

Time               Speed Units on right side                             Units on left side
                   [km/h]
15:14 – 15:23        50   G065, G055, G053, G056, TWOV,
                          TWOT, VIAT, CORA, VBOX
15:29 – 15:45        70   G065, G055, G053, G056, TWOV,
                          TWOT, VIAT, CORA, VBOX
15:57 – 16:09        70                                                  G065, G055, G053, G056, TWOV,
                                                                         TWOT, VIAT, CORA, VBOX
16:31 – 16:46        50                                                  G065, G055, G053, G056, TWOV,
                                                                         TWOT, VIAT, CORA, VBOX
16:47 – 16:55        30                                                  G065, G055, G053, G056, TWOV,
                                                                         TWOT(3), VIAT, CORA, VBOX
16:55 – 17:02        30      G065, G055, G053, G056, TWOV,
                             TWOT, VIAT, CORA, VBOX

   The surface for the test runs was ploughed, old System 2000 ice. There was new snow on
the surface in general as there were some light snow fall during the morning. Many test runs
were driven by Pon-Cat, ViaTech and Coralba devices during the morning, this testing
activity created two wheel tracks that had stable road grip, see Figure 21.




Figure 21   Monday 17th test track with tire tracks on old 2000 ice surface

   Test runs were also made in the new snow to the left of the wheel tracks. See Figure 22
below.




                                                 15
Figure 22   Monday 17th test track with new snow on old System 2000 ice.

   The order of the units was:
   1. RT3, G065
   2. RT3, G055
   3. RT3, G053
   4. RT3, G056
   5. TWO, TWOV
   6. TWO, TWOT
   7. ViaFriction, VIAT
   8. Coralba, CORA
   9. VBox, VBOX

   We did not change the order of the vehicles during Monday. For information about what
type the test units are, see Appendix A.
   Tests made on Monday were overlapped by tests the following days and results will not be
presented from this day.

4.6.2 Schedule and preparation for Tuesday 18th
   Table 4 lists start time, end time, speed and what side of the test track each vehicle used
during each run on Tuesday, March 18, 2008. Each test was driven in a triplicate of runs. The
four letter description represents a specific vehicle, details about each vehicle are found in
Appendix A. Numbers within brackets represent runs omitted due to data error.




                                                16
Table 4    Test runs made on Tuesday, March 18, 2008

Time               Speed [km/h]     Units on right side                   Units on left side
10:26 – 10:44           30          G065(2), G055, VIAT, CORA             G053, G056, TWOV, TWOT
10:45 – 10:57           50          G065(2, 3), G055, VIAT, CORA          G053, G056, TWOV, TWOT
10:57 – 11:10           70          G065, G055, VIAT, CORA                G053, G056, TWOV, TWOT
14:12 – 14:24           50          G065, G055, VIAT, CORA                G053, G056, TWOV, TWOT
14:25 – 14:34           70          G065, G055, VIAT, CORA                G053, G056, TWOV, TWOT
14:34 – 14:42           30          G065, G055, VIAT, CORA                G053, G056, TWOV, TWOT
15:08 – 15:18           70          G065, G055, VIAT, CORA,               G053, G056
                                    TWOV, TWOT
15:19 – 15:26            30         G065, G055, VIAT, CORA,               G053, G056
                                    TWOV, TWOT
15:27 – 15:37            50         G065, G055, VIAT, CORA,               G053, G056
                                    TWOV, TWOT

    It was decided that the TWOs would change from measuring on the left side of the test
track to the right side for the last nine runs, this to enable comparisons between the two sides.
    The preparation that we requested from Icemakers contained 3 sub sections for the runs on
Tuesday. The first section was ploughed ice that previously had been prepared with System
2000, the surface was old making the edges of the ice rounder. The furrows in the ice are
filled with fresh snow. The second was brushed old polished ice where snow and ice crystals
had attached onto the surface forming rough parts. The third section was a new System 2000
ice surface.

4.6.3 Schedule and preparation for Wednesday, March 19, 2008
   Table 5 lists start time, end time, speed and what side of the test track each vehicle used
during each run on Wednesday, March 19, 2008. Each test was driven in a triplicate of runs.
The four letter description represents a specific vehicle, details about each vehicle are found
in Appendix A. Numbers within brackets have been omitted due to data error.

Table 5    Test runs made on Wednesday the 19th.

Time               Speed [km/h]     Units on right side           Units on left side
09:58 – 10:21           30          G065(1,2), G055, CORA,        G056
                                    SRIS, VBOX
10:22 – 10:34            50         G065, G055, CORA,             G056
                                    SRIS, VBOX
10:35 – 10:53            70         G065, G055, CORA,             G056
                                    SRIS, VBOX
11:38 – 11:46           50                                       G065,G055,SRIS
11:46 – 11:55           30                                       G065,G055,SRIS
   Preparations on Wednesday had one major difference from preparations on Tuesday, this
time the polished ice was made a few hours before the tests started with a Zamboni ice
machine [2].

4.6.4 Schedule for extra activities Wednesday 19th
   After the main tests were done some extra tests were performed, this section describes
these tests and how they were executed.


                                             17
4.6.4.1 Brake test on polished ice
   Brake tests were made after that we had completed all regular test runs on Wednesday.
Regular cars were driven at 30 km/h and 50 km/h onto the polished ice. At a marker they were
told to apply the brakes fully, activating the ABS system in the vehicle. The instructions were
to brake until the vehicle came to a full stop. The stopping distance was measured as close to
the start point of the front wheels position when the brakes were applied and not from the
location of the marker. Each car made three test runs at each speed. The surface used was the
right side of the polished ice. It should be noted that the studded tires changed the surface
characteristics as the test progressed.

4.6.4.2 Brake test on old System 2000 ice
   Regular cars were driven at 30 km/h and 50 km/h on old System 2000 ice. At a marker
they were told to apply the brakes fully, activating the ABS system in the vehicle. The
instructions were to brake until the vehicle came to a full stop.

4.6.4.3 Round track test
   Two RT3 units were tested on a round track. The tests were performed on snow covered
old polished ice on the inner radius and on old System 2000 ice on the middle radius, see
Figure 114. The test runs were carried out with two speeds, 20 and 40 km/h, both counter
clockwise and clockwise.

4.6.4.4 Handling course test runs
   RT3 G065 was driven on the handling course at 20 and 40 km/h in both directions, see
Figure 114 for plots of the course.

4.6.4.5 Regular road test runs
   Several test runs where made with the devices on the road leading to the test area on lake
Kakel. The speed was 50 km/h during the test. These test where made in conditions with too
large variations on the test sections to be able to analyze the measurements. They will not be
further presented in this report.

4.7   Calibration of RT3
   The recommended calibration method for RT3 is basically:

   1. Set the display to use tread depth compensation for 9.2 mm.
   2. Drive the units down the dry asphalt and manually record data from the display.
   3. Change the angle of the RT3 relative to the vehicle to get an average reading of 100 on
      the display.

   The recommended procedure could not be followed as we did not have dry asphalt, we had
too short a distance and we did not have tires with temperature calibration data. The tire on
the RT3 is standard winter tire from Bridgestone. We had to settle for calibration on a wet
asphalt surface at Daimlers test site, see chapter 4.5.5. The temperature calibration was set to
100 as surface temperature was close to 0°C.

4.8   Calibration of Coralba
   The Coralba unit was taken to a test stretch with a verified distance of 1000 meters and
was calibrated. The process requires a reference road grip value. This value was measured
with a TWO unit, see chapter 4.5.5.


                                             18
4.9    Calibration of ViaFriction
   The calibration for the ViaFriction was made with a ROAR late autumn 2007 and is
required to be done each year. The ROAR is an official road grip reference device used to
calibrate other road grip measuring devices in Norway, and they are operated by the
Norwegian road administration, Vegvesen. One 1000 meter run was performed to verify tire
wear.

4.10 Calibration of TWO
    Calibration is done yearly according to Pon-Cat instructions and verified with a ROAR
unit, see 4.9. The TWO is calibrated by pressing the reference wheel to the ground and thus
locking it. A vertical strap mounted to a scale is wrapped around the rear measurement wheel
and used to hold the wheel off the ground, at a height of between 0.5 – 2 cm, see Figure 23.
The strap supplies a force transforming into tension in the chain between the wheels, and
normal force acting through the pivot point. The TWO is calibrated at a tension of 294 N
(66 pounds) and at 588 N (132 pounds), giving calibration constants for normal load and
friction force. A TWO device has shock absorber and are hydraulically lowered and elevated
from the test surface.




Figure 23   Calibration of TWO unit, D = Shackle, E = lifting strap, F = scale, and G = calibration stand.

    For detailed information please contact Pon-Cat directly.

5     Conditions during the test
   Weather data was collected during the test runs to see if there were any noticeable effects
on the measurements from weather. Temperature and humidity data were logged with a
portable weather station, as well as an USB logger. Temperature sensors on the weather
station seemed unaffected by the solar influx, whereas the temperature logger displays a bias
to solar influx. Even though the USB logger passes 0°C, the ice does not display signs of
melting. Ice temperature was measured with a calibrated IR sensor.




                                                 19
5.1        Conditions on Monday March 17, 2008
   Initially sunny, around 15:30 the sky became overcast; this can be seen in the ground
temperature measurements in Figure 24 logged by the USB logger. The logger is sensitive to
solar influx. We see how Ice temperature and ground temperature converge as the solar influx
decreases. Measurements are presented in Table 6 below.
      2                                                                                                        1


      1                                                                                                        0.9


      0                                                                                                        0.8


      -1                                                                                                       0.7
                                                                                            Air temp ('C)
      -2                                                                                    Ground Temp ('C)   0.6
                                                                                            Ice Temp ('C)
                                                                                            Air Humidity
      -3                                                                                                       0.5
                                                                                            Ground Humidity
                                                                                            Pressure (torr)
      -4                                                                                                       0.4


      -5                                                                                                       0.3


      -6                                                                                                       0.2


      -7                                                                                                       0.1


      -8                                                                                                      0
       14:00         14:30         15:00        15:30        16:00       16:30         17:00              17:30



Figure 24      Metrological observations during Monday the 17th.



Table 6        Meteorological data during Monday, March 17, 2008.

                       Air temp     Ground      Pressure     Wind      Air       Ground        Ice Temp
 Date                    [°C]      temp [°C]      [torr]     [m/s]   Humidity    Humidity         [°C]
 3/17/2008 14:05         -4.9         0.0          1.0        2.4     76%         58%             -2.0
 3/17/2008 14:51         -4.2         0.3          1.0        1.6     71%         59%             -1.9
 3/17/2008 15:11         -4.5         0.3          1.0        1.6     69%         59%             -1.9
 3/17/2008 15:23         -4.7        -0.5          1.0        0.0     73%         62%             -1.9
 3/17/2008 15:29         -4.0         0.0          1.0        0.0     72%         61%             -1.9
 3/17/2008 15:35         -4.1        -0.5          1.0        0.0     70%         61%             -1.9
 3/17/2008 16:00         -5.2        -2.0          1.0        0.0     75%         65%             -2.0
 3/17/2008 16:06         -4.8        -1.5          1.0        0.0     76%         67%             -2.6
 3/17/2008 16:35         -6.6        -3.5          1.0        1.2     76%         68%
 3/17/2008 17:21         -7.0        -5.5          1.0        0.0     83%         75%



5.2        Conditions on Tuesday, March 18, 2008
   Slight overcast all day, despite this the USB logger display much higher temperatures than
both the IR sensor and the weather station. This difference is believed to be due to high solar
influx in the invisible spectra (UV). Data from the day can be seen in Figure 25 and Table 7
below.


                                                  20
      6                                                                                                                     1


                                                                                                                            0.9
      4
                                                                                                                            0.8

      2                                                                                                                     0.7


                                                                                                                            0.6
      0

                                                                                                                            0.5

      -2
                                                                                                                            0.4


      -4                                                                                                                    0.3
                                                                                Air temp ('C)
                                                                                Ice Temp ('C)                               0.2
                                                                                Ground Temp ('C)
      -6
                                                                                Pressure (torr)
                                                                                                                            0.1
                                                                                Air Humidity
                                                                                Ground Humidity
      -8                                                                                                                    0
       09:30    10:00   10:30   11:00   11:30   12:00   12:30   13:00   13:30    14:00   14:30     15:00    15:30   16:00



Figure 25      Meteorological measurements during Tuesday, March 18, 2008. The curves have been
               smoothed for better approximation.



Table 7        Meteorological data Tuesday March 18, 2008.

                           Air temp       Ground   Pressure   Air     Ground                               Ice Temp
 Date                        [°C]        Temp [°C]  [torr]  Humidity Humidity                                 [°C]
 3/18/2008 09:43             -5.1           3.5     0.941     54%     44.5%                                   -4.6
 3/18/2008 09:46             -6.3           2.5     0.941     64%     45.5%                                   -4.2
 3/18/2008 10:13             -4.6            3      0.941     61%     46.5%                                   -3.9
 3/18/2008 10:45             -5.4            2      0.942     69%      49%                                    -3.9
 3/18/2008 11:02             -5.5            2      0.942     69%     49.5%                                   -3.9
 3/18/2008 11:10             -5.4           1.5     0.942     72%      52%                                    -3.8
 3/18/2008 11:32             -4.6            2      0.942     69%     50.5%                                   -3.7
 3/18/2008 13:02             -3.3            3      0.941     64%     49.5%                                   -3.2
 3/18/2008 14:32             -2.6            3      0.941     58%      50%                                    -3.1
 3/18/2008 16:16             -5.8          -5.5     0.942     68%     62.5%                                   -3.9


5.3        Conditions on Wednesday the 19th
   The night and morning were cold, followed by a clear day with high solar influx, only a
few clouds in the sky during the day. Despite this, no melting on the ice surface could be
observed. In Figure 26 and Table 8 meteorological data collected during March 19, 2008, can
be seen. Note that Ground temperature in Figure 26 is incorrect as high solar influx heated the
dark colored USB temperature and humidity logger.




                                                        21
      15                                                                                                                              1



                                                                                                                                      0.9
      10

                                                                                                                                      0.8


       5
                                                                                                                                      0.7



                                                                                                                                      0.6
       0

                                                                                                                                      0.5

       -5
                                                                                                                                      0.4



                                                                                                                                      0.3
      -10
                                                                             Air temp ('C)
                                                                             Ice Temp ('C)                                            0.2
                                                                             Ground Temp ('C)
      -15
                                                                             Pressure (torr)
                                                                                                                                      0.1
                                                                             Air Humidity
                                                                             Ground Humidity
      -20                                                                                                                             0
        09:20   09:50   10:20   10:50   11:20   11:50   12:20        12:50   13:20    13:50   14:20   14:50   15:20   15:50   16:20




Figure 26          Meteorological data Wednesday, March 19, 2008



Table 8            Meteorological data Wednesday, March 19, 2008.

            Date            Air temp         Ground             Pressure               Air            Ground          Ice Temp
                              [°C]          Temp [°C]             [torr]             Humidity         Humidity           [°C]
 3/19/2008 09:22               -16             5.5               0.942                                 0.41
 3/19/2008 09:58             -12.6             10                0.943                                 0.305              -7.6
 3/19/2008 10:36             -14.3             10                0.942                42%              0.31
 3/19/2008 11:37              -7.7              5                0.944                53%              0.335              -6.7
 3/19/2008 14:44              -3.9            -5.5               0.944                51%              0.64               -6.4
 3/19/2008 15:13              -2.3            -5.5               0.944                45%              0.69
 3/19/2008 15:27              -5.7             -6                0.944                49%              0.68
 3/19/2008 16:39              -5.6             -9                0.946                48%              0.725              -6.6

5.4         Night conditions
   The USB logger was kept outdoors from Monday morning to Wednesday evening, logging
temperature and humidity every minute, see Figure 27. The measurements show that
conditions never reached dew point. The night-time ground air temperature was more accurate
than the day-time ground air temperature, since solar influx has no effect on the USB logger
at night.

5.4.1 Monday to Tuesday
   Monday and Tuesday were relatively cloudy, the temperature was fairly stable, still, a
significant drop can be seen from day to night, mostly due to the USB logger’s ability to
absorb sunlight during the day.




                                                                22
5.4.2 Tuesday to Wednesday
   The cloudy sky from Tuesday broke up during the night, and allowed the ice to cool down
significantly as can be seen; a high solar influx also warmed the ice during the following day.
No melting could be observed though.

      15                                                                                                    100


                                                                                                            90
      10

                                                                                                            80
       5
                                                                                                            70

       0
                                                                                                            60


       -5                                                                                                   50


                                                                                                            40
      -10

                                                                                                            30
      -15
                                                            Temp(C)                                         20
                                                            Humidity(%rh)
      -20
                                                                                                            10


      -25                                                                                                    0
       Mon 17     Mon 17   Mon 17   Tue 18   Tue 18        Tue 18   Tue 18   Wed 19   Wed 19   Wed 19   Wed 19
        06:00      12:00    18:00    00:00   06:00         12:00     18:00    00:00    06:00   12:00     18:00



Figure 27       Temperature and humidity chart, from Monday 17th to Wednesday 19th.



6     Results and discussion
   Analysis of data recorded between March 17 and March 19, 2008, in Arjeplog revealed
several interesting avenues for further investigations. The results are sorted into multiple
sections to reflect the changing surfaces. They are also sorted to show correlation between the
different equipment used.

6.1    Test runs on two surfaces 17th
   The first type of condition we had available was old System 2000 ice with some new snow
on it. The snow cover was roughly 3 cm thick.
   The tests on Monday were replicated on Tuesday and Wednesday and are therefore
omitted.

6.2    Test runs over 3 surfaces 18th
    On Tuesday, March 18, 2008, several test runs were made, the surfaces available were:

    1. Ploughed old ice surface, first created with System 2000
    2. Brushed old Polished ice, cleared with rotational brush system
    3. Recently ploughed ice, System 2000 was used




                                                      23
    Those are hereby referred to as sub runs. For further information about surfaces see chapter
4.5. In this test we investigated the repeatability of an RT3 device. Comparisons with other
RT3s were also performed. We repeated the test runs with as close to the same conditions as
we could manage. It should be noted that the surfaces vary slightly and it is difficult to get the
measuring wheel to run across the exact same path. The ViaFriction device and the TWO
devices have the measuring wheel on the opposite side of the vehicle, this is the left side also
used by ROAR. As this is the case no direct comparison on exactly the same surface was
made. It also should be noted that steering has an impact on the readings from an RT3. There
is a steering sensor installed that can discontinue the data logging if the steering angle
supersedes a pre set value. This function can be turned on and off.
    We will compare measurements with other road grip measuring equipment such as Pon-
Cat TWO, ViaFriction, V-Box and Coralba equipped vehicles.

6.2.1 Test runs 70 km/h, G065
   Figure 28 below illustrates the results from test runs 7 – 9 for G065 at 70 km/h on the right
side of the test track. This is the third test speed on March 18, 2008, with 30 and 50 km/h runs
made previously. One interesting observation in Figure 28 is seen in sub run 3, this is the new
System 2000 ice section, and one can see that the road grip decreases towards the end of the
measurement, look for this in all measurements.


                                         RT3, G065 70 km/h 18th, Run 7-9


                  100
                   90
                   80
                   70
     Grip value




                   60                                                           RT3 G065 7
                                                                                RT3 G065 8
                   50
                                                                                RT3 G065 9
                   40                                                           RT3 G065 average
                   30
                   20
                   10
                    0
                    0
                        26
                             52
                                   78

                                    4
                                    1
                                    8
                                    5
                                    2
                                    0
                                    7
                                     4
                                    1
                                    8
                                     4
                                    1
                                  10
                                  13
                                  15
                                  18
                                  21
                                  24
                                  26
                                  29
                                  32
                                  34
                                  37
                                  40




                                          Distance from start [m]


Figure 28           Road grip on runs 7 – 9 for RT3 G065 at 70 km/h.

   It is clear that this RT3 has good repeatability of measurements and has good resolution at
the conditions tested. The frequency of the data acquisition is around 10 Hz. The data from
the load cell are sampled at 100 Hz and thus each road grip value is an average from roughly
10 samples.



                                                       24
    We have calculated standard deviation and average road grip values for each of the three
sections in each test run; data is presented in Table 9. The data is also presented with a graph
for easier overview, see Figure 29. One reason for the variations seen in Figure 28 that results
in larger standard deviation is the macro roughness of the test surfaces. This roughness effect
the RT3 measurement in a similar way to how most cars would be affected on bumpy roads.
A better suspension will increase average road grip and make road grip more even in macro
rough areas. We did no measurement how well the test devices maintained even force on the
measurement wheels. The TWO device measure the vertical forces on measurement tire and
reference tire.

Table 9 Average road grip values and standard deviation for runs 7 - 9 at 70 km/h with RT3 G065.

 Unit, run      Average        Standard Relative standard          Distance     Distance
 and sub run    Road grip      Deviation Deviation [%]             start        stop
 G065, 7, 1       64.1            5.83          9.1                    10            120
 G065, 8, 1       66.6             5.7          8.6                    10            120
 G065, 9, 1       66.3             6.3          9.5                    10            120

 G065, 7, 2         40.2          10.7               26.7             140           250
 G065, 8, 2         39.1          10.8               27.7             140           250
 G065, 9, 2         38.9          10.2               26.2             140           250

 G065, 7, 3         53.0           7.9               14.9             270           380
 G065, 8, 3         52.1           6.6               12.7             270           380
 G065, 9, 3         51.4           8.8               17.2             270           380

   In Table 9 we can see that the average road grip values are less in the last sub run of the
test track compared with the first section of the test track. This result was unexpected as the
last section should give higher road grip values due to its significantly more aggressive
structure with new sharp ice edges. We believe that the RT3 units, that use a very small
angular displacement, translating into a low lateral slip, may slide on loose pieces of ice and
therefore indicate lower road grip. If the lateral slip had been greater we would probably have
seen higher road grip values on the last section of the test track. Brake tests show that at high
longitudinal slip situations, high road grip values are generated on new System 2000 ice, see
chapter 6.7.5. The difference between lowest and highest average road grip value in Table 9
         Average road grip values and standard deviation for runs 7 - 9 at 70 km/h with RT3
G065., is 28 units or 71%.




                                                25
                            G065 runs 7-9 70 km/h Mean road grip with one standard deviation

                  80
                  70
                  60
      Road grip




                  50                                                                                          High
                  40                                                                                          Low
                  30                                                                                          Mean

                  20
                  10
                  0
                        G065,   G065,   G065,          G065,   G065,   G065,         G065,   G065,   G065,
                         7, 1    8, 1    9, 1           7, 2    8, 2    9, 2          7, 3    8, 3    9, 3
                                                     Vehicle, Run and sub run



Figure 29              Average road grip with one standard deviation on runs 7 - 9 for RT3 G065 at 70 km/h.

    In Figure 29 above the standard deviation is much larger for sub run two during the runs.
In Table 9 one can see that the standard deviation ranges from 26.2 to 27.7 for sub run 2.
When we search for an explanation, the answer is found in Figure 28; we can clearly see that
there is a large road grip variation recorded consistently by the G065 RT3 unit. This will
increase the standard deviation. Through these measurements we can see that the right side of
the test track doesn’t have constant stable road grip on the three sub sections. The most likely
reason why this dramatic change in road grip is measured is road grip tests with a Coralba
device performed simultaneously. The Coralba unit had studded winter tires that made marks
into the brushed old polished ice during the brake tests and these marks increased the road
grip where the brake tests where performed. Most cars had studded winter tires that would
increase the road grip measured by the friction tires (winter tires without studs) assembled on
all RT3s. It should only occur on the right side of the course, as vehicles measuring on the left
side only had friction tires.

6.2.2 Test runs 50 km/h, G065
   Figure 30 illustrates the road grip on runs 10 -12 for G065 at 50 km/h. We can see some
deviations in run 12, (yellow curve) before the 250 meter position. No explanation is
available.




                                                          26
Figure 30   Road grip on runs 10 – 12 for G065 at 50 km/h.

   Table 10 contains average road grip and standard deviation on runs 10 – 11 for RT3 G065
at 50 km/h. The relative standard deviations differs some compared with the once presented in
Table 9. No correlation with speed can be verified. The difference between the lowest and
highest average road grip in Table 10 is 35 units or 111%.

Table 10        Average road grip with one standard deviation on runs 10 – 12 for G065 at 50 km/h.

Unit, run       Average Standard          Relative standard     Distance Distance
and sub run     Road grip deviation        Deviation [%]          start    stop
G065, 10, 1      65.72      6.75                10.27              15      115
G065, 11, 1      63.99      7.93                12.39              15      115
G065, 12, 1      66.82      7.54                11.28              15      115

G065, 10, 2       34.74         7.69             22.14             145        230
G065, 11, 2       34.72         6.56             18.89             145        230
G065, 12, 2       31.73         6.06             19.08             145        230

G065, 10, 3       51.79         9.53             18.40             275        375
G065, 11, 3       52.78         9.12             17.28             275        375
G065, 12, 3       50.47         6.87             13.61             275        375




                                                27
Figure 31           Average road grip with one standard deviation on runs 10 – 12 for G065 at 50 km/h.


6.2.3 Test runs 30 km/h, G065
   Figure 32 illustrates results from runs 22 – 24 at 30 km/h on the right side of the test track.
This is the eight set of test runs on March 18, 2008. There is clear correlation between the
different runs.


                                             RT3, G065 30 km/h 18th, Run 22-24


                  120


                  100


                  80
                                                                                                                         RT3 G065 22
     Grip value




                                                                                                                         RT3 G065 23
                  60
                                                                                                                         RT3 G065 24
                                                                                                                         RT3 G065 average
                  40


                  20


                   0
                        0
                            26
                                 52
                                      77
                                           102
                                                 127
                                                       152
                                                             177
                                                                   202
                                                                         227
                                                                               253
                                                                                     278
                                                                                           303
                                                                                                 329
                                                                                                       354
                                                                                                             380
                                                                                                                   405




                                                  Distance from start [m]


Figure 32           Road grip for RT3 G065 on runs 22 – 24 at 30 km/h.




                                                                         28
   We have calculated standard deviation and average road grip values for each of the three
sections in each test run, data is presented in Table 11. The data is also presented with a graph
for easier overview, see Figure 33. The difference between the lowest average road grip and
the highest in Table 11 is 111%.

Table 11   Average road grip and standard deviation for RT3 G065 on runs 22 – 24 at 30 km/h.

Unit, run and Average Standard Relative standard                Distance       Distance
sub run       Road grip Deviation Deviation [%]                   start          stop
G065, 22, 1     71.3      6.7          9.4                         10            120
G065, 23, 1     71.5      6.8          9.6                         10            120
G065, 24, 1     70.9      6.7          9.5                         10            120

G065, 22, 2        38.0         7.7             20.3              140             250
G065, 23, 2        33.8         8.3             24.6              140             250
G065, 24, 2        38.1         8.3             21.9              140             250

G065, 22, 3        59.6        10.2             17.1              270             380
G065, 23, 3        61.5         9.4             15.3              270             380
G065, 24, 3        61.3        11.0             17.9              270             380

There are some significant changes between the results in Table 9 and Table 11. The average
road grip increased in sub sections 1 and 3 going from Table 9 to Table 11, while the road
grip decreased in sub section 2. The decrease in sub section 2 could be due to the fact that
rough parts of the brushed old polished ice were worn down as cars were driving over them. It
could also be due to loose snow that got dragged out onto the polished section by tires and
wind. The increased road grip could be as a result from compacting of loose snow and ice
particles on the two System 2000 ice surfaces.




                                              29
                               G065 runs 22-24 30 km/h Mean road grip with one standard deviation

                     90
                     80
                     70
                     60
         Road grip




                                                                                                                                                     High
                     50
                                                                                                                                                     Low
                     40
                                                                                                                                                     Mean
                     30
                     20
                     10
                      0
                           G065,       G065,         G065,                   G065,     G065,      G065,                  G065,       G065,   G065,
                           22, 1       23, 1         24, 1                   22, 2     23, 2      24, 2                  22, 3       23, 3   24, 3
                                                                       Vehicle, Run and sub run



Figure 33                  Average road grip with one standard deviation on run 22 - 24 for G065 at 30 km/h.


6.2.4 Test runs 70 km/h, G055
   Figure 34 illustrates the results from test runs 7 – 9 at 70 km/h on the right side of the test
track.


                                                           RT3, G055 70 km/h 18th, Run 7-9


                     100

                      90

                      80

                      70

                      60                                                                                                                RT3 G055 7
    Grip value




                                                                                                                                        RT3 G055 8
                      50
                                                                                                                                        RT3 G055 9
                      40                                                                                                                RT3 G055 average

                      30

                      20

                      10

                       0
                           0
                                27,6
                                       55,5
                                              83,3
                                                     110
                                                           136
                                                                 162
                                                                       187
                                                                              212
                                                                                     236
                                                                                           261
                                                                                                 285
                                                                                                       310
                                                                                                             335
                                                                                                                   359
                                                                                                                         383
                                                                                                                               408




                                                             Distance from start [m]


Figure 34                  RT3 G055 road grip values for runs 7 – 9 on March 18, 2008, at 70 km/h.



                                                                                      30
   RT3 G055 shows good repeatability and has good resolution in the conditions created.
Standard deviation and average road grip values have been calculated for each of the three
sections in each test run. That data is presented in Table 12 and in Figure 35. The difference
between the lowest and highest average road grip in Table 12 is 90%.

Table 12                    Average road grip values and standard deviation for RT3 G055 on runs 7 - 9 at 70 km/h.

      Unit, run, sub              Average        Standard        Relative standard Distance start Distance stop
      run                         Road grip      Deviation        Deviation [%]
      G055, 7, 1                    67.8            4.8                 7.1             10             120
      G055, 8, 1                    71.5            6.2                 8.7             10             120
      G055, 9, 1                    69.9            5.9                 8.4             10             120

      G055, 7, 2                      40.1           11               27.4                140               250
      G055, 8, 2                      37.7          13.7              36.3                140               250
      G055, 9, 2                       41           11.7              28.5                140               250

      G055, 7, 3                      57.8           9.2              15.9                270               380
      G055, 8, 3                      56.1          12.1              21.6                270               380
      G055, 9, 3                      54.7          10.5              19.2                270               380

   In Table 12 we can see that the standard deviation ranges from 11.0 to 13.7 for sub run
two. Standard deviation range for sub run three is 9.2 to 12.1. Why is the standard deviation
that large? The answer is found in Figure 34; we can clearly see that there is a large road grip
variation recorded consistently by the RT3 G055. This will increase the standard deviation
and is not a sign of uncertain measurements.


                       G055 runs 7-9 70 km/h Mean road grip with one standard deviation

               90
               80
               70
               60
   Road grip




                                                                                                            High
               50
                                                                                                            Low
               40
                                                                                                            Mean
               30
               20
               10
               0
                    G055,     G055,   G055,         G055,   G055,   G055,         G055,   G055,   G055,
                     7, 1      8, 1    9, 1          7, 2    8, 2    9, 2          7, 3    8, 3    9, 3
                                                  Vehicle, Run and sub run



Figure 35            Average road grip value with one standard deviation for RT3 G055 on runs 7 - 9 at 70 km/h.

    In Figure 35 above the standard deviation is much larger for sub run two and sub run three.
It looked like the runs for G055 were started at different positions; we shifted them to see if
they matched better. Figure 36 below show the shifted graphs and how they overlap much
better.


                                                            31
                         RT3, G055 70 km/h 18th, Run 7-9 curves shifted to match
                                                 better


               100
                90

                80
                70
                                                                                                                      RT3 G055 7
  Grip value




                60
                                                                                                                      RT3 G055 8
                50
                                                                                                                      RT3 G055 9
                40                                                                                                    RT3 G055 average
                30
                20

                10
                 0
                     0
                         21,1
                                51,2
                                       81,2
                                              110
                                                    138
                                                          166
                                                                193
                                                                      219
                                                                            246
                                                                                  272
                                                                                        299
                                                                                              325
                                                                                                    352
                                                                                                          378
                                                                                                                404
                                                    Distance from start [m]


Figure 36            Same curves as in Figure 34 with shifted start points for better view of correlation.

   There is less correlation between different curves from unit G055 then for unit G065.
There are many factors that can cause this. One is that the surfaces were not homogeneous. If
one changes the path ever so slightly the road grip values will be different. There were many
test vehicles on the test track and they changed the surface characteristics on the test track as
the test progressed.

6.2.5 Test runs 50 km/h, G055
   Figure 37 illustrates road grip on runs 10 – 12 for G055 at 50 km/h on the right side of the
test track. There is a distinct road grip increase in the brushed old polished ice on the right
side, a few meters before the 150 meter position, see Figure 37. This increase is seen in most
graphs for runs on the right side of the test track. The position varies slightly as different
drivers start the measurements at different positions. The most likely reason for the increase in
road grip is brake tests performed with the Coralba device.




                                                                            32
Figure 37   Road grip on run 10 – 12 for G055 at 50 km/h.

   The difference between the lowest average road grip and the highest is 37 units or 127%,
see Table 13.

Table 13    Average road grip with one standard deviation on run 10 – 12 for G055 at 50 km/h.

Unit, run and Average Standard Relative standard Distance Distance
sub run       Road grip deviation Deviation [%]    start    stop
G055, 10, 1     65.6       6.0         9.1          15      115
G055, 11, 1     66.7       5.9         8.9          15      115
G055, 12, 1     66.9       6.1         9.2          15      115

G055, 10, 2        32.1          6.2             19.3             145        230
G055, 11, 2        29.5          5.1             17.2             145        230
G055, 12, 2        30.8          6.6             21.5             145        230

G055, 10, 3        55.2          7.8             14.1             275        375
G055, 11, 3        55.7          8.6             15.4             275        375
G055, 12, 3        52.1         10.5             20.1             275        375




                                                33
Figure 38   Average road grip with one standard deviation on runs 10 – 12 for G055 at 50 km/h.


6.2.6 Test runs 30 km/h, G055
   Figure 39 illustrates road grip results from test runs 22 – 24 for G055 at 30 km/h on the
right side of the test track.




Figure 39   RT3 G055 test runs 22 – 24 at 30 km/h.

   It looks like run 23 (red) is shifted in this graph. Nothing else is noticed out of the ordinary.




                                                34
Table 14 shows the average road grip at 30 km/h for run 22 – 24 on March 18, 2008. The
difference between the lowest average road grip and highest is 153%.

Table 14    Average road grip values and standard deviation for RT3 G055 at 30 km/h.

Unit, run and         Average         Standard       Relative standard      Distance        Distance
sub run.              Road grip       deviation        Deviation %            start           stop
G055, 22, 1            67.92            7.97               11.73               15             115
G055, 23, 1            71.50            6.98                9.76               15             115
G055, 24, 1            67.07            7.63               11.37               15             115

G055, 22, 2              28.25           3.87              13.69               145            230
G055, 23, 2              30.80           5.41              17.58               145            230
G055, 24, 2              29.92           4.98              16.65               145            230

G055, 22, 3              66.04           7.94              12.03               275            375
G055, 23, 3              68.58           7.85              11.44               275            375
G055, 24, 3              64.68           7.19              11.12               275            375

   Figure 40 illustrates the average road grip with one standard deviation on runs 22 – 24 for
G055 at 30 km/h. There is nothing special to note, there are no significant differences between
the averages or the standard deviations.




Figure 40   Average road grip value with one standard deviation for RT3 G055 during runs 22 - 24 at 30
            km/h.


6.2.7 Test runs 70 km/h, VIAT
   These graphs are made from files produced through the software of the ViaFriction unit.
The ViaFriction unit operates with a 100 Hz sampling frequency, each data point is an
average created by a fixed distance, in this case it was set to 5 meter intervals. The “reported
distance” can be set and it is possible to produce reports for different distances to achieve

                                                35
different frequencies with the same base data. The ViaFriction device was driven on the same
side as the G055, G065 and the Coralba. However, the measuring wheel is on the driver’s side
(left side). This makes it difficult to compare directly with the two RT3 units on the right side
of the test track for runs 7 – 9 on March 18, 2008. In Figure 41 the measurements used for
calculations start at 140 meters, this is where the first “sergeant” was placed, see Figure 3. It
is very clear that the end of the brushed old polished ice is around 395 meters as the road grip
increases significantly at that location.


                                             VIAT 70 km/h 18th, Run 7-9


                   0,4
                  0,35
                   0,3
     Grip value




                  0,25                                                                     VIAT run 7
                   0,2                                                                     VIAT run 8
                  0,15                                                                     VIAT run 9

                   0,1
                  0,05
                    0
                     5
                    35
                    65
                                 95

                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                  5
                                12
                                15
                                18
                                21
                                24
                                27
                                30
                                33
                                36
                                39
                                42
                                45
                                48
                                51
                                                Distance from start [m]


Figure 41           ViaFriction runs 7 – 9 at 70 km/h.

   From the graphs in Figure 41 we can see that the ViaFriction has good correlation between
the different runs. It also shows that with the tire Trelleborg unitester 520 it is possible to see
difference between the surfaces. It is clear that the road grip is higher in the last section of the
test track. We can also see a clear road grip increase around 335 meter mark. This is possibly
due to the brake test performed by the studded winter tire equipped Coralba Mitsubishi L200.
The test tire footprint of the ViaFriction’s Trelleborg wheel can be seen in Figure 18.
   Average road grip values in Table 15 below show that the ViaFriction unit detects
differences in the different sub sections of the test track.




                                                         36
Table 15                 VIAT average road grip and standard deviations run 7 – 9 at 70 km/h.

   Unit, run,                         Average Standard Relative standard Distance Distance
   sub run                            Road grip Deviation Deviation [%]    start    stop
   VIAT, 7, 1                          0.291     0.009         3.0         140      250
   VIAT, 8, 1                          0.296     0.010         3.5         140      250
   VIAT, 9, 1                          0.300     0.016         5.2         140      250

   VIAT, 7, 2                             0.267             0.020                   7.6              270            380
   VIAT, 8, 2                             0.280             0.021                   7.4              270            380
   VIAT, 9, 2                             0.271             0.024                   8.7              270            380

   VIAT, 7, 3                             0.330             0.013                   4.1              400            510
   VIAT, 8, 3                             0.323             0.010                   3.2              400            510
   VIAT, 9, 3                             0.328             0.013                   4.1              400            490

   In Figure 42 we see that the difference between the average road grip values is small. The
ViaFriction is equipped with a tire that has no edges that will interact with the road surface in
the longitudinal direction. This decreases the sensitivity in changing roughness conditions.
The difference between the lowest average and the highest is 24%. It should be noted that the
standard deviation is very low. Remember that the standard deviation is not comparable with
an RT3 unit’s standard deviation, as the data points from the ViaFriction is an average of
sampled values over a 5 meter distance. To compare with approximately 0.19 m for the RT3s.

                                         VIAT runs 7-9 70 km/h Mean road grip with one standard deviation

                   0,4
                  0,35
                   0,3
      Road grip




                  0,25                                                                                                                 High
                   0,2                                                                                                                 Low
                  0,15                                                                                                                 Mean
                   0,1
                  0,05
                    0
                           VIAT, 7,   VIAT, 8,   VIAT, 9,           VIAT, 7,   VIAT, 8,   VIAT, 9,    VIAT, 7,   VIAT, 8,   VIAT, 9,
                              1          1          1                  2          2          2           3          3          3
                                                                    Vehicle, Run and sub run




Figure 42                ViaFriction average road grip with one standard deviation on run 7 – 9 at 70 km/h.


6.2.8 Test runs 50 km/h, VIAT
   Figure 43 illustrates the road grip on run 10 – 12 for ViaFriction at 50 km/h. The runs are
made on the right side of the test track.




                                                                        37
Figure 43   Road grip on run 10 – 12 for ViaFriction at 50 km/h.

   The transition between old System 2000 ice and brushed old polished ice is not as clear as
for the RT3s. The difference between the lowest average road grip and highest is 36%.

Table 16    Average road grip with one standard deviation on runs 10 – 12 for ViaFriction at 50 km/h.

Unit, run and Average         Standard       Relative standard       Distance      Distance
sub run       Road grip       deviation       Deviation [%]            start         stop
VIAT, 10, 1     0.25            0.018              7.10                 30           100
VIAT, 11, 1     0.28            0.015              5.26                 30           100
VIAT, 12, 1     0.28            0.022              7.94                 30           100

VIAT, 10, 2        0.21         0.017                8.04              160            230
VIAT, 11, 2        0.22         0.013                6.02              160            230
VIAT, 12, 2        0.22         0.014                6.43              160            230

VIAT, 10, 3        0.29         0.015                5.21              290            360
VIAT, 11, 3        0.30         0.019                6.36              290            360
VIAT, 12, 3        0.30         0.013                4.15              290            360




                                                38
Figure 44   Average road grip with one standard deviation on run 10 – 12 for ViaFriction at 50 km/h.

  Figure 44 illustrates the average road grip with one standard deviation. In these runs the
ViaFriction measures the road grip on old System 2000 ice as higher compared to new
System 2000 ice.

6.2.9 Test runs 30 km/h, VIAT
   Figure 45 illustrates the runs 22 – 24 from the ViaFriction unit at 30 km/h. The ViaFriction
unit was driven on the right side of the test track.




Figure 45   ViaFriction grip value for run 22 – 24 at 30 km/h.




                                                 39
   Table 17 lists the average road grip and standard deviation for runs 22 – 24 at 30 km/h.
This measurement is stable and the relative standard deviation is 3% to 11%. The difference
between the lowest and highest road grip value is 53%.

Table 17    VIAT average road grip and standard deviations runs 22 – 24 at 30 km/h.

 Unit, run       Average       Standard       Relative standard      Distance Distance
 and sub run     Road grip     deviation       Deviation [%]           start    stop
 VIAT, 22, 1       0.25          0.023              9.10                30      100
 VIAT, 23, 1       0.25          0.020              8.25                30      100
 VIAT, 24, 1       0.24          0.018              7.31                30      100

 VIAT, 22, 2         0.18         0.004              2.33               160           230
 VIAT, 23, 2         0.17         0.007              4.26               160           230
 VIAT, 24, 2         0.18         0.009              5.22               160           230

 VIAT, 22, 3         0.26         0.016              6.14               290           360
 VIAT, 23, 3         0.26         0.021              8.22               290           360
 VIAT, 24, 3         0.27         0.021              7.74               290           360

   In Figure 46 we can see that the measurements are very stable with more standard
deviation on the two System 2000 ice surfaces, this due to larger variation in the actual road
grip, see Figure 45.




Figure 46   ViaFriction average road grip with one standard deviation on runs 22 – 24 at 30 km/h.

   The average road grip has been reduced significantly compared with the tests performed
earlier at 70 km/h, see chapter 6.2.7. This is most likely due to speed sensitivity, see chapter
6.5.1 Speed dependencies for continuous devices.




                                                40
6.2.10 Test runs 70 km/h, CORA
   The Coralba equipped Mitsubishi L200 measured the following road grip values on the
right side of the test track during runs 7 – 9 at 70 km/h.

Table 18                 Coralba road grip on runs 7-9 at 70 km/h.

                                                                  Brushed old
Side, speed                           Old System 2000 ice         polished ice    New System 2000 ice
Right, 70 km/h                               N/A                      0.14               0.27
Right, 70 km/h                               0.22                     0.13               0.24
Right, 70 km/h                               0.21                     0.08               0.28
Average road grip                            0.22                     0.12               0.26
Standard deviation                          0.0071                  0.0321              0.0208

   The results show significantly lower road grip in the second section. As expected the road
grip is slightly higher on the new System 2000 ice than on the old System 2000 ice. The
difference from the lowest to the highest road grip value is 0.2 units or 250%. Note that the
0.08 value is one of the lowest values recorded during the whole test and compared with the
two others from that section it is likely that it is an outlier.


                                                CORA, 70 km/h 18th, Run 7 - 9

                        0,30

                        0,25
      Road grip value




                        0,20
                                                                                            Old 2000 ice
                        0,15                                                                Brushed old polished ice
                                                                                            New 2000 ice
                        0,10

                        0,05

                        0,00
                                     7                  8                  9
                                                      Run



Figure 47                Road grip values for Coralba equipped Mitsubishi L200 on right side of test track.

   Note that in Figure 47 there was one missing measuring point for old System 2000 ice on
run 7.

6.2.11 Test runs 50 km/h, CORA
   The Coralba equipped Mitsubishi L200 measured the following road grip values on the
right side of the test track during runs 10 – 13 at 50 km/h.




                                                             41
Table 19                       Coralba road grip on runs 10 – 12 at 50 km/h.

                                                                Brushde old
Side, speed                           Old System 2000 ice       polished ice    New System 2000 ice
Right, 50km/h                                0.28                   0.13               0.27
Right, 50km/h                                0.28                   0.13               0.29
Right, 50km/h                                0.30                   0.11               0.26
Average road grip                            0.29                   0.12               0.27
Standard deviation                          0.0115                0.0115              0.0153

   The results show significantly lower road grip in the second section. The road grip has
changed and is now higher for old System 2000 ice. The difference from the lowest to the
highest road grip value is 0.19 units or 172%.


                                                CORA, 50 km/h 18th, Run 10 - 12

                        0,35

                        0,30

                        0,25
      Road grip value




                        0,20                                                             Old 2000 ice
                                                                                         Brushed old polished ice
                        0,15                                                             New 2000 ice
                        0,10

                        0,05

                        0,00
                                      10                 11                12
                                                        Run



Figure 48                Road grip values on runs 10 – 12, for Coralba at 50 km/h.

  Figure 48 illustrates the road grip on runs 10 – 12, there is a small difference between old-
and new System 2000 ice.

6.2.12 Test runs 30 km/h, CORA
   The Coralba equipped Mitsubishi L200 measured the following road grip values on the
right side of the test track during runs 22 – 24 at 30 km/h.




                                                               42
Table 20                 Coralba road grip for 30 km/h, from runs 22 – 24.

                                                               Brushed old
Side, speed                         Old System 2000 ice        polished ice      New System 2000 ice
Right, 30km/h                              0.21                    0.12                 0.21
Right, 30km/h                              0.21                    0.11                 0.26
Right, 30km/h                              0.20                    0.10                 0.20
Average road grip                          0.21                    0.11                 0.22
Standard deviation                        0.0058                 0.0100                0.0321

   The results show significantly lower road grip in the second section. As expected the road
grip is slightly higher on the new System 2000 surface than on the old one. The difference
from the lowest to the highest road grip value is 0.16 units or 160%.


                                              CORA, 30 km/h 18th, Run 22 - 24

                        0,30

                        0,25
      Road grip value




                        0,20
                                                                                          Old 2000 ice
                        0,15                                                              Brushed old polished ice
                                                                                          New 2000 ice
                        0,10

                        0,05

                        0,00
                                    22                 23                24
                                                      Run



Figure 49                Road grip values for Coralba runs 22 – 24 at 30 km/h.

    In Figure 49 we see that there is one higher value in run 23 on the new System 2000 ice, all
others seem ok. There is no reason to believe that the value is incorrect as differences exist on
all surfaces.

6.2.13 Test runs 70 km/h, G053
   Figure 50 illustrates the results from test runs with G053 at 70 km/h on the left side of the
test track. This is the third test speed on March 18, 2008, with 30 km/h and 50 km/h runs
made previously.




                                                             43
                                                        RT3, G053 70 km/h 18th, Run 7-9


                  90
                  80
                  70
                  60
     Grip value




                                                                                                                                  RT3 G053 7
                  50                                                                                                              RT3 G053 8
                  40                                                                                                              RT3 G053 9
                  30                                                                                                              RT3 G053 average

                  20
                  10
                  0
                       0
                           25,1
                                   51,7
                                          79,2
                                                 108
                                                       135
                                                             163
                                                                   190
                                                                         217
                                                                               245
                                                                                     272
                                                                                           299
                                                                                                 326
                                                                                                       352
                                                                                                              378
                                                                                                                    404
                                                                                                                          429
                                                         Distance from start [m]


Figure 50              Road grip on runs 7 – 9 for RT3 G053 at 70 km/h.

   This RT3 G053 has good repeatability and has good resolution in the conditions created.
We have calculated relative standard deviation and average road grip values for each of the
three sections in each test run; data is presented in Table 21. The difference between the
lowest and highest average road grip is 33 units or 97%.

   Table 21                Average road grip values and standard deviation on runs 7 – 9 for G053 at 70 km/h.

Unit, run and                     Average Standard Relative standard Distance Distance
sub run                           Road grip deviation Deviation [%]    start    stop
G053, 7, 1                         63.08       9.28      14.71          15      115
G053, 8, 1                         66.90      10.16      15.19          15      115
G053, 9, 1                         64.87       9.24      14.24          15      115

G053, 7, 2                         41.23                4.14                    10.04                        145            230
G053, 8, 2                         34.00                3.89                    11.43                        145            230
G053, 9, 2                         35.79                4.73                    13.23                        145            230

G053, 7, 3                         50.51                5.69                    11.26                        275            375
G053, 8, 3                         49.58                7.07                    14.26                        275            375
G053, 9, 3                         52.67               10.04                    19.06                        275            375

   The data is also presented with a graph for easier overview, see Figure 51.




                                                                               44
   Figure 51   Average road grip value with one standard deviation on runs 7 – 9 for G053 at 70 Km/h.

   In Figure 51 we see some differences in the average value for sub run two in run 7. The
causes for this can be seen in Figure 50, see the blue curve at 150 meter. This could be due to
a shift in the starting point for the measurement on run 7.

6.2.14 Test runs 50 km/h, G053
   Figure 52 illustrates the road grip on test runs 7 – 9 for RT3 G053 at 50 km/h. Look
carefully at the yellow graph, around 180 – 200 meters on run 12, we can see a deviation that
also was detected by G056 on the 11th run at 50 km/h, see Figure 58. Maybe some unit
dropped accumulated snow onto the ice and this snow increased the road grip for a couple of
runs. The snow that may have fallen onto the ice could have contained sand as there were
some tests made on regular roads that morning. The difference between the lowest and highest
average road grip is 34 units or 107%.




                                               45
Figure 52   Road grip on runs 10 – 12 for RT3 G053 at 50 km/h.



Table 22    Average road grip with one standard deviation on runs 10 – 12 for G053 at 50 km/h.

Unit, run       Average Standard Relative standard Distance Distance
and sub run     Road grip deviation Deviation [%]    start    stop
G053, 10, 1       65.6       7.4        11.3          15      115
G053, 11, 1       65.5       7.7        11.8          15      115
G053, 12, 1       65.1       8.5        13.0          15      115

G053, 10, 2        31.7          3.2             10.0            145        230
G053, 11, 2        33.8          2.7              8.1            145        230
G053, 12, 2        37.1          4.2             11.2            145        230

G053, 10, 3        51.0          8.0             15.7            275        375
G053, 11, 3        54.3          7.6             13.9            275        375
G053, 12, 3        53.7          7.2             13.4            275        375




                                                46
Figure 53                Average road grip with one standard deviation on runs 10 – 12 for G053 at 50 km/h.

   We can clearly see in Figure 53 how the average road grip increased due to the deviation in
run 12. The standard deviation also increased for run 12. The standard deviation is larger in
the last sub run due to decreasing road grip throughout sub run 3, see Figure 52.

6.2.15 Test runs 30 km/h, G053
   Figure 54 illustrates the results from test runs 22 - 24 with G053 at 30 km/h on the right
side of the test track.


                                                       RT3, G053 30 km/h 18th, Run 22-24


                         90

                         80

                         70

                         60
                                                                                                                   RT3 G053 22
            Grip value




                         50                                                                                        RT3 G053 23
                         40                                                                                        RT3 G053 24
                                                                                                                   RT3 G053 average
                         30

                         20

                         10

                          0
                              0
                                  29
                                       59
                                            89
                                                 119
                                                       148
                                                             177
                                                                   205
                                                                         234
                                                                               261
                                                                                     289
                                                                                           316
                                                                                                 344
                                                                                                       372
                                                                                                             401




                                                         Distance from start [m]


Figure 54                Road grip on runs 22 – 24 for RT3 G053 at 30 km/h.




                                                                               47
   We have three runs where the RT3 detect distinct differences in road grip. The problem
with lack of constant frequency of data points are seen as the curves are starting to shift from
each other, which can be noticed by indicative bumps in the graph. See the later part of sub
run two in Figure 54. There we have three distinct measurements where a difference in road
grip is detected by the RT3 unit and it can be seen how they deviate from each other due to
the drift in sampling frequency.
   The standard deviation and average road grip values for each of the three sub runs in each
run is presented in Table 23. The difference between lowest and highest average road grip is
26.1 units or 66%.

   Table 23     Average road grip and standard deviation on runs 22 - 24 for RT3 G053 at 30 km/h.

   Unit, run and      Average Standard Relative standard              Distance        Distance
   sub run            Road grip Deviation Deviation %                   start           stop
   G053, 7, 1           63.2      6.2         9.8                        10             120
   G053, 8, 1           64.9      5.5         8.5                        10             120
   G053, 9, 1           65.7      5.5         8.3                        10             120

   G053, 7, 2            39.6         7.3             18.5               140             250
   G053, 8, 2            41.6         7.8             18.8               140             250
   G053, 9, 2            41.0         5.8             14.1               140             250

   G053, 7, 3            50.2         9.9             19.7               270             380
   G053, 8, 3            49.4         8.3             16.9               270             380
   G053, 9, 3            53.7        11.0             20.5               270             380

   Figure 55 illustrates the average road grip with one relative standard deviation on runs 22 –
24 for G053 at 30 km/h. We can see that the standard deviation is large for the last sub run on
new System 2000 ice. In Figure 54 we see that the road grip is decreasing throughout the last
sub run on new System 2000 ice and is causing the relative standard deviation to increase. We
have looked carefully at this trend for all continuous measuring devices and it is seen often.
One explanation for this could be that the sand contamination in the last sub run warms up,
warming the ice, and making it softer especially to slight lateral motions, see Figure 3. The
sand had melted down into the ice and that is the reason why we don’t believe that it would
contribute to the road grip.




                                               48
                             G053 runs 22-24 30 km/h Mean road grip with one standard deviation

                    80

                    70

                    60

                    50
        Road grip




                                                                                                               High
                    40                                                                                         Low
                                                                                                               Mean
                    30

                    20

                    10

                     0
                          G053,   G053,   G053,          G053,    G053,   G053,       G053,   G053,   G053,
                          22, 1   23, 1   24, 1          22, 2    23, 2   24, 2       22, 3   23, 3   24, 3
                                                       Vehicle, Run and sub run



   Figure 55                Average road grip value with one standard deviation on runs 22 - 24 for RT3 G053 at 30
                            km/h.


6.2.16 Test runs 70 km/h, G056
   Figure 56 illustrates the results from test runs 7 - 9 with RT3 G056 at 70 km/h on the left
side of the test track.


                                              RT3, G056 70 km/h 18th, Run 7-9


                    100
                     90
                     80
                     70
     Grip value




                                                                                                  RT3 G056 7
                     60
                                                                                                  RT3 G056 8
                     50
                                                                                                  RT3 G056 9
                     40
                                                                                                  RT3 G056 average
                     30
                     20
                     10
                      0
                                             4
                                             0
                                             6
                                             2
                                             7
                                             3
                                             8
                                             3
                                             8
                                             2
                                             6
                                             9
                       0
                      24
                      48
                                  73
                                  99
                                          12
                                          15
                                          17
                                          20
                                          22
                                          25
                                          27
                                          30
                                          32
                                          35
                                          37
                                          39




                                                  Distance from start [m]


Figure 56                Road grip on runs 7 – 9 for RT3 G056 at 70 km/h.

   This RT3 G056 has good repeatability and has a good range in the conditions created.
Standard deviation and average road grip values for each of the three sections in each test run



                                                             49
are presented in Table 24. The data is also illustrated with a graph for easier overview, see
Figure 57. The difference between the lowest and highest road grip value is 94%.

Table 24               Average road grip values and standard deviation on runs 7 – 9 for G056 at 70 km/h.

   Unit, run and                 Average        Standard         Relative standard
   sub run                       Road grip      Deviation         Deviation [%] Distance start Distance stop
   G056, 7, 1                      60.4            9.5                 15.8          10             120
   G056, 8, 1                      65.2            8.6                 13.2          10             120
   G056, 9, 1                      65.3           10.7                 16.4          10             120

   G056, 7, 2                       36.6           10.1                  27.6           140                   250
   G056, 8, 2                       33.7            9.2                  27.3           140                   250
   G056, 9, 2                       38.0           11.5                  30.3           140                   250

   G056, 7, 3                       54.2            8.3                  15.2           270                   380
   G056, 8, 3                       55.4            9.1                  16.4           270                   380
   G056, 9, 3                       58.8           10.3                  17.4           270                   380



                            G056 runs 7-9 70 km/h Mean road grip with one standard deviation

                  80

                  70

                  60

                  50
      Road grip




                                                                                                                    High
                  40                                                                                                Low
                                                                                                                    Mean
                  30

                  20

                  10

                  0
                        G056,   G056,   G056,          G056,     G056,   G056,       G056,    G056,   G056,
                         7, 1    8, 1    9, 1           7, 2      8, 2    9, 2        7, 3     8, 3    9, 3
                                                     Vehicle, Run and sub run



Figure 57              Average road grip value with one standard deviation for G056 run 7 – 9 at 70 km/h.

   In Figure 57 we see a stable measurement, only some variations in standard deviation and
average road grip value between the different runs.

6.2.17 Test runs 50 km/h, G056
   Figure 58 illustrates the results from test runs 10 - 12 with RT3 G056 at 50 km/h on the left
side of the test track.




                                                            50
Figure 58   Road grip on runs 10 – 12 with G056 at 50 km/h.

   There is a large deviation in the 11th run on sub run 2 (red curve) around 200 meter. We
did notice some small deviations in run 12 for RT3 G056 as well, measuring on this side of
the track, see Figure 52. The detection is not as large for G053, however the location is the
same. Possible reasons why the detection is less distinct could be that G053 only partially
drives over the disturbed surface. Another reason could be that the two TWO units have
passed the area between the two RT3s. Maybe some slush dropped from one of the devices.
This debris could then have been gradually spread over the surface, making the effects
decrease over the next run. The difference between the lowest and highest average road grip is
124%.

Table 25    Average road grip values and standard deviation on runs 10 – 12 with G056 at 50 km/h.

Unit, run and    Average Standard Relative standard Distance Distance
sub run          Road grip deviation Deviation [%]    start    stop
G056, 10, 1       66.17      11.30      17.08          15      115
G056, 11, 1       66.32       8.50      12.82          15      115
G056, 12, 1       67.00      10.50      15.68          15      115

G056, 10, 2         30.49        3.99             13.09            145        230
G056, 11, 2         42.44        9.74             22.96            145        230
G056, 12, 2         29.96        3.54             11.82            145        230

G056, 10, 3         54.89         8.82            16.07            275        375
G056, 11, 3         55.66         7.98            14.33            275        375
G056, 12, 3         57.26        10.42            18.20            275        375




                                               51
Figure 59   Average road grip value with one standard deviation for G056 runs 10 – 12 at 50 km/h.

   As we can see, the road grip disturbance in run 11 shows up very distinctly in Figure 59.

6.2.18 Test runs 30 km/h, G056
   Figure 60 illustrates the results from test runs 22 - 24 for RT3 G056 at 30 km/h on the left
side of the test track.




Figure 60   Road grip on runs 22– 24 for G056 at 30 km/h.

   Average road grip and relative standard deviation is presented in Table 26. The difference
between the lowest and highest average road grip is 124%.




                                                52
Table 26    Average road grip values and standard deviation on runs 22 – 24 with G056 at 30 km/h.


Unit, run and     Average Standard Relative standard             Distance Distance
sub run           Road grip deviation Deviation [%]                start    stop
G056, 22, 1        68.44      6.57        9.60                      15      115
G056, 23, 1        68.53      5.72        8.34                      15      115
G056, 24, 1        67.48      5.30        7.85                      15      115

G056, 22, 2         39.58         4.92            12.44             145          230
G056, 23, 2         38.28         5.90            15.40             145          230
G056, 24, 2         39.44         5.77            14.63             145          230

G056, 22, 3         52.61        11.04            20.99             275          375
G056, 23, 3         53.40        10.09            18.90             275          375
G056, 24, 3         54.03         9.39            17.39             275          375

   In Figure 61 and in Table 26 we see that the relative standard deviation is high in sub run
three, this can also be seen in Figure 60, it is clear that the road grip changes drastically over
the interval, causing the increase in relative standard deviation.




Figure 61   Average road grip value with one standard deviation for G056 runs 22 – 24 at 30 km/h.


6.2.19 Test runs 70 km/h, TWOV
    The tires on the Pon-Cat Two are Trelleborg Industries model 523. The tread pattern can
be seen in Figure 17. No winter tire rubber is available for this tire model, there are some
difficulties associated with the kind of rubber used for industrial tires. The most important one
is that the dynamic range of measurement decrease.



                                                53
   Pon-Cat TWOV shows good repeatability, see Figure 62 below. We can also see that this
Pon-Cat unit indicates that the last section of the test track has higher road grip than the first
section. That is what is expected in high longitudinal slip condition as the surface is sharper
than old System 2000 ice in the first section. Results from the Coralba equipped Mitsubishi
also confirmed this. These set of runs are the worst logged.




Figure 62   Road grip on runs 7 – 9 for Pon-Cat TWO, TWOV at 70 km/h, left side.

   The standard deviation shows high differences for each section. In the first sub run of the
seventh run we see a relative standard deviation of 21.5%, compared to 4.9% and 5.2% for the
two other runs. See Table 27 and in Figure 63 below. In the second sub run the relative
standard deviation is 9.1% – 15.7%. In the third section we see variations from 2.9% to 8.3%.
In Figure 62 above we can see two large deviations; the first is in run 7 at 170 meters. The
second is in run 8 at 274 meter. It is hard to tell why these variations have occurred; maybe
there were some real road grip differences at some local spots that this particular unit passed
on these runs.
   It should be noted that the standard deviation is based on data points that represent an
average of several data points that originally were recorded at 100 Hz. The reader should note
that these standard deviation values are not directly comparable with the standard deviations
calculated for the RT3s as there were less averaging made on the RT3 data. The difference
between lowest and highest average road grip, seen in Table 27, is 50%.




                                               54
Table 27        Average road grip values and standard deviation on runs 7 – 9 for TWOV at 70 km/h.


Unit, run and    Average       Standard      Relative standard Distance        Distance
sub run          Road grip     Deviation      Deviation [%]      start           stop
TWOV, 7, 1         0,16          0,03              21,52          30             100
TWOV, 8, 1         0,21          0,01               4,96          30             100
TWOV, 9, 1         0,20          0,01               5,20          30             100

TWOV, 7, 2          0,21          0,03              15,71            160          230
TWOV, 8, 2          0,18          0,02               9,10            160          230
TWOV, 9, 2          0,18          0,03              14,16            160          230

TWOV, 7, 3          0,23          0,02              8,28             290          360
TWOV, 8, 3          0,23          0,01              4,04             290          360
TWOV, 9, 3          0,24          0,01              2,88             290          360

   In Figure 63 the average road grip values have low variation for the different sub runs.
Average road grip value for the first sub run/section is very similar as those for the second sub
run. The road grip is slightly higher in the third sub run.




Figure 63   Average road grip value with one standard deviation for TWOV runs 7 – 9 at 70 km/h.

   Tests indicate that the devices using Trelleborg industrial type of tires would benefit from
getting tires made with winter tire rubber. The dynamic range measuring winter road
conditions would increase, as the hysteresis effects would be more noticeable.




                                               55
6.2.20 Test runs 50 km/h, TWOV
   Figure 64 illustrates road grip on runs 10 – 12. We see some increased road grip on the
brushed old polished ice on run 10; this could be an indication of the road grip issue that RT3s
G053 and G056 detected in later runs.




Figure 64   Road grip on runs 10 – 12 for TWOV at 50 km/h, left side.

   The difference between lowest and highest average road grip, seen in Table 28, is 56%.
Apart from run 10, sub run 2, the average values are relatively stable and as expected, with
the old System 2000 ice having slightly lower grip than new System 2000 ice, and brushed
old polished ice being the most slippery.

Table 28    Average road grip values and standard deviation on runs 10 – 12 for TWOV at 50 km/h.

Unit, run and     Average Standard          Relative standard     Distance Distance
sub run           Road grip deviation        Deviation [%]          start    stop
TWOV, 10, 1         0.20     0.0078               3.88               30      100
TWOV, 11, 1         0.20     0.0101               5.03               30      100
TWOV, 12, 1         0.22     0.0152               6.99               30      100

TWOV, 10, 2          0.18       0.0108               5.85               160   230
TWOV, 11, 2          0.16       0.0123               7.80               160   230
TWOV, 12, 2          0.16       0.0076               4.71               160   230

TWOV, 10, 3          0.24       0.0105               4.45               290   360
TWOV, 11, 3          0.23       0.0043               1.86               290   360
TWOV, 12, 3          0.25       0.0073               2.91               290   360

   Figure 65 illustrates standard deviation with one standard deviation.




                                                56
Figure 65   Average road grip with one standard deviation on runs 10 – 12 for TWOV at 50 km/h.


6.2.21 Test runs 30 km/h, TWOV
   Figure 66 illustrates road grip on runs 22 – 24. The TWO units have moved over to the
right side of the test track. We see some differences in road grip between runs 22 – 24, this is
most likely effects from brake test performed by the Coralba unit.




Figure 66   Road grip on runs 22 – 24 for TWOV at 30 km/h, right side.

   The difference between lowest and highest average road grip, seen in Table 29, is 80%.


                                               57
Table 29    Average road grip values and standard deviation on runs 22 – 24 for TWOV at 30 km/h.

 Unit, run and       Average        Standard         Relative standard    Distance Distance
   sub run           Road grip      deviation          Deviation %          start    stop
TWOV, 22, 1            0.23          0.0104                 4.49             30      100
TWOV, 23, 1            0.21          0.0236                11.25             30      100
TWOV, 24, 1            0.23          0.0225                 9.98             30      100

TWOV, 22, 2             0.15         0.0121                7.85              160        230
TWOV, 23, 2             0.18         0.0134                7.33              160        230
TWOV, 24, 2             0.18         0.0163                9.13              160        230

TWOV, 22, 3             0.27         0.0250                9.16              290        360
TWOV, 23, 3             0.26         0.0210                8.13              290        360
TWOV, 24, 3             0.24         0.0186                7.58              290        360

   Figure 67 illustrates average road grip values with one standard deviation.




Figure 67   Average road grip with one standard deviation on runs 22 – 24 for TWOV at 30 km/h.


6.2.22 Test runs 70 km/h, TWOT
   Pon-Cat TWOT shows good repeatability, see Figure 68. This measurement indicates that
the last section of the test track has higher road grip then the first section. There is noticeable
difference in road grip between the different sub sections.




                                                58
Figure 68   Pon-Cat TWO, TWOT at 70 km/h, left side.

   From Table 30 it can be gathered that the standard deviation was 9.0 – 10.6% for the first
sub run; the road grip is relatively stable. In the second section sub run the standard deviation
increases to 16.3 – 20.0%. The increase in standard deviation is most likely due to changes in
road grip, see Figure 68. In the last sub run the standard deviation is 4.5 – 6.9%. The
difference between lowest and highest average road grip is 56%.

Table 30    Average road grip values and standard deviation on runs 7 – 9 for TWOT.

Unit, run and     Average Road       Standard         Standard
sub run               grip           Deviation      Deviation [%] Distance start Distance stop
 TWOT, 7, 1           0.19            0.022            11.15           30             100
 TWOT, 8, 1           0.20            0.016             7.75           30             100
 TWOT, 9, 1           0.20            0.016             8.09           30             100

 TWOT, 7, 2            0.17            0.015             9.03              160           230
 TWOT, 8, 2            0.16            0.018             10.92             160           230
 TWOT, 9, 2            0.15            0.017             11.18             160           230

 TWOT, 7, 3            0.25            0.019             7.67              290           360
 TWOT, 8, 3            0.24            0.013             5.26              290           360
 TWOT, 9, 3            0.25            0.018             7.09              290           360

   Figure 69 illustrates that the average road grip values and standard deviations are different
between the sub sections.




                                               59
Figure 69   Average road grip value with one standard deviation for TWOT runs 7 – 9 at 70 km/h.


6.2.23 Test runs 50 km/h, TWOT
   Pon-Cat TWOT shows good repeatability, see Figure 70. This measurement indicates that
the last section of the test track has higher road grip then the first section. There is noticeable
difference in road grip between the different sub sections. Run 11 were omitted due to similar
names.




Figure 70   Pon-Cat TWO, TWOT at 50 km/h, left side.

   The difference between lowest and highest average road grip is 85%.


                                               60
Table 31    Average road grip values and standard deviation on runs 10 – 12 for TWOT.

Unit, run and Average          Standard      Relative standard    Distance      Distance
sub run       Road grip        Deviation      Deviation [%]         start         stop
 TWOT, 10, 1    0,19            0,0125             6,52              30           100
 TWOT, 11, 1    N/A              N/A               N/A              N/A           N/A
 TWOT, 12, 1    0,20            0,0174             8,77              30           100

 TWOT, 10, 2         0,14        0,0069             5,09             160          230
 TWOT, 11, 2         N/A          N/A               N/A              N/A          N/A
 TWOT, 12, 2         0,13        0,0107             8,11             160          230

 TWOT, 10, 3         0,23        0,0108             4,62             290          360
 TWOT, 11, 3         N/A          N/A               N/A              N/A          N/A
 TWOT, 12, 3         0,24        0,0117             4,85             290          360

   Figure 71 illustrates that the average road grip with one standard deviation.




Figure 71   Average road grip value with one standard deviation for TWOT runs 10 and 12 at 50 km/h.


6.2.24 Test runs 30 km/h, TWOT
   Pon-Cat TWOT shows good repeatability, see Figure 72. This measurement indicates that
the last sub run/section of the test track has higher road grip than the first sub run. There is
noticeable difference in road grip between the different runs locally. This is most likely due to
brake test with the Coralba unit.




                                               61
Figure 72   Pon-Cat TWO, TWOT at 30 km/h, right side.

   The difference between lowest and highest average road grip in Table 32 is 67%.

Table 32    Average road grip values and standard deviation on runs 22 – 24 for TWOT.

Unit, run and Average          Standard     Relative standard     Distance      Distance
sub run       Road grip        Deviation     Deviation [%]          start         stop
 TWOT, 22, 1    0,22            0,0188             8,46              30           100
 TWOT, 23, 1    0,24            0,0260            11,05              30           100
 TWOT, 24, 1    0,22            0,0356            16,08              30           100

 TWOT, 22, 2         0,15        0,0151             10,06            160          230
 TWOT, 23, 2         0,15        0,0152             10,15            160          230
 TWOT, 24, 2         0,15        0,0115              7,73            160          230

 TWOT, 22, 3         0,25        0,0230              9,35            290          360
 TWOT, 23, 3         0,23        0,0336             14,66            290          360
 TWOT, 24, 3         0,22        0,0301             13,72            290          360

   Figure 73 illustrates that the average road grip with one standard deviation.




                                               62
Figure 73   Average road grip value with one standard deviation for TWOT runs 22 - 24 at 30 km/h.


6.3   Road grip development during March 18, 2008
   One objective for the CASTT project is to investigate measurement techniques that will
enable forecasts of road grip. This requires reliable measurements over many controlled
conditions. In an effort to understand this we assembled the data from the different test runs of
Tuesday to try and see if any trends could be noticed.

6.3.1 Comparison of RT3 measurement changes on old System 2000 ice
   Two of the vehicles pulling RT3s, G053 and G056 were equipped with friction tires and
drove on the left side of the test track. The other two RT3’s vehicles, G055 and G065, had
studded tires and drove on the right side. It is hard to see a significant difference in the change
of road grip between the different sides of the test track on old System 2000 ice, see Figure
74. Maybe stud wear increased road grip a little, but this could also be because more vehicles
used the right side of the track. Another possibility is effects from braking tests performed
with the Coralba equipped unit.




                                               63
Figure 74   Comparison of changes on sub section one, caused by friction tires on the left side. Measured
            by G053 and G056. Compared with changes of sub section one on the right side by studded
            tires measured by G055 and G065. All measurements are presented with one standard
            deviation borders.

   On runs 19 – 27 the TWO units switched from driving on the left side of the track to the
right side, see Table 4. Significant changes appear in road grip measurements on sub section
two after the switch.

6.3.2 Comparison of RT3 measurement changes on brushed old polished ice
   Figure 75, illustrates road grip on the brushed old polished ice, and shows two different
trends. The left side shows an increase in road grip over the day, the right side shows a
decrease in road grip.




Figure 75   Comparison of changes on sub section two, caused by friction tires on the left side. Measured
            by G053 and G056. Compared with changes of sub section two on the right side by studded
            tires measured by G055 and G065. All data is presented with one standard deviation borders.



                                                 64
   The reasons for this can be numerous, one example is that the ice is worn down by the
studs scaring the surface, ice and snow particles torn from the ice alter the measurements.

6.3.3 Comparison of normalized road grip on brushed old polished ice
   During March 18, 2008 the TWO devices measured on the left side of the test track most
of the day. For comparisons between the right and left side of the test track the last nine runs
where made on the right side. Because general trends are most important, all values were
normalized to average road grip values, giving the same centre point. It can be seen in Figure
76 that the general trends do not deviate visibly. Road grip measurements made after the
TWOs moved over to the right side of the test track show significantly higher grip.




Figure 76   Normalized average road grip comparison on brushed old polished ice between TWOs
            TWOT and TWOV to RT3s G053 and G056, on the left side of the test track, March 18, 2008.
            All data is presented with one standard deviation borders.


6.3.4 Comparison of RT3 changes on new System 2000 ice
   Measurements on old System 2000 ice indicate a change by the end of the day; see Figure
77. While road grip on the left side of the track is about the same, an increase can be seen in
the road grip on the right side.




                                              65
Figure 77   Comparison of changes on sub section three, caused by friction tires on the left side.
            Measured by G053 and G056. Compared with changes of sub section three on the right side
            by studded tires measured by G055 and G065. All data is presented with one standard
            deviation borders.

  Over the entire day, a difference in how average road grip was measured in subsection one
and three could be noted between the longitudinal and transverse slip devices.

6.4   Relative road grip comparison of different devices
   The most prominent difference between the devices was how they measured road grip on
old System 2000 ice and new System 2000 ice. We compared all different speeds, normalized
to brushed old polished ice, since brushed old polished ice was assumed to have no directional
bias. Observing results in Figure 78, Figure 79 and Figure 80 the RT3s reports lower road grip
on new System 2000 ice compared to old System 2000 ice, the TWOs and ViaFriction
reported higher road grip on new System 2000 ice compared to old System 2000 ice.




                                               66
Figure 78   Relative road grip comparison normalized for brushed old polished ice for all systems with
            presentable data at 30 km/h.




Figure 79   Relative road grip comparison normalized for brushed old polished ice for all systems at 50
            km/h.




                                                67
Figure 80   Relative road grip comparison normalized for brushed old polished ice for all systems at 70
            km/h.

   The viable reason is that the TWOs and the ViaFriction measures road grip through high
longitudinal slip, the RT3 measures road grip through low lateral slip. The longitudinal slip
rate for the TWOs is about 18% and about 20% for ViaFriction. The slip rate is adjustable on
the ViaFriction unit and is set to 20% as this is the standard slip rate used in Norway. The
lateral slip rate of the RT3 is much less. The RT3s measures a larger relative difference
between the surfaces, compared to TWOs and ViaFriction. This can not be attributed to the
larger scale, since all values were normalized.

6.5   Speed dependency
   The two parameters with dominating effect on rubber behavior are temperature and
frequency [3]. As the frequency of which the rubber is excited is changed the characteristics
of the rubber change dramatically for some conditions. Every time the exciting frequency
changes with 10 times, like from 100 Hz to 1000 Hz rubber behaves as if the temperature
dropped 7 - 8 °C. If the rubber behaves as chilled it will get harder and hysteresis will not
work as well, this transition temperature generally happens very rapidly. We analyzed the data
to see if the conditions we tested at during March 18, 2008 showed indication that we had
passed into the glassy state of rubber.

6.5.1 Speed dependencies for continuous devices
   The data in this section are average road grip values for the three test speeds during March
18, 2008. Data is presented with one standard deviation. We have been informed that the
internal friction in the ViaFriction was not correctly compensated for during these tests. These
settings have been corrected at the time of the publication of this report.

6.5.1.1 Old System 2000 ice
    We compared average road grip on old System 2000 ice depending on speed, for all
devices and runs during March 18, 2008. Observing Figure 81 the RT3 devices winter friction
tire did not show any significant trends in this comparison. The TWOs did not show a


                                                68
significant speed dependency. ViaFriction show speed dependency, as the road grip increases
with increased speed. ViaTech state that this problem was due to internal friction and that it
has been corrected.




Figure 81   RT3’s G053, G055, G056, G065 as well as TWOT, TWOV and VIAT with relevant runs over
            the entire day on old System 2000 ice, sorted by speed. All measurements are presented with
            one standard deviation.




                                                69
6.5.1.2 Brushed old polished ice
   We compared average road grip on brushed old polished ice depending on speed, for all
devices and runs during March 18, 2008. As evident in Figure 82 most devices is unaffected
by speed on this surface. ViaFriction show speed dependency ViaTech has updated the unit to
compensate for inner friction that was the cause of this discrepancy.




Figure 82   RT3s G053, G055, G056, G065 as well as TWOT, TWOV and VIAT with relevant runs over
            the entire day on brushed old polished ice, sorted by speed. All measurements are presented
            with one standard deviation.


                                                70
6.5.1.3 New System 2000 ice
   We compared average road grip on new System 2000 ice depending on speed, for all
continuous devices and runs during March 18, 2008. As evident in Figure 83 most devices is
unaffected by speed. ViaFriction show speed dependency.




Figure 83   RT3s G053, G055, G056, G065 as well as TWOT, TWOV and VIAT with relevant runs over
            the entire day on new System 2000 ice, sorted by speed. All measurements are represented
            with one standard deviation.


                                               71
6.5.2 Speed dependencies of braking devices
   Some different reactions can be seen for different car/tire systems. Braking was not to a
full stop, the deceleration is mainly from the testing speed and lower.

6.5.2.1 Old System 2000 ice
   As seen in Figure 84 and Figure 85 displaying the average of viable measurements from
VBOX and CORA respectively, no dominant trend can be seen. Both devices however report
slightly better braking ability at 50 km/h compared to the other speeds.


                                        VBOX measurements on old System 2000

                   0,35

                    0,3

                   0,25
      Road grip




                    0,2

                   0,15

                    0,1

                   0,05

                     0
                                  30 km/h                      50 km/h                       70 km/h
                                                                Speed



Figure 84            Road grip values for VBox on old System 2000 ice at different speeds, with one standard
                     deviation.


                                  Grip values on old System 2000 for Coralba

                    0,30

                    0,25

                    0,20
        Friction




                    0,15

                    0,10

                    0,05

                    0,00
                                   30 km/h                      50 km/h                     70 km/h
                                                                Speed



Figure 85            Road grip values for Coralba on old System 2000 ice at different speeds, with one standard
                     deviation.



                                                         72
6.5.2.2 Polished ice
   Figure 86 and Figure 87 display the average of viable measurements from VBOX and
CORA respectively. No dominant trend can be seen for the CORA. The VBOX indicated
higher road grip with increased speed on the polished ice something that correlates with the
braking tests in chapter 6.7. Note that this ice is new, made by a Zamboni ice machine and
have much lower friction then the brushed old polished ice from the day before.


                                              VBOX measurements on polished ice

                  0,16

                  0,14

                  0,12

                      0,1
      Road grip




                  0,08

                  0,06

                  0,04

                  0,02

                       0
                                    30 km/h                       50 km/h                       70 km/h
                                                                  Speed



Figure 86              Road grip values for VBox on polished ice at different speeds, with one standard deviation.


                                       Grip values on Polished ice for Coralba

                      0,14

                      0,12

                      0,10
           Friction




                      0,08

                      0,06

                      0,04

                      0,02

                      0,00
                                      30 km/h                     50 km/h                      70 km/h
                                                                   Speed



Figure 87              Road grip values for Coralba on brushed old polished ice at different speeds, with one
                       standard deviation.




                                                            73
6.5.2.3 New System 2000 ice
   Figure 88 and Figure 89 display the average road grip from viable measurements made by
VBOX and CORA respectively. No dominant trend can be seen for the devices.


                                             VBOX measurements on polished ice

                    0,45
                     0,4
                    0,35
                     0,3
      Road grip




                    0,25
                     0,2
                    0,15
                     0,1
                    0,05
                      0
                                   30 km/h                      50 km/h                      70 km/h
                                                                Speed



Figure 88             Road grip values for VBox on new System 2000 ice at different speeds, with one standard
                      deviation.




                                  Grip values on new System 2000 for Coralba

                     0,30

                     0,25

                     0,20
         Friction




                     0,15

                     0,10

                     0,05

                     0,00
                                    30 km/h                     50 km/h                     70 km/h
                                                                 Speed



Figure 89             Road grip values for Coralba on new System 2000 ice at different speeds, with one standard
                      deviation.




                                                          74
6.5.2.4 Speed dependence SRIS.
   Figure 90 illustrating the average road grip for each sub run of the SRIS Volvo V70 on
Wednesday the 19th, displaying speed dependence. These results are questionable as the
amount of data is limited. The measurements on the two System 2000 ice surfaces correspond
with analysis of stopping distance tests done later during the day, see chapter 6.7. The
stopping distance tests on polished ice showed similar results. Probably an effect from the
ABS working with the low quality winter tires mounted on the SRIS Volvo V70.

                 0.45

                  0.4

                 0.35

                  0.3
                                                                                         New System 2000
      Friction




                 0.25
                                                                                         Old System 2000
                  0.2
                                                                                         Polished ice
                 0.15

                  0.1

                 0.05

                   0
                             30 km/h              50 km/h              70 km/h
                                                Speed [km/h]



Figure 90           Speed dependence based on SRIS measurements on Wednesday the 19th.


6.6   Test runs across three surfaces 19th
  In the test on Wednesday the 19th March we had three RT3s, one Coralba equipped
Mitsubishi, one V-Box equipped SAAB and one Volvo V70 SRIS. First we will present the
RT3’s results.

6.6.1 Test runs 30 km/h, G055
   Figure 91 illustrates the results from test runs 1 - 3 with G055 at 30 km/h on the right side
of the test track. The road grip of the first run is significantly higher than on the following
runs. This was the second run on the track, only G065 had made a run on the new polished
ice.
   It is possible that the contact between a clean surface with micro roughness and a well
adapted winter tire creates relative high road grip. As snow and ice particles get dragged onto
the track, road grip decreased to the fairly stabile low level seen on the following runs.
Another possible explanation is that there are some small sharp ice crystals on the surface that
get worn away early on as vehicles start to traffic sub section two.




                                                     75
Figure 91   Road grip on runs 1 – 3 for RT3 G055 at 30 km/h, with blue displaying the first run.

   Table 33 displays the averages and standard deviations of the different sections. The
difference between the lowest and highest average road grip is 45.4 units or 279%.

Table 33        Average road grip with one standard deviation on runs 1 – 3 for G055 at 30 km/h.

 Unit, run        Average       Standard        Standard         Distance     Distance
 and sub run     Road grip      deviation     Deviation [%]       start        stop
 G055, 1, 1        61.50           8.60           13.99             10          120
 G055, 2, 1        59.23          10.62           17.92             10          120
 G055, 3, 1        60.80           9.90           16.28             10          120

 G055, 1, 2         29.91          2.00               6.70          140          250
 G055, 2, 2         16.22          2.14              13.22          140          250
 G055, 3, 2         17.60          2.25              12.76          140          250

 G055, 1, 3         58.95          9.92              16.83          270          380
 G055, 2, 3         57.43          8.30              14.46          270          380
 G055, 3, 3         61.63          7.82              12.69          270          380




                                                76
Figure 92   Average road grip with one standard deviation on run 1 – 3 for G055 at 30 km/h.

   In Figure 92 it is clear that initial road grip is significantly higher on the polished ice.

6.6.2 Test runs 30 km/h, G056
    Figure 93 illustrates results from test runs 1 - 3 with G056 at 30 km/h on the left side of the
test track. We see the similar results as in chapter 6.6.1 regarding road grip on sub run 2 in the
first run.




Figure 93   Road grip on runs 1 – 3 for RT3 G056 at 30 km/h, with blue displaying the first run.

   Table 34 displays the averages and standard deviations of the different sections. The
difference between the lowest and highest average road grip is 33.5 units or 141%. This




                                                77
deviation is less, compared to the measurement made by G055 on the other side of the track,
the minimum road grip value is higher as well.

Table 34    Average road grip with one standard deviation on runs 1 – 3 for G056 at 30 km/h.

 Unit, run        Average      Standard        Standard         Distance     Distance
 and sub run     Road grip     deviation     Deviation [%]       start        stop
 G056, 1, 1        54.63         8.50            15.55             10          120
 G056, 2, 1        55.20         7.75            14.05             10          120
 G056, 3, 1        57.06         6.59            11.54             10          120

 G056, 1, 2         35.92         3.64            10.15            140          250
 G056, 2, 2         26.73         3.59            13.42            140          250
 G056, 3, 2         23.61         2.50            10.58            140          250

 G056, 1, 3         49.49         9.37            18.93            270          380
 G056, 2, 3         50.46         8.74            17.33            270          380
 G056, 3, 3         52.01         8.26            15.87            270          380

   Here the effect isn’t as pronounced as with the G055, but still very noticeable.

6.6.3 Test runs 50 km/h, G055
   Figure 94 illustrates the results from test runs 4 - 6 with G055 at 50 km/h on the right side
of the test track.




Figure 94   Road grip on runs 4 – 6 for RT3 G055 at 50 km/h.

   Table 33 displays the average road grip and standard deviations. The difference between
the lowest and highest average road grip is 46.3 units or 271%.




                                                78
Table 35    Average road grip with one standard deviation on runs 4 – 6 for G055 at 50 km/h.

 Unit, run        Average       Standard         Standard         Distance     Distance
 and sub run     Road grip      deviation      Deviation [%]       start        stop
 G055, 4, 1        62.76          10.26            16.35             10          120
 G055, 5, 1        62.45          10.21            16.35             10          120
 G055, 6, 1        62.90           9.77            15.54             10          120

 G055, 4, 2         18.17          2.15              11.83           140          250
 G055, 5, 2         16.98          2.24              13.19           140          250
 G055, 6, 2         19.14          1.99              10.38           140          250

 G055, 4, 3         62.43          8.43              13.49           270          380
 G055, 5, 3         62.31          8.13              13.05           270          380
 G055, 6, 3         63.03          6.78              10.76           270          380

   Results from measurements on runs 4 – 6 correlate well and represent good measurements.
All measurements with the RT3s during Wednesday show increased road grip just before and
after the Polished ice. This as System 2000 was used to create well defined ends, as explained
in chapter 4.5.2.




Figure 95   Average road grip with one standard deviation on run 4 – 6 for G055 at 50 km/h.

   Figure 95 illustrates the average road grips which are very stable, despite the increased
standard deviation mostly due to the heightened road grip before and after the flooded ice.

6.6.4 Test runs 30 km/h, SRIS
   Finding valid road grip measurements in the SRIS excel data sheet was hard. We utilized
time data and GPS coordinates in order to find some road grip values.




                                                79
   Table 36 shows SRIS road grip measurements made at 30 km/h. Average road grip values
show correspondence with the stopping distance tests, if the questionable 0.07 results are
treated as outliers, see section 6.7.

Table 36   SRIS measurements on the 19th at 30 km/h.

SRIS                                                                              Standard    Rel. Standard
30 km/h               1       2       3       4         5        6     Average    deviation   deviation [%]
Old System 2000      0.32    0.41    0.29    0.29      0.29             0.32        0.052        16.24%
Polished ice         0.07    0.11    0.07    0.07      0.11    0.12     0.09        0.024        26.20%
New System 2000

   We did not find valid road grip data from tests on new System 2000 ice at 30 km/h.

6.6.5 Test runs 50 km/h, SRIS
   Table 37 shows SRIS road grip measurements made at 50 km/h. Average values show
correspondence with stopping distance tests.

Table 37       SRIS measurements on the 19th at 50 km/h.

 SRIS                                                         Standard    Rel. Standard
 50 km/h               1       2       3      Average         deviation   deviation [%]
 Old System 2000      0.32    0.32    0.22     0.29             0.058        20.14%
 Polished ice
 New System 2000      0.32    0.41    0.29      0.34           0.062        18.37%

   We did not find valid road grip data from tests on polished ice at 50 km/h.

6.6.6 Test runs 70 km/h, SRIS
   Table 38 shows SRIS road grip measurements made at 70 km/h. We found road grip values
for each sub run from the SRIS at this speed.

Table 38   SRIS measurements on the 19th at 70 km/h.

 SRIS                                             Standard                         Rel. Standard
 70 km/h           1   2    3    4    5   Average deviation                        deviation [%]
 Old System 2000 0.18 0.16 0.23            0.19     0.036                             18.98%
 Polished ice    0.11 0.11 0.13 0.11 0.13  0.12     0.011                             9.28%
 New System 2000 0.29 0.32 0.32 0.22       0.29     0.047                            16.41%


6.6.7 Test runs 30 km/h, VBox
   Figure 96 illustrates measurements made with the VBOX equipped SAAB on Wednesday
the 19th at 30 km/h. Each road grip value is presented individually as well as with calculated
average and one standard deviation.




                                              80
                                               VBox 30 km/h run 1-3

                         0.4

                        0.35

                         0.3

                        0.25
            Road grip




                                                                                         VBox
                         0.2
                                                                                         Average
                        0.15

                         0.1

                        0.05

                          0
                               0           1               2             3
                                                     Sub run



Figure 96     All road grip values from VBOX at 30 km/h together with average and standard deviation.

   Road grip measures on polished ice show a good level of precision see Table 39.

Table 39      Average road grip values and standard deviation on runs 1 – 3 for VBOX at 30 km/h.

 Unit, runs                Average     Standard    Rel. standard   Distance   Distance
 and sub run               road grip   deviation   deviation [%]     start      stop
 VBOX, 1-3, 1                0.25        0.02          8.7%           15        115
 VBOX, 1-3, 2                0.08       0.004          5.6%           15        115
 VBOX, 1-3, 3                0.32        0.03         10.6%           15        115

6.6.8 Test runs 50 km/h VBOX
   Figure 97 illustrates measurements made with the VBOX equipped SAAB on Wednesday
the 19th at 50 km/h. Each road grip value is presented individually as well as with calculated
average and one standard deviation.




                                                      81
                                              VBox 50 km/h run 1-3

                         0.4

                        0.35

                         0.3

                        0.25
            Road grip




                                                                                         VBox
                         0.2
                                                                                         Average
                        0.15

                         0.1

                        0.05

                          0
                               0          1                2            3
                                                     Sub run



Figure 97     All road grip values from VBOX at 50 km/h together with average and standard deviation.

   Table 40 presents the average road grips and standard deviations of the measurements.

Table 40      Average road grip values and standard deviation on runs 1 – 3 for VBOX at 50 km/h.

 Unit, run and              Average    Standard    Rel. standard   Distance   Distance
 sub run                   road grip   deviation   deviation [%]     start      stop
 VBOX, 4-6, 1                0.26        0.03         11.2%           15        115
 VBOX, 4-6, 2                0.10        0.02         24.6%           15        115
 VBOX, 4-6, 3                0.37        0.01          2.5%           15        115

6.6.9 Test runs 70 km/h, VBOX
   Figure 98 illustrates measurements made with the VBOX equipped SAAB on Wednesday
the 19th at 70 km/h. Each road grip value is presented individually as well as with calculated
average and one standard deviation.




                                                      82
                                              VBox 70 km/h run 1-3

                        0.45
                         0.4
                        0.35
                         0.3
            Road grip




                        0.25                                                              Average
                         0.2                                                              VBox
                        0.15
                         0.1
                        0.05
                          0
                               0          1                2            3
                                                     Sub run



Figure 98      Individual road grip measurements with calculated average road grip and standard deviation
               for the VBOX at 70 km/h.

  Table 41 show that the standard deviation is similar for measurements made at 70 km/h as
measurements made at 30 km/h and 50 km/h.

Table 41       Average road grip values and standard deviation on runs 1 – 3 for VBOX at 70 km/h.

 Unit, run and              Average    Standard    Rel. standard   Distance   Distance
 sub run                   road grip   deviation   deviation [%]     start      stop
 VBOX, 7-9, 1                0.25        0.01          5.7%           15        115
 VBOX, 7-9, 2                0.13        0.01          9.2%           15        115
 VBOX, 7-9, 3                0.35        0.05         14.6%           15        115

6.7   Stopping distance brake test on the 19th
   The Swedish road administration requested that stopping distance measurements should be
performed to get information about how long the stopping distance was on the different
surfaces. A tape measure was used for distance measuring. We instructed the drivers to
engage brakes fully at a marked position, activating the Anti-lock Braking System, ABS. The
actual start of the braking was compensated for as there were some deviations in the driver’s
accuracy. This test was conducted in a simple manner and results must be considered to be
indications.
   In order to investigate correlation between stopping distance and RT3’s road grip values,
we used the RT3 G055 before the different brake tests. The speed used for the reference
measurement was 50 km/h.
   For information about tires and vehicles see Appendix A. It is reasonable to believe that
the difference in tires on the different vehicles and system performance have a large impact on
the stopping distance.
   The first brake test was performed on polished ice. The second brake test was on old
System 2000 ice with a thin layer of new snow. The third brake test was performed on a new
System 2000 ice surface.


                                                      83
6.7.1 Stopping distance on polished ice at 30 km/h and 50 km/h
   First we measured the stopping distance on a very low grip surface, polished ice. The
surface was created about 6 hours before the brake test. Table 42 shows the stopping distances
for each vehicle with average and standard deviation. There are some large deviations for the
SRIS Volvo V70, see the second brake test at 30 km/h. The SRIS Volvo may have passed
where the vehicles equipped with studded winter tires had been braking. It is not likely that
the stopping distance would have decreased by 21% between test runs, as the surface was
homogenous. Even larger deviations were seen between the second and third brake tests for
the SAAB equipped with VBOX system, the distance increased with 30%.
   The difference at 30 km/h between the vehicle with shortest and the vehicle with longest
average stopping distance is 38%. The difference at 50 km/h is 58%.

Table 42                                               Stopping distances in [m] with standard deviation on new polished ice.

                                                                                       Standard                                                 Standard
                                                             30       30   30 Average deviation                 50   50   50            Average deviation
                                                            km/h     km/h km/h 30 km/h 30 km/h                 km/h km/h km/h           50 km/h 50 km/h
VV70, ST                                                    23.0     23.9 24.4 23.77     0.71                  53.1 52.8 56.0            53.97    1.77
SRIS, FT                                                    34.7     27.5 35.9 32.70     4.54                  80.2 89.8 85.3            85.10    4.80
VBOX, ST                                                    23.0     27.0 28.7  26.23    2.93                    -  59.5 77.4            68.45    12.66
VWPA, ST                                                    33.0     30.3   -   31.65    1.91

  Figure 99 illustrates the average stopping distance and the standard deviation at 30 and 50
km/h on polished ice.


                                                                                  Stop distance on polished ice
      Vehicle, FT=Friction Tires, ST=Studded Tires




                                                     VWPA, ST




                                                     VBOX, ST

                                                                                                                                             50 km/h
                                                                                                                                             30 km/h

                                                      SRIS, FT




                                                     VV70, ST



                                                                 0   10      20     30     40      50     60      70     80       90   100
                                                                                           Stop distance [m]



Figure 99                                              Average stopping distance on polished ice, initial speed 30 and 50 km/h.

We can see that the average stopping distance shows large variations during the first brake
test. The standard winter tires equipped SRIS Volvo V70 had the longest average stopping


                                                                                            84
distance even with a questionably low sub run. The tires were of “friction type” also called
studless winter tires. The rented Volvo V70, (VV70) with studded winter tires had the shortest
average stopping distance.
   The average stopping distance increased between 127% and 161% on polished ice when
the speed was increased from 30 to 50 km/h, see Figure 100. It must be noted that there are
many factors in this test that impact the results. Initial speed at brake application is a major
source of error. Other sources are distance measurement, estimation of initial brake point,
variations on the test surface and changes in surface characteristics as the test progresses, etc.
The best way to test different tires are to mount them on the same car. This is important as
there are significant differences in the performances of different vehicles during a braking
sequence.


                          Increase of average stopping distance in % for each vehicle as
                             speed increased from 30 km/h to 50 km/h on polished ice

                    180

                    160

                    140

                    120
      Percent [%]




                    100
                                                                                           Increase of average stopping
                                                                                           distance
                    80

                    60

                    40

                    20

                     0
                              VV70, ST             SRIS, FT             VBOX, ST
                                 Vehicle, FT=friction tires, ST=studded tires



Figure 100 Increase of average stopping distance when speed is increased from 30 to 50 km/h, in percent.

   Reference runs with the RT3 produced average road grip values according to Table 43.

Table 43              Average road grip with one standard deviation for RT3 G055 on polished ice at 50 km/h.

                            Average      Standard      Relative standard
                            road grip    deviation      deviation [%]              Start     End
G055, run 1                   24.2         3.69              15.2                  172       258
G055, run 2                   27.7         1.74               6.3                  172       258

   The road grip distribution across the polished ice can be seen in Figure 101.




                                                              85
Figure 101 Road grip on polished ice with RT3 G055 at 50 km/h.


6.7.2 Stopping distance on old System 2000 ice at 30 and 50 km/h
   This type of surface represents a low intensity winter road where no asphalt is exposed and
some new snow has fallen during the night, most commonly found on small roads. The
surface can be seen in chapter 4.5.1. Average stopping distance is greatly reduced on this
surface compared with polished ice, see Table 44. Only one value seems to be out of place
and that is for the SRIS Volvo, in the last sub run going 50 km/h, the stopping distance is
much shorter than during the two other sub runs. The difference at 30 km/h between the
vehicle with shortest and the vehicle with longest average stopping distance is 68%. The
difference at 50 km/h is 83%.

Table 44   Stopping distances in [m] with standard deviation on old System 2000 ice with new snow.

                                                    Standard                                         Standard
                 30      30       30     Average    deviation     50       50      50     Average    deviation
               km/h     km/h     km/h    30 km/h    30 km/h      km/h    km/h     km/h    50 km/h    50 km/h
VV70, ST         5.9     5.1      7.7      6.23        1.33      17.0     15.9    18.9     17.27       1.52
SRIS, FT        11.3     9.5      9.5     10.10        1.04      30.6     30.0    23.2     27.93       4.11
VWPA, ST        4.8      7.4      5.8     6.00         1.31      15.0     16.7    14.1     15.27       1.32

    Figure 102 illustrates the average stopping distance with one standard deviation and there
is clear evidence of a problem with the SRIS equipped Volvo. The tires on the vehicle are
Goodyear Ultragrip Performance 225/55R16 friction tires.



                                               86
                                                                                     Stop distance on old System 2000 ice

           Vehicle, FT=Friction Tires, ST=Studded Tires



                                                          VWPA, ST




                                                                                                                                                   50 km/h
                                                           SRIS, FT
                                                                                                                                                   30 km/h




                                                           VV70, ST




                                                                      0          5        10              15     20          25   30        35
                                                                                                     Stop distance [m]



Figure 102 Average stopping distance on old System 2000 ice with new snow, initial speed 30 and 50
           km/h.

  The increase in stopping distance is stable on this surface ranging between 154% and
177%, see Figure 103.


                                                                Increase of average stopping distance in % for each vehicle as
                                                                   speed increased from 30 km/h to 50 km/h on polished ice

                                                          180

                                                          160

                                                          140

                                                          120
      Percent [%]




                                                          100
                                                                                                                                   Increase of average stopping
                                                                                                                                   distance
                                                          80

                                                          60

                                                          40

                                                          20

                                                           0
                                                                      VV70, ST                 SRIS, FT               VBOX, ST
                                                                          Vehicle, FT=friction tires, ST=studded tires



Figure 103 Increase of average stopping distance when speed is increased from 30 to 50 km/h on old
           System 2000 ice with new snow.



                                                                                                      87
   As for polished ice, there were reference runs made on sub section one with RT3 G055.
The average road grip values and standard deviations are illustrated in Table 45 and Figure
104. The average lateral road grip values were 67.5 units for the first run and 66 units for the
second run. Note that wet asphalt produced a road grip of 90 units after calibration.

Table 45    Average road grip with one standard deviation for G055 on old System 2000 ice at 50 km/h
            with new snow.

                   Average       Standard Relative standard
                   road grip     deviation deviation [%]            Start       End
G055, run 1          67.60         5.94          8.79                0          120
G055, run 2          65.98         7.22         10.94                0          120



      80

      70

      60

      50

      40

      30

      20

      10

       0
                           G055, run 1                                  G055, run 2



Figure 104 Average road grip with one standard deviation at 50 km/h, for two reference runs on old
           System 2000 ice with new snow.

   The results from the RT3 are consistent and the standard deviation indicates no problem
with the reference runs. This is verified by inspection of the road grip graph illustrated in
Figure 104.




                                                88
Figure 105 Road grip on sub section 1 with snow on the surface measured with RT3 G055 at 50 km/h.

   According to theory, friction tires should be equal to studded tires on snow and rough ice.
The main difference should be on polished ice. Still, the friction tires in this test are even
worse on the System 2000 surface than on polished ice, when compared to the relative
stopping distance of the studded tires. For a more conclusive test other friction tires should be
included.

6.7.3 Stopping distance on new System 2000 ice at 30 and 50 km/h
    This surface is perceived to have the best grip as there are many sharp edges and high
surface roughness for the tires to find grip on. See chapter 4.5.3. The brake test verifies this
and in Table 46 we see evidence of high tire grip as the stopping distances are short. The
difference at 30 km/h between the vehicle with shortest and the vehicle with longest average
stopping distance is 23%. The difference at 50 km/h is 45%. This is the closest the friction
tired equipped SRIS Volvo gets, compared to the other vehicles in the brake tests.

Table 46   Stopping distances [m], with standard deviation results on new System 2000 ice.

                                               Standard                                        Standard
              30       30      30      Average deviation 50           50       50      Average deviation
              km/h     km/h    km/h    30 km/h 30 km/h km/h           km/h     km/h    50 km/h 50 km/h
VV70, ST 4.9           4.8     4.8     4.83       0.06        17.5    15.0     18.0    16.83   1.61
SRIS, FT 5             6.4     6.4     5.93       0.81        24.4    24.4     24.2    24.33   0.12
VWPA, ST 5.1           6.4     5.7     5.73       0.65        15.0    15.8     19.8    16.87   2.57

   Figure 106 shows the average stopping distance on new System 2000 ice with one standard
deviation with initial speed of 30 and 50 km/h. The difference between the vehicles is small at
30 km/h.




                                               89
                                                                                   Stop distance on new System 2000 ice
      Vehicle, FT=Friction Tires, ST=Studded Tires




                                                     VWPA, ST




                                                                                                                                                      50 km/h
                                                      SRIS, FT
                                                                                                                                                      30 km/h




                                                     VV70, ST




                                                                 0             5            10           15           20         25            30
                                                                                                  Stop distance [m]



Figure 106 Average stopping distance on new System 2000 ice with new snow and one standard
           deviation, initial speed 30 and 50 km/h.

  The increase in speed resulted in a larger increase of the stopping distance for the SRIS
Volvo compared to the other vehicles, see Figure 106.


                                                           Increase of average stop distance in % for each vehicle as speed
                                                             increased from 30 km/h to 50 km/h on new system 2000 surface

                                                     350

                                                     300

                                                     250
                Procent [%]




                                                     200
                                                                                                                           increase of average stop distance
                                                     150

                                                     100

                                                      50

                                                       0
                                                                 VV70, ST              SRIS, FT           VWPA, ST
                                                                     Vehicle, FT=friction tires, ST=studded tires



Figure 107 Increase of average stopping distance when speed is increased from 30 to 50 km/h on new
           System 2000 ice.



                                                                                                  90
   Based on increase in kinetic energy from 30 km/h to 50 km/h stopping distance should
increase with 177%, assuming that road grip has no speed dependence. This makes the 310%
increase for the SRIS Volvo V70 questionable.
   The reference runs made with RT3 G055 on subsection three gave the average road grip
values and standard deviations that are illustrated in Table 47 and Figure 108.

Table 47    Average road grip with one standard deviation for G055 on new System 2000 ice at 50 km/h.

                 Average Standard Relative standard
                 road grip deviation deviation [%]              Start     End
G055, run 1        51.3       7.3        14.3                     0       120
G055, run 2        50.7       9.9        19.4                     0       120


      70


      60


      50


      40


      30


      20


      10


      0
                           G055, run 1                                  G055, run 2



Figure 108 Average road grip with one standard deviation at 50 km/h, for two reference runs on new
           System 2000 ice.

   The results from the RT3 are consistent and the standard deviation indicates no problem
with the reference runs. This is verified by inspection of the road grip graph illustrated in
Figure 109.




                                               91
Figure 109 Road grip on new System 2000 ice with RT3 G055 at 50 km/h.


6.7.4 Road grip, SRIS Volvo V70, polished ice
   The internal road grip measurement system in SRIS Volvo V70 reported an average road
grip coefficient of 0.12 with a standard deviation of 0.010 from braking tests on polished ice.
The values from SRIS can be seen in Table 48.

Table 48   Road grip values on polished ice derived from the SRIS Volvo V70 in the SRIS project.

Run          1       2      3        4       5         6    7    8          9       10   11   12
SRIS μ     0.14    0.13   0.12     0.12    0.11      0.13 0.11 0.13       0.11     0.12 0.11 0.13

   Analysis on the average road grip from the stopping distance measurements gives an
average road grip coefficient of 0.11. This average road grip coefficient was calculated for a
32.7 m average stopping distance on polished ice at 30 km/h. For information about the
stopping distances see Table 42.
   The analysis for the 50 km/h brake test yielded a road grip coefficient of 0.115. The
average stopping distance was 85.1 m for the SRIS Volvo V70 at 50 km/h.
   Analysis was also done for old System 2000 ice. SRIS V70 internal system gave an
average road grip coefficient of 0.27 with a standard deviation of 0.036 on old System 2000
ice, see Table 49.

Table 49   Road grip coefficients on old System 2000 ice as reported from the internal road grip
           coefficient analysis system in the SRIS V70.

Run           1         2          3
SRIS μ      0.30      0.28       0.23

  The SRIS V70 had an average stopping distance of 10.1 m on old System 2000 ice at 30
km/h. Analytical calculations gave an average road grip coefficient of 0.35. With initial speed



                                                92
50 km/h and an average stopping distance of 27.9 m the analytical average road grip
coefficient was also calculated at 0.35.
   To conclude, the SRIS road grip coefficient calculation seems accurate for polished ice.
There are some differences between the reported and calculated values on old System 2000
ice. The largest source of error comes from the actual speed of the vehicle and the fact that the
automated safety systems in cars do not necessarily produce an even road grip.

6.7.5 Summary of brake tests
   The RT3 measures road grip in lateral direction at a relatively low slip rate. This is clear as
the tire has a relatively low wear rate even on dry asphalt. Rubber generates the highest
amount of road grip through hysteresis at a slip rate of 10% to 30% (3). Car manufacturers
aim to make their ABS systems work in this slip region. RT3 measurements indicate that the
lateral road grip at a low slip rate is less on the new System 2000 surface than on the old
System 2000 surface. This could be interpreted in the following way, “a vehicle needs more
steering input to achieve a direction change on new System 2000 ice compared to the old
System 2000 ice”. However no tests were made to verify this.
   Figure 110 shows average stopping distances for all vehicles at 30 km/h and 50 km/h
together with the average road grip values measured with RT3 G055 at 50 km/h, for each of
the three surfaces on which the brake test was performed. We see that there is not a one to one
correspondence between the RT3 road grip value and the stopping distance.
   The observations of changes in stopping distance for the two Volvo cars, on new System
2000 ice were surprising, see Figure 107. The average stopping distance increased by 248%
for the VV70 Volvo and by 310% for the SRIS Volvo. These results should be investigated
further. It is possible that this is a measurement error, as the expected change is an increase of
177% when changing the initial speed from 30 km/h to 50 km/h.


                             80,00                                             80

                             70,00                                             70
                                                                                    Road grip value [N/A]




                             60,00                                             60                           Average stop distance for all
         Stop distance [m]




                             50,00                                             50                           vehicles at 30 km/h [m]
                                                                                                            Average stop distance for all
                             40,00                                             40
                                                                                                            vehicles at 50 km/h [m]
                             30,00                                             30                           RT3 G055 average road grip
                             20,00                                             20                           at 50 km/h [N/A]

                             10,00                                             10
                              0,00                                             0
                                     Polished ice Old system 2000 New system
                                                         ice       2000 ice
                                                Type of surface



Figure 110 Average stopping distances for all vehicles in the brake test at 30 km/h, Blue line and 50
           km/h, pink line. Average road grip value at 50 km/h measured with RT3 G055, green line.


6.8   Circle test runs March 18 and 19
   All continuous road grip testers are designed for linear measurements, indicating that
turning could alter the road grip measurement. This was tested for two of the devices, the RT3
and the TWO.


                                                                  93
6.8.1 RT3s G055 and G065 circular test
   In an effort to investigate how RT3s handle turns, a number of laps were driven on the
circular track, see chapter 4.6.4.3. Both left and right turns were done, followed by short
straight tests of the road grip. A portion of the data acquired is presented in Figure 111,
together with the average from the whole data with one standard deviation.




Figure 111 Road grip values from left turns, straight runs and right turns, respectively (see arrow
           directions), presented with average value and one standard deviation for RT3 G065 and
           G055.

   It is obvious that turning alters the road grip value. A curve radius of about 130 m alters
the values with about ±50%. The surfaces on the circular test track were covered in snow, see
Figure 8. This makes the road grip measurement difficult as the road grip variation is large.
Significant design changes are needed to enable the RT3 to measure road grip in curves.

6.8.2 TWOV circular test
   The TWO did a slightly different test on March 18, 2008, driving straight for a reference
distance and then turning into a spiral curve with an initial radius of 40 m and shrinking, as
can be seen in the GPS data presented in Figure 112.




                                                94
Figure 112 GPS plot of TWOV driving straight then turning left with increasing steepness.

   The road grip measurement seems to follow a steady decline as seen in Figure 113,
depending on the turning radii. This is expected as more of the road grip is used to produce
lateral grip through lateral slip and less road grip is thus available for longitudinal grip.




Figure 113 Road grip measurement with the TWO unit going straight and then into an increasingly steep
           left turn. It seems to follow a linear decline with the turning ratio.

   The dual wheel setup of the TWO units is more robust than the RT3 unit when it comes to
measuring road grip in turns. The RT3 measure lateral forces, which increase dramatically
when the vehicle is turning left and decreases when the vehicle is turning right. A TWO unit
measures the difference between the reference wheel and the measurement wheel in the
longitudinal direction and is there for not nearly as affected by lateral forces caused by turning
by the pulling vehicle.


                                                95
6.9   Handling course test runs on March 19th
   On Wednesday, March 19, we drove RT3 G065 on the handling course. The data is so
inconsistent that it is meaningless to try to get road grip information from the runs. It could be
possible to get useful information from an RT3 system, however additional position
measurement systems will be required. See Figure 114 for GPS data.




Figure 114 Plotted GPS data from the RT3s G055 and G065, from top to bottom, the circular test track,
           the straight test track and the handling course.


6.10 Rotational ice friction analysis
    During the tests several friction measurements were conducted with laboratory equipment,
in this case a rubber ring that is rotated on the ice. The normal force and torque applied to the
rubber ring is measured at 50 Hz. The device is hand operated and connected to a data
acquisition system with a laptop for data recording. The device can be seen in Figure 115.




                                               96
Figure 115 Rotational friction analysis device. Both the upper and lower handle can be used to apply
           torque.

    The recording procedure initiates in the laptop and then the device is turned. A typical
recording is seen in Figure 116. Torque is applied to the device as testing starts, torque is then
increased to a maximum value where the rings start to slip, and this is considered the static
friction. The kinetic friction is seen as a plateau in the recording when rotation is fairly stable.
On fresh snow a high pitched sound is heard during slipping.




                                                97
                                                           Rotational friction

                    7                                                                                    0.8

                    6                                                                                    0.7




                                                                                                               Normal force [N] / Friction [-]
                                                                                                         0.6
                    5
      Torque [Nm]




                                                                                                         0.5
                    4                                                                                                                            M (Nm)
                            Built up tension    Static friction                                          0.4                                     N (kN)
                    3                                                                                                                            Friction
                                                                                                         0.3
                                                                                  Kinetic friction
                    2
                                                                                                         0.2
                                                                                  Stop
                    1                                                                                    0.1

                    0                                                                                    0
                        0            0.5       1            1.5          2          2.5              3
                                                         Time [s]



Figure 116 Typical measurement of normal force and torque on polished Zamboni ice. An initial torque
           can be seen, which then builds up with increased torque to a maximum value for static
           friction. Further rotation gives the kinetic friction value. Rotation is stopped at 2.3 seconds.
           The yellow line represents the friction coefficient calculated from torque and normal force.

   It is interesting to note that brushed old polished ice has very large differences between
static and kinetic friction, whereas rough System 2000 ice has very similar values for both
kinds of friction. With this method there is not a large surface dependency for static frictions,
but kinetic friction differs greatly. A test was also conducted on polished (Zamboni) ice with a
fresh snow cover, and it can be seen that it has friction coefficients comparable to System
2000 ice.

Table 50                Rotational friction values on some surfaces.

                                                                       Static μ    Kinetic μ
 Brushed old polished ice test 1                                        0.18        0.110
 Brushed old polished ice test 2                                        0.14        0.097
 Polished ice test 1                                                    0.20        0.110
 Polished ice test 2                                                    0.17        0.070
 Polished ice test 3                                                    0.18        0.086
 Polished ice test 4                                                    0.16        0.085
 Polished ice test 5                                                    0.19        0.099
 New System 2000 ice test 1                                             0.17        0.160
 Old System 2000 ice with snow test 1                                   0.24        0.220
 Old System 2000 ice with snow test 2                                   0.25        0.210
 Brushed old polished ice with fresh snow
 cover test 1                                                           0.26          0.220




                                                                  98
6.11 Correlation between weather conditions and Road grip
   The weather was stable during the three day test, therefore no definite correlation between
weather and road grip could be established. To explore this further would require more data
taken over a large variety of weather conditions, on a dedicated stretch of road or test track.
   A second step would be to interconnect a large amount of different types of sensors, like in
the SRIS project. Gathering data and correlating those to road grip measurements as well as
dedicated meteorological measurements. This data could then be assembled in a database and
run through an expert system to predict the weather effects on traffic conditions.

6.12 Data and measurement issues
  Special events or circumstances that were noticed have been assembled for each vehicle
and in general for types of measuring devices. In many cases they hold the answer to
unexpected data.

6.12.1 RT3 G055
   The driver of RT3 G055 reported that the performance of his engine was unstable and
power came and went randomly. This affects the speed during a test. It can also take focus
from the driving and cause larger deviation when driving as straight as possible.

6.12.2 RT3 G053
   RT3 system G053 had display problems. It was changed in the beginning of Tuesday,
March 18, 2008.
   The system time was set wrong during the first half of Monday March 17, 2008.

6.12.3 Measurement challenges
   During the analysis of data we found the following issues:

6.12.3.1 RT3 issues
        •    The data logging system used with the RT3s does not have constant frequency.
             This produces a changing amount of data points per second. The variations were
             large, and ranged from 4 to 10 Hz. This makes the analysis of the specific Road
             grip issues during the run down the course hard and is the main reason why we
             can’t calculate exactly where the equipment is on the test track.
        •    Time data in the RT3’s files is recorded in full seconds, which is coarse.
        •    Time stamp of the RT3’s files is not marked with AM or PM.
        •    The speed data in the RT3’s files is logged in 1 mph resolution, which is coarse.
        •    There is speed data that reads among others 0 mph and 147 mph which is
             impossible in the specific position. The speed data that was clearly wrong was
             changed according to surrounding data.
        •    The GPS system has three main issues:
             • The resolution is low, values recorded from the GPS do not change for several
                  seconds.
             • There are several sequences where there was no data recorded.
             • The position recorded is not near Lake Kakel, it was several km north of our
                  actual location.




                                            99
         •    The RT3 doesn’t measure road grip accurately in turns, this has an impact if
              there is interest in complete road evaluation.
         •    During inspection the RT3s was found to be mounted differently than what was
              recommended in the manual from Halliday Technologies INC.

6.12.3.2 TWO
         •    Measurements are already averaged therefore no high resolution information is
              available.
         •    TWO does not measure road grip accurately in very sharp turns with radius
              below 40 meters.
         •    Sometimes it is hard to detect a change in road grip going from old System 2000
              ice onto brushed old polished ice.

6.12.3.3 ViaFriction
         •    Some data is rounded to a low precision. The data is detailed enough for regular
              use.
         •    Sometimes it is hard to detect a change in road grip going from old System 2000
              ice onto brushed old polished ice.
         •    At one time the wheel locked up due to snow caught in the system, and did not
              rotate.

6.12.3.4 Coralba
         •    The measurements have to be documented by hand.
         •    Only gives a single Road grip value with no standard deviation.

6.12.3.5 SRIS
         •    Only samples at given times, no control over the samples.
         •    Samples have to be retrieved from an off-site database.

6.12.3.6 VBox
         •    Propriety software is needed to access data.
         •    Propriety software is hard to use.

7    Conclusions
    These are the conclusions that can be drawn from the tests in Arjeplog:

    •   All RT3 units show good repeatability. The precision is good enough to separate the
        different surfaces.
    •   The RT3 units compare well with each other, while dynamic range and road grip
        values can differ, trends are the same. Good calibration routines are necessary to avoid
        discrepancies.
    •   The RT3 can not handle curved roads, which can be considered the most dangerous in
        winter traffic. The most interesting forces in those situations are the lateral (steering)
        forces that the RT3 measure on straight roads.




                                             100
      •    Both TWO units show good repeatability. The precision is not always enough to
           separate old System 2000 ice from brushed old polished ice with TWOV. TWOT had
           no problems.
      •    The TWO units compare well with each other. While dynamic range and road grip
           values can differ, trends are the same. Good calibration routines are necessary to avoid
           discrepancies.
      •    The ViaFriction units show good repeatability. The precision is good enough to
           separate the different surfaces in most of the cases.
      •    RT3 does not measure the same kind of road grip as TWO or ViaFriction.
               o Comparing old and new System 2000 surfaces gives different results on RT3
                  compared than TWO and Via Friction.
               o RT3 measures lateral (steering) forces at low slip rate.
               o TWO and ViaFriction measures longitudinal (braking) forces with high slip
                  rate.
      •    TWO and ViaFriction correlates well to Coralba regarding road grip between old
           System 2000 ice and new System 2000 ice, RT3 shows less correlation for these
           surfaces.
      •    No road grip changes were observed that only could be associated with weather.
      •    Speed had no noticeable and always present effect on road grip measurements during
           these conditions, with the RT3s and the TWOs. We did see speed dependency for the
           ViaFriction. Speed might have had an effect on the SRIS Volvo V70, during braking
           with winter tires of friction type.
      •    The TWO has some trouble with very narrow bends, but manages most country roads’
           curvature.
      •    All devices recognize polished ice as a slippery surface in general, whereas old and
           new System 2000 ice surfaces are generally deemed to have adequate road grip.

8     Outlook
   The Road grip project has at this stage only scratched the surface in the field of road grip
prediction and estimation. It would be possible to create a database that together with active
measurements would provide early and accurate information about road grip. This would
enable proper countermeasures to be executed to maintain a safe level of road grip for all
traffic.

8.1       Data base content
   A database shall contain information about outer condition patterns that will be
automatically detected through weather stations, direct measuring on the road surface, and
computer models. Data shall then be assembled to warn for potential risks that the road grip
levels are about to become unacceptable on a road.

8.2       Features of road grip measuring equipment
   What we need from a road grip measuring device are results that enable us to answer the
following questions:

          •   What is the road grip available for acceleration?
          •   What is the road grip available for steering?
          •   What is the effective deceleration road grip?




                                               101
      These questions must be answered under the following conditions:
         • The measurements shall not vary depending on driver or device. They should be
             stable enough to always give the same results for the same road condition. Can
             measure all stretches of road with minimum disturbance of traffic.
         • The range of friction coefficients of interest are between 0.1 and 1. If the friction
             coefficient is less than 0.1, safe transportation is nearly impossible. If the friction
             coefficient is more than 1 there is enough road grip for most transportation
             applications.


8.3      RT3 usage and data acquisition system
   During the analysis of the data it is clear that an improved data acquisition system is
needed. There were multiple runs in which GPS data were missing and the frequency of data
points showed variations. In some measurement series large deviations were detected, the data
points per second ranged from 4 to 10.

8.3.1 Start /stop triggering
   Automatic systems for starting and stopping test runs are needed. It is clear that there
would be a significant advantage to be able to start or at least mark the data logging
automatically at some set locations. An alternative is to log continuously, and keep track of
the location and time of all vehicles at all times to compare this to the vehicle timestamps and
get a distribution of road grip measurements over time.

8.3.2 Faster data acquisition
   The rate at which data is recorded must be as fast as the output from the RT3. In a research
set up it would be best to change the electronics and use the analog signal from the load cell
inside the RT3.

   The potential for further development is huge in modeling and simulation of the tire road
interface related problems such as ice and tire wear, micro and macro roughness, adhesion
and hysteresis forces and weather related changes. Some ideas for further research are:
       • Further improvements of realistic and useful models for tire tread patterns.
       • Further understanding of rubber-ice adhesion and rubber hysteresis.
       • Introduction of “on-board” monitoring of road grip.
       • Further development of weather models for road grip predictions.

9      References

[1]       Center for Automotive System Technologies and Testing – CASTT,
          www.ltu.se/tfm/cooperation/castt/about/CASTT
[2]       www.Zamboni.com 2008-06-05
[3]       Michelin (2001) The tire grip. France: Société de Technologie Michelin




                                                 102
Appendix A

Vehicle data

  RT3 G065, (G065)

  Mitsubishi L200 1998 Reg nr: RBS 822
  Tires on vehicle/type: Nokian Hakkapeliitta C2, 205 R 16 C/studded winter tires
  RT3 serial number: HTI-A-0603-E-001 F GEM 065
  Tire on RT3: Bridgestone Blizzak Nordic WN-01, manufactured week 50 in 2005.
  Type of tire on RT3: Friction
  Tire pressure on RT3: 1.9 bar
  Tread depth on RT3: 9.2 mm
  Internet address: www.hallidaytech.com




  RT3 G055, (G055)

  Mitsubishi L200 2005 Reg nr: URC 406
  Tires on vehicle/type: Nokian Hakkapeliitta C2, 205 R 16 C/studded winter tires
  RT3 serial number:
  Tire on RT3: Bridgestone Blizzak Nordic WN-01, manufactured week 50 in 2005.
  Type of tire on RT3: Friction tire
  Tire pressure on RT3: 1.9 bar
  Tread depth on RT3: 8.4 mm (the unit was set at 8.5)

                                           A
RT3 G053, (G053)

Reg nr: RUB 881
Type of tire on vehicle: Friction tires
RT3 serial number: F-T8 G053 D-6
Tire on RT3: Bridgestone Blizzak Nordic WN-01, manufactured week 50 in 2005.
Type of tire on RT3: Friction tire
Tire pressure on RT3: 1.85 bar
Tread depth on RT3: 9.0 mm

RT3 G056, (G056)

Reg nr: BKW 799
Vehicle: GMC Econoline
Type of tire on vehicle: Friction tires
RT3 serial number: F-T8 G054 D-6
Tire on RT3: Bridgestone Blizzak Nordic WN-01, manufactured week 50 in 2005
Type of tire on RT3: Friction tire
Tire pressure on RT3: 1.9 bar
Tread depth on RT3: 9.2 mm

Pon-Cat TWO, (TWOV)

Reg nr: XV 62829
Vehicle: BMW X5 with friction tires
TWO serial nr: 106040
Tires on Pon-Cat unit: Trelleborg Industries 523, 4R 8P.R(Layer), test wheel have
production week 34 year 2005.
Reference wheel production week 25 year 2006
Type of tire on TWO: Industrial
Tire pressure on TWO: 1 bar
Internet address: www.pon-cat.com/Start____10018.aspx




                                           B
Pon-Cat TWO, (TWOT)

Reg nr: DL 42373
Vehicle: Toyota Hiace with friction tires
TWO serial nr: 106001
Tires on Ponn-Cat unit: Trelleborg Industries 523, 4R 8P.R(Layer), test wheel have
production week 35, year 04
Reference wheel have production week 36, year 2006
Type of tire on TWO: Industrial
Tire pressure on TWO: 1 bar

ViaFriction 1001, (VIAT)

Reg. Nr: KT 72905
Vehicle: VW Transporter with studded tires
Driver: Tony Skattebo
Road grip measuring unit: ViaFriction with S/N 1001
Tire on ViaFriction unit: Trelleborg unitester 520 4-8
Type of tire on ViaFriction: Industrial
Software: ViaFriction
Internet address: http://viatech.no




                                            C
Coralba, (CORA)

Reg. Nr: SOA 545
Vehicle: Mitsubishi L200
Type of tires on vehicle: Studded winter tires
Coralba type: Coralba µ/3
Internet address: www.coralba.se




V-BOX, (VBOX)

Reg. Nr: FNJ 358
Vehicle: Saab 9-5 Linear SportCombi 2.0t Bio
Tires: Kleber Kapnor5 195/65/R15
Tire pressure: 2.2 Bar
Types of tires: studded
Wheel base: 2703 mm
Track width Front: 1528 mm
Track width Rear: 1528 mm
Weight: 1550+75=1625kg
Weight distribution: 60%/40% front/rear
GPS data logging system: VBOX III 100 Hz
GPS data logging system: VB20SL 20 Hz
Inertia measuring: RLVBIMU 02
Internet address: www.racelogic.co.uk




                                            D
Volvo SRIS, (SRIS)

Reg. Nr: FUS 616
Vehicle: Volvo V70
Tires: Goodyear Ultragrip Performance 225/55R16
Type of tires: Friction
Road grip measuring: Algorithm from Volvo utilizing the cars own systems
Internet address: www.sris.nu




Volvo V70, (VV70)

Reg. Nr: GNS 510
Vehicle: Volvo V70
Tires: Pirelli Carving, 205/60/16
Type of tires: Studded
VW Passat, (VWPA)
Reg. Nr: GCR 067
Vehicle: VW Passat
Tires: Khumo I/ZENWISKW19, 205/55/16
Type of tires: Studded




                                         E

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:3
posted:11/29/2011
language:Swedish
pages:115