Roughness Surveys

surveys.gif (4292 bytes)

	ROUGHNESS SURVEYS Introduction Roughness Meters Roughness Meter Installation and Calibration Setting Up ROMDAS for Roughness Surveys Executing a Roughness Survey Manual Roughness Surveys
TOP	INTRODUCTION Highway engineers are interested in road roughness as it is a good overall indicator of the quality of a pavement and influences road user satisfaction and vehicle operating costs. ROMDAS was originally developed for roughness surveys and this is still one of its greatest strengths.
TOP	ROUGHNESS METERS There are two ROMDAS roughness meters: Laser Profilometer: The ROMDAS laser profilometer is a non-contact roughness meter. A precision laser measures the elevation of the instrument above the pavement while an accelerometer is double integrated to get the motion of the instrument relative to datum. The difference between these two values is the road profile, from which the IRI is calculated. Bump Integrator: This is a response-type roughness meter. These are mechanical devices which measures the relative motion of the vehicle floor to its suspension. The photo below shows the ROMDAS Bump Integrator (BI). Most ROMDAS systems are used with response-type roughness meters, and ROMDAS has been used with each of the following: ROMDAS BI Farnell BI South African Linear Displacement Integrator (LDI) Australian NAASRA meter The ROMDAS BI has been designed to be superior to the others, particularly on smooth roads where reduced sensitivity of the other instruments means that there can be significant differences in roughness between sections due to coarseness of their readings. In a recent independent field test in Indonesia Lea Consultants compared the ROMDAS BI against two Farnell BIs. It was found that: "The [ROMDAS BI] has a better clustering of data and thus generated a higher level of confidence (roughly about 4-5% better). This was most prevalent on the low IRI test sections. [This was due in part to suspension/tyre differences but also] the ROMDAS BI generate[ing] ten fold the number of bumps per test section. The additional digit (increased sensitivity) on the low IRI sections contributes to a tighter clustering of data and the higher level of confidence." The improved performance is a reflection of the higher precision of the ROMDAS BI than others. The BI encoder returns 360 pulses per revolution. With a spindle diameter of 45 mm this corresponds to approximately 2.5 pulses per mm of vertical travel, or a resolution of approximately 0.4 mm. On rough roads ROMDAS also performs well, as evidenced by the comparison with the Farnell BI described in our technical papers. The ROMDAS laser profilometer offers much higher precision than the BI, but at a much higher cost. It is also limited to operations on paved roads with roughnesses of no more than about 8 IRI m/km. However, when precise measurements are required the profilometer is the instrument of choice. For further details on the laser profilometer please contact info@dcl.co.nz.
TOP	BUMP INTEGRATOR INSTALLATION AND CALIBRATION The installation of the BI depends on the type of suspension in the vehicle. If the vehicle has a solid rear axle it is common to mount the meter over the centre of the axle. As shown in the figure below, this gives what is termed a 'Half-Car' measurement. With independent suspensions it is best to measure two 'Quarter-Cars' and to average the readings of the two instruments. This correlates best with the International Roughness Index (IRI) which is based on a quarter-car simulation. It should be noted that even with solid axles using two meters for quarter-car measurements will give improved accuracy over a singe half-car meter. A less desirable option for independent suspensions is to use a single roughness meter. This will not give as accurate a result as two instruments since the wheelpath roughness is always different between the shoulder and centreline roughness. The photo below is an example of a single ROMDAS BI installed in a vehicle with its protective cover removed. Since each vehicle responds differently to roughness, the roughness meter(s) must be calibrated against a known reference roughness. This is done be using a reference instrument to accurately profile a number of test sections and establish the IRI. The roughness vehicle is driven over these sections and a regression equation is developed which relates the raw roughness to the IRI. A full description of how one calibrates roughness meters is given in our technical papers section. It is also possible to install ROMDAS to a trailer. This was done in Kenya and a photo of the system is here. The advantage of using a trailer is that it will not be used as frequently, and thus be requiring recalibration, and it can also be used with a wider selection of vehicles--often desirable if the servicability or availability of parts for the survey vehicle is at all in question.
TOP	SETTING UP ROMDAS FOR ROUGHNESS SURVEYS The more time spent preparing for a roughness survey the easier the survey will be for the operators to execute and the less time will be spent processing the data. Usually, every hour spent preparing for the survey will save 5 or more hours in processing. Planning. The routes to me measured should be plotted out on maps. Depending upon the country and road conditions, one can measure 150 - 300 km/day so the planning should be based on these sorts of figures. It is often best to plan for 150 km modules which can then be combined to larger modules if conditions allow. LRPs. Location referencing is essential to the successful execution of any survey. See our page on location reference surveys for details on why this is so. Road ID. ROMDAS is designed to make use of ROAD ID information. These are character strings defining a section to be surveyed. The Road Description file in ROMDAS can be created which associates key information against the ROAD ID, such as the file names to store the data, the road description, LRP reset options, etc. All the operator needs to do is to enter the ROAD ID for the route of interest and this additional data will be automatically supplied. This ensures consistency and data quality.
TOP	EXECUTING A ROUGHNES SURVEY Roughness surveys should be done by driving the vehicle at a constant speed. If it is necessary to operate at a range of speeds different calibration equations should be established to cover this range. The ROMDAS software will apply the appropriate equation given the survey speed. It is useful to run a regular check of the data. This consists of weekly running over a section of road measured previously and confirming that the results are not markedly different. The vehicle should also be recalibrated at 5-10,000 km intervals. Roughness data are collected over what is called a sampling interval. This is typically 100 m, although it can be anywhere from 50 to 5000 m. The roughness is summed over the sampling interval and a single value is given, for example 120 counts/100 m. At the end of the interval the roughness is reset to 0 and summed for the next interval. When the data are converted to IRI the units are adjusted to a per km basis. Instead of using regular intervals such as 100 m the operator can optionally manually define the end of the sampling interval. This is commonly done when there are km markers and the roughness between markers is of interest, but it is not the recommended practice. When measuring two-lane highways it is often necessary to measure both lanes. When completing a survey the user is offered the option of doing the Other Side. This tells the software that it is a two-lane highway and the sampling intervals are to be correlated. This means that the first sampling interval in the opposite direction is a shorter interval than usual, resulting in the same set of sampling intervals (see the (b) below). If this was not done the top figure (a) would arise. This feature is used when there are no LRP resets. In executing a roughness survey one of the most difficult components is the programming of the survey. It is often necessary to interrupt what should be a continuous run to measure roads running off from the primary survey route. This can be easily handled with ROMDAS, and an example of this is shown in the figure below. The operators should terminate the main survey at an LRP or other roadside feature which will be easy to return to (Run 1). It is generally recommended that this be past the point where the other surveys are to commence. They may then execute the other surveys (Run 2 and Run 3). Returning to where they ended the previous survey they enter the same Road ID as was used previously. They will then be given the option to Continue the previous run. The data collection will be started where they left off and, once processed, both components of Run 1 will be integrated. It is not always practical to run a separate LRP survey before the roughness survey. If this is the case, you can create the LRPs during the roughness survey by using the LRP Define option. At each LRP the operator presses the Insert key and is then prompted to type in the name of the LRP. Optionally, the LRPs can be predefined and a short form keycode (eg 100 for km 100) can be used. If defining LRPs the operator can end the survey and define the final LRP at the same point by pressing the F9 key. It is sometimes desirable not to measure the roughness over some sections, for example if the survey comes across a pavement which is being reconstructed. In this instance by pressing the F5 key the roughness meter measurements will be turned off until the key is pressed again. The software will optionally calculate the Equivalent Roughness which is simply the roughness over the interval which was measured, factored up to cover the whole interval. If it is likely that the vehicle will travel at low speeds, either due to congestion or having to stop at LRPs for measurements/photos, then the option to use ‘equivalent roughnesses’ at low speeds should be used. This sees the roughness measurements adjusted when the speed is below a user-specified minimum speed based on measurements at higher speeds. One of the greatest problems encountered when conducting annual surveys is in reconciling the survey lengths. This is because in spite of the utmost care, the measured lengths will never be identical in two successive surveys. In year 1 the road may be 41.153 km long; in year 2 it will be 41.168 km while in year 3 it may be 40.952 km. Since relational database systems expect all roads to have the same start and end points it is necessary to "rubber band" the data so that they match. ROMDAS is designed to do this "rubber banding" in real time. When the data are reset at the LRP the measured distance is replaced with the distance in the LRP file and the data adjusted accordingly. Thus, every year every survey will have the same start and end points as the LRP survey – thereby eliminating any post-processing of the data.
TOP	MANUAL ROUGHNESS SURVEYS The discussion above has focused on the ROMDAS system which uses a computer to record roughness data during surveys. It is also possible to use the low-cost ROMDAS Manual Roughness Counter which is a stand-alone unit which sees the operator manually record the roughness. Details on the Manual Roughness Counter are here.


	� 2006 Data Collection Ltd. All rights reserved. Last updated 28 August 2006