THE EFFECT OF VARIOUS DEBURRING MEDIA CONTAINING MINERAL
SILICON (SiO2) OR ALUMINUM OXIDE (A1,0.) ON CONTACT PERFORMANCE
Particle fragmentation and embedment of mineral silicon (SiO2) or aluminum oxide (A1,0,) introduced by deburring electrical contacts inhibit metallic conduction. The relationship between the concentration of mineral particles embedded on the rivet’s surface and contact performance was investigated. The fine silver rivets were tested in a dry circuit application and switched under various arcing properties to rate their performance.
The investigation elucidated that the attainment of low contact resistance in dry circuit applications was achieved independent of concentration and size of insulating particles embedded on the fine silver rivets at contact pressure less than 50 grams(0.5N). The contrary was observed when rivets were subjected to 12.5 VDC, 0.5 Amps and 2 Amps load. Degradation of the contact spot was directly related to the formation of a carbon film by absorption of organic vapor during switching, and not by decomposition of insulating particles (SiO, or Al2O2) on the rivets surfaces. The degradation process was significantly accelerated by varying arcing properties. The stable contact resistance obtained in dry circuit applications and erratic millivolt drop during electrical switching were explained as interactions among a concentration of insulating particles embedded on the fine silver rivets, particle size, and carbon film formed on the contact spot after switching.
Electrical contacts are subjected to numerous manufacturing steps, which can cause surface contamination. Rivets are polished to remove burrs and obtain smooth surfaces. To enhance the effect of this treatment, deburring media are utilized. A common source of contamination is mineral silicon used for deburring contacts after they have been fabricated. The presence of foreign matter on the surface of electrical contacts is usually the cause of high resistance failures. Electrical contact is a junction between two or more current-carrying members which provides electrical continuity at their interface . Often, the rivets appear contaminated and no failure results, or on the open circuit the increased contact resistance is detected without any visible contact contamination.
The most common types of silicon compounds that have been found as constraints on contacts are organic silicon and organosilicon compounds and silicone. Early research by various investigators has shown many ways for silicone contamination to take place on electrical contacts resulting in detrimental effects on contact performance. Moberly , showed silicone transfer by vaporization when heating silicone oil to 200°C. Witter and Leiper , showed that silicone mineral contamination caused a significant increase in contact resistance, but did not cause loss of electrical continuity in devices operating at 1 newton of pressure and more than 100 volts. Trächslin (4), described finding silicate particles embedded on composite contact rivets from several contract manufacturers.
Trace contamination from base metals ordinarily would not cause low-level failure if contaminants remained pure metal. It is the subsequent change to insulating oxides or nitrides that ultimately prevents conduction. Loose particles or unoxidized oil films are easy to remove, but the removal of metallic inclusions or highly compacted/embedded particles from deburring media is difficult indeed.
The effects of surface contamination on electrical contacts have been studied for many years. However, prevention of surface contamination caused by deburring media containing mineral silicon (SiO2) and aluminum oxide (Al2O2), by manufacturers of electrical contacts seemed to be so obvious that papers on this subject could not be found. Embedment of particulate debris from the above process will degrade contact resistance and cause premature failure of the electrical contacts. The objective of this investigation is to further elucidate that even when rivets are contaminated with insulating particles, the driving force for degradation of the contact spot occurs only when an arc is sustained, due to absorption of organic vapors.
2.1 Experimental Conditions
The surfaces of the deburred rivets were subjected to quantitative surface analysis using Energy Dispersion Spectroscopy (EDS) techniques. Each contact was analyzed
The rivets under consideration were manufactured from fine silver (99.9% Ag) under normal tool conditions [15 microinches (Ra)]. They had radius head configuration and contained spherical asperities from the tool surface. The surface finish of test specimens was measured with a Mitutoyo Surftest-201 surface measuring instrument .
Once the rivets were manufactured with the desired surface characteristics, residual oils on the contact surface from the cold heading operation were removed by a conventional metal finishing process (burnishing). To assure that rivets were free from particulate debris, they were etched in a cyanide solution for 2 minutes or until a frosty appearance was visible.
The rivets were subdivided into smaller lots for contamination through tumbling with deburring media containing mineral silicon (SiO2) and aluminum oxide (Al2O2). The deburring process was performed with an open tumbler half-filled with deburring media and deionized water.
Chemical analysis of deburring media and rivets surface
The deburring media used in the investigation were crushed to develop fine powders (0.1 microns). These powders were then subjected to analysis using energy dispersion spectroscopy (EDS) techniques, utilizing an electron beam having an accelerating voltage of 5Kev. Table 2.2.1 depicts the quantitative results of analyzed powders in weight percent.
Rivets were deburred in an open tumbler half-filled with media containing various percentages of mineral silicon (SiO,) and aluminum oxide (A1,0z) and deionized water for 1 hour. They were then submitted to burnishing for 30 minutes.
The surfaces of the deburred rivets were subjected to quantitative surface analysis using Energy Dispersion Spectroscopy (EDS) techniques. Each contact was analyzed with an accelerating voltage of 10 Kev.
Each specimen was tilted 22 degrees and an acquisition time of 500 seconds was used. An area of approximately 1000 microns square near the center of each rivet was analyzed. This area best characterized the entire contact surface. The results of these analyses were quantified and approximate weight percent values were determined for aluminum oxide and mineral silicon as depicted in Table 2.2.2.
2.3 Technique and Apparatus
Contact resistance was measured employing dry circuit techniques. The primary requirement for dry circuit testing of contacts has been previously described . Figure 2.3.1 illustrates the contact resistance apparatus. The apparatus consists of fixtures for holding specimens of various sizes and shapes, and a Keithly 580 micro-ohm meter. A similar set up was used by Antler  and Nobel . A mechanism applies a measurable load which can be increased, decreased or held constant. Only one contact spot was randomly assessed during testing. However, specific contact spots could be assessed employing the stage micrometer to move the probe holder. The reference surface (the probe) was manufactured from fine silver with a conical shape (gold diffused and gold treated).
The probe holder has been designed so force may be applied to the contact. An electrical load force gauge ACCU FORCE II is mounted on the top of the probe holder, A calibrated spindle on the side of the stage is advanced manually, while the electronic gauge displays the load applied to the sample. One of the micro-ohm meter voltage leads was attached to the probe holder and the other lead to the sample holder. A minimum of seven rivets from each interval was submitted to contact resistance probing. After each measurement, the probe was gently wiped with tissue assuring a debris-free surface similar to the technique used by Russell .
2.4 Life Test
The various groups of rivets were tested in the experimental switching device shown in Figure 2.4.1. A power source allowed for adjustment of electrical parameters (500 Amps, 40 Volts max). The experimental switch gave cross-bar type make and break with closing force control by pre-adjusted terminals after assembly was completed. The operation parameters utilized in these tests were:
The experimental device was mounted on a rotary table, and cam-actuated, to push the plunger to open a set of contacts. Each device was actuated and deactivated at a rate of 16 times per minute. The electrical load for each pair of contacts was achieved by adjusting one wire wound 25 ohms resistor to obtain 0:50 amps and another 6.25 ohms resistor to obtain 2 amps resistance load and an open-circuit voltage of 12.5 volts DC. The voltage drop across the contacts was measured by using a digital voltmeter across the terminals. The circuit continuity after contact closure was monitored initially and after 10 and 10,000 cycles.
3.1 Effects of mineral silicon contaminated rivets on static contact resistance in dry circuit applications.
Figure 3.1.1 depicts in a graphical form the average static contact resistance of deburred rivets in various media. The static contact resistance values of the deburred and burnished rivets were compared to the contact resistance values of those rivets that were only burnished. The static contact resistance value yielded at 25 grams of contact pressure was 1.15 milliohms. When the contact pressure was increased to 50 grams, the static resistance value decreased to 0.94 milliohms. After rivets were deburred for 1 hour with the various media, the static contact resistance values yielded for test conditions 2 through 5 were 1.32, 1.08, 1.13 and 1.19 milliohms. At 50 grams of contact pressure, the static contact resistance values yielded were 1.05, 0.91, 1.00 and 0.97 milliohms respectively.
The cause for such negligible static contact resistance is best explained by the surface analysis of the deburred rivets. Figure 3.1.2 depicts the typical sizes and concentration of fragmented deburring media embedded on the rivet’s surface. The 1000 microns square examined near the center of the rivet revealed insulating particles scattered throughout the area. The insulating particles embedded on the rivet’s surface were approximately 0.25 microns in diameter.
A graphical representation of the average millivolt drop is depicted in Figures 3.2.1A and 3.2.1B. The device was actuated at 0.5 amps or 2 amps at 12.5 VDC. The millivolt drop measured initially and after 10 cycles ranged from 4 milliohms to 10 milliohms respectively. The contact spot degradation was directly related to the arcing properties and number of operations used. After 10,000 operations the millivolt drop for the various conditions tested ranged from 200 millivolts to 500 millivolts, as illustrated in Figure 3.2.1A. Actuating the device with 2 amps of load yielded 18 millivolt drop initially for all test conditions and after 10 cycles ranged from 25 millivolts to 40 millivolts respectively. The millivolt yield after 10,000 cycles ranged from 150 to 200 millivolts for the various test conditions investigated, as illustrated in Figure 3.2.1B.
The contact spots of the rivets from selected test conditions were analyzed by energy dispersion spectroscopy (EDS) for carbon and mineral silicon. The peak ratio of carbon and relative weight percent of mineral silicon and aluminum oxide are depicted in Figures 3.2.2A and 3.2.2B. When the device was actuated with 0.5 amps, the carbon present in the contact spot was 0.063 peak percent for all test conditions. The concentrations of mineral silicon and aluminum oxide ranged from 0.2 to 0.58 and 0.0 to 0.2 weight percent, as illustrated in Figure 3.2.2A. As expected, mineral silicon and aluminum oxide were not detected in test condition 5. Increasing the load to 2 amps allowed for a substantial increase in the carbon formation on the contact spot. The carbon peak ratio increased to 0.125% as illustrated in Figure 3.2.2B. The mineral silicon and aluminum oxide at the contact spot remained exactly the same as the devices actuated with 0.5 amps.
3.3 Scanning Electron Microscope Examination
Energy Dispersion Spectroscopy (EDS) was performed at an acceleration voltage of 10 Kev. This provided identification of typical contaminant material. Light elements such as carbon, oxygen, and nitrogen could be identified with the use of an ultra-thin window.
Figures 3.3.1 A, 3.3.1 B, 3.3.1 C, and 3.3.1 D depict scanning electron microscope photographs of the contact spot after actuating the device at 0.5 amps and 2 amps. The contact spot exhibited a gradual increase in carbon formation directly related to the arcing properties used. Figures 3.3.2 A, 3.3.2 B, 3.3.2 C, and 3.3.2 D depict the contact spot after ultrasonic cleaning with acetone. The surface of the contact spot revealed some remaining carbon and very fine fragments of the insulating particles embedded on the surface. These mineral particles were observed on rivets in test condition 2 regardless of the arcing properties used, as illustrated in Figures 3.3.2 A and 3.3.2 B. The contact spot of rivets from test condition 5 did not exhibit any mineral silicon or aluminum oxide on its surface, as expected. Only scattered carbon debris was detected as illustrated in Figures 3.3.2 C and 3.3.2 D. The surface analysis of rivets from test condition 4 revealed identical surface morphology as observed in those rivets from test condition 2.
The results illustrated the behavior of the contacts when subjected to dry circuit applications or actuated with various arcing properties. The rivets tested in dry circuit applications attained low static contact resistance very readily. In fact, those rivets purposely contaminated with the various deburring media containing different concentrations of mineral silicon and aluminum oxide yielded identical static contact resistance values compared to those rivets manufactured and burnished in the appropriate solution.
In dry circuit applications, the most important factor contributing to the attainment of low static contact resistance is a clean contact surface. Although the surface of the rivets was contaminated with deburring media containing mineral silicon and aluminum oxide in various concentrations, the fragmented particles were scattered throughout the rivet’s surface as indicated in Figure 3.1.2. The actual fragmentation of the deburring media was 0.25 microns in diameter and very similar in all conditions tested. The latter condition allowed for the attainment of low static contact resistance regardless of the concentration of mineral silicon and aluminum oxide in the deburring media employed.
The initial electrical switching of the contacts produced insignificant millivolt drop but increased when tested under the various arcing properties. The contact spot progressively degraded as the number of operations increased to 10,000 cycles. Degradation of the contact spot was greatly influenced by the arcing properties used. As the current increased from 0.5 amps to 2 amps, the contact spot revealed evidence of some cleaning of the carbon film. Although the carbon film was slightly broken by subsequent arcing striking the rivet’s surface, the remaining carbon film adhered to the contact spot prohibiting metal to metal conduction. As a result, the millivolt drop measured at the end of 10,000 operations was considered an open circuit for both cases. However, the scanning electron microscope revealed that the concentration by weight percent of insulating particles presents on the contact spot prior to and after electrical switching was identical. This indicated that absorption of organic vapor during arcing produced the carbon film which adhered to the contact spot, instead of decomposition of the deburring media.
The effect of rivets contaminated with deburring media containing various concentrations of mineral silicon and aluminum oxide on dry circuit applications and degradation of contact spot during switching can be explained as follows:
The attainment of low static contact resistance was achieved in dry circuit applications regardless of the concentration of mineral silicon and aluminum oxide embedded on the rivet’s surface at contact pressure less than 50 grams.
During switching contact spot degradation was the result of absorption of organic vapor which formed carbon on the contact surface, instead of decomposition of the deburring media embedded on the rivet surface with subsequent arcing.
Degradation of the contact spot was accelerated by increasing the current from 0.5 amps to 2 amps and the number of cycles to 10,000.
Scanning electron microscopic examination revealed that insulating particles embedded on the rivet’s surface were approximately 0.25 microns in diameter and the concentration by weight percent was identical prior to and after actuating the device.