Distortion of ceramometal fixed partial dentures during the firing cycle

Distortion of ceramometal fixed partial dentures during the firing cycle David V. Bridger, D.D.S.,

M.S.D.,

Vancouver, B.C., Canada, and University

and Jack I. Nicholls, of Washington,

F

rameworks for ceramometal fixed partial dentures have a poorer intraoral fit after the application and firing of porcelain compared to the initial fit of the casting.‘.” This alteration in fit may be due to porcelain contamination of the internal aspects of castings or to distortion of the metal framework during the porcelain firing cycle. Distortion in a fixed partial denture framework is represented by an increased space between the restoration and the prepared tooth. This space provides a niche for bacterial plaque which may lead to caries and gingival inflammation.

LITERATURE

REVIEW

Several suggestions have been proposed in the scientific literature to explain the distortion resulting in metal frameworks after the various stages of the porcelain firing schedule;“.” these include (1) contraction of the porcelain with subsequent metal deformation,” (2) contamination of the casting, reducing its melting temperature,” (3) grain growth of the alloy, constricting the diameter of the crown,” (4) plastic flow and creep of the porcelain-gold alloy at high temperatures,” (5) reduction in the resiliency of the metal due to the rigidity of porcelain,” (6) improper support of the framework during firing,” (7) inadequate framework design at the gingival leve1,5. ’ and (8) inadequate design of the framework as a whole.“, “’ When considering metal framework distortion, two separate areas of investigation prevail-warpage of the thin metal margins and deformation of the framework body as a whole. Studies investigating warpage of thin margins have thus far been limited to single crowns.‘, ‘. ’ Shaffner’ noted a general trend toward poorer marHonorable Mention, American College of Prosthodontists, Research Award Competition, San Antonio, Tex. *Professor, Department of Restorative Dentistry.

0022-3913/81/050507

+ 08$00.80/O 0 1981 The C. V. Mosby Co.

Ph.D.*

School of Dentistry,

Seattle, Wash.

ginal adaptation in porcelain-fused-to-metal complete crowns after each successive stage in the firing schedule. His conclusion, however, was that this distortion was not clinically significant. Shillingburg et al.” found that shoulder and shoulder-bevel preparations provided some protection against metal distortion during porcelain firing. Faucher and Nicholls’ supported the findings of Shillingburg et al. stating that all margin preparation designs exhibit some distortion but that the incorporation of a shoulder in the preparation decreased the amount of metal distortion. Few studies have investigated the deformation of the fixed partial denture framework as a complete unit.‘. ’ Tucillo and Nielsen’ demonstrated that property changes in ceramic alloys at firing temperatures can result in a sag of the framework. Bryant and Nicholls* subjected straight long-span frameworks to the degassing temperature and observed insignificant differences between the distortions observed in cast and soldered joints. They found significant differences between end-supported and continuously supported fixed partial dentures and suggested that long-span frameworks be well supported during the firing cycle to prevent this sag. Dental technicians empirically use a variety of techniques to control distortion in their fixed partial denture frameworks. Such techniques include different types of sagger trays, different porcelain-application techniques, presoldering and postsoldering, and the use of one-piece castings. It is not yet known whether the loss of marginal integrity is a product of changes in the thin metal margins or the result of changes in the framework body. Nor is it known whether the apparent distortion occurs within the metal itself or as a result of contractile forces within the fired porcelain. The purpose of this study was to answer the following questions: 1. Does distortion occur in the body of a curved,

THE JOURNAL

OF PROSTHETIC

DENTISTRY

507

BRIDGER

Fig. 1. Master stone model of prepared typodont

ANU

NICHOLLS

teeth.

Fig. 3. Labial view of cast metal framework.

Fig. 2. Silicone model.

mold and wax framework

on master

long-span fixed prosthesis framework upon the firing of porcelain? 2. If distortion does occur, is it the result of changes in the metal only, or does it result from combined changes in the porcelain and metal? 3. At what stage of the firing schedule does the greatest distortion occur? 4. What is the clinical importance of the distortion? 5. What is the degree of reversibility of distortion if the porcelain is chemically removed from the framework?

MATERIAL

AND METHODS

A master model representing a clinical maxillary anterior fixed partial denture was developed from a

508

typodont.* The central incisors on the typodont were removed’to simulate missing teeth. Complete crown preparations were performed on the ivorine lateral incisors and canines. The complete crown preparations had shoulder finish lines with a 0.5 mm bevel. The dimensions of reduction were as follows: labiogingival, 1.3 mm; Iinguogingival, 0.8 mm; midlabial, 1.5 mm; midlingual, 1 mm; and incisal 2 mm. All reductions were within It 0.1 mm of the above values. A poured stone model of the typodont and prepared teeth served as the master model (Fig. 1). A six-unit fixed partial denture was waxed to final contour on the master model and cut back to provide room for the porcelain veneer. The wax framework was cast in one piece, and the resulting metal framework served as a prototype for the construction of a silicone mold. Subsequent frameworks were generated by pouring molten wax into the mald, which fitted over the master model (Fig. 2). Ten frameworks were constructed. The wax frameworks were sprued, invested in magnesium phosphate investment and cast as one*Columbia

Dentoform

Corp., New York, .XY.

MAY

1981

VOLUME

45

NUMBER

5

DISTORTION

IN FIRING

CYCLE

Fig. 4. Incisal view of cast metal framework, piece castings in ceramic alloy. The castings were then allowed to bench cool without quenching. Only visually nonporous castings were accepted. The frameworks had support arms extending from the distal aspect of the canines and measuring struts protruding from the midlingual of each tooth (Figs. 3 and 4). After metal preparation, interproximal joints measured 2 mm ( t 0.1 mm) in width and 3 mm (? 0.1 mm) in height. Metal surfaces to be veneered with porcelain were prepared, reducing the thickness of the metal to no less than 0.4 mm (to.1 mm). Superior surfaces of the lingual struts were machined flat, and triangular-shaped measuring markers were placed on these struts with a steel punch. Inferior surfaces of the supporting arms were machined flat to allow positive seating in the measuring jig (Fig. 5). X, Y, and Zcoordinates of each triangular measuring marker were determined after casting, degassing, opaque firing, first body porcelain bake, second body porcelain bake, glaze firing, and porcelain removal. X and Y coordinates were determined using a precision two-way table with micrometer adjustments. The triangular measuring depressions were viewed through a X40 magnifying tube mounted over this table (Fig. 5). Cross hairs within the magnifying tube were aligned directly over the apex of the triangle formed at the metal surface of the measuring depressions. Each micrometer had a measuring accuracy of 1 p., and X and Y coordinate readings were displayed on a digital display panel. Z coordinates were determined using a vertical height gauge with a measuring capability of 1 p. All readings were displayed on a digital display panel,

THE IOURNAL

OF PROSTHETIC

DENTISTRY

with lingual markers labelled A to F

Fig. 5. Precision X-Y table with display panel, and ~40 magnifying

micrometers, tube.

digital

and measurements were made at each triangular marker (Fig. 6). All X, Y, and Z measurements were read three times, the average of the three readings served as the experimental value. For convenience, raw data were stored in a minicomputer* for later analysis. *PDP

11/40,

Digital

Equipment

Corp.,

Maynard,

MA

509

BRIDCER

Fig. 7. Jig for standardizing tion on frameworks.

Fig. 6. Vertical height gauge and display panel for Z axis measurement.

A feasibility study was performed to determine the measuring error. For this, one point was measured 10 times. Readings were reproducible to within + 3 p in the X and Y directions and to within f 1 p in the Z direction. During all stages of the firing sequence, the frameworks were supported by individualized custom firing trays. These trays were fabricated from a commercially available support material.* After the initial measurements, the metal frameworks were cleaned ultrasonically in hydrofluoric acid for 10 minutes, in a surgical detergent for 10 minutes, and in distilled water for 5 minutes. The castings were degassed in a preheated porcelain oven beginning at 860” C with an elevation in temperature at a rate of 32” C per minute until a final temperature of 960” C was reached. Castings were held at this temperature for 8 minutes. The same *Prop,

510

J. Aderer,

Inc., New York,

NY.

AND

NICHOLLS

thickness of porcelain addi

rate of temperature increase was used for subsequent firing stages. The castings were allowed to air cooi, remounted on the measuring jig, and remeasured. Opaque porcelain was applied and condensed by vibration. The castings were dried on a hot plate at 350” C for 10 minutes and then placed in the oven at 700” C. The temperature was then elevated to 920” C in a partial v;Icuum (730 mm Hg). After cooling, a second opaque layer was applied and fired in a similar manner to a temperature of 910” C. The opaque layer was approximately 0.3 mm thick. The frameworks were then remeasured. A thick layer of body porcelain was applied and partially condensed on the frameworks. The frameworks were mounted one at a time on a jig to standardize the thickness of the added porcelain (Fig. 7). This jig was modified from the one used by Shillingburg et al.” The labioincisal third \+‘a~ replaced with incisal porcelain, and facial and lingual contours were reshaped to conform to the jig. Using a fine instrument, the porcelain was C’UI interproximally to the level of the opaque. Following condensation, the frameworks were transferred to the hot plate to dry for 15 minutes at 350” C. They were then placed in the porcelain oven at 700” C and raised to a firing temperature of 900” C under a partial vacuum of 730 mm Hg. After reaching 900” C, the vacuum was broken; the frameworks were then allowed to air fire for 1 minute at 900” C. After cooling, the castings were remeasured, and a

MAY

1981

VOLUME

45

NUMBER

5

DISTORTION IN FIRING CYCLE

Table I. Mean distortions

for seven experimental

Comparison stages A X distortions

A Y distortions

A Z distortions

frameworks

of

Experimental A

I to 2 (degassing) 1 to 3 (opaque) 1 to 4 (body I) 1 to 5 (body 2) 1 to 6 (glaze) 1 to 7 (porcelain removed) 1 to 2 (degassing) 1 to 3 (opaque) I to 4 (body I) I to 5 (body 2) 1 to 6 (glaze) 1 to 7 (porcelain removed) I to 2 (degassing) 1 to 3 (opaque) 1 to 4 (body I) I to 5 (body 2) I to 6 (glaze) I to 7 (porcelain removed)

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

B

marker

C

-7(8) -2(10) -3(9) -3(15) -8(15) 2(9) 19(21) 16(7) 19(9) 18(12) 34( 14) 20(18) 3(6) -l(5) 3(5) 6(7) 4(4) 3(5)

-6(12) -4(12) -2(14) -5(19) -14(17)

w3) 28(23) 24(17) 28(17) 26(25) 51(22) 22(19) 0 0 0 0 0 0

D

W3) 103) 805)

wo) -8(21) 6(15) 27(22) 26(16) 29(17) 28(20) 50(22) 25(18) 3(2)

WJ) 264 2(2) l(5) --1(l)

E -5(12) -1(13) -1(16) -1(23) -17(20) 4(10) lB(16) 12(11) 20(9) 19(13) 33(14) 17(9) --1(4) -2(7) -l(3) -1(16)

F -10(17) -2(15) -6(16) -4(27) -27(26)

WI

-4(3)

0 0 0 0 0 0 0 0 0 0 0 0

E

F

15(2) 503) ll(13)

804) -11(19) 5(20) 500) -21(18) -29(24) 0 0 0 0 0 0 0 0 0 0 0 0

--WI

Values in microns. Numbers in parentheses = SD.

Table II. Mean distortions

A X distortions

A Y distortions

A Z distortions

for three control

frameworks Experimental

Comprison of stages

A

B

C

1 to 2 1 to 3 1 to 4 1 to 5 1 to 6 1 to 7 1 to 2 1 to 3 1 to 4 1 to 5 1 to 6 1 to 7 1 to 2 1 to 3 1 to 4 1 to 5 1 to 6 1 to 7

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

lO(4)

X0(9) -4(12) 502) 4(15) -3(18) -5(22) 21(11) 24(8) lO(9) 16(9) 23(6) 24(8) 0 0 0 0 0 0

5(b) 9(l) ll(4) 5(4) 4(4)

8(6) 12(4)

w5) 6(3) 18(5) 15(10) -l(5) -3(3) O(5)

-42) -2(3) 2(5)

maiker D 8(7) O(W lO(15) 9(9) -1(16) -2(19) lB(20) 21(7) 13(18) 17(15) 24(14) 22(17) -2(4) -2(5) -5(4) -6(5) 4(6) -2(7)

13(8) -1(14) -2(17) 13(20) 15(16) 7(20) lO(17) 14(19) ~(24) -6(4) 2(3) 4(9) -4(l) -2(3) O(3)

Values in microns. Numbers in parentheses = SD.

bake of body and incisal porcelain was applied using the same technique as for the first porcelain application. The frameworks were remeasured after the second porcelain bake. To obtain a natural glaze, the frameworks were preheated to 350” C on the hot plate and then introduced into the porcelain oven at 860” C. The temperature was raised to a glazing temperature of

corrective

THE JOURNAL OF PROSTHETIC DENTISTRY

920” C and held for 2% minutes. After cooling, the frameworks were remeasured. Porcelain was removed from the frameworks by placing them in hydrofluoric acid in ari ultrasonic cleaner. Final measurements were taken after porcelain removal. Ten metal frameworks were constructed. Seven of the frameworks received a normal application of

511

BRIUGER

AND

NICHOLLS

1Y

X Fig. 8. Diagrammatic glaze stages.

representation

of alteration in shape of frameworks between initial and

porcelain and firing sequence; three served as controls and were subjected to the same firing sequence without the porcelain application.

RESULTS The distortion was determined from the X, U, and 2 values of all lingual markers for all frameworks at the seven measurement stages. Measuring points were labelled A to F for purposes of comparison (Fig. 4). Markers A, C, and F were chosen to lie in selected positions. Point A was chosen to be at the origin (X, Y, and 2 = 0). Point C was arbitrarily selected to lie in the XY plane (2 = 0), and point Flay on the X axis (Y and 2 = 0). The measured coordinates of all points were stored in a computer. Distortion computations were related to an individual framework, and no comparison between frameworks in the same experimental group was made. Distortion values for each fixed partial denture were determined by comparing X, Y and 2 values of the initial measuring stage with values of all subsequent stages. Thus at any stage, a change in shape of that particular fixed partial denture from its original dimensions could be determined. This change was determined as the AX, AY, and A.2 distortions of each of the measuring points in the X, Y, and Z axial directions. Tables I and II show the data for the seven experimental frameworks and the three control frameworks, respectively. The data in these tables show that distortion values were greatest in the Y axial direction and least in the Z axial direction. Virtually all mean hZ distortion values were within the experimental read-

512

ing error, indicating very little distortion in this axial direction. The AX distortion in Table I is relatively small for all stages, except at glazing (stage 6) when an average distortion of 27 ~1 occurred at F in the negative axial direction. The framework rebounded after chemical removal of the porcelain (stage 7). Table II shows that the control frameworks developed very little distortion in the X direction except for an aberrant negative distortion at point F during glazing. In the Y direction, the experimentaf frameworks underwent a positive distortion during degassing, averaging 28 h at point C and 27 p at point D (Table I). These markers did not subsequently distort until the glaze cycle, when mean distortions of 5 1 and 50 p were recorded for points C and D, respectively. After chemical removal of the porcelain, the frameworks tended to rebound to the shape they had attained after the degassing stage. Control frameworks demonstrated the same type of initial distortion after degassing, but they showed no outstanding changes at subsequent stages (Table II). Major distortions occurred between the initial and degassing stages and between the initial and glaze stages (Table I). Pictorial representation of this alteration in shape of the experimental frameworks between their original dimension and after the glazing cycle is shown in Fig. 8. The anterior portion of the fixed partial denture moved anteriorly an average of 50 p and closed an average of 27 p between the canines. Values after porcelain removal were 23 p anteriorly and 1 p between the canines. Glaze distortions in the Y direction were statisti-

MAY

1981

VOLUME

45

NUMBER

5

DISTORTION

IN FIRING

CYCLE

tally compared with the distortions occurring after porcelain removal for the experimental frameworks using a t-test. This comparison showed a statistically significant difference at the 95% confidence level. Similarly, glaze distortions in the Y direction were compared for the experimental and control fixed partial dentures. Again, statistical significance was obtained at the 95% confidence level. DISCUSSION This study has demonstrated that distortion did occur during the firing stages necessary to bake porcelain to a ceramometal fixed partial denture framework. The distortion followed a pattern, occurring in certain directions. All experimental frameworks tested exhibited similar distortion patterns. Not all frameworks distorted to the same degree as evidenced by the relatively large standard deviations. Most distortion was observed after the degassing and glazing stages. These two stages have the highest firing temperatures, and both stages have a rapid descent of the frameworks out of the porcelain oven muffle. The first large distortion observed after degassing may be due to sagging and to the relief of internal stresses incurred during the casting and cooling processes. Very little distortion occurred during the firing of the first and second body bakes. Body porcelain bakes differed from the glaze firing in that they occurred under a partial vacuum (730 mm Hg), the temperature was 20” C lower, and frameworks were slowly raised into and out of the muffle. The gradual descent of porcelain-fired frameworks would allow a slower porcelain cooling and thus an annealing of the porcelain.” Rapidly cooled glass has a greater tendency to contract than slowly cooled glass. Areas with high levels of concentrated residual stresses will result from rapid cooling.” The outer glazed layer of porcelain solidifies first, compressing this region, bringing the inner porcelain into tension in the final state. In addition, stresses resulting from any mismatch in the coefficients of thermal expansion between the metal and the porcelain are more pronounced during rapid cooling. A more gradual cooling might reduce distortion because annealing of the porcelain and creep of the metal may reduce the residual stresses. Distortion occurring in the metal framework at glazing is an elastic deformation which rebounds after the removal of porcelain. The final glazed fixed

THE JOURNAL

OF PROSTHETIC

DENTISTRY

% = D,Tan 04 8 = ANGLE OF CONVERGENCE Fig. 9. Relation between marginal discrepancy marginal distortion (d).

(D) and

partial denture therefore is in a state of equilibrium between stresses in both the porcelain and metal. If creep should occur within the metal or porcelain, there is a potential for a change in shape over time. If teeth are considered to be in a static relationship to one another, then marginal discrepancies resulting from framework distortion can be controlled to a large degree by the convergence angle of the preparations (Fig. 9). For example, if the convergence angle 8 of the canine preparations is 10 degrees and only the Lx distortion of 27 p (6) is considered, the theoretical marginal opening (D) is approximately 150 ~1. When the distortions incurred in all three axial directions are combined, the net marginal discrepancy would be even greater than 150 p. Christensen” states that a marginal opening of 39 p should be the limit of clinical acceptance. Teeth, however, are in a dynamic relationship to each other. The buccolingual physiologic mobility of teeth ranges from 56 to 108 p,“’ and physiologic intrusion of 28 p has been recorded for maxillary incisorsi Physiologic accommodation of the periodontal ligament would certainly aid in reducing marginal discrepancies from framework deformation, and temporary seating of the fixed partial

513

BRIDGER AND NICHOLLS

denture for several days would further nal discrepancies.

reduce margi-

CONCLUSIONS 1. Distortion does occur in the body of curved, long-span fixed partial denture frameworks during the porcelain firing cycle. 2. This distortion is a result of changes in the metal as well as the contraction of fired porcelain. 3. The greatest distortional changes occur during the degassing stage and the final glaze stage of the porcelain firing cycle. 4. Distortion incurred by the application and firing of the porcelain is reversible. When the porcelain is chemically removed from the framework, there is an elastic rebound. 5. Distortion is clinically important in that it may lead to detectable marginal openings. 6. The distortion pattern observed in the curved fixed partial denture is a closing of the posterior or lingual dimensions and labial movement in the anterior dimension. This indicates the effect of the contracting porcelain on the metal framework. REFERENCES 1. Shaffner, V. B.: Porcelain-fused-to-metal restorations. The effect of the firing schedule on the shoulder- and chamfertype restorations. M.S.D. Thesis, Indiana University School of Dentistry, 1972. 2. Howell, Ii. A.: Gold/porcelain bridgework. Br Dent J 116:80, 1964. 3. Mumford, G.: The porcelain-fused-to-metal restoration. Dent Clin North Am March 1965, pp 241-249.

514

4.

5.

6.

7. 8.

9. 10.

11.

12. 13.

14.

Silver, M., Klein, G., and Howard. ,%I.. An rvaluation anil comparison of porcelain fused TO cast. metals. ,J P~nsrrz~: DENT 10:1055, 1960. Faucher, R. R., and Nicholls, J. I : Distortion related !li margin design in porcelain-fused-to-metal restorations. ) PROSTHET DENT 43:149, 1980. Schillingburg, H. T., Hobo, S., and Fisher, D. W’.: I’reparation design and margin distortion in porcelain-fused-tometal restorations. J PROSTHEY DENT 29:276, 1973. Tucillo, J. J., and Nielsen, J. P.: Creep and sag propertie\ of a porcelain-gold alloy. J Dent Res 46:579. 1967. Bryant, R. A.. and Nicholls, J. I.: Measurrment ofdistorrionn in fixed partial dentures resulting from degassing. J PKOSTHET DENT 423515, 1979. Miller, I,. I..: Framework design in ceramometal reritorations. Dent Clin North Am October 1977, pp 699-716. Warpeha, J., and Goodkind, R.: Design and technique variables affecting fracture resistance of metal ceramic restorations. J PROSTHETDE3.r 35:291. 1976. Kingery, W. D., Bowen. H. K., and Uhlmann, D. K.: Introduction to Ceramics, ed 2. New York, 1976, John M’ile\ and Sons. Christensen, G. J.: Marginal fit ot gold inlay castinK>. ,J PROSTHETDENT 16:297, 1966. Rudd, K. D., O’Leary, T. J.. and Stumpf, A. J,: Horizontal tooth mobility in carefully screened subjects. Periodontics 2:65, 1964. Parfitt, G. J,: Measurement of the physiological mobility of individual teeth in an axial direction. .I Dent Res 39:fiOii. 1960.

Reprint requeststo: DR. JACK I. NICHOLLS UNIVERSITVOF WAS~~NCTON SCHOOLOF DENTISTRY SEATTLE, WA 98195

MAY 1981

VOLUME 45

NUMBER 5