Fatigue Case Study 4

Case Studies 4: Bridge Plate with Welded Transverse Stiffners, G40.21 350W (ASTM A572 Gr.50) Steel

Fatigue Design and Evaluation Committee
Web Page: https://fde.uwaterloo.ca/Fde/CaseStudies/GhahremaniMSc/ghahremaniCase4.html

Data Reference:
1. K.Gharemani "Predicting the Effectiveness of Post-Weld Treatments Applied under Load"
MSc. Thesis, Civil Eng., U.of Waterloo 2010.
2. K. Ghahremani and S. Walbridge, "Fatigue testing and analysis of peened
highway bridge welds under in-service variable amplitude loading conditions,"
Int.J.Fatigue 33 2011, pp.300-312.

Material

The material is a bridge steel CSA standard 350W. Plate rolling direction
was in same direction as specimen loading. The transverse stiffners seem to be
welded with full penetration and are stated to have been made "using the FCAW
welding process". CSA 350W is a mild structural steel. The standard replaces
a previous CSA G40.21 specification. The "350" is for minimum yield(mpa) and the
"W" implies weldable. From other literature the nominal chemistry is

 

Nominal Chem. for CSA 350W:  

       All elements are maximums allowed:

        0.23C wt% 0.4Si  0.50-1.50Mn  0.04P 0.05S
	
Crack Initiation and Propagation data for a simulated HAZ exist

                          here

The digital "Fitted file" is available here:
https://fde.uwaterloo.ca/Fde/Materials/Steel/Other/Astm-A588C/ghahremaniSimHAZ-350W.html

Constant amplitude Crack Propagation data was tested by Ghahremani at
R=0.1 ΔS = 270 MPa and by El-Haddad (for CSA G40.11 and G40.21 steels)
Figure 1 depicts the crack propagation data as extracted from Ghahremani's
crack length vs. cycles graphs of the constant amplitude test results. Crack
length was measured using a electric potential drop technique. The El Haddad[1]
data was observed with a travelling microscope. In both cases the points shown
in figure 1 were digitized from graphs and may contain errors.

Fig. 1: Crack Propagation Rate vs. Stress Intensity for Ghahremani Potential
Drop Tests of CSA 350W and El Haddad tests of CSA G40.11 and G40.21 Steels

The data points of figure 1 were fitted visually to obtain computation curves for
use in crack propagation calculations. The points are shown in figure 2 in black.

Fig. 2: Fitted digital points overlaid on the data points of Fig. 1.

Text files of these fitted curves are available:
da/dN vs ΔK CSA350W Steel R=0.1
da/dN vs ΔK CSA G40.21 Steel R=-1

Geometry of Specimens

Welded fatigue specimen cross-section

Specimens were cut from 9.5mm (3/8") plates. Transverse stiffner plates were welded
with full penetration to the centers of the plates. The stiffner plate strips were
also 9.5mm thick and rose above and below the main plate 25mm. Welds were 6.4mm (1/4)
heel/toe full penetration. Specimens were then cut from the plates with widths of 30mm,
and test gauge sections of approximately 75mm. Further details are available in the
references at the top of this page. The main plate rolling direction was the same
direction as the loading axis of the specimens.

Load History (repeated):

Constant amplitude(CA) tests were performed at Stress Ratios of R=-1, R=0.1 and R=0.4.
The specifics of the constant amplitude tests in terms of Smax and Smin nominal stresses
are listed in Tables 1 and 2 in the "Results" section below.

Constant amplitude plus underload (CA+UL)tests consisted of blocks of 1000 cycles, where
an underload (R=-1) sub-block of 10 cycles was followed by 990 cycles at R=0.1 with
the same Smax as the underloads. Test specifics are shown in the "Results section below.

Two variable amplitude (VA) histories were applied to several samples. The histories,
supplied by Ghahremani[2] are shown graphically in figures 3 and 4.

Fig. 3: Variable amplitude history ps-m40; Smax=300, Smin=30 mpa

Fig. 4: Variable amplitude history ps-r15; Smax=380, Smin=10 mpa

Digital Versions of Test Histories:

Test Results



                        Table 1
                #Kasra Ghahremani  MSc
       
    #CA tests,  As Welded
       #R= -1  
       #  Nf   DeltaS  Smax  Smin
        33417   400.   200  -200       #R= -1  Wilson
        49568   400.   200  -200       #R= -1  Wilson
        53635   400.   200  -200       #R= -1  Wilson
       #R= 0.1
        54919   270.   300    30       #R= 0.1
        58036   270.   300    30       #R= 0.1
        69577   270.   300    30       #R= 0.1
       201741   180.   200    20       #R= 0.1
       382113   180.   200    20       #R= 0.1
       403798   180.   200    20       #R= 0.1
       #R= 0.4
       170953   180.   300   120        # R= 0.4
       180655   180.   300   120        # R= 0.4
       
       
    #POL tests and VA1 and VA2 tests
       # As Welded specimens. "L" implies Large sub-block
       # Small subblock has 990 cycles,  Large sub-block has 10 cycles
       #  Nf  sDeltaS sSmax sSmin  LSmax  LSmin
        52964   270    300    30   300    -300     #CA+UL 
        60408   270    300    30   300    -300     #CA+UL
       210197   180    200    20   200    -200     #CA+UL
       275552   180    200    20   200    -200     #CA+UL
       #
    #Variable Amplitude Histories.
       #  Nf   equivDeltaS  Smax  Smin
       295579   161         380    10        #ps-r15
       345038   161         380    10        #ps-r15
       
       225492   159         300    30        #ps-m40
       331974   159         300    30        #ps-m40
       353163   159         300    30        #ps-m40
       
       
       
                              Table 2
       
       # Kasra Ghahremani MSc Thesis 2010.
       # Needle Peened tests
       
    #CA tests
       #R= -1  
       #  Nf   DeltaS  Smax  Smin
        923988   400.   200  -200   #CA  R= -1    #runout
        924000   400   #CA  R= -1    #runout
       3096270   400  #CA  R= -1    #runout
       
       #R= 0.1
        172717   270    300    30   #CA  R= 0.1
        418576   270    300    30   #CA  R= 0.1
        592659   270    300    30   #CA  R= 0.1
        688696   270    300    30   #CA  R= 0.1
       
       3394070   180    200    20   #CA  R= 0.1  #runout
       3615600   180    200    20   #CA  R= 0.1  #runout
       3912960   180    200    20   #CA  R= 0.1  #runout
       
       #R= 0.4
       1627410   180    300   120   #CA  R= 0.4
       1747370   180    300   120   #CA  R= 0.4
       1832230   180    300   120   #CA  R= 0.4
       
    #POL tests and VA1 and VA2 tests
       #  "L" implies Large sub-block
       # Small subblock has 990 cycles,  Large sub-block has 10 cycles
       #  Nf  sDeltaS sSmax sSmin  LSmax  LSmin
        111632   270    300    30   300    -300   #CA+UL
        308927   270    300    30   300    -300   #CA+UL
        480343   180    200    20   200    -200   #CA+UL
        269589   180    200    20   200    -200   #CA+UL   runout
        298244   180    200    20   200    -200   #CA+UL   runout
       
    #Variable Amplitude Histories.
       #  Nf   equivDeltaS  Smax  Smin
       1638690   161         380    10   #ps-r15
       2889630   161         380    10   #ps-r15   runout
       3375430   161         380    10   #ps-r15   runout
       
       3454800   159         300    30    #ps-m40   runout
       4130840   159         300    30    #ps-m40   runout
       5055650   159         300    30    #ps-m40   runout

Other References:

El Haddad, M.H., "A Study of the Growth of Short Fatigue Cracks Based on
Fracture Mechanics," PhD Thesis, Civil Eng., U.of Waterloo August 1978.
Ghahremani, K. Personal Communication, 2013.