Evaluation and Local Calibration of MEPDG EICM Model Using ...
Transcript of Evaluation and Local Calibration of MEPDG EICM Model Using ...
Conclusions• ACsurfaceproducedmarkedlyhighersurfacetemperatures
thanPCC
• Despitehighersurfacetemperatures,anACoverlayreduced
thermalgradientsinaPCClayer
• EICMsimulationsquantitativelyreproducedtheinsulat-
ingeffect,whichmaycontributetolongevityandimproved
PCCperformance
• Thermalpropertyinputscanhavesignificanteffectonpave-
mentperformancepredictions,andtheseinputsshouldbe
adjustedaspartofalocalcalibrationprocess
• EICMsimulationsproducedtemperaturedistributions
smallerthanthemeasureddistributionswhentheMEPDG
defaultkvaluewasused(1.25BTU/hr-ft-˚F)
• Severalkvaluesweretested,thebestagreementbetween
measuredandmodeleddatawask=0.94BTU/hr-ft-˚F
Acknowledgements• FHWATPF-5(149)Partners
-FederalHighwayAdministration
-MN,CA,andWADepts.ofTransportation
-MinnesotaLocalRoadResearchBoard
• Prof.RandalBarnes,UniversityofMinnesota
MEPDG and EICM Sensitivity to Thermal Conductivity• Measuredandmodeleddataplottedonacumulativefrequencydistributionchart
• ResultsindicatedtheEICMunderestimatedthetemperaturedistributionsina
PCCpavementwithdefaultkvalue,k=1.25BTU/hr-ft-˚F
-Unnecessarilyhighkvaluemaycontributeinparttounderestimation
• k=0.94BTU/hr-ft-˚Fproducedtheclosestmatchtomeasuredvalues
Figure 8a. July, k = 1.25 BTU / hr-ft-°F Figure 8b. July, k = 0.94 BTU / hr-ft -°F
Figure 9a. March, k = 1.25 BTU / hr-ft-°F Figure 9b March, k = 0.94 BTU / hr-ft-°F
• Thermalconductivity(k-value)affectsthetemperatureprofilethroughalayer
• Inputk-valueinfluencesperformancepredictions,especiallyinPCCslabs
EICM Predictions• EICMqualitativelypredictsthermalinsulatingeffect
• Visibleincumulativefrequencydistributionscomparingmeasuredandmodeleddata
• SupportthehypothesisthatanAClayersignificantlyaltersPCClayertemperaturedistributions
Objectives• Evaluatethemodelingofthermalbehaviorinconcreteand
compositepavementsbytheEnhancedIntegratedClimatic
Model(EICM)
• InvestigatebenefitsofthinACoverlaysonthermalcharac-
teristicsofPCCslabsusingMnROADData
• ValidateEICMpredictionsofthermalgradientsthrough
PCCslabs
• InvestigatetheeffectofMEPDGuserinputsforthermal
conductivityofthePCC
MnROAD DataData Collection
• Oneyearofhourlydatafromthermocouple“trees”inPCC
andAC/PCC
• OnlydifferenceinpavementswasthepresenceofAClayer
Figure 1. MnROAD Cells Used in Analysis
Data Filtering
• Subjecteddatatovariousteststoidentifymissingandinsuf-
ficientdata,sensoroutliers,datasubsetoutliers,etc.
Figure 2. Example of Filtered Data
• Suspectdatawereflaggedandexcludedfromanalysis
• Screeningallowedresearcherstomakecomparisonswith
confidence
Evaluation and Local Calibration of MEPDG EICM Model Using MnROAD DataLuke Johanneck, Derek Tompkins, Timothy Clyne, and Lev Khazanovich
MnROAD Data Analysis – Thermal Gradients in PCC LayerEffect of Albedo
• AChashighersurfacetemperature
• GreateroveralltemperaturegradientsinAC/PCCstructure
Thermal Gradients in PCC Layer
• LargerthermalgradientsincompositesystemdoesnotequatetolargerthermalgradientinPCClayer
• AClayerprovidesaninsulatingeffect
• Gradientsmostpronouncedinsummer(seefigureof2-weekdetail)
Figure 4. Thermal Gradients – Relative DepthsFigure 3. Surface Temperature – Albedo Effect
Figure 5. Thermal Gradients – Relative Depths Figure 6. Thermal Gradients – Absolute Depths
Sensor Locations
• Thermocouplesensordepthswererelativeto5”PCCslab
• Topsensorswere0.5”fromtopofPCCsurface
• Bottomsensorswere1”abovebottomofPCClayer
• Identicalverticaldistanceof3.5”
• Asimilarcomparison(Fig.6)wasdoneusingsensorsatsimilarabsolutedepthstoensurethatthediffer-
enceinthermalgradientswasduetoAClayer,andnotsensorpositioning
Legend
Red–AC/PCC
Blue–PCC
Figure 7. Effect of k on Transverse Cracking
Figure 10. Thermal Gradients – Measured vs. Modeled