Laser–matter interactions have numerous applications which are based on the generation of plasma and its interaction with the surrounding environment. Ultrahigh intensity laser–matter interactions (high energy femtosecond laser pulses) can produce secondary sources of radiation ranging from terahertz wavelengths to x rays and can realize table-top particle accelerators [1]. However, long-pulse nanosecond laser–matter interactions enable applications like narrowband short wavelength light sources for lithography [2], broadband sources for spectroscopy, laser-induced breakdown spectroscopy (LIBS) [3], as well as nanoparticle synthesis and material processing [4]. One of the most industry-relevant laser–matter interaction applications is laser shock peening (LSP) [5] of metals for improving their fatigue life and thus the overall operational life of the system. Nanosecond laser shock laser peening involves the use of high energy laser pulses (typically mJ- to J-level energies) focused on to a workpiece to initiate the formation of a plasma. This plasma then expands and is confined by a confinement layer (typically water or transparent material). Initially, the plume pressure is significantly higher than that in the surrounding environment and thus compresses the atoms of the confinement layer leading to the formation of a shock wave propagating outwards as well as within the material. If the strength of the shock wave propagating into the material is greater than the Hugoniot elastic limit of the material, it introduces compressive residual stresses and plastic deformation near the surface (to some depth) of the workpiece, thus reducing the micro cracks on the surface of the workpiece and improving its fatigue life.

The requirement of a confinement layer (such as water or a transparent material) for nanosecond LSP can have a detrimental impact in acceptance of LSP as a technique in industry owing to the increased complexity and accessibility for remote or on-the-job LSP applications. Recently, Sano et. al. [6] have demonstrated femtosecond laser peening of an Al2024-T361 aluminum alloy without absorption or confinement layer. Laser pulse energies of up to 600 µJ and 120 fs were used with multiple shots overlapping at the same spot. Peak compressive stresses comparable to nanosecond LSP were generated; however, the depth of peening was limited to 90 µm, one tenth of the peened depth obtained by nanosecond laser peening.

In this paper, we report on absorption and confinement layer free nanosecond laser shock peening (CLF-ns-LSP) of tungsten and its alloy TAM 7525. In other words, we demonstrate compressive residual stresses in tungsten and its alloy by application of a nanosecond laser and using ambient gas at atmospheric pressure as the confinement layer. We also compare the performance of CLF-ns-LSP to that of ns-LSP with water for pure tungsten and comment on its advantages over femtosecond confinement layer free LSP.

The process of laser ablation and plasma generation is a complex process and depends on the material used in the target, the surrounding environment, as well as the pulse duration or the interaction time. For example, in vacuum with short laser pulses τ_p ≤ 1 ps, the plasma formation takes place after deposition of the entire laser pulse energy on to the target. However, for long nanosecond pulses, the plasma is formed by the leading edge of the laser pulse and is heated efficiently by the rest of the laser pulse. The introduction of ambient gas to atmospheric pressure influences the evolution and expansion of the plasma. Instead of free expansion in vacuum, the plasma under atmospheric pressure experiences internal plume structures, plume splitting, and confinement [7]. Previous studies on plasma formation and expansion in ambient gas have shown that the plume dynamics heavily depends on the laser properties, pressure, and the focusing conditions [8]. Further, a shock wave is generated due to the huge difference between the initial plume pressure and the surrounding pressure, i.e., a denser surrounding environment will result in a higher shock strength, as it is difficult for the plasma to expand. This can alternatively be described as the plasma being more effectively confined. Thus traditionally, for nanosecond laser shock peening, an additional confinement layer (typically of water or transparent material) is used to enhance the shock strength and in turn relax the stringent requirement of laser parameters and control of the plasma expansion. However, depending on laser parameters, along with the target material’s properties like density and plasma threshold [9], it should be possible to achieve suitable conditions at atmospheric pressure to generate and expand the plasma such that the shock produced has a higher strength than the Hugoniot elastic limit of the material at nanosecond time scales.

Based on this idea, we decided to experimentally investigate tungsten as the target material for CLF-ns-LSP due to its high density (19.3 g/cm³), and melting and boiling points (3683°C and 5933°C) [9], which are the highest of all known materials and thus places tungsten at the other side of the spectrum compared to a soft material like aluminum. Owing to its hardness, thermal conductivity, and high melting point, tungsten and its alloys have numerous industrially relevant applications ranging from armaments (military), aviation, sports (NASCAR and Formula One) to nuclear fusion. Importantly, magnetic confinement fusion machines like ITER, France and the JET facility in the UK use tungsten as the divertor in the tokamak machine to extract heat and ash produced in a controlled thermonuclear reaction and protect the surrounding walls from thermal and neutronic loads. Thus, it will have significant impact on the fusion reactor’s lifetime and operating cost if the fatigue life of tungsten can be improved by introducing compressive residual stresses in situ using atmospheric pressure as the confinement layer for LSP or CLF-ns-LSP.

Confinement and absorption layer free nanosecond laser shock peening experiments were carried out using the DiPOLE laser system at the Central Laser Facility, UK. Detailed description on the DiPOLE concept, laser, and its performance can be found in the literature [10,11]. The main amplifier for the DiPOLE system is a cryogenic gas cooled multi-slab Yb:YAG amplifier, capable of generating temporally shaped pulses (2 ns to 10 ns in duration) at 1030-nm wavelength in excess of 10 J at 10 Hz. The output nearfield spatial profile has an 18 mm by 18 mm flattop footprint with a super Gaussian of n = 10. For all the experiments reported in this paper, the pulse shape was kept as a square flattop pulse (shown in Fig. 1, a typical 2-ns FWHM pulse), polarization state was circular, and repetition rate was limited to 1 Hz. Also shown in Fig. 1, as the inset, is the near-field spatial profile at the output of the DiPOLE main amplifier. The output profile was image relayed to the peening workstation through vacuum spatial filters (VSFs) and a set of 45° HR mirrors. Further, the beam was focused onto the workpiece using a 400-mm focal length lens to produce a 5 mm by 5 mm spot. The workpiece was angled at approximately 10° with respect to the input beam and was fixed on an automatic XY-stage with speed control to synchronize the overlapping of subsequent laser shots.

 

Fig. 1. Temporal profile from the DiPOLE laser (FWHM of 2 ns) used for laser shock peening experiments; the inset shows the spatial profile of the output beam, with a super Gaussian profile of n = 10.

Download Full Size | PDF

Pure tungsten (99.95%, W00032) samples were supplied by Goodfellow Cambridge Ltd. with dimensions of 50 mm by 50 mm and thickness of 10 mm. They were manufactured by cold isostatic pressing of tungsten powder into sheet ingots, which were subjected to sintering (exhausting gas and impurities) and cold rolling. Finally, the samples were cut and polished to a surface roughness of Ra ∼ 10 µm. Tungsten alloy TAM7525 (W 75%, Cu 25%) rods were supplied by Thames Stockholders Ltd. and were further cut and polished in-house to 50-mm diameter and 10-mm-thick samples with similar roughness values as pure tungsten. Figure 2 shows the picture of the tungsten target after CLF-ns-LSP at 5.2 J, 2 ns. Residual stresses introduced by LSP were measured by the incremental central hole drilling (ICHD) technique [12], which is an invasive, mechanical strain relief method. All ICHD measurements presented in this study were performed at StressCraft Ltd. in accordance with the ASTM837 standard [13]. Tungsten carbide inverted cone cutters were used for the orbital hole drilling (circular milling) at a speed of 15,000 rpm with a pre-set eccentricity to give the required hole diameter. A strain gauge EA-06-125RE-120 was used to measure the relied strain and back calculations were done to ascertain the residual stresses in longitudinal (σ1) as well as transverse (σ3) directions.

 

Fig. 2. Laser shock peening locations (bottom 7.5 J, 2 ns, and top 5.2 J, 2 ns) without any absorption and confinement layer for pure tungsten.

Download Full Size | PDF

Figure 3 shows the ICHD measurements results for tungsten laser shock peening without any absorption and confinement layer. Residual stresses for an un-peened sample (gray solid line) are plotted for reference. Figure 3(a) shows the residual stresses introduced by different laser energies for a 2-ns pulse duration. Note that a strong compressive stress with a peak of −349 Mpa and depth of ∼400 µm is observed in the tungsten target at 5.2 J, 2 ns. Laser energies above and below 5.2 J did not register any compressive stresses in tungsten. To confirm this observation, we measured the residual stresses at another location (see Fig. 2) of the peened area. The repeat ICHD measurement (red dotted line) showed a peak compressive stress similar to the original measurement, however, the depth of compressive stress was increased to ∼2000 µm.

 

Fig. 3. ICHD measurements for residual stresses in pure tungsten for single shot (L1) after CLF-ns-LSP: (a) variation of laser energy while keeping the pulse duration fixed at 2 ns; (b) variation of pulse duration; (c) residual stresses at longitudinal (σ1) and transverse (σ3) directions.

Download Full Size | PDF

Figure 3(b) shows the results for a change in pulse widths to 2 ns, 4 ns, and 10 ns. Again, compressive stresses were introduced only for the 2-ns, 5.2-J pulse and no other pulse width or energy could record compressive stress. This indicated that the optimum conditions for plasma generation, expansion, and confinement at atmospheric pressure, which produce shock strength higher than the Hugoniot elastic limit of tungsten, can only be generated at 5.2 J, 2 ns, and 0.25 cm². The plasma dynamics above and below this laser energy and pulse width are not conducive for laser shock peening. Furthermore, note that the compressive stresses are not uniform for longitudinal (σ1) and transverse (σ3) directions [Fig. 3(c)]. For tungsten, the longitudinal direction (σ1) displays compressive stresses as high as −349 MPa, however, no compressive stresses are developed in the transverse (σ3) direction. This can be attributed to the surface texture of pure tungsten which shows regular patterns of linear grooves [see Fig. 5(a)], which, on application of circular polarized light, generates unequal pressures applied by the plasma shock wave in two orthogonal directions.

Further, to determine the required pulse width and energy for CLF-ns-LSP of a tungsten alloy, assuming a linear dependence on the density of the target material and based on the successful experiments on tungsten, we defined a constant parameter $\xi = E/({{\tau_p}\ast A\ast \rho } )$, where E is the energy of the laser (J), τ_p is the pulse width (s), A is the area (cm²), and ρ is the density of the target material (g/cm³). We then back calculated the pulse width required for TAM7525 CLF-ns-LSP. Table 1 shows the various parameters used for estimation of the pulse width requirements for TAM-7525. The same assumption can also be extended to other industrially relevant metals for determining optimum energy, pulse width, and area based on the density of the target, however, that will need experimental verification and is out of the scope of this paper.

Table 1. Estimation of the Required Pulse Width for TAM7525

View Table

Figure 4 shows the ICHD measurement results for the tungsten alloy, TAM7525 (W 75% and Cu 25%), for a single shot (L1) CLF-ns-LSP based on the above calculations. The laser pulse duration was varied from 2 ns to 2.6 ns keeping the laser pulse energy constant at 5.2 J [Fig. 4(a)]. Note that the peak compressive stresses introduced in TAM7525 are enhanced for the 2.6-ns, 5.2-J pulse (from −309 MPa to −357 MPa), indicating an increase of the shock strength compared to the 2-ns, 5.2-J case. This experimental evidence points to the importance of considering the density of the target material for CLF-ns-LSP. Moreover, as shown in Fig. 4(b), the compressive stresses introduced in both the longitudinal (σ1) and transverse (σ3) directions are similar, unlike the case for pure tungsten.

 

Fig. 4. ICHD measurement results for tungsten alloy TAM7525 (W 75% and Cu 25%): (a) compressive stresses for 5.2-J energy and two different pulse widths of 2 ns and 2.6 ns, for single shot (L1) in the longitudinal (σ1) direction; (b) compressive stresses introduced in the longitudinal (σ1) and transverse (σ3) directions.

Download Full Size | PDF

 

Fig. 5. SEM image for tungsten at 2-µm resolution: (a) surface profile for an un-peened sample; (b) surface profile after CLF-ns-LSP with one shot (L1); (c) surface profile after CLF-ns-LSP with four 100% overlapping shots (L4); (d) surface profile after ns-LSP with water.

Download Full Size | PDF

Comparing these results with those from the tungsten ns-LSP with water [14], we note similar peak compressive stresses but with nearly twice the laser energy i.e., 10 J, 10 ns. Furthermore, if confinement free femtosecond peening (CLF-fs-LSP) of aluminum [6] is compared with the results in this paper, a significant difference is observed for the depth of compressive stresses. The reported compressive stress depth for CLF-fs-LSP of aluminum is 90 µm, which is approximately 4 to 10 times shorter compared to CLF-ns-LSP in tungsten which is ∼7 times denser. This highlights the advantage of peening at nanosecond time scales.

To further investigate the surface morphology of tungsten, before and after CLF-ns-LSP, scanning electron microscopy (SEM) was performed for the peened and un-peened samples. Figure 5(a) shows the SEM image of an un-peened tungsten sample showing regular patterns of linear grooves owing to its manufacturing process. The CLF-ns-LSP process for single shot (L1) [Fig. 5(b)] shows changes to the surface morphology akin to instantaneous melting and re-solidification. However, this thermal effect of melting and re-solidification is more pronounced in long-pulse (10 ns, 10 J) ns-LSP with water as the confinement layer, where the surface morphology is changed significantly compared to the un-peened sample. This observation shows the effect of different pulse durations in the nanosecond time scale on the surface morphology of the target material. Although compressive residual stresses are introduced, micro-cracks appear on the sample surface on application of four 100% overlapping shots (L4) for CLF-ns-LSP, as seen in Fig. 5(c). This formation of micro-cracks on tungsten is attributed to the ductile-to-brittle transition effect during the re-solidification process. Furthermore, this observation also suggests that multiple shots at the same site (as required for CLF-fs-LSP) might be deleterious to some special metals like tungsten for the introduction of compressive residual stresses.

In summary, we have demonstrated for the first time, to the best of our knowledge, the existence of a condition where ambient gas at atmospheric pressure can be used as confinement layer in nanosecond time scale laser irradiation for the introduction of compressive stresses in tungsten and its alloy TAM7525. Peak compressive stresses of −349 MPa and −357 MPa were measured in pure tungsten and TAM7525, respectively. Compressive stresses were introduced only in one direction (σ1) during CLF-ns-LSP in tungsten. However, TAM7525 (tungsten alloy) showed similar compressive stresses in both (σ1 and σ3) directions. This demonstration of absorption and confinement layer free laser shock peening in tungsten and its alloy opens up possibilities of using pressurized cells (with multi atmospheric pressures) to achieve CLF-ns-LSP in a softer material such as aluminum.