Submitted for to the 2015 Transportation Research Board Conference for Presentation and Publication

Transcription

1 0 0 ROUGHNESS INDEX MEASURED WITH LINE LASER AND TRIPLE POINT LASER IN TEXTURED AND NON-TEXTURED STRIPS By Bernard Igbafen Izevbekhai, P.E., Ph.D. Corresponding Author Research Operations Engineer Minnesota Department of Transportation 00 Gervais Avenue Maplewood MN 0 Bernard.izevbekhai@state.mn.us Phone: Fax: Jed Ohiremen Tamunodienye Ig-Izevbekhai Case Western Reserve University College of Arts and Sciences 000 Euclid Avenue Cleveland OH 0 Joi@case.edu Phone: Manshean (Sharon) Wong Student Worker Paraprofessional Minnesota Department of Transportation 00 Gervais Avenue Maplewood MN 0 manshean.wong@state.mn.us. Phone: 0 Fax: 0 # of Words = # of Tables = *0= 000 # of Figures = *0= 00 Equivalent Words = Submitted for to the 0 Transportation Research Board Conference for Presentation and Publication

2 Izevbekhai, Ig-Izevbekhai & Wong 0 0 ABSTRACT Practitioners have often wondered if during ride measurement with inertial devices, motion of the laser through pavement texture introduces non-representative values of International Roughness Index (IRI) particularly in certain textures. In response to this problem, a special texture study created a non-textured strip by a recession of the middle ft of a texturing broom dragged longitudinally behind the paver. The study measured IRI, other surface properties in the adjacent textured and non-textured strips using a lightweight profiler outfitted with the line laser (ROLINE) and Triple Laser arranged in tandem. IRI measurements were performed after sufficient concrete strength gain and repeated as soon as the joints were sawn. The same measurements were repeated after the joints were deployed. Results showed a significant difference between IRI of textured strip and that of nontextured strip. Further analysis indicated that though texture appeared to affect IRI, the effect is amplified by laser type used, as the triple laser appeared to exhibit higher IRI in comparison to the ROLINE. Although the ROLINE is not a reference profiler that is unaffected by texture, the prevalence of ROLINE and Triple Laser in construction acceptance testing is sufficient reason to be concerned about the difference inherent in the results obtained. CHI squared and t-test statistical analysis showed that laser type induced comparable and even higher IRI anomalies than the experimental drag texturing. Texture and laser-induced-iri-anomaly can be minimized by measuring smoothness after true joint deployment has occurred.

3 Izevbekhai, Ig-Izevbekhai & Wong INTRODUCTION Certain pavement smoothness specification had resulted in some undesirable riding conditions including the chatter phenomena that could not be penalized due to the use of the profile indices and blanking band filters. Other factors such as anomalous ride quality due to certain textures led the industry to enquire about the effects of texture on ride at that time. Additionally, contractors had expressed concern that that the zero blanking bands may result in strict penalties since texture effects on ride measurement had not been quantified for a corrective algorithm. To address this issue, many agencies changed from a two tenths blanking band to a zero blanking band specification and subsequently to an International Roughness Index (IRI) specification. A poll conducted showed that above every other requirement, most people want smooth riding pavements (). A study performed by the Smith et al () for NCHRP (Report I- of ) correlated increase in service life to various percentage improvements in ride quality. The study showed that some Portland cement concrete (PCC) sections in Alabama experienced increases of %, % and % respectively in service life for 0%, % and 0% increase in ride quality. Minnesota PCC experienced %, % and 0% increase in service life with the respective ride quality improvements. Recent published MnDOT reports showed that smooth pavements remain smoother and the rate of deterioration of poorly riding pavements is higher (). It is therefore evident that pavement smoothness should be a major infrastructure goal. To evaluate pavement performance through ride quality it is important to ensure that measured ride quality is indicative of actual pavement condition and that any error is quantified after the source has been identified. Some of these sources of errors have been indicated in previous research work (). A Minnesota Department of Transportation (MnDOT) 00 investigation created non-textured strips between astro-turf textured finished strips on a paving project on US trunk Highway between Olivia and Bird Island. A lightweight profiler and a California Profilograph were used to measure ride quality, on each strip before and after joint-establishment. Results showed consistent deviation of 0 to 0 inches per mile of IRI between the textured and un-textured strips. Subsequent diamond-ground surface exhibited inches per mile lower IRI than the non-textured surface. A ProVAL Power spectrum density analysis showed similar high preponderant wavelengths attributed to joints and string lines (). With advancements in wave analysis frequency fragmentation and intrinsic mode decomposition mode decompositions, we are now better able to quantify roughness inducing factors. The quest for a corrective algorithm for the effect of texture on IRI led to the development of a suggested texture ride optimization software () that was superseded by an implementation of IRI in program delivery as well as a combined IRI specification for construction acceptance () (). Although a transition was made from Profile Index (PI) in program delivery to IRI, the challenge of the degree to which texture influences measured smoothness had not been fully solved. Lightweight profiler used in 00 was at that time equipped with a single laser (). Studies later conducted () observed an anomalous difference between IRI measured with point lasers and that measured by line lasers in the same profiler. Given the equal usage of line lasers and point lasers, a comprehensive evaluation of the effect of texture must include the various laser types on adjacent textured and non-textured segments. Current research addresses this issue with applicable experimental design. EXPERIMENTAL DESIGN AND EQUIPMENT USED The test section consisted of 00 feet of the outside lane on northbound Interstate Highway near Midway Road in Duluth Minnesota. The southern limit of test section was approximately

4 Izevbekhai, Ig-Izevbekhai & Wong 0 0. feet from northern limits of Midway Bridge approach-panel on this Interstate Highway (FIGURE ). This section was part of a major construction project SP 0- of a pavement of 0-inch thick dowelled concrete with non-skewed joints at ft interval. The paving was followed by finishing and broom drag texturing. In the test section, the broom was indented at the middle (between wheel paths thus creating a ft wide non-textured strip between two ft wide textured strips. Texturing was followed closely by the application of a uniform layer of Alpha Methyl Styrene curing compound. When sufficient strength gain had occurred in hours, a lightweight profiler was able to get on the pavement. The first set of smoothness measurements were conducted on the textured strip as well as the non-textured strip. MnDOT had equipped the lightweight profiler with a tandem arrangement of the Triple Laser accelerometer and the line laser (ROLINE) accelerometer as shown in FIGURE with inset of magnified laser rays and beam. In this arrangement, the two accelerometers measured IRI simultaneously followed by repeat runs in each strip. These Pre-Saw ride measurements were identified as: mndotimidwaydlpresawbwpnontxt (runs -).erd + ft from CL of road (nontextured). mndotimidwaydlpresawrwptxt(runs -).erd + ft from CL of road (broom textured) (The names and file extensions were so chosen for continuity and easy access in a myriad of project and research files). As soon as the transverse sawing was performed (:00 AM next day), longitudinal sawing commenced and was completed at approximately :00 AM. Joints were washed and generally cleaned of excess slurry. Another set of IRI measurements was conducted on the textured and non-textured sections. Post Sawing Ride Files were identified as: mndotimidwaydlpostsaw..0bwpnontxt.erd Runs -).erd + ft from CL (nontextured) mndotimidwaydlpostsaw..0rwptxt.erd (Runs 0-) + ft from CL of road (textured w/ broom) 0 FIGURE : Location of Test Section on Highway near Duluth Minnesota

5 Izevbekhai, Ig-Izevbekhai & Wong Circular Track Meter FIGURE Lightweight Profiler outfitted with two lasers, accelerometers and Circular Track Meter ASTM E-. 0 Creation of Adjacent Strips Close Up View of Adjacent Strips FIGURE 000-ft of textured and non-textured strips & close-up view Ride was measured with the International Roughness Index (IRI). The International Roughness Index is based on the suspension algorithm of the quarter car () travelling at 0 miles per hour. Vertical acceleration of the Quarter Car is associated with displacements that are summed over the travelled distance as inches/mile or m/km. IRI is neither a slope of the profile nor a summation of slopes of elements of the profile but the average rectified value of the slope powerspectrum-density (PSD).

6 0 0 Texture measurements were conducted with the Circular Track Meter (ASTM E-) in order to evaluate the texture configuration associated with the textured and non-textured segments. The circular track meter, uses a charge coupled device (CCD) laser-displacement sensor to measure the profile of an.-inch diameter circle. The CCD is mounted on an arm that rotates at. inches above the surface. It is driven by a DC motor at a tangential velocity of. ft. /min in a counterclockwise direction. Measurements were performed in conformance to the ASTM E standard. Accordingly, three measurements were made at each test cell location. The output data is segmented into eight. inches arcs of samples each. The precision for the given standard deviation of the eight measurements on the test cell is 0.00 inches. Outputs given by the device is mean profile depth (MPD) of the eight segments. By using the in-house Visual basic Program PARSER developed for visualizing basic texture output texture wavelength and texture orientation were evaluated. RESULTS AND ANALYSIS Comparison of Texture Configuration Properties Mean Profile Depth (MPD) and an unscaled texture profile were obtained directly from the CTM. Additionally, texture direction, asperity interval and texture orientation were considered necessary for adequate characterization of the texture configuration. Asperity interval is defined as characteristic wavelength of a repeating texture patterns. The turf or broom drag texture used is an anisotropic texture with longitudinal asperities represents a longitudinal texture. Texture orientation (spikiness) () is the measure of the skewness of the amplitude distribution function of a texture. It is positive or negative texture depending on the skewness from equation. PARSER provided texture depth measurements for each segment for each of three separate runs. Skewness was computed as in equation. Skewness or Texture orientation = i= (Y i Y) (N )S () where Y = depth measured from reference N = Sample size S = Sample standard deviation 0 Texture orientation is a measure of texture spikiness in pavement surfaces. Negative textures are characterized by peaks and rounded valleys that indicates appearance of asperities projected above surface; while positive textures exhibit flat peaks and sharp valleys that indicated as depressions in surface (). MPD values were obtained for various measurement points at 0.00, 00+00, and at various offsets thus evaluating the textured and adjacent non- textured strips. Tables and thus show the MPD values obtained at station 0+00 and in these strips. The MPD for the non-textured strip ranged from 0. to 0. mm while the MPD for the textured strip ranged from 0. to.0 mm. These measurements were performed weeks after paving prior to opening to traffic but not without some light construction traffic. Consequently, if texture loss would have occurred it will be proportionately higher in the 0 textured strips. Values obtained for asperity interval showed that the textured segment had lower asperity intervals of.0 to. mm while the broom texture showed asperity interval of. to.0 mm. N

7 Izevbekhai, Ig-Izevbekhai & Wong Probability density function plotted by using frequency of peak heights shows that that spiky surface has positively skewed distribution, and non-spiky surface has negatively skewed distribution. Values obtained showed that the non-textured segments had a somewhat neutral texture orientation while some were slightly positive generally ranging from 0. to +0.. The textured segments exhibited a range of -0. to -0.0 indicating a more negative texture. Table and Figure show results of texture measurements and subsequent analysis for one of many sections in the study area. Table shows mean profile depth (MPD), texture skewness and texture wavelength of the textured and non-textured strips. Evidently, the non-textured surface was not void of asperities, having been finished with the paver and float in some areas and the paver in others. Typically, texturing is performed for adequate skid resistance but it occasionally provides negative texture that imparts other surface benefits. Pavement Smoothness Evaluation A gradual decrease in the difference between IRI of textured and non-textured strip is indicated by the Triple Laser measurement (Tables &) from paving to joint deployment. Initially before sawing, there is a difference of 0. inches per mile which became. inches/mile after sawing. The reduction in the net effect of texturing is attributed in part to changes in the megatexture and stress relief from built-in warp and curl due to the joint sawing. After standard joint sawing to (/ rd of the pavement thickness) the space beneath the joint is expected to crack to the slab bottom and provide load transfer through aggregate interlock. This phenomenon is referred to as joint deployment. Though it can be accelerated by heavy equipment, such loads may cause uncontrolled cracking. Consequently prior to traffic loading, shrinkage of the concrete and restraint of the base (or interlayer) facilitate joint deployment. However, the crack propagation (joint deployment) was evident weeks after paving in this test section. After joint deployment, a difference of. inches per mile was observed. The ROLINE in all cases exhibited a lower IRI value than the Triple Laser. This justifies the fact that the bridging of the texture asperities by a line laser may be more representative of tire footprint that is not necessarily influenced by the texture asperities. Initially before sawing, ROLINE showed an IRI difference of. inches per mile between textured and non-textured strips. After sawing this difference became. inches/mile. Crack deployment resulted in a difference of. inches per mile. TABLE shows that the difference in IRI arising from the sawing of the joints is almost insignificant in the ROLINE but remarkable with the Triple Laser. Examining the actual IRI values, the Triple Laser started at inches per mile in the non-textured strip and decreased slightly to. in/mile after joints were sawn but changed to. inches per mile when the joints were deployed in the non-textured strip. These numbers are within margin of error. In the textured strip, the Triple Laser started at. inches per mile and decreased slightly to. after joints were saw but increased slightly to.0 inches per mile when the joints were deployed. The ROLINE started (pre-saw) at.0 inches per mile in the non-textured strip and changed slightly to. inches per mile after joints were sawn but increased slightly to. inches per mile when the joints were deployed in the non-textured strip. In the textured strip, the ROLINE started (pre-saw) at. inches per mile and increased slightly to. inches/ mile after joints were sawn and changed slightly to. inches per mile when the joints were deployed. These show that laser type and texturing were more influential to the changes than the joints and the deployment. Relative importance of texture and laser is evaluated with statistical tools in a later section.

8 Izevbekhai, Ig-Izevbekhai & Wong TABLE Texture Test Results of Non-Textured and Textured Strips at Stations 0.00 ft. and 00 ft Peaks Texture Orientation Texture Wavelength (mm) Non Textured Non Textured Non Textured Textured Textured Textured Mean Profile Depth (mm) Non Textured Textured Test (Sta.0) Test (Sta.0) Test (Sta.0) Average Test (Sta.0) Test (Sta.0) Average Test (Sta.00) Test (Sta.00) Test (Sta.00) Average Test (Sta.00) Test (Sta.00) Test (Sta.00) Average inch =.mm

9 Elevation Elevation Elevation Izevbekhai, Ig-Izevbekhai & Wong CTM Sector Location 0) (a) PARSER- Profile of Non-textured Strip (Station (b) PARSER Profile of Textured Strip (Station 0) CTM Sector Location (a) PARSER- Profile of Non-textured Strip (Station (b) PARSER Profile of Textured Strip (Station 00) 00) FIGURE : PARSER Recreated profile on station 00 (a) non-textured (b) textured

10 Izevbekhai, Ig-Izevbekhai & Wong 0 TABLE Triple Laser and ROLINE Pre-Saw and Post-Saw Test IRI (inches/mile) Non-textured Strip Triple Laser ROLINE Difference Non-textured Strip (Pre-Saw) mndotimidwaydlpresawbwpnontxtr.. 0 mndotimidwaydlpresawbwpnontxtr.. 0. mndotimidwaydlpresawbwpnontxtr Mean Textured Strip (Pre-Saw) mndotimidwaydlpresawrwptxt.. mndotimidwaydlpresawrwptxt.. mndotimidwaydlpresawrwptxt.. Mean..0. Mean Difference (Textured -Non-textured) Non-textured Strip Post-Saw mndotimidwaydlpostsaw..0bwpnontxt.. -. mndotimidwaydlpostsaw..0bwpnontxt mndotimidwaydlpostsaw..0bwpnontxt Mean Textured Strip (Post-Saw) mndotimidwaydlpostsaw..0rwptxt... mndotimidwaydlpostsaw..0rwptxt.. -. mndotimidwaydlpostsaw..0rwptxt Mean Mean Difference (Textured -Non-textured).0..0 TABLE : ROLINE and Triple Laser IRI Post-Joint Deployment IRI File Triple Laser IRI (in/mile) ROLINE IRI(in/mile) Diff (in/mile) Non-Textured Strip Post-Joint Deployment DLpostpro..0BWPnontxtr DLpostpro..0BWPnontxtr DLpostpro..0BWPnontxtr Mean (Non-Textured) Textured Strip Post-Joint Deployment DLpostpro..0RWPtxtr DLpostpro..0RWPtxtr DLpostpro..0RWPtxtr Mean (Textured) Nontex. post crack propagation (Textured- Non-textured) inches/mile = I m/km

11 IRI (inches/ mile) Izevbekhai, Ig-Izevbekhai & Wong TABLE : Average IRI Pre-Saw, Post Saw and Post Joint Deployment Strip Pre-Saw IRI (in/mile) Post-Saw IRI (In/mile) Post Joint Deployment (in/mile) Triple Laser Mean Non-Textured Triple Laser Mean Textured...0 ROLINE Mean Non-Textured ROLINE Mean Textured.0.. Difference Triple Laser 0... Difference ROLINE.0.. Pre-Saw IRI (in/mile) Post-Saw IRI (In/mile) Post Joint Deployment (in/mile) Triple Laser Mean Non-Textured Triple Laser Mean Textured ROLINE Mean Non-Textured ROLINE Mean Textured FIGURE IRI in pre-saw, post-saw and post-joint-deployment

12 IRI Difference (inches/mile) Pre-Saw IRI (in/mile) 0 Post-Saw IRI (In/mile) Post Joint Deployment (in/mile) Triple Laser ROLINE FIGURE IRI difference between textured & non-textured strips STATISTICAL ANALYSIS OF EFFECTS OF LASER-TYPE, TEXTURE, SAWING & JOINT-DEPLOYMENT. This section examines texture the CHI squared (χ ) test and the t-test (0) as the chosen statistical tools to evaluate the relative importance of texture and laser in influencing IRI. The χ test (Equation ) first calculates a χ statistic using the formula: (A ij B ij ) r c i= j= () B ij where: Aij = actual frequency in the i-th row, j-th column Bij = expected frequency in the i-th row, j-th column r = number or rows c = number of columns A low value of χ is an indicator of independence. As can be seen from the formula, χ is always positive or 0 only if Aij = Eij for every i, j. This test returns the probability that a value of the χ statistic at least as high as the value calculated by the above formula could have happened by chance under the assumption of independence. This test uses the χ distribution with an appropriate number of degrees of freedom (df). df= (c-) (r-) () where c = number of columns (c >) and r = number of rows (r >)

13 Izevbekhai, Ig-Izevbekhai & Wong The χ statistic was first calculated in the comparison of the textured strip IRI to the non-textured strip IRI. Subsequently it was calculated between the pre-joint deployment and the post joint deployment IRI and between laser types. The results are shown in the first rows of Table. It is evident that proximity to is an indication of similarity in this case or dependence where applicable. In subsequent rows the χ statistic was obtained for other combinations of laser type and texture. The results are shown in the rd column. It shows that the laser type used has more influence on the IRI than the joint sawing and deployment. The t-test, based on the difference between means considered data spread and computed probability of overlap. The final formula for the t-test is shown in equation : t = [(x +x ) d] SE where x is the mean of textured or laser as applicable, x is the mean of non-textured or laser respectively) d is the hypothesized difference between population means, and SE is the standard error of the mean. The p-value is the probability of observing a sample statistic as extreme as the test statistic. Since the test statistic is a t-score, the t- distribution was used to assess the probability associated with the t-score, with the degrees of freedom computed below (Equation ). If the sample findings are unlikely, given the null hypothesis, the researcher rejects the null hypothesis. Typically, this involves comparing the p-value to the significance level, and rejecting the null hypothesis when the p-value is less than the significance level. The t-value will be positive if the first mean is larger than the second and negative if it is smaller. To test the significance, risk level (called the alpha level) (α) was set to 0.0. In the t-test, the degree of freedom is given by df= (n-) (or df = cr-) () where n is the population or c is number of columns and r is the number of rows Given the alpha level and degree of freedom (df), the t-value, was obtained from a standard table of significance. This test directly returned a p-value that was compared to a pivot of 0.0. It shows that joint sawing and joint deployment are not significant in the IRI distribution. It also shows that the laser type used has more influence on the IRI than the joint sawing and deployment based on this level of significance. Table elucidates the relative importance of the various factors (laser type, joint condition and texturing) on IRI. The various factorials shown in columns and were subjected to the two statistical tests described previously and the results ranked in columns, &. The combined ranking formed the basis of evaluation of the similarity or dissimilarity between various combinations of laser types test strips and joint conditions. Joint condition refers to presawing, post sawing and post-joint deployment in this context and not to degree of distress. It is therefore evident that irrespective of joint condition, the Triple Laser and the ROLINE measurements on textured strips to a % confidence level are dissimilar in spite of the fact that ()

14 Izevbekhai, Ig-Izevbekhai & Wong 0 ROLINE vs Triple Laser in the non-textured strips showed some similarity. It can also be deduced that the laser type may be more influential than the textured versus non-textured strips in the introduction of anomalies to ride measurements. However, in the non-textured strips the Triple Laser versus ROLINE appears similar to a % confidence level. The ROLINE on the texture strip vs the ROLINE on the non-textured strip was also found to be dissimilar. Texturing appears to have an effect but this effect is amplified by the laser types and by the correspondingly different laser effects. Table arranges the tests in order of significance based on each test and sums the rankings into a final rank of which the lowest number is the most significant. It identifies laser type and texture as very significant as accentuated by the laser effect being insignificant in the non-textured strip. TABLE Comparison of Significance of Texture and Laser Combinations RANK * = Combined X and T Rank (Arithmetic Sum) TEST VARIABLE P- Value RANK* EFFECT X T-TEST X T RANK* 0 Triple Laser Textured Vs Nontextured Triple Laser Vs ROLINE (Textured) ROLINE Textured Vs ROLINE Non Textured Post Deployment Vs Pre-Joint ( All Strips) Triple Laser Vs ROLINE (Non- Textured) Post-Joint Vs Pre- Joint ROLINE Vs ROLINE (Non Textured) Texture & Laser Laser (Texture) Texture with ROLINE Deployment Laser (No texture) Sawing Reference 0..E- Clearly Significant Clearly Significant Clearly Significant Significant Non- Significant Non- Significant Reference CONCLUSION AND RECOMMENDATION Conclusions A special texture study was performed on Interstate in July 0, in Duluth Minnesota. It created adjacent textured and non-textured strips and measured IRI using a lightweight profiler outfitted with the Line Laser and Triple Laser arranged in juxtaposition. The research also measured geometries of the textured and non-textured strips. Smoothness measurements were performed as soon as the lightweight profiler could ride on the concrete and repeated again as

15 Izevbekhai, Ig-Izevbekhai & Wong soon as the joints were sawn. The same measurements were repeated after the joints were deployed. The textured strips exhibited more negative orientation than the non-textured segments, implying that the broom imparted negative textures on the surface. The non-textured strip exhibited texture isotropy while the texture scan showed that the textured strip exhibited asperity alignment in the longitudinal direction. The broom texturing process appeared to have imparted a more negative texture orientation (skewness) to the generally neutral orientation of the non-textured strip. Results showed an IRI difference of 0. inches per mile between the non-textured and the textured strips adjacent strips with the triple point laser but. inches a per mile with the line laser before joint sawing. The difference after sawing of joints were respectively. and. inches per mile after joint sawing and subsequently. and. inches per mile after joint deployment Evidently, IRI measurements conducted after observable deployment of the joints indicated largely reduced difference between textured and non-textured strip IRI Results show a significant difference between IRI of textured strip and that of nontextured strip. Further analysis indicated that though texture appears to affect IRI this effect is amplified by laser type used as the triple laser appears to indicate higher IRI in comparison to the ROLINE. Although the ROLINE is not a reference profiler from which we may establish IRI values unaffected by texture, the prevalence of the ROLINE and Triple Laser in construction acceptance testing is sufficient reason to be concerned about the difference inherent in the results obtained. Difference obtained between joint sawing and join t deployment is expected to be very significant in many concrete paving projects including unbonded overlays with non-woven geotextile stress-relief layers. In this design the lesser restraint delays effective joint deployment occasionally beyond the time of pavement acceptance smoothness measurements. Recommendations The effect of texturing can therefore be minimized by measuring smoothness for acceptance at least two weeks after paving. However, adequate accommodation for texture may be interpolated from the values obtained in this study. The point at which acceptance IRI measurements are performed should be recorded and such acceptance tests must be performed with similar quality assurance and quality control devices so that anomalous laser effects are minimized. Based on findings of this research, a cursory assessment of the degree of joint deployment should be noted (where possible) during the pavement acceptance smoothness testing. Additionally, the texture type and configuration will affect IRI to various degrees. Consequently, further research is recommended to examine different texture types and different texture aggressiveness with the various laser types including the single laser which is still common in network monitoring. ACKNOWLEDGEMENTS Robert Orthmeyer, Senior Pavement Engineer of Federal Highway Administration facilitated the unique outfitting of MnDOT s lightweight profiler with ROLINE and Triple Laser to enable this research work. MnDOT s Steve Olson and John Pantelis performed on-site testing. MnDOT District, MnDOT s Maureen Jensen and Curtis Turgeon were instrumental to the conceptualization and realization of the research. Additionally, Gerard Geib (MnDOT) provided very helpful review.

16 Izevbekhai, Ig-Izevbekhai & Wong REFERENCES ) Swanlund M. Enhancing Smooth Pavements. FHWA Public Roads September /October 000 Vol No. ) Smith, K.L.; Smith, K.D.; Evans, L.D; Hoerner, T.E; Darter, M.I. Smoothness Specifications for Pavements (Final Report ) National Co-operative Highway Research Board, National Research Council. (NCHRP - ) Snyder M.B.: Lessons Learned From MnROAD Proceedings of the th International Conference on Concrete Pavements The Golden Gate to Tomorrow's Concrete Pavements Organized by ISCP San Francisco, California, USA August -, 00. ) Izevbekhai, B.I (00) a field investigation of the influence of pavement texture on Pavement Smoothness measurements. MnDOT MnROAD Reports. URL: Accessed //0 ) W. James Wilde. Implementation of an International Roughness Index for MnDOT Pavement Construction and Rehabilitation Report MN/RC-00-0 ) W. James Wilde and Thomas J. Nordstrom MnDOT Combined Smoothness Specification URL Accessed 0// ) Izevbekhai, B.I. and E. Lukanen Laser induced Roughness Index Anomalies in Longitudinal Box car Configurations URL: Accessed 0/0/ ) Izevbekhai, B.I. Tire Pavement Interaction Noise of Concrete Pavements. Dissertation for Doctor of Philosophy in Civil Engineering. University of Minnesota. May 0. ) Sayers, M.W. On The Calculation OF International Roughness Index from Longitudinal Profiles Transportation Research Board 00 Fifth Street, NW Washington, DC 000 USA URL: 0) Izevbekhai, B.I. Watson, M.W Evaluation of Concrete Texturing Practices in Minnesota: URL: Accessed 0// DISCLAIMER This paper does not express the opinion of Minnesota Department of Transportation, Federal Highway Administration or other agency. It is the result of research and analysis conducted by the authors, whose opinions are solely expressed.