Fowler Associates Labs

Procedures For The Design, Analysis And Auditing
Of Static Control Flooring/Footwear Systems

Stephen L. Fowler
Fowler Associates, 3551 Moore-Duncan Hwy, Moore, SC 29369
Tel: 864-574-6415, FAX: 864-576-4992, Email: sfowler@sfowler.com

William G. Klein
K&S Laboratories, 2026 Bay Rd., Stoughton, MA 02072
Tel: 617-341-8331, FAX: 617-341-8331, Email: kslabs@thecia.net

Larry Fromm
Hewlett Packard, 1501 Page Mill Rd., Palo Alto, CA 94304

Abstract - It is the purpose of this paper to show that the electrostatic performance of footwear/flooring systems, defined as the electrostatic potential of personnel arising out of the use of these systems, can be predicted with adequate precision based on component resistance data alone, and further to present resistance testing methodologies which are at once more relevant and more reproducible than most in common usage today. Conclusions, which are to a considerable extent a matter of opinion though based on hard data, suggest that inappropriately defined criteria and overly stringent specification are significant problems today to users, suppliers, and auditors.

Background and Introduction

It is generally accepted that the resistance and triboelectric properties of footwear and flooring materials together constitute the main parts of the system which limit the electrostatic body voltage of a person walking on the floor. Charge decay time may also be inferred from resistance data. Unfortunately, the use of resistance to predict or define quantitatively the electrostatic performance of flooring/footwear systems has been fraught with many problems. The problems are not trivial because the industry-wide failure to establish appropriate standards for the measurement of components and the performance of systems and the consequent use of various methodologies yielding significantly different values have often led to serious difficulties. These difficulties have resulted, for example, in both failed installations and expensive claims of failure not based on "true" criteria of performance. It is not possible to set up universally useful standardized criteria for product design, field performance, or auditing procedures without a significant degree of basic knowledge and an agreement within the ESD community as to the specific, quantifiable objectives toward which these criteria are aimed. With different manufacturers and different users playing by different rules, chaos has resulted.

Past work ^(2,3) has addressed some of the difficulties and has suggested solutions in defining relevant criteria for the design and evaluation of static controlled flooring, footwear, and flooring/footwear systems. It has not, however, had any noticeable effect on the way that floors and shoes are specified and tested, nor has there been any systematic effort to reconcile the troublesome differences. It is well understood that the rule of thumb for personnel voltages of a hard grounded worker are as follows: at 100 Meg W the personnel voltages may be over 100 Volts ; at 10 Meg W personnel voltages will usually be less than 100 Volts; at 1 Meg W the expected personnel voltages are in the 10 Volt range. Flooring and footwear systems are more complex than a simple wriststrap grounded situation. The current work expands on past work, introduces new methodologies which are both reproducible and relevant, and verifies the assertions made with quantitative data from both the laboratory and operative factory installations. A critical examination will be made as to how some currently used test methodologies fit into this picture, specifically ESD 7.1, ASTM F-150, NFPA 99, ESD 9.1 and IEC 1340-4-1. It is hoped that this work will be sufficiently intriguing to lead to a concerted effort to develop and promulgate a performance oriented set of specifications and more universally accepted evaluation and auditing procedures.

This work is aimed equally at those whose good fortune it is to have the freedom to write specifications for new installations and those tortured souls who must make critical choices and compromises in the utilization, modification, or scrapping of existing non-conforming flooring

Basic System Analysis

In order to intelligently approach the issue of specifying and evaluating the resistance properties of footwear and flooring as they affect body voltages, it is necessary to have some reasonable idea as to how these properties act and interact together with other pertinent system parameters. This is a highly complex problem for which there is no easy solution. Briefly, some common deficiencies of often cited analyses consist of:

1. the concept of linear, lumped parameters when they are in fact neither linear nor lumped;
2. neglect or inadequate definition of important variables; and
3. the use of a steady state approach to a dynamic event.

The list could go on. An example of this sort of specious reasoning is the popular representation of the human walker as a parallel plate capacitor with constant charge. Although it is tempting to use such a configuration, which sort of looks like a foot on the floor and is simple electrically, it suffers from all three deficiencies mentioned above and has virtually no analytical value. It is well beyond both the scope and purpose of this paper to delve deeply into this issue. Rather, we shall simply attempt to put the problem into a proper perspective so that the simplifications used to deal with the area of very low personnel static will be both credible and useful.

Figure 1: Equivalent Circuit

Figure (1) shows a highly simplified "equivalent" circuit⁷ of a person walking on a floor surface. It is presented here not as a model from which to make calculations, but as a demonstration tool to indicate the complexity of the general problem and as a basis for useful further simplification in the special case of interest here, a high degree of static control resulting in very low body voltages. The static potential on the individual is the result of the interfacial EMF’s due to triboelectrification, the surface neutralization at the foot/floor interface where the foot is down, and the flow of current through the same interface in response to a body potential to ground. While it is conventional to consider this flow, and therefore the body resistance to ground, to be the main controlling factor in limiting body voltage this is strictly true only for body charges originating from sources other than the shoe and floor. In order to segregate the effects of dissipation to ground and minimization of surface accumulation of charge, body voltages under dynamic conditions were measured with the various flooring/footwear combinations in a normal manner and also with the body insulated from ground by nonconductive shoe inserts. Maximum body voltages will be quantitatively characterized by shoe sole resistance, floor surface resistance, and body to ground resistance. It will be shown that, for the resistance levels used in static controlled systems, the surface resistances of the sole and floor are the main controlling factors.

Since our interest here is more than academic, it will be necessary to define and justify our test methodologies for both body voltage and resistances.

Test Methodologies

A. Measurement of Body Potential

Since body potential is taken in this work to be the criterion of electrostatic performance in a floor/shoe system, it is imperative that there be an understanding of the way that it is generated, measured, and reported. First, the measuring equipment must possess certain basic properties. It must not constitute in itself a significant electrical part of the body/footwear/floor system and its frequency response must be sufficient to capture the voltage pattern without distortion. Both of these conditions were satisfied by the equipment used in this work. It is of interest to note here that some test runs were performed with bandwidths of both 5Hz and 100KHz with no discernible difference in peak value readings. Peak values were determined in two ways, from a new versatile peak hold LED array instrument and by observation of an oscilloscope trace. For the body of the work, a 100KHz bandwidth was used.

The manner in which the test voltage is generated is important. Figure 2 is a sketch of the means.

A person holding a probe walks on a test surface wearing test shoes and his peak voltage is recorded. This procedure has been rather generally used for a large number of years. Except for the peak voltage criterion, it is the same as that required by AATCC TM-134¹², which calls for a damped response, and it is substantially the same as a Work in Process standard of the ESD Association. Basically, both say the same thing: if you want to find out what happens when you walk on a floor, walk on it and see.

Figure 2: Body Voltage Test Method

The differences among various methodologies presented in the literature reside in the prescribed method of walking for test purposes. While this is not usually a significant factor, it can be. In this work we have adopted a brisk natural gait for tests made in the field and a brisk, short, slightly high step for laboratory work where one can hardly walk naturally on a three or four foot square test specimen. This lab step produces, as observed on an oscilloscope, the highest voltages short of doing anything really wild.

These procedures clearly exaggerate the danger of any real human body discharge, at least statistically, as they involve high generation motions and define peak values of very short duration which, in any likely real life situation, would not be seen by a sensitive device.

B. Shoe Resistances

Three types of resistance measurement have been used to characterize the resistance properties of shoes. The ESD Association has a method S9.1⁸ which involves lining a shoe with foil, loading it with metal shot, and measuring the resistance to a base plate from the foil. We have found that this method does not adequately represent the body resistance from shoe to ground in normal wear. We have not used that method in this work. Previous work ^(2,3) has demonstrated the measurement of body resistance to ground using an actual body. This is also the general methodology of the ANSI Z-41⁹ standard and is one type of evaluation we use here. The third method is one we also use - the measurement of shoe sole surface resistance. This has been described in other work ⁽³⁾ but has been slightly modified by the use of a convenient step-on floor test unit which gives rapid, reproducible results and is suitable for on site testing. It may also be hand held as shown. This procedure is diagrammed in Figure 3.
Figure 3: Sole Surface Resistance Measurement

C. The Measurement of Flooring Resistance

It is critical that the measurement of the resistance of flooring be capable of producing reproducible results anywhere, any time, and by anyone, assuming samples of reasonably similar properties as this is a main criterion in virtually all specifications and audits. Today this is far from the case with the use of different methodologies as well as considerable variations even when nominally using the same method. As will be shown later, considerable variations in resistance properties may not be of great functional significance, but specifications must be made and standards must be met. Too often a material hovers between acceptance and rejection based on who makes the test and when. Four methods are in common use today, NFPA 99, ASTM F-150, ESD 7.1 and IEC 1340-4-1. ^6,10,11 All are quite similar and are based on NFPA 99, using identical 2 ½ inch, 5 pound electrodes (except for the IEC method which uses a 2 inch, 5 kg electrode). Aside from the differences in specified voltages and environmental conditions (100 or 500 volts, 12 or 50% RH), which could easily be reconciled by discussion, their fundamental flaw is the electrodes. They simply do not simulate the feet. The light weight and relatively hard surface of the NFPA electrodes render the resistance readings significantly sensitive to surface unevenness and even minor surface contamination. Much higher variability accompanies the use of a 100 volt test potential as in ESD 7.1, although this does seem like a reasonable level considering the intended purpose. Our test results from the lab and in the field indicate that more relevant and more reproducible results can be obtained by the use of heel electrodes rather than the 5 pound weights. The type of heel electrode used, which is now commercially available, is shown in Figure 4.

The area of the conductive contact is the same as that of the conventional electrodes and is made from heel grounder material. The pressure of the contact from body weight and the relative softness of the electrode both tend to produce a good electrical contact, just as a heel grounder or ESD shoe sole would do. Another advantage of this method is the great speed with which readings can be taken, thus encouraging large samplings which can be legitimately quantified statistically. Another important advantage is the virtual elimination of bias in floor resistance testing. Most auditors will, when they find a bad spot, wiggle the electrode a little or move it to an adjacent spot to try to get a good reading. This is probably legitimate from a functional point of view, but does leave too much room for personal judgment. It should be noted that, as described in previous work ⁽²⁾, significant surface homogeneity of resistance demands a modified electrode material for proper evaluation.

Figure 5 shows the surface to ground readings on a vinyl tile panel where 21 marked points were tested by the methodologies of ESD 7.1 and with heel electrodes, both at 100 volts.

Figure 5: Heel Electrode vs. S7.1 Electrode

The difference is obvious. It is not clear how much this difference would matter functionally, but it would certainly be important if you had a 1 or 2 Megohm specification to meet.

Figures 6 and 7 show comparative results on a much less conductive vinyl using both 100 and 500 volts with the ESD/NFPA electrodes and both surface-to-surface and surface-to-ground configurations.

It can be stated that in general that the heel electrode methodology using 100 Volts yields resistance readings which are lower and less variable than S7.1 type tests at 100 Volts. On average, the values obtained at 100 Volts using the heel electrodes tend towards those of the 500 Volt, 5 lb. electrode tests, but with less variability.

Figure 6: Heel Electrode vs. S7.1 Electrode

Figure 7: Heel Electrodes vs. S7.1 Electrodes

Figure 8 shows the ordered values for an audit-type run at a major electronic manufacturing facility.

Figure 8: Audit Data -- Surface-to-Ground

It contains approximately 285 observations made on a vinyl tiled floor. It is worthy of note for two reasons. First, because it contains so much more data than is usually amassed in a single area audit, a valid statistical analysis is possible; and second, because it takes one man with meter and computer in hand only about 30 minutes to do this job.

It should also be noted here that this installation outright failed the user’s specification for resistance and yet, with proper footwear easily met the performance requirement of body potentials below 100 volts. The quantitative explanation for this is to be found in the experimental work reported in the next section.

Body Voltage Test Results

The following data was generated in two ways. First there is laboratory work under controlled conditions and with limited sample size and second, in the field under normal working conditions. Figure 9 summarizes the laboratory data.

Figure 9: Flooring/Footwear Body Voltage Test Data

It shows peak voltages on a test individual with footwear and floor resistances defined and measured as described above under two different body /shoe relationships, normal wear and with both feet insulated from the shoes by means of high resistance polyethylene booties.

Figure 10 is a chart giving the conversion of resistance data from engineering notation to logarithms.

Figure 10: Resistance Conversion to Log R

We prefer the logarithmic form because it is a continuum which is easier for most people to grasp intuitively than powers of ten with multipliers. We further believe that it is preferable to avoid the use of such terms as "static dissipative" and "conductive" to define quantitative ranges. It would certainly be wrong to consider the nylon carpet in the laboratory to be static dissipative in the same sense as more conductive flooring, and a similar statement can be made about the plain polyurethane shoe soles, although also nominally in the SD category. The term "static dissipative" should be reserved for qualitative assessments.

Figure 11 shows quite impressively that the static development on the individual is strongly controlled by the combination of shoe sole surface resistance and floor surface resistance, as the body was effectively insulated from the floor and from ground for all of these tests (>10¹¹ Ohms).

Figure 11: Feet Insulated From Shoes

The control resides in mutual conductivity at the shoe/floor interface. This contention is reinforced by Figures 12a and 12b which for clarity have an expanded scale and the high resistance has been removed. In this case there is a significant conductive path from the body through the shoes.

Figure 12a: Shoes Worn In A Normal Manner

Figure 12b: Shoes Worn In A Normal Manner (detail)

Although the voltages are lower, they are not much lower. This brings into question the wisdom of characterizing shoe resistance by body to ground resistance alone, as is usual, when a major part of the shoe/floor control resides in the shoe/floor interface without regard to body-to-ground resistance. Body-to-ground resistance is a fairly good indicator of static performance in most cases but this is due to the usual correlation between sole surface resistance and resistance through the shoe. (See figure 9) Such a correlation need not be, however, and an ideal shoe may, in most situations, have a highly conductive sole and a very modest resistance from body to ground. There is no implication intended that vertical resistance is unimportant, simply that it may not play a major role in that static whose origin is the floor/shoe interface

Figure 13: Audit Data

Figure 13 is a depiction of the effects of foot grounding systems and mode of usage in two different carpeted assembly and test areas of a medical electronics manufacturing plant. The test subject was the ESD supervisor.

The legends on the graph are defined as:

1. "none" = no heel grounders -street shoes
2. "HG Norm" = heel grounders on both feet and walking in a normal manner
3. "HG Care" = heel grounders on both shoes and walking with extreme care to perform heel-toe movements.
4. "Toe/Heel" = each shoe had a combination heel and toe grounder attached , normal walking manner

Because the measured resistances were so far out of specification, heel grounders were required on the assumption that they would be more effective than ESD shoes. The supervisor reluctantly removed his grounders and was clocked at 1200 peak volts walking normally in the test bench area. With heel grounders his potential dropped to 250 in one installation and 450 in the other, these being the peaks when both heels left the floor, as they did with nearly every step. When care was taken not to allow simultaneous contact loss with both grounders the values were 40 and 170 volts respectively. When he used a sole grounder, which is essentially a combination of heel and toe grounders, the peak voltage in both cases dropped below 30 volts. A sole grounder creates what is really a conductive sole shoe with no opportunity for the basic, high generating sole to take part and no possibility of a momentary loss of contact. Controlling the source, the floor/shoe generator, is a major factor. Figure 14 graphs additional data from the same installation with additional footwear and operators.

Figure 14a:

Figure 14b:

Figure 14c:

From these data it can be seen that either of two pairs of shoes produced very little static. There is little choice electrically among the sole grounder and the two more conductive shoes. It is apparent that the more conductive soles were able to exploit the "local" surface conductivity of the carpet, but that’s another story.

How hazardous are the peak values associated with foot grounders? As before, we don’t really know. It’s a judgment call again, but what does a performance standard requiring personnel voltages to be maintained below 100 volts mean in a case like this, and how should it be audited?

While we have dwelled on the control of foot generated static and have shown that body-to-ground resistance plays only a minor role, other potential sources must not be neglected. If proper protocol is not followed, seating/clothing static, for example, can be considerable and calls for a body grounding mechanism whether ground strap or shoe/floor. If "good" clothing and "good" seating are used, the grounding needs are small, even as with a good shoe/floor combination. Figure 15 shows the grounding requirement for a very lively smock/seat combination.

Figure 15: Control of Clothing/Seating Static

The voltage on an individual wearing high resistance clothing was monitored as he quickly arose from a high resistance chair, keeping insulative shoes immobile on the floor, and with varying wrist grounder resistances. Peak voltages were of very short duration for resistance values below 10⁹ W and essentially disappeared for values below 10⁷ W. Does this mean that a value of less than 10⁷ W is necessary to protect against peak voltages? Again, probably not, but the offending furniture and clothing should be addressed.

Conclusions and Recommendations

Outer sole surface resistance should ideally be less than 10⁷ W . The more conductive the outer sole the better as more recombination of charge occurs (and thus lower personnel voltages). The only limiting factor should be the floor marking concerns of highly carbon loaded sole materials or other physical considerations.

The midsole construction of the shoe should be insulative enough to obviate any electrical safety concerns.

Users find it confusing that the resistance limits for defining conductive floors are very different from those for defining conductive packaging. Terms like "dissipative" and "conductive" are not necessary nor helpful. A continuous scale of resistance should used..

Present methodology for determining flooring resistance is inadequate.