Jump to navigation Jump to search
CTC Exeter Sunday Rides Coffee Pots Bike Bus Devon News

Surface Roughness Introduction Measurements Discussion Reference Testing Issues Belgium To Do

Experimental Surface Roughness Discussion

Some preliminary experiments suggest that modern smartphones can be used to assess the surface quality of cycle paths.

In this initial study a bike was ridden at around 10 mph over the route of interest and the vertical acceleration of the bike measured. If different speeds were used during monitoring then the size of the acceleration variations would be different. Thus the acceleration values should be analysed to derive the surface profile of the track before assessing the quality of the surface.

It is suggested that the measured 'quailty' be divided into different categories so that the quality can readily be understood. These categories might then be colour coded so that maps of surface quality can be produced. e.g.

Proposed Bike Path Quality Categories
Icon Category Colour Name Example RMS Acceleration
Range (m/s^2)
at 10 mph
Perfect Phone Stationary (~0.03) < 0.20
Bullet green.png 0 Green Ideal Topsham Road Service Road (~0.2) 0.20 - 0.45
Bullet yellow.png 1 Yellow Good Exe Trail (Double Locks) (~0.5) 0.45 - 1.00
Bullet orange.png 2 Orange Acceptable 1.00 - 2.25
Bullet red.png 3 Red Not Ideal Tesco Bridge Path (~3.2) 2.25 - 5.00
Bullet black.png 4 Black Dangerous Lakeside Avenue (~5.0) > 5.00

The proposed RMS acceleration values are on a log scale with each breakpoint a factor of x2.25 larger than the previous one.

Second Thoughts

Mercalli scale related to peak ground acceleration.png

It may be best to use a scale which increments in steps of x2. This would then accord with the Mercalli scale used to characterise the ground acceleration due to earthquakes.

On the Mercalli scale an acceleration of 5 m/s^2 (0.5g) is characterised as equivalent to a severe earthquake.


Roughness map.png
Roughness pop up box.png

Such measurements could conveniently be presented on an interactive map.

In the example (right) the measurements are presented as colour-coded recangles overlaid on a map. The data rectangles are aligned with the direction of travel at the time the data was collected. In this example data was sampled for 10 m of travel (about 2 seconds) then processed to give a representative RMS acceleration value.

This allows the user to readily identify problem areas. In the example (right) the service road adjacent to Topsham Road in the north is of good quality (green/yellow) whereas Lakeside Avenue in the south (red/black) is poor or dangerous.

If a data rectangle is clicked then more detailed information is revealed (see left).

In this case this includes

  • the date/time the measurements were made
  • the calculated RMS acceleration
  • the speed at which the data was collected, and
  • the direction of travel (this is important on wide roads)

Idealised Calculations

Roughness calculation for sinusoidal surface.png

The Belgian standard requires a vertical departure of no more than 5 mm in 3 m horizontally (i.e. a = 0.0025 m and L = 3.0 m). Assuming a cyclist travels at 10 mph (v = 4.47 m/s) the formula gives a maximum acceleration of 0.22 m/s^2. The RMS value would be √2 of this i.e. 0.16 m/s^2. This would be classified as 'ideal' above.

In the case of the Tesco path integration of the accelerometer measurements implied an amplitude of around 4 mm (a = 0.004 m) and observations suggest a wavelength of about 70 cm (L = 0.7 m). Using these figures the formulae gives a maximum acceleration of 6.4 m/s^2 - or 4.6 m/s^2 RMS. (this can be compared with measured RMS values of around 3.2 m/s^2).

Surface Roughness Introduction Measurements Discussion Reference Testing Issues Belgium To Do
CTC Exeter Sunday Rides Coffee Pots Bike Bus Devon News