Testing Post-Tensioned Steel Trunnion Rods for the U. S. Army ...

Testing Post-Tensioned Steel Trunnion Rodsfor the U. S. Army Corps of EngineersMark A. Cesare, , PhDmcesare@fdh-inc.comJ. Darrin Holt, PhD, PEholt@fdh-inc.comFDH Engineering, Inc. (919) 755 1012Raleigh, NCSt. Louis, MOBaton Rouge, LA

Trunnion Anchor Rods

The Problem• Some Catastrophic Rod Failures• Loss of Pre-Stressing (grip nut slippage,relaxation, etc.)• Unknown Initial / Existing Tensions• Typical Solution = Mechanical Lift-Off• New Solution = Nondestructive Testing (NDT)by Dispersive Wave Propagation (DWP)

Approach to the Problem• Feasibility investigation at West Point: Canthis even be done?• Lab models and full-scale trunnionassembly prototype.• Wide spread testing.• Mechanical lift-off.• Calibration / Validation• Dispersive Tension Equation (DTE).

Background on the Testing Method• Dispersive Wave Methodsas used for foundation testing• FHWA / NCDOT Researchin 1989 for Bridge Piles• Used in 50 States, PR,USVI, El Salvador• Used by several States andmany private corporations

The “Idea”Dispersive Flexural Wave MotionStrike with a HammerThe WaveGrip NutEmbedded EndFlexural wave velocities and response depend upon:MaterialCross-SectionTension

The Nondestructive Field Testing“Light” impacts create the dispersive waves.Wave motion captured and recorded.Data analyzed for response.Tension determined from response curves.

Field Testing

Accessibility Issues

Modeling the Steel Covers• Built full size cover boxwith PVC for rods• Helped develop customstrikers (“Hammer on astick”)• Developed and practicedtesting procedures

Tackling AccessibilityEach rod had to be measured& “pinged”– non-trivial problem thru theport hole openings– Remote camera had to beused

Trunnion Assembly Prototype• Reinforced concrete withtubes for rods and testequipment.• Designed and built at FDH

Allows for modeling ofmany damconfigurations.Trunnion Prototype

DWP Response with Tension23.5 in Live-End26 in Live-End28 in Live-EndResponse (Hz)30 in Live-EndMeasured (Known) Load (kips)West Point Configuration

DWP Response with TensionRobert F. Henry Configuration

Dispersive Tension EquationAfter performing lift-off tests, the Dispersive TensionEquations (DTE) were derived (1 st generation):“Long” Live-End Lengths (as West Point)T 181.8 * f1 123.1*C 24.55 * f1 2 0.67 * f 2 2 1099 * f 2 f 2f1 7.73* f1* f 2 2.272 * f1*C 14746.56“Short” Live-End Lengths (as RF Henry)T 12245.4 10.4C 0.0027 f 1 0.0019 f 2 4.9615 f 1 0.0179L 18888.0d 7314.7d 2 f 2

DTE Results – West Point, GA+ 8.6 k- 8.6 kMeasured (Known) Load (kips)West Point Standard Error : +/- 8.6 kips

DTE Results – Robert F. Henry+ 8.6 k- 8.6 kMeasured (Known) Load (kips)R.F. Henry Standard Error : +/- 2.6 kips

Greenup, KY104 rods testedIdentified one failed rodNo lift-offs as of yet

Advantages• Can be preformed w/ high level of safety.• No standing in front of rods.• Easily applied.• No additional stress added to the rods.• Can test 80 to 100 rods per day(approximately).

Thank You !

Response Results - R. F. HenryTypical Rods?Pier 3Typical RodsObserved frequency correlatedto exposed lengthKnown failures are outliersOthers are well below the“Typical” rodsPiers 1, 2, and 3 built bydifferent contractor than 4 thru12, and several years apartResponse (Hz)Length (in)Failed RodsLeft in damUnder-tensioned/possible failures

Greenup ResultsMost rods ~50 Hz10 3Intact RodBroken rod 43 HzLift-off test inplanning phaseAmplitude10 210 1Broken Rod10 010 2 10 3Response (Hz)

Greenup, KY Results56Observed responsecorrelated to exposedlengthOne extreme outlier forbroken rodResponse (Hz)F1 (Hz)545250484644424023.5 24.0 24.5 25.0 25.5(in)Length (in)

Trunnion Assembly ResultsRod typically testedat 25 different loadsfrom 0 to 115,000lbper setupAt no load – we getnoisy FFTsSmall load start tobring out peaksMode of vibration ofcantilever clearAll f’s increasingwith loadNo Load19.8 Hz55,150 lb46.92 Hz5000 lb33.95 Hz111,780 lb48.06 Hz

Results – Trunnion PrototypeLarge increases infrequency at lowloads60FrequencyAbove ~20,000 lbpattern of slow butsteady increase infrequency withtensionFrequency (Hz)504030204847.5104750000 70000 90000 11000000 20000 40000 60000Load (lb)80000 100000 120000

Results – Trunnion PrototypeMany tests done at different lengths and loads3000.02500.0Fully Loaded350300Response (Hz)2000.01500.01000.0500.0Lower Loads25020015010050Lower Loads0.000 1 2 3 4 25 26 27 28 29 30 31Length (in)

Preparing for liftoff tests• Test custom made gripping devices• Strain gages to determine actual pre-stressin rods• Generated mathematical relation betweenreported liftoff load and preset load in rod

Liftoff Tests Results• West Point - 109 rods liftoff tested– 7 lifted (110k to 115k)– 102 did not lift• R. F. Henry – 90 rods liftoff tested (not alldata reviewed)– 70 lifted !!!! (99k to 113k)– 20 did not lift• Greenup – 10 tests planned for 2011

West Point DataResponse (Hz)Length (in)

Length (in)R. F. Henry DataResponse (Hz)

The Dispersive Tension EquationAn equation relating tension to DWP responseThe “DTE” estimated tension with +/- 1 Std. Dev. (+/-8.0 K)

FDH Approach to Trunnion Rods• Non destructive to rods– Widespread lift-off not required (in general)– Load in rod never changes• Low risk to rods and personnel– Crews need only access to side of rods– Danger zone in front of rod avoided

FDH Approach to Trunnion Rods• Low impact on operations– Temporary structures not needed– Crane not needed• About 1/10 the cost of liftoff testing• Results +- 7.5 k

Current Approach• Measure all rods (length and diameter)• Record lateral acceleration from impact• Do liftoff tests on 5-10% of rods• Use Liftoff tests and analytic model tocreate a calibrated equation (specific to thedam)• Estimate loads in rods not liftoff tested

Future• With better modeling and ongoing datacollection eliminate need for liftoff tests• Once liftoff test are done calibration curvesand be reused for same dam• Real time monitoring of vibration frequency(passive test still had useable frequencycontent)

Thank you

Preliminary Tests• Built prototype testsetups• Gathered data forproof of concept

Greenup Rods• Grip nut integralwith face plate• “Insert” directly inface plate

Failures at West Point DamAll brittle failuresFlaw size at failure veryconsistentBackgroundFailures between 5 and15 feet from dead endWhat happens when arod having 115K oftension catastrophicallyfails?

Slipped Grip Nuts at R. F. Henry• Two with Grip nutseparated from faceplate• One slipped• Three with brittlefailures & ejection.• Rods were found tobe intact

A Methodical Research Project• FDH approached by Corps in 10/2008 with problem.• Preliminary experiments on Lab models• West Point – Feasibility tests (April 2009) – 37 rods• A Trunnion Assembly Full-Scale Model – Extensive testing• Collect field data at West Point (376 rods) and RF Henry (476 rods)• FDH testing / DYWIDAG demonstration lift-off tests on prototype• Liftoff Testing (West Point & RF Henry - 216 rods)• Greenup Dam, Greenup, KY testing (one pier)• Calibration of FDH results = Dispersive Tension Equation (DTE)

West Point Dam

Robert F. Henry Dam

Greenup, KY Dam

West Point Trunnion Rods26 to 30 inches exposedRods accessed through 24” by12” opening

R. F. Henry Trunnion Rods21 rods per group2 groups per Tainter Gate11 Gates2 end groups of 7 rods476 rods totalASTM A322-64, Grade 51601 ¼” Diameter145k Ultimate Tensile StrengthAnchor plate at far endRod Length 412” to 508”Rods are in tubes w/ NO-OXOX-ID GreaseRods from same manufacturer and lot asWest Point rods

Trunnion Model RodsVarious rod lengthsCouplers to attachactual rodfragments fromWest Point Dam

Lift-Off TestsOn site practice liftofftests at FDH facilityin Raleigh, NC• Dorsey & Dorsey• DYWIDAG• Corps ofEngineers

Preparing for liftoff tests• Close up of custom(DSI) grip-nutgripper• Feeler gagechecks for stretchin free part of rod

• Rods also tested by“Listening” as gateswere opened andclosed• Same dominantresponses observed• In fact, sameresponses observedby just attachingsensorsTesting the Rods

Modeling the Rods2-D, 3D (future) Beam Finite Element Model of RodSimple supports at face of trunnion beam and far endAdditional rotational springsTension between supports onlyDispersive Data AnalysisStrike with a HammerThe Wave

West Point ResultsCollectedacceleration timehistories ofimpactsUse DWP to findresponse contentIdentify first dominantmodes of motion2.5 x 104 21.510.50-0.5-1-1.5-2-2.50 0.5 1 1.5 2 2.5Ti ( )10 410 310 210 110 010 -110 -210 -310 -1 10 0 10 1 10 2 10 3 10 4

West Point ResultsObserved responsecorrelated to exposedlengthTwo rods are much shorterbut appear to fall in linewith responseNo extreme outliersAll rods appear to be intactResponse (Hz)Length (in)

R. F. Henry DamR. F. Henry rods have manymodes that are close to thelive-end response10 310 2Broken RodClear, single responsepeaks may not be seenDWP from broken and intactrods at R. F. Henry showsame pattern as no load andfully load case from mockupAmplitude10 110 010 -110 -2Intact Rod10 -310 0 10 1 10 2 10 3 10 4 10 5 10 6Response (Hz)

Trunnion Assembly Full-Scale Model

Greenup Trunnion Rods52 rods per group2 groups per Tainter GateOnly 1 pier tested1 1/8” Diameter~125,000 lb tensile capacityAnchor plate at far endAll rods same length (924”)Rods are in tubes with NO-OXOX-IDGrease

Preparing for Field Lift-Off Tests• West Point testssimulated (below)• 28 inch rods• Actual Rod salvagedfrom West Point• R. F. Henry testssimulated (above)• 2-4 inch rods• Gripper attaches to gripnut NOT rod directly

West Point ResultsObserved response correlated to exposed lengthNo extreme outliersAll rods appear to be intactResponse (Hz)Live-End (in)

Robert F. Henry ResultsTypical RodsResponse (Hz)Failed RodsLeft in damLive-End (in)Under-tensioned Rods

Trunnion Anchor Rod AssemblyTransfers Tainter gate forces to dam.Upwards to 115K post-tension load.