Elsevier

Surface and Coatings Technology

Volume 202, Issue 8, 15 January 2008, Pages 1443-1454
Surface and Coatings Technology

Standoff distance and bow shock phenomena in the Cold Spray process

https://doi.org/10.1016/j.surfcoat.2007.06.065Get rights and content

Abstract

Cold Spray involves the deposition of metallic powder particles using a supersonic gas jet. When the nozzle standoff distance is small, a bow shock is formed at the impingement zone between the supersonic jet and the substrate. It has long been thought that this bow shock is detrimental to process performance as it can reduce particle impact velocities. By using computational fluid dynamics, Particle Image Velocimetry and Schlieren imaging it was possible to show that the bow shock has a negative influence on deposition efficiency as a result of a reduction in particle velocity. Furthermore, the existence of the bow shock was shown to be dependent on the length of the nozzle's supersonic potential core. Experiments were carried out with aluminium, copper and titanium powders using a custom-made helium nozzle, operating at 2.0 MPa and 20 °C, and a commercial nitrogen nozzle operating at 3.0 MPa and 300 °C. In all cases, it was found that there is a direct relationship between standoff distance and deposition efficiency. At standoff distances less than 60 mm, the bow shock reduced deposition efficiencies by as much as 40%.

Introduction

The nozzle-substrate standoff distance (SoD) is one of the most important parameters in the Cold Spray process, yet its effects are only partially understood. Computational fluid dynamic (CFD) models have shown that, at short SoDs, the impact velocity of small particles is reduced as a consequence of the bow shock formed at the impingement zone; however, experimentally measured deposition efficiencies (DE), which are intrinsically linked to particle velocity, have often contradicted this phenomenon. This research aims to clarify the effects of SoD and the bow shock on particle velocity and DE.

In Cold Spray, metallic powder particles are accelerated to high velocities (600–1000 m/s) in a supersonic gas jet and directed towards a substrate. If the particle impact velocity is above a material-dependent critical value, then massive plastic deformation occurs in both the incident particles and underlying material [1]. This process disrupts thin surface films such as oxides and exposes fresh, active material; which, when brought into intimate, conformal contact under high localised pressures, undergoes adiabatic shear instabilities to form strong atomic bonds [2]. The continual bombardment and deposition of these high-speed particles leads to material build-up and eventually coating formation [3]. Coatings can be produced at low temperatures without the detrimental effects of high-temperature processing such as material oxidation, large residual stresses, poor mechanical properties, unwanted phase transformations and part distortion [4].

While nozzle SoD has been the subject of much research in the Cold Spray community there is no general consensus on its effects. As SoD is increased, the free gas jet shows a continual reduction in velocity away from the nozzle as a result of shockwaves, viscous effects and ambient mixing [5], [6], [7]; thus, the gas velocity at impingement decreases. However, the entrained particle velocity can either rise or fall outside of the nozzle depending on the gas flow conditions and the gas–particle relative velocity, Eq. (1).Fd=CdApρg(VgVp)22

Here Fd is the drag force, Cd is the drag coefficient, Ap is the particle frontal area, ρg is the gas density, (Vg  Vp) is the gas–particle relative velocity. Dykhuizen and Smith [8] showed analytically that this relative velocity determines particle acceleration, which can continue outside of the nozzle. Gilmore et al. [9] sprayed a copper powder with helium and found that the particle velocity increased from 600 to 650 m/s outside of the nozzle. Particle velocities only began to decrease when the SoD was increased beyond 50 mm. Stoltenhoff et al. [10] confirmed this observation using CFD; as SoD increased, the negative influence of the slowing nitrogen jet became too strong and the entrained copper particles began to decelerate. Conversely, Karthikeyan et al. [11] sprayed a titanium powder with helium and found that experimentally measured DEs decreased with increasing SoD. This was presumed the result of a gradual reduction in particle velocity outside of the nozzle. In contrast, Jodoin [12] used CFD to show that impact velocity of copper particles sprayed with air increased with increasing SoD – most notably for the smaller-sized particles (< 5 μm) – due to the weakening strength of the shockwave at the substrate. Although his model predicted that a SoD of 50 mm would yield the highest particle velocities and hence DEs, this was not confirmed experimentally. In fact, his DEs actually decreased as the SoD increased. Hence, although particle velocity may increase outside of the nozzle, the presence of the shockwave at the substrate – or bow shock – can theoretically reduce it.

Alkhimov et al. [13] found that when spraying with air and helium, the thickness of the compressed layer – formed between the bow shock and substrate – depended weakly on SoD; the smaller the SoD the thicker the compressed layer. Furthermore, calculations showed that aluminium and copper particles less than 5 μm in diameter could be decelerated in the compressed layer; however, this was not proven experimentally. In a separate but complimentary study, they supposed that the thicker the compressed layer, the greater the particle deceleration within it [14]. Gilmore et al. [9] and Dykhuizen et al. [15] also predicted that the smallest particles (< 5–15 μm) could be decelerated and even deflected away from the substrate by the bow shock; in both cases however, no experimental data was obtained to validate the theory. Given that DE is dependent on particle velocity, it may be inferred from these studies that DE increases with SoD.

It is clear then, that according to the theory at least, particle impact velocity and hence DE are dependent on SoD. However there are two competing mechanisms that affect particle velocity outside of the nozzle: particle acceleration/deceleration due to the free gas jet and particle deceleration due to the presence of the bow shock. Which of these mechanisms is dominant seems yet to be decided. This paper aims to clarify these matters. Information is provided on the bow shock itself, how it was modelled and predictions of its effects. In addition, the experimental equipment used to conduct spraying and measure particle velocity and the bow shock are discussed. Finally, results are presented pertaining to the effects of SoD on particle velocity, the bow shock and DE.

Section snippets

The Bow Shock

Shockwaves occur as a result of the adjustment of a supersonic flow to downstream conditions or perturbations. In Cold Spray, the downstream flow perturbation is the substrate. Fig. 1 shows a schematic diagram of the impingement zone between the supersonic gas jet and the substrate. As gas molecules in the primary jet flow impact with the substrate there is a general change in molecular energy and momentum, which is transmitted to other regions of the flow by infinitesimal pressure waves

Experimental equipment and procedures

Spraying was conducted using the Cold Gas Dynamic Manufacturing (CGDM) system [23], shown schematically in Fig. 3. This system is capable of spraying both helium and nitrogen at 3.5 MPa and 400 °C. A unique feature of this system is that it utilises helium recycling, which can operate at efficiencies in excess of 80%. During this study, two different axi-symmetric nozzles were used: the first was a custom-made linear helium nozzle, 100 mm long, with a design pressure of 2.0 MPa and a throat

Results and discussion

This research program was split into three sections: the first involved the use of PIV to measure in-flight particle velocity, the second involved the use of Schlieren imaging to visualise the bow shock, while the third involved conducting spray trials to measure DE. In each case the effects of SoD were observed.

Conclusions

The bow shock formed at the impingement zone plays a critical role in the Cold Spray process; not only does it reduce the velocity of the gas, but also that of the entrained particles. Therefore at small standoff distances, when the strength of the bow shock is high, deposition performance is reduced. While at large standoff distances, when the bow shock has disappeared, deposition can continue unhindered. Deposition efficiency was shown to be strongly dependent on standoff distance as a result

Acknowledgements

This research was carried out within the framework of a project sponsored by the Engineering & Physical Sciences Research Council (EPSRC) UK.

References (26)

  • H. Assadi

    Acta Mater.

    (2003)
  • J. Pattison

    Int. J. Mach. Tools Manuf.

    (2007)
  • T. Schmidt

    Acta Mater.

    (2006)
  • A.P. Alkhimov et al.

    Dokl. Akad. Nauk SSSR

    (1990)
  • T.H. Van Steenkiste

    Key Eng. Mater.

    (2001)
  • C. Amon
  • J.D. Anderson

    Modern Compressible Flow: with historical perspective

    (2003)
  • A.N. Papyrin
  • V. Shukla et al.

    J. Therm. Spray Technol.

    (2000)
  • R.C. Dykhuizen et al.

    J. Therm. Spray Technol.

    (1998)
  • D.L. Gilmore

    J. Therm. Spray Technol.

    (1999)
  • T. Stoltenhoff et al.

    J. Therm. Spray Technol.

    (2002)
  • J. Karthikeyan
  • Cited by (196)

    View all citing articles on Scopus
    View full text