Editorial

Citation

Bailey, C. (2002), "Editorial", Soldering & Surface Mount Technology, Vol. 14 No. 1. https://doi.org/10.1108/ssmt.2002.21914aaa.001

Editorial

This special issue of SSMT brings together seven papers demonstrating the latest achievements in the applications of computational modelling technology to soldering processes and solder joint reliability.

Why use computational models?

The performance of soldering materials during product assembly is governed by complex interacting physical phenomena. Although experimentation can be used to provide an insight into the mechanisms taking place it is often impossible to use such techniques for all situations. The development and use of computational models provides a way of understanding solder material behaviour that would be impossible or far too expensive to achieve through experimentation alone. Computational models also provide the ability to reduce the overall cost of designing and manufacturing new products. This is due in part to increased accuracy in the mathematical models, larger and faster computer systems, and better graphical interfaces allowing easy use of modelling technology by engineers. Using these technologies to test new ideas and analyse current processes helps take the guesswork out of industrial process design and offers very attractive benefits to industry – in terms of financial cost savings. The use of such software tools in the context of soldering and assembly is the theme behind this special issue of SSMT.

What is computational modelling?

At the heart of any computational modelling analysis are the mathematical equations that relate, for example, the heat transfer properties of a material to its surroundings. Ever since ancient times people have used mathematics to help understand the world around them. For example, the mathematical masters of Greece provided the world with Euclidean geometry, which provided the foundations for navigation and exploration. The 17th century saw a major leap in mathematical representation of physical phenomena with the development of calculus by Newton. This provided the basis for other scientists to develop the classical equations of mathematical physics throughout the 18th and 19th centuries, such as: Heat Transfer (Fourier), Fluid Flow (Bernoulli, Navier, Stokes), Solid Mechanics (Navier, Hooke), and Electromagnetism (Faraday and Maxwell).

The arrival of commercially available digital computers in the late 1950's and the introduction of high level computing languages such as Algol and Fortran marked the true birth of computational modelling as we know it today. With the introduction of computers it became possible to solve many simultaneous equations whose number was only limited by the memory available. In the early days the number of simultaneous equations (representing, for example, the unknown temperatures across a printed circuit board) was in the 100's. If the growth in computing power had remained linear over the last forty years, then the scientific community will still be very limited in what it could accomplish today with computers. Fortunately advances in computer chip architecture have grown at such a rate that computing power has doubled every 18 months. In fact, some estimates state that our problem solving capability has increased by a factor of a million during the last thirty years.

Solving the governing equations of physics using numerical techniques embedded in software provides the opportunity to investigate, in detail, phenomena taking place both during the soldering process and subsequently during the lifetime of the solder interconnects.

How are models currently used?

This issue of SSMT demonstrates how computational modelling tools are currently being used to investigate different aspects of the soldering process from solder paste print to solder joint reliability.

The first paper by Durairaj and co-workers demonstrates models that predict the flow of solder paste material during the printing process. Given appropriate rheology data such models, based on computational fluid dynamics, can be used to predict the velocity, pressure and shear rates in solder as it rolls across a stencil. Of interest here is the different performance predicted for both tin-lead and a lead-free solder paste and the pressures along the stencil.

Solder paste coalescence is the topic discussed in the second paper by Mannan. This illustrates the use of computational fluid dynamics to predict the coalescence of solder paste particles during the reflow process. These very challenging simulations, which also have the ability to predict the breakdown of oxide layers, are beginning to provide an insight into the manner in which solder particles break down and coalesce to form the required interconnect.

Laser soldering is the theme of the third paper by Beckett and co-workers. The models used in this analysis accurately represent the energy from the laser and its effect on soldering temperatures. Comparisons are provided with thermocouple readings showing the close integration between experimentation and modelling.

The fourth paper by Whalley and Hyslop details a modelling approach to predict temperatures across a printed circuit board as it travels through a reflow oven. Based on process measurements these simple models provide very accurate results and can be used quickly by analysts to ensure that specific oven conditions will result in appropriate board temperatures.

The reflow process is again the topic in the fifth paper by Yu and Kivilathi. Very detailed computational fluid dynamics models are used to predict the temperature through the oven including the effects of airflow. Again good comparisons are provided between the model and measured data.

In the sixth paper by Lee and Huang a model is presented for the shear test of a solder bump based on the JEDEC standard, JESD22-B117. Here the stress in the solder is predicted for different shear rates where the models are identifying optimal shear speeds.

The final paper by Stoyanov and co-workers uses models to identify optimal conditions for solder joint reliability during thermal cycling. The modelling methodology used is based on integrating thermo-mechanical finite element analysis with optimisation tools, such as design of experiments.

The papers in this issue clearly illustrate the breadth in which computational modelling analysis tools are being used by researchers investigating the soldering process. We should note that such tools do not replace experimentation. When possible these tools are used in harmony with experimentation and skilled judgement to provide a sound basis to scientific investigation.

Dr Chris BaileyGuest EditorComputing and Mathematical Sciences University of Greenwich London SE10 9LS UKc.bailey@gre.ac.uk