Scand. J. Dent. Res. 1975: 83: 245-253 (Key words: denial amalgam)
Creep of dental amalgam S. ESPEVIK AND S. E. SORENSEN NIOM, Scandinavian Institute of Dental Materials, Oslo, Norway ABSTRACT - The Steady-State creep rates of dental amalgams were measured and the creep rates were correlated with the microstructure of the amalgams. The influence of manipulation variables on creep rate and microstructure was evaluated. The samples were either mechanically or hand triturated, hand condensed, and stored at 20°C for 7 d. The specimens were subjected to a constant tensile load for 24 h, and the elongation was measured with a displacement transducer. The smallest creep rate was found on a spherical amalgam, and the highest on an amalgam made from a preamalgamated fine grain alloy. A correlation between microstructure and creep rate could be demonstrated. The amount of 72 phase did not seem to influence the creep rate. Amalgams that had large closely packed y particles surrounded by a small volume fraction of yi phase exhibited a low creep rate. Amalgams that had small broken up y particles surrounded by a large volume fraction of 71 exhibited a high creep rate. Abusive manipulation of lathe-cut amalgam alloys resulted in high creep rates, small broken up 7 particles and a high volume fraction of 71. {Received for publication 25 January, accepted 1 March 1975)
Dental amalgam has been found to exhibit three types of creep phenomena: (1) instantaneous elastic strain, (2) transient creep (retarded elastic strain), and (3) steady-state creep (OGLESBY, DICKSON, RODRIGUEZ,
DAVENPORT
&
SWEENEY
1968). The steady-state creep behavior of dental amalgams subjected to tensile stresses of 500 to 4,000 psi over the temperature range of 23 to 52°G were studied by
DICKSON,
OGLESBY
&
DAVENPORT
(1968). They found that the steady-state creep rate was dependent upon stress, temperature, and constants of the material over the temperature and stress ranges studied. Creep of amalgam depends on alloy brand (JORGENSEN 1973), structure of original alloy powder (VRIJHOEF, SCHARREHBERG & DRIESSENS 1974),
mercury content (MAHLER & EYSDEN 1969), and manipulation variables (OsBORNE, PHILLIPS, SWARTZ & NORMAN 1974). The effect of the manipulation variables on the creep rate varied considerably between alloy brands. Most of these parameters are known to affect the microstructure of dental amalgams, especially the relative amounts of y, yi and y-2 phases (OTANI & JORGENSEN 1967). A correlation between the creep rate of amalgam and the incidence of marginal breakdown of amalgam restorations has been demonstrated (MAHLER, TERKLA, EYSDEN & REISBICK 1970,
viLE &
KASLOFF
1971,
DUPERON,
BINON,
NE-
PHILLIPS,
SWARTZ, NORMAN & MEHRA 1973,
MOFFA
1974). These observations have focussed considerable importance and interest on creep as a parameter for the screening of
246
ESPEVIK AND SORENSEN Table 1 Dental amalgam alloys used in the investigation
Code 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Alloy name Spheraloy Powder, zinc-free Shofu Spherical, zinv-free Shofu Spherical
Alloy-to-mercury ratio
Manufacturer Kerr, Michigan, U.S.A.
Shofu Dental Mfg. Co. Ltd., Kyoto, Japan Shofu Dent. Mfg. Co. Ltd., Kyoto, Japan DAB AB Svenska Dental Instrument, Standard alloy Stockholm, Sweden New True S. S. White Limited, Dentalloy England Revalloy S. S. White Limited, England Dispersalloy Western Metallurgical, Edmonton, Canada *Standalloy F Degussa, Geschaftsbereich, Dental, Frankfurt, Germany *Standalloy Degussa, Geschaftsbereich, Dental, Frankfurt, Germany *Amalcap F. G. IVOCLAR AG, SchaanLiechtenstein *Amalcap IVOCLAR AG, SchaanLiechtenstein *DAB Argos AB Svenska Dental Instrument, Alloy, non-zinc Stockholm, Sweden *Royal Dental Alloy 5 Guld- & Amalgambolaget AB, Extra langsam Stockholm, Sweden *Royal Dental Guld- & Amalgambolaget AB, Alloy 20 Medium Stockholm, Sweden *Royal Dental Alloy Guld- & Amalgambolaget AB, 30 Extra snabbt Stockholm, Sweden *Ardent Powder Ardent, Stockholm, Sweden *A. N. A. 68 AB Nordiska Affineriet, Stockholm, Sweden
Batch no.
5:5
1355
5:4.2
562
5:4.2
30
5:7
730326
5:7
507321
5:5.5 5:6
057144 2093
5:5
70213
5:5
50201
5:6
5:5
240873 308544 100973 274559 730919
5:5
25.73
5:5
21.73
5:5
17.73
5:5 5:5
5:4
18
*S. T. A. 68
Guldsmeds AB, Malmo, Sweden
5:5
19
*DAB Argos Alloy
5:5
20 21
*Svedia 68 *Real Expansion Alloy *Solila F. C.
AB Svenska Dental Instrument, Stockholm, Sweden Svedia, Enkoping, Sweden Vastra Sveriges Dentaldepot, Goteborg, Sweden De Trey Freres S. A., Ziirich, Switzerland
730903 29.8 Nov. 1972 2829 730316
5 :5 5:5
3 1/4 08 841801
5:5
OEIOEL
22
"•Preamalgamated
amalgams. T h e purpose of the present
IMateriai and methods
study was to measure the steady-state creep
SAMPLE PREPARATION
rate of dental amalgams, and to correlate it with the microstructure.
Twenty-two amalgam alloys were employed in the study and the alloys and mercury were pro-
CREEP OF DENTAL AMALGAM portioned according to manufacturer's directions (Table 1). All alloys were purchased in the latter part of 1973. Trituration was done by hand with a mortar and pestle. Eight samples for each alloy were hand condensed immediately after trituration in a stainless steel mold with a condensation pressure of approximately 1.0 kp/mm^ and condensation was completed in 3 min after end of trituration (later referred to as correct manipulation). Alloys were also triturated mechanically 5 seconds in the Silamat®. The condensation of these alloys started 2 min after the start of trituration, and was completed in 5 min after end of trituration with a pressure of approximately 0.5 kp/mm^ (later referred to as abusive manipulation). The mold design employed was essentially that of MAHLER & MITCHEM (1964). The crosssectional area of the specimen was 6 mm^ and the gauge length 10.2 mm.
TEST PROCEDURE
After the specimens had been stored 7 d at 20° C they were subjected to a constant load of 12 kg at 20 ± r C for 24 h. The elongation of the samples was measured with a Hewlett Packard displacement transducer. When amalgam is subjected to a stress below
247
its fracture strength, it creeps. Fig. 1 is a typical creep curve for an amalgam. In the present study the secondary steady-state creep rate, represented by the inclination to the straight portion of the creep curve (Fig. 1), was recorded for eight samples from each amalgam alloy both when correct and when abusive manipulation were used in the sample preparation.
METALLOGRAPHY
The amalgam alloy powders were embedded in an ordinary metallographic resin and polished using standard lnetallographic techniques. The amalgam specimens were mounted in a resin (Epofix®, Struers) with a low temperature increase during setting. The specimens where then ground and polished using standard metallographic techniques (SiC paper and diamond paste with particle size down to 1 [xm). The specimens were examined in the scanning electron microscope (Jeol 50A) u.sing both secondary and back-scattered electron images. The phases were identified and analyzed using an energy dispersive detector (EDAX) attached to the scanning electron microscope, and a microprobe (ARL). Results CREEP RATES
I I M K , HOURS
Fig. 1. Typical creep curve for an amalgam. Initial portion of the curve (I) is primary creep, linear slower part of creep curve (II) is secondary steady-state creep, and increased creep (III) is tertiary creep. Secondary steady-state creep rate represented by the inclination to straight portion of creep curve (II) was recorded for specimens in present study.
The Steady-State creep rates of the amalgams listed in Table 1 using correct manipulation are given in Fig. 2. It should be noted that the highest creep rate observed was about 10 times higher than the lowest rate. A comparison between the steady-state creep rates of amalgam specimens fabricated using correct manipulation and specimens fabricated using abusive manipulation is shown in Fig. 3. It is apparent that the creep rate of lathe-cut alloys depends on manipulation variables. Alloy No. 10, which had the hardest alloy particles and highest 1-h compressive strength (ESPEVIK, unpublished data) of the lathe-cut alloys, showed the largest increase in creep rate from correct to abusive manipulation and had the highest creep rate when abu-
248
ESPEVIK AND SORENSEN
16 17 18 19 20 21 22 Fzg. 2. Steady-state creep rates of 22 amalgams in Table 1 are illustrated using correct manipulation. Mean and standard deviation for eight samples for each alloy are shown. A manual sample preparation was employed and the large standard deviation was expected because standardization was difficult.
sive manipulation was employed. Alloy No. 2 is a spherical alloy and only minor effects of manipulation were obsers^ed.
METALLOGRAPHY
Optical photomicrographs of the unreacted amalgam powder are shown in Fig. 4 A-D. Fig. 4 A is from a spherical alloy, and the particle size is approximately 3040 \m\. The surfaces of the spheres are smooth. The particle size in Fig. 4 B is about the same as in Fig. 4 A, and the surfaces of the particles are smooth. The amalgam powder in Fig. 4 C has a rough surface, but the particle size is about the same as in Fig. 4 A and B. The particle size in Fig. 4 D is larger, about 50-70 ptm, but the surfaces are not as rough as the surfaces of the particles in Fig. 4 C though rougher than in Fig. 4 A and B. The microstructure of specimens from, a
spherical amalgam is shown in Fig. 5. The phases were identified by microprobe and an energy dispersive detector, and the compositions thus obtained were similar to those previously reported ( SUTFIN & OGILVIE 1972, MAHLER, ADEY & EYSDEN 1973). The 7, 71 and 72 phases can be distinguished. In a back-scatter image (Fig. 5) the brightness of the different phases is a function of the average atomic number of the phase. The round gray areas are the remaining 7 particles. The lighter gray matrix phase is the 71, and the 72 phase can be seen in the y\ and usually in contact with 7. There appears to be more 71 and 72 in Fig. 5 B than in 5 A, and the 7 spheres seem to be packed closely with little 71 in between, especially in Fig. 5 A. No }'2 phase could be seen in back-scattered electron images from samples made from alloy No. 7 by either correct or abusive manipulation.
CREEP OF DENTAL AMALGAM
249
siderable Yl phase is present along with some Cu-Sn phase (Fig. 7). Substantially more Cu-Sn phase was generally found in the Y and YI, for alloys with higher copper content and somewhat lower creep rate (Fig. 8). SURFACE MORPHOLOGY OF ALLOY POWDER
Alloy powder with a low surface-to-volume ratio (Fig. 4 A, B) showed less tendency to break up during trituration and condensation. Alloy powder with a higher surfaceto-volume ratio, i.e. with cracks and rough surface (Fig. 4 C ) , generally resulted in smaller and broken up 7 particles in the amalgam (Fig. 8), and in high creep rate. Preamalgamated alloy powder usually had a higher surface-to-volume ratio, and a high creep rate. Discussion
2B 5A
5B iOA 10B
Fig. 3. Steady-state creep rates of three of the alloys in Table 1 are illustrated using correct (A) and abusive (B) manipulation. Correct manipulation refers to specimens which were hand triturated and hand condensed immediately after trituration with a pressure of approximately 1.0 kp/mm^. Condensation was completed in 3 min after end of trituration. Abusive manipulation refers to specimens which were triturated 5 s in the Silamat® and hand condensed starting 2 min after initiation of trituration and completed 5 min after end of trituration with a condensation pressure of approximately 0.5 kp/mm^.
The microstructure of Fig. 6 A shows that much of the original y phase is intact, and a relatively small volume fraction of 71 is present in this lathe-cut alloy. When abusive manipulation was employed in the sample preparation, the volume fraction of both Yl and 72 was increased (Fig. 6B). The microstructure of an amalgam with a high creep rate is shown in Fig. 7. The Y phase is small and broken up, and con-
MICROSTRUCTURE AND CREEP RATE
Qualitative microstructural observations of samples from different amalgam alloys indicated that a low creep rate was associated with substantial amounts of y phase, i.e. the Y particles showed little bulk reaction with mercury to produce the YI a-nd y^ phases and had maintained their original size and shape (Fig. 5 A, 6 A). A high creep rate was associated with small broken up y phase, and a larger volume fraction of YI (Figs. 7-8). AsHBY (1970) pointed out that second phase particles can slow down contributions to steady-state creep from: grain boundary sliding, diffusional creep and dislocation creep. He also found theoretically that large particles would be more effective in reducing grain boundary sliding than the same volume of small particles. ANSELL & WEERTMAN (1959) developed a quantitative model for creep of dispersion hardened alloys in which they assumed the rate-controlling process to be asso-
250
ESPEVIK AND SORENSEN
Fig. 4. Optieal micrographs of unreacted alloy powder. A is alloy No. 2, B is alloy No. 6, G is alloy No. 8 and D is alloy No. 10 in Table 1.
CREEP OF DENTAL AMALGAM
20wm
251
20 um
Fig. 5. Back-scattered electron images of amalgam made from alloy No. 2. Correct manipulation was used for the sample in A. Creep rate for this particular sample was 1.8 X 10-^ min-^. Abusive manipulation was used for sample in B which had a creep rate of 2.5 x 10-^ min-^. White matrix area (L) is the Yl phase, round dark phase (M) is 7 phase, and 73 phase (N) is pointed at by an arrow in Fig. 5 B.
20 um.
20 um
Fig. 6. Back-scattered electron images of amalgam samples made from alloy No. 5. Correct manipulation was used when sample A was made, and creep rate for this particular sample was 3.7 X 10-^ min-^. Abusive manipulation was used when -sample in B was made and creep rate for that particular sample was 9.4 x 10-^
252
ESPEVIK AND SORENSEN
pears that the observed slow down of the creep rate by large y particles could be a reduction of grain boundary sliding, or diffusion-controlled climbing of dislocations over y particles. Cu-Sn phases dispersed in the Yi phase may have the same effect on the creep rate as the y particles.
MANIPULATION VARIABLES, MICROSTRUCTURE AND CREEP RATE
Fig. 7. Back-scattered electron image of amalgam sample made from alloy No. 8. Correct manipulation was used and creep rate was 2.0 X 10-'' min--*-. The Cu-Sn phases are pointed at by arrow.
dated with climb of dislocations over hard particles. The model predicted the creep rate to decrease as the particle size was increased. DicKSON et al. (1968) concluded that creep of dental amalgam in the temperature range studied was probably diffusioncontrolled climbing of dislocations. It ap-
Fig. 8. Back-scattered electron image of amalgam made from alloy No. 12. Correct manipulation was used and creep rate for this particular sample was 15.5 x lO-*^ min-^. The Cu-Sn phases are pointed at by arrow.
Manipulation affected the microstructure and a correlation between the creep rate and the microstructure could be demonstrated. The delayed condensation with a low condensation pressure of the abusive manipulation resulted in an increased creep rate, which was associated with an increase in the amount of yi and a decrease in the y particle size and volume fraction. The abusive manipulation gave more mercury for the reaction with y, and more yi and yy resulted. The spherical alloy showed a slight increase in the volume fraction of y and the lowest increase in the creep rate between correct and abusive raanipulation. The y particles of the spherical alloy packed well even with a low condensation pressure. The abusive mianipulation increased the creep rate substantially for the lathe-cut alloys, and the amount of y phase was decreased and the volume fraction of yi was increased. The filings had more surface area for reaction with mercury, and a higher condensation pressure was necessary for the y particles to be packed closely together. A fast setting lathe-cut, preamalgamated alloy showed the largest increase in the creep rate when the abusive manipulation was used. Fast setting amalgams have previously been found to incorporate more mercury during setting (JORGENSEN 1974). These observations suggest that manipulation of fast setting amalgams is critical.
CREEP OF DENTAL AMALGAM References ANSELL, G. S. & WEERTMAN, J.: Creep of a dispersion-hardened aluminium alloy. Trans. Am. Inst. Mech. Eng. 1959: 215: 838-843. AsHBY, M. F.: The mechanical effects of a dispersion of a second phase. In: Second International Conference on the Strength of Metals and Alloys, Vol. 2. Ameriean Society for Metals 1970, p. 507-542. BiNON, P., PHILLIPS, R . W . , SWARTZ, M . L., NORMAN, R . D . & MEHRA, R . : Clinical be-
havior of amalgam as related to certain me, chanical properties. Annual Meet. Int. Assoc. Dent. Res. 1973, Abstr. No. 509, p. 186. DicKsoN, G., OGLESBY, P . L . & DAVENPORT, R .
M.: The steady-state creep behavior of dental amalgam. / . Res. Natl. Bur. Stand. C. Eng. lustrum. 1968: 72C: 215-218. DuPERON, D. F., NEVILE, M . D . & KASLOFF,
Z.: Clinical evaluation of corrosion resistance of conventional alloy, spherical-particle alloy, and dispersion-phase alloy. /• Prosthet. Dent. 1971: 25: 650-656. JORGENSEN, K . D . : Plastiske fyldningsmaterialer: Klinisk-teknologisk (2). Kursusnaevnet, Dansk Tandlaegeforening 1973, p. 32. JORGENSEN, K . D . : Den tale amalgamer. In: Nor disk Klinisk Odontologi. A/S Forlaget for Faglitteratur, Copenhagen 1974: 1, p. 7A-I-4. MAHLER, D . B. & EYSDEN, J. v.: Dynamic creep of dental amalgam. / . Dent. Res. 1969: 48: 501-508. MAHLER, D . B . & MITCHEM, J. C : Transverse strength of amalgam. /. Dent. Res. 1964: 43: 121-130. Address: NIOM Scandinavian Institute of Dental Materials Forskningsveien 1 Oslo 3 Norway
253
MAHLER, D . B., ADEY, J. D. & EYSDEN, J. v.:
Microprobe analysis of amalgam. I. Effect of surface preparation. / . Dent. Res. 1973: 52: • 74-78. MAHLER, D . B., TERKLA, L . G., EYSDEN, J. v. & REISBICK, H . M . : Marginal fracture vs.
mechanical properties of amalgam. / . Dent. Res. 1970: 49: 1452-1457. MoFFA, J.: Personal communication 1974. OGLESBY, P . L., DIGKSON, G., RODRIGUES, M . L., DAVENPORT, R . M . & SWEENEY, W . T . :
Viseoelastic behavior of dental amalgam. / . Res. Natl. Bur. Stand. C. Eng. Instrum. 1968: 72C: 203-213. OsBORNE, J. W., PHILLIPS, R . W . , SWARTZ, M . L. & NORMAN, R . D . : Static creep as affected by trituration time and condensation pressure. Annual Meet. Int. Assoc. Dent. Res. 1974, Abstr. No. 30, p. 62. OTANI, H . & JORGENSEN, K . D . : Structure study of amalgam. IV. Quantitative determination of the phases in silver amalgam. Acta Odontol. Scand. 1967: 25: 105-109. SuTFiN, L. V. & OGILIVIE, R . E . : Scanning electron microscopy and energy dispersion microanalysis of metallographically polished dental amalgam surfaces. / . Dent. Res. 1972: 51: 1048-10.54. VRIJHOEF, M . M . , SCH.^RSCHMIDT, J., REHBERG,
H. H. & DRIESSENS, F . C . M . : Microstruc-
tural characteristics of amalgam and their relation with creep properties. Annual Meet. • Int. Ass. Dent. Res. 1974, Abstr. No. 226, p. 111. WING, G. & RYGE, G.: Setting reactions of spher-
ical-particle amalgams. / . Dent. Res. 1965: 44: 1325-1333.
Creep of dental amalgam S. ESPEVIK AND S. E. SORENSEN NIOM, Scandinavian Institute of Dental Materials, Oslo, Norway ABSTRACT - The Steady-State creep rates of dental amalgams were measured and the creep rates were correlated with the microstructure of the amalgams. The influence of manipulation variables on creep rate and microstructure was evaluated. The samples were either mechanically or hand triturated, hand condensed, and stored at 20°C for 7 d. The specimens were subjected to a constant tensile load for 24 h, and the elongation was measured with a displacement transducer. The smallest creep rate was found on a spherical amalgam, and the highest on an amalgam made from a preamalgamated fine grain alloy. A correlation between microstructure and creep rate could be demonstrated. The amount of 72 phase did not seem to influence the creep rate. Amalgams that had large closely packed y particles surrounded by a small volume fraction of yi phase exhibited a low creep rate. Amalgams that had small broken up y particles surrounded by a large volume fraction of 71 exhibited a high creep rate. Abusive manipulation of lathe-cut amalgam alloys resulted in high creep rates, small broken up 7 particles and a high volume fraction of 71. {Received for publication 25 January, accepted 1 March 1975)
Dental amalgam has been found to exhibit three types of creep phenomena: (1) instantaneous elastic strain, (2) transient creep (retarded elastic strain), and (3) steady-state creep (OGLESBY, DICKSON, RODRIGUEZ,
DAVENPORT
&
SWEENEY
1968). The steady-state creep behavior of dental amalgams subjected to tensile stresses of 500 to 4,000 psi over the temperature range of 23 to 52°G were studied by
DICKSON,
OGLESBY
&
DAVENPORT
(1968). They found that the steady-state creep rate was dependent upon stress, temperature, and constants of the material over the temperature and stress ranges studied. Creep of amalgam depends on alloy brand (JORGENSEN 1973), structure of original alloy powder (VRIJHOEF, SCHARREHBERG & DRIESSENS 1974),
mercury content (MAHLER & EYSDEN 1969), and manipulation variables (OsBORNE, PHILLIPS, SWARTZ & NORMAN 1974). The effect of the manipulation variables on the creep rate varied considerably between alloy brands. Most of these parameters are known to affect the microstructure of dental amalgams, especially the relative amounts of y, yi and y-2 phases (OTANI & JORGENSEN 1967). A correlation between the creep rate of amalgam and the incidence of marginal breakdown of amalgam restorations has been demonstrated (MAHLER, TERKLA, EYSDEN & REISBICK 1970,
viLE &
KASLOFF
1971,
DUPERON,
BINON,
NE-
PHILLIPS,
SWARTZ, NORMAN & MEHRA 1973,
MOFFA
1974). These observations have focussed considerable importance and interest on creep as a parameter for the screening of
246
ESPEVIK AND SORENSEN Table 1 Dental amalgam alloys used in the investigation
Code 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Alloy name Spheraloy Powder, zinc-free Shofu Spherical, zinv-free Shofu Spherical
Alloy-to-mercury ratio
Manufacturer Kerr, Michigan, U.S.A.
Shofu Dental Mfg. Co. Ltd., Kyoto, Japan Shofu Dent. Mfg. Co. Ltd., Kyoto, Japan DAB AB Svenska Dental Instrument, Standard alloy Stockholm, Sweden New True S. S. White Limited, Dentalloy England Revalloy S. S. White Limited, England Dispersalloy Western Metallurgical, Edmonton, Canada *Standalloy F Degussa, Geschaftsbereich, Dental, Frankfurt, Germany *Standalloy Degussa, Geschaftsbereich, Dental, Frankfurt, Germany *Amalcap F. G. IVOCLAR AG, SchaanLiechtenstein *Amalcap IVOCLAR AG, SchaanLiechtenstein *DAB Argos AB Svenska Dental Instrument, Alloy, non-zinc Stockholm, Sweden *Royal Dental Alloy 5 Guld- & Amalgambolaget AB, Extra langsam Stockholm, Sweden *Royal Dental Guld- & Amalgambolaget AB, Alloy 20 Medium Stockholm, Sweden *Royal Dental Alloy Guld- & Amalgambolaget AB, 30 Extra snabbt Stockholm, Sweden *Ardent Powder Ardent, Stockholm, Sweden *A. N. A. 68 AB Nordiska Affineriet, Stockholm, Sweden
Batch no.
5:5
1355
5:4.2
562
5:4.2
30
5:7
730326
5:7
507321
5:5.5 5:6
057144 2093
5:5
70213
5:5
50201
5:6
5:5
240873 308544 100973 274559 730919
5:5
25.73
5:5
21.73
5:5
17.73
5:5 5:5
5:4
18
*S. T. A. 68
Guldsmeds AB, Malmo, Sweden
5:5
19
*DAB Argos Alloy
5:5
20 21
*Svedia 68 *Real Expansion Alloy *Solila F. C.
AB Svenska Dental Instrument, Stockholm, Sweden Svedia, Enkoping, Sweden Vastra Sveriges Dentaldepot, Goteborg, Sweden De Trey Freres S. A., Ziirich, Switzerland
730903 29.8 Nov. 1972 2829 730316
5 :5 5:5
3 1/4 08 841801
5:5
OEIOEL
22
"•Preamalgamated
amalgams. T h e purpose of the present
IMateriai and methods
study was to measure the steady-state creep
SAMPLE PREPARATION
rate of dental amalgams, and to correlate it with the microstructure.
Twenty-two amalgam alloys were employed in the study and the alloys and mercury were pro-
CREEP OF DENTAL AMALGAM portioned according to manufacturer's directions (Table 1). All alloys were purchased in the latter part of 1973. Trituration was done by hand with a mortar and pestle. Eight samples for each alloy were hand condensed immediately after trituration in a stainless steel mold with a condensation pressure of approximately 1.0 kp/mm^ and condensation was completed in 3 min after end of trituration (later referred to as correct manipulation). Alloys were also triturated mechanically 5 seconds in the Silamat®. The condensation of these alloys started 2 min after the start of trituration, and was completed in 5 min after end of trituration with a pressure of approximately 0.5 kp/mm^ (later referred to as abusive manipulation). The mold design employed was essentially that of MAHLER & MITCHEM (1964). The crosssectional area of the specimen was 6 mm^ and the gauge length 10.2 mm.
TEST PROCEDURE
After the specimens had been stored 7 d at 20° C they were subjected to a constant load of 12 kg at 20 ± r C for 24 h. The elongation of the samples was measured with a Hewlett Packard displacement transducer. When amalgam is subjected to a stress below
247
its fracture strength, it creeps. Fig. 1 is a typical creep curve for an amalgam. In the present study the secondary steady-state creep rate, represented by the inclination to the straight portion of the creep curve (Fig. 1), was recorded for eight samples from each amalgam alloy both when correct and when abusive manipulation were used in the sample preparation.
METALLOGRAPHY
The amalgam alloy powders were embedded in an ordinary metallographic resin and polished using standard lnetallographic techniques. The amalgam specimens were mounted in a resin (Epofix®, Struers) with a low temperature increase during setting. The specimens where then ground and polished using standard metallographic techniques (SiC paper and diamond paste with particle size down to 1 [xm). The specimens were examined in the scanning electron microscope (Jeol 50A) u.sing both secondary and back-scattered electron images. The phases were identified and analyzed using an energy dispersive detector (EDAX) attached to the scanning electron microscope, and a microprobe (ARL). Results CREEP RATES
I I M K , HOURS
Fig. 1. Typical creep curve for an amalgam. Initial portion of the curve (I) is primary creep, linear slower part of creep curve (II) is secondary steady-state creep, and increased creep (III) is tertiary creep. Secondary steady-state creep rate represented by the inclination to straight portion of creep curve (II) was recorded for specimens in present study.
The Steady-State creep rates of the amalgams listed in Table 1 using correct manipulation are given in Fig. 2. It should be noted that the highest creep rate observed was about 10 times higher than the lowest rate. A comparison between the steady-state creep rates of amalgam specimens fabricated using correct manipulation and specimens fabricated using abusive manipulation is shown in Fig. 3. It is apparent that the creep rate of lathe-cut alloys depends on manipulation variables. Alloy No. 10, which had the hardest alloy particles and highest 1-h compressive strength (ESPEVIK, unpublished data) of the lathe-cut alloys, showed the largest increase in creep rate from correct to abusive manipulation and had the highest creep rate when abu-
248
ESPEVIK AND SORENSEN
16 17 18 19 20 21 22 Fzg. 2. Steady-state creep rates of 22 amalgams in Table 1 are illustrated using correct manipulation. Mean and standard deviation for eight samples for each alloy are shown. A manual sample preparation was employed and the large standard deviation was expected because standardization was difficult.
sive manipulation was employed. Alloy No. 2 is a spherical alloy and only minor effects of manipulation were obsers^ed.
METALLOGRAPHY
Optical photomicrographs of the unreacted amalgam powder are shown in Fig. 4 A-D. Fig. 4 A is from a spherical alloy, and the particle size is approximately 3040 \m\. The surfaces of the spheres are smooth. The particle size in Fig. 4 B is about the same as in Fig. 4 A, and the surfaces of the particles are smooth. The amalgam powder in Fig. 4 C has a rough surface, but the particle size is about the same as in Fig. 4 A and B. The particle size in Fig. 4 D is larger, about 50-70 ptm, but the surfaces are not as rough as the surfaces of the particles in Fig. 4 C though rougher than in Fig. 4 A and B. The microstructure of specimens from, a
spherical amalgam is shown in Fig. 5. The phases were identified by microprobe and an energy dispersive detector, and the compositions thus obtained were similar to those previously reported ( SUTFIN & OGILVIE 1972, MAHLER, ADEY & EYSDEN 1973). The 7, 71 and 72 phases can be distinguished. In a back-scatter image (Fig. 5) the brightness of the different phases is a function of the average atomic number of the phase. The round gray areas are the remaining 7 particles. The lighter gray matrix phase is the 71, and the 72 phase can be seen in the y\ and usually in contact with 7. There appears to be more 71 and 72 in Fig. 5 B than in 5 A, and the 7 spheres seem to be packed closely with little 71 in between, especially in Fig. 5 A. No }'2 phase could be seen in back-scattered electron images from samples made from alloy No. 7 by either correct or abusive manipulation.
CREEP OF DENTAL AMALGAM
249
siderable Yl phase is present along with some Cu-Sn phase (Fig. 7). Substantially more Cu-Sn phase was generally found in the Y and YI, for alloys with higher copper content and somewhat lower creep rate (Fig. 8). SURFACE MORPHOLOGY OF ALLOY POWDER
Alloy powder with a low surface-to-volume ratio (Fig. 4 A, B) showed less tendency to break up during trituration and condensation. Alloy powder with a higher surfaceto-volume ratio, i.e. with cracks and rough surface (Fig. 4 C ) , generally resulted in smaller and broken up 7 particles in the amalgam (Fig. 8), and in high creep rate. Preamalgamated alloy powder usually had a higher surface-to-volume ratio, and a high creep rate. Discussion
2B 5A
5B iOA 10B
Fig. 3. Steady-state creep rates of three of the alloys in Table 1 are illustrated using correct (A) and abusive (B) manipulation. Correct manipulation refers to specimens which were hand triturated and hand condensed immediately after trituration with a pressure of approximately 1.0 kp/mm^. Condensation was completed in 3 min after end of trituration. Abusive manipulation refers to specimens which were triturated 5 s in the Silamat® and hand condensed starting 2 min after initiation of trituration and completed 5 min after end of trituration with a condensation pressure of approximately 0.5 kp/mm^.
The microstructure of Fig. 6 A shows that much of the original y phase is intact, and a relatively small volume fraction of 71 is present in this lathe-cut alloy. When abusive manipulation was employed in the sample preparation, the volume fraction of both Yl and 72 was increased (Fig. 6B). The microstructure of an amalgam with a high creep rate is shown in Fig. 7. The Y phase is small and broken up, and con-
MICROSTRUCTURE AND CREEP RATE
Qualitative microstructural observations of samples from different amalgam alloys indicated that a low creep rate was associated with substantial amounts of y phase, i.e. the Y particles showed little bulk reaction with mercury to produce the YI a-nd y^ phases and had maintained their original size and shape (Fig. 5 A, 6 A). A high creep rate was associated with small broken up y phase, and a larger volume fraction of YI (Figs. 7-8). AsHBY (1970) pointed out that second phase particles can slow down contributions to steady-state creep from: grain boundary sliding, diffusional creep and dislocation creep. He also found theoretically that large particles would be more effective in reducing grain boundary sliding than the same volume of small particles. ANSELL & WEERTMAN (1959) developed a quantitative model for creep of dispersion hardened alloys in which they assumed the rate-controlling process to be asso-
250
ESPEVIK AND SORENSEN
Fig. 4. Optieal micrographs of unreacted alloy powder. A is alloy No. 2, B is alloy No. 6, G is alloy No. 8 and D is alloy No. 10 in Table 1.
CREEP OF DENTAL AMALGAM
20wm
251
20 um
Fig. 5. Back-scattered electron images of amalgam made from alloy No. 2. Correct manipulation was used for the sample in A. Creep rate for this particular sample was 1.8 X 10-^ min-^. Abusive manipulation was used for sample in B which had a creep rate of 2.5 x 10-^ min-^. White matrix area (L) is the Yl phase, round dark phase (M) is 7 phase, and 73 phase (N) is pointed at by an arrow in Fig. 5 B.
20 um.
20 um
Fig. 6. Back-scattered electron images of amalgam samples made from alloy No. 5. Correct manipulation was used when sample A was made, and creep rate for this particular sample was 3.7 X 10-^ min-^. Abusive manipulation was used when -sample in B was made and creep rate for that particular sample was 9.4 x 10-^
252
ESPEVIK AND SORENSEN
pears that the observed slow down of the creep rate by large y particles could be a reduction of grain boundary sliding, or diffusion-controlled climbing of dislocations over y particles. Cu-Sn phases dispersed in the Yi phase may have the same effect on the creep rate as the y particles.
MANIPULATION VARIABLES, MICROSTRUCTURE AND CREEP RATE
Fig. 7. Back-scattered electron image of amalgam sample made from alloy No. 8. Correct manipulation was used and creep rate was 2.0 X 10-'' min--*-. The Cu-Sn phases are pointed at by arrow.
dated with climb of dislocations over hard particles. The model predicted the creep rate to decrease as the particle size was increased. DicKSON et al. (1968) concluded that creep of dental amalgam in the temperature range studied was probably diffusioncontrolled climbing of dislocations. It ap-
Fig. 8. Back-scattered electron image of amalgam made from alloy No. 12. Correct manipulation was used and creep rate for this particular sample was 15.5 x lO-*^ min-^. The Cu-Sn phases are pointed at by arrow.
Manipulation affected the microstructure and a correlation between the creep rate and the microstructure could be demonstrated. The delayed condensation with a low condensation pressure of the abusive manipulation resulted in an increased creep rate, which was associated with an increase in the amount of yi and a decrease in the y particle size and volume fraction. The abusive manipulation gave more mercury for the reaction with y, and more yi and yy resulted. The spherical alloy showed a slight increase in the volume fraction of y and the lowest increase in the creep rate between correct and abusive raanipulation. The y particles of the spherical alloy packed well even with a low condensation pressure. The abusive mianipulation increased the creep rate substantially for the lathe-cut alloys, and the amount of y phase was decreased and the volume fraction of yi was increased. The filings had more surface area for reaction with mercury, and a higher condensation pressure was necessary for the y particles to be packed closely together. A fast setting lathe-cut, preamalgamated alloy showed the largest increase in the creep rate when the abusive manipulation was used. Fast setting amalgams have previously been found to incorporate more mercury during setting (JORGENSEN 1974). These observations suggest that manipulation of fast setting amalgams is critical.
CREEP OF DENTAL AMALGAM References ANSELL, G. S. & WEERTMAN, J.: Creep of a dispersion-hardened aluminium alloy. Trans. Am. Inst. Mech. Eng. 1959: 215: 838-843. AsHBY, M. F.: The mechanical effects of a dispersion of a second phase. In: Second International Conference on the Strength of Metals and Alloys, Vol. 2. Ameriean Society for Metals 1970, p. 507-542. BiNON, P., PHILLIPS, R . W . , SWARTZ, M . L., NORMAN, R . D . & MEHRA, R . : Clinical be-
havior of amalgam as related to certain me, chanical properties. Annual Meet. Int. Assoc. Dent. Res. 1973, Abstr. No. 509, p. 186. DicKsoN, G., OGLESBY, P . L . & DAVENPORT, R .
M.: The steady-state creep behavior of dental amalgam. / . Res. Natl. Bur. Stand. C. Eng. lustrum. 1968: 72C: 215-218. DuPERON, D. F., NEVILE, M . D . & KASLOFF,
Z.: Clinical evaluation of corrosion resistance of conventional alloy, spherical-particle alloy, and dispersion-phase alloy. /• Prosthet. Dent. 1971: 25: 650-656. JORGENSEN, K . D . : Plastiske fyldningsmaterialer: Klinisk-teknologisk (2). Kursusnaevnet, Dansk Tandlaegeforening 1973, p. 32. JORGENSEN, K . D . : Den tale amalgamer. In: Nor disk Klinisk Odontologi. A/S Forlaget for Faglitteratur, Copenhagen 1974: 1, p. 7A-I-4. MAHLER, D . B. & EYSDEN, J. v.: Dynamic creep of dental amalgam. / . Dent. Res. 1969: 48: 501-508. MAHLER, D . B . & MITCHEM, J. C : Transverse strength of amalgam. /. Dent. Res. 1964: 43: 121-130. Address: NIOM Scandinavian Institute of Dental Materials Forskningsveien 1 Oslo 3 Norway
253
MAHLER, D . B., ADEY, J. D. & EYSDEN, J. v.:
Microprobe analysis of amalgam. I. Effect of surface preparation. / . Dent. Res. 1973: 52: • 74-78. MAHLER, D . B., TERKLA, L . G., EYSDEN, J. v. & REISBICK, H . M . : Marginal fracture vs.
mechanical properties of amalgam. / . Dent. Res. 1970: 49: 1452-1457. MoFFA, J.: Personal communication 1974. OGLESBY, P . L., DIGKSON, G., RODRIGUES, M . L., DAVENPORT, R . M . & SWEENEY, W . T . :
Viseoelastic behavior of dental amalgam. / . Res. Natl. Bur. Stand. C. Eng. Instrum. 1968: 72C: 203-213. OsBORNE, J. W., PHILLIPS, R . W . , SWARTZ, M . L. & NORMAN, R . D . : Static creep as affected by trituration time and condensation pressure. Annual Meet. Int. Assoc. Dent. Res. 1974, Abstr. No. 30, p. 62. OTANI, H . & JORGENSEN, K . D . : Structure study of amalgam. IV. Quantitative determination of the phases in silver amalgam. Acta Odontol. Scand. 1967: 25: 105-109. SuTFiN, L. V. & OGILIVIE, R . E . : Scanning electron microscopy and energy dispersion microanalysis of metallographically polished dental amalgam surfaces. / . Dent. Res. 1972: 51: 1048-10.54. VRIJHOEF, M . M . , SCH.^RSCHMIDT, J., REHBERG,
H. H. & DRIESSENS, F . C . M . : Microstruc-
tural characteristics of amalgam and their relation with creep properties. Annual Meet. • Int. Ass. Dent. Res. 1974, Abstr. No. 226, p. 111. WING, G. & RYGE, G.: Setting reactions of spher-
ical-particle amalgams. / . Dent. Res. 1965: 44: 1325-1333.
