Low-cycle fatigue tests were carried out on heat treated and phosphated samples at stress amplitude of 904 MPa and plastic strain amplitude of 5.65⳯ 10ⳮ3. The unphosphated samples with-stood 680 cycles, while the phosphated samples, 590 cycles. Con-sidering the usual scatter in fatigue tests, the difference is not
marked. On the fracture surfaces, the ratio of fatigue crack area to the overload failure area was approximately 60:40.
Discussion
The chemical composition of the failed bolts conformed to the specifications. The cleanliness and microstructure of the steel were satisfactory. The hardness and microstructure suggested that cor-rect heat treatment was given to the bolts. The intergranular oxi-dation in the bolts is an abnormality even though the failure is not directly related to this defect.
The presence of beach marks and striations on the fracture sur-face indicates clearly that the failure was due to fatigue. The
dif-1 mm
Fig. CH17.10 Marks resembling erosion tracks with the origin at the bolt hole
10 m
Fig. CH17.11 Fatigue striations in a failed bolt, at the fracture origin
10 m
Fig. CH17.12 Dimple rupture at the central region of the fracture with double origins
5 m
Fig. CH17.13 Mixed intercrystalline and transcrystalline fracture ahead of the beach marks in a bolt
100 m
Fig. CH17.14 Cracks on the barrel surface of a failed bolt, close to the fracture surface, propagating in intercrystalline and trans-crystalline modes
Fig. CH17.15 Intergranular oxidation on bolts chosen at random
ferent striation spacings coexisting on the same fracture surface suggest crack propagation initially in low-cycle mode and inter-mittently in high-cycle mode. This is in agreement with the fact that the torque generated at the time of engine start is high and it is stabilized to approximately one-third of the initial torque during continuous running of the engine.
The fretting marks on some bolts observed in conjunction with the area of contact between the two gears is indicative of flexing of the bevel gear over the spur gear. This is further corroborated by the presence of radial erosion marks on the spur gear near the bolt caused by the squeezed-out lubricating oil laden with solid particle debris. The oil squeezing is the result of flexing action of the bevel gear. The telltale marks on the gear assembly clearly indicate that there existed a cyclic relative motion between the bolt and the bevel gear, leading to fatigue failure. Poor surface finish inside the bolt hole, high interference tolerances and, probably, the existence of a burr at the chamfer at the interface could have ac-celerated the failure.
Conclusion
The dowel bolts failed due to fatigue. Fatigue was initiated by fretting in some bolts.
Recommendations
● Prevent the gap between the bevel and spur gears.
● To ensure specified contact area, blue ink check should be car-ried out even during overhauling.
● Chamfer edges should be smoothened after reboring during overhaul.
● Change of the bolt diameter from 8.5 to 9.2 mm is not desirable because the bolts have not failed because of overload or shear, and using the same diameter nut and the head section for the increased diameter of the shank without appropriate changes in the bolt hole configuration results in decreased load bearing areas.
● Decrease the interference tolerances. Higher interference will increase the fretting tendency if relative motion between the bolt hole and the shank is not prevented.
● Ensure the specified surface finish in the bolt holes.
● Ensure proper alignment of the bolt with respect to the hole as specified.
● Effective steps should be taken to prevent intergranular damage at the surface. The salt bath and phosphating baths need atten-tion with respect to the bath composiatten-tion and temperature.
● Quality control checks should be ensured to prevent occasional surface blemishes observed on the shank of unused bolts.
CASE 18
Failure of a Tail
Rotor Blade in a Helicopter
Summary
The outboard rib of a tail rotor blade in a helicopter became separated in flight, causing severe vibrations. The rivets for fitting the rib to the blade were also missing. The skin of the blade was found torn off at a number of rivet holes. The separation of the rib is attributed to stress-corrosion cracking (SCC) in the regions ad-jacent to the rivets. Erosion of the skin and the rivets could also have contributed to the failure.
Background
A helicopter experienced severe vertical vibrations during flight.
After landing, it was noticed that the outboard rib of a tail rotor blade was missing. The set of three tail rotor blades had completed cumulatively 1621 hours before the incident, compared with an expected life of 2500 hours.
Visual Examination of General Physical Features
The rib is fitted on the tail rotor blade by a set of six rivets, each passing through a pair of holes, indicated as A-A, and so forth, in the skin of the rotor blade (Fig. CH18.1). In the blade under in-vestigation, in addition to the rib, all six rivets were missing. The skin of the blade was found torn off at seven of the 12 rivet holes.
These tears are identified as A⬘, B, C, C⬘, D, D⬘, and F in Fig.
CH18.1. Figure CH18.2 shows a typical tear in the rivet hole.
The inner surface of the skin near the tip where the rib was missing was found coated with soot. The red paint on the top skin near the leading edge had been abraded and the thickness of the skin had decreased due to erosion.
Testing Procedure and Results
Scanning Electron Fractography
The two tear surfaces shown as A and B in Fig. CH18.2 were examined in a SEM. One of them, A, was found corroded. The
corrosion products and the cracks therein are shown in Fig.
CH18.3. The other tear surface, B, appeared fresh. Dimples char-acteristic of ductile overload failure were seen on this surface (Fig.
CH18.4). There was no evidence of fatigue failure on these frac-ture surfaces.
In another tail rotor blade from the same helicopter, a crack was noticed starting from the rivet hole A⬘ nearest to the leading edge and proceeding outboard to the edge of the tip. This crack was visible to the naked eye. The two rivets nearer to the leading edge had suffered more erosion than the skin.
The tip was sawed off from this blade and the rivets were moved carefully from the rib by drilling. The rib itself was re-moved in order to examine the inner side of the rivet holes.
A
Fig. CH18.1 Sketch showing the fastening of the rib to the blade
DOI:10.1361/faes2005p107 www.asminternational.org
The inner surface of the skin and the outer surface of the rib were found coated with soot. The crack emanating from the rivet hole A⬘, viewed from the inner side of the skin, is shown in Fig.
CH18.5. Figure CH18.6 shows a magnified image of this crack.
Secondary crack branching from the main crack can also be seen in this figure. In Fig. CH18.5, another small crack also emanating from the rivet hole can be seen.
Chemical Analysis
X-ray chemical analysis indicated that the skin was an alumi-num-copper alloy and the rivets were made of an
aluminum-mag-nesium alloy. The rib was made of a composite material and weighed about 55 g.
Microhardness
The microhardness of the skin was 124 HV and that of the rivet was 86 HV.
Discussion
The skin was made of an aluminum-copper alloy heat treated to high-strength condition. The rivet holes had been flared before driving the rivets. During flaring, further work hardening of this region took place. In this condition, the alloy is prone to SCC.
20 m
Fig. CH18.3 SEM fractograph showing corrosion products on the fracture surface A shown in Fig. CH18.2
20 m
Fig. CH18.4 SEM fractograph showing ductile overload failure on the frac-ture surface B shown in Fig. CH18.2
10 mm A
B
Fig. CH18.2 A typical tear in a rivet hole of a blade
Fig. CH18.5 Cracks emanating from a rivet hole in a blade
The soot on the inner surface of the skin indicates that this region had been subjected to the action of corrosive exhaust gases from the engine.
Fig. CH18.6 Magnified image of the larger crack in Fig. CH18.5
Due to erosion, the skin had thinned and the rivet heads had been reduced in size. The erosion on the rivet head was due more to lower hardness.
By a combination of these factors, SCC had set in on the skin near the rivet holes. This is evidenced by the intergranular nature of the crack. There was no evidence of fatigue failure. Due to the centrifugal forces and the already-cracked skin near the rivet holes, the outboard rib with the rivets had been pulled out of the skin.
Conclusions
The separation of the outboard rib was due to SCC in the regions adjacent to the rivets. Erosion could also have contributed. The exhaust gases from the engine provided the necessary corrosive environment.
Recommendation
Frequent inspection of the skin near the rivet holes is essential.
Care should be taken to prevent formation of a gap between the skin and the rib. Upon development of a visible crack on the skin, the blade must be withdrawn from service.