The Use of Cavitation Peening to Increase the Fatigue Strength of Duralumin Plates Containing Fastener Holes

An effective method for improving the fatigue life of Duralumin plates with fastener holes, such as those used in the construction of aircraft, is to introduce a compressive residual stress around the fastener holes. Cavitation peening is a novel peening method that uses the cavitation impact produced when a high-speed water jet is injected into a water-filled chamber. In this paper, Duralumin plate specimens with holes were treated by cavitation peening under various conditions, and the fatigue strength of the specimens was determined using a plate bending fatigue test. It was revealed that a compressive residual stress was introduced not only on surfaces perpendicular to the axis of the cavitating jet but also on the walls of holes which were parallel to this. It was found that a 51% improvement in fatigue strength could be achieved by cavitation peening. Note that this is first report demonstrating an improvement in the fatigue life of Duralumin plates with fastener holes by cavitation peening.


Introduction
The development of fatigue cracks around fastener holes in aircraft components is a serious problem that affects the life of these components.However, the introduction of a compressive residual stress around the fastener holes, for example by cold expansion [1], can improve the fatigue life.Laser peening has also been proposed for this purpose [2]- [5].Peening methods that use the cavitation impact induced by the shock waves that arise when bubbles collapse [6], known as "cavitation shotless peening" or "cavitation peening", have previously been pro-posed for enhancing the fatigue life of various materials [7]- [9].
In cavitation peening, the cavitation bubbles are generated by injecting a high-speed water jet into a waterfilled chamber through a nozzle.Cavitation clouds shed periodically from the nozzle and spread out across the surface being treated, and when the bubbles collapse they produce impacts which can deform the material [10].A cavitating jet can be used to treat surfaces set parallel to the jet axis [11] and also introduce compressive residual stress at depth beneath the surface [12].Thus, it may be possible to use such a jet to treat the surfaces of the walls of fastener holes.The surface roughness produced by cavitation peening is less than that produced by shot peening [13], and the chemical cleaning required to remove iron stuck to the Duralumin plate after shot peening is not needed after cavitation peening.As cavitation peening can also improve the fretting fatigue strength of Ti-6Al-4V [14], the number of applications for cavitation peening would increase if it could be demonstrated that the fatigue strength of plates with holes could be improved.
As is well known, cavitation is a hydrodynamic phenomenon [6], in which both the development and collapse of cavitation bubbles are affected by the shape of the surface being treated.Although the cavitating regions around hydrofoils or impellers can be obtained by numerical simulation, quantitative simulation of the cavitation impact cannot be done.Thus, an experimental study of the effects of cavitation peening on holes needs to be done.A relatively low pressure plunger pump with an injection pressure of 30 MPa is sufficient for this purpose [15].The expensive plunger pumps normally used for water jet cutting and peening [16] [17] are not required in this case.Recently, a fourfold increase in the aggressive intensity of a cavitating jet using a newly developed nozzle was successful demonstrated [18].It was proposed to use this new nozzle for the experiments done in this work.
Plate bending fatigue tests were used to measure the fatigue strength of specimens consisting of Duralumin plates with fastener holes before and after cavitation peening.Since the main factor in improving the fatigue strength by peening is the introduction of compressive residual stress, the residual stresses at the surface of the plates and in the walls surrounding the holes were evaluated by an X-ray diffraction method.

Experimental Facilities and Procedures
Figure 1 shows a schematic diagram of the peening system utilizing a cavitating jet [18].The cavitating jet is produced by injecting a high-speed water jet into a water-filled chamber, i.e., tank A. The test water is stored in tank B and pressurized by a plunger pump with a maximum pressure of 35 MPa and a maximum discharge of 3.0 × 10 −2 m 3 /min.In order to avoid suction vortices, a floating plate is placed near the nozzle.After the cavitation bubbles collapse, residual bubbles [19] remaining in the test water are recirculated.These residual bubbles give rise to a cushion effect that reduces the shock wave produced by cavitation bubbles collapsing.Thus, a partition plate is placed in tank B to eliminate these.Figure 2 shows a test nozzle equipped with an optimized cavitator and an optimized guide pipe for the cavitating jet [18].The geometrical parameters of the nozzle are given in Table 1.The cavitator feeds cavitation nuclei to the nozzle plate where the main cavitation is produced, and the guide pipe increases the size of the cavitation cloud.The cavitator and guide pipe enhance the aggressive intensity of the cavitating jet by about a factor of four [18].In the present experiment, the injection pressure p 1 was 30 MPa and the standoff distance from the upstream end of the nozzle plate to the surface of the specimen was 262 mmas shown in Table 2.These conditions are the same as those used in previous work [18].The material used in the plate bending fatigue tests was Duralumin Japanese Industrial Standards JIS A2017-T3.The chemical composition and mechanical properties of this are shown in Table 3 and Table 4, respectively.The specimens were each 90 mm long, 30 mm wide and 4 mm thick with a 12 mm diameter hole at the center, as shown in Figure 3.The holes were carefully drilled by a milling drill with 3 blades using a milling cutter to make the burr as small as possible.The edge was not deburred in order to demonstrate treatment of the sharp edges by cavitation peening.
In these experiments, various cavitating jet flow patterns were considered.Six different experimental set ups were examined as shown in Figure 4.In each case the specimen was placed on a base plate.In Case (a), the specimen was set on a base plate with a 14 mm diameter hole in it and the nozzle was scanned across the specimen at a constant speed v.In Case (b), the specimen was set on a base plate without a hole and scanned.In cases (c) and (d), the specimens were treated for fixed exposure times with a stationary nozzle.The base plates in these cases were open and closed, respectively.In cases (e) and (f), in order to treat the walls of the holes only, a covering plate with a 12 mm diameter hole was placed on each specimen.In the present experiment, the processing time per unit length, t p , is defined by the scanning speed, v, and the number of scans, n, as follows.
It has been demonstrated that the effects of cavitation peening, such as improvements in the fatigue strength   [20] and suppression of hydrogen embrittlement [21], can be quantified by plate bending fatigue tests.Therefore, a displacement control plate bending fatigue test machine was used to evaluate the fatigue strength of the test specimens in this work.The residual stress in the walls surrounding the holes and at the surfaces was evaluated by X-ray diffraction.Figure 5 shows the coordinate system used on the specimens.For the case of the walls around the holes, the sin 2 ψ method was used without cutting the specimen.The incident X-rays were directed onto the wall from one side of the specimen and the diffracted X-rays were collected from the other side as shown in Figure 6.The residual stresses at the center point, i.e., point C, and 1.5 mm from the center, i.e., points A and B, were measured.The direction of the residual stress in the walls was around the circumference of the hole.For the surface, a 2D method with ω-scanning was used in order to make precise measurements [22].For both methods, Kα X-rays from a tube with a Cr target operated at 35 kV and 40 mA were used.The X-rays were directed through a 0.8 mm diameter collimator and an incident monochromator.The lattice plane, (h k l), used was the Al (3 1 1) plane and the diffraction angle without strain was 139 degrees.The diffraction ring from the specimen was detected at several angles using a 2D-PSPC.

Results
Figure 7 shows apure aluminum surface subjected to a stationary cavitating jet for one minute.A ring with many plastically deformed pits induced by cavitation impact can be observed.The outer and inner diameters of the ring are about 60 mm and 15 mm, respectively.The cavitation clouds induced by the cavitating jet developed on the surface and then collapsed generating impacts, producing a ring pattern with the region at the center undamaged [23].The sizes of the large plastically deformed pits are one or two mm in diameter.
Figure 8 shows the results of fatigue tests carried out on peened specimens from the six different set ups shown in Figure 4, and these are compared with the results for a non-peened (NP) specimen.The exposure times for the stationary jets, i.e., (c)-(f), and the processing times per unit length for the scanning jets, i.e., (a) and (b), are shown in the figure.The dotted line in Figure 8 is for the non-peened data and is given by the following equation.
In all cases, the fatigue life was extended by cavitation peening.In order to compare the lifetimes at constant bending stress, the estimated number of cycles to failure n E at σ a = 250 MPa was calculated by the following procedure.First, we assumed the slope of the dotted line in Figure 8 can also be applied to the peened specimens.The intercept, b, at n = 1 for each condition was obtained from the bending stress σ a and the number of cycles to failure n.

94.45log
Then n E was obtained using Equation (4). Figure 9 shows the variation in the estimated number of cycles to failure,n E , at σ a = 250 MPa with respect to exposure time, t e .In Figure 9, t p is converted to t e as follows.As the outer diameter of the ringregion is about 60 mm then the peening width is also about 60 mm; thus, a processing time per unit length t p = 1 s/mm, equivalent to 60 mm/min, corresponds to t e = 1 minute.n E for the non-peened specimen is 64,748 as shown by the arrow in Figure 9. n E for all the peened specimens is larger than the non-peened one.For each experimental condition, n E generally increases with t e .The closed symbols are generally placed higher than the open symbols, signifying that the fatigue strength is increased by using a base plate without a hole.When a base plate with a hole is used, many of the cavitation bubbles pass through the hole.On the other hand, without a hole in the base plate, the cavitation bubbles collapse on the surface of the base plate and on the wall of the hole.n E for the covered specimens, (e) and (f), is 92,700 and 210,223, respectively.This means that, compared with the non-peened specimen, cavitation peening has improved the fatigue life by factors of 1.4 and 3.2, respectively.These results suggest that a cavitating jet parallel to the wall around the hole can be used to treat the surface of the wall.
For practical applications a scanning jet is used, since the overall surface, as well as the fastener hole, can be peened.Thus, a scanning jet and a base plate without a hole, i.e., as in Figure 4(b), were chosen for the fatigue test to obtain the fatigue strength.One side of the specimen was treated first, followed by the other.The processing time per unit length, t p , was set to 1 s/mm.As is well known, a peened surface becomes convex after peening.With t p = 1 s/mm, the arc height was too large for the specimen, when turned over to treat the other side, to be fixed to a flat plate without the possibility of plastically deforming the specimen.In order to minimize plastic deformation of the specimen, the front side was scanned first at v = 2 mm/s, i.e., t p = 0.5 s/mm, followed by scanning of the back side at v = 1 mm/s, i.e., t p = 1 s/mm, and finally further scanning of the front side at v = 2 mm/s.In this process, the maximum difference between the processing times per unit length between the front and back sides was 0.5 s/mm; with these conditions, plastic deformation of the specimen when fixing it to the base plate could be avoided.After the process, both sides were treated by cavitation peening at t p = 1 s/mm.
Figure 10 illustrates the relationship between the amplitude of the bending stress and the number of cycles to failure for the non-peened and cavitation peened specimens, which shows that cavitation peening has improved the fatigue life.As shown in Figure 10, although the non-peened specimen with σ a = 120.1 MPa broke after 1.48 × 10 6 cycles, the cavitation peened specimen with σ a = 164.1 MPa survived for 10 7 cycles.Using Little's method [24], the fatigue strength of the non-peened specimen for 10 7 cycles is 113.5 MPa and that of the cavitation peened specimen is 171.0MPa.That is, cavitation peening has improved the fatigue strength by 51%.
In order to investigate the mechanism by which the fatigue strength is improved by cavitation peening, Figure 11 shows photographs of the surface around the hole.On the surface of the cavitation peened specimen, many plastic deformation pits can be seen around the hole.These pits introduce compressive residual stress.The sizes of the large pits are between 1 and 2 mm in diameter as shown in Figure 11(b).Figure 12 and Figure 13 show the fracture surfaces of the non-peened and cavitation peened specimens, respectively.For both the non-peened and cavitation peened specimens, the cracks begin at the edges of the hole on the surface, where the maximum bending stress occurs.Also, as shown in Figure 12(b), plastic deformation pits can be seen on the walls of the hole.This means that the cavitating jet has treated the surface parallel to the jet axis.However, the sizes of the pits on the wall surrounding the hole are a few hundred micrometers in diameter, somewhat smaller than those on the surface.
Figure 14 shows the residual stress, σ R , on the surface of the non-peened and cavitation peened specimens.The direction of the residual stress in Figure 14 is in the longitudinal direction of the specimen, i.e., the y-direction, as the tensile stress in the plate bending fatigue test was applied in this direction.In the case of the non-peened specimen, σ R varies from about −35 to −70 MPa.On the other hand, σ R for the cavitation peened specimen varies from about −170 to −230 MPa.Thus, cavitation peening has increased the compressive residual stress by more than 100 MPa.The compressive residual stress of the cavitation peened specimen near the hole is somewhat larger than at the edge of the specimen.
Figure 15 shows the residual stress, σ θ , in the surface of the wall around the hole before and after peening.σ θ before peening varies between about −20 and −120 MPa.After peening, σ θ is about −200 MPa.Thus, cavitation peening has increased the compressive residual stress by about 100 MPa.We conclude, therefore, that increasing the compressive residual stress at the surface and in the wall surrounding the hole enhances the fatigue strength.

Conclusions
In order to demonstrate that the fatigue strength around fastener holes in aircraft components can be improved by cavitation peening, specimens of Duralumin plates with holes were treated by cavitation peening and the fatigue strength was evaluated using a plate bending fatigue test.The residual stresses on the surface and in the wall surrounding the hole were measured by X-ray diffraction.The main results obtained can be summarized as follows.
1) The fatigue life of a Duralumin plate with a hole in it was improved by cavitation peening.The degree of im-  provement in the lifetime depended on the flow pattern used for the cavitating jet.For equal exposure times to the jet, the improvement when a base plate without a hole was used was much larger than when a base plate with a hole was used.
2) In the present experiments, cavitation peening increased the fatigue strength of a Duralumin plate with a hole by 51%. 3) Cavitation peening can be used to introduce a compressive residual stress both into the surface of the plate and into the wall surrounding the hole.The compressive residual stress was increased by about 100 MPa using the present conditions.

Figure 1 .
Figure 1.Schematic diagram of the cavitation peening system.

Figure 3 .
Figure 3. Geometry of the test specimen.

Figure 5 .
Figure 5. Coordinate system used for the specimen.

Figure 6 .
Figure 6.Setup of specimen on XRD system.

Figure 12 (Figure 9 .
Figure9shows the variation in the estimated number of cycles to failure,n E , at σ a = 250 MPa with respect to exposure time, t e .In Figure9, t p is converted to t e as follows.As the outer diameter of the ringregion is about 60 mm then the peening width is also about 60 mm; thus, a processing time per unit length t p = 1 s/mm, equivalent to 60 mm/min, corresponds to t e = 1 minute.n E for the non-peened specimen is 64,748 as shown by the arrow in Figure9.n E for all the peened specimens is larger than the non-peened one.For each experimental condition, n E generally increases with t e .The closed symbols are generally placed higher than the open symbols, signifying that the fatigue strength is increased by using a base plate without a hole.When a base plate with a hole is used, many of the cavitation bubbles pass through the hole.On the other hand, without a hole in the base plate, the cavitation bubbles collapse on the surface of the base plate and on the wall of the hole.n E for the covered specimens, (e) and (f), is 92,700 and 210,223, respectively.This means that, compared with the non-peened specimen, cavitation peening has improved the fatigue life by factors of 1.4 and 3.2, respectively.These results suggest that a cavitating jet parallel to the wall around the hole can be used to treat the surface of the wall.For practical applications a scanning jet is used, since the overall surface, as well as the fastener hole, can be peened.Thus, a scanning jet and a base plate without a hole, i.e., as in Figure4(b), were chosen for the fatigue test to obtain the fatigue strength.One side of the specimen was treated first, followed by the other.The processing time per unit length, t p , was set to 1 s/mm.As is well known, a peened surface becomes convex after peening.With t p = 1 s/mm, the arc height was too large for the specimen, when turned over to treat the other side, to be fixed to a flat plate without the possibility of plastically deforming the specimen.In order to minimize plastic deformation of the specimen, the front side was scanned first at v = 2 mm/s, i.e., t p = 0.5 s/mm, followed by scanning of the back side at v = 1 mm/s, i.e., t p = 1 s/mm, and finally further scanning of the front side at v = 2 mm/s.In this process, the maximum difference between the processing times per unit length between the front and back sides was 0.5 s/mm; with these conditions, plastic deformation of the specimen when fixing it to the base plate could be avoided.After the process, both sides were treated by cavitation peening at t p = 1 s/mm.Figure10illustrates the relationship between the amplitude of the bending stress and the number of cycles to failure for the non-peened and cavitation peened specimens, which shows that cavitation peening has improved the fatigue life.As shown in Figure10, although the non-peened specimen with σ a = 120.1 MPa broke after 1.48 × 10 6 cycles, the cavitation peened specimen with σ a = 164.1 MPa survived for 10 7 cycles.Using Little's method[24], the fatigue strength of the non-peened specimen for 10 7 cycles is 113.5 MPa and that of the cavitation peened specimen is 171.0MPa.That is, cavitation peening has improved the fatigue strength by 51%.In order to investigate the mechanism by which the fatigue strength is improved by cavitation peening, Figure11shows photographs of the surface around the hole.On the surface of the cavitation peened specimen, many plastic deformation pits can be seen around the hole.These pits introduce compressive residual stress.The sizes of the large pits are between 1 and 2 mm in diameter as shown in Figure11(b).Figure12and Figure13show the fracture surfaces of the non-peened and cavitation peened specimens, respectively.Figure12(a) and Figure13(a)

Figure 10 .
Figure 10.Improvement in the fatigue strength by cavitation peening using the conditions shown in Figure 4(b).

Figure 11 .
Figure 11.Photographs of surface around the hole before the plate bending fatigue test.

Figure 12 .
Figure 12.Photographs of the fracture surface and inside the hole of the non-peened specimen (σ a = 175.7 MPa).

Figure 13 .
Figure 13.Photographs of the fracture surface and inside the hole of the cavitation peened specimen (σ a = 177.9MPa).show overall views and Figure 12(b) and Figure 13(b) show close ups of the inner surfaces of the holes.For both the non-peened and cavitation peened specimens, the cracks begin at the edges of the hole on the surface, where the maximum bending stress occurs.Also, as shown in Figure12(b), plastic deformation pits can be seen on the walls of the hole.This means that the cavitating jet has treated the surface parallel to the jet axis.However, the sizes of the pits on the wall surrounding the hole are a few hundred micrometers in diameter, somewhat smaller than those on the surface.Figure14shows the residual stress, σ R , on the surface of the non-peened and cavitation peened specimens.The direction of the residual stress in Figure14is in the longitudinal direction of the specimen, i.e., the y-direction, as the tensile stress in the plate bending fatigue test was applied in this direction.In the case of the non-peened specimen, σ R varies from about −35 to −70 MPa.On the other hand, σ R for the cavitation peened specimen varies from about −170 to −230 MPa.Thus, cavitation peening has increased the compressive residual stress by more than 100 MPa.The compressive residual stress of the cavitation peened specimen near the hole is somewhat larger than at the edge of the specimen.Figure15shows the residual stress, σ θ , in the surface of the wall around the hole before and after peening.σ θ before peening varies between about −20 and −120 MPa.After peening, σ θ is about −200 MPa.Thus, cavitation peening has increased the compressive residual stress by about 100 MPa.We conclude, therefore, that increasing the compressive residual stress at the surface and in the wall surrounding the hole enhances the fatigue strength.

Figure 14 .
Figure 14.Residual stress at the surface.

Figure 15 .
Figure 15.Residual stress in the wall surrounding the hole.

Table 3 .
Chemical composion of the material under test.

Table 4 .
Mechanical properties of the material under test.