VTT notat
No 32A-1996 Utgivningsår: 1996
Title: Relationship between the functional properties of road surface and traffic safety.
A State-of-the- Art Report.
Author: Peter Wretling
R&D Programme: Maintenance/Öperations - Effects Project No: 80100
Project title: Art. Road Surface - Effects.
Sponsor: Swedish National Road Administration (SNRA) Distribution: Free div Väg- och transport-forskningsinstitutet ä
Foreword
This literature survey was financed by the Swedish National Road Administration (SNRA).
At present, there is no overall survey ofR&D needs in the research field concerning the functional properties of the road surface. This survey is one of three parts of a project to survey national and international research in the field, also suggesting future research efforts. The different parts of the project are:
0 The importance of the road surface for traffic safety and driving speeds. 0 Stress and service life consumption of vehicle components.
0 Effects of road texture on friction, tyre wear, fuel consumption/rolling resistance and noise.
Eva Lundberg has been the contact person at the Swedish National Road
Administration and Gudrun Öberg the project leader at the VTI.
Valuable opinions were provided by Eva Lundberg (SNRA) and Gudrun Öberg (VTI).
Linköping, April 1996 Peter Wretling
Innehållsförteckning
1 Introduction
2 Relationships between road surface conditions and
traffic accidents andl or speed
3 " Relationships between skidding resistance and
traffic accidents
4 Future research work
5 References
VTI notat 32A-1996
13
25
27
1 Introduction
Low standard of the pavement can cause accidents and discomfort for drivers and also influence travel speed. Knowledge of the correlation between the state of the pavement and traffic safety is therefore of importance. Road users costs, described in terms of accidents, journey time and vehicle Operating costs, are much greater than the cost of investment in maintenance operation for the Road Administrators (l). See Figure 1. It is possible that improved pavement standard can reduce the total socio-economic costs of road transport.
Milliard SEK! year
,l
r* r-
: l' P '
I . | I 20 < ' l l ll
I
I I 10« a ll
I'
I'
0
I
i a
Investment Journey Vehicle ACCidcms Environment
maintenace time Operating
operation
Figure I Socio-economic costs per yearon the national road network, (1980
price level).
Pavement Management (PM) Systems may assist road administrators either to minimize the socio-economic costs for road transportation or to formulate guidelines for an acceptable road standard at the lowest cost. In a literature review of PM systems by Wallman et al. (2) two systems from Sweden, SABU and PUB, are described. According to SABU, the rut depth will not affect travel speed if it is less than 40 mm. Furthermore, resurfacing is assumed to increases the travel speed by 1-2 km/h and a comfort value greater than 5.5 will increase the number of accidents by 2 %. PUB recommendeds maximumvalues for acceptable rut depth and unevenness. The rut depth is defined as the average value measured over a distance of 400 metres. The evenness of the road is based on a measure of the longitudinal evenness. Both systems use data from the RST (Road Surface Tester). In the early eighties, Arnberg et al. (3) surveyed the need for research on the maintenance of road and street pavements. They considered the following six
research tasks to be most important:
l. Methods/ models for predicting long-term changes in pavement condition when implementing various surfacing strategies.
2. Efficient methods for determining the structural condition (bearing capacity) ofroads.
3. Measurement of current traffic loads and prognosis of future traffic loads (especially axle loads).
4. The influence of surface condition on vehicle wear. 5. The influence of surface condition on road accidents. 6. Evaluation of effects.
This State-of-the-Art Report deals with the relationships between the properties of the road surface and traffic safety. Chapter 2 deals with the relationships between road surface conditions and traffic accidents and/or speed, Chapter 3 with the relationships between Skid resistance and traffic accidents and finally Chapter 4 deals briefly with riding comfort.
References (4,5,6,7) are other relatively new State-of-the-Art reviews or
documentation discussing the relationship between road surface properties and traffic safety, and are mentioned only in the reference list.
2
Relationships between road surface conditions
and traffic accidents andl or speed
Two studies in Sweden concerning the relationships between road wearing course and traffic accidents by Schandersson (8) and by Björketun (9) showed a tendency for the accident rate to be lower on roads with surface treatment compared to roads with asphalt concrete.
Schandersson reported that this was the case if the roads have low traffic volumes.
Schandersson also concluded that the proportion of accidents on wet roads or on icy/snowy roads no significant differences were found for different wearing courses and that both the accident rate and the injury rate were shown to be approximately the same for different conditions of the wearing courses.
Furthermore, no strong relationships were found between indirect measures of the
condition (age of the wearing course, number of pairs of axles since the most
recent maintenance) and the accident rate.
Björketun included a factor describing the weather conditions. This was classiñed a-s fair or poor depending upon the amount of precipitation and mean temperature.
For roads with surface treatment, the analysis showed small differences in accident rate for fair and poor weather respectively, while the accident rate for roads with asphalt concrete was always higher in poor weather conditions. Furthermore, asphalt concrete has a significantly lower accident rate in fair weather conditions than has surface treatment. In poor weather conditions, no significant differences were found.
As stated before these two studies showed a tendency for the accident rate to be lower on roads with surface treatment compared to roads with asphalt concrete. However, the data were not very extensive in these two studies. Therefore, the relationship was investigated again in a new study by Björketun (10), this time including the total number of accidents on the paved national road network during
1977-1980.
This study showed, that the surface treatment roads, in the southern and central parts of Sweden, had an accident rate 10 0/0 lower than the accident rate for asphalt
concrete roads, when wild life accidents were excluded. This percentage was somewhat higher in the winter and somewhat lower in the summer. In northern Sweden no differences were found.
If personal injury accidents alone were included, no differences were found. This indicated that there might be differences in accident reporting frequency between the two pavement types.
In a study, divided into two phases. the relationship between road surface and travel speed was investigated in Sweden. Phase 1 by Linderoth (ll) concerned surface treatment of roads in bad surface condition.
The selected road sections were divided into two groups, an experimental group consisting of road sections selected for resurfacing treatment and a control group composed of road sections not to be resurfaced.
To eliminate or to keep constant the influence of interrelated factors besides surface Characteristics, pairs of nearly equivalent road sections were formed. Each pair was made up of one experimental and one control group member and had the same speed limit and approximately the same road width and traffic flow.
The results showed that there was a significant increase in median speed for passenger cars on resurfaced sections, in the time interval 15-09 (3 pm-9 am), of 1.7 km/h. Corresponding control group speeds decreased by 0.5 km/h. The difference between groups was significant at the 1 % level. In the time interval 09-15 there were no significant changes.
For the entire measuring day, 24 hours, there was a significant increase in the median speed for passenger cars of 1.4 km/h. The control group s median speed was almost unchanged.
The: conclusion was that there is no evidence of reduced speed due to unevenness. This can be explained by the comparatively good surface conditions initially found in the experimental group. However, in the experimental group during the time interval 15-09 a weak interrelation between speed change and change in unevenness was observed.
The next phase described by Kolsrud and Nilsson (12) compared surface treatment to plant mix.
Analysis ofspeed data from 17 locations before and after resurfacing and simultaneously from 16 control locations, on unchanged sections, did not show any significant differences between the two techniques or between various other surface Characteristics.
However, compared to the controls, a significant rise of 1.4 km/h was observed after resurfacing. These results are in accordance with the results of phase 1.
In a joint Nordic research project in Denmark, Finland, Norway and Sweden by Hemdorff, Leden, Sakshaug, Salusjärvi and Schandersson (13) traffic safety was investigated on roads with different pavement surfaces and surface wear. The study consisted of four substudies: the effect of pavement age on traffic safety, the safety effects of the condition of the pavement, the effects of road surface friction on accident rates (see Chapter 3), and the effects of repaving on driving speeds on dry road surfaces.
The accident rates increased with ageing pavements on days with more than 10 mm precipitation. On average, accident rates increased with ageing pavements. In
some regions, however, the rates decreased.
On inferior road pavements, the accident rates were about 7 % lower than on good pavements, caused by the lower rates during days without precipitation. On the other hand the accident rates on very rainy days, i.e. with more than 10 mm
precipitation, were higher on inferior road surfaces than on good road surfaces, see
Figure 2.
Precipitation
per day
All classes 10.6 (1349 acc.)
'-7 :-:-:-: :1:-:-:-:-:-:-. '-:-:-:-:-:-:A:'z-:-:-:«:-:-:-:v:-..-:tz-:c 9.3 (564 acc.)
Slightty worn ('good') < 0'1 mm
pavement surfaces 0.1 -10 mm ---' 11.9 (705 acc.)
> mm : : :-:-' 109 (60 acc.)
All ctasses 9.9 (1537 acc.)
' 3 8.6 (659 acc.)
Considerably worn ('poor') < Ol mm
pavement sudaces 0.1 -10 mm *'^'-'-*'-" *'**' ' v - ' 10.9 (785 acc.)
mm 144 (93 acc;
10.0 Accxdent rate (inj. acc./ 100 M vhkm)
Figure 2 Accident rates (injury accidents per 100 million vehicle kilometres) for good " ana' "poar pavement conditians. On average andfor
different precipitation classes.
By grouping the data into summer and winter, the results showed that the lower accident risk for the inferior road pavements was due to differences during summer. See Figures 3 and 4.
Precipitation
per day
All classes 9.8 (717 acc.)
9.9 8 .
Slightly worn ('good') < 0-1 mm (3 2 m)
pavement surfaces 01-10 mm 10.3 (311 acc.)
10 mm 6.1 (24 acc.)
All ctasses 8.6 (766 acc.)
.4 40 . Considerably worn ('poor') < 0'1 mm 8 ( 2 acc)
pavement surfaces 0.1-10 mm 8.4 (305 acc.)
> 10 mm 12.4 (59 acc.)
\
i I
100 Acadent rate
(inj.acc./1OO M vhkm)
Figure 3 Accident rates in summer för "goad " ana' "poor pavement sur/ace conditians. On average and .för diü'erent precipitation
classes.
Preczpltation
perday
All classes 11.6 (632 acc.)
sughuy wom (.good. < 0.1 mm 8.4 (202 acc.)
pavement surfaces 0.1- 10 mm 13.7 (394 aCC-l
10 mm 23.4 (36 acc.)
All classes 11.7 (771 acc.)
considerably wom (.poor.) < 0.1 mm 8.9 (257 acc.)
pavement surfaces 0.1-10 mm 135 (480 ace-l
mm 19.9 (34 acc.)
\
_ I
10 o Accadent rate (inj. acc./ 100 M vhkm)
Figure 4 Accident rates in winterfor good and "poor pavement surface
conditions. On average and_for different precipitation classes. Figures 2, 3 and 4 are taken from VTI Reprint 231 Road Pavement Condition and Traffic Safety. Some Results and Conclusions from the Nordic Project TOVE by Schandersson.
Regarding the effects of repaving on driving speeds on dry road surfaces, no effect was found in Finland or Norway. In Sweden, the mean speed increased by slightly more than 1 km/h as a result of repaving. No results were obtained from Denmark.
One conclusion from this study was that repaving is not necessarin a safety measure. The effect depends on the climatic conditions and especially on the frequency of rain.
Anund (14) used speed data and let the IRI (International Roughness Index)
and rut depth describe the quality of the pavement when fitting a multiple linear regression to the speed data in order to investigate the relationship between road surface and speed.
The speed data were divided into three different types of vehicles: passenger cars, trucks and trucks with trailers. The measuring sites were located on straight and level sections of main roads. The relationship was estimated by multiple linear regression in the following form.
Y=G+B|X| +B3X3+ . . . ..+BÖX(3+8
where
Y = mean speed Xl = rut depth, mm X2 = IRI, mm/m
X3 = I if the speed limit=90 km/h, otherwise O X4 = 1 if the speed limit=l 10 km/h, otherwise O X5 = 1 if the road width is less than 8 m, otherwise O X6 i l if the l lO-roads are a motorway, otherwise O 8 = error
The model showed that the speed reduction for passenger cars was 1.6 km/h if the rut depth increased by 10 mm and 2.2 km/h if the IRI increased by 1 mm/m. The corresponding figures in the time interval 09-15 were 1.9 km/h and 3.0 km/h respectively. The results were significant.
For trucks with and without trailers, no significant speed reduction with increased road surface or rut depth was found.
In the early eighties, an experimental investigation to determine whether surface deterioration significantly influences traffic speeds was carried out in Great Britain by Cooper, Jordan and Young (15). Speeds before and after the resurfacing of three test sites were measured and analysed.
Previous investigation and research at TRRL had shown that, in addition to road surface Characteristics, the following factors may influence traffic speeds on major roads: road type, road layout, traffic composition and density, time of day, weather conditions, fuel costs and speed restrictions. The experiment was designed to eliminate the effects of these factors as far as possible.
At two of the test sites, the condition of the road surfaces had deteriorated, both
profile and texture, but the road structures were considered to be sound. The third test site, structural deterioration was suspected. A decision was taken to reconstruct the road down to the upper level iof the cement-bound granular base.
The investigation showed that:
(i) traffic speeds are not affected by the resurfacing of a road whose profile has deteriorated to a level where the (52 of the unevenness is less than 3 mm2 over profile features less than 5 metres long;
.. .. . 7 . 7
(11) when a profile has deteriorated to a 0' of at least 8 mm about a 5 metre moving-average datum, an increase in speed of 2.6 km/h may be achieved from the resurfacing of such roads;
(iii) a change in surface texture has no significant effect on traffic speed providing all other factors are constant.
Craus et al. ('16) classified the Israeli roads network into five pavement state
groups, from 0 (the worst state) to lOO (excellent state). The average accident rate per vehicle-km was computed for each group and for three types of roads: single carriageway, dual carriageway and both single and dual carriageway. In order to test whether or not the accident rates of the five groups were from the same population, the Kruskal-Wallis H-test was utilized.
The conclusion, only applicable in the general case of the whole network, was that the pavement condition did not influence the accident rates for any of three types of roads.
Summary
Two studies in Sweden concerning the relationships between road wearing course and traffic accidents showed a tendency for the accident rate to be lower on roads with surface treatment compared to roads with asphalt concrete.
In another study in Sweden the relationship between road surface and travel speed was investigated. The results showed that there was a significant increase in median speed for passenger cars on resurfaced sections in the time interval 15-09 (3 pm - 9 am), of 1.7 km/h. For the entire measuring day there was a significant increase in the median speed of 1.4 km/h.
Some results from the TOVE-project:
- The accident rates increased with ageing pavements on days with more than 10 mm precipitation.
- On inferior road pavements the accident rates were about 7 % lower than on good pavements.
- In Sweden, the mean speed increased by slightly more than 1 km/h as a result of repaving.
One conclusion was that repaving is not neccessarily a safety measure.
A study from Sweden used speed data and let the IRI and rut depth describe the quality of the pavement when fitting a multiple linear regression to the speed data in order to investigate the relationship between road surface and speed. The model showed that the speed reduction for passenger cars was 1.6 km/h if the rut depth increased by 10 mm and 2.2 km/h if the IRI increased by 1 mm/m.
3
Relationships between skidding resistance and
traffic accidents
The effects of road surface friction on accident risks were studied within the TOVE-project (13). Only friction data from Denmark were obtained. The tendency is that the accident rate decreases with increasing road friction coefñcient, see Figure 5, and that this can be seen both for injury accidents and accident with property damage only. The results also showed that this tendency is most pronounced for two-lane highways and is greater on roads with surface dressing than on flexible or rigid pavement types.
Personskadauheld pr 10 mio vognkm
8.0 r 6.0 -4,0 3-7 2.0 1.7 1.6 y//// oo 0'0 ' 0.40- 0.50- o.so- 0.70- 0.80- 0.90-0.49 0 59 0.69 0.79 0.89 1.00 FRIKTION
Figure 5 Relationship between injury accident rate andfrictionfor all road types.
Roe, Webster and West (17) outline the basic principles of texture depth and skidding resistance and their measurement in a report from Great Britain. Furthermore, an inter-relationship between texture, skidding resistance and accidents was developed.
It has not been possible to assess the extent to which poor texture depth contributes to accidents because texture depth could not be measured on a large scale on inservice roads. The development of the laser-based, highspeed texture meter (HSTM) has made such measurements possible. Values from the high-speed texture meter were used to calculate the average texture depth for each 10 m length of road, known as sensor-measured texture depth (SMTD).
The skidding resistance for the accident site is represented by the mean summer
SCRIM coefficient (MSSC) which is the mean of three or more SCRIM
coefficients obtained for the same length of road spaced out over the summer testing period of a single year. (The sideway-force technique for measuring skidding resistance has been develOped into a routine test vehicle known as SCRIM.)
The accidents were classified in four groups: dry, no skidding; dry, skidding; wet, no skidding; wet, skidding. Road networks in three counties were designed specially for the study, where network A was included as a wide sample of road
types, network B was included because SCRIM measurements were already being made and finally network C was included as an example of asphalt concrete roads and roads carrying large numbers of commercial vehicles.
To test the null hypothesis that texture at the accident sites was distributed in the same way as texture on the network. a chi-square test was used. The test showed that the distributions were significantly different. This confirms that the accidents occur on the lower textures at a greater frequency than would be expected on the basis of the existence of such textures on the road. See Figure 6.
A similar test was carried out for the skidding resistance values in order to assess the influence of microtexture. This test showed no significant difference between the distributions for networks B and C and just significant at the 5 % level for network A. See Figure 7.
96 of ro ad ne two vk in S M T D ci as s 1,201.20& 1.00 1.10 0.90 SMTD (mm) 97 // /l // M/ // l/ 7/ // A llll llll |lll llll llll lllI llll
0.80 0.70 0.60 0.50 ? K W " m m llll llll llll i'll lIll llll llli llll ul W W illl |lll llll llll llll mlli li Illl ll 1 below 0.40 0.40
o
,
m
.9.
52
0
:s ep O l wg ua : m m ;0 sxua pça oe ;o 94, % of ro ad n e t wo r k in S M T D cl as s aboveal Network A. averaged over 7982 - 84, a// roads
120 1.20 81 .10 1 .00 1 .90 80 SMTD (mm)
b/ Network 8, 7.984 all road:
I F' . I, 0 I I I / ,9 I r_ . ' O L ue çr ñåxe g . . o f * -.\ » W J W M W M M uum wm m n I_ lum nm nuun m 8 0' 0.50 O *7. L L s k a xxx O S 9. 1 O O .0 m ;s ep a l ws ut ; m m ;0 s wa p o o a e ;o 94, 15 % of ro ad n e t w0 r k in S M T D cl as s 0 1.20 8: above above - 20 1 .20 network .10 1 å Dry, no skidding 4 30 Wet, skidding ü Dry, skidding D Wet, no skidding --- Proportion of .00 1 0.90 0.80 SMTD (mm)
cl Network C, 7984 a// road:
0.70 0.60 0.50
Proportiorz ofaccidents ofdi/fferent types at different texture levels compared with distribution oftexture levels on road network.
below 0.40 0.40 0.40 Sta l:) Q l wg un : m m ;0 sl ua pn oa e ;o 94,
VTI notat 32A-1996
Figure 7 16
50 i- n
ae
U 0 8 40 4 40 1? :n o 5 Fr_ 8_ E i 3 " \ 1P\ -4 2 8_ 30 » i' . 30 Z Z. V: 1,533 9 3 or' '1 f. 2 20 äs ' _ 20 :J G V: .2 g§3
' ' 3 ,,i
g
3
'L
O
' 10 if, -< 10 9_5
i 2:
g
ae ,2: - '\ l 0 " ' j 0 below 0.30 0.35 0.45 0.50 0.55 0.60 0.65 0.70 0.70 8: 0.30 aboveMean Summer SCRIM coefficient
a) Network A, averaged over 7.982 - 84, all road:
50 - .. 50
.å
o .-L
*
o 40 - « 40 9" g 3 2 _ ä : ---\ 3 I; "' \\ '* åfå
2
3. 20 - . 20 3 g \\ J ä '0 \ m a _ \ o å 10 - x '* 10 Q. .. \ m 0 \\\ 1 a ae - X 0 H 14--, I _ T ' r m n r ----I-_ T O below 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.70 & 0.30 aboveMean Summer SCRlM coefficient
bl Network B, 7984 all road:
a 50 -
;1
. .
50
g Dry, no sklddmg 32 0 . 9.. 8 40 g Dry, skidding .. 40 3 å ää
30 D Wet, no skidding d 30 'ä'3
2
' r....N
,
;-5 ' Wet,skidding 5' 2 \\\ '* m 5 _ x ---- -- Proportion of 5, ?3 < \ network (0; U ?är , \\ m 10 ;3 F1 - 10 2 *5 ;1; It \ å se ?m 0 0 E i m--ä----l 0 I I T I T below 0.30 0.35 0.50 0.55 0.60 0.65 0.70 0.70 81 0.30 aboveEstimated mean Summer SCRlM coefficient
cl Network C, 7986 a/l road: (single-survey only)
Proportion of accidents ofdzjferent types at different MSSC levels compared with distribution ofMSSC levels on road network.
One approach to estimate suitable levels of the texture is to determine the texture level at which the proportion of accidents exceeds the proportion of textures on the network. This level falls between 0.6 mm and 0.8 mm, indicating that the risk of accidents is greater for roads with an average SMTD below about 0.7 mm than for those above this level. see Figure 8.
g
20
G)3 *så
u 3.
,
;7:- 10 P_ \\ Network B 4 5 5 -a-J Network A 0 2 CD __ 4 5 a; 50 c ;2*3:5 2 0
å se i' - -4 5 E Network C : l 1 L L L 1 l l l 0 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 SMTD (mm)Figure 8 - Diference between accident sites and network texture
distributions.
An alternative approach is to consider accident rates (number of accidents per year per kilometre of road) at different texture levels. It is evident in Figure 9 that the three networks have different accident rates at the lowest textures. This follows the hypothesis that increasing the texture depth where it is low will reduce the number of accidents.
Network C Network A No , ot ac ci de nt s/ ye ar /k m of ro ad at gi ve n te xt ur e le ve l *- Network B - - _1411.... .. O 1 1 4 L d 4 J_ 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 SMTD (mm)
Figure 9 Variation of'average annual accident rate with texture level.
The authors end by concluding that there is a need for further research to
confirm the hypothesis that improving texture reduces accidents.
Hosking (18) examined the value of using the seasonal variation in skidding resistance as a basis for estimating the effect of changes in skidding resistance on the frequency of skidding accidents. Categories studied include road usage, traffic speed, class of road, class of vehicle and region of Great Britain.
The main findings were:
(i) That there existed a very high significant correlation between month-to-month changes in wet-road skidding resistance and the wet-road skidding rate for all categories.
(ii) Month-to-month changes in traffic volume and light conditions have relatively little effect on skidding frequency.
(iii) Estimated reduction in skidding frequency per unit change in skidding resistance was similar for most categories.
(iv) Good correlation was found between wet-road skidding resistance and
dry-road skidding accident frequency. Summary
The effects of road surface friction on accident risks were studied within the TOVE-project and the results showed that the tendency is that the accidents rate decreases with increasing road friction coefficient. A study undertaken in Great Britain showed that accidents occurred on the lower textures at a greater frequency than would be expected and that the risk of accident is greater for roads with an average texture depth below about 0,7 mm than for those above this level.
4 Future research work
Friction - texture -- behaviour - traffic safety
The influence of surface friction is of significance for traffic safety.
In a study undertaken in Great Britain, one result, among others, was that accidents occurred on the lower textures at a greater frequency than would be
expected on the basis of the existence of such textures on the road. Also, it was
found that the risk of accident is greater for roads with an average SMTD (texture depth) below about 0.7 mm than for those above this level.
The relationship between road friction and accident rate was investigated within the TOVE project. The results showed that the accident rate decreases with increasing road friction.
Future work is needed to develop knowledge about how variations in longitudinal friction properties and transverse friction affects traffic safety and driving behaviour. Driving speed, the lateral position of vehicles, acceleration and retardation can be used as measures of behaviour.
Furthermore, it is necessary to investigate whether accidents occur at sites with
lower textures at a greater frequency than would be expected. Also, it is desirable to estimate suitable levels of the texture at which the proportion of accidents exceeds the proportion of that texture. The condition of the road can be measured with a Laser Road Surface Tester and the accident rate can be calculated as a function of the macrotexture.
There is a lack of accident studies directed towards the problem of the effects on traffic safety of differential friction caused by marking materials and also of split friction caused by lower friction in the ruts than outside the ruts. Reference (4) describes a design procedure for determining the maximum permissible differential friction between pavement and material, given the length of the marking on the pavement for safe operation of cars.
Pavement markings present a problem in wet conditions for motorcycles. Roughness - ruts - behaviour - traffic safety
Another subject of interest is how road roughness and ruts affects traffic safety. One conclusion from the TOVE project was that the lower accident rate for poor surfaces was most probably caused by effects related to periods without snow. Studies from Sweden detected minor differences in speed, 1-3 km/h, for different types of surfaces and standards.
More research about how road roughness affects traffic safety and driving behaviour as well as riding comfort will be needed. It is also necessary to investigate whether different measures of the longitudinal evenness and the cross profile can be used in accident studies.
Driving behaviour on uneven roads with ruts is of interest. Measurements of driving speed and lateral postion of vehicles on selected sites for different road types, with and without ruts, can be used to detect whether there are any variations in the lateral position of vehicles due to light conditions - daylight and darkness, weather conditions -- good and bad. road surface -- dry and wet.
A future study concerning ruts should start with two descriptive investigations. The main purpose of these studies is to gain knowledge about important factors
(different conditions) for drivers. safety and comfort and then to design an experimental study with a sample of test persons driving a car under varying
conditions.
l. Measurements on roads with greater rut depths.
2. Measurements on roads with more normal rut depths.
3. Experimental study.
5 References
10. 11. 12.
Inriktningsprogram 1991-1995 för forsknings- och utvecklingssamarbe-tet mellan Vägverket och VTI. Vägverkets publ 1991:05, 1991.
Wallman Carl-Gustaf, Björckebaum Fredrik and Yngvesson Tora: System för beläggningsunderhåll (PMS) -- en litteraturs'ammanställning. VTI Notat 22-95, 1995.
Arnberg Peter W, Carlsson Gunnar, Djärf Lennart, Land Per-Gunnar, Magnusson Georg, Schandersson Rein, Simonsson Bo and Wiman Leif: Underhåll av belagda vägar. Kunskapsläge och FoU-behov. VTI Med-delande 406, 1984.
Ivey L Don et al: The influence of roadway surface discontinuities on safety. Transportation Research Board, 1984.
Väghållningsåtgärder - ytegenskaper - trafikanteffekter. Dokumen-tation från seminarium 16-17 november 1988. VTI Meddelande 598, 1989.
Öberg Gudrun: Road maintenance and traffic safety. VTI Notat T43, 1992.
Beläggningsteknik och beläggningsyta -- inverkan på trafik och om-givning. Dokumentation från NKTF-seminarium 5-6 oktober 1992. VTI Meddelande 723, 1993.
Schandersson Rein: Samband mellan vägbeläggningar och trafikolyckor 1977. VTI Meddelande 242, 1981.
Björketun Urban: Samband mellan vägbeläggningar, väderlek och tra-fikolyckor 1977. VTI Meddelande 317, 1982.
Björketun Urban: Samband mellan vägbeläggningar och trafikolyckor vid olika väderlek. VTI Meddelande 393, 1984.
Linderoth Ulf: Samband mellan vägyta ochreshastighet. Etappl.
Beläggningsunderhåll på hårt slitna vägar. VTI Meddelande 273, 1981. Kolsrud Björn and Nilsson Göran K: Samband mellan vägyta och reshas-tighet. Etapp 2. Jämförelse mellan ytbehandling och massabeläggning. VTI Meddelande 277, 1981.
Hemdorff Stig, Leden Lars, Sakshaug Kristian, Salusjärvi Markku and Schandersson Rein: Trafiksäkerhet och vägytans egenskaper (TOVE).
Slutrapport. VTT Tiedotteita 1075, 1989.
14.
16.
17. 18.
22
Anund Anna: Vägytans inverkan på fordonshastigheter. VTI Meddelande 680, 1992.
Cooper D R C, Jordan P G and Young J C: The effect on traffic speeds of resurfacing a road. TRRL Supplementary Report 571, 1980.
Craus Joseph, Livneh M and Ishai Ilan: The effect of pavement condition
on highway accidents. Paper presented at the international conference on new ways and means for improved safety in Tel Aviv, 1989.
Roe P G, Webster D C and West G: The relation between the surface
texture of roads and accidents. TRRL Research Report 296, 1991.
Hosking J R: Relationship between skidding resistance and accident
frequency: estimates based on seasonal variation. TRRL research Report
76, 1986.
