J. Dent. 1991; 19: 312-316
Strain gauge method for measuring polymerization contraction of composite restoratives R. L. Sakaguchi, Biomaterials Minneapolis,
C. T. Sasik, M. A. Bunczak and W. H. Douglas
of Oral Science,
ABSTRACT Post-gel polymerization contraction of composite restoratives produces a volumetric change in phase with the development of a modulus of elasticity and distributes contractile stresses through the resin hard tissue interface into the tooth. A new method for monitoring the polymerization contraction of composite restoratives utilizes electrical resistance strain gauges. The strain gauge system was calibrated with dial gauge measurements of the bulk expansion of gypsum products. Three composite types (microtilled, hybrid and posterior) were evaluated for polymerization exotherm, contraction during curing, and contraction for various shades. A 60-s curing time was used. The posterior composite (P-50) demonstrated the lowest exotherm and polymerization contraction. The contraction for Silux Plus dark grey was significantly lower than all other shades of all materials. The strain gauge method appears to be well suited for real-time measurement of the curing process and provides a means for studying the kinetics of polymerization. KEY WORDS: Composite restoratives, Polymerization, Technique J. Dent. 1991; 19: 312-316 1991)
(Received 28 August 1990;
reviewed 30 November
accepted 3 April
Correspondenceshouldbeaddressedto: Dr R. L. Sakaguchi, Biomaterials Research Center, University of Minnesota School of Dentistry, 16-212 Moos Tower, 515 Delaware Street, S.E., Minneapolis, MN 55455, USA.
INTRODUCTION Composite restoratives have the inherent property of polymerization shrinkage (Hansen, 1982; Davidson and deGee, 1984; Davidson et al.. 1984; van Noort et al., 1988), which produces a volumetric change and distributes contractile stresses through the resin hard tissue interface into the tooth. The part of the shrinkage which occurs before the polymerization gel point does not induce stress, and the volumetric change can be compensated for by immediate flow of the composite paste. However; following gel formation, the polymerization process is accompanied by a rapid increase in elastic modulus, which means that subsequent shrinkage can induce stress within the polymer and distribute it to the boundary layers. This post-gel shrinkage is a major problem in the clinical use of composites. Gaps occur at the interface between the restoration and the remaining tooth structure when the enamel or dentine bond strength is inadequate to withstand the polymerization contractile stresses. These gaps result in bacterial microleakage, which may be an aetiological factor in pulpal pathology (BrannstrGm, @ 1991 Butterworth-Heinemann 0300-57 12/9 1/OS0312-05
1984). If the bond strength is adequate, the contraction stresses are transmitted to the remaining tooth structure and may result in enamel microcracks at the cervical areas. These cracks may propagate as the tooth is cyclically loaded. Polymerization shrinkage has also been implicated in postoperative sensitivity of composite restoratives (Eick and Welch, 1986). The magnitude of the post-gel shrinkage is therefore an important criterion in the assessment of new composite materials and in the optimization of bond strengths for dental adhesives for use in bonded restorative systems. Strain gauges are extremely sensitive to linear dimensional changes. When the gauge is bonded to a substrate, the linear dimensional changes in the substrate are efficiently transferred to the gauge and readily measured. This linear dimensional change is only transferred when the substrate has a measurable modulus (post-gel) to induce stress on the gauge. It may therefore be applicable to the measurement of post-gel shrinkage. This study explores a new method for monitoring the polymerization contraction of composite resins utilizing
et al.: Strain gauge measurement
Extension arm connected to movable stop which engages dial gauge piston
fig. 1. ‘V’trough by a dial gauge
for the measurement and strain gauge.
electrical resistance strain gauges. The gauges bond readily to the surface of the composite materials sample and may be better suited for measurement of the clinically significant components of polymerization shrinkage than other methods. This paper describes the use of a strain gauge system for test configuration, calibration and preliminary studies on post-gel linear dimensional change in setting dental materials. Strain gauges are symmetrical in that they record expansion as easily as contraction, and in general the performance characteristics of the gauge apply equally well to positive and negative linear change. This symmetry is particularly useful in validation because bulk expansion of materials, such as gypsum products, is much easier to measure than contraction of polymeric materials and is well documented. Further, the measurement of bulk expansion by traditional means can be carried out with simultaneous measurement by a strain gauge on the same specimen. Thus the study design of the present report includes a general validation of strain gauge assessment of setting expansion (i.e. positive dimensional change) of impression plaster before proceeding to the more difficult measurement of composite resin shrinkage on smaller samples.
Comparison of the strain gauge methodology with traditional methods for measuring expansion As already indicated, strain gauges report tension (positive values) as the test sample expands and compression (negative values) as they contract. Further, the expansion behaviour of gypsum materials is well defined (Sweeney and Taylor, 1950; Mahler, 1955; Hollenback, 1963). A traditional ‘V’trough was designed which allows for bulk placement of gypsum materials (Fig. I). One end is fixed and the other end includes a ‘V’plate which is freely
fig. 2. Experimental of composite
movable and attached to a dial gauge. Resistance to expansion is reduced by separating the gypsum material from the surface of the ‘V’ trough through the use of petroleum gel and rubber dam material. As the material expands, the position of the movable plate is measured via the dial gauge. Impression plaster (No. 2 Snow White, Kerr Manufacturing Co., Romulus, MI, USA) was mixed according to the manufacturer’s instructions, but modified by substitution of 8 cm3 of polyvinyl white glue for 8 cm3 of water to increase the adhesion between the aqueous plaster and the hydrophobic strain gauge backing (20 cm3 water + 8 cm3 polyvinyl white glue/50 g powder). This mix was immediately placed in the ‘V’trough so that the mix engaged opposing ends of the trough. The dial gauge was zeroed and the lapse of time from the moment of mixing was recorded by a digital timer. In a second experiment, a strain gauge (CEA-06032UW-120, Measurements Group, Raleigh, NC, USA) was placed on the surface of a 15 mm diameter, 2 mm thick sample of freshly mixed impression plaster and embedded slightly (Fig. 2). The strain gauge output had been previously connected to a conditioner (Measurements Group, 2100 Series, Raleigh, NC, USA), which had been balanced to zero. Deformation of the sample was limited to the plane parallel to the strain gauge (Fig. 2). Flat plates of glass (microscope slides) and separating sheets are used to maintain the sample in one plane, while allowing freedom of movement parallel to the gauge. Independent measurements were made of the expansion of impression plaster via the dial gauge and strain gauge technique. The expansion was plotted as a function of time for each of the samples of impression plaster using the two methods.
19: No. 5
‘, $600 li 400
6 7 Time (min)
Fig. 3. Setting expansion of impression plaster: a, measured by a dial gauge; b, measured by a strain gauge,
Strain gauge shrinkage
In a technique identical to that described above, the polyimide-backed electrical resistance strain gauges were placed on the surface of discs (6 mm diameter, 1.0 mm thickness) of composite restoratives. The strain gauge is sensitive to the deformation only after a significant modulus has been attained, and measures the dimensional change of the freely polymerizing composite sample along a single axis. Surface adhesion between the gauge and composite transfers the post-gel polymerization shrinkage. Multiple runs of each variable (n > 5) were performed to enable statistical evaluations. A posterior composite (P-50, 3M Company, St Paul, MN, USA), a microfilled composite (Silux Plus, 3M, St Paul, MN, USA) and a hybrid composite (Herculite XR, Kerr Manufacturing Co., Romulus, MI, USA) were evaluated for polymerization contraction. Three representative shades plus the universal shade were evaluated for the three composite types.
RESULTS Results for the gypsum expansion are shown in Table I, which reports the final linear dimensional change of impression plaster in microstrain for the two methods. There was no significant difference between expansion values for gypsum as determined by the two methods (P = 0.86). The results for the strain gauge method are shown in Fig. 3b and the dial gauge method in Fig. 3a. The strain gauge output shows measurable change before the dial reading, by about 2 min. This was typical for all specimens. Fig 4a and b show the measurement of the polymerization contraction in phase with the rise in temperature for P-50 (universal shade). The polymerization parameters are shown for the three materials for the study in Tables II-IV. Table ZZshows the relative exotherms, and Tables ZZZand IV show the shrinkage as a function of material type and material shades. All the data in Tables ZZ-ZVwere respectively treated as an ANOVA In each case significant differences were found, and the exact source of the
difference was identified by the Student-Newman-Keuls multirange test.
DISCUSSION The strain gauge measurements of gypsum expansion shown in Table Z are not significantly different from those of the dial gauge method. It should be noted that the sample geometry for the strain gauge measurement of plaster expansion was similar to that used for measurement of composite shrinkage. This was done to validate the use of sample geometry to which strain gauges were applied. The expansion values are consistent with those attained by a dilatometer method (0.07-0.10 per cent linear expansion, deGee et al., 1981). The strain gauge measures expansion immediately (real-time) after placing the gauge on the surface of the plaster. These subtle changes are sufficient to deform the strain gauge but are not detected by the mechanical system until 2-5 min after the measurements commenced. This indicates the sensitivity of the strain gauge instrumentation. Adhesion between the aqueous impression plaster mix and the hydrophobic polyimide backing of the strain gauge had originally presented a problem in making reliable measurements of expansion throughout the entire Table 1. Measurements of setting expansion of impression plaster using a dial gauge and strain gauge
‘V’ trough dial gauge (microstrain)
2 3 4 5 6 7 8 9 Mean s.d. Probability(two-tailed) P = 0.86.
1040 952 709 1008 617
Strain gauge (microsrrain) 1000
1000 850 733 843 745 925 860 950 871 111
et al.: Strain gauge
I%. 4. Exotherm and polvmerization contraction as a function of time for a tvnical run. a. Typical exotherm of P-50 (universal shade) as a function of curing time. b, Typical polymerization contraction of P-50 (universal shade) as a function of curing time. I
setting reaction. The substitution of 8 ml aqueous solution of polyvinyl white glue for 8 ml of water in the plaster mix provided adequate adhesion between the aqueous plaster and the hydrophobic strain gauge backing for stress transfer while not interfering with the setting time or expansion of the plaster (P = 0.12). Adhesion is more than adequate between the composite sample and the polyimide backing of the strain gauge because of the inherent surface tack of the uncured composite. A polymerization contraction curve for P-50 is compared with the corresponding exotherms in Fig. 4. The heat contribution to the composite comes from the chemistry of the polymerization with a contribution from the radiant heat of the curing light source. Heat resulting from the polymerization exotherm causes a linear expansion of the composite sample, which is not an artifact but a real change that should be measured. Application of the curing light creates expansion of the composite due to the direct radiant heat. This is evident in the rapid contraction that takes place as soon as the light source is turned off. This effect is not recorded in the final net shrinkage value. Heat also directly affects the strain gauge by affecting its sensitivity. Fortunately the strain gauges used in this study (self-temperature compensating) demonstrate a very flat response in the temperature range of interest (15-40°C). The results show that typically the temperature rises sharply within a 10-s time period and is maintained at about 35 “C for the next 2 min. Fig. 4b also shows that 83 per cent of the linear post-gel contraction is achieved within a 60-s time period. The differential exotherms for three materials are shown in Table ZZ.This refers to temperature rise and not Tab/e II. Exotherm (“C) of Herculite XR, P-50 and Silux PIUS* n
P-50 Silux Plus
1 .o 3.1
Student-Newman-Keuls test at P < 0.05: P-50 was significantly different from Herculite XR and Silux Plus. *1 mm thick samples cured for 60 s.
instantaneous temperature. P-50 is shown to be significantly lower than the other two compounds, a microfil and a hybrid. This is expected since the two latter materials exhibit a higher filler loading and contain more uncured resin, the curing of which is responsible for the temperature rise. Although the results in Table ZZZdo not demonstrate a significant difference in the shrinkage of the three materials, P-50 was expected to show less shrinkage since the resin is responsible for the shrinkage and P-50 contains relatively less resin than the other two materials. However, the shrinkage differences between Silux Plus and Herculite XR should be studied with the awareness that Silux Plus demonstrates a relatively lower degree of cure. While the relatively lower degree of polymerization can explain the gross differences, a more sophisticated approach may be necessary to explain the other differences. It should be remembered that shrinkage accompanies the polymerization process and one simple, but trivial, way to reduce shrinkage is simply to reduce the degree of polymerization. This would have pronounced effects on the reduction of other physical properties. However, it is probably true that some smaller differences in shrinkage values can be explained by a reduction in the degree of polymerization, other things being equal. The effects of composite shades on post-gel shrinkage shown in Table IV are probably best rationalized by the ability of composite shading to attenuate the 470 nm light necessary for the polymerization process. The attenuation is achieved in three ways: (1) yellow colouring will spectrophotometrically remove the 470 nm band; (2) Table 111.Strain gauge measurement of the percentage linear polymerization contraction for the universal shade of each composite* Mean sample thickness
Herculite XR P-50 Silux Plus
0.224 0.208 0.218
0.025 0.013 0.027
0.9 0.9 1
Student-Newman-Keuls test at P < 0.05; no significantdifference. *60 s curing time.
J. Dent. 1991; 19: No. 5
Table IV. Strain gauge measurement of the percentage for three shades of each composite*
n Herculite XR Dark grey Dark yellow Light P-50 Grey Extra light Yellow Silux Plus Dark grey Grey Light
8 8 10
16 7 16
0.215 0.209 0.212
0.019 0.016 0.02 1
0.149 0.198 0.175
0.015 0.018 0.032
Student-Newman-Keuls test at P < 0.05: Silux Plus dark grey significantly different from all other composite types and shades. Silux Plus light significantly different from P-50 yellow and grey. *60 s curing time.
white additives, which increase the brightness (or value), will splinter, scatter and reflect the light, so that it does not reach the resin; and finally, (3) dark additives, which decrease the brightness (or value), absorb the light so that it does not reach the resin. The effects of all these shading parameters would be expected to reduce the shrinkage due to a reduced degree of polymerization under standard conditions. These considerations are most marked in the Silux Plus light and dark gray shades. However, the effects are much less marked in the P-50 and Herculite XR ranges. These findings follow those reported by Eick and Welch (1986) Baharav et al. (1988) and Walls et al. (1988). The strain gauge method of measuring composite shrinkage is complementary to dilatometry methods, which offer a total volumetric measurement of shrinkage (Hawken and Stevens, 1979; deGee et al., 1981; Bausch et al., 1982; Goldman, 1983; Davidson and deGee, 1984; Penn, 1986). The linear dimensional changes measured here are substantially lower than those reported by others (3-4 vol. per cent). It should be emphasized that this method records only post-gel changes in the linear dimension. In the introduction to this report, it was pointed out that before the composite resin reaches the gel state, it is able to compensate for the shrinkage by flow (Davidson and deGee, 1984). Contraction stresses are not placed on adjacent bonded structures until a sufficient modulus has been reached. This is the point when shrinkage becomes clinically significant (Bausch et al., 1982). Strain gauge methodology is able to separate out this clinically significant part of the total volumetric shrinkage (Sakaguchi and Douglas, 1989). The strain gauge methodology also has other advantages. The small size of the gauge allows it to measure localized shrinkages because the gauge can be precisely located. This becomes very useful in the restored tooth where stress transfer to the hard tissue due to the bonded composite can be measured in simulated clinical conditions in the laboratory. Finally, it is possible to measure
shrinkage in real-time throughout the curing process and use the methodology as a means for studying the kinetics of polymerization.
Acknowledgements This study was supported in part by USPHS Research Grant T32 DE07098 from the National Institute of Dental Research, Bethesda, MD 20892. The authors wish to thank Dr E. Brust and S. Nelson for their contributions in the measurements and MS Marilynn Erickson for preparation of the manuscript.
References Baharav H., Abraham D., Cardash H. S. et al. (1988) Effect of exposure time on the depth of polymerization of a visible light cured composite resin. J. Oral Rehabil. 15, 167-172.
Bausch J. R., de Lange K., Davidson C. L. et al. (1982) Clinical significance of polymerization shrinkage of composite resins. _I Prosthet. Dent. 48, 59-67. Brannstrom M. (1984) Communication between the oral cavity and the dental pulp associated with restorative treatment. Oper. Dent. 9, 57-68. Davidson C. L. and deGee A. J. (1984) Relaxation of polymerization contraction stresses by flow in dental composites. J. Dent. Res. 63, 146-184. Davidson C. L., deGee A. J. and Feilzer A (1984) The competition between the composite-dentin bond strength and the polymerization contraction stress. J Dent. Res. 63, 1396-1399. deGee A. J., Davidson C. L. and Smith A (1981) A modified dilatometer for continuous recording of volumetric polymerization shrinkage of composite restorative materials. J. Dent. 9, 36-42. Eick J. D. and Welch F. H. (1986) Polymerization shrinkage of posterior composite resins and its possible influence on postoperative sensitivity. Quintessence Int. 17, 103-111. Goldman M. (1983) Polymerization shrinkage of resin-based restorative materials. Aust. Dent. J. 28, 156-161. Hansen E. K. (1982) Visible light-cured composite resins: polymerization contraction, contraction pattern and hygroscopic expansion. Stand. J. Dent. Res. 90, 329-335. Hawken M. B. and Stevens A. L. (1979) The small volume plethysmograph: an improved instrument for measuring the setting shrinkage of composite dental restorative materials. J. Biomed. Eng. 1, 116-119. Hollenback G. M. (1963) The physical properties of gypsum plasters. J. S. Calif: Dent. Assoc. 31, 47. Mahler D. B. (1955) Plaster of Paris and stone materials. Znf. Dent. J. 5, 241. Penn R. W. (1986) A recording dilatometer for measuring polymerization shrinkage. Dent. Mater. 2, 78-79. Sakaguchi R. L. and Douglas W. H. (1989) Strain gauge measurement of polymerization shrinkage. J. Dent. Res. 68, 977 (Abstr. 885). Sweeney W. T. and Taylor D. F. (1950) Dimensional changes in dental stone and plaster. J. Dent Res. 29, 749. van Noort R., Cardew G. E. and Howard I. C. (1988) A study of the interfacial shear and tensile stresses in a restored molar tooth. J. Dent 16, 286-293. Walls A. W. G.. McCabe J. F. and Murray J. J. (1988) The polymerization contraction of visible-light activated composite resins. J. Dent. 16, 177-181.