The Journey

I am a Materials Engineer with a BS in Material Science. In the mid '70's found myself working first a technician, then as a materials engineer at Teledyne Wah Chang, a Zr, Ti, Hb, V producer, in Albany, Oregon USA. A large part of our market was Zirconium Alloys used in the Nuclear Reactors.

In the 80's there was considerable interest in improving service life of Nuclear Fuel cladding. Surface Science techniques were gaining interest. I was using primarily autoclave tests and optical microscopy to understand both uniform and nodular oxide growth.

During that time I heard of an emerging technique being promoted by Digby MacDonald at SRI called Electrical Impedance Spectroscopy, EIS. At that time we were collecting data using an x/y plotter and an oscilloscope to measure lissajou figures to obtain the current response to a voltage perturbation and plotting impedance as Nyquist and Bode Plots. Then came the first applications to fit the impedance data to electrical circuit models. The first models were Randles models describing steady state corrosion. The data I was collecting on high impedance oxide films (>10^10 w ohms) was quite different than at low impedance. The available models weren't sufficient to either characterize or correlate measurements to in-service behavior.

In 1991, I started work at Atomic Energy of Canada, Chalk River Ontario, Canada. My focus changed from fuel cladding oxidation to Deuterium pickup of CANDU pressure tubes, which dictated the life of the pressure tube. Between 1991 and 2004 the concepts behind the PEDRA approach began to take shape. The spectra of the oxide films contained three attributes; the barrier oxide thickness itself, internal oxide structure that allows ingress of electrolyte, and low frequency resistance often referred to as the polarization resistance.

In the summer of 1995, as a summer student program, the PEDRA application was developed. We used WaveMetrics IGOR PRO as a development tool so that we could alter the program as our knowledge grew. In the ensuing years till spring 2004 an extensive amount of data had been collected. However, a concise method to relate multiparameter results to in-service behaviour eluded us and EIS work was disbanded.

In September 2012, after retiring I continued to refine the PEDRA technique on data we had collected and explore its application to the broader range of impedance measurements. I compared the concept of using a signal expandable model to the traditional approach adopting a model to fit impedance data. I compared the PEDRA application to other commercial impedance software, reviewed published work and had conversations with software developers and EIS users.

Convinced that the PEDRA approach had distinct advantages over traditional methods, I employed byte physics, Physics Software Development to revise and improve PEDRA.

The result, PEDRA 9.0, is not a finished project, but an offering to the EIS community to improve upon. The Expandable Electrical Circuit Model, EECM, describes individual dielectric response that can be supported by the data (currently up to four responses can be included).

In addition to the EECM, is the concept that the three boundary contributions that are observed at: 1) high frequency inductance (e.g. batteries and super capacitors), 2) high frequency capacitance (e.g. polymer and oxide barrier films), and 3) low frequency inductance (steady state active dissolution), can be considered as external to the EECM. Each describes result of 1) high frequency/high current (inductance), 2) high frequency/low current (capacitance), and 3) low frequency/moderate current charge polarization (inductance), respectively.

This is my legacy. PEDRA is an approach aimed at determining responses that the data can support. It maybe sufficient to meet the need to correlate EIS characterization to in-service performance alone, or in conjunction with traditional ECM modelling by first identifying responses in the spectra.