VOLUMETHREE
FailureAnalysesofSixCylinderAircraftEngineCrankshafts
Source
Robert W. Hinton Failure Analyses of Six Cylinder Aircraft Engine Crankshafts Journof of This case history was adapted from:Failure Analysis and Prevention December 2007 Vol 7 Isue 6 p 407413. DOl: 10.1007/S11668-007-9085-6
Abstract
Results of foilure onolyses of two aircraft cronkshofts ore descnibed. These cronkshofts wereforged from AMS 6414 (similor position to AISI 4340) vocuum orc remelted steels with sufur contents of 0.003% (low sulfur) and 0.0005% (ultra-low sulfur). A grain boundory sulfideprecipitote wos coused by overheat of the low sulfur stee and an incipient melting ofgrain boundry junctions ws coused by overheat of the ultra-low su/fur steeThe precipitates ond incipient me/ting in these two foiled cronkshofts were obsenved during the exominotion. Asfacets along separated grain boundanes with o smoll sphenical manganese sulfide expected impoct froctures from the low sufur steel crankshoft contoined planor dimpledprecipitotes within eoch dimple. In contrast pfanar dimpled focets along separated grainuou ps ys a wo sa saupnog jo junoo ous o pu uog uo uooq uaou yo fussu sapgd ayds jous yosasaniuysaiad aps asauuu pa saupo aoua/o machined frm theutro-low sulfur steel crnkshft filed intellyt planar grainboundyfacets. Some of the focets were covered with itrogen boron ron nd carbon fm while otheray pauyap sapns uoy puauadxa yo sqnsayfoano yns jo aau /jangoa aam sa times ond temperotures required to produce incipient melting overheot ond focets ot groinboundory junctions of ultra-low sulfur AMS 6414 steels.
Keyword:rcraft crankhafts fatigue fracture Forging defects L-sulfur steel landimpled facets Tension-to-tension fatigue testing Notch impact toughness
Material: AMS 6414 alloy steel UNS G43400
Failure types: Fatigue fracture
each pin for planar dimpled facet (PDF) fractures or facets at (SEM) was used to search one side of the entire fracture ofmagnifications over one hundred times (100×) actual size. Ifa planar dimpled facet (PDF) was found on either pin fracture the crankshaft had to be replaced in the existing motor. Thisresulted in hundreds of crankshaft rejections in those identi- fied engine lots. If no PDF was found in either pin sample the grounded plane was released with the existing crankshaft in place.
Introduction and Background
In 1999 and 2000 15 crankshafts failed within 15 h to1 254 h of service after being installed in six-cylinder recipro- cating engines of airplanes. After the aircraft engine manufac-approved the manufacturer’s issue of Mandatory Service Bul- ) o ay a d s pe letin 00-5 (MSB 00-5) [1] listing the affected serial numberedto evaluate the potential for failure of the crankshafts in these engines and prescribing a nondestructive sampling methodengines. The MSB 00-5 test method consisted of trepanning (hollow-drilling) two 6.4 mm (0.25-inch) diameter pins fromthe propeller flange at the end of the crankshaft of a groundedambient (room) temperature. A scanning electron microscope airplane. These two pins were circular-notched and broken at
Although steelmaking and residual element contents of thespecified Aerospace Material Specification (AMS) 6414 crank- shaft steels were originally suspected as a root cause of thisforging studies contained herein demonstrate that steel problem results of the failure analyses and subsequentmakers and ingot-to-billet hot-work converters were not
responsible for this problem. The analysis demonstrated that high temperature forging preheat was responsible.
Fracture Analyses
Three of the 15 failed crankshafts that had subsurfaceFCI sites were located below the hard nitrided case at a dis- fatigue crack initiation (FCI) sites were not damaged. All three tance of 0.64 mm (0.025-inch) to 1.2 mm (0.047-inch) from the surface of the crankshaft.
Steel Quality Product Check of Failed Crankshafts
within position ranges that are similar to a vacuum arc Crankshaft steels are made to AMS 6414 steel specificationsremelted (VAR) 4340 steel position. The product check positions of the two failed crankshafts in this study aregiven in Table 1.
Macrographs of the broken half of the fifth LF5 (ULS)crankshaft to fail are shown in Figs. 1 and 2. This failed crank- shaft had a subsurface fatigue crack initiation (FCl) site at adistance of 1.1 mm (0.043-inch) below the surface and near the fillet of the rear crankpin where cyclic service tensilestresses are relatively high. The scanning electron microscope(SEM) micrograph (Fig. 3) of the FCI site on the LF5 fracture is a grain boundary facet or flat contact surface between twograins. This LF5 FCI facet is approximately 0.1 mm (0.004-inch)
quality requirements of AMS 6414 steel. The R1 failed crank- Both failed crankshafts met the product check chemicalshaft had a residual sulfur content of 0.003% by weight and no intentional vanadium (0.005%). R1 steel is character-ized as low sulfur (LS) steel above 0.002% sulfur content. Infur level of 0.0005% and an intentional vanadium content contrast the LF14 failed crankshaft steel had a residual sul-of 0.074%. Note that this intentional vanadium content is not specified but is allowed by AMS 6414. This LF14 (four-teenth to fail) crankshaft steel is identified as an ultra-low sulfur (ULS) steel. Fourteen (14) of the 15 (15) failed crankshafts were made from steels with ultra-low sulfurcontents.
A macrograph of the fatigue fracture of LF14 with contactdamage on the fatigue crack initiation (FCl) site is shown in Fig 4.
Impact Test Fractures of Failed Crankshafts
Charpy *V" notch (CVN) samples were machined from theken at ambient (room) temperature. A bined scanning cheeks of R1 (LS) and LF14 (ULS) failed crankshafts and bro-electron microscope and Auger spectrometer capable of ana- lyzing approximately a 15 atom depth of surface position
Both failed R1 (LS) and LF14 (ULS) crankshafts had corehardness values beneath the nitrided case within the manu-facturer's specified core hardness range of Brinell 311 to 363 (6 0 20 o)
Chemical position of failed crankshaftspercent by weightTable 1
c/sa c Mn P Cu Ni cr Mo Al V N BR1 LF14 0.42 0.76 0.69 0.009 0.006 0.003 0.0005 0.28 0.16 1.84 0.76 0.80 0.26 0.25 0.061 0.017 0.005 0.074 0.002 0.0004AMS specifications 0.38 0.43 0.43 0.56 0.90 0.010 0.010 0.31 0.15 0.35 0.35 0.07 1.65 2.00 1.87 0.70 0.90 0E'0 0.20 Nsb NS NS NS
a C/SFailed Crankshaft Identification b NSnot specied by Aerospace Material Specification
Fig. 1 Macrograph shows fatigue fracture of failed crankshaft LF5. 2×
Fig. 2 Macrograph shows fatigue crack initiation site (white) of failed crankshaft LF5. 10×
were used to observe the planar dimpled facets within the fracture and to analyze spherical particles at the center of some of these fracture dimples.
cal [2] planar dimpled facets (PDF) of overheating prior to The R1 (LS) crankshaft CVN impact fracture reveals the classi-the dimples originated at a spherical manganese sulfide particle. hot-working as shown in Figs. 57. The microv-oids producing
The LF14 (ULS) crankshaft CVN impact fracture reveals pla-nar dimpled facets (PDF) shown in Fig. 8 that look similar to the classical PDF with manganese sulfide but most of thesespherical particles consist of iron-nitrogen-boron-carbon (Fig. 9) and some oxygen even after the spherical surfacee o a s se e smmare in adjacent dimples on the same planar dimpled facet. layers of surface (Table 2). Some manganese sulfide particles
the LF14 crankshaft forging blanks were heat-treated and CVN Small archived billet samples of the AMS 6414 steel heat forimpact-tested. No PDF features were found on these fracture
Fig. 3 SEM micrograph shows grain facets (boundary surfaces)at fatigue crack initiation site of failed crankshaft LF5. Facets A and B are on the same grain. 300 ×
Fig. 4 Macrograph shows damaged fatigue crack initiation site of failed crankshaft LF14. 10 ×
billets were relatively large even for 121 mm (4%4-inch) × surfaces. The ASTM 3.03.5 prior Austenite grain sizes of these121 mm (43%4-inch) round cornered-square billets. The hot-work steel mill converter that produced the ingot-to-billet shape exposed the steel to high heat thereby growing thegrains but did not cause the facets.
Fatigue Test of Failed LF14 Crankshaft
larger cheek (No. 8) of failed crankshaft LF14 to provide a Tension-to-tension fatigue samples were machined from the6.38 mm (0.25-inch) diameter × 38.1 mm (1.5-inch) test lengthresulting in a fatigue lest volume of 1 218 mm? (0.074-inch?). The fatigue lest samples were cyclic-load tested within a tensilestress range of 45 Mpa (6.5 Ksi) to 897 Mpa (130 Ksi) to failure or jeaqu ue jo ypeas o (aunje ou) no-unu ap uol ot e oplanar facet that would fatigue crack propagate to failure. Since
Fig. 5 SEM micrograph shows Charpy V* notch impact planardimpled facet from failed crankshaft R1. 300×
Fig. 6 SEM micrograph shows small smooth particles in dimplesof facet from CVN fracture of R1. 10 000 ×
Fig. 7 SEM micrograph and Auger spectrum of R1 CVN fracturesshow manganese sulfide with oxygen carbon and iron. 35 000×
Fig. 8 SEM micrograph shows planar dimpled fracture facet (PDF) on CVN impact fracture of LF14. 5 000 ×
Fig. 9 SEM micrograph and Auger spectrum of LF14 CVN impact fracture show surface position includes oxygen iron carbon nitrogen and boron in the spherical particle. (See Table 2 for a number of Auger analyses of particles.) 20 000 ×
Table 2 Concentration of elements detected on particles in planar dimple facets (in atom %a)
Sample ID o Fe c N B MnSensitivity factors LF14 CVN1-4 0.212 0.178 0.076 0.161 0.101 0.161 0.652Sphere #1 Sphere #2 44 23 21 13 17 30 10 17 8 ndb ndSphere R3 as rec'd Rhombi 1 33 17 21 10 21 17 34 nd 34 nd 18 nd nd 8 nd 5 nd2.5 min sputter 1.5 min sputter 9 8 32 41 12 10 26 30 18 16 nd nd nd ndRhombic #1 LF14 CVN1-1 15Rhombic #2 Sphere #1 39 34 20 12 22 38 28 16 28 3 3 pu pu 14 pu 8 8 pu 6 6
a AES does not detect hydrogen and hlium and al oncentrations are normalized to 100%b nd-not detected
