1

Johan Granlund

VEHICLE AND HUMAN VIBRATION

DUE TO ROAD CONDITION

A Report from Studies on Low Volume Roads in the EU Northern Periphery of:
 Demonstrations of Truck Ride Measurements in ROADEX Partner Areas
 Effectiveness on Isolation of Ride Vibration by a Tyre Pressure Control System
 Impact from Road Maintenance Standard on Truck Drivers Working Environment
 2

ABSTRACT
This report is a follow-up on the ROADEX III report “Health Issues Raised by Poorly Maintained
Road Networks”, which stated: “Northern European road users may be exposed to unacceptable
health and safety risks, in terms of ride vibration and skid accidents”. That report included a case
study of a timber logging truck that showed that the test truck drivers on Road 331 in the summer
of 2007 had been exposed to daily vibration doses higher than the Action Value A(8) in Directive
2002/44/EC. The EC-directive includes limits to allowed vibration at work with respect to health
and safety risks. The ROADEX III test truck drivers were also exposed to significant spinal
compression stress when their trucks had passed over severe road damage. Truck vibration
modes included lateral buffeting during transient roll vibration in road sections with non-uniformly
deformed weak shoulders. These transient side forces are recognized for increasing the skid crash
risk on wet ice and other slippery road surfaces.

The ROADEX III study also validated a new road condition parameter, Rut Bottom Cross Slope
Variance (RBCSV), as an indicator of road damages that cause “truck roll motion”. The validation
was done by collating RBCSV road data with heavy truck ride data on transient roll vibration and
roll-related lateral buffeting. A limit value of 0.30 % for RBCSV was derived on the basis of several
inputs, including skid crash records, truck driver perceived skid risk and RBCSV statistical
distribution on good and bad roads.

This present ROADEX IV follow-up report describes results from demonstration projects in
Scotland, Finland, Norway and Sweden over the period 2010 to 2012, where the case study from
the Beaver Road 331 has been reproduced. This document also reports on a study of the influence
of road maintenance standard on truck ride vibration and vehicle internal noise, with special focus
on winter road condition. Furthermore it reports on a study on vibration isolation from road to truck
driver’s seat, by use of a Tyre Pressure Control System (TPCS). All measurements were carried
out in the period 2010 - 2012.

A high repeatability between similar truck round trips was confirmed when using the ROADEX
method to assess truck drivers’ daily vibration exposure A(8). Results from measurements of truck
ride quality during the demonstrations included:
 Unacceptably high levels of driver´s daily vibration dose A(8) were recorded in all of the
ROADEX Partner areas (the Norwegian E6-measurements were lower than the others but
still at about the EU Action Value).
 Significant compression stress in the truck drivers spine were recorded at severe road
damages, such as sharp frost heaves, settlements at bridges and culverts, improper
road/bridge joints and uneven transversal joints at both old and new asphalt repairs.
 In all Partner areas, intense truck roll vibration and lateral buffeting was recorded. This
confirmed a special health and safety problem in the EU Northern Periphery (NP) cold
climates.
 The pavement condition parameter RBCSV (a “truck roll vibration indicator”) was further
validated in addition to the previous ROADEX III study in Sweden. At sites with very high
RBCSV, there is a risk that cargo latches might break due to high lateral acceleration.
 Winter conditions in the NP can result in significant corrugations in thick ice covering non-
salted roads, and extremely uneven frost heaves. The project results show that these
conditions can make the ride vibration and noise much worse than during summer conditions.
 The use of a Tyre Pressure Control System (TPCS) has previously been shown to reduce
ride vibration. The present study used a more detailed analysis to quantify the TPCS vibration
isolating effect. Results from Scotland and Sweden show that TPCS was very efficient in
isolating “shake” vibration from short wave road roughness (megatexture < 0.5 m) such as
potholes and corrugated ice surfaces.
 Vehicle body “bounce” vibration with lower frequencies (1 – 3 Hz) were not isolated by the
TPCS. Such low frequency bounce vibration can only be reduced by pavement maintenance.
 3

 High side friction demand due to improperly banked horizontal curves was found to be a
contributing factor behind many loss-of-control crashes, including rollovers. Several curves
with tragic crash records were found to be improperly banked despite being newly resurfaced.

A number of overall conclusions and recommendations for NP road agencies and truck
operators are offered:
 Truck hauliers, both self-employed and those with employees, who find themselves operating
under conditions comparable to those in this study – typically representative for the EU
Northern Periphery (NP) – are obliged by their national laws under the 2002/44/EC Directive
“to make risk assessment for health and safety issues raised by exposure from vibration as
well as mechanical shock”. For winter operations this should also include assessing the skid
risk associated with lateral buffeting on slippery ice.
 The assessment of NP truck drivers’ exposure to daily vibration at work should take account
of winter operation. In this study, the daily vibration in the winter was +39 %, and the daily
compression stress dose in the spine was doubled, compared to same operation (truck,
route) despite significant higher driving speed in the summer.
 Foreign hauliers should be subject to the same level of supervision of vibration risk
assessment as native hauliers to avoid unfair competition. In Sweden, the current share of
foreign heavy trucks in long haulage is estimated to 40 % and it is rapidly growing.
 It was concluded that a significant share of the ride vibration problems identified in the
surveys could be eliminated in a relatively short time by improved road construction and
maintenance practices as they originated from manmade sources, such as culvert trenching,
transversal joints at bridges and at improperly performed resurfacings.
 The new pavement condition parameter RBCSV is a useful tool to identify hazardous sites
with high risk for instability crashes such as rollovers. High RBCSV can also explain crashes
involving strapped heavy goods on trucks or trailers that have slipped and/or overturned.
 An under-estimated source of poor ride quality is corrugation in gravel roads and in thick ice
covered paved roads. This study showed that the current smoothness specification for winter
roads in Sweden is unable to penalize unacceptable corrugations. More research is needed
to establish a measurement method and a limit value for maximum acceptable corrugation.
 Public road network condition surveys with laser profilers are traditionally only carried out in
the summer time in the NP area. In regions with “white winter roads” (not the Scottish
Highlands or Ireland), frost actions may locally cause significant differential ground heave,
resulting in a seasonal increase of road roughness that can be dramatic. The road condition
assessment procedures of the ROADEX Partners may therefore need revision in order to
become credible for seasons other than the summer (which is short in the Northern
Periphery). Norway already has started to incorporate road condition data such as IRI-values
from the teleløsning (“spring thaw”) season in their Pavement Management System. Sweden
is currently developing an improved practice for the measurement of frost-related road
roughness.
 Trucks that regularly operate on rough surfaces should be equipped with a TPC-System and
be frequent checked and corrected for wheel imbalance. This will significantly reduce daily
vibration A(8), as required by law.
 Trucks without a TPC-System should employ “continuous correction of wheel imbalance” by
balancing powder. The use of “Counteracting Balancing Beads” or similar products will
significantly reduce daily vibration A(8), as required by law.
 The traditional methods of repairing roads have been developed for dealing with structural
condition and with short wave roughness (up to some 5 m waves). These methods are not
however appropriate for dealing with the health and safety issues covered by this report. For
these, road agencies will also need to implement a practice for repair and improvement of the
pavement geometry, so unevenness with wavelengths up to some 30 m as well as improper
cross slope is eliminated. This calls for a modernized asphalt work process, with re-design of
road surface geometry and computerized asphalt machine control.
 4

 The study identified horizontal curves with tragic crash records that were still improperly
banked after resurfacing. Improper superelevation results in an excessive need for side
friction between road and tyre, which in turn is a key factor behind loss-of-control crashes
including rollover of heavy vehicles. Up to some 15 % of all fatal crashes (in Sweden: 50
deaths/year) on the road network could be prevented, by implementing a strategy for repair
and improvement of pavement geometry. Such a strategy could identify what curves need to
be improved, and ensure that the necessary improvement of pavement geometry (and
improved warning signs for sharp curves) is made at a minimum of cost in conjunction with
traditional pavement resurfacing or reinforcement works. Critical curves may need to be
elevated some 2 – 5 dm at the outer side compared to the road centreline. By doing this on
the improperly banked curves, the need for side friction typically drops by some 50 - 80 %.
This will significantly reduce the disproportionally high risk for fatal single-vehicle crashes in
outercurves. The cost/benefit of a scheme for such road safety improvement is likely to be so
high, that it could set a new reference for deliverable rates of return for road agencies on low
volume roads.

KEYWORDS
Human whole-body vibration (WBV), perception, comfort, driver performance ability, health, heavy
goods vehicles (HGV), suspension, isolation, amplification, damping, stiffness, mass, inertia,
dynamic force, jerk, jolts, transients, mechanical shocks, tyre pressure control system (TPCS),
unbalanced wheels, counteract balancing beads, road roughness, IRI, texture, side friction, rut
bottom cross slope variance (RBCSV), roll-related lateral buffeting, improperly banked curves,
adverse camber, drainage gradient, skid crashes, rollover (overturning), road safety, traffic safety,
pavement bearing capacity, frost heave, permafrost, corrugation, dirt roads, road maintenance,
repair and operation, winter, thick ice, thin black ice, ride quality, noise, low frequency vibration,
motion sickness, replacement of frost susceptible road materials, drainage, long wave
unevenness, pavement slopes, 3D, laser scanning, computer aided re-design (CAD) of road
surface geometry, computer aided manufacturing (CAM), asphalt machine control, building
information modelling (BIM), cost/benefit, sustainability.
 5

PREFACE
This is a final joint report from Tasks D3 and RE6 of the ROADEX IV “Implementing Accessibility”
project, a technical trans-national cooperation project between The Highland Council, Forestry
Commission Scotland and the Western Isles Council of Scotland; The Northern Region of The
Norwegian Public Roads Administration; The Northern Region of The Swedish Transport
Administration and the Swedish Forest Agency; The Centre of Economic Development, Transport
and the Environment of Finland; The Government of Greenland; The Icelandic Road
Administration; and The National Roads Authority and Department of Transport of Ireland. The
lead partner for the project was The Northern Region of The Swedish Transport Administration and
the project consultant was Roadscanners Oy (Finland), with sub-consultant Vectura Consulting AB
(Sweden).

This report gives an insight into ride vibration as a health and safety risk, by relating ride
experienced by professional drivers to the European legislation on vibration at work. The report
includes daily vibration exposures, A(8)-values, sampled during representative transport tasks on
typical Northern Periphery (NP) low volume roads in Finland, Scotland, Norway and Sweden. The
report includes illustrations of typical road damages found to yield health risks due to high spinal
compression stress. The report also relates ride to road standards, with special focus on how
winter conditions can affect both vehicle internal noise and ride vibration. This demonstrates that
ride quality is dependent even more on high standards for roads in a winter perspective (i.e. design
of pavement frost protection and operations such as snow-ploughing and grading icy roads), than
in a summer perspective. The new pavement condition parameter “Rut Bottom Cross Slope
Variance” demonstrated by the ROADEX III in Sweden is further validated in good correlation with
lateral buffeting of trucks. Since vehicle lateral buffeting strongly increases the risk for skid crashes
on ice-slippery roads, the new RBCSV parameter can be used as a tool to help prevent skid
crashes by identifying those road sections that need maintenance and/or repair. The effect of
vibration isolation by a Tyre Pressure Control System on a truck driver was quantified. It was found
that a TPCS improves the ride in a vehicle, especially on potholed and on corrugated dirt roads
and ice. However the use of TPCS is not a generic solution for poor truck ride on paved highways.
The report was prepared by Johan Granlund, Chief Technology Officer at Vectura Consulting AB.
Road profilometer data were taken from road network surveys in Finland (Destia Oy), Norway
(NPRA) and in Scotland (WDM Ltd). Vectura’s road profilometry in Norway and Sweden, and ride
measurements were undertaken by Max Risberg, Marcus Wettermark, Fredrik Stensson, Fredrik
Lindström and Johan Granlund, all of Vectura Consulting. The analyses were carried out by
Fredrik Lindström and Johan Granlund. Ron Munro, Project Manager of the ROADEX
“Implementing Accessibility” Project, checked the language. Mika Pyhähuhta of Laboratorio
Uleåborg designed the report layout.

The author expresses his gratitude to Jukka and Paavo Kurth (CEO and Founder at Kurth Oy), Ari
Kilponen (The Centre of Economic Development, Transport and the Environment of Finland),
Alisdair Ferguson (CEO at Ferguson Transport, Spean Bridge, Ltd), Richard Evans (Area Roads
and Community Works Manager, The Highland Council), Rune Damm (Technical advisor at the
Norwegian Haulier’s Association), Reidar Kulseng-Hansen (CEO Kulseng-Hansen Thermo-
Transport A/S), Per Otto Aursand (The Norwegian Public Roads Administration), Sylve Nordin and
Torbjörn Brorsson (CEO and Founder at Brorssons Åkeri AB). Special gratitude goes to the
professional truck drivers; Juho Kela (Kurth), Barry Thomson (Ferguson Transport), Jean
Pettersen (Kulseng-Hansen Thermo-Transport), Anders Sundin, Staffan Magnusson, Lars-Göran
Sidén and Jerker Edholm (Brorssons Åkeri) as well as to the garage mechanics at Ferguson
Transport and at Brorssons Åkeri. All of their assistance is greatly appreciated.

Finally the author would like to thank Timo Saarenketo of Roadscanners, Ron Munro of
Munroconsult Ltd, the ROADEX “Implementing Accessibility” Project Partners, and the Project
Steering Committee for their guidance and encouragement in this work.
 6

Copyright © 2012 The ROADEX “Implementing Accessibility” Project

All rights reserved.

ROADEX “Implementing Accessibility” Lead Partner: The Swedish Transport Administration,

Northern Region, Box 809, S-971 25 Luleå. Project co-ordinator: Mr Krister Palo.
 7

CONTENTS
ABSTRACT ......................................................................................................................................................... 1 
PREFACE ........................................................................................................................................................... 5 
CONTENTS ........................................................................................................................................................ 7 
1. INTRODUCTION ............................................................................................................................................ 9 

1.1. RIDE VIBRATION - A SERIOUS HEALTH AND SAFETY RISK ........................................................... 9 

1.1.1. Bumps and ride vibration cause stress, poor health and road crashes ...................................... 10 

1.2. THE ROADEX PROJECT .................................................................................................................... 12 

1.2.1. Objectives of the demonstration and research work on vibration and health ............................. 13 
2. INSTRUMENTS ............................................................................................................................................ 14 

2.1. RIDE QUALITY MEASUREMENT INSTRUMENTS ............................................................................ 14 

2.2. ROAD ROUGHNESS PROFILOMETRY ............................................................................................. 15 

3. DEMONSTRATION PROJECTS .................................................................................................................. 16 

3.1. OVERVIEW OF THE MEASURED A(8)-VALUES ............................................................................... 17 

3.2. POOR TRUCK RIDE QUALITY IN FINNISH LAPLAND ...................................................................... 18 

3.2.1. The route, the transport task, the truck and its instrumentation .................................................. 18 

3.2.2. High daily vibration exposure; A(8) exceeds the EU Action Value ............................................. 20 

3.2.3. Thawing permafrost is one of the root causes of high ride vibration .......................................... 22 

3.2.4. Occasional high spinal compression stress ................................................................................ 23 

3.2.5. Observations of hazards and various road features in Lapland.................................................. 26 

3.3. POOR TRUCK RIDE QUALITY IN THE SCOTISH HIGHLANDS ....................................................... 32 

3.3.1. The routes, the transport task, the truck and its instrumentation ................................................ 32 

3.3.2. High daily vibration exposure; A(8) exceeds the EU Action Value ............................................. 37 

3.3.3. Systematic high spinal compression stress ................................................................................ 40 

3.3.4. Undulating profile at bridges is a root cause of high ride vibration ............................................. 40 

3.3.5. Rough ride on patch work ........................................................................................................... 44 

3.3.6. Bumpy transversal joints ............................................................................................................. 44 

3.3.7. Bounce resonance at long wave unevenness ............................................................................ 45 

3.3.8. Precision in the measurements of truck ride quality ................................................................... 46 

3.3.9. Observations of hazards and various road features in Highlands .............................................. 47 

3.4. HAZARDOUS TRUCK OPERATION IN NORTHERN NORWAY ........................................................ 49 

 8

3.4.1. The route, the transport task, the truck and its instrumentation .................................................. 49 

3.4.2. Daily vibration exposure and spine compression stress ............................................................. 54 

3.4.3. Rough road profile at bridges is a root cause to high ride vibration ............................................ 55 

3.4.4. Bumpy transversal joints ............................................................................................................. 56 

3.4.5. Bumps at culverts ........................................................................................................................ 59 

3.4.6. Poor truck ride quality at deformed weak road sections ............................................................. 60 

3.4.7. Noise from thermo-cooler may disturb the drivers sleep rest ..................................................... 61 

3.4.8. Narrow road width and other hazards in Nordland/Trøndelag .................................................... 62 

3.5. POOR TRUCK RIDE QUALITY IN SWEDISH NORRLAND................................................................ 69 

3.5.1. The routes, the transport task, the truck and its instrumentation ................................................ 69 

3.5.2. The measurements ..................................................................................................................... 76 

3.5.3. High daily vibration exposure; A(8) exceeds the EU Action Value ............................................. 77 

3.5.4. Systematic high spinal compression stress ................................................................................ 78 

3.5.5. Rough road profile at bridges is a root cause to high ride vibration ............................................ 79 

3.5.6. Several years later: Still no warning sign at identified hazardous sites ...................................... 81 

3.5.7. Remarkably poor quality in road repair works ............................................................................. 82 

3.5.8. Observations of hazards and various road features in Norrland................................................. 86 

4. RELATING RIDE TO ROAD STANDARD .................................................................................................... 90 

4.1. COMPARING VEHICLE DRIVER VIBRATION TO IRI-VALUES......................................................... 91 

4.1.1. Previous studies relating vehicle drivers WBV to IRI-values ...................................................... 91 

4.1.2. Relating truck driver WBV to IRI-values in the current ROADEX IV study ................................. 93 

4.2. LATERAL BUFFETING – A SEVERE CRASH RISK ........................................................................... 97 

4.3. SEVERE FROST RELATED ROUGHNESS ...................................................................................... 100 

4.4. CORRUGATION IN THE ICE LAYER CAUSE NOISE AND VIBRATION ......................................... 101 
5. TPC-SYSTEMS REDUCE SHAKE VIBRATION ........................................................................................ 103 

5.1. TPCS REDUCE RIDE VIBRATION.................................................................................................... 103 

5.2. TPCS PREVENTS DRIVING WITH UNDER-INFLATED TYRES...................................................... 104 

6. CONCLUSIONS ......................................................................................................................................... 105 
7. RECOMMENDATIONS .............................................................................................................................. 108 
8. ACCURACY AND UNCERTAINTY ............................................................................................................ 110 
9. REFERENCES ........................................................................................................................................... 111 
 9

1. INTRODUCTION

1.1. RIDE VIBRATION - A SERIOUS HEALTH AND SAFETY RISK

Road roughness is source of vibration in vehicles and a well-known cause of wear and damage to
sensitive payloads1, to the vehicle itself, as well as to bridges and pavements. Vehicle vibration in
turn, also brings whole-body vibration (WBV) exposure to drivers and passengers. WBV is a root
cause of work-related accidents (in road traffic: crashes) and work-related diseases. Hence
Directive 2002/44/EC has set limits to the allowed vibration at work, with respect to health and
safety risks. These limits are implemented in all EU member states, and also by the Schengen
Agreement in the ROADEX partner areas in Norway, Iceland and Ireland.

This report is a follow-up report on the ROADEX III report “Health Issues Raised by Poorly
Maintained Road Networks”: That report concluded “Northern European road users may be
exposed to unacceptable health and safety risks, in terms of ride vibration and skid accidents”
based on a case study that mapped the risks on the Beaver Road 331 in Sweden. The report
demonstrated methods to efficiently prevent or reduce risks on similar roads. The Beaver Road
case study in particular showed that truck drivers on the road were, in summer time, exposed to
daily vibration doses higher than the Action Value A(8) in Directive 2002/44/EC. The truck drivers
were also exposed to significant spinal compression stress, when their trucks passed over
potholes, uneven bridge joints and road sections with various forms of severe deformations. The
ROADEX III study validated the new pavement condition parameter Rut Bottom Cross Slope
Variance (RBCSV) by collating it with data of transient roll vibration and of roll-related lateral
buffeting in a timber logging truck. Furthermore, a limit value of 0.30 % for RBCSV was derived on
basis on several inputs, including how lateral buffeting increased the risk of skid crashing on wet
black ice and other slippery road surfaces.

It is now recognised that professional truck drivers who frequently drive on rural low-volume roads
in poor condition can be exposed to human whole-body vibration (WBV) higher than the Action
Value set by EU directive 2002/44/EC. These drivers may suffer risks of stress related heart
diseases and of musculoskeletal problems in the neck, shoulders and back. Furthermore they can
be at high risk of being involved in instability crashes, where other road users may also be severely
injured when colliding with the heavy vehicle being out of control.

The main causes of ride vibration in vehicles are road defects. While vehicle suspension systems
are engineered to efficiently isolate the chassis from wheel vibration with higher frequencies, they
typically tend to amplify vibration frequencies somewhat lower than 4 Hz. Such vibrations are
excited from pavement deformation comparable to, or even longer/wider than, the vehicle
dimensions.

While most previous research projects on truck ride vibration have focused on vertical axis and
pitch axis oscillations, recent truck ride measurements on roads in the Northern Periphery (NP) of
the European Union have shown surprisingly high levels of both quasi-static and transient lateral
vibration. Unexpected high lateral forces in “egg-shaped” sharp improperly banked curves, and
roll-related lateral buffeting, are of major concern for traffic safety in cold climates, as they can
initiate skidding on ice-slippery surfaces. Buffeting in ambulance vehicles can also give rise to
health issues for vulnerable patients.

The Swedish National Institute of Public Health (2008) found that the most common types of
preventable mortality in Sweden were lung cancer (death rate of 17.1), suicide (15.4) and cerebral-

1
An example of payload damage can be seen in the road transportation of chilled salmon from the
Norwegian Northern Periphery to the consumer markets in the South. Due to frost actions in the late winter,
roads such as the E6 in Nordland can become very rough. The resulting transient ride vibration (mechanical
shock-loads) at road bumps can collapse the bottom layer of the polystyrene thermo-containers and destroy
the valuable salmon cargo.
 10

vascular disease (11.8). Among the therapeutic treatable death causes, diabetes mellitus was the
worst “big killer” with a death rate of 4.5. However, motor vehicle crashes were worse, with a death
rate of 4.9. This still places Sweden in the “top-three” position in the EU when it comes to safe road
traffic. However, there are large regional, and urban versus rural, differences in the risk of being
killed in a road traffic crash. While the metropolitan areas of Stockholm, Gothenburg and Malmoe
have a Standardized Mortality Ratio (SMR) of 70 for vehicle crashes, the rural areas in Sweden
have a SMR of 177. This means that vehicle users in the rural areas have (177 – 70) / 70 = 153 %
higher risk in ending up in a fatal crash, as compared to urban vehicle users. In the rural areas of
Northern Sweden, vehicle crashes take 39 % more lives than diabetes does.

1.1.1. Bumps and ride vibration cause stress, poor health and road crashes
Back disorders are costly to society and are the main causes of sick leave in the working
community. They cause great pain to those suffering, and are a significant economic burden to
society. Professional drivers are a group of workers that have been found to be at high risk for
back disorders. Many epidemiological studies have been made on the relationship between back
disorders and vehicle operation with vibration exposure. The results show overwhelming evidence
of a relationship that is consistent and strong, which increases with increasing exposure, and is
biologically plausible. Numerous back disorders are involved, including lumbago, sciatica,
generalized back pain, and intervertebral disc herniation and degeneration. The risk is elevated in
a broad range of driving occupations, including truck and bus drivers. Elevated risk is consistently
observed after five years of exposure; see Teschke et al (1999).

Two recent Swedish reviews2 of the scientific evidence of the relation between vibration at work
and adverse health effect are summarized in Table 1:

Table 1 Status of scientific knowledge for health effects by high vibration exposure at work
Low Back Pain: Confirmed!
Sciatica / Herniated discs: Confirmed!
Arthrosis: More research needed.
Miscarriage: Special regulation3 is implemented.
Male fertility: Indicated, more research needed
Viscus/Guts: No scientific support.
Heart: Several findings, more research needed.
Prostate cancer: Handful studies, more research needed
Motion sickness: Confirmed!
Performance: Several findings, more research needed
Mortality: Complex findings, more research needed

Severe negative consequences from ride vibration on the performance, drowsiness and safety for
professional drivers of road vehicles has been concluded in several studies, such as Arnberg &
Åström (1979) and Gillespie et al (1982).

Amongst older commercial drivers, musculoskeletal problems and cardiovascular diseases are the
primary reasons for changing their occupation. An increased risk of myocardial infarction among
professional drivers was first reported about 50 years ago, and has been reported repeatedly since
then. Stress under certain driving conditions is considered to explain the raised level of stress
hormones found in commercial drivers, and is believed to cause a large proportion of the health
problems, see Hedberg (1993). Bigert et al (2004) showed that the high incidence of certain heart
disease among Swedish truck drivers was constant over time.

2
Swedish Work Environment Authority Report 2011:8 and Arbete & Hälsa (Work & Health) Report 2012:46.
3
Regulation in Sweden: AFS 2007:5 Gravida och ammande arbetstagare. Arbetsmiljöverket.
 11

McFarlane & Sweatman (2003) studied the lane-keeping behaviour of heavy trucks on rough road
sections and found that where the road width was narrow, lateral “bumpsteer” disturbances could
require the driver to increase concentration into a stress level significant for driver fatigue.

Opinions of professional road users on road service levels across the EU NP area were mapped
by Saarenketo & Saari (2004). Truck drivers stated that the worst road sections had bumps at
culverts, weak pavement shoulders, poor road alignment and incorrect cross slope with respect to
road curvature. Improperly banked sharp curves yield high cornering lateral forces and very high
demand for side friction between tyre and road. They also reported continual stress when driving
on some long routes that the road agency believed to be in good driving condition. This happens
when unexpected poor road conditions force the driver to drop the vehicle speed far below the
planned speed for safety reasons. The result is a stressing conflict within the driver, between
making a delayed delivery and causing a major traffic safety risk.

Bray et al (2006) studied physiological stress responses to vehicular buffeting during a 5 minute
mild ‘off road’ exposure in a motion simulator, producing transient low frequency roll vibration with
1 m/s2 lateral vibration (root-mean-squared value). This level is not unusual during normal truck
driving on rural roads in the EU NP. The controlled exposure provoked an increase in heart rate
and blood pressure and a significant hypocapnia of PETCO2 34 mm Hg caused by tachypnea,
which took the test persons 5 minutes to recover. The authors concluded that buffeting in everyday
transport can affect people with cardiovascular disease.

The ROADEX III project made an in-depth assessment of truck driver’s exposure to vibration, see
Granlund (2008). Measurements were made in a timber logging truck during ten roundtrips of 140 -
170 km, with most time spent on Road 331 between the Swedish inland forest area and the coast.
The results showed that for all measured working days, the daily vibration exposure A(8) was
above 0.65 m/s2, including normal pauses with zero vibration, and that A(8) = 0.76 m/s2 was a fair
estimate for an 8 hour shift on this kind of route. This is significantly above the EU Action Value of
A(8) = 0.5 m/s2.

Employers of truck drivers performing long and bumpy driving are required by EU harmonized laws
to take necessary technical and/or organizational actions to minimize drivers’ exposure to
vibration. EU employers are also obliged to perform a special risk assessment for workers exposed
to repeated mechanical shock, such as from bumpy rides.

Unless a proper risk assessment has been made, or the required action taken, the standard fine in
Sweden has been set at 100,000 Euros following a pioneering case of bus drivers driving over
speed bumps in the City of Täby. Directive 2002/44/EC is implemented in national laws, which
apply to domestic employers. There is an open question on how the supervision of foreign truck
companies should be made.

One of the first cases in the EU where employees won annuities after a court case about adverse
health effects from vibration at work ended in 30 September 1997, when the UK High Court
awarded £127,000 compensation to 7 miners. The judge ruled that British Coal had been negligent
in not taking preventative actions against vibration since 1975. After losing an appeal to the 1997
ruling, the UK government set up the world's biggest ever compensation scheme. By the time it
closes the scheme is expected to have dealt with over 750,000 compensation payments to former
miners and their families, paying out an estimated £4.1 billion. Lawyers specialized on insurance
issues are now comparing diseases caused by “vibration at work” with casualty catastrophes such
as Exxon Valdez, the Asbestosis scandal and Agent Orange4.

4
Harding, et al. Casualty Catastrophes. (2009). General Insurance Convention. Internet 2012-05-08:
http://www.actuaries.org.uk/research-and-resources/documents/casualty-catastrophes
 12

1.2. THE ROADEX PROJECT

The ROADEX Project is a technical co-operation between road organizations across northern
Europe that aims to share road related information and research between the partners. The project
was started in 1998 as a 3 year pilot co-operation between the districts of Finland Lapland, Troms
County of Norway, the Northern Region of Sweden and The Highland Council of Scotland. It was
subsequently followed and extended with a second project, ROADEX II, from 2002 to 2005, a third,
ROADEX III from 2006 to 2007 and a fourth, ROADEX IV “Implementing Accessibility” from 2009
to 2012.

Figure 1 The Northern Periphery Area and ROADEX Partners

The Partners in the ROADEX “Implementing Accessibility” project comprised public road
administrations and forestry organizations from across the European Northern Periphery. These
were The Highland Council, Forestry Commission Scotland and the Western Isles Council from
Scotland, The Northern Region of The Norwegian Public Roads Administration, The Northern
Region of The Swedish Transport Administration and the Swedish Forest Agency, The Centre of
Economic Development, Transport and the Environment of Finland, The Government of
Greenland, The Icelandic Road Administration and The National Roads Authority and The
Department of Transport of Ireland. The partner areas are presented in Figure 1.

The aim of the project was to implement the road technologies developed by ROADEX on to the
Partner road networks to improve operational efficiency and save money. The lead partner for the
project was The Swedish Transport Administration and the main project consultant was
Roadscanners Oy of Finland with sub-consultant Vectura Consulting AB. The project was awarded
NPP funding in September 2009 and held its first steering Committee meeting in Luleå, November
2009.

A main part of the project was a programme of 23 demonstration projects showcasing the
ROADEX methods in the Local Partner areas supported by a new pan-regional “ROADEX
Consultancy Service” and “Knowledge Centre”. Three research tasks were also pursued as part of
the project: D1 “Climate change and its consequences on the maintenance of low volume roads”,
D2 “Road Widening” and D3 “Vibration in vehicles and humans due to road condition”. All of the
reports are available on the ROADEX website at www.ROADEX.org.
 13

1.2.1. Objectives of the demonstration and research work on vibration and health
The demonstration projects and research work on vibration and health in the Partner areas of
Finland, Scotland, Norway and Sweden had the overall objective to reproduce the ROADEX III
case study on the Beaver Road 331 in Sweden (2008). This included:
1. Measuring truck drivers daily vibration exposure and comparing the A(8)-values to the
Action Value 0.5 m/s2 in the EU Directive 2002/44/EC.
2. Measuring spine compression stress, Sed, caused by jolts at severe bumps and compare
the values to the 0.5 MPa stress limit in ISO 2631-5 (used as Action Value for bumps in
Sweden).
3. Relating truck roll & lateral buffeting in heavy trucks with a high Centre-of-Gravity to laser-
scanned non-uniform deformation at the pavement edge (the latter quantified by the
pavement condition parameter “Rut Bottom Cross Slope Variance”).

The selected test roads were measured with the normal profilometers locally used for road network
condition surveys. One goal was to implement the use of the Rut Bottom Cross Slope Variance
parameter in the Partner areas.

The overall objective for the ROADEX IV research project was to identify standards of road
maintenance & operations that give acceptable effects on health.

Vibration data obtained for winter conditions, such as from corrugations in a thick ice layer on the
road surface, were normalized to summer roughness. By this means the differing winter
operational component could be quantified.

A secondary research objective was related to the benefits of tyre pressure control systems on the
transfer of vibration from the tyre footprint to the driving seat. The research goal for this part of the
study was to quantify the transfer as function of vibration frequency (roughness wavelength x
speed).
 14

2. INSTRUMENTS

2.1. RIDE QUALITY MEASUREMENT INSTRUMENTS

Vectura instrumentation was used for recording truck ride vibration and ride quality as per the
international standards ISO 8041, ISO 2631-1 and ISO 2631-5. These were:
 A multi-functional instrument “Dewetron 3020” for 16 channel data logging up to 200 kS/sec
per channel, signal conditioning and filtering via dynamic differential signal amplifiers and
digital anti-aliasing filters, see Figure 2;
 A tri-axial seat-accelerometer “MMF KB103” at the pan of the drivers´ seat;
 A tri-axial accelerometer “MMF KS813B”, attached to the base of the drivers´ seat;
 A six-axial “Oxford Technical Solutions RT3000” precision inertial and GPS system for
measuring motion, position and orientation, mounted on an extensional carbon-fibre strut in
the truck cab, with dual GPS antennas on magnetic feet at the truck roof;
 A set of miniature accelerometers “Brűel & Kjaer TEDS 4507 B” (z-axis) mounted with glue
on axles and on the truck frame;
 A GPS with single antenna on the truck cab roof was used in the second measurements in
Norway and in Sweden, as the six-axial unit and its dual GPS antennas were omitted.
 A class 1 microphone “MMF MI17” for interior noise in the truck cab;
 A webcam for right-of-way pictures taken from the test truck dashboard;
 A Canon EOS550D 18 megapixel digital single-lens reflex camera with a Sigma 18 – 200
mm compact superzoom lens for freehand photos.

Figure 2 A Dewetron 3020 is the heart of Vectura’s ride quality measurement system

Vibration sampling was carried out at 1 kHz, except for the RT3000 system which operated at 100
Hz. The quality of the “main” ride measurements were very high, as can be seen in the section
“Precision in the measurements of truck ride quality” which reports the results from repeated
measurements.

Initial vibration analysis was made in MatLab software. Final analysis was made in MS Excel,
using the “Vibration Calculator” developed by the UK Health and Safety Executive5.

Coarse indicative mapping of ride quality problems due to poor winter road maintenance was also
carried out in the test vehicles with an Apple Iphone, using the apps Buller (“Noise”, provided by

5
UK HSE Vibration Calculator on Internet 2012-05-15: http://www.hse.gov.uk/vibration/wbv/
 15

the Swedish Work Environment Authority) and Vibration (provided by DLD-LLC in AppStore).
These measurements do not comply with the ISO standards listed above and have a much lower
accuracy than the “main measurements”.

2.2. ROAD ROUGHNESS PROFILOMETRY

In Finland and at the first trip in Norway, road alignment, slopes and road condition parameters, for
roughness and texture, were measured with a laser/inertial Profilograph from Greenwood
Engineering A/S (GE), such as the one shown in Figure 3 (left). The Profilograph typically makes
some 20,000 samples per metre, of which about 400 data values are stored. The accuracy is very
high. For many parameters the inaccuracy is in fractions of millimetres, as certified by a third party.

The road roughness on the Swedish routes was measured with Vectura’s portable GE LaserProf
System, shown in Figure 3 (right). The LaserProf fulfils the requirements for a Class 1 profiler in
the ASTM E950 standard. The road data in Finland was sampled by Destia OY with a Profilograph
system identical with Vectura’s P45.

In Norway, the road data was measured in 2011 and 2012 (frost weakening season in April) was
measured with a ViaPPS sweeping laser system (similar to the system at the bottom of Figure 3)
operated in-house by Norwegian Public Road Administration.

Figure 3 Vectura’s GE Profilograph P45, GE LaserProf and ViaPPS systems

The road condition on the test routes in Scotland was measured with a profilometer somewhat
similar to a Profilograph, operated by WDM Ltd. The accuracy of the profilometer used in Scotland
is unknown to the author.
 16

3. DEMONSTRATION PROJECTS
The ROADEX IV demonstration projects in Scotland (512 km), Finland (763 km), Norway (636 km)
and Sweden (1.867 km) aimed to reproduce the 10 roundtrips on the 2 x 140 km long ROADEX III
case study on the Beaver Road 331. The Beaver Road case study was in ROADEX IV also
repeated again on a 70 km section on Road 331 south of Ramsele. The locations of the
demonstration projects are shown in Figure 4.

Figure 4 Locations of the ROADEX IV demonstration routes for truck ride quality

The demonstrations included:

 Truck driver´s daily vibration dose A(8), as per Directive 2002/44/EC.
 Compression stress in the truck driver’s spine, as per the ISO 2631-5 standard.
 Roll vibration and/or related lateral buffeting, a root cause to many skid crashes on icy roads.

Based upon recorded seat vibration values, the truck driver´s daily exposure A(8) was assessed
for various transport tasks and normalized to daily work patterns as per standard EN 14253, using
the “Vibration Calculator” developed by the UK Health and Safety Executive6.

Previous research results show that the magnitude of spinal compression stress is related to
transient shock loads by an exponent of about 6. This high exponent makes assessment of spinal
compression stress very susceptible to the accuracy of the measurement of the transient seat
vibration. There have been several cases where one single shock load has caused compression
fracture in the spine of humans, i.e. when riding with a bus driver that drove over a traffic calming
speed bump at speed. Furthermore there can be huge differences between the peak stress of
various vehicles and even in human individuals. Bearing in mind such facts, the author, and this
report, will focus on the importance of preventing high compression stresses, rather than
concentrating on a (relatively) small number of transient sampling. Hence the report will aim to
emphasize illustrations of risk situations / road features, rather than comparing exact values of
compression stress.

A main source of truck roll vibration is deformation at the pavement edge, caused by high traffic
loads on edges with poor bearing capacity. Data on truck roll vibration / lateral buffeting will be
used for validation of the new road condition parameter RBCSV as described in the later section
“Relating ride to road standard”.

6
UK HSE Vibration Calculator on Internet 2012-05-15: http://www.hse.gov.uk/vibration/wbv/
 17

3.1. OVERVIEW OF THE MEASURED A(8)-VALUES

The daily vibration exposures A(8) of each of the demonstration routes were computed after
normalization to daily work patterns as per the European standard EN 14253 (2007)
“Measurement and calculation of occupational exposure to whole-body vibration with reference to
health”. An overview of the results of the demonstrations is given in Figure 5. These results show
that the all of the truck drivers tested had a daily exposure that exceeded the EU Action Value of
A(8) = 0.5 m/s2. An exception is the result from the new truck on European Highway E6 in Northern
Norway, which had an A(8) = 0.47 m/s2 being slightly below the Action Value. However when
normalizing this to the effects of driving on winter road conditions (where additional roughness is
present almost half the year), driving a worn truck, and having higher vibration when driving with
low/no payload, the exposure of Norwegian drivers is also expected to exceed the Action Value of
0.5 m/s2. Unacceptably high daily exposures of ride vibration are therefore a confirmed problem for
professional truck drivers in the ROADEX partner areas.

Figure 5 Resulting A(8) values form Lapland, Highlands, Northern Norway and Sweden
 18

3.2. POOR TRUCK RIDE QUALITY IN FINNISH LAPLAND

The demonstration project in Finnish Lapland resulted in A(8)-values in the range 0.56 – 0.59 m/s2,
after normalizing to daily work patterns. This daily exposure exceeds the EU Action Value of 0.5
m/s2. Truck hauliers in Lapland with employees working under similar or worse conditions* to those
in this study, should carry out a risk assessment with associated measurements to clarify if actions
have to be taken to protect their drivers from health and safety risks caused by ride vibration and
mechanical shocks.

*Road roughness IRI 1.8 mm/m (plus worse roughness due to frost actions and corrugated ice-cap
in winter / early spring), vehicle type & condition, speed and daily hours at the steering wheel.

The average spinal compression stress during the two days of testing was modest; Sed = 0.42
MPa. The value 0.42 MPa is below the health caution value of 0.5 MPa in the standard ISO 2631-5
on human response to transient vibration. However, occasional observations were made of spinal
compression stress over 0.5 MPa, as per calculations from transient accelerations in the seat pad.
The worst contributor to the average compression stress came from the bumpy Road 956/957 at
Raattamaa. Truck drivers operating on similar roads for a large part of their working day, especially
during winter/spring periods with additional frost related roughness, are likely to suffer special
health risk due to high spinal compression stress.

A special root cause of high ride vibration on the route in Lapland was extreme settlements in
sections were the road was founded on permafrost, which is exposed to long-term thawing due to
solar heat captured by the black road surface. Another problem on many roads in cold climate is
freezing boulders rising up through the road.

Many of the unhealthiest bumps come from man-made problems: improper edges of patch repairs,
and improperly built transitions between bridges and road. These kinds of problems are of course
preventable, and can be relatively quickly eliminated by improving road work practice.

3.2.1. The route, the transport task, the truck and its instrumentation
The location of the studied route in Finnish Lapland is shown in Figure 6.

Figure 6 Location of the Lapland demonstration route

The demonstration vehicle carried a cargo of brewery products and groceries. The demonstration
route was 763 km long as shown in Figure 7. The demonstration started on 21 September 2012 on
National Highway 21 (a.k.a. European Highway 8) northbound from the Kurth Oy truck garage in
 19

Pello up to Kilpisjärvi, on the border to Norway. From Kilpisjärvi the route returned south on
Highway 21 to Palojoensuu, where it went east on Highway 93 to a stop-over in Enontekis. On the
second road day, the truck was driven from Enontekis south-east on Road 956/957 via Raattamaa
and then Highway 79 westbound to Muonio. In Muonio, the route again took Highway 21
southbound to the starting point in Pello at the Kurth Oy garage. A note of historical interest: The
main part of Highway 21 was built by prisoners of WW2, while Finland was under German
occupation.

Figure 7 Map of the Finnish demonstration route

The haulage partner for the demonstration project was Kurth Oy from Pello, operating the Scania
124 L420 seen in Figure 8.

Figure 8 The Scania truck used for the ride quality study in Finland

The truck license plate was “EYP 258”, and it had some 677,000 km on the meter at the test. The
driver seat was a Scania original mounted air-suspended ISRI seat. All tyres were made by
Michelin. The steer axle had XZE2+ 315/70 R22.5 tyres, inflated to 7.85 and to 8.15 bar pressure.
The drive axles had pair-mounted X-IceGrip STUDLESS SNOW 295/80 R22.5 XDW. The bogie
axle had pair-mounted X Pilote XZA 295/80 R22.5. A remarkable note is that on both drives and on
bogies there was no realistic “easy access” to the tyre air valves. This is further discussed in the
section “TPCS prevents driving with under-inflated tyres”.

The transport task was such, that the trailer was detached and parked at a stop in Muonio. Thus,
the trailer was only included on Highway 21 from Pello to Muonio and from Muonio back to Pello.

The truck was instrumented as per Figure 9 with accelerometers on left and right side at both the
front axle and on the truck frame, tri-axial accelerometers at the base and on the pan of the driver
seat, a webcam, an interior microphone and a six-axial inertial/GPS unit at the centre of the cab
with dual antennas on the roof of the cargo cabinet.
 20

Figure 9 Instrumenting the Scania test truck in Pello

3.2.2. High daily vibration exposure; A(8) exceeds the EU Action Value
The first day of the trip was spent on Highway 21 northbound from Pello up to Kilpisjärvi. From
Kilpisjärvi the route returned southbound on Highway 21 to Palojoensuu, where it went east on
Highway 93 to a stop-over in Enontekis.

Normalization of the test journey to daily work patterns was done using the UK Health and Safety
Executive´s “Vibration Calculator”. The resulting A(8) value for the 512 km driven on the busy day
1 was 0.56 m/s2, as per the calculation in Table 2. This is above the EU Action Value A(8) = 0.5
m/s2.

Table 2 Truck driver’s vibration exposure, day 1 (7 hour driving time)

Vibration intensity Exposure time Partial exposure
Distance [km]
m/s² hours minutes m/s²
Pello - Muonio Truck w trailer 0.65 1 45 0.304 144
Muonio - Palojoensuu Only truck 0.55 0 40 0.159 48
Palojoensuu - Karesuvanto Only truck 0.48 0 35 0.130 37
Karesuvanto - Kilpisjärvi Only truck 0.63 1 28 0.270 109
Kilpisjärvi - Karesuvanto Only truck 0.63 1 28 0.270 109
Karesuvanto - Enontekis Only truck 0.53 0 53 0.176 65

Pause, non-driving time 0.00 1 12 0.000 0

Daily exposure value, m/s² A(8) 0.56 512

On the second day, the truck travelled south-east on the small Road 956/957 from Enontekis via
Raattamaa and then westbound on Highway 79. In Muonio the route again took Highway 21, now
southbound, back to Pello. The A(8) value for the 251 km driven on the half day was 0.42 m/s2, as
per the calculation in Table 3. After normalization to an 8 h “normal full day” work pattern of similar
conditions, as per Table 4, the A(8) value for a daily route of about 500 km would be about 0.59
m/s2. This is above the EU Action Value A(8) = 0.5 m/s2. Truck hauliers with employees working
under conditions similar to those in this study, should make a risk assessment with associated
measurements to clarify if actions must be taken to protect their drivers from health and safety
risks from ride vibration and mechanical shock.
 21

Table 3 Truck drivers vibration exposure, day 2 (only 3.5 h driving time)
Vibration intensity Exposure time Partial exposure
Distance [km]
m/s² hours minutes m/s²
Enontekis - Raattamaa Only truck 0.65 58 0.226 63
Raattamaa - Särkijärvi Only truck 0.68 30 0.170 33
Särkijärvi - Muonio Only truck 0.53 8 0.068 11
Muonio - Pello Truck w trailer 0.65 1 45 0.304 144

Pause, non-driving time 0.00 4 36 0.000 0

Daily exposure value, m/s² A(8) 0.42 251

Table 4 Truck drivers vibration exposure, day 2 (normalized to 8 hours work; 7 hours of driving)

Vibration intensity Exposure time Partial exposure

Distance [km]
m/s² hours minutes m/s²
Enontekis - Raattamaa Only truck 0.65 58 0.226 63
Raattamaa - Särkijärvi Only truck 0.68 30 0.170 33
Särkijärvi - Muonio Only truck 0.53 8 0.068 11
Muonio - Pello Truck w trailer 0.65 1 45 0.304 144
Similar driving 0.63 3 30 0.417 251

Pause, non-driving time 0.00 1 6 0.000 0

Daily exposure value, m/s² A(8) 0.59 502

 22

3.2.3. Thawing permafrost is one of the root causes of high ride vibration
Many sections of Highway 21 are very undulating, with frequent dips and crests. The subarctic
climate is very cold and some parts of the road are built on permanently frozen (“permafrost”) soils.
In frozen condition, soft soils such as peat and clay can have high bearing capacity almost
comparable to cured Portland cement concrete. After paving the road with black asphalt, increased
absorption of solar energy makes the ground warmer. This causes the soil beneath the road
structure, otherwise permanently frozen, to begin to thaw during the warmest summer days. As the
peat slowly thaws, it dramatically loses bearing capacity. As a result thawing permafrost can bring
extreme settlements, as seen on the photographs in Figure 10, and thus very poor ride quality.

Figure 10 Thawing permafrost under Highway 21 gives severe settlements and a rough ride
 23

3.2.4. Occasional high spinal compression stress

For the measurements in Finnish Lapland, the average spinal compression stress over the two
days was modest; Sed = 0.42 MPa. The worst contributor to the average stress came from the
bumpy Roads 956 and 957 from Enontekis, via Raattamaa, to Särkijärvi. The value 0.42 MPa
measured is below the health caution value of 0.5 MPa in the standard ISO 2631-5 on human
response to transient vibration. However, occasional observations were made of spinal
compression stress over 0.5 MPa using calculations of transient accelerations in the seat pad.
Truck drivers operating on roads like Roads 956 and 957 for a large part of their working day,
especially during winter/spring periods with additional frost related roughness, are likely to suffer
health risks due to high spinal compression stress. Some illustrative examples of root causes to
unhealthy7 truck ride in Lapland are given below.

Two large man-made problems causing unhealthy truck ride are improper transversal joints (road
profile steps) and high road profile slope variance at patch repairs and at transitions between
bridges and road. The problem with improper joints at patches is illustrated in Figure 11. The graph
at the bottom shows the vertical acceleration (green) and the root-mean-square of x,y,z vibration
(blue) along 2 minutes of the ride. The yellow marker shows the position along the graph where the
webcam picture was taken. Analysis of the webcam pictures showed that all of the transients
noticeable in the bottom green graph had occurred at freshly made asphalt patch works (here only
a sample is showed).

Good practice in the ROADEX Partner areas recommends that the transversal edge of the new
asphalt should not be placed “layer-on-layer” with the old asphalt, but cut down into the same
elevation as the surface of the old asphalt. For this, good asphalt patch works require the use of a
small milling machine. Such machines are available on the market, but are currently not used in all
highway pavement maintenance projects.

Figure 11 Patch repair zone at Highway 21: Intense mechanical shocks on the driver seat

7
Note: The graphs do not report spinal compression stress [MPa], only seat accelerations [m/s2] – the input
used when computing spinal compression stress. The compression stress goes high in sections where seat
acceleration is high. Thus, high seat acceleration indicates high spinal stress.
 24

At many bridges, the backfill behind the bridge front wall is inadequately compacted during
construction. This can result in severe settlement at the bridge joint, and thus also to intense ride
vibration in vehicles. Measurements taken in the study show numerous cases of high compression
stresses in the truck driver’s spine at improper bridge joints, such as seen in the 1 minute graph at
the bottom of Figure 12.

Figure 12 Intense ride vibration at an improper bridge backfill

There are of course also non-manmade root causes to road features that excite high spinal
compression stress. One example is pavement edge slump, where the lack of a robust pavement
shoulder has led to deformation of the pavement edge as seen in Figure 13. This kind of road
damage can result in intensive ride vibration as shown in the 1 minute graph at the bottom of
Figure 14.

Figure 13 Exterior of test truck approaching the pavement edge damage seen in Figure 14
 25

Figure 14 Intense truck ride vibration at deformed pavement edge on Road 956/957
 26

3.2.5. Observations of hazards and various road features in Lapland

3.2.5.1. LATERAL ACCELERATION EXCEEDS CARGO SECURING DESIGN?
An objective for cargo securing is to assure that the goods and the vehicle are kept tight together,
in other terms: that they accelerate similarly. The accelerations of an object under transport can be
classified in three groups:
 Single event shock / transient vibration
 Low-frequency motion
 High-frequency vibration
The design acceleration values in the IMO/ILO/UN ECE Guidelines for cargo securing have been
developed with reference to accelerations with durations that can be described as quasi-static, or
at least as having low frequency. All bodies that are not ideally stiff will experience oscillating
elastic deformations during transport. Acceleration in most points on the object will therefore have
peak values higher than the Centre of Gravity (CoG) of the object. In order to slide or tilt the object,
the CoG acceleration must exceed certain intensity. The Guidelines refer to this intensity related to
the CoG of the goods rather than to its surface. When measuring vibration in one or a handful of
points of the object for high frequencies, the recorded amplitudes will most certainly be significantly
higher than the amplitudes at the CoG. However, it is unclear what the relation is between
amplitudes at the CoG and at any other point of the object, depending on acceleration duration or
frequency. In railway transport, recorded peak amplitudes at frequencies above 8 Hz are filtered
down to as little as 5 %. It is not clarified how to filter amplitudes at frequencies between 2 and 8
Hz, but below 2 Hz no filtering should be applied. A simple conclusion is that if lateral acceleration
exceeds the 0.5 g limit of the Guidelines, (0.5 g = 5 m/s2), there is a risk of overloading the
capacity of the cargo latches. This risk should not be dismissed, until the character of the
accelerations has been investigated in detail. An example of this from a truck passing over edge
damage at 80 km/h can be seen in Figure 15, where the truck cab experienced a transient lateral
acceleration of 5.9 m/s2 as seen in the bottom graph. This clearly exceeds the 0.5 g lateral
acceleration to be considered when latching loads on road vehicles, as per the IMO/ILO/UN ECE
Guidelines for cargo securing.

Figure 15 Hazardous truck cab lateral vibration at deformed pavement edge on Roads 956/957
 27

A similar example of lateral buffeting from Highway 21 is shown in Figure 16. Here the cab lateral
acceleration peaked at 7.5 m/s2, with an intensity of about 3.5 m/s2 during a half second.

At roads with pavement edge damages as severe as those seen in

Figure 15 Hazardous truck cab lateral vibration at deformed pavement edge on Roads 956/957
Figure 15 and Figure 16, accelerations can be so high that there may even be a risk that cargo
latches break. As result of this, and the motion of the heavy cargo (slide or tilt), the truck may
crash. These accelerations should therefore be investigated beyond the scope of this report.

Often investigations of truck crashes do not include competent analysis of the role of the properties
of the road. This is in fact true also in many investigations of fatal crashes. Given this context, the
new RBCSV road condition parameter could be a useful tool to identify road damages that may
have contributed to a rollover crash.

Road sections with severe edge damage can yield a RBCSV value above the limit value 0.30 %,
which was established in the ROADEX III project on “Health Issues Raised by Poorly Maintained
Road Networks”. Pavement edge damages worse than RBCSV 0.30 % should therefore be
repaired as soon as practically possible, with a warning sign being raised until the road repair is
completed. Truck operators should also educate their drivers to drive slowly at this kind of
pavement damage.

Figure 16 Hazardous lateral truck vibration at deformed pavement edge on Highway 21

 28

When driving on many roads in the cold climates of the EU Northern Periphery, other sources of
both roll-related lateral buffeting and high compression stress in the driver’s spine are freezing
boulders rising up through the road. Every winter the road structure in cold climates heaves due to
frost. In case of a large boulder in the structure, during the spring thaw period, fine material may
migrate down through the soil via voids and come to rest under the boulder. This makes it
impossible for the boulder to sink down as much as the rest of the road. This process is repeated
annually. So as years pass by, the boulder rises more and more by the accumulation of fines
beneath. Eventually the rising boulder will create such a large bump, that it will require to be dug
out by an excavator. This is a very costly road repair action in paved roads but, as seen in example
in Figure 17, a rising boulder can result in hazardous lateral buffeting of the truck (here 6.3 m/s2).
The high cost for the repair of rising freezing boulders can therefore be paid back by the large
benefit of preventing hazardous lateral buffeting of heavy trucks and buses with a high positioned
centre of gravity (CoG).

Figure 17 Lateral buffeting of test truck at a rising boulder on the road centreline on Highway 21
 29

3.2.5.2. IMPROPERLY BANKED CURVES

Several outercurves on the demonstration route in Finnish Lapland were perceived to be
improperly banked. One example was at the sharp curve at Kilpisjärvi, as seen in Figure 18.

Figure 18 Test truck at an improperly banked outercurve on Highway 21 in Kilpisjärvi

With accurate data from a laser/inertial Profilograph (see Figure 3) averaged over 1 m steps, it is
relatively easy to objectively identify sharp outercurves with adverse cross slope (camber). Such
curves have a high need for side friction, which can bring a high risk for loss-of-control crashes.
Risky outercurves are recognized by having a combination of negative curvature and negative
cross slope, as seen in the left hand graph in Figure 19. Curvature expressed as 1000/radius [m] is
plotted on the x-axis, while crossfall [%] is plotted on the y-axis. By filtering the Profilograph
database, it is easy to find the coordinates of the improperly banked road curve, such as marked in
the right hand map of Figure 19.

Figure 19 Profilograph data reveals an improperly banked sharp outer-curve at Kilpisjärvi

 30

3.2.5.3. WILD ANIMALS AND HAZARDOUS ROAD WORK SITES

Reindeers are nice and cute in front of Santa´s sledge, but they do post a serious crash risk when
in front of a car at highway speed on an icy road. Wild reindeers, as seen in Figure 20, are a
common crash risk in Lapland. Roadside fencing may be an effective measure against wild
animals on these roads.

Figure 20 Wild reindeers are a crash risk on Lapland roads

At several road work zones, collision hazards were present without relevant warnings for stationary
or low speed heavy road construction machines on the road. An example is shown in Figure 21.

Figure 21 An excavator blocking part of the National Highway 21 without any warning signs
 31

3.2.5.4. NOISY ROAD SURFACES AND BUMPY FROST DAMAGES

Long sections of the Finnish roads surveyed showed surface damage “ravelling”, as illustrated in
Figure 22. This kind of damage brings many waves in the 50 – 500 mm Megatexture band.
Megatexture is known to cause high tyre/road noise, bringing vehicle internal noise as well as
external noise.

Figure 22 Pavement surface ravelling at Highway 21

Severe frost-related permanent damages as seen in Figure 23 are also common in Lapland.

Figure 23 Road damages caused by frost action in the late winter season
 32

3.3. POOR TRUCK RIDE QUALITY IN THE SCOTISH HIGHLANDS

The demonstration project in the Scottish Highlands resulted in A(8)-values in the range 0.66 –
0.90 m/s2 at two repeated round-trip routes near Ft William. A single one-way trip from Corpach to
Inverness resulted in 0.55 m/s2 for a half day of driving. However, after normalizing to a full 8 h
working day, driving from Ft William to Inverness and back on A82 resulted in A(8) = 0.65 m/s2. For
all three tested routes, after normalizing to 8 hour daily work patterns, the resulting exposures
exceeds the EU Action Value of 0.5 m/s2. Therefore truck hauliers in the Highlands with employees
working under similar or worse conditions* to those in this study, should carry out a risk
assessment with associated measurements to clarify if actions need to be taken to protect their
drivers from health and safety risks caused by ride vibration and mechanical shocks.

*Road conditions, vehicle type & condition, speed and daily hours at the steering wheel.

The average daily spinal compression stress during the three days of measurements in the
Scottish Highlands was systematically high; Sed = 0.9 – 1.1 MPa. This is far above both the health
caution value of 0.5 MPa and above the “high health risk” value of 0.8 MPa in the standard ISO
2631-5 on human response to transient vibration. One of the worst contributors to the average
compression stress came from Road B8004 from Gairlochy to the A82 (bound for Spean Bridge).
This road had a high incidence of long-wave unevenness.

A root problem of the high ride vibration in Scotland is long-wave undulating road profiles. This
problem is particularly pronounced at ancient bridges, many of which have a slight arc-shaped
road profile.

3.3.1. The routes, the transport task, the truck and its instrumentation
The location of the demonstration area in Scottish Highlands is shown in Figure 24.

Figure 24 Location of the Highland demonstration area

The vehicle for the demonstration was a timber truck hauling timber on round-trips between timber
storage areas, at either Loch Arkaig or South Laggan, and the BSW Sawmills (Kilmallie) in
Corpach. The final test trip was a one-way journey from the timber storage at Loch Arkaig to the
Norboard OSB factory in Inverness.

The three demonstration routes were in total 238 km long as shown in Figure 24. Due to repeated
round-trips on two routes, the total measured distance was 512 km. The first route was a round-trip
of about 60 km between BSW Sawmills (Kilmallie) Ltd in Corpach and Loch Arkaig, mainly on
roads B8004 and B8005, as seen in Figure 25. The second route was a round-trip of about 70 km
between the same sawmills and South Laggan, mainly on road A82 as seen in Figure 26. The third
 33

route was a one-way drive of 153 km from Banavie via Loch Arkaig to Inverness, mainly on road
A82 as seen in Figure 27. The measurements were made on 4th to 6th October 2010.

Figure 25 Overview of the demonstration route between Corpach and Loch Arkaig

Figure 26 Overview of the route between Corpach and South Laggan

 34

Figure 27 Overview of the demonstration route from Loch Arkaig to Inverness

The haulage partner was Ferguson Transport (Spean Bridge) Ltd, operating the 480 hp Volvo FH
seen in Figure 28, and some 30 other trucks.

Figure 28 The TPCS-equipped Volvo truck used for the ride quality study in the Highlands

The truck license plate was “N40 AFT”, and it had some 187,000 km on the meter at the test. The
driver seat was a Volvo original mounted air-suspended seat. The truck carried a Jonsered 1080
timber crane.

The steer axle had Michelin XZE2+ 315/80 R22.5 tyres, inflated to 7.95 and to 8.15 bar pressure.
The drive axles had pair-mounted tyres. Left side had Michelin 315/80 R22.5; maybe model XDY3
(scratch damages made it impossible to read clear). Right side had Bridgestone M729 315/80
 35

R22.5. The bogie axle had pair-mounted Bridgestone R297 315/80 R22.5. The tyres on the drive,
bogie, as well as on the triple axles on the wagon were connected to a “Tireboss” Tyre Pressure
Control System (TPCS). The inflation pressure in all tyres, except on the steer axle, was varied
during the ride vibration measurement trips. The TPCS had 6 different settings that the driver could
set with a quick press on the buttons to adapt to truck operation mode; low/high speed,
loaded/unloaded and on/off highway.

The truck was instrumented as per Figure 29 with accelerometers on the left and right side of both
the front axle and on the truck frame, tri-axial accelerometers at the base and on the pan of the
driver seat, a webcam, an interior microphone and a six-axial inertial/GPS unit at the centre of the
cab with dual antennas on the roof. Note the deep crack in the rubber bumpstop, marked with a
red circle, indicating an intensive bump exposure before the ROADEX tests.

Figure 29 Instrumenting the Volvo test truck at Ferguson Transport’s excellent workshop
 36

3.3.1.1. PHOTOS FROM THE DEMO-ROUTES

The road to Loch Arkaig was very narrow and the timber storage was on a stony and very rough
dirt road, see Figure 30.

Figure 30 Photos from the Loch Arkaig route

The timber storage area at South Laggan was on a muddy and thus very soft dirt road. There was
a road work zone on the A82 public road on the route back to the BSW sawmill in Corpach, see
Figure 31. The test truck had to wait in line, as traffic was directed to pass by the zone where
asphalt was being milled off. This caused a bias in the measured ride vibration exposure, reducing
repeatability between round trips.

Figure 31 Photographs from the South Laggan route

 37

3.3.2. High daily vibration exposure; A(8) exceeds the EU Action Value
The measurements commenced with several round-trips between the BSW Sawmills in Corpach
and the forest stockpiles at Loch Arkaig. The first trips were taken just as the driver normally did,
using the tyre pressure control system to adapt the inflation pressure to the truck operation mode
(speed, load etc). Then a round-trip was made with full inflation pressure. During this round-trip,
the vibration logger bounced from the truck dashboard and fell hard on to the cab floor. This gave
a tangible proof of a harsher ride when the TPCS was not in use. The floor crash caused some
damage to some cable parts in the vibration logging system. Afterwards, the data from the
subsequent trips to Loch Arkaig with full inflation pressure were found to be corrupted and had to
be discarded. After changing cables, the measurements were completed under normal
circumstances.

For three daily typical round-trips between the BSW sawmill and Loch Arkaig, under normal use of
the TPCS, the A(8) value was 0.77 m/s2, as per the calculation in Table 5. This is above the EU
Action Value A(8) = 0.5 m/s2.

Table 5 Truck drivers’ vibration exposure, Sawmill – Loch Arkaig with active TPCS
TPCS active Vibration intensity Exposure time Partial exposure
Distance [km]
m/s² hours minutes m/s²
Sawmill - Loch Arkaig Empty 1.00 37 0.278 25
Dirtroad Empty 1.53 9 0.210 2.3
Loading - 0.00 30 0.000 0
Dirtroad Loaded 1.28 9 0.175 2.3
Loch Arkaig - Sawmill Loaded 0.79 37 0.219 25
Unloading - 0.00 20 0.000 0
Sawmill - Loch Arkaig Empty 1.00 37 0.278 25
Dirtroad Empty 1.53 9 0.210 2.3
Loading - 0.00 30 0.000 0
Dirtroad Loaded 1.28 9 0.175 2.3
Loch Arkaig - Sawmill Loaded 0.79 37 0.219 25
Unloading - 0.00 20 0.000 0
Sawmill - Loch Arkaig Empty 1.00 37 0.278 25
Dirtroad Empty 1.53 9 0.210 2.3
Loading - 0.00 30 0.000 0
Dirtroad Loaded 1.28 9 0.175 2.3
Loch Arkaig - Corpach Loaded 0.79 37 0.219 25
Unloading - 0.00 20 0.000 0
0.000 0
Pause, non-driving time 0.00 54 0.000 0

Daily exposure value, m/s² A(8) 0.77 164

For three daily typical round-trips between the BSW sawmill and South Laggan, under normal use
of the TPCS, the A(8) value was 0.70 m/s2, as per the calculation in Table 6. This is above the EU
Action Value A(8) = 0.5 m/s2. As can be seen from the calculation, the three round-trips to South
Laggan make a long day. The measurements on Day 2 included three such round-trips and took
11 hours, pauses included.

For three daily typical round-trips between the BSW sawmill and South Laggan, without using the
TPCS to reduce the inflation pressure when appropriate, the A(8) value was 0.86 m/s2, as per the
calculation in Table 7. With TPCS engaged, the A(8) was lowered from 0.86 to 0.70 m/s2, a
reduction of approximately 18 %. This figure is based on a few measurements and at very low
speeds, so the uncertainty is significant. Using the TPCS, the vibration was 6 % higher when the
 38

truck was unloaded, than when loaded. The same relationship without TPCS was 13 %. This
makes sense, since the TPCS is able to reduce vehicle body pitch resonance from rear wheel
hopping in the unloaded condition. The steer axle was not fitted with TPCS.

Table 6 Truck drivers’ vibration exposure, BSW sawmill – South Laggan with active TPCS
TPCS active Vibration intensity Exposure time Partial exposure
Distance [km]
m/s² hours minutes m/s²
Sawmill - S Laggan Empty 0.87 31 0.221 31.5
Dirtroad up & down, incl loading Empty / loaded 0.66 1 14 0.259 9
S Laggan - Sawmill Loaded 0.82 32.5 0.213 31.5
Unloading - 0.00 20 0.000 0
Sawmill - S Laggan Empty 0.87 31 0.221 31.5
Dirtroad up & down, incl loading Empty / loaded 0.66 1 14 0.259 9
S Laggan - Sawmill Loaded 0.82 32.5 0.213 31.5
Unloading - 0.00 20 0.000 0
Sawmill - S Laggan Empty 0.87 31 0.221 31.5
Dirtroad up & down, incl loading Empty / loaded 0.66 1 14 0.259 9
S Laggan - Sawmill Loaded 0.82 32.5 0.213 31.5
Unloading - 0.00 20 0.000 0
0.000 0
Pause, non-driving time 0.00 6 0.000 0

Daily exposure value, m/s² A(8) 0.70 216

Table 7 Truck drivers’ vibration exposure, BSW sawmill – South Laggan without using TPCS

TPCS off Vibration intensity Exposure time Partial exposure

Distance [km]
m/s² hours minutes m/s²
Sawmill - S Laggan Empty 0.93 31 0.236 31.5
Dirtroad up & down, incl loading Empty / loaded 0.97 1 14 0.381 9
S Laggan - Sawmill Loaded 0.82 32.5 0.213 31.5
Unloading - 0.00 20 0.000 0
Sawmill - S Laggan Empty 0.93 31 0.236 31.5
Dirtroad up & down, incl loading Empty / loaded 0.97 1 14 0.381 9
S Laggan - Sawmill Loaded 0.82 32.5 0.213 31.5
Unloading - 0.00 20 0.000 0
Sawmill - S Laggan Empty 0.93 31 0.236 31.5
Dirtroad up & down, incl loading Empty / loaded 0.97 1 14 0.381 9
S Laggan - Sawmill Loaded 0.82 32.5 0.213 31.5
Unloading - 0.00 20 0.000 0
0.000 0
Pause, non-driving time 0.00 6 0.000 0

Daily exposure value, m/s² A(8) 0.86 216

 39

For a single one-way trip from Corpach/Banavie via Loch Arkaig to the Norboard factory in
Inverness, the exposure was 0.55 m/s2, as per the calculation in Table 8. Despite being only a half
working day, still the exposure is above the EU Action Value A(8) = 0.5 m/s2.

Table 8 Truck drivers’ vibration exposure, Banavie - Loch Arkaig - Inverness with active TPCS
TPCS active Vibration intensity Exposure time Partial exposure
Distance [km]
m/s² hours minutes m/s²
Banavie - Loch Arkaig Empty 1.00 37 0.278 24
Dirtroad Empty 1.53 9 0.210 2.3
Loading - 0.00 30 0.000 0
Dirtroad Loaded 1.28 9 0.175 2.3
Loch Arkaig - B8004 (Gairlochy) Loaded 0.84 28 0.203 17
B8004 Loaded 0.94 5 0.096 4
A82 (1) Loaded 0.63 32.5 0.164 28
A82 (2) Loaded 0.77 34 0.205 37
A82 (3) Loaded 0.70 27 0.166 27
A96 Loaded 0.56 9 0.077 9
Unloading - 0.00 20 0.000 0
0.000 0
Pause, non-driving time 0.00 4 0.000 0

Daily exposure value, m/s² A(8) 0.55 151

Using the above result of “6 % higher ride vibration when the TPCS truck was unloaded”, a rough
estimate of a full working day on a round-trip Spean Bridge (Ft William) – Inverness has been
calculated. The estimated A(8) value is 0.65 m/s2, as per Table 9. This is significantly above the
EU Action Value A(8) = 0.5 m/s2.

Table 9 Truck drivers’ vibration exposure, Ft William – Inverness round-trip on A82

TPCS active Vibration intensity Exposure time Partial exposure
Distance [km]
m/s² hours minutes m/s²
A82 (1) Loaded 0.63 32.5 0.164 28
A82 (2) Loaded 0.77 34 0.205 37
A82 (3) Loaded 0.70 27 0.166 27
A96 Loaded 0.56 9 0.077 9
Unloading - 0.00 20 0.000 0
A96 home Empty 0.59 8 0.077 9
A82 (3) home Empty 0.74 26 0.173 27
A82 (2) home Empty 0.82 32 0.211 37
A82 (1) home Empty 0.67 31 0.170 28
Loading - 0.00 20 0.000 0
A82 (1) Loaded 0.63 32.5 0.164 28
A82 (2) Loaded 0.77 34 0.205 37
A82 (3) Loaded 0.70 27 0.166 27
A96 Loaded 0.56 9 0.077 9
Unloading - 0.00 20 0.000 0
A96 home Empty 0.59 8 0.077 9
A82 (3) home Empty 0.74 26 0.173 27
A82 (2) home Empty 0.82 32 0.211 37
A82 (1) home Empty 0.67 31 0.170 28
0.000
Pause, non-driving time 0.00 18 0.000 0

Daily exposure value, m/s² A(8) 0.65 404

 40

3.3.3. Systematic high spinal compression stress

The average daily spinal compression stress during the three days of the measurements in the
Scottish Highlands was systematically high; Sed = 0.9 – 1.1 MPa. One of the worst contributors to
the average compression stress came from Road B8004 from Gairlochy to the A82 (bound for
Spean Bridge). This road had several long-wave bumps. These bumps are difficult to see on
photographs, but are easy to “perceive” when watching the video from the webcam on the truck´s
dashboard.

The measured spinal compression stress exceeded both health caution values in the standard ISO
2631-5 on human response to transient vibration. Exposures below 0.5 MPa are claimed by the
standard to bring “Low probability of an adverse health effect” and over 0.8 MPa to bring “High
probability of an adverse health effect”. In Sweden, Sed = 0.5 MPa is frequently used as the action
value for maximum compression stress. Truck drivers operating like the test driver in this Highland
study, with Sed = 0.9 – 1.1 MPa, are at high health risk due to intense spinal compression stress.

Truck hauliers with employees working under conditions similar to those in this study, should carry
out a risk assessment with associated measurements to clarify if actions need to be taken to
protect their drivers from health and safety risks from ride vibration and mechanical shock.

3.3.4. Undulating profile at bridges is a root cause of high ride vibration

The round-trips of the study included crossings of several ancient and often beautiful small
bridges. Many of these bridges have an arch-shaped road profile, with high amplitude and short
ramps; such that at normal low volume rural road speeds the road profile acts as a speed hump
causing vehicle body bounce. These small bridges can be extreme “hot-spots” when it comes to
poor truck ride quality. A further bridge problem encountered during the demonstration was a
damaged and rough bridge deck. Below are some examples of unhealthy8 truck ride over bridges.

The first “unhealthy” example is a bridge with a severely damaged deck on Road B8005, as shown
in Figure 32. The graph at the bottom of Figure 33 shows the vertical acceleration (green) and the
root-mean-square of x,y,z axis vibration (blue) along 30 seconds of the truck ride at 36 km/h. The
yellow marker shows the position where the webcam picture was taken. The maximum vertical
vibration (frequency-weighted as per ISO 2631-1) was 6.1 m/s2. The Maximum Transient Vibration
Vector Value (MTVVV) averaged over 1 sec was 5.1 m/s2. Both these values are extremely high
and the latter value corresponds to “extremely uncomfortable” (exceeding 2 – 2.5 m/s2) as per the
discomfort scale in the ISO 2631-1 standard.

8
Note: The graphs do not report spinal compression stress [MPa], only seat accelerations [m/s2] – the input
used when computing the spinal compression stress. The compression stress goes high in sections where
seat accelerations are high, so high seat acceleration is an indicator of high stress in the spine.
 41

Figure 32 Damaged bridge deck at Road B8005 from Banavie to Loch Arkaig

Figure 33 Powerful bumps at one of the ancient bridges on Road B8005

 42

The second example of an “unhealthy” bump is an arch-shaped bridge, also on Road B8005, as
shown in Figure 34 and Figure 35. Travelling northbound at 56 km/h, the truck almost went
airborne, when it was accelerated vertically with 0.9 G. Take-off occurs at 1 G (10 m/s2). When the
truck went southbound at the lower speed of 43 km/h, the driver incurred a maximum vertical
vibration of 5.7 m/s2, as seen in the 40 second long graph. Note that the intensive vibration started
tens of metres before the bridge itself, when riding southbound. This indicates that the causes of
the truck ride problem included settlements in the backfill behind the bridge abutment.

Figure 34 The Northbound truck almost went airborne (0.9 G) at the bridge on B8005

Figure 35 The truck felt severe bumps also when Southbound on B8005
 43

The problem with bumpy rides at bridges was present also on national road A82, as seen in Figure
36. The frequency-weighted vertical acceleration of the driver seat peaked at 5.9 m/s2; almost 0.6
G. This can be compared to the 0.7 G (unweighted) vertical acceleration being used in many
countries as design target value for traffic calming speed bumps on urban low speed streets.

Figure 36 The truck incurred severe bumps on a bridge on the A82 bound for Inverness
 44

3.3.5. Rough ride on patch work

Patch repair can be necessary before an overlay is funded. However, the result of the patch work
is often still a rough road, as can be seen in the results from highway A82 in Figure 37.

Figure 37 Rough truck ride over patches on A82 bound for Inverness

3.3.6. Bumpy transversal joints

In the Highlands, as well as in the other ROADEX Partner Areas, a frequent bumpy problem is the
transversal joint at the start of overlays and patch repairs. See the example in Figure 38.

Figure 38 Impulse-response in the truck at a rough transversal joint on A82

 45

3.3.7. Bounce resonance at long wave unevenness

As reported in the ROADEX III case study on Road 331 in Sweden, truck suspension systems can
magnify bounce, pitch and roll vibration at frequencies in a range between 0.7 and 3 Hz due to
resonance. At highway speeds around 80 km/h (50 mph), these frequencies are excited by
unevenness with wavelengths in a range of about 7 - 31 m. An example of truck resonance on an
uneven section of highway A82 from Ft William to Inverness is shown in Figure 39.

Figure 39 The truck suspension shows resonance at low frequencies / long wave unevenness

This kind of long wave unevenness is difficult to repair with a traditional asphalt overlay, since the
paver tends to “roller-coast” in undulations of more than some 10 m long. Good practice in repair of
such unevenness is careful filling of the hollows, and preferably also gentle milling of the ridges,
before resurfacing. The procedure should preferably be done after laser-scanning the undulations,
computer aided design of the undulation repair, and computer aided control of the milling works.
Extensive experience from Sweden shows that the final paving layer should preferably not be laid
with computer aid due to the complexity of controlling the paving plants floating screed. The final
paved wearing course layer should be laid on the prepared surface with correct slopes, preferably
created with a milling machine controlled with computer aid.
 46

3.3.8. Precision in the measurements of truck ride quality

The three series of measurements of truck seat vibration data recorded on A82 with full payload
from South Laggan bound for Corpach are shown in Figure 40. It can be seen that there is
noticeable variance between the three series of data, with standard deviation 0.13 m/s2 in an
arbitrary section. Such variance is inevitable however, given the complex nature of the truck driving
task with differences in truck speed and in the lateral driving position on the roads (roads are
generally rougher closer to the pavement edge). However, the differences between the three
series are averaged out when a longer section is analysed instead of only a single section. The
average values for the three series of 12 km were 0.744, 0.734 and 0.745 respectively. The
minimal difference shows that there was no significant bias between the three measurements.

From an overall perspective, the demonstrated precision was very good down to the second
decimal place. Any single measurement, longer than a couple of kilometres, gives a robust
estimate of truck ride quality on the measured route. There is no practical need for repeated
measurement, unless the ride conditions are changed (which is a matter of reproducibility, not of
repeatability). Three examples of significant change in conditions are: 1. resurfacing the road, 2.
rideability during summertime and during wintertime on a road with severe frost actions, and 3.
differences between a driver who avoids potholes and a driver who does not avoid road surface
damages by yawing and/or braking.

Figure 40 Three repeated measurements of truck seat vibration

 47

3.3.9. Observations of hazards and various road features in Highlands

3.3.9.1. IMPROPERLY BANKED CURVES
Just as with the route in Finnish Lapland, the route in the Scottish Highlands included several
outercurves that were seen to be improperly banked. One example was the sharp curve shown in
Figure 41. Rollover and various other types of loss-of-control crashes cluster at this kind of “flat
curve”, indicating that they are extremely hazardous. With accurate data of crossfall and curvature
from a road network condition survey, it is easy to identify curves with adverse camber, such as
demonstrated in the data from Finland in Figure 19. A more advanced analysis of improperly
banked outer-curves, relating to design codes for geometry of new roads, was presented in the
ROADEX III study “Health Issues Raised by Poorly Maintained Road Networks” (2008).

Figure 41 A82 has many improperly banked and thus hazardous sharp outer-curves

3.3.9.2. NARROW ROADS AND SHARP CURVES

The test truck and trailer were modern vehicles with fairly high capacity and thus large dimensions
and weight. These types of wide long vehicles can be incompatible with narrow lanes and sharp
curves, as seen in evidence in the form of the deformed grass verge in Figure 42. The lack of
horizontal clearance at many sections of the test roads is visible also in Figure 43 and Figure 44.

Figure 42 Water pooling in deformed shoulder at the exit of a sharp inner-curve on narrow A82
 48

Figure 43 Hazardous rock edges at the pavement edge on Road B8005 at Loch Arkaig

Figure 44 Narrow sections cause frequent scratching on tyre sidewalls and fenders
 49

3.4. HAZARDOUS TRUCK OPERATION IN NORTHERN NORWAY

The demonstration project in Northern Norway resulted in an A(8) value of 0.47 m/s2, normalized to
daily work patterns. This is just below the EU Action Value of A(8) = 0.5 m/s2. The measurement
was made during very good conditions on European Highway E6 with a new truck, a full payload
and in the autumn. When considering the influence of factors like lower payload (empty trucks
vibrate more than full loaded), vehicle ageing and wear, corrugated thick ice on the road surface
and frost actions in the pavement, an annual generic estimate would be rather higher than the
Action Value. The average spinal compression stress during the two driving shifts was moderate;
Sed = 0.44 MPa. This is below the health caution value of 0.5 MPa in the standard ISO 2631-5 on
human response to transient vibration.

Truck hauliers operating in Nordland/Trøndelag with employees working under somewhat worse
conditions (late winter with frost damages, worn truck…) to those in this study, should still make a
risk assessment with associated measurements to clarify if actions need to be taken to protect
their drivers from health and safety risks caused by ride vibration and mechanical shocks.

An obvious risk for any truck driver on the tested E6 route is rollover crashing. This was firmly
demonstrated by the original test truck being crashed the night before the ROADEX
measurements. Roll-related lateral buffeting on slippery road surfaces is a serious risk factor that
all hauliers in Norway should address in their risk assessment. High rollover risk is, together with
high speed and poor driving, related to lateral forces (both quasi-static and transient), low friction,
as well as a lack of sufficient road width to recover control of the vehicle when an incident occurs.
Root causes to rollover crashes include sharp, long and egg-shaped curves, improper banking of
outercurves, lack of sufficient road width to recover control of the vehicle, deformation of weak
pavement shoulders and insufficient road friction management. The high rollover risk can be
reduced by implementing road signs with a radar-based display, giving warning for vehicles with
high curve approach speed.

3.4.1. The route, the transport task, the truck and its instrumentation
Originally, the ROADEX demonstration project in Norway was scheduled for a trip on the E6 Bodø
– Tromsø road, with a fish-truck from Transportsentralen Lillestrøm (TSL) on Saturday 15 October
2011. However, on the cold night between 13 and 14 October, the truck from TSL was destroyed
in a rollover-crash near Svartåstjønna, located between Smalåsen/Smeelehaesie and
Namsskogan. The crashed truck is shown in the left photograph (by Anne Brekkvassmo) of Figure
45. The crash site is showed in the right photograph, taken by J Granlund while riding the
replacement test truck. Fortunately the truck driver survived the crash with minor injuries.

Figure 45 The original test truck crashed between Smalåsen and Namsskogan
 50

With the kind assistance of the Norwegian Haulier’s Association (NLF), Kulseng-Hansen Thermo-
Transport A/S was identified as an alternative haulage partner with a replacement truck. Luckily
the owner, Reidar Kulseng-Hansen, had read the ROADEX III report “Health Issues Raised by
Poorly Maintained Road Networks” on the case study on Road 331 in Sweden. Since Mr Kulseng-
Hansen appreciated the ROADEX research, he was delighted to marshal one of his trucks to
reproduce the study; now on European Highway E6 in Norway. By this means the measurements
were able to be completed without delay and additional costs, just a change in route. With respect
to the ROADEX objectives, the final route was as good a test section as the originally planned test
route E6 Bodø – Tromsø. Since the final route was much longer than the original, the results are
even more representative for long haulage in the Northern Periphery.

The location of the studied route on the E6 from Fauske (Nordland) down to Trondheim
(Trøndelag) in Norway is shown in Figure 46. The test truck was carrying a cargo of chilled fish
from the Nordlaks Atlantic fish-harbour in Stokmarknes south to the market in Oslo.

Figure 46 Location of the Nordland/Trøndelag demonstration route on E6

The route Fauske – Trondheim, see Figure 47, was in total 636 km long. The traffic volume was
about 1300 AADT over the northern mountains (of which some 25 % were heavy trucks), and
some 18 000 AADT (10 % heavy trucks) close to Trondheim in the south. The speed limit was
mainly 80 km/h. About 1 hour drive south of Fauske, the E6 goes over Saltfjellet. This high
mountain area regularly causes problems in the winter. The road is typically totally closed 10 – 25
times per winter, with a similar number of occasions per winter when traffic is restricted for safety
reasons to travelling only in organized columns.

Figure 47 Overview of the demonstration route on E6 between Fauske (north) and Trondheim (south)
 51

The demonstration measurements were carried out on 14 and 15 October 2011. The ROADEX
haulage partner was Kulseng-Hansen Thermo-Transport A/S, operating a Volvo FH16 540 as seen
in Figure 48. The truck had license plate “KH 76414”, and was almost brand new. The driver seat
was a Volvo original mounted air-suspended seat. The steer axle had Michelin ENERGY 385/55
R22.5 tyres. The drive axles had pair-mounted Michelin XDN2 GRIP 315/70 R22.5. The length
between the first and the second axle was 3.40 m.

Figure 48 The Volvo truck used for the ride quality study on E6 in Nordland/Trøndelag

The truck was equipped with advanced equipment for safety and traction, such as a video camera
monitoring the “hidden angle”, on-board sand gritting aggregate, and a trailer rollover-stability
system (TRS) made by Haldex. It also carried snow chains, but these had never been used. As the
driver, Mr Pettersen, said “Chains are not good. They cause extreme noise and make any ice-
 52

capped road surface extremely rough, so all following trucks will get a harsher ride. When the road
condition is so poor that it calls for chains, then I simply pull over, park the truck and phone the
road maintenance contractor responsible for ploughing and gritting”.

Best practice for climbing icy long steep upgrades, proven over decades by native truckers in
Northern Scandinavia, is to increase the ground pressure on the drive axle by temporary raising
the bogie axle. This way traction is increased without time-consuming and hazardous near-
roadside work with mounting on/off snow chains. Despite the increased tyre/road pressure, this
practice is not recognized as causing additional damage to the road. The explanation is that when
the road is icy and stiff-frozen, it simply can withstand extreme traffic load. In fact, road agencies in
the Nordic countries consider abrasion by studded passenger car tyres as the only significant
mode for traffic-related wear of pavements in wintertime. The traditional practice of “bogie-lift” at
stiff-frozen icy upgrades is also not recognized as significantly damaging to truck components such
as axles and tyres, since bogie-lift is done during short periods and at modest speed. On the other
hand, snow chains are recognized for damaging the road surface, especially when used on a drive
axle with insufficient load. The result is wheel spin destroying the ice cap, often also causing
permanent damage to the underlying asphalt surface from the spinning steel chained wheels.

Mr Pettersen drove the route between Stokmarknes and Oslo 2 – 3 times per week throughout the
year. He had a very long driving experience, including several years transporting aid supplies in
Arctic conditions in Russia. His strategy is to drive with a good margin on the speed limit, to get a
relaxed but efficient ride with a minimum of braking and acceleration. This kind of driving results in
minimized fuel consumption and emissions. Mr Petersen’s experience is that the retarder (one of
the brake systems in the truck) tends to cause the wheels to lock on wet asphalt if set in position 3
and 4. In the winter, Mr Pettersen adjusts the settings in the truck computerized control system so
the retarder is in position 1 for the smoothest retardation, and hence the lowest skid risk, on
extremely slippery wet thin ice.

The truck was instrumented with tri-axial accelerometers at the base and on the pan of the driver
seat, webcam and an interior microphone, as seen in Figure 49.

Figure 49 Instruments in the Volvo test truck used on European Highway E6 in Norway
 53

3.4.1.1. PHOTOS FROM THE E6 DEMO-ROUTE

The E6 in northern Norway can be a very demanding road, as seen in Figure 50. In winter time,
many Norwegian professional drivers prefer, if possible, to use the 100 km longer route on the
European Highway E45 through Sweden instead.

As seen in the bottom right photograph, a newly resurfaced section was also in a hazardous
condition. The wearing course had delaminated from the old surface, obviously when a heavy
vehicle had braked hard. It is of no use to have good brakes on heavy trucks if pavement layers
aren’t glued with enough amount of tack-coat to each other, sufficient to withstand the brake
forces.

Figure 50 Photographs from E6 Fauske - Trondheim

 54

3.4.2. Daily vibration exposure and spine compression stress

The ROADEX measurements started just before midnight on 14th October 2011 in Fauske. Earlier
the same day, the truck driver had started his work with about a 5 hour long drive from the fish
harbour in Stokmarknes down to Fauske. Normally, long haulage road transport, as in this study,
contains a minimum of non-driving time. For the 8 hours of driving during the two working shifts
from Fauske down to Trondheim, the A(8) value was 0.47 m/s2, as per the calculation in Table 10.
This is below the EU Action Value A(8) = 0.5 m/s2. The longest section with “high partial exposure
per hour” was the first two hour drive South of Mosjøen. In theory it could be possible to exceed
the EU Action Value by driving repeatedly back and forth on this section. In reality however, no
transport task is designed for such “medium length haulage” without pauses for loading on/off the
truck and thereby lowering the vibration dose. So, at first glance, the truck operations on the E6
seem to be below the EU Action Value.

Table 10 Truck drivers’ vibration exposure on E6 Fauske – Trondheim; Loaded trailer

Vibration intensity Exposure time Partial exposure
m/s² hours minutes m/s²
Fauske to Mosjöen, stage 1 Loaded 0.41 0 0 0.000
Fauske to Mosjöen, stage 2 Loaded 0.53 7 0.064
Fauske to Mosjöen, stage 3 Loaded 0.57 37 0.158
Fauske to Mosjöen, stage 4 Loaded 0.42 1 51 0.202
Fauske to Mosjöen, stage 5 Loaded 0.43 19 0.086
Sleep rest in Mosjöen 0.000
Mosjöen to Trondheim, stage 6 Loaded 0.53 1 51 0.255
Mosjöen to Trondheim, stage 7 Loaded 0.52 44 0.158
Mosjöen to Trondheim, stage 8 Loaded 0.41 1 26 0.174
Mosjöen to Trondheim, stage 9 Loaded 0.43 32 0.111
Mosjöen to Trondheim, stage 10 Loaded 0.41 10 0.059
Mosjöen to Trondheim, stage 11 Loaded 0.47 8 0.061
Mosjöen to Trondheim, stage 12 Loaded 0.44 12 0.070

Pause, non-driving time 0.00 0 0.000

Daily exposure value, m/s² A(8) 0.47

However, the tests on the E6 in Norway were made under favourable conditions not fully
representative of a full round-trip operation, not for full year operation, and not for “worn truck”
operation.
The circumstances of the demonstration were:
 The new Volvo truck was in excellent condition.
 The road was neither covered with a thick rough ice layer, nor exposed to frost effects.
 The road was in its smoothest condition of the year. The measurements were done in the
autumn after maintenance and resurfacing operations during the summer.
 The trailer had a full payload. The return payload can be much lighter, as the fish harbours in
the north are much smaller freight receivers than Oslo, and other cities in the South. With
low/no payload, the northbound truck ride becomes much harsher.
Based on previous measurements, a modest estimate is that the ride would be at least 12 % worse
in empty mode compared with having a full payload. Including also the return trip with low/no
payload from the market down South back to the fish harbours in northern Norway, the A(8)
estimate would be 0.53 m/s2 (+12 % intensity) instead of the measured 0.47 m/s2 at full payload.
 55

Together with any corrections for higher vibration as the truck wears with use, and for rougher road
in the other seasons, the long term A(8) can be estimated to be definitely above the EU Action
Value A(8) = 0.5 m/s2.

For the measurements in Norway, the average spinal compression stress during the two driving
shifts was moderate; Sed = 0.44 MPa. This is below the health caution value of 0.5 MPa in the
standard ISO 2631-5 on human response to transient vibration.

In order to comply with the EU/Norwegian health and safety legislation on vibration at work, truck
hauliers operating in Nordland and Trøndelag with employees working under slightly worse
conditions (winter road, frost damages, worn truck…) to those in this study, should make a risk
assessment with associated measurements to clarify if actions need to be taken to protect their
drivers from health and safety risks from ride vibration. An obvious risk for any truck driver on the
tested E6 route is loss-of-control and rollover crashing. Roll-related lateral buffeting on slippery
road surfaces is a serious risk factor that hauliers in Norway should address in their risk
assessment.

3.4.3. Rough road profile at bridges is a root cause to high ride vibration
Just as in the ROADEX demonstration projects in Finland, Sweden and Scotland, the test on the
E6 in Norway showed that road/bridge joints can be extreme “hot-spots” when it comes to poor
truck ride quality. The problem is preventable however, as properly maintained pavements do not
give a ride problem. At the problem sites, the cause is typically settlements in backfills behind the
bridge abutment, or an improper road profile over the steel beam transversal joint. The latter is
often in the form of “steps” between the asphalt and the bridge joint, or the asphalt surface being
below the joint beam, instead of properly being some 3 - 8 mm above the joint.

Two typical examples with an unhealthy truck ride over rough road/bridge joints on the E6 are
shown in Figure 51 and Figure 52.

Figure 51 Unhealthy truck ride at a bridge on E6

 56

Figure 52 Unhealthy truck ride at a another bridge on E6

3.4.4. Bumpy transversal joints

A frequent bumpy problem, common to Norwegian Nordland/Trøndelag and other ROADEX
Partner areas, happens at transversal joints at the starts and ends of overlays and patch repairs.
Two examples of bad transversal joints are shown in Figure 53. The left photograph is of a fresh
made joint, while the right photograph is of a several years old and well-worn joint. These two
rough joints resulted in unhealthy mechanical shock in the truck driver seat, as can be seen in
Figure 54 and Figure 55. Another example of a bad new joint is given in Figure 56. While “joint
shocks” are typically lower than shocks at “natural” potholes and bumps, the joint shocks are more
or less man-made. Thus the joint shocks may be perceived by road users as more “unnecessary”
and frustrating, while they are obviously easy to prevent by improving the practice for asphalt
repair works on highways.

Figure 53 Improper transversal joints at new and old asphalt patch repairs on E6
 57

Figure 54 Mechanical shocks in the truck driver’s seat at a bad asphalt-joint

Figure 55 Mechanical shocks in the truck driver’s seat at the old bad asphalt-joint
 58

Figure 56 Mechanical shocks in the truck driver’s seat at another bad asphalt-joint
 59

3.4.5. Bumps at culverts

A common reason for road roughness is improperly installed drainage pipes or culvert crossings
under the road. In order to avoid bumps at culverts, the construction needs to follow certain
procedures. According to Scandinavian road construction codes, the trench for a piped culvert
should have the specified bottom width and excavated side slopes, making the trench upper width
typically some 25 – 45 m. Such large trenches and associated backfills are rather costly and hence
many culverts are installed with minimal trenches, often simply a vertical trench just wide enough
for the culvert. The result may be acceptable in some cases, but during the late-winter-to-early-
spring frost action period the results can often be very bumpy sections at poorly installed pipes. An
example of a bumpy culvert on the E6 is seen in Figure 57. This ROADEX measurement was done
in October, without any frost action, so this culvert bump is likely to be much worse in the spring.

Figure 57 Bumpy truck ride due to settlements at a culvert

 60

3.4.6. Poor truck ride quality at deformed weak road sections

While efficient road transports are carried out with long, wide and heavy trucks and buses, large
proportions of the Northern Periphery (NP) road network pavements lack the necessary bearing
capacities for these traffic loads. The ubiquitous result of this “overloading” is deformation of the
weakest pavement sections. While transversal deformation can be fairly homogenous, in the form
of rutting, longitudinal deformation is almost always heterogeneous. Thus weak pavements quickly
develop a rough road profile, which excites intense ride vibration. An example of a deformed weak
pavement section on the E6 is seen in Figure 58. A common reason for insufficient bearing
capacity in the NP is poor road drainage in the form of clogged or insufficient sized ditches.

Figure 58 Bumpy truck ride at a deformed and alligator-cracked weak pavement section
 61

3.4.7. Noise from thermo-cooler may disturb the drivers sleep rest
Excessive working time is a major cause of stress, depression and illness, and the purpose of the
driving and rest time regulation is to protect professional driver's health and safety. It is assumed
that during rest hours, a professional driver will get a good rest of at least 11 hours in any 24
hours. However, during the sleep rest between the long measurements in the demonstrations on
the E6 in Norway, the sleep in the truck cab was periodically seriously disturbed by noise from the
diesel-powered freight thermo-aggregate (the refrigeration unit at the front of the trailer). The in-
cab noise was measured at some 25 – 35 dB, with some transients up to 50 dB from cars stopping
and starting at a nearby gas station as seen the noise records in Figure 59 and Figure 60.
However, as the thermo-aggregate started up every 1½ hour, the noise rose to a sleep-stopping
level of 63 - 67 dB! Conventional thermo-coolers typically operate at 74 – 78 dB outdoor noise, but
the rear wall of the truck cab does provide some noise isolation. When wakened several times
during the stipulated sleep hours, the driver does not get proper rest and may become a drowsy
and hazardous driver. A fairly simple solution would be to install electric power outlets at truck rest
areas, to permit thermo-trailers aggregates to be operated on electricity (quiet mode) instead of on
diesel fuel (noisy mode) during the sleep hours. Providing infrastructure in format of power outlets
is indeed a road safety related service that could be provided by public authorities, rather than
organized by truck operators. The cost for the electricity should of course be paid by the hauliers
who use the facility, and who in turn will make corresponding savings on reduced diesel fuel
consumption by the thermo-cooler during the driver rest hours.

Figure 59 The driver´s sleep can be disturbed by noise from the freight thermo-cooler

Figure 60 Interior cab noise 65 dB from thermo-cooler disturbed the sleep at time 3700 – 4400 sec
 62

3.4.8. Narrow road width and other hazards in Nordland/Trøndelag

The E6 highway is a remarkably narrow national road, with long sections of only some 6 m width
and several bridges and railway underpasses with width down to 4 and 5 m.

When two long haulage trucks of 2.6 m width each meet in a sharp curve on a 5 - 6 m wide
highway, as seen in Figure 61, there is not much of safety margin for undesired motions. Some
heavy vehicle combinations also suffer from significant trailer off-tracking and lateral displacement
of the trailer, caused by rearward amplification of lateral accelerations from the prime moving truck
due to wind bursts or deformed pavements. Such phenomena can cause a lateral displacement of
up to some 0.8 m. As a result of low road standard, the E6 north of Trondheim suffers from
unusually frequent and severe crashes, both in meeting collisions and single-vehicle road
departure crashes.

Figure 61 The E6 highway north of Trondheim is remarkably narrow for long haulage transports
 63

3.4.8.1. IMPROPERLY BANKED SHARP CURVES, MANY ALSO “EGG-SHAPED”

Severe truck crashes are frequent on the road section from Stortjønna (in the north) via Smalåsen
(a.k.a. Smeelehaesie) to Namsskogan (in the south). Despite a moderate number of vehicles,
1330 vehicles/day AADT, there are severe crashes each and every year on this section.

When approaching the sharp curve shown in Figure 62, the truck driver said “This is a very
hazardous place”. The oncoming large truck seen in the photograph was forced out on to the weak
2.5 dm narrow unsealed shoulder to a full stop in order to avoid a collision of the outer mirrors of
the two trucks. The curve had been newly resurfaced. Despite this, it was clearly perceived from
the truck cab as being insufficiently banked. This view was confirmed by objective pavement
measurement. The data recorded on 1st Sept 2011 with a ViaPPS road profiler is presented in
Figure 63. Note that reference boxes in the figure are from the Swedish road design code, while
the Norwegian road design code calls for a higher superelevation rate (max 8 – 9.5 %, instead of
the Swedish max 5.5 – 6 %), and allows much sharper curves than the Swedish road design code.
The data in the graph shows that the curve has a severe adverse camber. If the negative cross
slope in the outer-curve is corrected to superelevation (see red arrows pointing up), the need for
side friction between road and tyres would be dramatically lower and hence the crash risk would
be much lower. In Norway, the maximum allowed superelevation when designing curves is 8 %
(plus an extra 1.5 % = 9.5 % when considering construction tolerance). The difference from the
current -3.3 % cross slope is therefore about 11 units of per cent. For a 2.75 m wide lane, the outer
edge needs to be lifted some 0.11 x 2.75 = +0.3 m in the worst 20 m section to reach the target +8
% superelevation. When raising this outercurve (at distance 1230 m in the Road Data Bank), the
new safety barrier seen in the photos requires to be dismantled and refitted after the pavement
works.

Figure 62 Narrow road for truck meeting on the sharp and improperly banked curve at Stortjønna

Figure 63 The Stortjønna outercurve needs to be banked up from -3.3 % into high superelevation
 64

At a short S-curve some 3 km south of Smalåsen, there have been three fatal crashes within 44 m,
as well as other fatal, and non-fatal crashes within the full length of the curve. Several of the
crashes in this curve involve heavy goods vehicles on long haulage. The curve has recently had
the safety upgraded, with new crash barriers and curve warning signs, as seen in Figure 64.
However, some parts of the recently resurfaced southbound inner-curve are adversely cambered.
At 14/720 km on the Road Data Base, the data from 25th August 2011 shows road geometry where
vehicles tend to be “pushed into the oncoming traffic”. Several of the crashes in the curve,
including two fatal, were head-on crashes. Road geometry data show that these sections need to
be corrected as per the long red arrows in the graph in Figure 65. There are also some sections
both in the innercurve and in the adjacent outercurve that need to be lifted up at the road edge in
southbound direction. This will strongly reduce the need for side friction, and hereby also reducing
the risk for crashes.

Figure 64 The extremely hazardous curve 3 km south of Smalåsen

Figure 65 The Smalåsen adversely cambered inner-curve needs reconstruction including lowering

Many of the curves on the E6 are “egg-shaped”, i.e. they go from straight into first wide radius and
then gradually into much sharper main curve section. This kind of road alignment is often designed
on purpose with a geometric element called an Euler-spiral or clothoid. Spirals were implemented
on railways in 1829, to decrease the wear of the rail and train wheels at the transition between
straight sections and curved segments. Spiral-shaped roads give a benefit to car drivers by
reducing the steering wheel torque needed, which is also believed to increase road safety. In the
USA, many states use spirals while others have never implemented them in their road design. A
study by Tom (1995) compared crash rates with and without spirals. The study showed that spirals
increase the crash rate, instead of decreasing the crash risk. One conclusion was that spirals give
vehicle drivers the impression of the curve being less sharp, and thus many tend to overestimate
 65

the maximum safe curve speed. Drivers entering curves with too high a curve entrance speed
must of course brake hard while cornering. Spirals may also “steal space”, so that the main curve
segment must be made sharper (and thus more hazardous) with spirals, than if spirals are omitted.
Today all multi-wheeled highway vehicles have power steering and thus the steering comfort is as
good without spirals on fairly wide roads.

The truck crash in October 2011 seen in Figure 45 took place at an S-curve some 600 m north of
the island Storholma, at about the same latitude as the small Svartåstjønna Lake. This curve has
been the scene of several other crashes in the past. However, the crash in October 2011 is not
registered in the database. The truck driver was a Swede working in Norway. He survived the
crash with minor injuries, while the truck had severe damages. This S-curve has several risk
factors. One of these is the poorly synchronized transition between the inner and outercurves, as
seen in the graph in Figure 66. “Fragments” of superelevation from the innercurve continue into the
outercurve, with a cross slope of -0.1 % at curvature 1.4 (radius 731 m). At another section of the
innercurve, the cross slope is adversely banked into +0.8 % superelevation. With such road
geometry, vehicles tend to be “pushed” into the oncoming traffic. To correct this, the cross slope at
the entrance of the outercurve, at about RDB-distance 6130 m, should be lowered by
reconstruction down to some -3 %. This is illustrated by red arrows in the graph. “Safe
combinations” of cross slope and curvature form a rather large zone, briefly illustrated by the blue
transparent field marked in the figure. Data in this blue area are proper combinations, as they
result in a modest need for side friction between tyre and road. The equation between side friction
demand and the factors horizontal curvature, cross slope and speed is derived and discussed
thoroughly in the ROADEX III report by Granlund (2008).

High skid risk due

to insufficient
Drainage Gradient
in sharp curve
section

Figure 66 Poorly synchronized transition at the hazardous Svartåstjønna S-curve

The current pavement geometry at the Svartåstjønna curve also raises an issue of insufficient
Drainage Gradient at the entrance of the innercurve, where the cross-slope is close to 0 % despite
a significant curvature. Unless there is a significant longitudinal grade in this section, the Drainage
Gradient (resultant to cross slope and grade) will also be very close to 0 %. This means that water
run-off is not assured, and that there can be a wide area of deep water during and after rainfall. At
freezing temperatures, this creates a risk for a local, but large, ice-lens formation at this section.

The driver of the test truck from Kulseng-Hansen, Mr Jean Pettersen, had a long experience of
truck driving. Mr Pettersen had experienced one truck crash during his career. This crash took
place on 5th January 2011 on the old deformed road in the S-curve seen in Figure 67 (photograph
by Google in April 2009). Both truck and trailer were totally destroyed as they hit the ditch at the
end of the S-curve. Also this truck crash is not registered in the Norwegian crash-database.
 66

Figure 67 The hazardous S-shaped and narrow crash-curve on E6, 2 km north of Fjerdingen

The photograph in Figure 68 was taken during the ROADEX measurement on 15th October 2011,
after new resurfacing. The data on the curvature and cross slope is taken from the Norwegian
Rosita database and include measurements from both 15 August 2011 and 19 April 2012. The
graph in Figure 69 shows that the geometry of the road is at least as hazardous in April 2012 as it
was in August 2011. Some mint-green dots from April are even in a worse position than the pink
squares9 from August. This may a result of frost heave in April, on resurfacing works being done
after August that did not accord with the road geometry, or on errors in the chain from the
Norwegian road profilometer to the Rosita database client.

Figure 68 The Fjerdingen S-curve is improperly banked after resurfacing 2011

9
Each point in this graph represents an average value for 20 m of road. The 20 m averaging in the two
series recorded different years has most likely started with an offset (1-19 m), so the points from nearby
sections may not be fully comparable in a curvy road section like this.
 67

Figure 69 The hazardous S-curve is improperly banked also after resurfacing

3.4.8.2. WILD ANIMALS INCREASE THE CRASH RISK

Wild animals, such as the reindeer, as seen in Figure 70, are a common crash risk on the curvy,
narrow and often ice-slippery roads in Northern Norway. Roadside fencing may be an effective
measure against these types of large wild animals on the roads.

Figure 70 Wild reindeers and other animals bring crash risk

A safety risk unique on high mountain areas was encountered on the Saltfjell. In the dark night,
lemmings suddenly crossed the E6 in such large numbers, so that there was no other option but to
run over countless of these animals. This caused a skid risk locally due to low friction between the
truck tyres and the road surface. Since the lemmings are so small, fencing would not be an option.

3.4.8.3. MOTION SICKNESS AND OTHER HAZARDS

When passing Trofors, the author felt clear symptoms of motion sickness, a product of
accumulated low-frequency (0.1 – 0.5 Hz) oscillations, like in a boat in rough sea. Motion sickness
syndrome negatively affects human performance long before the syndrome goes from diffuse
inconvenience, pallor and cold sweat to become so severe that the subject has to vomit.
Furthermore the negative performance effect may remain for many hours after the difficult ride.
Thus motion sickness syndrome may be underestimated as contributing factor in traffic crashes. It
is remarkable that a European highway, such as the E6, has such a poor alignment that an adult
man can experience motion sickness when travelling within the legal speed.
 68

The E6 route over Hamarøya north of Fauske (not included in the ROADEX demonstration route)
often has extreme problems with frost related roughness. Mr Pettersen said that these frost actions
in the pavement often cause severe damage to the bottom layers of Styrofoam boxes with fish
cargo, especially for drivers less familiar with the road and its bumps. Figure 71 shows a truckload
of fish in boxes (left), the principle of crushing the bottom layer of boxes (middle) and damaged fish
boxes as the truck arrives to Oslo.

Figure 71 Payload damage on frost damaged road sections

[Left pic: J Granlund; Mid pic: Dr K Chatti; Right pic: DB Schenker, Thermoterminalen Oslo]

Mr Pettersen had also experienced frost problems in tunnels, especially in Røros where freezing
leaking water and frost lift within the pavement reduced the posted 4.3 m vertical clearance to the
extent that 4.17 m high vehicles had been scratched. Mr Pettersen also called for more overtaking
lanes in flat terrain as well as at grades, since this would reduce stress, road rage, crashes, travel
time and fuel consumption.

Driving in the dark is a considerable issue according to Mr Pettersen, especially in the subarctic
region with almost no daylight at all in the cold and dark winter. Many urban road sections lack
streetlights. Rural sections often lack roadside poles. Many guardrails lack proper reflectivity. After
new asphalt has been paved, it often takes several months before the road markings are
repainted. In Norway, the truck owner can be fined if a truck has more than 2 extra headlights. This
is due to an EU regulation that is interpreted in a particular way in Norway than in, for example,
Sweden. Norway is actually not an EU-member, but is obliged to follow EU-regulations under the
Schengen Agreement. A typical Swedish truck with 6 extra headlights can be seen in Figure 74.
 69

3.5. POOR TRUCK RIDE QUALITY IN SWEDISH NORRLAND

The demonstration project in Swedish Norrland showed large differences in A(8) between late
winter / spring and autumn, as well as between driving with full tyre pressure at all times and using
a Tyre Pressure Control System to reduce the pressure when driving off-highway, at low speed
and when driving unloaded. The A(8) value was 0.91 m/s2 in the springtime test “without TPCS” on
the main route, while it was 0.86 m/s2 with the TPCS active and driving the same route 5 km/h
faster. In the autumn, the A(8) value of the main route was 0.66 m/s2. All of these values are above
the EU Action Value of A(8) = 0.5 m/s2. They also confirm similar magnitude of A(8) as the 0.76
m/s2 measured in the case study on partly the same route in ROADEX III in 2007/2008.

The average spinal compression stress during the springtime test was very high; Sed = 1.2 MPa on
the main route. This is far above both the health caution value of 0.5 MPa and the “High probability
of an adverse health effect” value of 0.8 MPa in the standard ISO 2631-5 on human response to
transient vibration. In the autumn test, the spinal compression stress on the same route was much
lower; Sed = 0.6 MPa despite a much higher speed. The lower stress level in the autumn test is still
above the Action Value Sed < 0.5 MPa used by the Swedish Work Environment Authority.

Truck hauliers operating in Norrland with employees working under similar conditions to those in
this study, are obliged by the Swedish AFS 2005:15 Vibrationer (“Vibration at Work”) regulation to
make a risk assessment with associated measurements to clarify if actions need to be taken to
protect their drivers from health and safety risks caused by ride vibration and mechanical shocks.

3.5.1. The routes, the transport task, the truck and its instrumentation
The location of the demonstration routes in Swedish Norrland is shown in Figure 72. A map of the
438 km surveyed main route is shown in Figure 73.

Figure 72 Location of the Norrland demonstration routes

 70

Figure 73 Map of the Norrland demonstration main route

The demonstration vehicle in the Norrland demonstration was a timber truck hauling logs from
inland forest areas to sawmill and paper pulp facilities located mainly on the Swedish east coast.
The 438 km long main route started at Ramsele, marked (I) on the map in Figure 73. The initial
section was on the Beaver Road 331 down to Österforsse (B). Then on National Highway 87 to
Sollefteå (C), a short section of National Highway 90, then along Road 335 to Överhörnäs (D), and
finally European Highway E4 to SCA Timber’s sawmill in Rundvik (E). The return trip from Rundvik
was on E4 south back to Överhörnäs (D), and then on to Road 348 to Bredbyn (F). From Bredbyn
the route was on secondary Road 1035 and Road 983 to Junsele (G), Road 346 and Road 977 to
Nordankäl (H) and finally Road 331 to the truck garage in Ramsele (I) again.

In addition to the main demonstration route, long single trip measurements were undertaken
between the forest areas in Edsele to sawmill in Kälarne, as well as in Bollsta. These however are
not shown on the map.

The ROADEX haulage partner in Sweden was Brorssons Åkeri AB from Ramsele, the same
partner as in the ROADEX III study of truck ride vibration at the Beaver Road 331. While most of
Brorssons 20 trucks are driven in a group served by a separate crane, the test truck was designed
for solitary operation. It carried a Jonsered Loglift for independent loading. Carrying a crane of
course increases the deadweight and reduces the payload. Sometimes the crane is left behind in
the forest to enable the full highway payload of the truck to be utilized. The Gross Vehicle Weight
of the truck was 60 tonnes, with some 40 tonnes payload capacity (when not carrying the Loglift
crane). The test truck also had a Tyre Pressure Control System. Thanks to the TPCS, this truck
was allowed to operate on many roads with reduced bearing capacity.

The test truck was the Scania R500 seen in Figure 74, with license plate “WUW 416” and license
plate “BJG 570” on the trailer. It had some 506,000 km on the meter at the spring time test, and
approximately 615,000 km at the time of the autumn test.
 71

Figure 74 The TPCS-equipped Scania truck used for the ride quality studies in Sweden

All tyres on the demonstration vehicle were manufactured by Michelin and all, including the steer
axle and trailer axles, where connected to a Tyre Pressure Control System, see Figure 75. The
steer axle had XZY3 385/65 R22.5 tyres. The twin drive axles had pair-mounted XDN2 GRIP
295/80 R22.5. The length between the first and the second axle was 4.90 m. The truck also had an
on-board sand gritting aggregate.

The trailer was fitted with XTY2 265/70 R 19.5 tyres.

 72

Figure 75 The TPCS controls all tyres on Brorssons truck, including on the steer axle
 73

The truck was instrumented as shown in Figure 76 with accelerometers on the left and right side at
both the front axle and on the truck frame, tri-axial accelerometers at the base and on the pan of
the Scania original mounted air-suspended ISRI seat, a webcam, an interior microphone and a
GPS unit.

Figure 76 Instrumentation in the Scania test truck

 74

3.5.1.1. PHOTOS FROM THE DEMONSTRATION ROUTES

The condition of Swedish forest roads varies greatly. Thick snow, as seen in the upper left
photograph in Figure 77, can in winter smear out potholes and other roughness on dirt-roads.
Private bridges are often simple and in poor condition. Before the commencement of any larger
haulage operations, the forest roads are normally reinforced and graded, as seen on the middle
photograph. In fact, private forest roads are often in better condition during timber logging
operations than public local roads are. The two bottom photographs in Figure 77 are from public
local roads in Edsele. The left photograph shows corrugated thick ice, while the right shows a
combination of corrugated thick ice with high steps up/down to the bare asphalt, with its severe
deformations and wide frost-related cracks. As the truck tyre hits such a step, a mechanical shock
is created and sent up through the truck and into the driver’s seat. The corrugations in the thick ice
on the poorly maintained public road caused terrible noise in the truck cab as well as to the exterior
environment.

Figure 77 Late winter on forest roads and local public roads in the Ramsele / Edsele areas
 75

During the late winter / early spring tests, numerous sections of the tested highways had severe
additional roughness (see examples in Figure 78) caused by frost actions in the subgrade soil
under the pavement.

Figure 78 Roughness with several decimetre amplitudes, caused by frost action

All road engineers in the Nordic know that the freezing of water leads to a volume increase of 9 %.
As a result many think that this increase in the volume of freezing water in the soil pores explains
the ground lift. However, this assumption is fundamentally wrong. When considering normal values
of void content, water content and soil compression, only a ground lift of 2-3 % can be explained in
this way. A ground lift of 2-3 % corresponds to only a few cm, as the frost frontier penetrates down
to some 2 m depth in the soil. As seen in the photographs above, already the differential in frost
heave between adjacent road sections can be several decimetres, and can thereby not be
explained by only pore water volume expansion of the frozen soil layers.

Janson discovered in 1914 that the explanation of large frost heaves is the result of additional
water flowing into the frozen section of fine-grained soils. This was later confirmed by other
researchers, such as Taber, Penner, Hou and Phukan. Taber also showed that the ground will
keep swelling when a soil is frozen from top to bottom when water can migrate through the soil
profile from an underlying water table. Soil continuously subjected to a negative temperature will
keep lifting. The ground surface will keep rising, due to the formation of segregated ice lenses, as
long as “the zero degree isotherm” - the freezing front - penetrates deeper into the soil. Beskow
 76

(1935) found out that most important parameter governing the ability of a soil material to swell due
to frost actions is its capillary characteristics given by its grain size distribution (Berglund, 2009).
Later research has shown that the grain size distribution affects the matric suction of the soil
material, while the amount of ionic compounds (i.e. road salts) affects the level of osmotic suction.
In cold climates the phenomenon of cryosuction also occurs. The features of soil that affect the
amount of unfrozen water available for freezing into ice-lenses and thus causing ground heave are:
the mineralogical properties of soil, the salt content, the soil granularity, the specific area of the soil
particles and surface tension. Further information on this subject can be found in the ROADEX e-
learning lesson on “Drainage”10, section 2: “Water in road materials and subgrade soils”.

When closely examining the frost-related road damages in Figure 78, it is clear that the damage
goes from side to side of the pavement, and that the soil has swelled much less in this section than
in the sections before and after. The explanation in this case is that a culvert has been installed
under the pavement in an improper way. A large differential frost heave has occurred as some 2 m
of road at the culvert lies founded on good road material, while the pavement in the adjacent
sections lies on frost-susceptible soil. A correct culvert installation in Sweden involves excavating a
25 – 45 m long “transition zone” and refilling with proper road materials, so that the difference in
frost heave is spread over an acceptable roughness wave instead of the abrupt and hazardous
“steps” seen in Figure 78.

3.5.2. The measurements

The spring time instrumentation was installed on 15 April 2011 and the first measurements were
carried out on 16 and 17 April 2011. The season can be described as “late winter” in the inland
forest areas of Ramsele, Edsele and Kälarne, while it was rather “early spring” at the East coast
with its industries.

The autumn instrumentation was installed on 25 September 2011, with second measurement
carried out on 26 September 2011.

For the springtime measurements, the 438 km long main route was measured with the TPCS
“active” (i.e. reduced tyre pressure on forest roads, at low speed, and when driving empty) and
then reproduced with the TPCS “inactive” (standard full tyre pressure on the whole route). In
addition, long single trips were also taken from several forests in the Edsele area to the Kälarne
sawmills, as well as the Bollsta sawmills. In total 1,417 km were measured in the spring time test.

In the autumn test, the 438 km main route was measured for reproduction, with the TPCS
inactivated at all time. The decision not to use TPCS low pressure modes was not made
voluntarily, but had to be taken to prevent a constant alarm noise from the TPCS control box as a
consequence of an “accident”. A new practice at the tyre garage servicing all Brorssons trucks was
to install so called balancing powder in the tyres to prevent tyre imbalance, instead of mounting
old-fashioned static balancing weights on the rims. Following this new routine, the TPCS-equipped
truck also had the powder installed (Brorssons only have one TPCS truck). The powder however
clogged the air valves of the TPCS so badly, that the air flow dropped to a minimum. Due to the
unusually slow airflow, the TPCS considered the exceptionally slow changes of tyre pressure as
indicators of a leaking puncture. This caused the TPCS alarm to sound for long periods as soon as
the TPCS settings were changed by the driver. In order to avoid this terrible alarm, the decision
was taken not to use the TPCS low pressure settings in the autumn test. This spoiled the planned
analysis of how the TPCS could isolate vibration as function of frequency (road roughness
wavelength x truck speed). However the overall effect of TPCS on vibration isolation was still
possible to assess.

10
Read more in the ROADEX e-learning on water in road materials and subgrade soils, Internet 2012-06-01:
http://www.roadex.org/index.php/drainage2
 77

Timber was picked up at various sites in the Ramsele and Edsele areas, and a number of short
forest roads in various conditions were driven. While many forest roads can be in terrible condition
for decades, they normally get an upgrade before a large timber logging operation is executed.
This results in quite smooth forest roads during haulage operations, such as seen on the middle
photograph in Figure 77. (The practice is obviously different in Scotland as seen in the
photographs from forest roads in the ROADEX demonstration at Fort William.)

For reason of precision, the ride vibration data from forest roads were omitted when comparing
data from the spring and the autumn tests in Sweden.

3.5.3. High daily vibration exposure; A(8) exceeds the EU Action Value
The A(8) value for the 438 km driven during late winter with TPCS “off” (i.e. constant full tyre
pressure) was 0.91 m/s2, as per the calculation in Table 11. This can be compared to the 0.86 m/s2
measured when reproducing the route at 5 km/h higher speed with the TPCS “on” (using lower tyre
pressure when not driving with full payload on highways). In both cases, the A(8) was clearly
above the EU Action Value A(8) = 0.5 m/s2. However the daily vibration was 6 % higher without the
TPCS, despite 5 km/h higher average speed when the TPCS was active.

These results show that the TPCS not only reduces ride vibration, but also makes it possible to
significantly increase productivity due to higher truck operating speed. However, the drivers’
exposure to ride vibration still is too high. A likely reason is that the road profile has high amplitude
at long wavelengths that the TPCS cannot isolate. The previous case study on Road 331 in the
ROADEX III project showed that this test road had a lot of unevenness with 7 – 31 m wavelengths,
causing resonance at low frequencies in several of the truck suspension systems. Such
unevenness cannot be isolated by TPCS, or regular truck suspension systems, but has to be
repaired by the road agency.

In the autumn measurements the roads were much smoother, since there were no frost-related
additional roughness due to swelling frozen soil under the pavement, and no corrugated thick layer
of ice on the road surface. On the very same route as studied in the late winter, Ramsele –
Rundvik, the A(8) was measured to be 0.66 m/s2 with the TPCS off, see Table 13. This value can
be compared to the 0.91 m/s2 measured with TPCS off in the spring. The increase in average daily
vibration was +39 % due to winter condition.

Table 11 Ramsele – Rundvik route, late winter / spring, TPCS off (driving time 6.4 hours)
Spring; pauses within most stages Vibration intensity Exposure time Partial exposure
Distance [km]
Without TPCS m/s² hours minutes m/s²
Ramsele - Rd 331 - Rd 87 Loaded 1.17 1 3 0.424 68
Rd 87 - Sollefteå - Rd 335 (break) Loaded 0.84 1 3 0.304 53
Rd 335 - E4 - Övik (break) Loaded 0.83 1 3 0.301 57
Övik E4 - Rundvik, unloading Loaded 0.84 1 3 0.304 49
Rundvik E4 Southbound (lunch) Empty 0.70 1 3 0.254 48
E4 - Rd 348, Rd 1035 Empty 1.00 1 3 0.362 70
Rd 1035, Rd 346, Nordankälvägen Empty 0.97 1 3 0.352 71
Nordankälvägen, Rd 331 - Ramsele (break) Empty 1.06 25 0.242 20
0.000

Av speed 68 km/h Daily exposure value, m/s² A(8) 0.91 438

 78

Table 12 Ramsele – Rundvik route, late winter / spring, TPCS active (driving time 6.0 hours)
Spring; pauses within most stages Vibration intensity Exposure time Partial exposure
Distance [km]
TPCS active m/s² hours minutes m/s²
Ramsele - Rd 331 - Rd 87 - SLÅ Loaded 1.02 1 3 0.370 73
Sollefteå - Rd 335 Loaded 0.93 1 3 0.337 75
Rd 335 - E4 - Övik (dinner) Loaded 0.48 1 3 0.174 17
Övik E4 - Rundvik, unloading Loaded 0.73 1 3 0.265 55
Rundvik E4 Southbound Empty 0.83 1 3 0.301 57
E4 - Rd 348, Rd 1035 Empty 0.95 1 3 0.344 62
Rd 1035, Rd 346, Nordankälvägen Empty 1.08 1 3 0.391 83
Rd 331 - Ramsele Empty 0.88 15 0.156 15
0.000

Av speed 73 km/h Daily exposure value, m/s² A(8) 0.86 438

Table 13 Ramsele – Rundvik route, autumn, TPCS off

Autumn Vibration intensity Exposure time Partial exposure
Distance [km]
Without TPCS m/s² hours minutes m/s²
Ramsele - Rd 331 - Rd 87 Loaded 0.96 57 0.331 74
Rd 87 - Sollefteå - Rd 335 Loaded 0.75 59 0.263 76
Rd 335 - E4 - Övik Loaded 0.77 15 0.136 17
Pause Loaded 0.00 40 0.000 0
Övik E4 - Rundvik, unloading Loaded 0.60 46 0.186 55
Pause Empty 0.00 20 0.000 0
Rundvik E4 Southbound Empty 0.69 45 0.211 56
Övik to Bredbyn Empty 0.79 32 0.204 42
Bredbyn Empty 0.61 5 0.062 6
Ödsbyn to Junsele Empty 0.76 39 0.217 47
Klappsjö to Lillterrsjö Empty 0.82 16 0.150 20
Klappsjö to Lillterrsjö II Empty 0.61 15 0.108 18
Tågsjöberg to Lillterrsjö Empty 0.95 7 0.115 8
Tågsjöberg to Lillterrsjö II Empty 0.84 7 0.101 7
Nordankäl to Ramsele Empty 0.53 10 0.077 13
0.000

Av speed 75 km/h Daily exposure value, m/s² A(8) 0.66 439

3.5.4. Systematic high spinal compression stress

The average spinal compression stress during the springtime test was very high; Sed = 1.2 MPa.
This is far above both the health caution value of 0.5 MPa and the “high health risk” value of 0.8
MPa in the standard ISO 2631-5 on human response to transient vibration. In the autumn test, the
spinal compression stress on the same route was much lower; Sed = 0.6 MPa. The lower stress
level in the autumn test is still above the Action Value for bumps, Sed < 0.5 MPa, used by the
Swedish Work Environment Authority.

Truck hauliers with employees working under conditions similar to those in this study should carry
out a risk assessment, with associated measurements, to clarify if actions need to be taken to
protect their drivers from health and safety risks from ride vibration and mechanical shock.
 79

3.5.5. Rough road profile at bridges is a root cause to high ride vibration
In Sweden, there are several specifications for road profile unevenness at transversal joints
between roads and bridges. The requirements differ remarkably between new constructions and
maintenance. Furthermore the requirements have changed significantly over the last few years.
However one common line that has stood the test of time in all requirements is that “The pavement
surface shall be located X mm above the steel beam at the joint”. (The value of X varies between 0
mm and 3 – 8 mm across the various specifications). An example of a joint that does not meet any
of these basic requirements can be seen in Figure 79. Not only is the pavement far below the steel
beam instead of properly being above it, but the paving contractor has also been so sloppy that a
thick puddle of asphalt has been spread onto the beam, making the step at the joint even higher
and rougher. This very rough joint can cause high vertical acceleration and shock load into the
spine of truck drivers, as seen in the response data in Figure 80 recorded in the late winter test run
at the modest speed of 39 km/h.

Figure 79 A very rough bridge joint at Nämforsen

Figure 80 Shock load into the truck drivers spine when passing the bridge joint at Nämforsen
 80

Another cause of truck ride problems at bridges is settlement in the backfill behind the abutment
wall. This kind of roughness brings harm to drivers, and also damage components in the truck.

The graph in Figure 81 shows the vertical acceleration of the outer wheel of the test truck as it
passed over the backfill at the bridge on Road 331 in Österforsse at 56 km/h. The acceleration
peaks at 83 m/s2, about 8 G. This kind of shock loads into the wheel assembly generates noise
and cause intensive wear of many components in the vehicle.

Figure 81 High acceleration of the truck wheel assembly at settlement in bridge backfill
 81

3.5.6. Several years later: Still no warning sign at identified hazardous sites
The ROADEX III case study in 2007/2008 identified several hazardous sites (HS) on the Beaver
Road 331. One of the worst HS was the edge deformation at Åkerö, reported in section 5.3.2.3
“Rock ´n Roll at HS Åkerö” in the ROADEX III report. In the 2007 survey, the pavement edge
deformation was 69 mm deep (see Figure 82) and exposed the timber logging truck to a transient
roll vibration of 5 /s, with lateral buffeting of 2 – 3.5 m/s2. This level of lateral buffeting could be
enough to cause a half-empty tanker truck with bad suspension to have a rollover crash. The
Åkerö HS was very clearly highlighted as a road safety hazard in the ROADEX III report,
underlined by the fact that a photograph of the Åkerö damage included in the report revealed
remains of an exploded truck tyre at the site. However, at the time of the ROADEX IV
demonstration on 26th Sept 2011, the Åkerö damage had still not been repaired. In fact, there was
not even a warning sign raised at the site.

Figure 82 The Åkerö edge deformation in 2007 [Source: The ROADEX III project]

The new response measurements in the ROADEX IV demonstration showed that the driver,
despite yawing to avoid the deepest part of the road damage, was exposed to 0.7 G peak vertical
acceleration within fractions of a second. This is the very same intensity used as design criteria for
passenger cars going at 30 km/h over traffic calming speed bumps on urban streets. The running
RMS over a 1 second time window of the seat vibration measured at the edge damage in 2011
was much higher than in 2007, as shown in Figure 83.

Calculations in MatLab show that the single pass with less than a second duration over the deep
road deformation in 2011 resulted in a shock load in the driver´s spine corresponding to a daily (8
hour) stress of 0.24 MPa. It will be enough with a handful of such shock loads to reach the Action
Value of 0.5 MPa on a daily 8 hour basis as used by the Swedish Work Environment Authority.

It is recommended that the local road organization should re-read the report of the ROADEX III
case study on Road 331 and, as soon as possible, raise the necessary warning signs and start
planning for road improvement at the documented hazardous sites.
 82

Figure 83 The seat vibration at the edge deformation in Åkerö was worse in 2011 than 2007

3.5.7. Remarkably poor quality in road repair works

The ROADEX IV main test route included a long part of Road 1035. A 5 km section of this road
west of Bredbyn had previously been rated by the County Council as unacceptably hazardous for
public traffic due to severe frost-related roughness and was given “heavy maintenance” during
2010/2011. The 5 km road repair work commenced with local reconstruction (excavation of frost
susceptible fine-grained soil down to a frost free depth and backfill with gravel) of some local
sections with extreme frost damages in the autumn 2010. In the summer of 2011 the entire 5 km
section was resurfaced.

In the ROADEX IV demonstration in March 2011, the instrumented timber logging truck was, for
reference purposes, followed by a passenger car with a portable laser/inertial road profiling
system, as shown in the bottom left photograph in Figure 84. The GE Laser Prof System comes in
a suitcase (upper photograph) and includes an advanced odometer. This wheel encoder is firmly
attached with strong magnets to the bolts of the left hand non-driven wheel of the car. The encoder
measures the travelled position, speed and longitudinal acceleration 20,000 times per revolution of
the wheel. In order to reach this high resolution, the sophisticated sensor uses optical photoelectric
scanning. This method detects even very fine lines, no more than a few microns wide, reading
periodic structures known as graduations from an etched disc of glass (discs illustrated in the
upper right photograph). This expensive and sensitive sensor was destroyed, see lower right
photograph, at one of several extreme bumps at the reconstructed Road 1035. The story of the
event is as follows:

During the 17th March 2011, some 600 km of frost-damaged rough roads had been measured by
the time the truck was unloaded in Rundvik. Before the trip back to Ramsele via Road 1035, the
mounting of the LaserProf System was inspected without any remarks. From Rundvik the truck
and the profiling car drove on the E4 to Överhörnäs and Road 348 to Bredbyn without problem. On
Road 1035 from Bredbyn there were several long-wave bumps/hollows with extreme amplitude at
sections that had been reconstructed just a few months before. At one of these bumps, the
powerful magnets could not keep the odometer in position, and it bounced off the car and was
destroyed, see lower right photograph in Figure 84.
 83

Figure 84 The portable LaserProf System and the odometer after being “killed” at a bump

The roughness of the “reconstructed” road sections of Road 1035, and the 3 dm “odometer killer
bump” in particular, is illustrated by the recorded truck driver seat response in Figure 85. In this
instance the driver was exposed to greater vibration levels than those measured in buses
traversing 10 cm speedbumps in city streets, and this was just some 6 months after the road had
been reconstructed.

Figure 85 Several high acceleration events at newly reconstructed sections of Road 1035

In the summer of 2011, the entire 5 km repair project section on Road 1035 from Bredbyn was
resurfaced. However, the new road surface became unacceptably rough. The truck seat vibration
during the demonstration on 26th Sept 2011 is shown in Figure 86. The average vibration intensity
was 0.6 m/s2. If the driver was to spend the whole day driving back and forth on this 5 km section,
the vibration exposure would exceed the EU Action Value A(8) = 0.5 m/s2.
 84

Figure 86 Road 1035 still gives a rough truck ride just after resurfacing 2011
Another example of a poor result following a road repair is given in Figure 87. Despite the truck
yawing to avoid the deepest part of the edge deformation in the new road surface, the truck was
exposed to a transient vibration approximately 20 times higher than the average ride vibration.

Figure 87 Transient vibration at an edge deformation on a newly repaired section of Road 335

Following the ROADEX III case study on the Beaver Road 331, more than 10 km of the road
between Nordankäl and Ramsele was reconstructed in 2008. Despite expensive reconstruction,
this section continues to show problems. The ISO 2631 frequency weighted truck seat vibration in
 85

2011 is shown in Figure 88. The average value was 0.46 m/s2, just below the EU Action Value of
0.5 m/s2.

Figure 88 Average vibration 0.46 m/s2 despite the road being reconstructed in 2008
 86

3.5.8. Observations of hazards and various road features in Norrland

3.5.8.1. SHARP AND IMPROPERLY BANKED HORIZONTAL CURVES
As discussed in the previous section, the reconstruction of some 10 km of the Beaver Road 331
between Nordankäl and Ramsele in 2008 resulted in a poor quality road in terms of roughness.
However, the reconstruction also resulted in an improperly banked curve, as seen in the
measurement results in Figure 89. Despite a sharp curvature of -1.9, corresponding to a curve
radius as tight as -526 m, the superelevation is only +0.6 % while the Swedish road design code
requires +5.5 % (see vertical arrows in the figure). The lack of superelevation in a sharp outercurve
causes a high need for side friction between tyre and road. One of the truck drivers at Brorssons
had a terrible experience there in the winter of 2009. A passenger car skidded and oversteered
before it gained grip again, then went straight across the road just in front of the timber truck with a
gap of only a handful of metres. This “close shave” may just as well have resulted in a fatal head-
on crash. With a proper cross slope, the car would have needed some 60 % less side friction,
which of course would have meant a much lower risk for the skid crash.

The section Nordankäl – Ramsele also has a very sharp right hand bend, with a curvature of up to
+4.7 (see red oval marking in Figure 89). This is remarkably sharp and hazardous for a 90 km/h
road.

Figure 89 One outercurve became hazardous low superelevation after reconstruction 2008

The ROADEX III report by Granlund (2008) included a discussion of the improperly banked Roos
curve. As seen in Figure 90, crashes have clustered there, showing that the curve is multiple times
more hazardous than the average for the road. This curve was unchanged at the time of the
demonstration survey in 2011. A photograph of the swaying test truck of 2011 is shown in Figure
91. As can be seen, the inner wheels almost have a lift-off due to severe weight transfer in the
hazardous outercurve.
 87

Figure 90 Crashes have clustered at the improperly banked Roos curve

Figure 91 The “Roos curve” on Road 331 between Österforsse and Helgum

A graph from the ROADEX III report by Granlund (2008) is repeated here in Figure 92. This figure
shows that the Roos outercurve needs to be banked up from -1 % cross slope up to the maximum
allowed superelevation +5.5 % (see vertical red arrow in the figure). Today, up to +7 %
superelevation is allowed when improving old sharp bends in Sweden.
 88

Figure 92 Profilograph data from the improperly banked, sharp Roos Curve

As seen in the results from the case study in Norway, the same circumstance is also quite
common with banked innercurves on the E6 in Northern Norway. Banked innercurves are very
hazardous, as they tend to “push the vehicle over into the oncoming traffic”. This kind of road
hazard is quite rare in Sweden. Instead, the main problem with innercurves in Sweden is excessive
negative cross slope, which may be nice for over-speeding traffic but can give a problem for
unskilled and insecure drivers at low speed on slippery roads. It is far from natural to have to turn
the steering wheel counter clockwise in order to take a right hand bend. However, innercurves do
not need to be banked into superelevation in order to be hazardous. It is in fact often enough to
have too little (negative) cross slope to make an innercurve dangerous. One example of this is
given from a section of the Beaver Road 331, as seen in the photo in Figure 93 and the
corresponding plot of road curvature and cross slope data in Figure 94. The photograph shows
that the trailer rear inner wheels almost have a lift off (caused by weight transfer to the outer
wheels). Note that the photograph is taken at a roll angle from a trailing car at speed in the sharp
curve, as can be realized from the lack of vertical alignment of the trees in the background of the
picture. As seen in the plot below the photograph, the negative cross slope in this sharp curve at
Mo should be increased from current 3.5 to 4.5 %, to (at least) 5.5 % in order to better comply with
the Swedish road design code for cross slope of innercurves.
 89

Figure 93 Unusual hazard: Insufficient crossfall of INNER-curve at Road 331 at Mo

Figure 94 Profilograph data from the sharp and too flat innercurve at Mo
 90

4. RELATING RIDE TO ROAD STANDARD

Most national public road administrations worldwide are surveying the condition of roads in their
road network, with respect to criteria such as Road durability, Winter operations, Noise, Debris,
Mobility, Road safety, Ride Quality, Wear and Tear and Energy Consumption. See the headers of
the vertical columns in Figure 95.

Each criterion is affected by several road properties, as seen in the horizontal rows of the figure.
Longitudinal roughness causes ride vibration. Due to its large impact on criteria such as Road
durability, Mobility, Road safety, Ride quality, Wear and tear as well as Fuel consumption, the by
far most important is the longitudinal roughness having most red-marked cells in Figure 95. When
considering normal conditions, megatexture (waves 50 – 500 mm), friction between tyre and road,
macrotexture (waves 0.5 – 50 mm), edge deformation, water pooling and bearing capacity are
secondary in importance. Rut depth11, crack index and other parameters are third in importance;
having mainly yellow, green or even white (no importance) marked cells in Figure 95.

For road safety and ride quality, the focus aspects of this report, the most important road
properties are longitudinal roughness, megatexture, edge deformation, water pooling and friction.
Cross slope has been given a low rank in the matrix. If it had been combined with horizontal
curvature and speed limit as well as longitudinal grade, or together with the mentioned properties
been transformed into “Need for side friction” and “Drainage gradient”, it would surely been top
ranked for road safety (as reflected in road design manuals world-wide).

Criteria
Wear and tear
Road durability Winter operations Noise Debris Mobility Road Safety Ride Quality Vehicle Tyre Goods Fuel consumption
Bearing capacity 3 3
Surface stiffness 1 2
Longitudinal roughness 3 2 2 1 3 3 3 3 2 3 3
Road property

Megatexture 2 2 3 1 2 2 3 3 2 3 3
Macrotexture 2 2 3 1 2 1 1 3 3
Cross slope 2 1 1 2 1 1 1 1
Edge slump 3 2 1 3 3 3 2 1 2 1
Rut depth 3 2 1 2 2 2 2 1 1 1 1
Water pooling 1 1 1 3 2 3 2 1
Friction 3 1 3 3 2 3 3
Retroreflection 2 2 2

Importance
Large 3
2
1
Small 0

Figure 95 Influence of road properties on road durability, safety, road user costs etc [After VTI]

11
Rut depth in Norway is measured (or rather: calculated) by a fundamentally different definition than in
countries following the “wire method” defined in the EN 13036 standard. The Norwegian definition can give
very high “rut depth” values for roads with a convex-shaped cross section, and for newly laid asphalt layers
without any ruts at all. This may lead to serious confusion, both in road contracts and in research.
 91

4.1. COMPARING VEHICLE DRIVER VIBRATION TO IRI-VALUES

The overall single most important road property of the matrix in Figure 95 is longitudinal road
roughness. There are several available scales for road roughness. The most used scale globally is
the International Roughness Index, IRI. This index is calculated with a computer-based “quarter
car model” as illustrated in Figure 96. IRI was defined by the World Bank in 1986 and describes
the vertical motion between the suspended vehicle body and the unsuspended wheel mass, when
driving at the reference speed 80 km/h. When calculating IRI, a set of reference parameters for
stiffness, damping and mass are used. These parameters define the “Golden car” used for the IRI-
calculation. See Ahlin & Granlund (2002) for details.

IRI may be computed from road profile data achieved by various methods, including static rod and
level. However road profile sampling must be made at maximum 3 dm long steps (see ASTM E
1364) and with very high vertical accuracy. For practical reasons, road condition measurements
are normally made with mobile high-speed laser/inertial profilometers as per the ASTM E 950
standard.

Figure 96 Quarter car model used for calculation of IRI

Since many road agencies are measuring and managing the condition of their road network based
on the IRI-scale, it is relevant to relate truck drivers’ exposure from whole-body vibration (WBV) to
IRI-values / road quality. The following sections report on relationships between WBV and IRI
established in some previous studies as well as in ROADEX IV.

4.1.1. Previous studies relating vehicle drivers WBV to IRI-values

4.1.1.1. PASSENGER CAR DRIVER VIBRATION
Mucka & Granlund (2012) made a review of reports from studies of relations between ride vibration
in passenger cars and road roughness on the IRI-scale. Most studies have resulted in linear
relationships as per Equation 1.

Equation 1 Vehicle seat or cab floor vibration aw as function of IRI-values

aw [m/s2] = a + b * IRI [mm/m]

The review of car studies showed ranges for the intercept “a” of 0 < a < 0.0768 and for the slope
factor “b” of 0.0665 < b < 0.27. Most studies had a coefficient of determination R2 better than 0.90,
a high value which indicates that IRI is a relatively good indicator of ride quality in passenger cars.

Ahlin and Granlund (2002) derived the physical relationship between IRI and cab floor vertical
acceleration in the “Golden car” reference model for the IRI computation. The derived relationship
in Equation 2 concerns the vehicle model´s “driving speed” v and the waviness w of the road
 92

profile elevation. Typical values for waviness are 1.6 < w < 2.4. For v = 80 km/h and w = 2, the
slope factor “b” equals 0.16. The intercept a-value was 0 (zero). Note that this study did not include
the effect of megatexture waves, which is a likely cause to why the intercept was a = 0.

Equation 2 Golden car cab floor vibration aw as function of IRI-values

aw [m/s2] = 0.16 * (v/80)((n-1)/2) * IRI [mm/m]

4.1.1.2. DRIVER VIBRATION IN HEAVY TRUCKS WITH TRAILERS

Ahlin et al (2000) measured seat vibration (as per the ISO 2631 standard) in timber logging trucks
and collated the drivers’ vibration exposure at 75 km/h to road roughness on the IRI-scale. The
collation was carried out using a linear relationship as per Equation 1.

Exploratory data analysis (EDA) showed that there were greater variances in seat vibration at road
roughness levels above IRI = 3 mm/m. This may be explained by the fact that while modest road
roughness is difficult to see, severe road damages are more visible. Hence most truck drivers try to
avoid traversing the worst roughness by changing the truck´s lateral position on the carriageway
(even to the extent of driving on the wrong side of the road), and by braking to lower the speed in
order to reduce vibration. Another explanation is that rough road sections with IRI > 3 mm/m cause
much more rotational vibration in roll and pitch modes, with associated lateral and fore-aft
vibration.

Regardless of the exact reason why there is a breakpoint for WBV (x, y, z vector) at IRI = 3 mm/m,
the EDA led to the conclusion that different regression analysis had to be made for IRI above and
below 3 mm/m. Results from Ahlin et al (2000) are given in Equation 3, Equation 4 and Figure 97.

Equation 3 Seat vibration (weighted x,y,z vector) in a truck with a trailer at IRI < 3 mm/m
WBV [m/s2] = 0.18 + 0.30 * IRI [mm/m]

Equation 4 Seat vibration (weighted x,y,z vector) in a truck with a trailer at IRI > 3 mm/m
WBV [m/s2] = 0.35 + 0.22 * IRI [mm/m]

Truck with trailer (75 km/h)

2,0

1,6
Dz [m/s 2]

1,2

0,8
Measured
0,4 Regression Fit

0,0
0 1 2 3 4 5 6 7 8 9 10
IRI [mm/m]

Figure 97 Truck drivers WBV (vertical direction only) as function of IRI. From Ahlin et al (2000).
 93

Note that according to Equation 3, and assuming an 8 hour daily driving time, the truck drivers
exposure to WBV exceeded the EU Action Value of A(8) = 0.5 m/s2 with as little as an IRI slightly
above 1 mm/m. Roads with IRI = 1 mm/m are considered by typical passenger car occupants to be
in good condition and give a very smooth ride. Note also that a significant part of the truck seat
vibration is related the equations´ a-term (the intercept), rather than to the b-term (the slope factor)
associated with road roughness reflected by the IRI-scale. With the a-value being as high as 0.18
(not to mention the 0.35 in Equation 4), the intercept makes a great proportion of the EU Action
value = 0.5 m/s2. If the intercept can be lowered, a higher IRI-value (rougher roads) can be
accepted without exceeding the EU Action Value for truck driver´s exposure to whole-body
vibration during the truck operation.

Experience from several other studies show that the factors a & b in Equation 1 vary between road
sections and between trucks. Typical ranges are for the intercept “a” of 0.1 < a < 0.35 and for the
slope factor “b” of 0.2 < b < 0.3. Bad roads and bad trucks results in higher factors.

A very important question is

-Why does the seat vibrate significantly despite no road roughness being present? Why isn´t the
intercept a = 0 (zero) at IRI=0?

Recalling the previous section, it was shown by Ahlin & Granlund (2002) that a = 0 for the idealized
“Golden car”. When studying real cars however, the intercept ranged up to a = 0.08 for passenger
cars. This may possibly be explained by the fact that IRI does not reflect megatexture waves < 0.5
m at all, while megatexture causes vibration in cars. Hence tendencies of washboarding, ravelling
and pothole formations may be present and cause ride vibration despite low IRI-values in both cars
and heavy trucks.

For heavy trucks, there are several other possible explanations why the intercept is not at zero:
• On weak roads, especially during the spring thaw period, “soft spots” in the road can give
high variance in pavement deflection under a heavy truck wheel (as indicated by falling
weight deflectometer testing of pavements). This causes ride vibration but is not measured
by IRI, as IRI is calculated assuming a 100 % stiff road profile.
• Truck wheels may suffer from severe unbalance, caused by geometric and stiffness
eccentricity. While wheel unbalance is more or less impossible to become unaware of in
any modern passenger car, the resulting vibration in a heavy truck may be masked by other
vibration and may hence not be easily discovered.
• Truck frames are flexible compared to stiff car bodies. Long wave road unevenness may
cause vibration due to “frame beaming”, while IRI is not sensitive to these long waves.
• The truck engine´s powerful combustion pulses may not be efficiently isolated from the
frame, cab and driver seat.

4.1.2. Relating truck driver WBV to IRI-values in the current ROADEX IV study
Data from two adjacent sections of road 331 recorded in the very same truck at constant speed are
presented in Figure 98. While the slope factor b is similar for both road sections (0.25 á 0.26), the
intercept a-factor differs by a factor close to 2 (0.12 versus 0.21). This indicates a large difference
in road properties. The section with IRI of 1.0 to 1.5 mm/m and intercept a-factor 0.21 had been
fairly recently resurfaced, but has probably low and varying pavement stiffness due to being built
on a subgrade of soft soil.
 94

Figure 98 Truck drivers WBV (weighted x,y,z vector) as function of IRI. Current ROADEX IV study.

4.1.2.1. USING IRI TO ESTIMATE TRUCK RIDE VIBRATION

This section offers an example on how to use IRI to estimate truck ride vibration, based on the
background in previous sections in this chapter. The road section example is from Hw 21 in
Finland, heading Northbound for Kilpisjärvi. Data in the Finnish PMS show that this section had IRI
= 1.81 mm/m. The example vehicle is a grocery truck. It is of a comfortable premium model, but old
and worn. To reflect this, ”low mid-range” values for a & b-factors have been inserted into Equation
1, resulting in WBV ≈ 0.18 + 0.23 * IRI, i.e. WBV ≈ 0.18 + 0.23 x 1.81 = 0.60 m/s2.

In reality, ROADEX IV measured the truck driver WBV to be 0.58 m/s2 for this case, so the
estimate was very close the measured value. The author of this report is convinced that with a fair
selection of a & b when using Equation 1, IRI can give WBV typically within some 15 %12.

The final step in assessing the daily A(8) vibration exposure is to use the UK HSE A(8)-calculator13
to normalise the WBV-value to daily driving hours.

12
For truck speed far below 80 km/h, IRI may overestimate ride vibration at rather long waves. This makes it
unsuitable to use IRI for estimating truck ride quality at undulating secondary/tertiary low speed roads.
13
UK HSE Vibration Calculator on Internet 2012-05-15: http://www.hse.gov.uk/vibration/wbv/
 95

4.1.2.2. USING THE EU ACTION VALUE TO IDENTIFY ROAD ROUGHNESS LIMITS

An overall objective for road management should be to keep pavements in such good condition
that it is possible to operate normal trucks over normal working days behind the steering wheel,
and still keep the A(8) under the EU Action Value 0.5 m/s2.

Many road agencies have limits for road roughness on the IRI-scale in their pavement
management. In most (all?) cases, these limits have been set with respect to passenger car
drivers comfort, and not with respect to the working environment of professional truck drivers. This
has led to a serious overestimation of acceptable road roughness, as shown below.

The discomfort scale in the ISO 2631-1 standard says that vibration is clearly “uncomfortable” at
0.8 m/s2. As seen in previous sections, the vehicle vibration response to road roughness (the b-
factor in Equation 1) for the reference “Golden car” is 0.16 at 80 km/h. With an intercept of a = 0,
road roughness can be as high as IRI = 5.0 mm/m before WBV corresponds to the “uncomfortable”
WBV = 0.8 m/s2 in normal cars.

When changing focus from car drivers comfort to truck drivers working environment and the A(8)-
concept implemented in Directive 2002/44/EC, a first step is to account for the truck driver’s one
hour of breaks with zero vibration.

Equation 5 Assessing the daily vibration exposure [AFS 2005:15 Vibrationer / Directive 2002/44/EC]

T
A(8)  A(T ) *
8

The effect of reducing driving time from 8 to 7 hours in Equation 5 is rather small. This is because
A(8) is affected by exposure time only in a “root” relationship, and is affected by vibration intensity
in a linear relationship. Using the UK HSE A(8)-calculator it is easy to compute that the acceptable
A(8) = 0.50 m/s2 at 8 vibration exposure hours daily, corresponds to 0.53 m/s2 after reducing the
exposure time T to 7 hrs/day. The increase in allowed vibration was only 0.03 m/s2 or 6 %, despite
a reduction of driving hours T by 12.5 %.

Inserting the ”low mid-range” values of “a” = 0.18 and “b” = 0.23 into Equation 1 and solving for
WBV < 0.53, the result is IRI < 1.52 mm/m.

By changing focus from car comfort to healthy and safe professional heavy truck operation, the
acceptable road roughness drops from being about 5 mm/m14 to some 1.5 mm/m!

Of course IRI < 1.52 mm/m is a target difficult to reach without a massive increase on road
maintenance funding. Thus two relevant questions are:
1. Is the EU Action Value of A(8) = 0.5 m/s2 really justified?
2. What else can be done to comply with the Action Value, besides shortening the interval
between asphalt overlays to prevent high IRI-values?
The first question is simple to answer; -Yes, the Action Value of 0.5 m/s2 is justified! Even without
reference to scientific knowledge reviews, it is enough to realize that WBV = 0.5 m/s2 corresponds
to “fairly uncomfortable” as per the ISO 2631-1 standard. You do not need to be expert on how
stress affects human health, in order to realize that negative health effects are likely in the long
term among many of those who are working in an environment that is “fairly uncomfortable” on the
average during the whole working day, year after year.

14
An example is STA Pavement Management standard 2011, available in Swedish on Internet (2012-06-30):
http://publikationswebbutik.vv.se/upload/6636/2012_074_Underhallsstandard_belagd_vag_2011.pdf
 96

The second question calls for more complex but also innovative answers. Truck seat vibration can
be lowered by many actions besides shortening the interval between asphalt overlays:
 Increasing pavement stiffness and decreasing pavement deflection variance (“dynamic road
roughness”). This can be achieved by better pavement bearing capacity management and
ultimately pavement reinforcement.
 Improving the initial asphalt smoothness and eliminating long wave unevenness by using
modern laser-based technologies for management of pavement geometry (crossfall and
profile) when resurfacing roads.
 Reducing initial asphalt megatexture; 50 – 500 mm waves.
 Improving the truck by equipping it with the best available seat and other technical solutions,
such as a tyre pressure control system or counteract balancing beads (or similar products)
for best possible and continuous balance of the wheel ends.

By combining the above bulleted actions, the road roughness can be allowed to go higher than IRI
= 1.52 mm/m also at roads with truck operating speed of 80 km/h.
 97

4.2. LATERAL BUFFETING – A SEVERE CRASH RISK

Road warpiness is a serious traffic safety problem for vehicles with a high Centre-of-Gravity (CoG),
especially on slippery icy roads. At sections where the lanes cross slope oscillates within a spatial
“window” of up to some 25 m length, vehicles can be exposed to roll vibration (both wheel axles
and vehicle body) and roll-related lateral buffeting.

An example of a road section with severe warpiness on Road 956 in Finnish Lapland is shown in
Figure 99. The weak and deformed pavement edge seen in the upper photograph has been
repaired, but has settled even more after the repair. The result is a strong transient roll and lateral
vibration in the ROADEX IV test truck, as seen in the screen dump below. High transient roll
vibration / roll-related lateral vibration were measured in all ROADEX IV demonstration projects.
This confirms that lateral buffeting is an issue in Scotland, Finland, Norway as well as Sweden.

Figure 99 Pavement edge deformation is a common cause of road warpiness

Lateral buffeting on icy roads may trigger the vehicle to skid. The lateral buffeting in other cases
can fool the driver that there is a skid, requiring a quick response with evasive manoeuvring
(against a non-existing skid) occasionally thereby initiating a real skid. See the photograph taken
by Niklas Thunborg in Figure 100 for a tragic example that occurred 20th October 2010 on
extremely slick thin black ice at a Hazardous Site with remarkably high road warpiness identified in
the ROADEX III project.
 98

Figure 100 A fatal skid crash on “black ice” at a section of Rd 331 with RBCSV >> 0.3 %

A new pavement condition parameter Rut Bottom Cross Slope Variance (RBCSV) was defined15 in
Sweden in 2006, aiming to identify road sections with severe road warpiness, causing truck cab
roll vibration and related lateral buffeting. The ROADEX III project validated the RBCSV-parameter
and drafted a limit value of max acceptable RBCSV = 0.3 % or 3.0 BAC.

In ROADEX IV the RBCSV-parameter has been further evaluated. Data on truck cab roll vibration
has been collated with data on pavement RBCSV from Road 331 in Sweden in Figure 101. The
road had severe pavement edge damages at about section 112,250 m, which are clearly detected
by both the truck roll sensor and by the RBCSV parameter. Truck cab roll vibration has also been
collated with data on pavement rut depth (Figure 102) and with road roughness IRI (Figure 103).
As seen by comparing these figures, pavement RBCSV is a good predictor of truck cab roll
vibration as opposed to both the rut depth and the IRI parameters.

Figure 101 Collations of Truck Cab Roll Vibration and Pavement RBCSV

15
The RBCSV algorithm is presented in English in the paper by Granlund (2010).
 99

Figure 102 Collations of Truck Cab Roll Vibration and Pavement Rut Depth

Figure 103 Collations of Truck Cab Roll Vibration and Road Roughness on the IRI-scale

Due to a phenomenon called rearward amplification, the highest lateral acceleration is typically
seen at the end of the trailer [Hurtig 2010]. Hence it can be argued that an indicator of road
warpiness should be designed to match trailer end lateral vibration, rather than truck cab vibration.
However, the amount of rearward amplification is greatly affected by truck and trailer properties,
and these properties are subject to vehicle regulations. In order to not over-tune the road
warpiness indicator to a certain type of truck and trailer combination, as well as keeping maximum
correlation to truck drivers’ exposure to ride vibration, the concept of keeping the RBCSV-
parameter designed to match cab vibration is justified.
 100

4.3. SEVERE FROST RELATED ROUGHNESS

This ROADEX IV project measured and compared truck ride vibration during summer and winter
condition in Sweden. The winter condition included mixed impacts of a corrugated ice layer on
some “late winter” inland road sections, and of uneven frost heave action at “early spring” roads,
see Figure 104.

The measurements were made in the same truck in both winter and summer, and comparison was
made on the very same route. This 438 km “main route” started with the loaded truck and trailer at
Brorssons truck garage in Ramsele, a journey to the Rundvik Sawmill were the timber was
unloaded, and finally back to Ramsele again. During the winter, the truck average speed was for
safety reasons 7 km/h slower than during the summer measurements.

Figure 104 The winter conditions included driving on corrugated ice layers and uneven frost heaves
The resulting daily vibration exposure A(8) values were:
 Winter, severe frost A(8) = 0.91 m/s2 (68 km/h)
 Summer, no frost A(8) = 0.66 m/s2 (75 km/h)

The truck drivers´ daily vibration was +39 % during the winter, despite a 10 % slower speed.

The problem with additional road roughness / ride vibration during winter conditions is present also
in Northern Finland and Northern Norway. The Norwegian Public Roads Administration has started
to include data on wintertime road roughness in their Pavement Management System database
“Rosita”. For a 10 km long section on European Highway 6, the measurements show a +17 %
increase in road roughness in the winter compared to in the summer as seen in Table 14.

Table 14 Road roughness at E6 Fauske – Trondheim section 17/000 – 27/000 km

IRI [mm/m] Measurement date
2011‐08‐25 2012‐04‐19
Average value 2.4 2.8

It is clear from this that additional winter roughness should be reflected in strategies for road
condition surveys in the Scandinavian part of the EU Northern Periphery. The road network
condition / rideability needs to be measured also when the roads are frozen and covered with an
ice layer.
 101

4.4. CORRUGATION IN THE ICE LAYER CAUSE NOISE AND VIBRATION

The Swedish Ambulance service reported16 in December 2011 that they had to avoid driving
ambulance transports on National Highway 90 north of Sollefteå, due to unhealthy interior noise for
both patients and staff. Noise measurements showed average sound pressure of 100 dB(A) with
peaks up to 120 dB(A), and both patients and staff needed earmuffs to prevent permanent hearing
damages. The noise source was extreme corrugation in the thick ice layer covering the highway.
The upper left photograph in Figure 105 was taken several days after the corrugations had been
smoothed with a heavy grader.

Figure 105 Corrugated ice caused noise in ambulances [Photo: S Engblom, Wiman Amb., J Granlund]

In this project, interior noise was measured in a modern Ford Galaxy family car in good condition
while driving at 90 km/h on the corrugated ice layer, but only AFTER it had been graded.
Reference measurements in the Ford Galaxy car were made at 90 km/h on small roads with a non-
corrugated ice layer. These resulted in noise pressures of about 65 dB(A). The measurements on
the graded Highway 90 revealed noise pressures with an average over 75 dB(A) and peaks above
100 dB(A). The EU work environment regulation calls for earmuffs at a noise level of 75 dB(A).
Clearly not only ambulance car occupants needed earmuffs before the ice layer was graded, but
also normal passenger car occupants.

Since the graded ice layer was still not smooth enough to be considered as acceptable (with
respect to the 75 dB noise), the graded corrugations were inspected. The sections generating
highest peak noise had 3-4 dm wavelengths and amplitudes up to 1.5 cm (after grading), as seen
in Figure 106.

16
Internet 2012-06-30: http://allehanda.se/start/solleftea/1.4224782-ambulans-undviker-riksvag-90
 102

Figure 106 Remaining corrugations; note grooves created by the grader [Photo: J Granlund]

According to the Swedish Transport Administration´s specification for winter maintenance

standard, roughness in thick ice layers shall be measured as deviations from a 60 cm long straight
edge. Roughness, “ojämnhet” in Swedish, is allowed to have amplitude of up to 1.5 cm, see Figure
107.

Figure 107 STA spec for roughness of ice-covered roads

The STA specification has been in use for over a decade and experience has proved it to be
relevant for ice-ruts and for singular roughness. However graded corrugations with less than the
allowed 1.5 cm amplitude also resulted in over +10 dB noise, which is not acceptable. This shows
a need for an additional roughness specification to include corrugations that cause resonance in
normal car wheel suspension systems. Additional research is needed to establish a reasonable
specification on corrugations. This is relevant for both ice-covered roads and for dirt-roads as well.
 103

5. TPC-SYSTEMS REDUCE SHAKE VIBRATION

5.1. TPCS REDUCE RIDE VIBRATION

One of the tasks for ROADEX IV was to evaluate the effectiveness of a Tyre Pressure Control
System (TPCS) on vibration isolation from road roughness to the truck driver’s seat.

Truck tyre standard inflation pressures are required to be set high enough to comply with
maximum load, maximum speed and the worst road condition. Thereby the tyres are normally
over-inflated during many operations, such as when driving with low or no payload, as well as at
low speed. Hence a normal timber logging truck may be operating with suboptimal settings, and
thus experiencing excessive ride vibration, during up to 65 % of the driving cycle [Munro &
McCulloch].

The upper left photographs in Figure 108 show that the tyre footprint can be increased by use of a
TPCS. With a larger footprint, the tyre has better enveloping effect on the road megatexture
(waves with 50 – 500 mm length). A significantly lowered tyre pressure also reduces the stiffness
of the tyre carcass, especially its walls. This affects the eigenfrequency and further contributes to
improved ride quality at low speed on mega-textured surfaces on potholed or stony dirt-roads and
on corrugated ice layers.

Figure 108 TPCS isolates the driver from mega-textured roads [Photo: J Granlund & P Granlund]

The first ROADEX IV experience from studying TPCS in Scotland was very tangible. When
measuring on the Loch Arkaig dirt-road with TPCS in low pressure, the truck ride was quite harsh.
But when measuring with high pressure, the truck cab shook so hard that the measurement
computer fell on to the floor!

Numeric results confirm a great effect from the TPCS on truck driver´s ride quality. During the full
South Laggan round-trip, including both A82 + several km of dirt-road, the A(8) was 0.80 m/s2, with
high tyre pressure for the whole round-trip. The same round-trip but with low pressure except when
driving at highway speed with full payload resulted in A(8) = 0.66 m/s2; a reduction of some 17 %!
However, the driver´s exposure to vibration was still over the EU Action Value of 0.5 m/s2.

The TPCS-study in Sweden was made on a very long round-trip mainly on paved state highways.
With full tyre pressure during the whole route, the A(8) was 0.91 m/s2 with an average speed of 68
km/h. With reduced tyre pressure at a significant share of the route, A(8) was lower at 0.86 m/s2 on
a higher average speed of 73 km/h. Despite +7 % increased speed, driver vibration was reduced
by 7 % which is impressive good. Still though, the vibration was far too high to be acceptable.
 104

5.2. TPCS PREVENTS DRIVING WITH UNDER-INFLATED TYRES

Maintaining correct tyre pressure is crucial to achieve optimum tyre performance with minimal
energy consumption and impact on the environment. With correct tyre inflation pressure, the driver
can experience good comfort, while the tyre and vehicle durability is maximized. Tyre deflection
(the tread and sidewall flexing where the tread comes into contact with the road) remain as
originally designed and excessive sidewall flexing and tread squirm is avoided. Heat build-up is
prevented and rolling resistance kept low. Proper tyre inflation pressure also stabilizes the overall
tyre structure, blending the tyre response, traction and handling.

An over-inflated tyre is stiff, while the size of its footprint in contact with the road is reduced. If a
vehicle's tyres are over-inflated, they could be damaged more easily when running over potholes
or foreign objects on the road. Higher inflated tyres cannot isolate the road´s megatexture well,
causing a harsh ride.

An under-inflated tyre loses its shape and becomes too flat while in contact with the road. Lower
inflation pressure will allow the tyre to deflect more. This can however build up internal heat,
increasing the rolling resistance and causing a reduction in fuel economy. The vehicle can also
experience a significant loss of steering precision and cornering stability.

Tyre pressure must be checked with a quality air gauge, as the pressure cannot be accurately
estimated through visual inspection. In both of the ROADEX demonstration projects on test trucks
with traditional tyre solutions (Finland and Norway), the lack of realistic access to the tyre air
valves on pair-mounted wheels on both the drive axle and on the bogie axle was noted. These and
similar trucks suffer high risk for poor tyre pressure management and hence high risk to be
operated unintentionally under-inflated.

With a TPCS mounted, such as on the test trucks in Scotland and Sweden, the risk for operating
under-inflated is minimized.

However, the TPCS test truck in Sweden suffered from a serious problem. In the mid 1990´s, a
new technology for heavy vehicle tyre balancing was introduced. With this technology, an
absorbent balancing powder is placed directly inside truck and bus tyres. Due to the powder´s
ability to re-distribute itself when the vehicle is in motion, it provides maximum tyre and wheel
balance at all times, during the entire lifetime of the tyre. This technology is routinely used on
standard truck tyres in the Västernorrland region and unfortunately the tyre workshop erroneously
applied it also to the tyres on the TPCS-equipped truck. The result was that the TPCS air valves
clogged with the powder, thereby obstructing the TPCS to deflate the tyres at normal pace, until
the tyres and the TPCS got cleansed. Obviously there is a need to implement means (warning
signs, introduce a check for TPCS at the tyre workshops) to prevent this kind of problem in the
future.
 105

6. CONCLUSIONS
The ROADEX IV demonstration projects in Finland, Scotland, Norway and Sweden had the overall
objective to reproduce the ROADEX III case study from the Beaver Road 331 in Sweden (2008):
1. Measuring truck drivers daily vibration exposure and comparing the A(8)-values to the
Action Value 0.5 m/s2 in 2002/44/EC.
2. Measuring spine compression stress, Sed, caused by jolts at severe bumps and comparing
the values to the 0.5 MPa stress limit in ISO 2631-5 (used as Action Value in Sweden).
3. Relating truck roll & lateral buffeting in heavy trucks with high Centre-of-Gravity to laser-
scanned non-uniform deformation at the pavement edge (the latter quantified by the
pavement condition parameter “Rut Bottom Cross Slope Variance”).

A high repeatability was confirmed between similar truck round trips, when using the ROADEX
method to assess truck drivers’ daily vibration exposure A(8).

Results from the first task, concerning the A(8), are summarized in Figure 109. The only measured
A(8) below 0.5 m/s2 was for European Highway E6 in Norway. However, that measurement was
taken under best possible conditions. After adjustment to average conditions, the A(8) for E6 in
Northern Norway is also expected to exceed the EU Action Value 0.5 m/s2.

Figure 109 Daily vibration exposure A(8) exceeds EU Action Value 0.5 m/s2

Results from the second task are summarized in Figure 110. The drivers of timber logging trucks
both in Scotland and in Sweden were exposed to unhealthy spinal compression stress.

Figure 110 Timber truck drivers had high spinal compression stress Sed
 106

Measurement results confirmed that roll vibration and roll-related lateral buffeting in heavy trucks is
an issue in Scotland, Finland, and Norway, as well as in Sweden. The traditional road condition
parameters of “Rut Depth” and “IRI” were showed to have poor correlation with truck cab lateral
buffeting. The pavement condition parameter “Rut Bottom Cross Slope Variance” (RBCSV) defined
in 2006 was again shown to be a good indicator of “road warpiness” and to truck lateral buffeting.

This project did not reach the goal to implement the use of the Rut Bottom Cross Slope Variance
(RBCSV) parameter in the Partner areas. A reason for this was that Profilometer data from
network surveys in Norway and Scotland was not available in the step length of 1 m required for
calculating the RBCSV parameter. The RBCSV parameter has been tested with good results in
both Finland (by the consulting company Destia OY that carries out the network condition survey,
see Hurtig 2010) and by Vectura in several projects in Sweden. The research institute VTI is
currently making an in-depth evaluation of the RBCSV parameter as indicator of road warpiness.
Preliminary results from VTI are that the RBCSV show high correlation with both pavement edge
deformation and with vehicle roll vibration.

This ROADEX IV project showed that there is a need to define a new measure and establish
associated limit values for corrugation in both ice layers and in dirt-roads. Current specifications
deal only with rutting and with singular roughness. The Swedish specification could not be used to
penalize corrugations that made National Highway 90 unusable for normal transports (incl
ambulance) due to extreme interior noise.

An overall objective for road management should be to keep pavements in such good condition
that it is possible to operate normal trucks at normal working days behind the steering wheel, and
still keep A(8) under the EU Action Value 0.5 m/s2.

Many road agencies have limits for road roughness on the IRI-scale. In most (all?) cases, these
limits have been set with respect to passenger car drivers comfort and not with respect to the
working environment of professional truck drivers. This has led to a serious overestimation of
acceptable road roughness. By changing the focus from car comfort to a healthy and safe heavy
truck operation, the acceptable road roughness drops from being about 5 mm/m17 to some 1.5 - 2
mm/m. This report discusses several means of lowering truck seat vibration besides shortening the
interval between asphalt overlays, for example improved balancing of truck wheel ends with
“balancing powder”.

Winter conditions in the Northern Periphery include extremely uneven frost heaves and severe
corrugation in the ice layers covering non-salted18 roads. The project results show that these
conditions can make the ride vibration significantly worse than during summer conditions. Hence
assessment of Northern Periphery truck drivers’ exposure to daily vibration at work must be
adjusted for difference in summer and winter.

Road network condition surveys with laser profilers are currently made only in summer time. The
above finding shows that such condition assessments can seriously underestimate the winter
season problems with road roughness. The road condition assessment procedures of the
ROADEX Partners Norway, Sweden and Finland (maybe also Iceland?) therefore need to be
reviewed.

A secondary research objective concerned mapping the benefits of tyre pressure control systems,
regarding vibration transfer from tyre footprint to driving seat. One research goal was to quantify
the transfer as function of vibration frequency (roughness wavelength x speed) from data recorded

17
An example is STA Pavement Management standard 2011, available in Swedish on Internet (2012-06-30):
http://publikationswebbutik.vv.se/upload/6636/2012_074_Underhallsstandard_belagd_vag_2011.pdf
18
Road salt is used as de-icing agent in order to increase road friction and traction. At temperatures below
about -9 á -13 C however, the salt resin re-freezes and can create an extremely slick road surface and
hence increase instead of decrease the skid risk.
 107

in Sweden. This goal could not be met for several reasons. In the summer test, the test truck
TPCS air valves became clogged, so the system was not useable. In the winter test, the portable
laser profiler was damaged at an extreme bump at a road section that had been reconstructed 6
months previously.

The report from the ROADEX III case study 2007/2008 on road 331 in Sweden included
identification of about a dozen very hazardous sites. These have not been addressed at the date of
this report, neither have warning signs been raised at these hazardous sites.

Similarly a number of actions need to be taken to address defective practices highlighted in the
study. Examples include, bumpy “steps” at bridge joints and at transversal joints at asphalt patch
repair, extreme roughness in the first winter following reconstruction of frost damaged road
sections, new resurfacings exceeding the EU Action Value for an 8 hour daily truck ride, etc.
Current practices in some geographical areas are not yet acceptable with respect to the health and
safety of professional truck drivers.

There is a need to foster a renewed spirit in road professionals and a focus on sustainable road
management. The historic background of underfunding of rural roads over the years, and the
resultant backlogs of work, may be dispiriting. However, focused engineers can still make large
difference in assuring that Northern road networks continue to remain valid for the challenges
ahead.
 108

7. RECOMMENDATIONS
Current limits for road roughness on the IRI-scale have been set with respect to passenger car
drivers´ comfort and not with respect to the working environment of professional truck drivers. This
focus needs to be changed, from car comfort, to healthy and safe heavy truck operation. This calls
for much smoother roads.

In addition, several other means to lower truck seat vibration also need to be implemented, for
example improving the balancing of truck wheel ends with “balancing powder” or installing a TPCS.
The effectiveness of the proposed actions should be further investigated.

Roll vibration and roll-related lateral buffeting in heavy trucks is an issue in Scotland, Finland, and
Norway as well as in Sweden. The pavement condition parameter “Rut Bottom Cross Slope
Variance” has again been validated as a relevant measure of road warpiness and should be
considered for implementation in all Partner areas.

Truck cab transient lateral acceleration may exceed the 5 m/s2 (0.5 g) lateral acceleration to be
considered when latching loads on road vehicles, as per the IMO/ILO/UN ECE Guidelines for
cargo securing, when passing over pavement edge damages at 70 - 80 km/h. The latching
guidelines may need a review because of this.

Pavement edge damages worse than RBCSV 0.30 % should be repaired ASAP, with a warning
sign raised until the road repair is completed. Truck operators should also educate their drivers to
drive slowly at this kind of pavement damage.

Investigations of truck crashes do not always include competent analysis of the role of road
properties. This is in fact true also in many investigations of fatal crashes. Given this context, the
new RBCSV road condition parameter could be a useful tool to identify of road damages that may
have contributed to a rollover crash.

The assessment of Northern Periphery truck drivers’ exposure to daily vibration at work, and road
network condition surveys, must be adjusted for difference in summer and winter.

There is an urgent need to establish a new measure and associated limit values for corrugation in
paved roads with ice layers and in dirt-roads.

Several of the root causes to high vibration are man-made. Actions need to be taken to assure that
road repairs are effective and not worse than before. It is recommended the road condition
(alignment, slopes and surface condition) should always be measured by a certified inspector after
the completion of any major road-repair works using an approved ASTM E950 “Class 1” road
profiler.

Transversal edges of new asphalt overlays should not be placed “layer-on-layer” with the old
asphalt, but cut down into the same elevation as the surface of the old asphalt. This will require the
use of a small milling machine. Such machines are available on the market, but not always
implemented in highway pavement maintenance.

Long wave unevenness is difficult to repair with a traditional asphalt overlay, since the paver tends
to only go “roller-coasting” in the up to more than 30 m long waves. A sustainable solution contains
careful filling in the hollows and preferably also gentle milling on the ridges, before resurfacing.
Preferably this procedure should be done after laser-scanning the undulations, computer aided
design of the undulation repair, and computer aided manufacturing of the milling works. The final
resurfacing should preferably not be done with computer aid, with respect to the complexity of
controlling the paving plants floating screed.
 109

Improperly banked horizontal curves are a common traffic safety problem in the Northern
Periphery. A method for easy detection of hazardous flat outercurves as well as banked
innercurves has been demonstrated in this report. It is recommended that this is implemented by
all ROADEX Partners with road management responsibility, as well as by investigators of rollover
or other instability road vehicle crashes.
 110

8. ACCURACY AND UNCERTAINTY

The repeatability of A(8)-assessment using Vecturas portable vibration measurement system was
demonstrated in three runs between Corpach and South Laggan in Scotland, with very good
likeness (almost to the second decimal) between the runs, see Figure 40. The reproducibility was
also surprisingly good when comparing results between the measured data from the Beaver Road
331 between Ramsele and Ärtrik at the Åkerö edge damage recorded in ROADEX III (2007)
respectively in ROADEX IV (2011), see Figure 83.

Some measurement data files were found to be damaged due to severe mechanical shocks as
described in the report. Those data were discarded and not used when preparing this report.
 111

9. REFERENCES
 Directive 2002/44/EC on The minimum health and safety requirements regarding the exposure
of workers to the risks arising from physical agents (vibration). The European Parliament and
the Council.
Internet 2010-01-30: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2002:177:0013:0019:EN:PDF
 Granlund, J. (2008). Health Issues Raised by Poorly Maintained Road Networks. The ROADEX
III project.
Internet 2010-01-30: http://www.roadex.org/Publications/docs-RIII-EN/Health%20Issues%20-%20RIII.pdf
Videolecture 2010-01-31: http://videolectures.net/surf08_grandlund_rhirt/
 Westerling, R. (2008). Åtgärdbar dödlighet som en indikator i den folkhälsopolitiska
uppföljningen. (Avoidable mortality as an indicator in the evaluation of public health politics).
Statens folkhälsoinstitut, Rapport 2008:03.
Internet 2010-01-30: http://www.fhi.se/PageFiles/3122/Atgardbar_dod_webb_0802.pdf
 Teschke, K., Nicol, A-M., Davies, H. & Ju, S. (1999). Whole Body Vibration and Back Disorders
Among Motor Vehicle Drivers and Heavy Equipment Operators: A Review of the Scientific
Evidence. Report to Workers' Compensation Board of British Columbia.
Internet 2010-01-30: http://www.cher.ubc.ca/PDFs/WBV_Report.pdf
 Systematiska kunskapsöversikter: 2. Exponering för helkroppsvibrationer och uppkomst av
ländryggssjuklighet. Arbete & Hälsa Report 2012:46
 Arbete och helkroppsvibrationer – hälsorisker. Arbetsmiljöverket. Rapport 2011:8
 Arnberg, P. & Åström, G. (1979). Vägojämnheters inverkan på bilförares prestation och trötthet.
Statens Väg- och Trafikinstitut, Rapport 181.
 Gillespie, T.D., Segel, L., Ervin, R.D., Schneider, L.W. & Campbell, K.L. (1982). Truck Cab
Vibrations and Highway Safety. Highway Safety Research Institute, University of Michigan.
FHWA report RD-82/093.
 Hedberg, G.E., et al. (1993). Risk indicators of ischemic heart disease among male
professional drivers in Sweden. Scand. J Work Environ Health, Vol 19, p 326-333.
 Bigert, C. Klerdal, K., Hammar, N., Hallqvist, J. & Gustavsson, P. (2004). Time trends in the
incidence of myocardial infarction among professional drivers in Stockholm 1977-96. Occup
Environ Med; Vol 61, p 987 - 991. Internet 2010-01-30: http://oem.bmj.com/content/61/12/987.abstract
 McFarlane, S. & Sweatman, P.F. (2003). Investigation into the Specification of Heavy Trucks
and Consequent Effect on Truck Dynamics and Drivers: Final Report. Roaduser International
Pty Ltd.
 Saarenketo, T. & Saari, J. (2004). User perspective to ROADEX II test area’s road network
service level. The ROADEX II project.
Internet 2010-01-30: http://www.roadex.org/Publications/docs-RII-EN/1_1%20User%20Perspective_l.pdf
 Bray, A., Bronstein, A.M., Green, D.A., Gresty, M. A. & Golding, J. F. (2006). Tachypnea and
hypocapnia are induced by ‘buffeting’ in vehicles. Clin Auton Res, 16:281–285.
 EN 14253+A1 (2007). Mechanical vibration – Measurement and calculation of occupational
exposure to whole-body vibration with reference to health – Practical guidance. European
Standard.
 European Best Practice Guidelines on Cargo Securing for Road Transport. European
Commission Directorate – General for Energy and Transport.
Internet 2012-04-15: http://ec.europa.eu/transport/road_safety/vehicles/doc/cargo_securing_guidelines_en.pdf
 Högfors, C. (2001). Snabba accelerationers inverkan på gods under transport.
Internet 2012-04-18: http://mariterm.se/download/Rapporter/Snabba_accelerationers_inverkan.pdf
 Rutevise utredninger for riksvegnettet, Riksvegrute 7, E6 Trondheim - Fauske med
tilknytninger. (2011). Statens Vegvesen, Norway.
Internet 2012-05-15: http://www.vegvesen.no/_attachment/263397/binary/467332
 Tom, G.K.J. (1995). Accidents on Spiral Transition Curves. ITE Journal.
 Berglund, A. (2009). Tjäle – en litteraturstudie med särskilt focus på tjällossning. Luelå
Tekniska Universitet. Internet 2012-05-15 http://pure.ltu.se/portal/files/3333867/Tjale_Litteraturstudie.pdf
 Ahlin, K., Granlund, J. & Lundström, R. (2000). Whole-Body Vibration When Riding on Rough
Roads – A Shocking Study. Swedish Road Administration, Vol. 31E. Internet 2012-05-30:
http://publikationswebbutik.vv.se/upload/3054/2000_31E_Whole_body_vibrations_when_riding_on_rough_roads.zip
 Ferguson, R., Prem, H., Ramsay, E. & McLean, J. (2000). A Road Profile Based Truck Ride
Index (TRI). Austroads Inc., Australia. Internet 2012-05-30:
 112

http://www.rahiran.ir/data/reports/australia/heavy_vehicle/a%20road%20profile%20based%20truck%20ride%20index/377_AP-
R177.pdf
 Mucka, P. & Granlund, J. (2012). Is the Road Quality Still Better? Journal of Transportation
Engineering, Vol. 138, No. 12. Internet 2012-06-30: http://dx.doi.org/10.1061/(ASCE)TE.1943-5436.0000460
 Ahlin, K., and Granlund, J. (2002). Relating road roughness and vehicle speeds to human
whole body vibration and exposure limits. Int. J. Pavement Eng., 3(4), 207–216.
 Granlund, J. (2010). Reducing Health and Safety Risks on Poorly Maintained Rural Roads.
11´th International Heavy Vehicle Transport Technology Symposium, Melbourne, Australia.
 Hurtig, E. (2010). Raskaan kaluston kallistelun voimakkuuden arvioiminen tienpinnan profiilista.
MSc thesis. Aalto University School of Science and Technology, Finland.
The thesis in English summary on Internet 2012-04-15: http://tierakennusmestari.com/lehdet/Hurtig.pdf
 Munro, R. & McCulloch, F. (2008). Tyre Pressure Control on Timber Haulage Vehicles. The
ROADEX III project.
Internet 2012-05-30: http://www.roadex.org/uploads/publications/docs-RIII-EN/TPCS%20Trial%20-%20RIII.pdf
 Hurtig, E. (2010). Raskaan kaluston kallistelun voimakkuuden arvioiminen tienpinnan profiilista.
MSc thesis. Aalto University School of Science and Technology, Finland.
The thesis in English summary on Internet 2012-04-15: http://tierakennusmestari.com/lehdet/Hurtig.pdf
 113

ROADEX PROJECT REPORTS (1998–2012)

This report is one of a suite of reports and case studies on the management of low volume
roads produced by the ROADEX project over the period 1998-2012. These reports cover a
wide range of topics as below.
 Climate change adaptation
 Cost savings and benefits accruing to ROADEX technologies
 Dealing with bearing capacity problems on low volume roads constructed on peat
 Design and repair of roads suffering from spring thaw weakening
 Drainage guidelines
 Environmental guidelines & checklist
 Forest road policies
 Generation of ‘snow smoke’ behind heavy vehicles
 Health issues raised by poorly maintained road networks
 Managing drainage on low volume roads
 Managing peat related problems on low volume roads
 Managing permanent deformation in low volume roads
 Managing spring thaw weakening on low volume roads
 Monitoring low volume roads
 New survey techniques in drainage evaluation
 Permanent deformation, from theory to practice
 Risk analyses on low volume roads
 Road condition management of low volume roads
 Road friendly vehicles & tyre pressure control
 Road widening guidelines
 Socio-economic impacts of road conditions on low volume roads
 Structural innovations for low volume roads
 Treatment of moisture susceptible materials
 Tyre pressure control on timber haulage vehicles
 Understanding low volume pavement response to heavy traffic loading
 User perspectives on the road service level in ROADEX areas
 Vehicle and human vibration due to road condition
 Winter maintenance practice in the Northern Periphery
All of these reports, and others, are available for download free of charge from the ROADEX
website at www.ROADEX.org.

Copyright © 2012 The ROADEX IV Project, EU Northern Periphery Programme. All rights reserved.