Non-destructive Tests In Roads And Airfields A Study Of .

2y ago
13 Views
2 Downloads
246.06 KB
10 Pages
Last View : 18d ago
Last Download : 2m ago
Upload by : Evelyn Loftin
Transcription

Non-destructive tests in roads and airfieldsA study of the Falling Weight DeflectometerEdgar Alexandre Chong CardosoInstituto Superior Técnico, Lisbon, PortugalOctober 2017AbstractRoad infrastructure is a high value asset in the development of modern society where itsperceived quality translates into a fundamental role in security, economy, competitivity andsustainability of the free flow of goods and merchandise. The gradual degradation of that qualitythrough time should be evaluated in such a manner that maintenance and rehabilitation effortscan be timely planned and carried out to maintain its specified minimum quality requirements. Inpavement condition assessment, there are several parameters that gauge pavement quality.For structural surveys, the Falling Weight Deflectometer (FWD), is the main non-destructivetesting equipment used to assess the bearing capacity of road and airfield pavements. Thistest’s results are very relevant in several contexts, for example, a survey for bearing capacity inexisting road or airfield pavements requiring rehabilitation intervention.The present dissertation’s objective is to assess the precision and uncertainty performance inmeasuring deflection and to analyze its influence in the quality of results from the testingcampaign, therefore assessing the structural capacity of existing pavements (backanalysis), inview to evaluate the structural quality and support a rehabilitation project.The adopted methodology consisted in a proficiency test scheme (PTS) field test compliant withISO/IEC 17043 featuring a fleet of three FWD from different manufactures and owned byportuguese operators. The obtained deflection data was firstly processed for repeatability andreproducibility, and afterwards analyzed for uncertainty quantification. Lastly, the resulting datawas used for a sensitivity analysis featuring the uncertainty of the measured deflection influenceon the mechanical properties (elastic moduli) estimated from the field survey (backanalysis) onflexible pavements.The experimental research results confirmed a satisfactory repeatability of deflectionmeasurements. In contrast, the reproducibility is difficult to achieve in most cases.Consequently, the uncertainty levels revealed to be high. Uncertainty and precision revealed tobe dependent of pavement type and deflection magnitude. Uncertainty presented high valuesfor flexible pavement and for high deflections. Regarding to the sensitivity analysis on theuncertainty’s influence on the FWD results interpretation, it was concluded that the flexiblepavements presented higher sensibility to uncertainty mainly when gauging for stiffness on thefoundation layers.1

Keywords: falling weight deflectometer, pavement, repeatability, reproducibility, uncertainty,backcalculation.1IntroductionLooking at the European Union statistic data, the Trans-European networks in transport (TENT) plays a vital role to promote people and goods circulations between member states(Eurostat, 2014). By promoting business and easy people circulation, transportation strategiesare an effective way to tackle inclusion of state members and its citizens. Eurostat data referringto modal split of transportation in EU (Eurostat, 2017) shows that road transport is still by far themost common, representing about 75% of total tonne-kilometers of freight transported. Dataforecasts expect a continuing rise of freight transport by road in the foreseeable future.Consequently, new and existing infrastructure assets can benefit from planned maintenance toprolong its life span and reduce the involved financial costs. In a report commended by theEuropean Commission (Steer Davies Gleave, 2009), EU countries invested in total 859 billionin its transport infrastructure sector between 2000 and 2006. A significant portion of the budgetwas used towards road maintenance to keep existing infrastructures at an acceptable level ofservice. This sector has proven its significance given the large sum invested thus incentivizingpavement engineering to continually improve (COST, 1997).Maintenance and rehabilitation (M&R) requires both minimizing administration and user costswhile still maintaining infrastructures at high level of service (Meneses & Ferreira, 2012). Tomanage pavements at network level, administration rely on pavement management systems(PMS) which aggregate road condition data by road sections. This information is in turnanalyzed to clearly prioritize interventions to the most critical road sections (Fwa, et al, 2000).Road and airfield administrations are the main clients for the services provided by pavementcondition assessment companies. These survey proceedings are regulated by internationalstandards (ASTM, 2008, 2009) using equipment capable of measuring and recording pavementparameters to assessment its condition. One of the most used equipment today is the FallingWeight Deflectometer (FWD), a stationary impulse load deflectometer which will be studied inthe following sections of present dissertation.Although equipment manufacturers guarantee high reliability and repeatability levels throughperiodic calibration, generally, equivalent models from different manufacturers are less likely toreproduce each other’s measurements. Several authors research (Garg, 2002; Murphy, 1998;Rocha et al, 2004) mention the repeatability and reproducibility issues associated with the FWDwhich should be taken in consideration and carefully assessed in practice. To empoweradministration decision makers with informed decisions while executing pavement surveys, it isnecessary to experimentally analyze the actual reliability level of existing FWD fleets in currentavailable service providers and thus clearly quantifying existing differences.This dissertation aims to investigate the precision performance of a FWD fleet under acontrolled environment, and mainly to quantify the level of uncertainty in deflectionmeasurements which ultimately influence the quality of the backcalculation process. It is crucialfor the administration to have a good understanding of the uncertainty involved in this processwhich may lead to rehabilitation project designs that may prove to be ineffective and financiallyinefficient.2

22.1Falling Weight DeflectometerOperation principleThe basic principle behind the deflectometer is a mechanism of hydraulic lifters that elevate apredetermined mass of weights to a certain height then drops. This mass generates a force onimpact thought a set of rubber bumpers producing a load cycle equivalent to a vehicle wheel innormal traffic speeds. The FWD is highly mobile when compared to other type of static androlling wheel equipment giving administration entities the flexibility necessary to perform surveysin a broad area in limited time. Given its operation principles and a computerized user interface,the FWD has been recognized as the preferred method to perform deflection measurements.Prior to testing the pavement’s load-carrying capacity it is necessary to specify parameters onwhich the test is to be conducted, essentially to define the test protocol: test location and itsstructural constitution, load force values, loading plate diameter, geophone positions andpavement surface temperature.Load forceThe load force necessary for a pavement test depends mainly on the pavement type (flexible orrigid). Higher load produces more impulse on the pavement and thus higher deflection readings.Depending on the pavement constitution, a rigid pavement will require higher load to generate adeflection value within the system’s geophone resolution and range.Dampening systemThe weight dropping mechanism generates a pulse of force that is transmitted to the groundthrough a dampening system. This load pulse is comparable to the action of a wheel axle on thepavement. The load pulse transmitted to the pavement is shaped as a half-sine curve similar tothe actual impulse produced by a wheel axle. During the development of the FWD system, therubber bumpers acting as dampers for the falling weight plates were identified as determinant tothe force curve shape generated (Bohn, 1989). For these reason, the configuration of weightplates and the number rubber buffer in the system may significantly change the shape of theload pulse generated and, consequently, the value of load pulse time and the resultingdeflections.Load pulseLoad pulse is the time that the FWD takes to fully deploy the impulse load on to the pavement.This force cycle is configured to be shaped as a sine curve and the duration and magnitude ofthe force applied by the FWD is representative of the load pulse that would be induced by avehicle in movement (Garg, 2002). The load pulse is a parameter measured in milliseconds andcan vary between 25 and 60ms, depending on what kind of wheel axle is being simulated. Pulsetime is particularly important to control in multi-layered pavements that are flexible, cohesivesoils or saturated soils, for it may influence to some degree the obtained deflectionmeasurements. The dampening system comprises from rubber materials which means that itsbehavior change depending on the conditions tested on: temperature, load level and even thebuffer physical shape change the spring effect “constant”.GeophonesThe FWD can have up to 9 geophones attached to the trailer. These sensors are evenly spacedand directed radially away from the center of impact. The geophones capture minute surfacedisplacement (analog signals) and interpret them as electronic signals enabling computers torecord even small amount of surface movement. The resulting array of deflection measurement3

from the impact center produce a graph named the Deflection Basin which helps visualize thestructural capacity of the layers below surface.2.2FWD AccuracyAlthough FWD are commonly requested by administrations for routine campaigns, severalauthors (Van Gurp, 1991 and Murphy, 1998) have given evidence of lack of reproducibilitybetween a FWD fleets. Researchers Rocha et al (2004) presented a thorough literature reviewon the accuracy and precision of FWD. The main possible sources of uncertainty in FWDmeasurements most commonly reported in the literature are related to its buffers and thepavement stiffness. The shape, size, age and stiffness of rubber buffers impact the peak load,the rise time and the load pulse shape, and in consequence the magnitude of the deflections(Chen et al, 1999; Lukanen, 1992). Having calibrated load cells and calibrated deflection sensordoes not offset the different equipment characteristics, such as, rubber buffers shape andhardness, the thickness and quality of rubber pad under the load plate, the type of deflectionsensor, sensor positioning in the frame, and other factors that impact the load pulse shapes anddeflection readings. FWD time histories produced by one equipment are different for anotherFWD, producing different peak force values. This implies that data collected from different FWDare not intercomparable, even in the case of fully calibrated equipment.3Proficiency testA fleet of three FWD joined the Proficiency Test Scheme compliant with ISO/IEC 17043 (ISO,2010) (Table 1). The test site pavement layer composition was obtained with a coring sampleand from archived design plans (Table 2). A test protocol was followed requiring several roundsof drop tests to be performed (ASTM, 2009), with only the last three drops were considered forlater data analysis (ASTM, 2008).Table 1 - Main specifications of FWD equipmentManufacturerModelYear of acquisitionLoad range [kN]Load pulse time[milliseconds]Diameter of loadplate [cm]Type of deflectionsensorsDeflection sensorrange [ m]Relative accuracy ofdeflection sensors(*) Not availableCarl BroCarl Bro PRI 2100(*)7-250KUABKUAB 240200430– 240DynatestDynatest 800220027-12020-303020-3030 and 4530 and 4530 and 45SeismometersGeophonesGeophones2.2(*)(*)1 m 2%(*)2 m 2%Table 2 – Characteristics of the test sitesSitePavement1Asphalt seSubgrade(1)4- Unbound granular material;(2)- UnknownMaterialAsphalt concreteUGM (1)Soil-cementSandy soilConcrete slabPaving stonesSoil-cementSandy soilThickness [cm]52015152012(2)(2)

Table 3 – Test protocolSiteLoad peak [kN]165, 90290Load plate diameter [cm]0, 30, 45, 60, 90, 120, 150, 18030Distance [mm]300 600 900 1200 1500 1800000100100200200300300400500600700FWD 1800FWD 2900FWD 31 000Deflection [ m]Deflection [ m]0Deflection sensors distance [cm]400500600700FWD 1800FWD 2900FWD 31 000a) 65kN, flexible pavement0Distance [mm]300 600 900 1200 1500 18000b) 90kN, flexible pavementPavement deflection [ m]7503 Weight plates5 Weight plates7 Weight plates100200Deflection [ m]Distance [mm]300 600 900 1200 1500 1800725300400500700600700FWD 1800FWD 2900FWD 31 000c) 90kN, rigid pavement67515,020,025,0Load pulse [ms]30,0d) Load pulse in relation to deflection valuesFigure 1 - Deflection charts and Load Pulse influence in deflection measurementsThe deflection results were plotted in graphs representing complete deflection basins for eachFWD measurement in the same marked location. One of the equipment (FWD 1) showed clearbias by recording values consistently shifted by a constant factor in relation to the rest of thetest sample. It was later investigated and discovered that technicians had preset the computersoftware to allow smoothing filtering of the pulse load signals at a frequency different from therecommended in the manufacturer’s manual (150 Hz instead of 60 Hz).5

4Data Analysis4.1Repeatability and reproducibilityAnalyzing Figure 2, repeatability limit for both N 3 and N 2 seems to scatter uniformly settlearound r 4 and r 2, respectively for Site 1 (flexible pavement) and Site 2 (rigid pavement).Figure 3 presents the reproducibility limits with trend lines and their respective expressions. ForSite 1, being a flexible pavement, the limit for N 3 increase rapidly when deflection values alsoincrease towards peak values in the center of test impact. Serving as an indicative example, fora given deflection D 300 m, R 169. With the N 2 scenario, without FWD 1 contribution,reproducibility limits clearly drop to acceptable values and trend line stay almost flat in the entiredeflection range. In the same line of example as above, for a D 300 m, limit is returned asR 13. Same analysis is valid for Site 2 reproducibility limits, although predictably, for rigidpavements, the trend line presents a smaller slope, for given D 300 m, R 93 (N 3) and R 25(N 2).N 3N 2303025252020Repeatability rRepeatability rN 215r 410N 31510r 25500020040060080010000100Deflection D [ m]200300400500Deflection D [ m]a) Site 1b) Site 2Figure 2 – Repeatability limits for deflection measurementN 2N 3N 2450450400400350Reproductibility R350Reproductibility RN 3y 1.702D0.8064R² 0.99300250200150y 0.1163D0.8257R² 0.37100300250200y 0.9251D0.8081R² 0.99150100y 0.7906D0.6042R² 0.905050000200400600Deflection D [ m]a) Site 180010000100200300Deflection D [ m]c) Site 2Figure 3 - Reproducibility limits for deflection measurements6400500

4.2Uncertainty assessmentWhen comparing with the N 3 case, the critical values are significantly smaller for N 2,representing the case in which a FWD fleet presented satisfactory repeatability andreproducibility. Figure 4 presents the graph for the deflection mean of the means values as afunction of their respective critical values. The model indicates a good fit (R2 0.99) mainly fromN 3 cases. It is noticeable a positive trend for which the critical value increases with thedeflection magnitude. To express this relationship resulted from the PTS experiment, a trendline was plotted and its governing equation will serve as the model behavior rule to calculatecritical values for given any deflection.300N 280N 3N 2N 370250Uncertainty U [ m]Uncertainty U [ m]60200y 1.171D0.8075R² 0.99150100y 0.0015D1.50250y 0.6397D0.8084R² 0.99y 0.605D0.6069R² 0.90403020R² 0.4450100005001000Deflection D [ m]0200400Deflection D [ m]b) Site 2a) Site 1Figure 4 - Uncertainty of deflection measurements4.3Sensitivity analysisThe aim is to extrapolate the uncertainty data and apply it in a sensitivity analysis onstandardized flexible pavements suggested in the portuguese pavement design manual, knownas MACOPAV (JAE, 1995). These pavement models are used in real world practice bypavement designers and so the results should constitute a close approximation to a real-worldapplication. As the scope of this study is only flexible pavements, the uncertainty critical valueswill be obtained with equation 1:ܻ 1.171 ܦ .଼ ହ(1)The analysis makes use of standardized flexible pavements models suggested by MACOPAVdesign manuals. Each pavement model has attributed reference moduli values for each type oflayer and finally, through BISAR multilayer elastic linear pavement design software (Shell,1995), the models deflection basin were calculated.Table 4 resumes part of the backcalculated elastic moduli for the class F3 subgrade pavementssubjected to 65 kN and 90 kN surface loads. It shows in a clear manner that the MACOPAVmodels presented convergent estimation of layer moduli for either 65 kN or 90 kN. In this way,the FWD fleet measurement uncertainty is now represented in the form of layer moduli, with thehigher and the lower limits of the interval. In this form, it is possible to assess the range of7

moduli uncertainty when performing field surveys and even evaluate financially the sameuncertainty although this study is outside the scope of the present dissertation.Table 4 – Interval of pavement layer moduli, F3 subgrade class, 65 kN and 90 kNAsphalt bound base, T6, 65 kNRMS 3.6%Emax(1)4700300175ReferenceEAsphalt bound base, T1, 65 kNRMS 0.9%(2)Emin4000200100RMS 3.6%(3)Emax3000160704500400200Asphalt bound base, T6, 90 kNRMS enceE(2)4000200100RMS 2.3%Emin (3)300014070Asphalt bound base, T1, 90 kNRMS 2.7%(2)(1)RMS 200100RMS 1.2%Emin (3)350015075Units in MPa; (1) – Layer modulus for stiffer limit layer composition; (2) – Layer modulus initially assumed;– Layer modulus for softer limit layer composition.(3)The variation in moduli values necessary to adjust the initial reference moduli to the criticalvalues boundaries were expressed in function to the asphalt concrete material content inrelation to the pavement total thickness (in percentage). From top to bottom, Top layer - Base Sub-base - Subgrade, both the granular material layer required more adjustment, reaching ashigh as 80%, leading to stiffer layers for pavements with higher content of asphalt concrete(more than 40%). Pavement models with higher AC percentage present higher sensitivity toFWD uncertainties4.4Test site backcalculationRegarding test site 1, the PTS results confirmed the findings included in chapter 4.3. Thebackcalculation of deflection measurements resulted in three different pavement designs, eachinfluenced by the uncertainty already mentioned. In the event of a rehabilitation project, theseresults could directly affect the possible project solutions, namely when determining thenecessary thickness of the overlay design, poising financial impacts on the project’s final cost.Table 5 and Table 6 show deflection plot charts with BC deflection curve layered on top. RMSE obtainedwere between 6.0 to 8.4%. High RMSE values were obtained representing that the pavement model mightnot exactly match the actual pavement.Table 5 – Backcalculation results for Site 1, 65 eedmoduli40009009005030FWD 1FWD y soil20090Sandy soil4050RMS%7.46.5AC – Asphalt concrete; UGM - Unbound granular material; Root Mean Square Error (%)8FWD 3700050090075406.6

Table 6 – Backcalculation results for Site 1, 90 kNThicknessSeedFWD 1FWD 0Soil-cement15

pavement condition assessment, there are several parameters that gauge pavement quality. For structural surveys, the Falling Weight Deflectometer (FWD), is the main non-destructive testing equipment used to assess the bearing

Related Documents:

the development of an on-site non-destructive testing technique to assess the performance of sealants on building facades. Tests were carried out on various . Non-destructive Non-destructive Vertical Horizontal simulated y 1.0858x - 10.229 0.9702 Non-destructive Semi-destructive Vertical - simulated y 0.6159x 33.565 0.8375 .

Roads in Malaysia are classified into two broad categories, namely Federal Roads and State Roads. 2.1 Federal Roads Federal roads are all roads declared under the Federal Roads Ordinace (1959) and the major interurban roads joining the state capitals and roads leading to points of entry to and exit from the country.

surprise that there are differences between outcomes of destructive and non-destructive testing as well. It is important to keep the differences in mind: Table 1: Important properties compared Destructive testing (Peel decohesion) Non-destructive testing (PAUT) Tests the strength Checks for deviations Only small part of the joint is investigated.

The rebound (Schmitz) hammer is one of the most popular non-destructive testing (NDT) methods used to test the strength of concrete. This is due to its relatively low cost and simplicity in use (Luke, 2012). Although the non-destructive testing (NDT) results are much quicker compared to the destructive methods, they are more of an approximation .

1 DESIGN STANDARDS FOR INDUSTRIAL ROADS 1.1 Design Standards 1.1.0 Industrial Estate Roads have been categorised as follows: i. Major Industrial Roads (Major IR) ii. Minor Industrial Roads (Minor IR) In general only culs-de-sac of less than 200m in length should be considered as Minor Industrial Roads with all others being Major Industrial Roads.

EN 571-1, Non-destructive testing - Penetrant testing - Part 1: General principles. EN 10204, Metallic products - Types of inspection documents. prEN ISO 3059, Non-destructive testing - Penetrant testing and magnetic particle testing - Viewing conditions. EN ISO 3452-3, Non-destructive testing - Penetrant testing - Part 3: Reference test blocks.

NFP121 5 Sommaire 1) Tests et tests unitaires - Outil : junit www.junit.org une présentation Tests d'une application - Une pile et son IHM Tests unitaires de la pile Tests de plusieurs implémentations de piles Tests d'une IHM Tests de sources java Invariant et fonction d'abstraction comme tests - Tests en boîte noire - Tests en boîte blanche

reliable type of testing, while non-destructive testing is a wide group of analysis techniques used in science and industry to evaluate the properties of a material, component or system without causing damage. Non-destructive testing can be applied to both old and new structures (Jedidi Malek, 2014).