Investigation of the Fatigue Life of Steel Base Plate 

to Pole Connections for Traffic Structuresto Pole Connections for Traffic Structures

Pooled Fund Project 9‐1526

F lt D K l F kFaculty: Dr. Karl FrankGraduate Students: Nick Richman, 

Andrew Stam Stephen PoolAndrew Stam, Stephen Pool,James Kleineck, Luca Magenes

OutlineOutline

• Test Method• Test Method

• Weld Details Tested

• High Mast Results

• Design Guidelines• Design Guidelines

• Galvanizing!?

• Weld Repair

Mast Arm Setup

x

M

P

M0 10 1

16'

High Mast ResultsHigh Mast Results

Typical Socket ConnectionFillet Welded Connection

Ø1.87" See Detail

Ø30.00"Ø36.00"

Variable Thickness

Socket with External Collar

Ø1.88" See Details

Ø30.00" 12.00" SleeveThickness=

Ø36.00"3.00"

0.375"

Stool Connection

Full Penetration Weld DetailsFull Penetration Weld Details

T W iSee Detail

Texas Wyoming

0.75"

.2545 degrees

.75 x .31

.62 x .37

.25

Seal Weld

0.25" ThickBacking

Variable Thickness

BackingRing

0.25"

3.00"

Full Penetration With External Collar/Ground SleeveCollar/Ground Sleeve

11/16" x 3/8"

1/4"

45°

0.88"0.38"

7/8" x 3/8"0.69"0.38"

.25"30°11/16" x 3/8"

0.69"

0.88"0.38"

7/8" x 3/8"

0.25"0.25" ThickB ki

.38"

0.25"

BackingRing

A100

Sockets: Base Plate Thickness

AB

CD

EEE'

e (ksi)

1.5"

10

ess Rang

2"

Str

3"

Among 8‐Bolt Specimens

1

10,000 100,000 1,000,000 10,000,000 100,000,000

Cycles

A100

Sockets: Base Plate Thickness & Bolts

AB

CD

EEE'

e (ksi) 8 Bolt 1.5"

12Bolt 1 5"

10

ess Rang 12 Bolt 1.5

8 Bolt 2"

12B lt 2"

Str 12 Bolt 2"

8 Bolt 3"

1

10,000 100,000 1,000,000 10,000,000 100,000,000

Cycles

A100

Sockets: Base Plate Thickness & Bolts

AB

CD

E 8B l 1 5"EE'

e (ksi)

8 Bolt 1.5"

12 Bolt 1.5"

10

ess Rang 8 Bolt 2"

12 Bolt 2"

Str

8 Bolt 3"

16 Bolt 3" 

1

(Lehigh)

10,000 100,000 1,000,000 10,000,000 100,000,000

Cycles

Bending of Base Plate and Mast WallBending of Base Plate and Mast Wall

Full Pen Details ‐ FractureFull Pen Details  Fracture

Full Pen DetailsFull Pen Details

FractureFracture

Wyoming Detail

Full Pen DetailsFull Pen Details

F tFracture

Texas Detail

A100

Socket vs. Full Pen Details (Phase 1)

AB

CD

EEE'

e (ksi)

10

ess Rang

Socket

Str

Wyoming

A 8 B lt 2" S i

1

Among 8 Bolt 2" Specimens

10,000 100,000 1,000,000 10,000,000 100,000,000

Cycles

A100

Full Pen Details

AB

CD

EEE'

e (ksi)

10

ess Rang Texas

WY Thick‐wall

Str

1

12 Bolt 3" Details

10,000 100,000 1,000,000 10,000,000 100,000,000

Cycles

External Collar DetailsExternal Collar Details

Socket External CollarSocket External Collar

External Collar DetailsExternal Collar DetailsSmaller Hole

Wyoming External CollarWyoming External Collar

External Collar DetailsExternal Collar Details

Texas External CollarTexas External Collar

External Collar Details ‐ FractureExternal Collar Details  Fracture

Failure at base weld toe: Socket EC, Texas EC

External Collar Details ‐ FractureExternal Collar Details  Fracture

Stiff Connection

Wyoming External Collar

Typical Weld Toe Fatigue CrackTypical Weld Toe Fatigue Crack

A100

External Collar Details

AB

CD

EEE'

e (ksi)

Socket EC10

ess Rang Socket EC

WY EC

Str Texas EC

1

12 Bolt 3" Details

10,000 100,000 1,000,000 10,000,000 100,000,000

Cycles

Finite Element AnalysisFinite Element Analysis

Local Hot Spot Stresses at Bends

l

High Mast Results

Connection DetailArm 

Diameter (in)

Base Plate Thickness (in)

No. of Specimens

Average A Value (x108)

A Val. Standard Dev.

Average Fatigue Category

Socket24

1.5 2 0.36 0.18 X

2 2 1.64 1.18 X

3 4 9.45 6.90 E' 

Total 8 5.23 6.43 E' 

242 1 2.31 0.00 X

Full Penetration3 6 11.31 6.25 E

32.625 3 5 5.23 2.89 E' 

Total 11 8.03 5.76 E'Total 11 8.03 5.76 E  

External Collar 24 3 5 19.65 12.75 E

Full Penetration External Collar 24 3 3 68.58 9.63 C

Stool 24 2 8 24.12 10.01 D

Total 35

Mast ArmsMast Arms

Socket Connection

Full PenetrationFull Penetration

Full PenetrationFull Penetration

BackingBackingBar

Crack

Full PenetrationBacking Bar Weld Failure

Full Penetrationk kBacking Bar Crack

Crack

BackingBar

Full PenetrationFull Penetration

Tack Welds

Crack

Mast ArmConnection Detail

Arm Diameter (in)Base Plate 

Thickness (in)No. of 

SpecimensAverage A Value (x108)

A Val. Standard Dev.

Average Fatigue Category

'

Socket10

1 3 5.85 3.86 E' 1.5 11 4.68 1.72 E' 1.75 2 2.39 0.11 X2 12 14.59 23.72 E3 2 10 10 5 09 E'3 2 10.10 5.09 E  

13 2 3 1.35 0.15 XTotals 33 8.98 15.40 E' 8 2 2 110.84 10.62 C

1 8 26 10 7 36 D

Full Penetration

10

1 8 26.10 7.36 D1.5 2 12.86 14.93 E2 5 95.15 29.88 C3 12 118.09 95.18 C

122 2 56.41 7.12 C

123 2 42.94 3.55 D

13 2 1 7.85 0.00 E' Totals 34 77.80 72.50 C

Full Penetration (peened) 10 3 3 44 96 25 37 CFull Penetration (peened) 10 3 3 44.96 25.37 C

External Collar

8 2 2 34.00 5.79 D

101.5 2 55.68 22.43 C1.75 5 57.64 24.35 C

External Collar2 8 93.91 64.13 C

12 2 4 21.72 4.68 ETotals 21 62.17 49.06 C

Totals 91

10" Diameter Mast Arm with 2" Base Plate Connection Comparisons

100%

120%

ceed

ing 

Connection ComparisonsSocket Connections (12)Full Penetration Connection (5)External Collar Connections (8)

80%

eting or Ex

40%

60%

cimen

s Mee

Category

20%

40%

age of Spe

c

0%

≥X ≥E'  ≥E ≥D ≥C ≥B ≥APercen

ta

Fatig e CategorFatigue Category

120%

Mast Arm Connection Comparisons

100%

Exceed

ing  Socket Connections (33)

Full Penetration Connection (34)

External Collar Connections (21)

80%

Meeting

 or E

y

60%

ecim

ens M

Category

20%

40%

ntage of Sp

0%

20%

Percen

≥X ≥E'  ≥E ≥D ≥C ≥B ≥A

Fatigue Category

Conclusions of ResearchConclusions of Research

• Fatigue Strength Increases with:Fatigue Strength Increases with:– Thicker Base Plates– Smaller Internal Holes in Full Penetration Details

• High Mast‐Combination of Thick Base Plates, Small Hole, 12 anchor rods, and External , ,Collar Produces Category C Fatigue Strength

• Bolt Pattern Not Significant on Mast Armsg• Full Penetration Weld on Mast Arm‐ Category D or better with 2 in. base platep

Thirty‐three Inch Diameter High M t P f SMast Performance Summary

The Last Specimens?The Last Specimens?• Pair of 33” HMIP for last tests of 

l d f d jpooled fund project

• Poles identical in fabrication

• One specimen, 33‐3‐12‐TX‐SG‐A, was galvanized, the other 33‐3 12 TX SB B was left black3‐12‐TX‐SB‐B was left black

• Ultrasonic Testing revealed that the 33‐3‐12‐TX‐SG‐A specimenthe 33‐3‐12‐TX‐SG‐A specimen contained small cracks

Pre Test Inspection

• 33‐3‐12‐TX‐SG‐A

Pre Test Inspection

33 3 12 TX SG A had cracks on every bend except atbend except at seam weld

• Testing resulted in• Testing resulted in a lower‐than‐anticipated fatigueanticipated fatigue performance

Typical Initial CrackTypical Initial Crack

Test Results

• 33‐3‐12‐TX‐SG‐A

Test Results

33 3 12 TX SG A only cycled 81,326 times before developing large fatigue 

kcracks• Tested back to b k i h bl kback with black specimen

Research History• 33‐3‐12‐TX‐SB‐B UT indicated no signs of initial cracking

• 33‐3‐12‐TX‐SB‐B did not k fcrack after

81,326cyclescycles

• Two Additional Galvanized SpecimensGalvanized Specimens Tested From Another Supplier

Initial Test ResultsInitial Test ResultsA

B

100

CDE

E'

(ksi)

10

Stress Ran

ge (

Nom

inal S

33‐3‐12‐TX‐SG

1

33‐3‐12‐TX‐VG

33‐3‐12‐TX‐VG (Flipped)

1

10,000 100,000 1,000,000 10,000,000 100,000,000

Cycles

Fatigue Crack Bend 5Fatigue Crack Bend 5

Fracture Surface Bend 5Fracture Surface Bend 5

Fatigue Crack

Initial Crack

Crack Depth at ToeCrack Depth at Toe‐3/32 in.

Initial Crack Bend 5Initial Crack Bend 5

Initial Crack at Bend 12 OpenedNo Fatigue Damage

Dark Area‐Initial Crack

Cracks at Weld SeamTested after Cracking at Bend 5

Shallow Initial CracksShallow Initial Cracks

Longitudinal Seam Weld at CornerNo Initial Cracks!

Possible Cracking CausesPossible Cracking Causes

• Source currently unknownSource currently unknown– State of art described in report written by Thomas J Kinstler: Current Knowledge of the Cracking ofJ. Kinstler:  Current Knowledge of the Cracking of Steels During Galvanizing

– Largely speculative with general rules of thumbLargely speculative with general rules of thumb which are not fully developed scientifically

Possible Cracking Causes

• Investigated five potential causes of cracking– Bend radius of shaft

– Base Plate Volume to Shaft Volume

– Liquid Metal Embrittlement / Chemistry of Bath and Base Metal

– Hardness testing

– Thermal Stress Analysisy

Bend Radius Study

• Variable radii per sample

Bend Radius Study

Variable radii per sample

• Variable radii per manufacturer

G d b S d i i f• Governed by ASTM and American Institute of Steel Construction Design Manual– For hot‐dip galvanized structural steel products, ASTM 143 recommends bend radius ≥ three times the member thicknessthe member thickness

Bend Radius StudyBend Radius StudyHigh Mast/Pole Data

Specimen NameThickness  Avg. Inside Bend 

Radius/ThicknessSpecimen Name(in)

gRadius (in)

Radius/Thickness

33‐3‐12‐TX‐SG‐A* 0.313 1.45 4.6433‐3‐12‐TX‐SB‐B 0.313 1.35 4.3233‐3‐12‐TX‐VG‐A* 0.313 1.28 4.0833‐3‐12‐TX‐VG‐B* 0.313 1.23 3.9224‐3‐16‐WY‐PG 0.313 0.66 2.1124 3 16 TX PG 0 313 0 69 2 2024‐3‐16‐TX‐PG 0.313 0.69 2.20

24‐3‐16‐SEC‐PG‐A 0.313 0.40 1.2924‐3‐16‐SEC‐PG‐B 0.313 1.09 3.4924‐2‐8‐STL‐VG‐A 0.313 4.00 12.8024‐2‐8‐STL‐VG‐B 0.313 4.00 12.80

VII‐6 0.250 0.79 3.17VII‐7 0.250 0.83 3.32

*Asterisk denotes existence or indication of initial cracks

Bend Radius StudyMast Arm DataThickness Avg Inside Bend

Bend Radius Study

Specimen NameThickness 

(in)Avg. Inside Bend 

Radius (in)R/T

10‐2R‐EC‐PG‐A 0.179 0.48 2.67

10‐2R‐EC‐PG‐B 0.179 0.35 1.98

10‐2S‐WY‐PG‐A 0.179 0.65 3.60

10‐2S‐WY‐PG‐B 0.179 0.67 3.72

10‐3R‐WY‐PG‐A 0.179 0.42 2.33

12‐2R‐EC‐PG‐A 0.179 0.42 2.32

12‐2R‐EC‐PG‐B 0.179 0.35 1.98

12‐3R‐WY‐PG‐A 0.179 0.42 2.32

VII 1 0 188 0 47 2 49VII‐1 0.188 0.47 2.49

VII‐2 0.188 0.47 2.49

VII‐3 0.188 0.47 2.49

VII‐6 0.188 0.68 3.61

VII‐7 0.188 0.65 3.49

Bend Radius Study ConclusionsBend Radius Study Conclusions

Cracked SpecimensCracked Specimens

Specimen Name Radius/Thickness

33‐3‐12‐TX‐SG‐A* 4.64

33‐3‐12‐TX‐VG‐A* 4.08

33‐3‐12‐TX‐VG‐B* 3 92

• Cracked Specimens Had Larger R/t ratio then 

33‐3‐12‐TX‐VG‐B 3.92

many uncracked specimens

• Radius/Thickness ratio does not seem to be a contributor to cracking

Base Plate to Pole Wall Volume Ratio Comparison

• Relationship between base plate weight (volume) and bottom 12” of shaft weight (volume) noticed in comparing designs that did and did not result in 

ki i T C k i B Pl t W ld 30 Ycracking in Toe Cracks in Base Plate Welds – 30 Years Later by Richard Aichinger and Warren Higgins 

• Shows that as this ratio increases the “Event• Shows that as this ratio increases the  Event Probability” also increases in a nearly linear trend

• We have interpreted “Event Probability” as number• We have interpreted  Event Probability  as number of cracks divided by number of inspected bends

Percentage of Cracks Found in the Field and Laboratory versus Volume Ratio of Base Plate to Shaft

150'‐80mph60%

70%

Cracks

33‐3‐12‐TX

40%

50%

Bend

s with C

30%

40%

f Inspe

cted

 B

100'‐80mph

125'‐80mph

175'‐80mph

10%

20%

ercentage of

175 80mph

24‐3‐16‐TX/WY0%

0 2 4 6 8 10 12 14

Pe

Base Plate to Shaft Volume (Bottom 12") Ratio

• No Correlation Found Between Volume Ratio and Cracks Discovered.

Chemistry AnalysisChemistry Analysis

• Pelco Structural and Steel Ameron UnionPelco, Structural and Steel, Ameron, Union Metals, and Valmont samples taken from high masts for chemical testing at Chicago Spectromasts for chemical testing at Chicago SpectroService Laboratory, Inc

• Steel chemistry differed slightly from• Steel chemistry differed slightly from fabricator to fabricator

Chemistry AnalysisChemistry Testing 2 (%)Chemistry Testing 2 (%)

Chemistry AM UM PW *ST VS *VTAmeron Union Pelco Structural Valmont Old Valmont New

0.22 0.2 0.15 0.04 0.2 0.15Manganese 0.85 0.75 1.07 0.94 1.3 1.05

eg

Phosphorus 0.013 0.013 0.009 0.013 0.009 0.008Sulfur 0.005 0.006 <0.005 <0.005 <0.005 <0.005Silicon 0.02 0.02 0.02 0.21 0.02 0.02Nickel 0.05 0.05 0.04 0.05 0.04 0.04Chromium 0.06 0.06 0.03 0.03 0.03 0.04

Base Molybdenum 0.01 0.01 0.01 0.01 0.01 0.01

Copper 0.07 0.07 0.02 0.03 0.02 0.04Aluminum 0.04 0.04 0.04 <0.01 0.04 0.04Nitrogen 0.003 0.007 0.004 0.004 0.006 0.005Boron <0.0005 <0.0005 <0.0005 <0.0005 <0.0005 <0.0005

dVanadium <0.01 <0.01 <0.01 <0.01 <0.01 <0.01Titanium <0.01 <0.01 <0.01 <0.01 <0.01 <0.01Niobium <0.01 <0.01 <0.01 <0.01 <0.01 <0.01Iron Balance Balance Balance Balance Balance BalanceCadmium <0.001 <0.001 <0.001 <0.001 <0.001 <0.001Al min m <0 005 <0 001 <0 001 <0 001 0 001 0 001

Coating

Aluminum <0.005 <0.001 <0.001 <0.001 0.001 0.001Lead 0.005 0.005 <0.005 <0.005 0.005 0.005Tin <0.005 <0.005 <0.005 <0.005 <0.005 <0.005Nickel <0.001 <0.001 <0.001 <0.001 0.001 0.001Indium <0.005 <0.005 <0.005 <0.005 <0.005 <0.005Bismuth <0 005 <0 005 <0 005 <0 005 <0 005 <0 005

*Specimen with initial cracks

C Bismuth <0.005 <0.005 <0.005 <0.005 <0.005 <0.005Iron <0.005 <0.005 <0.005 0.005 0.005 0.005Copper 0.001 0.001 0.001 0.005 0.005 0.005Zinc Balance Balance Balance Balance Balance Balance

Zinc In CracksZinc In Cracks

• Electron microscopy tests reveal localElectron microscopy tests reveal local elemental concentrations

Micro Hardness TestingMicro Hardness Testing

• Utilized to determine estimated tensileUtilized to determine estimated tensile strength of material Micro Hardness 

Test Pass E

Micro hardness sample taken from bend 7 of S&S

• No significant hardness variation between weldMicro Hardness Testing

• No significant hardness variation between weld, HAZ and base metal

Hardness through Weld Profile

200

250

Hardness through Weld Profile

150

200

rdne

ss

100

Vickers Ha

0

50

0

1 2 3 4 5 6 7 8 9 10Weld Heat Affected Zone Base Metal

Weld RepairWeld Repair

• Crack detection techniquesCrack detection techniques

• Weld repair specificationsSh i– Shop repair

– Field repair

• Repair observations

• Repair performance

Crack Detection TechniquesCrack Detection Techniques

• Magnetic Particle TestingMagnetic Particle Testing– Advantages

• Little training necessaryLittle training necessary

• Easily portable

• Makes cracks subjected to jtension on smooth, surfaces easily visible

Crack Detection TechniquesCrack Detection Techniques

• Magnetic Particle TestingMagnetic Particle Testing– Disadvantages

• Necessitates smooth surfaces with few if anyNecessitates smooth surfaces with few, if any, geometric discontinuities

• Does not indicate the depth of crack

• High compressive loading across crack will reduce effectiveness, which we proved experimentally

If k i t li htl i ibl d t ti ill• If crack is not slightly visible, no detection will occur

• Will not detect small initial cracks

Crack Detection TechniquesCrack Detection Techniques

• Ultrasonic TestingUltrasonic Testing– Advantages

• Accurate at detecting both initial and fatigue cracksAccurate at detecting both initial and fatigue cracks

• Will locate cracks regardless of weld geometry

– DisadvantagesDisadvantages• Requires specialized training to accurately use

Weld Repair SpecificationsWeld Repair Specifications

• Surface preparationSurface preparation– Clean surface

Grind out crack– Grind out crack• Leave 1/16” root face on through cracks (field repair)

• Grind ½ wall thickness minimum (shop repair)Grind ½ wall thickness minimum (shop repair)

• Grind crack 2” beyond extents of indication

– Verify crack location identified with mag. particlee y c ac oca o de ed ag pa c e

– Remove galvanizing with flap wheel

Shop Weld Repair Specifications Shallow Cracks Discovered After Galvanizing

• FCAWFCAW– Developed to repair cracks due tocracks due to galvanizing in shop

– Can execute in oneCan execute in one pass

– Results in favorable weld profile

FCAW Characteristics• Filler metal

– AWS E71T‐1

– 1 stringer pass, 1/8” electrode

• Weld Parameters– 170 to 370 Amps.

– 21 to 28 V, DC+,

• Shield gas25% CO2 75% Ar– 25% CO2, 75% Ar

– Flow rate 45 cubic feet per hour

P iti• Position– Horizontal (1G)

Field Weld Repair SpecificationskDeeper Fatigue Cracks

• SMAWSMAW– Developed to repair crack located in erected poles

– Accessible for all welderswelders

– Slow processWill require tighter– Will require tighter quality control to properly execute

SMAW CharacteristicsSMAW Characteristics• Filler metal

– AWS E7018• 1 stringer root pass, 3/32” electrode

• 3 stringer passes, 1/8” electrode

– Interpass cleaning with wire brush

• Weld Parameters– 80 to 100 Amps.

• Position– Vertical (2G)Vertical (2G)

Weld Repair PerformanceWeld Repair PerformanceA

BC

100

CD

EE'

e (ksi)

10

al Stress R

ange

24‐3‐12‐TX‐VG

Nom

ina

24‐3‐16‐TX‐PG

33‐3‐12‐TX‐SG

33‐3‐12‐TX‐VG

33‐3‐12‐TX‐VG (Flipped)

1

Field Repair

Shop Repair (Did Not Fail)

10,000 100,000 1,000,000 10,000,000 100,000,000

Cycles

Did Not Fail2,000,000

Number of Cycles to Failure at a Stress Range of 12 ksi for Various TxDOT Details

1,600,000

1,800,000

D

1 200 000

1,400,000

G

E

1,000,000

1,200,000

Num

ber o

f Cycles

24‐16‐TX

‐PG

VG‐A

VG‐B

G‐B 

d) ‐TX‐SG

‐B 

paired)

E' E'600,000

800,000

2‐TX

‐VG‐A

24‐12‐TX

‐VG‐B

G‐A

33‐12‐TX

‐V

33‐12‐TX

‐V

12‐TX‐VG

‐A 

(Flipped)

‐12‐TX

‐VG‐B

(Flipped)

33‐12‐TX

‐VG

(Repaired

33‐12

(Re p

E'

E'

X

E' E'

200,000

400,000 24‐1 2

33‐12‐TX

‐S

33‐ (

33‐

0

24" 32.625" Field Repair Shop Repair

Weld Repair PerformanceWeld Repair Performance

• Weld repair specimens’ performanceWeld repair specimens  performance exceeded shop fabricated specimens

• Field Repair SMAW• Field Repair ‐ SMAW– More than 1,467,734 cycles

F il f d b f ld ( b l )– Failure forced to base of weld (at base plate)

• Shop Repair ‐ FCAW– Nearly 1,893,306 cycles, deemed class D run out

Thermal StudyThermal  Study

KHF

Thermal StudiesThermal Studies

• Galvanizing InstrumentationGalvanizing Instrumentation– Thermocouples attached to identical high masts during galvanizing

– Data recorded fromtwo separategalvani ersgalvanizers

– Indications of severetemperature gradienttemperature gradientbetween shaft andbase plate

81 of 10

Current WorkCurrent Work

• Finite Element Thermal Stress AnalysisFinite Element Thermal Stress Analysis– Thermal Conductivity Studies to Develop Heat Flow Between Bath and High Mast PolesFlow Between Bath and High Mast Poles

– Parametric Studies to Look at Effect of Base Plate Mass to Pole Mass Upon Thermal StressesMass to Pole Mass Upon Thermal Stresses

• Additional Full Size Specimens to Measure Temperature Response of Ground SleeveTemperature Response of Ground Sleeve Specimens and Development of Cracks

QuestionsQuestions