The Australian Standard AS 4663 on slip resistance of pedestrian surfaces is currently undergoing review. This has caused me to reflect on the limitations of the prescribed testing method. Whilst the pendulum is a well designed instrument, it necessarily has limitations and I offer caution in relation to testing on high profiled surfaces, some TGSIs and slopes.

Tribology, the core science behind all slips and trips, is defined simply as the engineering art and science of i. friction, ii. lubrication and iii. wear.  It involves, first off, the study of topography, i.e., the nature of a surface, and, from there, the nature of a surface interacting in some way with another surface. 

We are all familiar with profiled surfaces from the diagrams in AS 3661.1 of 1993 and onwards.  The nature of the film of water on those surfaces is, at best, very fine, since they are narrow, and the fluid is most definitely not saturated.  The peaks (asperities) of the topmost sections of those surfaces prohibit saturation of water. AS 4663 is explicit that we need to have a saturated ponded film of fluid.  Why saturated? The very design of the pendulum, and its rectangular cuboid slider, is, and only is, for the use of fully saturated fluid films.  That is, the water will, and must be seen to be, ponded. 

From the work of Professor Desmond Moore[1] (under whom I studied tribology), we know and accept that the thickness of a fluid film, h, is directly proportional to the coefficient of friction, f.  So, mathematically, dh/dt applies at all times as being directly correlated with the coeff. of friction, f.  That is the specifics. What about the general theory.  The transition between different lubrication regimes (boundary, mixed, elastohydrodynamic, and  hydrodynamic) is well described by the Stribeck curve.  The Stribeck curve is well known amongst all tribologists.  It is scientifically correct and objective.  So, as per Stribeck, it is scientifically correct to note dh / dt is proportional to f.  The following website contains a valuable introduction to the Stribeck Curve https://www.tribonet.org/wiki/stribeck-curve/  The lowermost portion of the curve is where lubrication is at its best, and that happens to be where the fluid film is at its thinnest.  The term Stribeck curve is used to describe a plot showing the frictional characteristics of a liquid lubricant over conditions usually spanning the Boundary, Mixed and Hydrodynamic regimes. Each regime is defined by the ratio of the film thickness to the surface roughness, or the λ ratio] The pendulum is only applicable to the upper right hand side i.e.., part, but not all, of the straight line.  Thin films, which require a different device’s design for friction measurement, behave very differently to thick, saturated films. 

Limitations of the pendulum. 

The design of the pendulum complies closely, and has to comply, with Reynolds Equation – a complicated, partial differentiation equation.  After all, the pendulum measures, in combination, friction and (only saturated) lubrication. 

There are some limitations to the Pendulum in the following specific circumstances:

1. Contrary to what is outlined in AS4663, the use of the pendulum on highly profiled surfaces is in error. Why? No saturated fluid film, a necessary requirement for use of the pendulum to begin with, is possible. Peaks of a surface topography prohibit the creation of thick, fully ponded, saturated fluid films.  Unless there is some specific flatness, in area terms, about the peak that allows for ponded, saturated water to remain.  Few peaks on highly profiled surfaces will allow such a ponded film.  The issue is by demonstration of the Stribeck curve that you obtain a lower value for friction on thin films with the pendulum, which is erroneous.  The pendulum’s rectangular cuboid slider has only been designed for saturated films, not thin ones. 
2. Probably errors in testing some TGSI’s in the wet condition. Unless the expert user of AS 4663 can prove that there is a fully thick, saturated, ponded fluid film of water on the topmost surface of the TGSI (and not beaded in form which is just as erroneous) then it is highly probable that only a thin layer of fluid film will eventuate. [Some, but not all,  TGSI’s have concentric rings making the water meniscus potentially liable to thin and so not be fully saturated.  ]  How does one prove one has a fully thick film,  one cannot#  So, ideally, in these circumstances, we need to flag for the user of AS 4663 that, by way of a note for example, there is a risk of non-compliance with the water dispensed on the surface because it may not be fully saturated. 

iii. Use of the pendulum on slopes.  Currently, and erroneously, in AS4663, we state that the precise process/method of test be specified.  In fact, from an engineering standpoint, looking at the design of the pendulum, it cannot be used on a slope.  There are a number of reasons for this.  i. One is you are not able to move the scale plate to the new and changed 6 o’clock position (at the sloped 5 o’clock position or the sloped 4 o’clock position), nor ii. is one able to have a fully saturated thick film – since gravitational discharge at even 3 degrees flows the water off, leaving a thin film with all the consequences demonstrated in the Stribeck curve.  One should test the horizontal, and then only rely on the slope-correction tables in the annexures of AS 4663.  Note, when testing on a sloped surface iii.  the slope gives a kick to the slider foot since the potential energy is still within the system.  So to use the pendulum on a slope, just like a highly profiled surface, one obtains a lesser value, which is erroneous.  A plaintiff solicitor may well wonder, if he looses a case due to an error in the standard, that he relied on and that is not of his, or his experts making, then contemplates it is possible to seek redress by suing Standards Australia as an organisation. 

Finally, given we cannot test for grass, sand, mud, gravel, and so forth, how does one scientifically come up with a coefficient of friction for an interface with them, and indeed similarly so for slopes (with no horizontal member) and highly profiled surfaces.  Well, that is where several chapters of the above literature resource (see footnote) comes in to assist, and allows one to give an accurate value for the c. of f. 

[Patrick Donohue M.A., M.E.D., CPE, CPEng, is a jointly qualified tribologist and ergonomist, and sits, on AS4663 on behalf of the Institution of Engineers.] 

[1] ‘Principles and Applications of Tribology’,  by Prof. Desmond Moore, Pergamon Press, 1975.