Dent Mater 10:269-274, July, 1994

Wear characteristics in a two-body wear test

Robert W. Wassell, John F. McCabe, Angus W.G. Walls

Department of Restorative Dentistry, University of Newcastle Upon Tyne, England, UK

ABSTRACT Objectives. A previous report compared spherical steatite (ceramic enamel substitute) abraders with those of natural enamel in a two-bedy wear test. The wear rates and coefficients of fric- tion of the two abraders against various composites and an amal- gam were well correlated although the wear rates were slightly higher with steatite.This report investigates the characteristics of the wom abrader and specimen surfaces. Methods. Scanning electron microscopy and laser profilometry were used. Results. Similar wear characteristics were found for the two types of abraders. Adhesive wear was evident for the amalgam, Dispersalloy (Johnson & Johnson), and the heat/pressure-cured microfill composite, Isosit (Ivoclar-Vivadent). Abrasion was seen with the hybrid composite, Occlusin (ICI), and, to a lesser extent, the microfill composite, Heliomolar (Ivoclar-Vivadent). The ap- pearance of the worn small particle hybrid composite, Brilliant Dentin (Colt~ne), suggested that fatigue and delamination were involved. Laser profilometry showed that the hybrid composites caused much greater wear to the abraders than either the microfill composites or amalgam. The Ra values of the worn abraders and specimens were similar, suggesting conformal contact between them and endorsing the well controlled conditions of the wear test. Significance. The results of this and other publications suggest that steatite can be used as an alternative to enamel in perform- ing two-body wear tests on dental composites. This should help significantly in materials evaluation and development.

INTRODUCTION Enamel abraders are unsuitable for standardized wear test- ing. To overcome the inconsistencies due to differences in morphology, composition and microstructure, an alternative abrader material was sought. Steatite (a ceramic enamel analogue) abraders, available as 9.5 mm spheres, were cho- sen as they had a Wickers hardness similar to enamel. The spherical abrader shape was considered superior to a cylin- drical or conical one as it avoided edge loading and plowing of the specimen (Burwell and Strang, 1952).

A two-body wear test for restorative materials, devised to be carried out in a universal testing machine, has been re- ported (Wassell et al., 1994). That report compared spherical steatite abraders with those of natural enamel. The initial findings showed that steatite might provide a good enamel substitute: the wear rates and coefficients of friction of the two abraders against various composites and an amalgam were well correlated although the wear rates were slightly higher with steatite.

This report is to supplement those wear rate measurements by examination of the worn abrader and specimen surfaces. Such an examination may elucidate the principal wear mecha- nisms and help determine if similar wear mechanisms resulted with the two types of abrader. Clearly, if a wear test is to be acceptable for general use, the substitute enamel abrader should produce wear in a sfinilar way to enamel.

Scanning electron microscopy (SEM) has been used previ- ously for the examination of worn surfaces (McKinney and Wu, 1982; Lutz et al., 1984; Rice et al., 1984). Mechanical profilometry has also been used (Atkinson et al., 1978), but this technique is difficult to use on spherical abraders because of the problems of reliably tracking a stylus across a curved surface. More recently, however, laser profilometers have been introduced which should be much better suited to profiling the spherical abraders used in this study.

Thus, the aims of this study were to examine representa- tive specimen wear scars and abrader facets with both SEM and laser profilometry to determine the wear characteristics, measure the surface roughness and estimate the amount of abrader wear by superimposing the contour tracing of an unworn abrader over that of the worn abraders.

MATERIALS AND METHODS Two hybrid and two microfilled composites were selected (Table 1) along with a dispersed phase amalgam (Dispersalloy, Batch No. 031288A, Johnson & Johnson, East Windsor, NJ, USA). The composites were chosen primarily to give a wide variety of matrix systems; one material was based on a Bis-GMA matrix (Brilliant Dentin), another on a urethane

Dental Materials~July 1994 269

TABLE 1: COMPOSITES USED IN THE STUDY

Composite Type Shade Manufacturer Occlusin Large particle hybrid S ICI, Macclesfield, UK

Light-cured

Brilliant Dentin Small particle hybrid B2 Colt~ne Ag., Altst&tten Light-cured Switzerland

Heliomolar RO Microfilled composite U Ivoclar-Vivadent, Schaan Light-cured Liechtenstein

Isosit Microfilled composite 3A Ivoclar-Vivadent Heat and pressure-cured

sputter~oated for SEM. These results were prepared for illustrative purposes and were

Batch No. _ not subjected to statistical analysis. UM61A Laser profilometry is a new technique

which avoids the problems of stylus con- tact. The UBM 16 (Ulrich Breitmeier

180989-01 Messtechnik GMBH, Ettlingen, Germany) is a laser profilometer used in industry. The principle of this instrument is as

436901 follows: light from a semiconductor laser BG 32 in the sensor is directed onto the object sur-

face as a focused spot. A moveable lens, 20285031 suspended in the sensor is continuously ad-

jnsted to ensure that the focal spot of the beam is always coincident with the object surface. As a consequence, the laser spot

size does not vary over the measurement range. The sensor is designed such that any displacement of the object surface in the direction of the beam causes a displacement of the image of the surface incident spot on a photodetector. The output of this photodetector is then used to control the position of the moveable lens, and a second measurement system connected to the lens provides the desired surface displacement mea- surement (UBM technical data). This laser profilometer has the capacity to show surface contour and provide roughness parameters (ISO 468-1982) as well. Surface contours can be displayed as a three-dimensional representation on a monitor and selected transverse profiles made.

Specimens were aligned so that the profilometer scanned transversely to the direction of abrader movement. Each specimen wear scar and abrader facet was scanned in 50 prn increments and the three-dimensional representation dis- played on a monitor. The center of the wear scar or facet was determined. The stored electronic data from the scan in this area was then used to print a transverse profile and provide roughness characteristics (width limited electronically) of the specimen wear scar and abrader wear facet.

The specimens were then sputter-coated with 20 nanom- eters of gold and viewed by secondary emission electron imag- ing and backscattered electron emission to obtain representa- tive micrographs at 27x and 1 K magnification (Cambridge Stereoscan $240, Leica Cambridge, Cambridge, UK ).

TABLE 2: MEAN WEAR RATES (pro/1000 CYCLES) AGAINST STEATITE AND ENAMEL (STANDARD DEVIATIONS)

Material Steatite Enamel

n=3 n=3

Brilliant Dentin 3.73 (0.38) A 1.86 (0.34) a

Occlusin 2.94 (0.38) B 1.52 (0.70) ab

Heliomolar 2.01 (0.45) c 1.30 (0.38) ~b

Isosit 1.41 (0.08) ° 1.25 (0.07) ab

Dispersalloy 0.89 (0.21)D 0.62 (0.05) b

Means with same superscript are not significantly different (p < 0.05).

dimethacrylate (Occlusin) and two based on Bis-GMA/ure- thane dimethacrylate blends (Heliomolar and Isosit). Isosit was heat-and-pressure-cured while the other three materials were light-cured. Disc specimens (9 mm diameter, 2 mm deep) were made in acrylic blocks. A detailed description of specimen preparation and the wear machine has been given elsewhere (Wassell et al., 1994). Essentially, the wear machine consisted of a device which could be mounted in a screw-driven universal testing machine (Instron, High Wycombe, UK). The abrader was attached to the crosshead (speed: 50 mm/min) and the specimen immersed in a water bath (37 ° C) supported by the load cell. The abrader was ap- plied against the specimen with a 15 N force (produced using a 20 N vertical force and a frictionless hinge) and reciprocated up and down with an amplitude of 2 mm. Frictional forces were derived from the chart recording of the load cell output. Coefficient of friction was readily calculated from the frictional force (recorded throughout the test) and the normal force (con- stant). Measurements of specimen wear scar depth were made every 1000 cycles up to 10,000 cycles with a mechanical profilometer (Surfometer Type SF 101, Planer Products, Sunbury on Thames, UK). Wear rates (pm/1000 cycles) between 1000 and 10,000 cycles were derived from the gradi- ents of the regression analyses of depth of wear scar against the number of cycles. The mean wear rates measured for each material for beth enamel and steatite abraders were com- pared using one-way analysis of variance and a Student Newman Keuls multiple range test. Significant differences were determined at the 5% level. Following wear testing, a representative example of opposing specimen and abrader from each material were profiled with a laser prefilometer and then

RESULTS

A summary of the wear rates of the composites and amalgam is shown in Table 2. The laser profilometer was designed to operate primarily with essentially flat specimens. A deviation from "flatness" could readily be tolerated, but surfaces curv- ing more than 0.5 mm above or below the horizontal line of travel of the laser spot could only be profiled over a limited area. This made it difficult to profile the unworn aspects of the enamel abraders. Profile results were therefore restricted to the steatite abraders and their opposing specimens. The trans- verse profiles of the specimen wear scars and steatite abrader facets were assembled to form composite tracings of each opposing specimen and steatite abrader (Fig. 1).

To indicate how much of the abrader had been lost during the experiment, a profile of an unworn abrader was superim- posed as a dotted line. The two hybrid composites (Brilliant Dentin and Occlusin) were clearly very destructive to the abrader. A small amount of abrader wear was noted with the

270 Wassell et aL/Two-body wear test

.-g ,_..=

'° 1 60

50

40

3 0 "

2 0 '

1 0

0

- 1 0

- 2 0

- 3 0

- 4 0

- 5 0

- 6 0

- 7 0 DEN

E o

o Q.

7o 1 60

50

40 , ,

30 , ,,

20 ~, '

10 x ' • /

-'10 [mml

-20

-30

- 4 0

-50

-60

-7o O C C

70

60

50

40

3 0 -

2 0 -

1 0 -

• 0

"~ -lo. - 2 0 -

- 3 0 -

- 4 0 -

- 5 0 -

- 6 0

- 7 0

~ s 70

60

50

40

30

2O

1 0

, I J 0 3 r=

[mm] ~ - l o

-20

-30

-40

-50

- 6 0

HM -to

[rnrn]

ISO

Fig. 1. Laser Profilometry: Transverse profiles of worn steatite abraders positioned above their respective specimens. The outline of an unworn abrader is superimposed (dotted line) to show the amount of abrader wear. No abrader wear was detected for Dispersalloy. Den = Brilliant Dentin, Occ = Occlusin, Hm = Heliomolar, Iso = Isosit.

microfilled composites (Heliomolar and Isosit) and virtually none with the amalgam.

Despite the problem of obtaining transverse profiles of the unworn aspects of the enamel abraders, it was possible to get surface roughness profiles of the wear facets of both types of abrader. Roughness values were calculated automatically, and Table 3 shows the Ra values of some abraders and their corresponding specimens. The roughness parameters varied between materials, but similarities were noticed between some abraders and their opposing specimens. This similarity was especially evident where the SEM had shown evidence of

material transfer between speci- men and abrader.

Typical micrographs of the wear scar for a hybrid composite and the resulting abrader facet (Brilliant Dentin/Steatite Ball) are shown in Fig. 2. The scars for the microfilled composites and amalgam were smaller and smoother. Such sur- face morphology was best demon- strated with secondary electron emission imaging.

Backscattered electron emission was used to show material trans- fer from specimen to abrader. Back- scattered emission shows areas con- taining elements with a high atomic number (e.g., mercury) as light, while areas of low atomic number (e.g., carbon) are dark. Transfer of amalgam onto steatite is shown in Fig. 3 as a white area, and this feature was also seen with enamel. Dark areas, representing a resin smear, were seen on the ste- atite abraders worn against Isosit (Fig. 4). Representative electron photomicrographs at 1000x mag- nification were recorded (Figs. 5 to 7), first, to compare the effects of enamel and steatite abraders, and second, to help determine wear mechanisms.

Brilliant Dentin (Fig. 5) showed a remarkably similar appearance against both enamel and steatite.

The surface was granular, revealing the filler structure, but with smooth patches overlayed. These patches had irregular edges, a cracked surface and were approximately 1 lxrn thick. These smooth areas represented the smeared and degraded surface prior to delamination as wear particles. Wear par- ticles were seen on the specimen and on both types of abrader (platelet-like, 1 - 5 ]am across). The abraders were gouged by tracks aligned with the direction of movement. These tracks were deeper in steatite than in enamel, which may be ex- plained by the release of large abrasive particles from the ste- atite as a result of localized surface breakdown.

Fig. 2. SEM micrographs of steatite abrader (left) worn against Brilliant Dentin specimen (right). Notice the elliptical wear facet in the spherical abrader and the elongated wear scar in the fiat specimen (field widths 4 ram).

Fig.3. SEM micrograph of a steatite abrader facet worn against Dispersalloy (2 mm field). Back- scattered electron imaging shows the transfer layer of amalgam (light patch).

Dental Materials~July 1994 271

Fig. 4. SEM micrograph of a steatite abrader facet worn against Isosit (field width t mm each side). Backscattered electron imaging (right side) revealed the smear layer as a dark patch.

Fig. 5. SEM micrographs of Brilliant Dentin (A) worn against an enamel abrader (B) and Brilliant Dentin (C) worn against a steatite abrader (D) (Horizontal field widths 75 pm).

Fig. 6. SEM micrographs of Occlusin (A) worn against an enamel abrader (B) and Occlusin (C) worn against a steatite abrader (D). The backscattered electron im- ages of the specimen (E and F) highlight the worn filler particles. (Horizontal field widths 75 pm).

Occlusin (Fig. 6) also gave a similar picture of wear against the two types of abrader. Again, there were granular lower areas with slightly raised smooth zones. Backscattered elec- tron imaging clearly demonstrated the pattern of filler par- ticles beneath the smooth zones and suggested that there was a thin resin smear across the large filler particles worn against enamel. This smear was not present in the specimen worn against steatite. Both scars had localized areas of severe sur- face disruption where large filler particles had been wrenched out of the composite surface: cracks were evident around some of the filler particle/matrix junctions indicating the potential for further disruption. Wear particles of up to 10 pm across were seen on the steatite abrader, although the enamel abrader retained only small particles of up to 2 pm within its smooth grooved surface. Large wear particles were observed on the steatite abrader, again due to localized breakdown of its other- wise smooth surface.

@ Fig. 7. SEM micrographs of Heliomolar (A) worn against an enamel abrader (B) and Heliomolar (C) worn against a steaUte abrader (D). (Horizontal field widths 75 pm).

Heliomolar (Fig. 7) gave a different pattern of wear to the two hybrid composites. The specimen worn by the enamel abrader showed abrasion marks both running with and per- pendicular to the direction of movement. Most of the surface was smooth with a high density of wear particles (1 - 5 pm). In other areas, however, the main wear grooves were transected by crescent-like ridges of the type described by Atldnson et al. (1978) in the wear of ultra high-molecular weight polyethylene used for joint protheses. They suggested that this pattern was due to the rolling of wear particles. The specimen worn by the steatite abrader had a similar pattern to that described above although the surface was more heavily gouged in the direction of movement. The resulting wear particles were accreted on the surface of the steatite abrader. No such accretions were seen on the enamel abrader although there was a suggestion that thin striae of smeared material had formed in the fine wear grooves.

272 Wassell et aL/Two-body wear test

Isosit gave similar results to Heliomolar, but with two im- portant differences. First, on specimens worn against beth types of abrader, there were no signs of the transverse abra- sion grooves seen in Heliomolar. Second, the density of wear particles on specimens and abraders was less. Backscattered electron imaging was used to good effect to demonstrate the resin-rich smear layer on the surface of the steatite abrader (Fig. 4).

Dispersalloy specimens did not provide any meaningfxfl results as some chemical reaction had taken place on the wear scar surface which obscured detail. On the other hand, beth types of abrader, as already illustrated, revealed amalgam transfer.

DISCUSSION The micrographs provided a qualitative assessment of the wear mechanisms (Lansdown and Price, 1986) and, for each restor- ative material, the two abraders showed a similar picture. However, the surfaces of the steatite abraders and their op- posing specimens were more heavily grooved, possibly as a result of the coarser abrader surface. This finding may ex- plain the higher wear rate associated with the steatite abrad- ers.

Adhesive wear was evident for Dispersalloy, Isosit and to a lesser extent, Heliomolar. Backscattered electron imaging clearly demonstrated the transfer of material from specimen to abrader which is characteristic of this type of wear. Another characteristic of adhesive wear is a relative lack of wear par- ticles which was certainly seen with Isosit. Heliomolar, how- ever, had a relatively large number of wear particles on its surface which gave strong evidence of abrasion.Abrasion also affected the two hybrid composites as shown by the numer- ous wear particles and grooving of the specimen wear scar.

There was no sign of the onset of fatigue wear as shown by the sudden change in wear rate reported by Atldnson et al. (1978) and McKinney and Wu (1982). Nevertheless, the worn surface of Brilliant Dentin (Fig. 5) suggested that surface fa- tigue, through a process ofdelamination (Suh, 1977), had taken place. The delamination theory proposes that as a rider (abrader) moves over a surface, the surface is plastically de- formed and microcracks are generated within the material. These microcracks eventually coalesce, resulting in a crack parallel to the surface and the eventual loss of a wear particle. In this process, the Hertzian stresses developed below the sur- face will obviously depend on the applied force and geometry of the contacting bodies (Reid et al., 1990). Further work, per- haps using silver staining (Wu and Cobb, 1981; Wassell et al., 1992), is needed to determine if subsurface damage had oc- curred and if so, how the damage relates to applied abrader force.

Laser profilometry offers considerable advantages over con- ventional profilometry.. Conventional profilometry, which per- forms well on flat specimens, cannot easily be used on curved surfaces such as the spherical abraders employed in this study. Moreover, the slight plowing effect of the profilometer stylus may influence the roughness values obtained. The non-con- tact technique of the laser profilometer overcame these prob- leins and provided useful data on the amount of wear of the abrader and the roughness of the worn surfaces.

Willems et al. (1992) used a laser profilometer to monitor the wear produced by a vibrational wear machine using a spherical chromium steel counterbody against various dental

TABLE 3: Ra VALUES (pm) FOR ABRADERS AND CORRESPONDING SPECIMENS

Steatite Enamel Abrader Specimen Abrader Specimen

Brilliant Dentin 1.16 0.71 0.47 0.29

Occlusin 1.62 1.41 040 0.52

Heliomolar 0.39 0.39 0.77 0.37

Isosit 0.40 0.34 0.35 0.23

Dispersalloy 0.48 0.46 0.78 0.82

The Ra values shown in bold indicate a similarity in surface roughness between specimen and abrader.

composites. That study, although well controlled, produced only limited intraoral simulation; the steel abrader was run at 8 Hz under dry conditions. Interestingly, Willems et al. (1992) calculated the wear volume of the spherical abrader assuming this represented a segment of a sphere, in other words, assuming the abrader facet was flat in cross section. In the present study, laser profilometry (Fig. 1) clearly demon- strated that the abrader facet was either curved or irregular, but never flat. Using the same assumption as Willems et al. (1992), the abrader wear volume results from this study would have been grossly overestimated for Heliomolar and Isosit. However, because of the problems outlined and the limited number of abraders profiled, wear volume was not calculated. The amount of abrader wear was estimated simply by super- imposing a profile through the center of an unworn abrader over the central profile of each worn abrader.

The abraders opposing the hybrid composites showed much heavier wear than those opposing the microfills or amalgam. These findings agree with those of Burgoyne et al. (1991). The heavy wear of the steatite abraders opposing the hybrids was associated with a release of wear particles from the abrader which may have contributed to the higher rate of specimen wear than seen with enamel. Such wear particles may have contributed to the slightly rougher composite specimen sur- faces opposing the steatite abraders (Table 3). Nevertheless, despite these minor differences, the readeris reminded (Wassell et al., 1994) that wear rates and friction against enamel and steatite were well correlated (r = 0.94 and r = 0.98, respec- tively).

Table 3 gives typical values of Ra for matched pairs of abraders and specimens. In this case, the individual results are cited in preference to mean values and ranges. It is the direct comparison of roughness for each abrader and oppos- ing specimen that is important here. Some of the materials showed Ra values similar to those of their opposing abraders. This similarity was particularly striking with steatite opposing amalgam, Isosit and Heliomolar, and with enamel opposing amalgam. On the micrographs, all these combina- tions showed evidence of material transfer from specimen to abrader. The similarity of Ra values may be explained by this transfer in combination with a well controlled experimental set-up producing conformal contact. Conformal contact im- plies that the surfaces were closely opposed and that no plow- ing, through edge loading, had occurred.

Dental Materials/July 1994 273

This investigation has helped elucidate the types of wear characteristics which can occur in a two-bedy wear experi- ment. Steatite and enamel abraders showed similar charac- teristics. Indeed, the results reported here offer more support for the use of steatite as a model material for studying the wear of composites opposed by enamel. Further work is needed to determine how these wear patterns relate to those experienced intraorally.

ACKNOWLEDGMENTS

This study was based on part of a Ph.D. thesis submitted to and accepted by the Subfaculty of Dentistry, University of Newcastle upon Tyne. The authors would like to thank Dr. John Nunn, Dr. Paul Reynolds and Christine Shuttle, of Courtaulds Coatings, Gat~head, Tyne & Wear for their kind help with the laser profilometry. The SEM work was carried out by Dr. Trevor Booth in the Biomedical Electron Micros- copy Unit, University of Newcastle Upon Tyne. The compos- ites were kindly donated by the respective manufacturers.

Received May 3,1994/Accepted June 23, 1994

Address correspondence and reprint requests to: Robert W. Wassell Department of Restorative Dentistry The Dental School University of Newcastle Upon Tyne Newcastle Upon Tyne NE2 4BW England, UK

REFERENCES

Atkinson Jl~ Brown KJ, Dowson D (1978). The wear of high molecular weight polyethylene. Part I: The wear ofisotro- pic polyethylene against dry stainless steel in unidirectional

motion. Trans Am Soc Mech Eng (J Lubric Tech) 100:208-218.

BurgoyneAR, Nicholls JI, Brudvik JS ( 1991).In vitro two-body wear ofinlay-onlay composite resin restoratives. JProsthet Dent 65:206-214.

Burwell JT, Strang CD (1952). On the empirical law of adhe- sive wear. J Appl Physiol 23:18-28.

ISO 468 (1982). Surface roughness - Parameters, their values and general rules for spedfying requirements. International Organization for Standardization, Geneva, Switzerland.

Lansdown AR, Price AL (1986). Materials to Resist Wear. Ox- ford: Pergamon Press, 3.

Lutz F, Phillips RW, Roulet JF, Setcos JC (1984). In vivo and in vitro wear of potential posterior composites. J Dent Res 63:914-920.

McKinney JE, Wu W (1982). Relationship between subsur- face damage and wear of dental restorative composites. J Dent Res 61:1083-1088.

Reid CN, Fisher J, Jacobsen PH (1990). Fatigue and wear of dental materials. JDent 18:209-215.

Rice SL, Bailey WF, Roto M, Wayne SF (1984). Wear behavior of a composite restorative and various styli in sliding con- tact. J Dent Res 63:932-935.

Suh NP (1977). An overview of the delamination theory of wear. Wear 44:1-15.

Wassell RW, McCabe JF, WallsAWG (1992). Subsurface defor- mation associated with hardness measurements of com- posites. Dent Mater 8:218-223.

Wassell RW, McCabe JF, Walls AWG (1994).A two-body fric- tional wear test. JDent Res 73:1546-1553.

Willems G, Celis JP, Lambrechts P, Braem M, Roos JR, Vanherle G (1992). In vitro vibrational wear under small displacements of dental materials opposed to annealed chro- mium-steel counterbodies. Dent Mater 8:338-344.

Wu W, Cobb EN (1981). A silver staining technique for inves- tigating wear of restorative dental composites. J Biomed Mater Res 15:343-348.

274 Wassefl et aL/Two-body wear test