Home

Introduction & purpose - Minnesota Local Road Research Board

1. 14 4 3 1 Compare with Specifications aisl 15 4 3 2 Xieostatisacal Ad di 16 3 User s Manual 22i arieso tette mordet a ree 17 O o 6c e Er e i e dei i ie 17 5 2 Software Installation and Setups saeco desc o eiim sey sn eave avn i a 17 5 9 Data Import and Data Entry iii tanda cad sansa ini ER ERR 17 o Data A de pod as e a its 17 2 2 2 Data Entity iet ia 20 04 E Creatas se eeu E c E 21 a A AO 21 5 5 Basic ArcInfo Procedures ica de da da 22 Sol Basemap and Shapebilesss oes aan dba Desi Pec a ao 22 55392 15 Valid Procedures Said A ticos 24 5 5 3 Target Value Classification and Coloring IC Data eee 26 5 5 4 Calculating Percent Passing sitos ide diia eo iv Cae isaac 29 5 6 Geostatistical Analysis of IC Di lic 29 5 6 1 Interactive Historia 30 3 0 2 Interactive Trend Analysis divinidad cdas candado echa d sade arica 31 210 3 G statistical Wizard tetitas eii 33 5 6 3 1 Raw vs Interpolated IC Data 33 5 6 3 2 Geostatistical Parameters foe eae eee Miia es 35 5 6 3 3 Gradient of Compaction Value eese 40 5 7 Improvement Analysis Procedures for Comparing Overlying Coverages 40 Skk E 40 PS A TTL 40 DELO A cordon ond etis Pe a tatu colonies dearly ahead nae Ue 41 RU A M 45 5 7 4 1 Reduction in Subsequent Roller Compaction Value 45 5 7 4 2 Percent Improvement S
2. 6 04 4 53 302 151 A T 10 i z o i i WES CHAT 54 145 E55 T 9 Tetraspherical 99085 Pentaspherical ET aney Exponential Gaussian hino Range E4 Rational Quadratic e Hole Effect K Bessel J Bessel Stable 0 o Es Ene iam aii TS ia ea Distance h Parameter Ej f Partial Sil ig 25 1 53 m Semivariogram Covariance Surface Show Search Direction Iv Nugget Angle Direction oo f Hele ed 21 393 Error Modeling Angle Tolerance 450 IV Shifts Bis Bandwidth lads 60 sy zz Pas Semivariogram Covariances Lag Number var amp Varl y Size 8 2295 ofLags 12 P5 159 Spherical 54 145 21 393 Nugget lt Back Finish Cancel Figure 28 Semivariogram and Covariance Dialog Box Showing Raw Data Red Dots Model Right Top Model Parameters Right and Model Results Yellow Line 4l The Semivariogram Covariance Modeling dialog box initially opens with a best estimate of a model and model parameters that match the spatial variability of the sampled data In the upper left corner of the dialog box shown in Figure 28 is a semivariogram with sampled data as red dots and the best estimate model shown as the yellow line On the right side of Figure 28 is the control panel for the model which is shown as the yellow line on the Semivariogram The default model an isotropic Spherical model and
3. Technical Report Documentation Page Report No 2 3 Recipients Accession No MN RC 2009 35 4 Title and Subtitle 5 Report Date Mn DOT Intelligent Compaction Implementation Plan December 2009 Procedures to Use and Manage IC Data in Real Time EAT 7 Author s 8 Performing Organization Report No D Lee Petersen Jeff Morgan Andrew Graettinger 9 Performing Organization Name and Address 10 Project Task Work Unit No CNA Consulting Engineers 2800 University Avenue SE Suite 102 11 Contract C or Grant G No Minneapolis MN 55414 c 91076 CompNet Concepts 2348 Timberlea Dr Woodbury MN 55125 University of Alabama Civil Construction and Environmental Eng Dept Box 870205 Tuscaloosa AL 35487 0205 12 Sponsoring Organization Name and Address 13 Type of Report and Period Covered Minnesota Department of Transportation Final Report Research Services Section 14 Sponsoring Agency Code 395 John Ireland Boulevard Mail Stop 330 St Paul Minnesota 55155 15 Supplementary Notes http www lrrb org pdf 200935 pdf 16 Abstract Limit 250 words Mn DOT research has indicated that intelligent compaction IC will improve construction quality and efficiencies for the Contractors and Mn DOT field staff Experience thus far has illustrated that quality control quality assurance and research activities were problematic using the software provided by the roller manufacturers The manufacturers software is proprieta
4. Figure 18 An Example Coverage with IC Data Above 120 of Target Value is Highlighted From the compaction value data shown in Figure 17 and Figure 18 it is clear that the majority of the data is above 90 of the target value for this coverage but it is also clear that most of the data above 90 of the target value is also above 120 of the target value Large areas that are in red in Figure 17 and Figure 18 may be good locations to perform spot checks of compaction The raw compaction value data can also be viewed thematically by going to the Symbology tab on the Layer Property dialog box On the Left side of the Symbology tab select Quantiles and Graduate Colors In the Fields area of the Symbology tab choose Compaction as the Value to thematically map To adjust the classification break values the user should click the Classify button and the Classification dialog box will appear An example of the classification dialog box is shown in Figure 19 21 ix r Classification Classification Statistics Method Natural Breaks Jenks y Count 10000 Minimurn 0 70000 Classes 5 gt Maximum 63 40000 ata Poon Sum 213966 30000 Mean 21 39663 Exclusion Sampling Median 22 10000 7 62144 Standard Deviation Columns 100 lt Show Std Dev Show Mean Break Values x S8 S s S8 S8 400 8 8 8 8 S 12 20000 e 00 z 9 m 18 60000 a 24 30000 30 00000 69 40000 0 70000 17
5. Cancel Figure 40 Create New Shapefile Dialog Box to Create the Centerline Polyline Shapefile 2 Draw centerline Return to ArcMap and add the new CenterLine shapefile to the map Begin an editing session set the target layer to CenterLine and set the Task to Create New Feature Select the Sketch Tool pencil and move the cursor to the beginning of the roller compaction data pick points along the centerline Double click at the end of the line to finish the polyline centerline feature An example of a centerline polyline is shown in Figure 41 wee CenterLine Join Output Figure 41 Centerline of Compaction Area Represented by a Polyline 3 Change centerline polyline into points along the centerline Select HawthsTools gt Animal Movements gt Convert Paths To Points lines to points Use the CenterLine 47 polyline shapefile and the Input line feature the Unique ID field can be left blank and set the Interval Between Points to 1 which is one foot Deselect the Add turning angles to output table Enter an output point shapefile name CenterLine Points and click OK This produces a shapefile of points separated by one foot along the centerline of the road Spatially join the CenterLine Points shapefile to the roller compaction points which will join the nearest centerline point having the distance down the center line to the roller compaction point data Right click on the roller compaction point data and selec
6. 50 Coverage 2 s Coverage 1 Elevation ft 971 I T T T T T 0 50 100 150 200 250 300 Station ft Figure 47 Elevation Data for Coverage 1 and 2 5 7 4 5 Average Variation along Centerline This analysis is used to identify trends in data longitudinally along the compacted area Previously roller compaction point data was spatially joined to the nearest centerline station point Summarizing averaging the roller compaction data based on station provides an average of transverse data at each station Plotting these average values against the centerline station produces figures that can be used to identify trends and variability along the longitudinal direction 1 Summarize roller compaction data based on centerline station Open the roller compaction data and right click on the centerline station column Select Summarize to open the summarize dialog box Select any data field of interest and check average Fields from either coverage may include drum left and right elevation roller compaction value thickness and percent improvement Save this summary output table CenLine Ave and add it to the map This data can be joined back to the roller compaction points thereby providing the average transverse data value to compare each data point against The dbase file can also be opened in Excel and graphed Examples of the centerline analysis graphs are shown in Figure 48 Figure 49 F
7. Figure 27 Unsampled or Invalid IC Data Filled with Interpolated Data Circled Areas 35 Figure 28 Semivariogram and Covariance Dialog Box Showing Raw Data Red Dots Model Right Top Model Parameters Right and Model Results Yellow Line 36 Figure 29 Anisotropic Semivariogram Covariance Modeling Dialog Box 37 Figure 30 Major and Minor Range vs the Calculation Distance seeeeee 39 Figure 31 Partial Sill Nugget and Sill vs Calculation Distance eee 39 Figure 32 Kriged Surface Gradient Red Large Gradient Green Small Gradient 40 Figure 33 Overlapping Roller Compaction Coverages eene 41 Figure 34 Hawth s Tools Convert Paths to Points Dialog Box eee 42 Figure 35 Roller Compaction Data Represented by the Original Lines and the New Points 42 Figure 56 Spatial Jom Dialog BOX ime ii ieeitdecer ta tetris tt IA ERR e I set odi qr debo Ede dta ded 43 Figure 37 Spatially Joined Roller Point Data that Joined to Point 1 Foot 44 Figure 38 Highlights Indicate Where Overlying IC Data is Less than Underlying Data 45 Figure 39 Summary Statistics on Percent Improvement Column eee 46 Figure 40 Create New Shapefile Dialog Box to Create the Centerline Polyline Shapefile 47 Figure 41 Centerline
8. IC Data Represented by Both Lines Green and Points Black 5 6 1 Interactive Histogram To generate an interactive histogram of the IC data for a coverage select the Explore Data button on the Geostatistical Analyst pull down menu and then select the Histogram Tool A histogram will open and in the lower left corner of the dialog box change the Attribute to CompactionValue Typically coverages have enough data to increase the number of bars in the histogram from the default of 10 bars to 50 or even more An example histogram of compaction value data is shown in Figure 22 In this figure summary statistics are shown in the upper right corner of the histogram and include data count min max mean and 1 median and 3 quartile values for the data set The histogram is in scientific notation and is interactive meaning that a selection on the histogram will highlight data on the map This can be seen in Figure 22 where the low compaction value bars of the histogram have been selected and the corresponding data locations are highlighted on the map This feature may be used to identify locations where the compaction value may be unacceptable 30 Frequency 10 Skewness 0 25924 11 8 20 Kurtosis 2 6535 1 st Quartile 17 2 21 855 Median 22 5 1 6 8036 3 rd Quartile 26 7 9 44 7 08 4 72 2 36 0 o a 0 03 0 46 0 89 1 32 1 75 2 18 2 61 3 04 3 47 3 9 4 33 Data 10 Figure 22 Interactive Histogram with Sele
9. performs the validation checks populates the geodatabase and writes GIS shapefiles The software is written to permit implementation of a different importer for each manufacturer The software was developed in Visual Studio 2008 and targets the Version 3 5 NET framework Open Standards were followed throughout project development 4 2 2 2 Validation Validation processes are needed to ensure that irrelevant or erroneous IC data is not included in the GIS visualization A list of validation criteria some of which are applied by the import routine and some by the end user follow Criteria that may be applied by the import routine are typically numerical checks regarding data limits and machine operating parameters End user criteria generally address the spatial limits of coverages IC data is valid only if the machine operating parameters fall within manufacturer specific ranges Each IC record is checked against these criteria during import Invalid data is imported and maintained in the database but is flagged accordingly The following is an example of validation criteria that may be used during the import process the listed criteria are Caterpillar roller specific Roller compaction value is numeric and between the ranges of 0 and 150 Roller speed is less than 4 mph Valid GPS position must be Yes GPS mode must be RTK Fixed RMV must be between zero and 17 Roller vibration frequency must be between 28 hz and 34 hz Machin
10. right click on the shapefile of interest in the table of contents and click Properties This opens the Layer Properties dialog box Select the Symbology tab and then click the Import button A predefined layer symbology has been set up and is stored in the GIS Files directory in the project directory Click the browse icon on the Import Symbology dialog box and navigate to the GIS Files directory Select IsValid Symbology lyr and click the Add button Click OK on the remaining dialog boxes and the IsValid Symbology will be applied to the IC data as shown in Figure 15 With this layer symbology Valid data is shown in green Invalid data is shown in red and Marginally Invalid data is shown in black From this analysis it is clear that the majority of the coverage shown in Figure 15 is valid but some areas of the coverage have invalid or marginally invalid IC data The invalid and marginally invalid areas should not be included in the IC analysis and should therefore be removed ISVALIDREC Valid Invalid Marginally Invalid Figure 15 Map Showing Valid Invalid and Marginally Invalid IC Data A predefined definition query has been set up and can be used to automatically remove invalid and marginally invalid data from an analysis The query is stored in the GIS Files directory in the project directory Right click on the shapefile on interest in the table of contents and click Properties This opens the Layer Properties dialog bo
11. 87500 35 05000 52 22500 69 40000 Snap breaks to data values Cancel Figure 19 Layer Symbology Classification Tab The classification dialog box shows a histogram of the IC data basic statistics of the data in the upper right corner and break values on the right side The largest value in the break values should match the maximum value shown in the classification statistic window The break values are automatically calculated but can be adjusted by clicking on the number in the break value window and typing in a new value or by selecting one of the blue lines on the histogram and adjusting the break location For large data sets typing in the break number is more efficient The mean and standard deviation of the Compaction Value data can be turned on by checking the boxes directly above the histogram This information may assist a user in determining or adjusting the break values A graduated color thematic map of compaction value data having the distribution shown in Figure 19 is shown in Figure 20 where large compaction values are in dark red and small compaction values are in pink 28 STAPLES 10 9 5 2007 17 35 40 COMPACTION 0 70000 12 20000 12 20001 18 60000 18 60001 24 30000 24 30001 30 00000 30 00001 69 40000 Figure 20 Thematic Map of Compaction Value Data 5 5 4 Calculating Percent Passing Right click on a layer name in the table of contents and click Open Attribute Table At the bottom
12. 914 46 301 0 46 301 8 25 200 199 92 32 394 28 922 20 298 49 22 4 50 200 199 86 25 201 32 97 15 974 48 944 8 50 400 399 4 45 031 23 263 27 063 50 326 4 100 400 399 81 42 527 24 393 25 958 50 351 8 100 800 799 62 85 053 18 349 33 766 52 115 The geostatistical parameters of interest are plotted in Figure 30 and Figure 31 Figure 30 shows how the major range blue and minor range cyan increase with increased calculation distance Figure 31 shows how the partial sill decrease as the nugget increases as the calculation distance increases The combination of the partial sill and the nugget is the sill which increases slightly as the calculation distance increases It can be seen from Figure 30 that the range partial sill and nugget are sensitive to the calculation distance while the sill is relatively insensitive to the calculation distance It should be noted that the typical use of the semivariogram model is to produce an interpolated grid of data Because of the very dense grid spacing of the compaction data any of the models presented in Table 2 interpolates the data approximately the same Therefore if the goal of using the Geostatistical Wizard is to determine the range partial sill and nugget then the results are highly sensitive to user input If on the other hand the goal is to interpolate data then the results are insensitive to the model parameters because of the density of the sampled data In either case a geostatistical analysis add
13. CalibrationArea Yes No 1 ProofLayer Yes No 1 CreatedDateTimeStamp Date Time 8 UpdatedDateTimeStamp Date Time 8 Approved Yes No 1 ApprovedDateTimeStamp Date Time 8 Reworked Yes No 1 BorrowMaterialID Text 255 CoverageSequenceNumber Double 8 LayerThickness Long Integer 4 LayerThicknessUnit Text 50 FoundationCondition Text 255 MoistureContent Long Integer 4 TargetCompactionValue Double 8 SpecificationID Text 255 StandardDeviation Double 8 CoefficientVariation Double 8 Mean Double 8 StatisticalTargetValue Double 8 A 4 Table Geography Name Type Size ID Long Integer 4 ProjectID Text 255 PassID Text 255 X1 Long Integer 4 X2 Long Integer 4 Y1 Long Integer 4 Y2 Text 255 Description Memo DescriptiveName Text 255 CreatedDateTimeStamp Date Time 8 UpdateDateTimeStamp Date Time 8 Table Project Name Type Size ID Text 255 ProjectName Text 255 ProjectDescription Memo SponsorProjectNumber Text 100 ContractorProjectNumber Text 255 SponsorSubDivision Text 255 ProjectStartDate Date Time 8 ProjectEndDate Date Time 8 ProjectPhases Long Integer 4 ProjectCoordinateSystem Text 255 ProjectDatum Text 255 CreatedDateTimeStamp Date Time 8 UpdatedDateTimeStamp Date Time 8 Table QA Name Type Size ID Text 255 TestType Text 255 TestDateTimeStamp Date Time 8 CoverageID Long Integer 4 QATestValue Memo P
14. InternalField Text 255 CriteriaRequired Yes No 1 CheckSequence Long Integer 4 A 7
15. Options button Choose Select By Attribute Select points where the percentage improvement is less that zero and apply this selection Highlighted points show a decrease in roller compaction value which may indicate a need to recompact or could be used to direct field compaction testing Figure 38 shows an example set of spatially joined coverages where the subsequent roller compaction value is less than the previous compaction value first coverage second coverage Join Output Figure 38 Highlights Indicate Where Overlying IC Data is Less than Underlying Data 5 7 4 2 Percent Improvement Statistics This analysis is used to understand the overall improvement of an overlying coverage compared to an underlying coverage Open the attribute table of the joined point coverage shapefiles and right clicking on the percent improvement column Per Imp then click on statistics A dialog box containing summary statistics and a histogram will open an example of which is shown in Figure 39 The statistics dialog box shows minimum and maximum percent improvement as well as the mean improvement which is the average point by point improvement over the coverage A mean percent improvement of 300 indicates that the second coverage has an average roller compaction value that is three times higher than the first coverage 45 Statistics of Join Output 2 x Field Per Imp z Frequency Distribution Statistics Count 8160 Min
16. Screen uela op is 21 Figure 11 Typical Shapetile Names ti 22 Figure 12 Typical Startup Screen for the IC GIS eseeeseseeeseseessiseresresseserssressesererresseseresreeseese 23 Figure 13 Basemap with Aerial Photo Layer Turned On eee 24 Figure I4 Raw IC Shapehle Data 00 pe pe este Races A Unt A a Bae uae 24 Figure 15 Map Showing Valid Invalid and Marginally Invalid IC Data 25 Figure 16 Map Showing Only Valid IC Data eese nennen 26 Figure 17 Example Coverage Where IC Data Below 80 of Target Value is Highlighted 26 Figure 18 An Example Coverage with IC Data Above 120 of Target Value is Highlighted 27 Figure 19 Layer Symbology Classification Tab eese 28 Figure 20 Thematic Map of Compaction Value Data esee 29 Figure 21 IC Data Represented by Both Lines Green and Points Black 30 Figure 22 Interactive Histogram with Selected Bars Highlighted on the Map 31 Figure 23 Trend Analysis Dialog Box with IC Data esee 32 Figure 24 Compaction Value Trend along Left and Across Right the Coverage 33 Figure 25 Selected Points in the Interactive Trend Analysis Highlight on the Map 33 Figure 26 Raw IC Data with Interpolated Data by Inverse Distance Weighting amp Kriging 34
17. and light green data points will appear on the Trend Analysis graph The input data are shown as light green points in the lower portion of Figure 25 These data can be selected on the Trend Analysis graph and the resulting locations will be highlighted on the map as shown in the upper portion of Figure 25 Figure 25 Selected Points in the Interactive Trend Analysis Highlight on the Map 5 6 3 Geostatistical Wizard The geostatistical wizard is a tool that was designed to assist users in interpolating data values between sampled locations Because IC data fundamentally covers all of a compacted area there is no benefit in interpolating data to unsampled locations and there may be misrepresentation of the data if interpolation is used The geostatistical tool however may be used to visualize spatial statistic parameters that may help in understanding and evaluating IC data Unfortunately some of these geostatistical parameters are sensitive to user input and therefore should be used with caution or not at all 5 6 3 1 Raw vs Interpolated IC Data Compaction value data is shown in Figure 26 as a thematic map Raw IC Data top a deterministic interpolation Inverse Distance Weighting middle and a geostatistical 33 interpolation Kriging bottom It can be seen in this figure that the results of all of these methods are virtually the same The Raw IC Data map in Figure 26 is each IC line colored based on the compaction value wh
18. being joined that are closest to it and a count field showing how many points are closest to it How do you want the attributes to be summarized Ayverat Minimum F Standard Deviation I Sum Maximum Variance Each point will be given all the attributes of the point in the layer being joined that is closest to it and a distance field showing how close that point is in the units of the target layer 3 The result of the join will be saved into a new layer Specify output shapefile or feature class for this new layer Es EARCH MN DOT Improvement_Spec Join_Output shp 5 About Joining Data Cancel Figure 36 Spatial Join Dialog Box Next limit the allowable distance for the spatially joined data Because a spatial join will match points from the second coverage to the nearest point in the first coverage there is a potential to have points join that are separated by a large distance Set up a Definition Query by right 43 clicking on the spatially joined output layer and selecting Layer properties to limit the allowable spatial join distance Select the Definition Query tab and limit the allowable X Y distance to be less that 1 foot You can use the Query Builder or enter Distance 1 into the text box After applying this query only points on the second coverage that match a point on the first coverage within 1 foot are displayed An example of the limited point data is shown in Figure 37 and can b
19. lag size multiplied by the number of lags and represents the total distance that sampled data are compared to produce the semivariogram Although there is no standard method to select either the lag size or the number of lags a rule of thumb is the lag size should be about the average sample spacing and the number of lags should be about half the total sample distance divided by the lag size For the compaction data shown in Figure 26 the average longitudinal data spacing is about 1 foot while the average transverse spacing is about 8 feet Therefore two lag sizes of 4 feet and 8 feet were investigated and are shown in the first column of Table 2 The compaction data used in this analysis is about 80 feet across and about 900 feet long therefore the number of lags was varied from 12 to 100 as seen 37 in column 2 of Table 2 This produced a calculation distance from 48 to 800 feet as shown in column 3 of Table 2 The second row of Table 2 shows the default values produced by the Geostatistical Analyst tool The geostatistical model parameters of major and minor range partial sill nugget and sill are shown in columns 4 through 8 of Table 2 respectively Table 2 Influence of Lag Size and Number of Lags on Geostatistical Parameters Lag Number Calculation Major Minor Partial Nugget Sill size ofLags Distance Range Range Sill 4 12 48 47 413 13 421 41 899 0 41 899 8 2295 12 98 754 97 546 25 91 32 967 14 151 47 118 4 25 100 99 96 12
20. map may require rerolling because a major portion of the area was not properly measured The same area after Inverse Distance or Kriging interpolation benefits the coverage by producing high compaction value data at locations that actually were not measured properly Raw IC Data Inverse Distance Weighting Figure 27 Unsampled or Invalid IC Data Filled with Interpolated Data Circled Areas 5 6 3 2 Geostatistical Parameters The Geostatistical Wizard can be use to evaluate spatial parameters unfortunately these parameters are highly dependent on user selected input and therefore are not consistent Begin by clicking the Geostatistical Wizard on the Geostatistical Analyst pull down The Input Data should be the point shapefile representing the center of the drum and the Attribute should be the CompactionValue Select Kriging from the methods box in the corner of the Geostatistical Wizard Choose Input Data and Method dialog box On the Step 1 of 4 dialog box choose Ordinary Kriging Prediction Map and then click Next The Step 2 of 4 dialog box contains interactive semivariogram and covariance models that can be manipulated to produce geostatistical parameters of interest An example of the Step 2 of 4 dialog box is shown in Figure 28 35 Geostatistical Wizard Step 2 of 4 Semivariogram Covariance Modeling 2 xl View Madels Semivariogram Covariance v Model 1 Ir Model 2 Model 3 Major Range E e
21. measurement Each measurement record contains a link to the project roller and coverage table This table also includes the time and location information about the measurement The coordinates of the end of the drum are imported and stored in this table In order to facilitate geostatistical analysis in GIS the center of the drum is calculate during import and stored in this table This table is linked to the project roller and coverage tables 4 Coverage information table The project scope includes only data from the final or measurement coverage on a proof layer However this table was named the coverage information table in anticipation of expanding the functionality to all coverages and layers in the future For the current scope this table contains one record for each final coverage on a calibration area or proof layer Note that the data stored in this table will permit the user to distinguish between calibration areas and proof layers In the future the data structure would also permit distinguishing between any coverage Quality assurance table This table is used to the information about quality assurance tests performed on the unbound layers Each record provides the quality assurance test results the time and location of the test and the IC coverage corresponding to the test Borrow material table This table contains a record for each separate unbound material used on the project where separate means having a target va
22. of this window click the Options button and then click Select By Attributes The Select By Attributes Dialog Box will appear and the user should double click the field percent of compaction value field PRENTOFCV and then the greater than sign The user then needs to enter the 90 which is 9096 of the target value that was determined during data loading This query is also stored in the GIS Files directory and can be loaded from the Select by Attributes dialog box by clicking the Load button browsing to the GIS Files directory and clicking on the Above 90 Target Value Expression exp Clicking Open will load the predefined Above 90 Target Value Expression query Clicking Apply will select all the records in the coverage that are above 90 of the target value The Attribute table will reappear and rows that meet the selection query will be highlighted At the bottom of the Attribute table is the number of records meeting the criteria shown as Records X out of Y Selected where X is the number of records selected and Y is the total number of records in that coverage By dividing X by Y and multiplying by 100 the percentage of records above the 90 target value can be determined 5 6 Geostatistical Analysis of IC Data Several tools exist in the add on Geostatistical Analyst extension that can be employed to evaluate and understand IC data and data trends These tools include an interactive histogram where bars on the histogram can be selected and corr
23. the end user to simply open the project in ArcGIS and add a layer to show the current level of compaction 5 2 Software Installation and Setup The software is installed on a workstation by the executing the setup package which will install all the necessary files The application requires the NET Framework to be installed the setup package will install the Framework if missing on the workstation There are no prerequisites needed for the software to run After installation is complete the user must edit the configuration file to specify the location of the database and output Open the importer exe config file with a text editor and change the two file paths as appropriate for the user s path structure The two locations to be edited are 1 The connection string defines the location of the database Change the path C Users jjmorgan Documents CNA IntelligentCompaction ICDB mdb to the desired location 2 The output directory string defines where the output files will be placed shapefiles Change the path c junk gis shapefiles to the desired location Be careful during editing to not change other text in the configuration file 5 3 Data Import and Data Entry 5 3 1 Data Import Start the application to set up a project import IC data or enter QA data On the main screen Figure 6 you will choose the project you are importing to which coverage you want to update or if you want to add a new layer you would choose to Ad
24. to compaction of both unbound and bound materials used to construct the entire flexible pavement structure This final report describes the target functionality terminology geodatabase structure import and filtering software and ArcInfo geographic information system GIS platform processes 1 Introduction 11 Purpose and Historical Development This document is the final report for Mn DOT Agreement No 91076 Intelligent Compaction Implementation IC Data Management Under a prime agreement with CNA consulting Engineers the team of CNA Jeff Morgan of CompNet Concepts and Andrew Graettinger of the University of Alabama developed software and processes to load filter rework store manipulate and visualize the large quantities of data produced by intelligent compaction equipped rollers The focus of the contract was shifted as more was learned about the nature of IC data and the requirements for GIS import and data linking became clear The original concept was to focus on developing methods guidelines and procedures to import IC data into a GIS compatible database import the database into ArcInfo and conduct quality assurance QA activities at construction sites using tough laptops Software development was necessary to accommodate the variety of data formats provided by the roller manufacturers to filter and validate the data and to write GIS shapefiles in addition to the database files Mn DOT s purpose also changed from developi
25. 04 Intelligent systems for QA QC in soil compaction Proceedings of the Annual Transportation Research Board Meeting Washington D C CD ROM Terzaghi K Peck R and Mesri G 1996 Soil Mechanics in Engineering Practice 3rd Ed John Wiley amp Sons New York NY Thompson M and White D 2006 Estimating compaction of cohesive soils from machine drive power Journal of Geotechnical and Geoenvironmental Engineering ASCE Reston VA submitted for review Thompson M and White D 2007 Field calibration and spatial analysis of compaction monitoring technology measurements Transportation Research Record Journal of the Transportation Research Board National Academy Washington D C submitted 7 31 06 for review Thurner H and Sandstr m A 1980 A new device for instant compaction control Proceedings of International Conference on Compaction Vol II Paris France pp 611 614 White D Jaselskis E Schaefer V Cackler E Drew I and Li L 2004 Field Evaluation of Compaction Monitoring Technology Phase I Final report Iowa DOT Project TR 495 Des Moines IA September 55 White D Jaselskis E Schaefer V and Cackler E 2005 Real time compaction monitoring in cohesive soils from machine response Transportation Research Record Journal of the Transportation Research Board National Academy Press No 1936 Washington D C pp 173 180 White D Thompson M J
26. 6 04 4 apu waar waar d m A we Exponential V Anisotropy te 453 Gaussian Minor Range 4 Rational Quadratic 2581 3 02 z Hole Effect S K Bessel 151 J Bessel Direction B ry A E ya Stable 80 7 0 12 5 25 355 50 625 75 92 5 100 ios deci Parameter Eg PataSi ag m Semivariogram Covariance Surface 32 967 Show Search Direction lv Nugget m2 Anale Direction fo o 14 1 51 Error Modeling Angle Tolerance 45 0 B nale Tolerar 4 v Shifts Bg T 6 E sandwidth lags le D Bandwidth lags 6 0 E ae Y Semivariogram Covariances Lag Number var amp Varl Size 8 2295 ofLags 12 32 967 Spherical 37 546 25 91 90 7 14 151 Nugget Back Finish Cancel Figure 29 Anisotropic Semivariogram Covariance Modeling Dialog Box 41 Geostatistical parameters of interest include the major and minor ranges and the sill which is a combination of the partial sill and the nugget Unfortunately these parameters are sensitive to the lag size and the number of lags which are user inputs As shown in Table 2 various lag sizes and number of lags were investigated to evaluate the affect these input parameters have on the geostatistical parameters of interest those being major range minor range partial sill and nugget Table 2 is ordered by Calculation Distance from smallest to largest The calculation distance is the
27. 9 34 29 csv 12 3 2007 2 53 PM Microsoft Office E 1 437 KB n 1 Computer 1550667J003RS STAPLES 10 9 7 2007 15 03 13 csv 12 3 2007 2 53 PM Microsoft Office E 64 KB E Pictures 0667J003RS STAPLES 10 9 8 2007 07 33 53 csv 12 3 2007 2 53 PM Microsoft Office E 922 KB E H E Music 0667J003RS STAPLES 10 9 10 2007 09 46 09 csv 12 3 2007 2 53PM Microsoft Office E 2105 KB p echo 0667J003RS STAPLES 10 9 11 2007 06 52 28 csv 12 3 2007 2 53 PM Microsoft Office E 2 578 KB y 0667J003RS STAPLES 10 9 11 2007 07 46 32 csv 12 3 2007 2 53 PM Microsoft Office E 1 600 KB di Public 0667J003RS STAPLES 10 9 12 2007 07 02 57 csv 12 3 2007 2 53PM Microsoft Office E 2 800 KB f amp i 0667 003RS STAPLES IO O47 OU Eset ES Type Microsoft Office Excel TEE Separated Emm File Xs 6 0667 003RS STAPLES 10 9 12 2007 17 33 44 csv Size 2 73 MB KB 155 0667J003RS STAPLES 10 9 13 2007 07 17 13 csv Date modified 12 3 2007 2 53 PM KB Folders 155 0667J003RS STAPLES 10 9 13 2007 12 55 50 csv 12 3 2007 2 53 PM Microsoft Office E 2 041 KB File name y Import Files csv E om Figure 7 File Explorer Dialog Box Once the file has been read and converted the application will make some initial calculations with the data and prompt the user for additional information First it will calculate the Coefficient of Variation COV to help in determining if the data is statistically useful or not A low C
28. Environmental Geotechnics Melbourne Australia pp 245 250 Anderson T Embacher R A Graettinger A J Morgan J and Petersen D L 2009 Software and Processes for Intelligent Compaction Data Analysis Proceedings of the Transportation Research Board Annual Meeting Washington D C Brandl H and Adam D 1997 Sophisticated continuous compaction control of soils and granular materials Proceedings of the 14th International Conference on Soil Mechanics and Foundation Engineering September Hamburg Germany pp 1 6 Briaud J and Seo J 2003 Intelligent Compaction Overview and Research Needs Report December Texas A amp M University College Station TX ESRI 2007 ArcGIS Desktop version 9 1 380 New York Street Redlands CA 92373 8100 Fleming P Frost M Lambert J 2006 Sustainable earthworks specifications for transport infrastructure Proceedings of the Annual Transportation Research Board Meeting Washington D C CD ROM Forssblad L 1980 Compaction meter on vibrating rollers for improved compaction control Proceedings of the International Conference on Compaction Vol IL Paris France pp 541 546 Petersen D Siekmeier J Nelson C and Peterson R 2006 Intelligent soil compaction technology results and a roadmap toward widespread use Proceedings of the Annual Transportation Research Board Meeting Washington D C CD ROM Sandstr m A and Pettersson C 20
29. LN Deed eSI eA SE LUN D Erit 2 2 2 2 A ce ee Oe uri mtetlq bu aan aoe O cedens 2 RO Mae S etu et t 2 2 24 Quality Control and Quality Assurance eec dd 2 22x33 Target ET n 3 2 2 6 Test Pad Layer Calibration Area and Coverage ocoooococinccconncconoconcnonncconccconacinoss 3 3 Implementation Background Information encendido e reete bep em ts ee ipes 5 Dol Geodatabase Description eese terit eiit RI DIE 5 SA sGeodatabase Concert tii 5 3 1 2 Geodatabase Tables and Structure isis ini eas dete ce UG 6 2 Geographical Representan u c ria 8 4 Functional and BEoCeSses A A A A needs 9 Al Introduction aissein ed 9 4 2 Custom Software Funcional cc d pasat 9 4 2 1 Setup and General Information Entry eese 9 4 2 2 Import and Validation of IC Data os 10 Z2 ME EN aN 810 u ocior E E E E ood E duds 10 42 212 Validation ui tt 11 4 2 2 3 Concept of Attempted Compacti0N oooococnnocccnoncccnoncnnonncnonnncnonnncnnnos 11 4 2 2 4 Implementation of Validation Criteria eene 12 4 2 3 Statistical Caleulations nen da c dre tede iia ebur detracta 12 4 2 4 Creation OF GIS Shapefiles sess Cortez ee koe dades Ta ERN E EUM amends 13 A RA bosco a a tende talem deis Sa tests 13 424 Slapelless A pat I eG end eu ta ehe tus 14 42 5 Quality Assurance Test Data Entry ttes t drid oc usd goo 14 4 3 GIS Functional m
30. OV indicates more uniform compaction 18 File Project Roller Coverage Help Project Staples Highway 10 y Coverage Number Add New Coverage Pass Roller 1 V Create Additional Shapefiles Coefficient of Variation is 0 138374880460444 Do you want to save this import Figure 8 Coefficient of Variation Display If you chose YES the system will write out the values to the IC database for future reference Once that is completed the system will calculate what would be a good target value for this coverage based on the data presented The suggested target Value is 1 28 standard deviations below the mean based on the rationale described in Section 4 2 3 You may enter any value for a target value 19 MEC Project Man B X j File Project Roller Coverage Help Project Staples Highway 10 v Coverage Number Add New Coverage e Pass v Target Value Figure 9 Target Value Prompt Once the process completes the import the data will be in the database and the shapefiles will be in the output directory 5 3 2 Data Entry Data entry in the system is mostly driven via free text in a textbox a drop down list or menu driven as shown on the project data entry dialog box in Figure 10 20 ud Project Information gt 4 all 1 of1 TX ld Clicking on the Calendar button Project Name Staples Highway 10 Project Start Date 5 7 2009 gv brings up a selectable Project End Date V 7 Ba
31. Pealendar Sponsor Project Number Created Date Time Stamp 4 May 2009 Contractor Project Number Updated Date Time Stamp Sun Mon Tue Wed Thu Fri Sat 26 27 28 29 30 1 2 Sponsor Sub Division a 4 5 f 8 9 10 11 12 13 14 15 16 Project Phases 17 18 19 20 21 2 23 Project Coordinate System on pn zh 2 2 2 a d 3 4 35 Project Datum C Today 5 7 2009 Project Description Thisis an update Textbox a for data entry Figure 10 Project Data Entry Screen You can see the textbox labeled on screen and the calendar control as well Near the top you have a navigation menu that lets you add new records delete records and save updates 5 4 Shapefile Creation In the import process up to three shapefiles may be created The main shapefile that is always created is one that has the line data to represent the roller and the work being done If you check the option on the main import screen to create other shapefiles there will be two additional files created One will be a representation of 9 points along the drum contact point that can be used for Percent Improvement analysis see Section 5 7 The second file is the center of the drum that is also used for statistics There may be a delay in the processing if you chose to create additional shapefiles as it takes time to calculate the points and update the shapefile The larger the dataset the longer the import will take 5 4 1 Shapefile Names The names of the
32. Select the Symbology tab and then click the Import button Click the browse icon on the Import Symbology dialog box and navigate to the GIS Files directory Select the of Target Value Symbology lyr and click the Add button Click OK on the remaining dialog boxes and the of Target Value Symbology will be applied to the IC data as shown in Figure 17 With this layer symbology IC data shown in green exceeds 90 of the target value while IC data shown in red is below 80 of the target value STAPLES 10 9 5 2007 17 35 40 PRCNTOFTCV 8096 80 90 90 100 100 120 gt 120 Figure 17 Example Coverage Where IC Data Below 80 of Target Value is Highlighted Both the classification criteria and the coloring of the classes can be fully controlled by the user Five classes are automatically defined and based on the Target Value they are less than 80 80 90 90 100 100 120 and greater than 120 The five classes are initially colored red for the lowest two classes that are less than 90 of the target value and green for the three classes greater that 90 of the target value At any time a user can change the color of any class by simply clicking on the line in the table of contents and choosing a different color For example Figure 18 shows the greater than 120 class in yellow STAPLES 10 9 5 2007 17 35 40 PRCNTOFTCV lt 80 80 90 90 100 100 120 gt 120
33. Value from Coverages 1 and 2 40 Coverage 2 35 Coverage 1 100 per Mov Avg Coverage 1 30 100 per Mov Avg Coverage 2 25 gt O 20 igi 5 tc 15 V 7 10 5 V d U 0 T T T T T 0 50 100 150 200 250 300 Station ft Figure 43 100 Point Moving Average of Roller Compaction Value from Coverages 1 amp 2 2000 4 1500 1000 500 Improvement Station ft Figure 44 Percent Improvement of Coverage 2 over Coverage 1 49 2000 100 per Mov Avg Improvement Improvement 1500 1000 500 Improvement 500 Station ft Figure 45 100 Point Moving Average of Percent Improvement in Compaction Value 5 7 4 4 Lift Thickness Analysis Lift thickness can also be analyzed by subtracting the Z coordinate of the lower coverage from the Z coordinate of the upper coverage The thickness can then be thematically mapped as shown in Figure 46 where thick areas of the lift are red and thin areas are blue The thickness data can also be analyzed with respect to the centerline as was done with the percent improvement in Section 5 7 4 3 An example of the thickness trends along the center line is shown in Figure 47 Thickness ft 1 1650 1 4000 1 4001 1 5250 1 5251 1 6200 18201 1 7350 17351 1 9550 Figure 46 Thematic Map of Thickness between Coverage 1 and 2
34. alue Classification and Coloring IC Data see Section 5 5 3 Calculating Percent Passing see Section 5 5 4 Compare with Specifications Intelligent compaction performance specifications continued to mature during the life of this project The 2007 Specifications established three numerical limits to be applied to proof layer IC measurements The following excerpts list these limits For acceptance of compaction at each proof layer during general production operations all segments of the granular treatment shall be compacted so that at least 90 of the IC compaction parameter measurements are at least 90 of the applicable corrected IC TV prior to placing the next lift The Contractor shall re compact and dry or add moisture as needed all areas that do not meet these requirements Additional IC compaction parameter measurements shall be taken for acceptance of the re compacted areas If localized areas have an IC compaction parameter measurement of less then 80 of the corrected IC TV the Contractor shall re compact and dry or add moisture as needed these areas to at least 90 of the IC TV prior to placing the next lift If a significant portion of the grade is more than 20 in excess of the selected corrected IC TV the Engineer shall re evaluate the selection of the applicable calibration area corrected IC TV If an applicable corrected IC TV is not available the Contractor shall construct an additional calibration area to reflect the po
35. and attribute information contained in the geodatabase The shapefile contains objects e g points lines and polygons i e a vector format and the attributes associated with these objects e g compaction value In this project the principal shapefile includes these following attributes the record ID the unique field the compaction measurement target value the compaction measurement machine speed 1 2 3 4 a flag indicating if the measurement is valid invalid or marginally valid 5 6 measurement time stamp and 7 the percentage of the compaction target value Three shapefiles may be created 1 The principal shapefile containing the information described in the preceding paragraph which represents the data as a line from one end of the drum to the other 2 A shapefile with nine points along the drum including one at each drum end and one in the center Each of the nine points has the same compaction value this shapefile is created if the percent improvement assessment described in Section 5 7 is planned 3 A shapefile with one point at the drum center This shapefile is a created if geostatistical analysis is planned see Section 5 6 4 2 5 Quality Assurance Test Data Entry Companion tests for quality assurance may be entered into the geodatabase then later viewed and analyzed in GIS Even though the geodatabase table and fields refer to quality assurance tests any test may be entered and stored i
36. ane while the right view shows the same data projected on the Y Z plane It can be seen from the green trend line that the 3l compaction value trends slightly upward as one moves from left to right down the coverage The blue trend line in Figure 23 and Figure 24 shows the average compaction data value is larger along the center line of the coverage and drops off along either side of the coverage This pattern of higher compaction vales along the center line of a coverage is typical Trend Analysis ES mma IV Legend Rotation Angles Location 0 3D Graph Horizontal 120 Vertical 11 5 Rotate Locations ao LEEEM Perspective a gt Tip Glick or drag over points to select Add to Layout Graph Options gt Number of Grid Lines IV Projected Data x fis Y fi 4 E on Projections ticks ze 3 v Axes Input Data Points Grid Line Width 1 7 Data Source Layer Attribute CompactionData Events Compactionvalue Figure 23 Trend Analysis Dialog Box with IC Data 32 Figure 24 Compaction Value Trend along Left and Across Right the Coverage The Trend Analysis tool is interactive which allows data selections made in the tool to be highlighted on the map To select data in the Trend Analysis tool the Input Data Points have to be turned on In the lower right corner of the dialog check the Input Data Points can be checked on
37. assTest Yes No 1 Station Text 255 Offset Text 255 Latitude Text 255 Longitude Text 255 CreatedDateTimeStamp Date Time 8 UpdatedDateTimeStamp Date Time 8 A 5 Table Roller Name Type Size ID Text 255 ProjectID Text 255 RollerSerialNumber Text 255 RollerModel Text 255 RollerManufacturer Text 255 RollerPower Long Integer 4 RollerPowerUnits Text 255 RollerOperatingMass Long Integer 4 RollerOperatingMass Units Text 255 DrumWidth Long Integer 4 DrumWidthUnits Text 255 DrumDiameter Long Integer 4 DrumDiameterUnits Text 255 DrumMass Text 255 DrumMassUnits Text 255 SpecialEquipment Text 255 RollerDescription Memo RollerNumber Text 255 CreatedDateTimeStamp Date Time 8 UpdatedDateTimeStamp Date Time 8 Table tbl Line Points Name Type Size RecID Double 8 LineID Long Integer 4 X Double 8 Y Double 8 Z Double 8 CompactionValue Double 8 Table tbl Polygon Name Type Size ID Long Integer 4 GeoID Long Integer 4 XCoordinate Double 8 YCoordinate Double 8 Valid Yes No 1 CreatedDateTimeStamp Date Time 8 UpdatedDateTimeStamp Date Time 8 A 6 Table tbl Validation Criteria Name Type Size ID Text 255 ProjectID Text 255 RollerID Text 255 CriteriaName Text 255 CriteriaValue Text 255 CriteriaType Text 255 CriteriaDescriptor Text 255 CriteriaBound Text 255
38. ation wide leader in IC many of the reports and studies recently completed include Mn DOT demonstration and pilot projects starting with three demonstration projects in 2005 and more in subsequent years The next step is to implement IC as a Quality Control tool for the contractor and to review the IC data as part of Mn DOT s acceptance program The Department implemented intelligent compaction IC for unbound materials on a construction project contractual basis on four projects during the 2007 construction season In 2008 one 2007 project continued and two additional unbound material projects and one hot mix asphalt were completed Experience illustrated that quality control quality assurance and research activities were problematic using the software provided by the roller manufacturers The manufacturers software is proprietary expensive subject to change and generally did not provide the functionality required by Mn DOT Hence the Department chose to develop software and processes fitting their specific needs The work included 1 Development of database structures for managing and archiving IC data 2 Software to import and validate IC data populate the database and write geographic information system GIS shapefiles 3 Processes and tools to manage display and evaluate IC data within ArcInfo GIS software Intelligent compaction generates vast amounts of data which required special handling The end product is equally suited
39. cted Bars Highlighted on the Map 5 6 2 Interactive Trend Analysis A three dimensional trend analysis can be performed by clicking on Explore Data on the Geostatistical Analyst pull down menu and then select the Trend Analysis button In the lower right hand corner of the Trend Analysis dialog box shown in Figure 23 change the Attribute value to CompactionValue A cloud of input data will appear at coordinates X Y and Z where the Z coordinate is equal to the CompactionValue On the left hand side of the Tread Analysis page deselect Sticks and Input Data Points under Graph Options What remains is the data projected on the X Y plane in red which is a plan view the X Z plane in green which is the distance along a coverage vs compaction value and the Y Z plane in blue which is the distance across a coverage vs compaction value Projected compaction value data are shown in Figure 23 in a perspective view along with data trend lines The green trend line shows the compaction value trend along the coverage while the blue line shows the compaction value trend across the coverage The perspective view of the graph can be adjusted by selecting Graph from the Rotate pull down menu just below the graph on the dialog box Using the slider bars below the lower right corner of the graph the graph can be rotated to different views as shown in Figure 24 Two views are shown in Figure 24 the left view shows the compaction value data projected on the X Z pl
40. d analysis This section describes the basic procedures for data loading display and analysis It is assumed that the user has basic ArcGIS skills but details of the procedures are provided 5 5 1 Basemap and Shapefiles An ArcGIS project basemap will be in the project directory The map file extension is mxd Clicking on this file will open ArcGIS and the project map file Initially the project map will consist of a basemap that was previously setup by the MNDOT GIS group The basemap will consist of portions of the project design file line work that include edge of pavement The basemap may also contain an aerial photo layer of the project area The basemap is not projected and the units are in feet for easting and northing This coordinate system is the same coordinates that are output by the IC rollers 22 The initial GIS screen will look similar to the screen shown in Figure 12 which is an example of a basemap This figure shows the basemap in the map window and the available layers are in the table of contents along the left side of the screen By checking layers on and off in the table of contents different layers will be displayed in the map window This can be seen by comparing Figure 13 with Figure 12 where Figure 13 has the additional aerial photo layer turned on Loading IC shapefiles and the project geodatabase tables is done through the Add Data button on the standard toolbar in ArcGIS The shapefiles will be located on the disk
41. d new coverage and finally the roller that you are importing for At the bottom of the form is a checkbox that if chosen will generate 17 additional shapefiles that have a point representing the center of the drum and another shapefile that has several points along the drum to be used for statistical calculations IC Project Manager Ll ese File Project Roller Coverage Help Project Staples Highway 10 Coverage Number Add New Coverage Pass y 1 Figure 6 Import Main Screen Once you hit the import file button the software will prompt you as shown in Figure 7 to locate the data file that you will to import Hit open once the file is selected to import it ron _ oe J E v 55 f Search GU de gt Jeffery J Morgan Documents CNA Intelligent Compaction Data THIO Staples Data Roller Data gt SS By Organize y ews fm NewFolder G Favorite Links Name Date modified Type Size Tags E Reine 1550667J003RS STAPLES 10 8 24 2007 16 42 22 csv 12 3 2007 2 52 PM Microsoft Office E 3107 KB 155 0667J003RS STAPLES 10 9 5 2007 17 35 40 csv 12 3 2007 2 53 PM Microsoft Office E 5 384 KB i B erect hekorl ee R 0667J003RS STAPLES 10 9 6 2007 15 54 59 csv 12 3 2007 2 53PM Microsoft Office E 3041 KB Recent Places E 0667 003RS STAPLES 10 9 7 2007 08 37 03 csv 12 3 2007 2 53PM Microsoft Office E 1 408 KB il W Desktop N 0667J003RS STAPLES 10 9 7 2007 0
42. e Gradient Green Small Gradient 5 7 Improvement Analysis Procedures for Comparing Overlying Coverages 5 7 1 Purpose This section provides detailed procedures and techniques to compare and analyze the roller compaction values of two or more overlying coverages The act of compaction should improve soil properties which are related to the roller compaction value Subsequent lifts of soil when properly compacted should 1 show an increase in roller compaction value at a specific location and 2 show less variability in the roller compaction value as the thickness of compacted material in a fill increases Analyzing roller compaction data based on a comparison of subsequent data at the same location provides insight into the quality of compaction possible location for field verification and potential locations requiring additional compaction effort Procedures to prepare the roller compaction data are presented herein along with several analyses and example outputs 5 7 2 Setup Software It is assumed that the raw roller compaction data is in the form of lines representing the drum soil contact These line data are converted to point data which allows for spatial joins the basis of this approach Required software includes 1 ArcMap 40 2 ArcCatalog 3 HawthsTools freeware available at www spatialecology com htools download php 4 Excel or other graphing program 5 7 3 Procedures First identify the two layers of roller com
43. e compared to the point locations in Figure 35 2 a oa first coverage second coverage Join Output Figure 37 Spatially Joined Roller Point Data that Joined to Point 1 Foot Finally calculate the percent improvement of the second coverage over the first coverage Open the attribute table and click on the Options button Select Add Field and the Add Field dialog box will open Name the column with less than 10 characters e g Per Imp choose Double for type set the Precision to be 8 and the Scale to be 2 and click OK Right click on top of the new column in the attribute table and click calculate values As shown in 3 to calculate the percent improvement 46 imp select the roller compaction value field from the second coverage RCV gt 2 and subtract off the compaction value from the first coverage RCV and then divide by the first compaction value and multiply the result by 100 to give the values in percentage improvement RCV RCV x100 RCV improvement 3 44 5 7 4 Analysis 5 7 4 1 Reduction in Subsequent Roller Compaction Value This analysis is used to identify locations where the roller compaction value on an overlying lift is less than the underlying lift To select points where a decrease in roller compaction value occurred on a subsequent coverage select all points where the percent improvement is less than zero Open the attribute table of the spatially joined coverages and click the
44. e gear must be Forward Vibration must be On E a dS ER Vibration amplitude must be less than 0 5 mm 10 Automatic mode must be Manual All roller makes and models will have similar criteria defining valid data 4 2 2 3 Concept of Attempted Compaction Several visualization processes are important to assessing and approving coverages Two processes are typically done together viewing the physical extent of a coverage and understanding where the roller operator was attempting to take compaction measurements i e attempted compaction Figure 5 shows the measurement locations for a coverage with the data colored to indicate where the roller operator was or was not attempting to collect measurements Referring to the figure key Invalid measurements are where the machine produced unreliable compaction measurements For example the roller was not vibrating or was backing up and turning around at each end of the coverage and was not producing reliable data The other two categories Valid and Compaction attempted indicate that the operator was attempting to collect valid measurements For example at the starts and ends of a pass where the roller was speeding up or slowing down the collected measurements were not reliable Within passes there 11 are also locations were measurements were attempted but loss of GPS signal or other aberrations prevented reliable measurement Finally there are location
45. ect table is included in the geodatabase Data Types Database Table Name Project Not included Not included Coveragelnformation Not included CompactionData Figure 2 Relationship between Possible IC Data Types and Database Tables 3 1 2 Geodatabase Tables and Structure The three data tables contain the majority of the data in the geodatabase six additional tables contain additional information necessary to implement the desired functionality The geodatabase was developed using Microsoft Access primarily because of anticipated limited resources on field laptop computers However an open data layer allows for a SQL server or any other ODBC database Data keys are included in fields to allow multiple projects to be combined into a master database for archiving Appendix A lists all geodatabase tables field names data types and field sizes The tables and their purpose are l Project table This table contains project information like project name project number dates datum coordinate system etc Every major database table contains the unique project ID from this table 2 Roller table This table contains information about each roller used with a unique field linked to the IC measurement table Fields in this table include roller model horsepower weight etc di Compaction measurement table This table is the principal intelligent compaction data table and contains one record for every
46. ed in the geodatabase table for the project includes the project name likely including the S P number start and end dates contractor any project phases and the project coordinate system and datum The project coordinate system and datum are stored in the geodatabase to facilitate archival and subsequent retrieval in a state wide IC system Many of the data items in the project table should conform to some standards about content and format Appropriate standards should be developed prior to widespread use Database amp GIS Shapefile Key Figure 4 Data Flow and Principal Processes for Assessing Intelligent Compaction Data 4 2 2 Import and Validation of IC Data 4 2 2 1 Import Typically IC data is stored on a memory card on the compactor The data is downloaded to a laptop or desktop computer and then exported to a comma delimited ASCII file using proprietary software developed by the roller manufacturer The Department has specified a standard format for the structure and fields of the comma delimited file but has not been successful in enforcing adherence to this format each manufacturer has their own preferred format The custom import and validation software has been designed to facilitate implementation of new input routines to import data that is not in the required format A new input routine must be programmed for each data format 10 The software reads manufacturer s exported data formats typically comma delimited
47. er by the roller A pass may be in forward or reverse vibrating or not vibrating etc A pass produces IC data but the data may not be valid unless the criteria listed in Section 4 2 2 2 are satisfied Coverage One or more passes that constitute complete rolling of a lift Typically several coverages are necessary to compact a lift The specifications are applied to the IC data from the final coverage of a lift designated as a proof layer Hierarchy The hierarchy of compaction data relationships is illustrated in Figure 1 and described below see also Section 3 1 1 a Many roller measurements are taken during a pass Roller measurements are stored in the database b One or more passes constitutes a coverage However passes are not tracked in the database One or more coverages are necessary to compact a layer Coverage information is stored in the database d One or more layers are necessary to construct a subgrade Layer information is not stored in the database One or more of the layers are designated as proof layers The final coverage on a proof layer is designated as being a proof layer Several Coverages Compactive Effort for a Lift Some Lifts are Proof Layers Figure 1 Measurement Pass Coverage Terminology 3 Implementation Background Information 3 Geodatabase Description 3 1 1 Geodatabase Concept The geodatabase concept for this project was developed by considering the data types inherent in intel
48. ere higher values have darker color The Inverse Distance Weighting map in Figure 26 is a made by a deterministic method that calculates a grid cell value based on a summation of the inverse distance times the compaction value of neighboring samples points Finally Kriging is a geostatistical method that like the inverse distance method weights nearby neighboring samples higher than distant samples The Kriging weights are produced through spatial statistical methods Both the Raw IC data and the Inverse Distance method show similar values while Kriging smoothes the IC data and removes the extremes in the data set This can be seen by looking at dark colored areas in the Raw and Inverse Distance maps compared to the same location in the Kriging map Raw IC Data Inverse Distance Weighting Kriging Figure 26 Raw IC Data with Interpolated Data by Inverse Distance Weighting Kriging Interpolating data does have the affect of filling in data at unsampled locations For examples red ellipse on the Raw IC Data map shown in Figure 27 highlights an area of a coverage where IC data was not collected or was invalid That same area is shown with relatively high compaction values dark color on the Inverse Distance and Kriging maps This is because the surrounding compaction value data was high and when interpolating nearby data is assumed to represent unsampled locations In this example it is clear that the highlight area on the Raw IC Data
49. esponding points on the map will be highlighted an interactive 3 D trend analysis where IC data is presented in X Y Z format where the Z coordinate is the Compaction Value and a geostatistical wizard that can be employed to calculate semivariogram parameters such as range and sill The geostatistical analyses described in this section require a point data set To create an event file of points from Compaction Data the CompactionData table from the ICDB database must be in the map Add the data table CompactionData from the ICDB database using the Add Data button from the standard toolbar To create point events go to the tools menu and select the Add X Y Data button and choose the CompactionData table In the X Field and Y 29 Field choose center of the drum DrumCenterX and DrumCenterY To create a point layer for the entire CompactionData table simply click OK on the Add X Y Data dialog box Apply a query in the Definition Query tab of the Layer Properties dialog box to reduce the point layer event file to just valid IC data for the specific coverage of interest The following query will reduce the point data CoverageID desired coverage number AND IsValid 0 An example of the resulting display is show in Figure 21 where the lines connect the ends of the roller drum and the black points are the center of the drum After the creation and filtering of the IC point data the geostatistical analyses can be performed Figure 21
50. ge M Unique ID field Beo s ptions Interval between points fi interval must be in coordinate system units e g meters for UTM IV Add Y coordinates to output table J7 Add turning angles to output table Output Output point shapefile CARESEAR CHAMN DOT Improvement SpecMirst coverage points shp tl Web Hep DK Exit DEA ae Figure 34 Hawth s Tools Convert Paths to Points Dialog Box first_coverage second coverage o first coverage points 2 second coverage points Figure 35 Roller Compaction Data Represented by the Original Lines and the New Points Next join the roller coverage line attribute data to the new roller coverage points so each point will have all the information of the roller line Right click on one of the point coverages and select Joins and Relates gt Join In the top selection box on the Join Data dialog box select Join 42 attributes from a table In box 1 choose the RecID field in box 2 choose the line coverage layer and in box 3 choose the RecID i e the record identification Open the attribute table to make sure the line attributes are joined to the points Repeat this attribute join for the other coverage Then export the point layers that have been joined to the line attribute data This will create a point file with all the attribute data in a file rather than being generated through an attribute join on the fly Next spatially join the fi
51. igure 50 and Figure 51 51 981 980 979 978 977 c 2 976 S 975 E 974 D Average Drum Left Coverage 1 M Average Drum Left Coverage 2 PS pt Average Drum Right Coverage 1 972 ES Average Drum Right Coverage 2 971 0 50 100 150 200 250 300 Station ft Figure 48 A verage Elevation of the Right and Left Side of Drum from Coverage 1 and 2 1 80 1 60 t 1 40 1 20 O c F 0 60 Average Thickness 0 50 100 150 200 250 300 Station ft Figure 49 A verage Thickness of Soil between Coverage 1 and 2 52 e Ave RCV Coverage 2 s Ave RCV Coverage 1 0 50 100 150 200 250 300 Station ft Figure 50 A verage Roller Compaction Value and a 10 Point Moving Average 3000 4 2500 gt Ave Per Imp 2000 gt E 9 5 1500 2 e E 1000 A e E x Oe e e 2 i M st 500 2 7 lt ib es o UEM A cr e Ya u 5 v Me e tue ope e ws We a A E 0 ee T T c i RC add T 50 100 150 200 250 300 500 Station ft Figure 51 A verage Percent Improvement Coverage 2 Value over Coverage 1 Value 5 7 4 6 Moment of Improvement This analysis presents pairs of compaction values from coverage 1 on the X axis and coverage 2 on the Y axis as shown in Figure 52 The 45 line in Figure 52 represents no change in the compaction value from coverage to 2 at a specific location Points above
52. imum 85 000000 Maximum 4960 000000 Sum 2484423 840000 Mean 304 463706 Standard Deviation 393 125490 85 893 1871 2849 3827 4805 404 1382 2360 3338 4316 Figure 39 Summary Statistics on Percent Improvement Column 5 7 4 3 Roadway Center Line Trends This analysis is used to show trends in roller compaction data or percent improvement along the centerline of the roadway Each point of roller compaction data is joined to a station 1 foot separation along the compaction area centerline The roller compaction data can then be analyzed as a graph where the X axis is the station offset and the Y axis is the roller compaction data of interest To accomplish this analysis a centerline must be drawn and turned into one foot station points This station points can then be joined to the roller compaction data 1 Creating a centerline shapefile Open ArcCatalog right click on the project folder click on New and then select Shapefile On the Create New Shapefile dialog box name the shapefile CenterLine and set the feature type to Polyline as shown in Figure 40 and click OK to create an empty CenterLine 46 Create New Shapefile ax Name CenterLine Feature Type Polyline m r Spatial Reference Description Unknown Coordinate System d Show Details E dit Coordinates will contain M values Used to store route data Coordinates will contain Z values Used to store 3D data
53. in the user defined directory determined during the IC data loading procedures The IC database will be in the main project directory and any table in the database can be loaded with the Add Data button Once the desired shapefiles are loaded into the project map they will appear on the map as lines with a randomly selected color An example of raw IC data in the shapefile format is shown in Figure 14 ho m mant arria arenda 0 itini f DM p pt jes p Wedon Dv Deu e 3E TILA u x aS QQuurees Pork oases 44 e TJ A mews 11 O AAA 2 STE A tar Map Window Table of i Contents 4 cw ae D cd zj Cre tome mmn Jeacz v y j oft aem ne cart me LIT d Figure 12 Typical Startup Screen for the IC GIS 23 Be Edt Yow Insert Selection Tos Widow tiep E WEE oe me ex oo ve m s eenw m x t mmm n e QQuize es sPBukoAaun jaa r X S ae reme gt DRA EwAwa iTMERMTeE Figure 13 Basemap with Aerial Photo Layer Turned On Figure 14 Raw IC Shapefile Data 5 5 2 IsValid Procedures Roller compaction data should meet minimum validation criteria to be included in a GIS analysis The data loading routing calculates an IsValid value for each IC record This IsValid data field in combination with a predefined ArcGIS layer style is used to evaluate if a sufficient amount of a coverage has been sampled and that the compaction value is valid 24 To perform the IsValid analysis
54. ing the second row in Table 1 as an example the target value is TV 1 1 280 0 9 C 4 2 4 Creation of GIS Shapefiles 4 2 4 1 Rationale The original concept for GIS implementation was to import the complete geodatabase then perform the database joins necessary to create the desired shapefiles on the fly This concept worked well during early evaluations of the import software geodatabase structure and GIS processes using limited datasets However a production size data set consisting of sixty coverage files and 416 000 compaction measurement records was slow to form the joins necessary and was prone to crashing Investigation revealed that the geodatabase after import was about 150 megabytes and that the shapefile containing the line work representing the compaction measurements was of similar size These usability issues were resolved by creating the desired GIS shapefiles using the import validation software The shapefiles have the database joins in place and may be directly imported in ArcInfo GIS In most instances the geodatabase would no longer be imported into GIS The result is a much faster import process and much smaller GIS files The import validation software prepares both the shapefiles and the geodatabase as illustrated in Figure 4 and consequently database joins not implemented in the shapefile may be accomplished as needed 13 4 2 4 2 Shapefiles A GIS shapefile is a condensed form of the geographical
55. jectID Text 255 RollerID Text 255 CoverageID Long Integer 4 RollerTimestamp Date Time 8 DrumLeftX Double 8 DrumLeftY Double 8 DrumleftZ Double 8 DrumCenterX Double 8 DrumCenterY Double 8 DrumCenterZ Long Integer 4 DrumRightX Double 8 DrumRightY Double 8 DrumRightZ Double 8 CompactionValue Double 8 Compaction ValueQuality Double 8 PercentageOfTargetValue Double 8 MaterialTemperature Double 8 TravelDirection Text 255 VibrationOn Text 5 TravelSpeed Double 8 TravelSpeedUnit Text 255 Frequency Double 8 Mode Text 255 GPSMode Text 255 Peak VerticalAmplitude Double 8 PeakVerticalAcceleration Double 8 PeakVerticalForce Double 8 IsValid Text 255 RCV Percentage Double 8 ErrorMessage Text 255 CreatedDateTimeStamp Date Time 8 UpdatedDateTimeStamp Date Time 8 A 3 Table Contact Name Type Size ID Text 255 Title Text 255 FirstName Text 255 LastName Text 255 Address Text 255 City Text 255 State Text 2 ZIP Text 5 ZIP 4 Text 4 OfficePhone Text 255 HomePhone Text 255 CellPhone Text 255 Email Text 255 CompanyName Text 255 IsActive Yes No 1 Table Coverage Information Name Type Size ID Long Integer 4 ProjectID Text 255 GeoID Text 255 MaterialID Text 255 GISDescription Text 255 ImportFileName Text 255 Description Memo CoverageStartTime Date Time 8 CoverageStopTime Date Time 8
56. ligent compaction The left hand side of Figure 2 depicts a pyramid of data types from the most populated at the bottom to the least populated at the top The data types are l 5 6 Compaction measurement millions of compaction measurements may be taken on a single project Pass As illustrated in Figure 1 many compaction measurements constitute a pass Coverage One to several passes constitute a coverage Layer Typically several passes are necessary to adequately compact a layer of unbound materials Embankment Several layers are necessary to produce an embankment Project A project may have one or more embankments The right hand side of Figure 2 illustrates data types stored in geodatabase tables 1 Compaction measurement A compaction measurement is the primary data type and must be stored in the database Pass In common practice a pass is not identified or tracked separately so pass data is not stored in the database tables Coverage A coverage which may also be the final compaction on a proof layer is tracked by a database table Layer Tracking the coverages that constitute a layer would have some benefit However a practical means to identify and link such coverages was not found so layers were from the database structure Embankment No significant benefits were identified for tracking the layers in an embankment or to identify the layers forming an embankment Project A proj
57. lly about 5 Hz or about 4 inches to 12 inches apart in the direction of compactor travel The data records from intelligent compaction may be represented in several ways depending upon the purpose and analysis method The representations used here are l Line representation The line representation is the principal method of depicting IC data used in this project This representation best fits the nature of an IC measurement since a single measurement results from the drum contact with the ground Depending upon the scale of the view in GIS the line thickness may be varied The line representation is depicted at different scales in Figure 5 Figure 14 and Figure 33 Multi point representation This representation is used to conduct the percent improvement assessment described in Section 5 7 Nine points each having the same compaction value are spaced uniformly across the width of the roller drum Using this representation one point is at the left end another point is at the right end and one point is in the center of the drum Single point representation This representation is used for geostatistical analysis A single point at the center of the drum best represents the data for this type of analysis These three representations are written to separate GIS shapefiles during the import process More information about these shapefiles may be found in Section 4 2 4 2 4 Functionality and Processes 4 1 Introduction Figu
58. lue distinct from other materials Columns include the target value material type optimum moisture content and links to the project calibration area and proof layer Validation criteria table This table is used to store the validation criteria specific to a particular roller The table is constructed to accommodate any sort of validation criteria numeric or text and validation bounds less than less than or equal to equal to greater than not equal to etc One criteria and the associated bound are stored per record in the table The end user enters this information manually and there is no practical limit to the number of criteria Refer to Section 4 2 2 for the criteria for a Caterpillar roller Calibration area table This table is used to store information about calibration areas where target values are determined Figure 3 illustrates the relationship between materials calibration areas and proof layers Other tables Several other tables are used internally to store results necessary to prepare the information contained in shapefiles The geodatabase is units neutral Any consistent set of units may be used A N A vA N vA SN Y ls Figure 3 Relationship between Materials Calibration Areas and Proof Layers 3 2 Geographical Representation Intelligent compaction vibratory rollers generate data at the drum vibration rate typically about 30 Hz However output data records may be provided at a slower rate typica
59. me ProjectDescription SponsorProjectNumber ContractorProjectNumber SponsorSubDivision ProjectStartDate ProjectEndDate ProjectPhases ProjectCoordinateSystem ProjectDatum CreatedDateTimeStamp UpdatedDateTimeStamp RollerID CoveragelD RollerTimestamp DrumLeftX DrumLeftY DrumleftZ DrumCenterX DrumCenterY DrumCenterz DrumRightX DrumRightY DrumRightZ CompactionValue CompactionValueQuality PercentageOfTargetValue MaterialTemperature TravelDirection VibrationOn TravelSpeed TravelSpeedUnit Frequency Mode GPSMode PeakVerticalAmplitude PeakVerticalAcceleration PeakVerticalForce were Table Borrow Material Name Type Size ID Text 255 ProjectID Text 255 Title Text 255 Description Memo TargetCCV Double 8 MaterialT ype Text 255 OptimumMoisture Double 8 CalibrationAreaID Text 255 CreatedDateTimeStamp Date Time 8 UpdatedDateTimeStamp Date Time 8 CalibrationAreaNumber Long Integer 4 ProofLayerNumber Long Integer 4 Table Calibration Area Name Type Size ID Long Integer 4 CalibrationAreaName Text 255 MaterialID Text 255 CreatedDate Date Time 8 CCV Long Integer 4 Station Text 255 CreatedDateTimeStamp Date Time 8 UpdatedDateTimeStamp Date Time 8 A 2 Table Compaction Data Name Type Size RecID Long Integer 4 Pro
60. measures the deflection and estimates the modulus based on the force required to generate a given deflection for the type of granular material Materials Embankment Grading Materials all grading materials placed in the roadbed subgrade as indicated in the plans The embankment is the zone under the base pavement and curb structures bounded by the roadbed slopes shown in the Plan or 1 1 slopes from the shoulder PI point of intersection 1 0 vertical to 1 5 horizontal slopes for fills over 10 meters 30 feet in height Quality Control and Quality Assurance Quality Control a procedure whereby the Contractor develops utilizes and documents Quality Control activities that govern how the embankment is constructed on this project Quality Assurance a procedure where the Engineer monitors the Contractor s Quality Control activities and performs assurance monitoring and or testing for final acceptance of all embankment construction 2 2 5 4 2 2 6 Target Values Intelligent Compaction Compaction Target Value IC CTV The target compaction parameter reading obtained from calibration areas compacted with the IC roller for each type of granular treatment used on the project The IC CTV is determined for specific drum amplitude drum frequency roller speed and roller direction Modulus Test Value At each LWD test location the LWD drop is repeated six 6 times without moving the plate The first second and third drops a
61. mits listed in Table 1 maybe described as follows l Only localized areas may fall below 80 percent of the target value In applying the specification for this project localized areas were assumed to be equivalent to 2 percent ps Only 10 percent of the IC measurements may fall below 90 percent of the target value 12 3 If more than 80 percent of the data is above the target value then the target value must be reconsidered Assuming that the data follows a normal distribution which is true in many cases it is possible to calculate the number of standard deviations below or above the mean target value associated with each specification limits from u au Po b where 4 is the mean or target value o is the standard deviation a is the specification limit 80 percent 90 percent 120 percent is the number of standard deviations below or above the mean Values of f are listed in the third column of Table 1 Negative values indicate the number of standard deviations below the mean and positive values indicate the number of standard deviations above the mean Column 4 of Table 1 lists the coefficient of variation necessary to meet each of the three specification limits Many IC data sets available during this project did not meet even the least stringent of these variability limits For a given set of data working values for the target value may be based on any one of the three specification limits Us
62. ncluding laboratory tests not conducted in the field Key information that may be entered includes test type and date the coverage where the test was conducted either station offset or latitude longitude the test value and if the test passed Details about the geodatabase table and fields for quality assurance tests may found in Item 5 on page 7 above and in Appendix A 4 3 GIS Functionality The detailed procedures for conducting geographic information systems functions are provided in the User s Manual starting on page 17 As noted in the introduction to this section the geographical information system provides the following end user functionality visualizing IC data including percent improvement assessing calibration areas comparing proof coverages against the specifications and conducting geostatistical analysis Visualizing and Verifying Visualizing and verifying IC data is a primary function of the Mn DOT initiative Visualization is essential for assessment of the ongoing compaction work e g accepting coverages and identifying QA test locations for the establishment of target values and investigation of the 14 compaction measurements and for the researcher s development of IC practice The principal building blocks for all GIS functionality are the following 1 2 3 4 4 3 1 Importing base maps and shapefiles see Section 5 5 1 Verifying the Coverages Using IsValid Procedures see Section 5 5 2 Target V
63. ng a field QA tool to developing an office tool for research and development 2 Consistent Terminology 2 1 Background Intelligent compaction terminology in common use by various stakeholders varies and is sometimes in conflict For example some stakeholders used pass to mean a single traverse by the roller while others used pass to mean a complete coverage on a lift This section provides definitions of important intelligent compaction terminology Some of the glossary terms are from Mn DOT s 2007 Intelligent Compaction Specification This section was revised as necessary during the project 2 2 Definitions 2 2 1 l 2 2 2 2 2 3 2 2 4 General Intelligent Compaction IC This process involves measuring and recording the time location and compaction parameters of the granular treatment during the compaction process with a vibratory roller that is equipped with an accelerometer based measuring system and a global positioning system Equipment Intelligent Compaction IC Roller Rollers are vibratory with an accelerometer based measurement system and capable of recording the compaction parameter measurements Portable Light Weight Deflectometer LWD a hand operated device that uses a sensor to measure the deflection of the 200 mm 8 inch diameter flat plate impacted by a falling weight to measure the stiffness of the granular treatment The LWD has one sensor directly below the falling weight The device
64. nger University of Alabama Civil Construction and Environmental Engineering Department December 2009 Published by Minnesota Department of Transportation Research Services Section 395 John Ireland Boulevard MS 330 St Paul MN 55155 1899 This report represents the results of research conducted by the authors and does not necessarily represent the views or policies of the Minnesota Department of Transportation CNA Consulting Engineers CompNet Concepts or the University of Alabama This report does not contain a standard or specified technique The authors the Minnesota Department of Transportation CNA Consulting Engineers CompNet Concepts and the University of Alabama do not endorse products or manufacturers Trade or manufacturers names appear herein solely because they are considered essential to this report Acknowledgments The authors gratefully acknowledge the funding provided by the Minnesota Department of Transportation The technical and contract liaisons and the project Technical Advisory Panel provided invaluable insight into the Department s Intelligent Compaction Implementation Plan and fostered a team based working environment without which this work would not have produced the desired outcomes Table of Contents Ms CER 1 1 1 Purpose and Historical Development esses enne enne 1 2 Consistent Lennon ici lese 2 2 Backoround MR MC CC 2 22 DE e eene da 2 272 General Rem er ee ene e a a x
65. of Compaction Area Represented by a Polyline sss 47 Figure 42 Roller Compaction Value from Coverages 1 and 2 sees 48 Figure 43 100 Point Moving Average of Roller Compaction Value from Coverages 1 amp 2 49 Figure 44 Percent Improvement of Coverage 2 over Coverage 1 sse 49 Figure 45 100 Point Moving Average of Percent Improvement in Compaction Value 50 Figure 46 Thematic Map of Thickness between Coverage 1 and 2 sees 50 Figure 47 Elevation Data for Coverage 1 and 2 e ione dt tt 51 Figure 48 Average Elevation of the Right and Left Side of Drum from Coverage 1 and 2 52 Figure 49 Average Thickness of Soil between Coverage 1 and 2 sess 52 Figure 50 Average Roller Compaction Value and a 10 Point Moving Average 53 Figure 51 Average Percent Improvement Coverage 2 Value over Coverage Value 33 Figure 52 Moment of Improvement Comparing Compaction Values from Two Coverages 54 Executive Summary Intelligent Compaction IC began in Europe in the late 1970s and early 1980s Manufacturers have developed and tested these rollers for 20 years Past and on going research conducted by the Minnesota Department of Transportation Mn DOT has indicated that IC will improve construction quality and efficiencies for the contractors and Mn DOT staff Mn DOT is a n
66. on measured values Current implementation plans focus on the improvement of measured properties from coverage to coverage see Section 5 7 This software and GIS procedures support many types of specifications 4 3 2 Geostatistical Analysis Processes were developed for conducting geostatistical analysis of IC data based on the capabilities of ArcGIS Geostatistical Analyst The procedures include both assessment of spatial continuity and prediction of values at unsampled locations The procedures are described in Section 5 6 16 5 User s Manual 5 1 Introduction This project has produced custom software that performs Setup and General Information Entry see Section 4 2 1 Import and Validation of IC Data see Section 4 2 2 Creation of GIS Shapefiles see Section 4 2 4 and Quality Assurance Test Data Entry see Section 4 2 5 In summary the custom software 1 Read manufacturer s data files and places the data into a Microsoft Access or SQL server database 2 Filters and validates the data Most manufacturers use unknown algorithms for filtering validation interpolation and display With the software developed in this project the filter and validation criteria may be defined and changed as necessary 3 Prepares the data for visualization with GIS software The import application reads the raw data files from the manufacturer writes database tables and creates ESRI shapefile s of the data at the same time This allows
67. ovaag K Morris M Jaselskis E Schaefer V and Cackler E 2006 Field Evaluation of Compaction Monitoring Technology Phase II Final report Iowa DOT Project TR 495 Des Moines IA March White D Thompson M and Vennapusa P 2007 Evaluation of Intelligent Compaction Systems Final report Iowa State University CTRE Project Ames IA January 56 Appendix A Geodatabase Description V DD TestType TestDateTimeStamp CoveragelD QATestValue PassTest Station Offset Latitude Longitude CreatedDateTimeStamp UpdatedDateTimeStamp tblValidationCriteria V DD ProjectID RollerlD CriteriaName CriteriaValue CriteriaType CriteriaDescriptor CriteriaBound InternalField CriteriaRequired CheckSequence Figure A 1 Geodatabase Tables Coveragelnformation y 1D GeoID MaterialID GISDescription ImportFileName Description CoverageStartTime CoverageStopTime CalibrationArea ProofLayer CreatedDateTimeStamp UpdatedDateTimeStamp Approved ApprovedDateTimeStamp y Roller 9 ID V ProjectID RollerSerialNumber RollerModel RollerManufacturer RollerPower Project BorrowMaterial Description TargetCCV MaterialType imumMoisti Y RollerPowerUnits RollerOperatingMass RollerOperatingMassUnits DrumWidth DrumWidthUnits DrumDiameter DrumDiameterUnits DrumMass DrumMassUnits SpecialEquipment RollerDescription RollerNumber CreatedDateTimeStamp UpdatedDateTimeStamp Y 1D ProjectNa
68. paction data line work that represent overlapping coverages If necessary multiple shapefiles can be merged together In ArcToolbox go to Data management tools gt General gt Merge If a point shapefile representing points along the drum was created during data import simply add the point shapefiles for two overlying coverages These shapefiles will contain the compaction value therefore the next step will be to spatially join the two coverages Figure 33 shows two overlapping roller compaction coverages represented by lines first coverage second coverage Figure 33 Overlapping Roller Compaction Coverages Second convert roller compaction lines into points for each overlying coverage Select HawthsTools gt Animal Movements gt Convert Paths To Points lines to points which opens the Convert Paths to Points dialog box shown in Figure 34 Use the RecID as the unique ID field and set the Interval between points to be 1 which is 1 foot This will create approximately 7 8 points per line Turning angles are not needed in the output table so this can be deselected Provide a name for this new point shapefile e g first_coverage_points shp Repeat this step for both roller compaction data layers The new point layers representing the roller compaction data will display on the screen an example of which is shown in Figure 35 41 H Convert Paths To Points Xj Input Input line feature path layer first covera
69. re 4 illustrates the functionality and processes between the roller the geodatabase shapefiles and the geographic information system The roller is the source of the intelligent compaction data In most instances the data from the roller is converted by custom proprietary software into an importable form typically files containing ASCII comma separated values This project has produced custom software that performs Setup and General Information Entry see Section 4 2 1 Import and Validation of IC Data see Section 4 2 2 Creation of GIS Shapefiles see Section 4 2 4 and Quality Assurance Test Data Entry see Section 4 2 5 The geodatabase and shapefiles store and transfer project data and permit long term archiving The geographical information system provides the end user functionality visualizing IC data including percent improvement assessing calibration areas comparing proof coverages against the specifications and conducting geostatistical analysis Procedures for conducting these functions are described in the User s Manual in Section 5 4 0 Custom Software Functionality This section describes the functionality provided by the custom software developed by this project 4 2 1 Setup and General Information Entry For each new project users are required to run the software set up the system for the project and enter project general information Details of running the software may be found in Section 5 3 The information stor
70. re designated as seating drops The LWD Modulus Test Value is the average modulus estimated from the fourth fifth and sixth drops in the testing sequence Portable Light Weight Deflectometer Compaction Target Value LWD CTV The target modulus measurement for each type of embankment grading material determined by LWD testing on the calibration area s constructed on this project Test Pad Layer Calibration Area and Coverage Intelligent Compaction Test Pad a location where the Contractor will demonstrate to the Engineer that the IC roller and the intelligent compaction instrumentation meet all the requirements of this specification The Contractor and Engineer will agree on a location s within the project to construct the IC test pad Layer a quantity of embankment grading material placed spread and compacted before placement of additional material Calibration area locations where the contractor places spreads and compacts embankment construction materials to determine the Intelligent Compaction Target Value IC CTV for each type and or source of material Proof layer a predetermined layer that requires Quality Control measurements by the Contractor and Quality Assurance measurements by the Engineer to ensure compliance with the IC amp LWD Compaction Target Values prior to placing successive lifts Measurement the fundamental IC data representing the conditions sensed by the drum Pass a single traverse along a lay
71. rst roller coverage points to the second coverage points A spatial join will join each point from the first coverage to the nearest point in the second coverage Right click on the second coverage and select Joins and Relates gt Join to open the join dialog box shown in Figure 36 In the top selection box on the Join Data dialog box select Join data from another layer based on spatial location Select the first coverage points layer to join to the second layer and choose the bottom radio button so that each point in the new joined layer will have the attributes from both point layers Click OK to create the output layer this is necessary for transition to the next step below The Join Output layer will have the same number of points and in the same location as the original points in the second coverage layer x Join lets you append additional data to this layer s attribute table so you can for example symbolize the layer s features using this data What do you want to join to this layer Join data from another layer based on spatial location m 1 Choose the layer to join to this layer or load spatial data from disk first coverage points join 2 You are joining Points to Points Select a join feature class above Y ou will be given different options based on geometry types of the source feature class and the join feature class Each point will be given a summary of the numeric attributes of the points in the layer
72. ry expensive subject to change and generally did not provide the functionality required by Mn DOT Hence the Department chose to develop software and processes fitting their specific needs The work included 1 Development of database structures for managing and archiving IC data 2 Software to import and validate IC data populate the database and write geographic information system GIS shapefiles 3 Processes and tools to manage display and evaluate IC data within ArcInfo GIS software The end product of this research is equally suited to compaction of both unbound and bound materials used to construct the entire flexible pavement structure This final report describes the target functionality terminology geodatabase structure import and filtering software and ArcInfo geographic information system GIS platform processes 17 Document Analysis Descriptors 18 Availability Statement Intelligent compaction IC Quality control QC Database No restrictions Document available from Geographic information systems GIS National Technical Information Services Springfield Virginia 22161 19 Security Class this report 20 Security Class this page 21 No of Pages 22 Price Unclassified Unclassified 74 Mn DOT Intelligent Compaction Implementation Plan Procedures to Use and Manage IC Data in Real Time Final Report Prepared by D Lee Petersen CNA Consulting Engineers Jeff Morgan CompNet Concepts Andrew Graetti
73. s little to the understanding of compaction data 38 Variance 900 800 700 600 80 70 60 50 40 30 20 10 Range Vs Calculation Distance Major Range a Minor Range Linear Major Range Linear Minor Range 100 200 300 400 500 600 700 800 900 Calculation Distance lag size X number of lags Figure 30 Major and Minor Range vs the Calculation Distance Variance Vs Calculation Distance e Partial Sill Nugget a Sill Log Partial Sill Log Nugget Log Sill 2 4 100 200 300 400 500 600 700 800 900 Calcualtion Distance lag size X number of lags Figure 31 Partial Sill Nugget and Sill vs Calculation Distance 39 5 6 3 3 Gradient of Compaction Value The variability of compaction value data along or across a coverage can be evaluated by examining the gradient between neighboring points The slope of a kriged surface of compaction value data is shown in Figure 32 In this figure large gradients are shown in red while small gradients are shown in green Black dots in Figure 32 represent the location of the center of the roller drum The linear pattern visible in Figure 32 indicates that compaction data is similar along the direction of travel green and dissimilar from one pass to the next red Figure 32 Kriged Surface Gradient Red Larg
74. s with no data at all these often seem to be related to GPS signal No data IS VALID Valid Invalid Compaction attempted Figure 5 Visualization of Valid IC Compaction Measurements using GIS 4 2 2 4 Implementation of Validation Criteria Validation criteria are entered by the user and are stored in one table of the geodatabase see item 7 on page 7 As noted validation criteria are defined by roller and project There are no practical limitations on the number of criteria that may be applied The end user must define validation criteria prior to importing data The validation process is conducted during import by the custom software in order to relieve pressure on the GIS engine 4 2 3 Statistical Calculations Mn DOT IC specifications in use during the start of this project established three limits with respect to the desired target value The first column of Table 1 lists the specification limit in terms of target value and the second column lists the allowable percentage Table 1 Relationship between Specification Limits and Allowable Percentages Specification Allowable Percentage Number of Standard Coefficient of Variation Limit Deviations Below or Required to Meet Above the Mean 8 Specification Limit 80 of TV localized areas 2 2 055 0 078 assumed 9096 of TV 1096 1 280 0 097 12096 of TV 8096 30 845 0 237 The three specification li
75. shapefiles will be the concatenation of the CoverageID date and time If you chose to create additional files the name will be prefixed with either a C or a P which stands for Center and Points respectively The date and time come from the first record in the dataset 21 Name Date modified Type Size P3 9 13 2007 7 19 52 dbf 5 7 2009 9 36 AM ACT Database 4 076 P3 9 13 2007 7 19 52 shp 5 7 2009 9 36 AM SHP File 3 660 P3 9 13 2007 7 19 52 shx 5 7 2009 9 36 AM SHX File 666 C3 9 13 2007 7 19 52 dbf 5 7 2009 9 33 AM ACT Database 454 C3 9 13 2007 7 19 52 shp 5 7 2009 9 33 AM SHP File 407 C3 9 13 2007 7 19 52 shx 5 7 2009 9 33 AM SHX File 75 83 9 13 2007 7 19 52 dbf 5 7 2009 9 32 AM ACT Database 5 389 3 9 13 2007 7 19 52 shp 5 7 2009 9 32 AM SHP File 814 3 9 13 2007 7 19 52 shx 5 7 2009 9 32 AM SHX File 75 m H Figure 11 Typical Shapefile Names In Figure 11 you can see that there are three shapefiles created for this coverage The coverage is number 3 and it was started on 9 13 2007 at 7 19 52 am The files that start with a P is just like the file that is the line data file except that it has 9 points along the drum rather than the line itself The file that starts with the C has just one point for each line at the center of the drum 5 5 Basic ArcInfo Procedures Both a database and coverage shapefiles are created during the IC data loading process These data sources can be added to a GIS for display an
76. t Join and Relate gt Join In the top selection box on the Join Data dialog box select Join data from another layer based on spatial location Select the CenterLine Points layer to join to the roller compaction points layer Choose the bottom radio button so that each point in the new layer will have the attributes from both point layers and name the new joined layer CenterLine Join The Join Output layer will have the same number of points and in the same location as the original points in the roller compaction layer Plot the roller compaction data vs the StepID column The StepID column is the station distance along the centerline of the road in feet Using the StepID as the X coord and any other column as the Y coord provides the basis for the roadway centerline trends This data can be imported into Excel for analysis by opening the dbase file associated with the Centerline Join point data Example roadway trend analyses are presented in the following figures Figure 42 raw roller compaction values from subsequent lifts Figure 43 100 point spatial moving average of compaction values Figure 44 Percent improvement in roller compaction values for subsequent lifts and Figure 45 100 point spatial moving average of percent improvement of roller compaction value from subsequent lifts 40 y Coverage 2 35 Coverage 1 A 0 50 100 150 200 250 300 Station ft Figure 42 Roller Compaction
77. tatistics eese 45 5 7 4 3 Roadway Center Line Trends 2eieoeeeetseo eese etie eet boten ert nee e 46 5 1 44 Lift Thickness Analysis etre tite ener lender enda ao corea 50 5 7 4 5 Average Variation along Centerline esses 51 5 7 4 6 Moment of ImprOVvetmoent oce sepan tere Gave ener e eoe Goes 53 wb un ZEN 54 A A EMO 35 Appendix A Geodatabase Tables List of Tables Table 1 Relationship between Specification Limits and Allowable Percentages Table 2 Influence of Lag Size and Number of Lags on Geostatistical Parameters List of Figures Figure 1 Measurement Pass Coverage Terminology eese 4 Figure 2 Relationship between Possible IC Data Types and Database Tables 6 Figure 3 Relationship between Materials Calibration Areas and Proof Layers 8 Figure 4 Data Flow and Principal Processes for Assessing Intelligent Compaction Data 10 Figure 5 Visualization of Valid IC Compaction Measurements using GIS 12 Figure 6 Import Main Screen as anes Sa teeta eee ada cna it eo repetenda an dtet oes 18 Figure 7 File Explorer Dialog BOx ied deu eet Aa ede i AS 18 Figure Coefficient of Variation Display 4er te eoe ied tento tea twenties Pelo seen aq raid 19 Figure 9 Target Value Prompt eet rei rr ae C RERRAEN daria 20 Figure 10 Project Data Entry
78. tential changes in compaction characteristics Each of these limits is different but each requires similar assessments to be performed The consultant team developed the following processes necessary to compare proof layer measurements with the specifications l Ninety percent of measurements are 90 percent of the target value This specification is written in terms of the number of measurements rather than the area covered by those measurements Hence the spatial nature of the measurements is not important and the assessment may be performed by creating a histogram of measurements and applying the 90 percent of 90 percent criterion Localized areas less than 80 percent of the target value This specification is spatial since it addresses localized areas The inspector and others will require a tool that identifies the size of areas that have measurements less than 80 percent of the target value The tool will also sum the size of all areas in the proof layer All assessments will be done on measured values the same as used to produce surface covering documentation Significant areas more than 120 percent of the target value This specification is analogous to item 2 above simply reversed so that the requirement is for identifying the 15 size of areas that have measurements greater than 120 percent of the target value Like item 2 the tool will also sum the size of all areas in the proof layer and assessments will be done
79. the 45 line indicate an improvement in compaction value while points below the line indicate a decrease in the 33 compaction value associated with coverage 2 The greater the distance away from the 45 line the larger the magnitude in the difference in roller compaction value between coverage 1 and 2 40 35 30 25 20 15 RCV Coverage 2 10 0 10 20 30 40 RCV Coverage 1 Figure 52 Moment of Improvement Comparing Compaction Values from Two Coverages 5 7 5 Summary The Improvement Analysis is based on comparing roller compaction data from an overlying coverage to an underlying coverage at the same locations This was accomplished through a GIS spatial join of point data that represent each compaction value record from each coverage From this joined coverage data a comparison of roller compaction data from subsequent coverages can be made Locations where a decrease in compaction value occurred can be identified and overall coverage improvement statistics can be calculated Through the creation of a centerline along the compacted area a spatial join can be made that provides a roadway station coordinate for every roller compaction data point This station can be used to plot summarize and analyze roller compaction data along the roadway 54 6 References Adam D 1997 Continuous compaction control CCC with vibratory rollers Proceedings of the Ist Australia New Zealand Conference on
80. with the calculated Lag Size and Number of Lags at the bottom of the control panel the model parameters Major Range Partial Sill and Nugget are reported Below the Semivariogram on the left side of Figure 28 is a Semivariogram Covariance Surface that shows similar data blues and dissimilar data reds It can be seen in Semivariogram Covariance Surface that the compaction value data is anisotropic having strong similarity or correlation along the direction of roller travel and weak similarity or correlation across the direction of travel Anisotropy can be selected on the Semivariogram Covariance Modeling dialog box below the Major Range on right side of dialog box shown in Figure 29 With Anisotropy checked on both the major and minor ranges are calculated for the model Multiple yellow lines are shown on the Semivariogram representing the different ranges at different angles across the data The minor 36 range is the steepest yellow line while the major range is the shallowest yellow line The Semivariogram Surface lower left of Figure 24 shows the range as an ellipse in light blue Geostatistical Wizard Step 2 of 4 Semivariogram Covariance Modeling 2 xl View Models Semivariogram Covariance V Model 1 Ir Model 2 Model 3 Major Range E e 4 T 10 i i qu i a Uu a 97 546 T55 l 2 Keine iced n Po 9 Tetraspherical i i 9908 0 Pentaspherical
81. x Select the Definition Query tab and then click the Query Builder Click the Load button and navigate to the GIS Files directory Select the Is Valid Expression epx and click Open Click OK on the remaining dialog boxes and the Is Valid query will only display valid IC data as shown in Figure 16 Comparing Figure 16 to Figure 15 it can be seen that the invalid data at the ends of the coverage have been removed the marginally invalid data lower left have been removed and the invalid data in the center of Figure 15 red has valid IC data below which is now showing as valid green data in Figure 16 Figure 15 and Figure 16 should be used to determine if a sufficient amount of a coverage has been measured with valid IC data In areas where large holes exist the in the IC data the contractor should be asked to recompact that area and record valid IC data 25 ISVALIDREC Valid Invalid Marginally Invalid Figure 16 Map Showing Only Valid IC Data 5 5 3 Target Value Classification and Coloring IC Data To view and analyze compaction value data compared to the target value entered during data loading a predefined symbology has been created and is stored in the GIS Files directory Before performing this analysis the user should eliminate invalid IC data by following the procedures in Section 5 5 2 Right click on the shapefile of interest in the table of contents and click Properties This opens the Layer Properties dialog box

Introduction & purpose - Minnesota Local Road Research Board

Contents

Download Pdf Manuals

Related Search

Related Contents

Introduction &amp; purpose - Minnesota Local Road Research Board

Contents

Download Pdf Manuals

Related Search

Related Contents

Introduction & purpose - Minnesota Local Road Research Board