Welding wave on the contact spot of solids

https://doi.org/10.1016/S0301-679X(98)00018-8Get rights and content

Abstract

In this paper the theoretical model of adhesion between clean solid surfaces is presented. Material behaviour is described by hydrodynamic equations for viscous liquid and by the Gruneisen equation of state. This approach is quite suitable for quantitative analysis of mechanical processes running under high pressures. The model states that growth of a bonded area can occur in a regime of self-propagating welding wave (WW). WW originates at any active point where initial bonding takes place and further moves at velocity Uww∼γ/η (γ is surface energy and η is viscosity of material). According to the model presented surface energy plays the key role in adhesion.

Introduction

The seizure phenomenon is caused by the welding of solid bodies (parts) over contacting area during friction or joint deformation. To avoid seizure in most cases lubricants are used in tribology systems. At the same time the seizure phenomenon is used in technologies of metal joining by roll, pressure and other kinds of cold welding.

There presently exist different views on the mechanism of solid-state-welding. In Semenov[1]and Krasulin[2]the opinion was stated that the bonding process proceeds similarly to heat explosion. Seizure occurs as a result of a self-sustaining topochemical reaction which sometimes proceeds violently, and the diffusion process can play a substantial role on complete reaction only. It is emphasised by Semenov[1]that interatomic bonding occurs between atoms which escape the certain energetic barrier (activated atoms). This activation is stimulated by heat transfer from bonded to unbonded area. The first acts of bonding appear at `active points', which in many cases are the dislocation outlets on the solid surface[2].

Tabor and his co-authors[3]consider that at least for metals activation of surface atoms is not needed for bonding, and the moving force of the seizure process is connected with the existence of surface energy. There are the other thoroughly made experimental works which confirm this point of view4, 5.

To our opinion, the work most close to reality is the consideration given by Buche[6], who stated that the mechanism of seizure depends on materials as well as on conditions under which these materials are treated. To provide bonding in some cases it is enough to destroy surface layers and to force clear surfaces into direct contact. In other cases activation of surface atoms is needed as an additional condition.

In any case all researchers agree that surface contaminations are the main obstacles preventing bonding. There are a lot of publications on the behaviour of lubricants in tribology systems, and on the effect of oxides, hydroxides and other layers on the cold welding process (see for example references in Semenov[1]). Therefore we do not consider this question in the present paper. It should be borne in mind only that surface layers have an effect on seizure, even in such extremely intensive processes as explosive powder compacting and explosive welding7, 8. Actually, in the model below we assume that surface contaminations are somehow removed from the contact zone and surface atoms are driven in direct contact. Thus we study the mechanism of clear surface bonding.

Section snippets

Theoretical considerations

Perhaps everyone has watched the coalescence of mercury balls. The moving force of this process is the surface tension of material. Coalescence occurs very quickly because mercury has a small shear strength and viscosity. For all solids which have positive surface energy the bonding over the contact zone is thermodynamically justified, but this process is usually suppressed by material strength (low mobility of atoms). Experience shows that this obstacle (strength) is overcome by the action of

Isothermal process

In this case χ is found from Eq. (17)directly, as we have T2=T1. An isothermal process is possible when thermal conductivity of material is of great value and all the released heat is quickly removed from the bonding zone. Eq. (17)can be transformed into the form ofχ2+1+αv20T1v21−2γv0dKv1χ−2γv0Kdv1=0

In Table 1 the physical parameters for several metals are given. As can be seen for all of them, approximate equality has place2γKd≈0.1

Besides, we can take v1≈v0 because the bulk modulus K is very

Adiabatic process

In this case q=0 and Eq. (17)should be solved together with , . Besides we should take into account that the flow line consists of two atomic layers (see assumption 3 above) and that released surface energy should be included in consideration. It means that the zero level of energy w0 in Eq. (16)is different for material before (state 1) and behind (state 2) the WW front. Denoting these energies as w01 and w02 we havew01−w02=γdρ0

Making a number of operations we get the quadratic equation for

Discussion

It can be seen from Table 2 that Uww depends much on the rate of deformation ϵ′ because material viscosity is very sensitive to this flow parameter. As mentioned above ϵ′ is the given parameter characterising the state of material under certain loading conditions. Viscosity decreases with the growth of ϵ′ and this provides less time for complete welding. This time can be estimated if the surface density of active points is known. According to Krasulin[2]dislocation outlets play the role of

References (15)

  • O Nishikawa et al.

    Some FIM observations on cold-welding of metals

    Surface Science

    (1971)
  • A.P Semenov

    Skhvatyvanije metallov i metody ego predotvrashenija pri trenii

    Trenije i Iznos

    (1980)
  • Krasulin, Yu. L., Vzaimodeistvije Metalla s Poluprovodnikom v Tverdoi Faze. Nauka, Moscow,...
  • N Gane et al.

    Adhesion between clear surfaces at light loads

    Proc. Roy. Soc., London

    (1974)
  • K.I Johnson et al.

    Effect of contamination on the adhesion of metallic couples in ultra-high vacuum

    J. Appl. Phys.

    (1967)
  • Buche, N. A., Podshipnikovyje Splavy Dlja Podvizhnogo Sostava. Transport, Moscow,...
  • Shtertser, A. A., Combustion, explosion, and shock waves, Vol. 29. Plenum Publ. Corp., New York, 1993, pp....
There are more references available in the full text version of this article.

Cited by (8)

View all citing articles on Scopus
View full text
Copyright © 1998 Elsevier Science Ltd. All rights reserved.