Thermal gradients and residual stresses in veneered Y-TZP frameworks

Abstract

Objectives

The occurrence of “chipping” of all-ceramic restorations with Y-TZP frameworks has resulted in various designs and cooling procedures recommended for reducing such behavior. In this paper the temperature gradients during fast and slow cooling for conventional and anatomical designs are compared as well as an optical procedure to directly compare the influence of cooling rate on residual stress.

Methods

This investigation quantifies the temperature gradients between the inner and outer surfaces of crowns measured with thermocouples during two different cooling methods with uniform and anatomical frameworks. In the first method the crown was removed from the furnace after commencement of cooling whereas for the second method the crown was cooled to the glass transition temperature (600 °C) before removal. Direct observation of the residual stresses was made with an optical polarimeter and thin slices of veneered copings.

Results

This study observed that slow cooling decreases the temperature differences but still differences of up to 88 °C were observed. For the fast cooled crown, temperature differences of more than 100 °C for the uniform and 140 °C for the anatomical framework at temperatures above the glass transition temperature were recorded. Optical polarimeter observations indicated much lower stresses within the porcelain layer upon cooling by removing the crown below the glass transition temperature.

Conclusion

Slow cooling during the final veneering of dental restorations with zirconia frameworks reduces the temperature gradients and residual stresses within the porcelain layer, which represent one possible cause for chipping. An anatomical designed framework did not show the same reduction extent.

Introduction

The introduction of metal–ceramic system in 1962 enabled dentists and dental technicians to produce highly esthetic dental restorations. More recently with the development of dental all-ceramic systems especially with yttria partially stabilized zirconia (Y-TZP) or alumina as framework materials, esthetics have been further improved. These enhanced dental restorative materials come ever closer to that of natural teeth. In 2008 in Germany over 40% manufactured dental restorations were all-ceramic, but according to the broad-based consumer surveying company since then the demand has decreased to less than 25%. The reason for the decline is the clinical concern about failures and long-term stability of restorations especially with Y-TZP frameworks.

This apprehension, especially the occurrence of “chipping” of all-ceramic restorations with Y-TZP frameworks, has generated considerable research within the dental materials community . Failures within the porcelain layer or breakage at the interface in metal–ceramic restorations are well known since their introduction. But this special “chipping” failure mode appears to arise predominantly with Y-TZP all-ceramic systems especially in the cusp area. The term “chipping” according to the dictionary definition implies that splinters occur, which realistically describes the characteristic features of this failure mode. The crack propagation occurs exclusively within the porcelain layer and does not appear to run to the interface between the veneering and the framework material. Added to the outer surface of a zirconia framework is generally a thin veneering layer beneath where the chipping event occurs .

With all-ceramic framework materials, especially with Y-TZP, the processing guidelines have been modified by most commercial suppliers compared with the well-known technique of the metal–ceramic systems. The consensus, although not as yet genuinely verified, is that residual stresses develop within the porcelain during rapid cooling and these contribute to chipping induced fracture. Residual stresses can be introduced during the firing process inside the porcelain layer and can be of two major origins; due to thermal expansion mismatch and tempering stresses associated with temperature gradients during cooling . These residual stresses, which remain after the cooling in the porcelain layer, are one possible explanation for the differences of “chipping” failures between metal and Y-TZP based all-ceramic restorations. The materials parameters of all-ceramic restoration such as Young’s modulus or coefficient of thermal expansion of the material partners as well as the glass transition temperature of the porcelain do not differ significantly from those of the well-known metal–ceramic porcelains. It has been argued that the drastically different thermal conductivity of the framework materials (gold has approximately 150-times higher thermal conductivity than zirconia) may be the origin of this special failure mode .

One goal of this investigation was to determine experimentally the temperature differences arising during the firing process of all-ceramic crowns with Y-TZP frameworks by the use of a rotational symmetric crown model. In addition two slightly different models (one framework with constant wall thickness and the other with anatomical supported structure) were cooled after firing in two different ways (fast and slow cooling). Another objective was the determination and visual presentation of the stress distribution with an optical imaging polarising system that can measure the optical polarization state of light transmitted through the porcelain. It is mainly based on the ability to build effective polarization state analysers that acquire the Stokes vectors corresponding to each pixel in the image . This is done traditionally by rotating birefringent optical elements such as quarter-wave plates in front of a fixed polarizer .

Reductions of residual stresses within the veneering porcelain will possible result in more secure zirconia framework based dental restorations. The hypothesis investigated in this paper is that temperature gradients produced at temperatures above the glass transition temperature of the porcelain during cooling generate higher residual stresses. Optical polarimeter images are used to verify this hypothesis.

Materials and methods

Crown/sample model

A rotationally symmetric crown form similar to that previously invested by Lenz et al. was selected to ensure the production of identical samples and to simplify the experimental determination of the surface temperature corresponding with the geometry of a premolar tooth. Anatomically precise crowns, while possible to evaluate in a similar manner would have drastically increased the difficulty of production, especially regarding thermocouple placement and subsequent cross-sectioning of similar veneered copings.

The copings (four of each group) were designed by the inLab software (inLab 3.80, Sirona, Germany) and milled using an inLab milling machine (inLab MC XL, Sirona, Germany). The Y-TZP partially sintered frameworks (VITA In-Ceram YZ) were sintered to full density for 2 h at 1530 °C in a furnace (VITA ZYrcomat, VITA, Germany). After sintering the wall thickness was on average 0.7 mm with constant thickness frameworks. The anatomically supported frameworks had a wall thickness of 0.7 mm at the cervical region, which increased within the middle and cusp range from 1.3 mm and within the occlusal region to 1 mm. The overall height of all copings, constant wall or supported framework was 6.95 mm (see Fig. 1 , crown with constant wall thickness).

Fig. 1
A section of the axially rotational crown showing the thermocouple measurement points.

The first veneering step with the dentin porcelain (VM9; VITA, Germany) was the application of a very thin so called “wash-dentin layer” and fired at 950 °C. The first build up dentin firing was at 910 °C and the second was fused at 900 °C. Finally a glaze firing of the porcelain was carried out in each case at 900 °C in a Vacumat 4000 Premium T (VITA, Germany).

The veneer thickness of all samples was identical. At the cusp, complete thickness (framework and porcelain) of 1.4 mm was developed. Thus the porcelain thickness resulted in constant wall thickness framework copings of 0.7 mm on average. The porcelain of the anatomically supported crown had a maximum layer thickness of 0.4 mm.

Thermal measurement configuration

As a basis for supporting the firing of the samples a refractory honeycomb tray was used (Renfert, Germany). The honeycomb tray was drilled centrically to which a thermocouple protective glass with a length of approx. 23 mm was fixed. This length was selected, as the cervical edge of the samples was positioned a height above the tray of approx. 6 mm, which corresponds to the reality in dental labs.

To add retention to the crown a bent wire with protective glass sheath to support the thermocouples was included. The thermocouples were then placed on the internal and external selected measuring points (see Fig. 1 ) prior to being connected to the appropriate channels of data logging system (SCXI 1303, National Instrument, USA). For data recording and evaluation of the results the software LabView 8.5 was used

To enable comparative measurements of the cooling profiles with the attached thermocouples, a different furnace was used, namely a Programat P95 (Ivoclar Vivadent, Liechtenstein). The sample was located in this furnace more securely than in a furnace with mobile base. The more rapid mouth opening like mechanism of this furnace enabled the sample to be more quickly exposed to air movements of the environment than for the plinth lowering furnaces.

The thermocouples wires were led out of the firing chamber of the Programat P95, sealed with an insulating firing pillow, so that no air movement would falsify the measurement and vacuum could be used during the firing process. During all measurements the temperature at positions 1 and 10 were used as references to ensure comparability. Due to the measurement setup and the possible errors of temperature measurement in the furnace by using too many thermocouples, simultaneous measurement at all 10 points was not accomplished. Instead there were only 3 thermocouples in addition to the reference ones at positions 1 and 10 at the same time in the furnace. The temperature was recorded in total at five different measuring points during a firing process. Each sample was heated up and cooled down twice for the thermocouples placed at every measurement point. Altogether 48 measurements were accomplished per group.

Every sample was placed in the middle of the furnace chamber and the furnace was heated to the final temperature (840 °C) at a heating rate of 60 °C/min and held there for 10 min to achieve a homogeneous initial test temperature. The selected final temperature was 70 °C below the recommended dentin porcelain firing temperature to enable sintering and retention of the thermocouples into the soft porcelain on the surface as without this step the results would be inaccurate. After 10 min at the maximum temperature the chamber was opened directly and the furnace was switched off. The only difference between the slow and fast cooling was that for the former the furnace remained completely closed, until it reached 600 °C (the glass transition temperature of VITA VM9 is 603 °C), and thereafter was opened. The samples were not touched until they reached room temperature.

Polarimeter method

The uniform thickness framework samples for the polarimeter (StrainMatic ® M3/250, ilis GmbH, Germany) were produced following exactly the same firing steps and in the same sizes as for the thermo-couple measurements. After the firing, a thin slice 0.5 mm thick was cut using a water-cooled diamond saw from the mid-point of the coping and mounted in the polarimeter. With the polarimeter the turn value of the light through the crown was determined. The change of the angle of the linear polarized light was determined with passage through the samples .

This value along with the characteristic stress optical coefficient, which carried the designation optical activity, enabled determination of the principal residual stresses perpendicular to the optical path within the veneering porcelain. It could not determine the stresses within the zirconia framework . The differences between the rapid and slow cooled samples are observable in a color-coded residual stress distribution. The residual stresses revealed are those of a radial nature, as such the magnitude of these stresses at the external surface is minimal. In addition the act of sectioning the axial symmetric crowns would have modified the hoop stress components.

Materials and methods

Crown/sample model

A rotationally symmetric crown form similar to that previously invested by Lenz et al. was selected to ensure the production of identical samples and to simplify the experimental determination of the surface temperature corresponding with the geometry of a premolar tooth. Anatomically precise crowns, while possible to evaluate in a similar manner would have drastically increased the difficulty of production, especially regarding thermocouple placement and subsequent cross-sectioning of similar veneered copings.

The copings (four of each group) were designed by the inLab software (inLab 3.80, Sirona, Germany) and milled using an inLab milling machine (inLab MC XL, Sirona, Germany). The Y-TZP partially sintered frameworks (VITA In-Ceram YZ) were sintered to full density for 2 h at 1530 °C in a furnace (VITA ZYrcomat, VITA, Germany). After sintering the wall thickness was on average 0.7 mm with constant thickness frameworks. The anatomically supported frameworks had a wall thickness of 0.7 mm at the cervical region, which increased within the middle and cusp range from 1.3 mm and within the occlusal region to 1 mm. The overall height of all copings, constant wall or supported framework was 6.95 mm (see Fig. 1 , crown with constant wall thickness).

Fig. 1
A section of the axially rotational crown showing the thermocouple measurement points.

The first veneering step with the dentin porcelain (VM9; VITA, Germany) was the application of a very thin so called “wash-dentin layer” and fired at 950 °C. The first build up dentin firing was at 910 °C and the second was fused at 900 °C. Finally a glaze firing of the porcelain was carried out in each case at 900 °C in a Vacumat 4000 Premium T (VITA, Germany).

The veneer thickness of all samples was identical. At the cusp, complete thickness (framework and porcelain) of 1.4 mm was developed. Thus the porcelain thickness resulted in constant wall thickness framework copings of 0.7 mm on average. The porcelain of the anatomically supported crown had a maximum layer thickness of 0.4 mm.

Thermal measurement configuration

As a basis for supporting the firing of the samples a refractory honeycomb tray was used (Renfert, Germany). The honeycomb tray was drilled centrically to which a thermocouple protective glass with a length of approx. 23 mm was fixed. This length was selected, as the cervical edge of the samples was positioned a height above the tray of approx. 6 mm, which corresponds to the reality in dental labs.

To add retention to the crown a bent wire with protective glass sheath to support the thermocouples was included. The thermocouples were then placed on the internal and external selected measuring points (see Fig. 1 ) prior to being connected to the appropriate channels of data logging system (SCXI 1303, National Instrument, USA). For data recording and evaluation of the results the software LabView 8.5 was used

To enable comparative measurements of the cooling profiles with the attached thermocouples, a different furnace was used, namely a Programat P95 (Ivoclar Vivadent, Liechtenstein). The sample was located in this furnace more securely than in a furnace with mobile base. The more rapid mouth opening like mechanism of this furnace enabled the sample to be more quickly exposed to air movements of the environment than for the plinth lowering furnaces.

The thermocouples wires were led out of the firing chamber of the Programat P95, sealed with an insulating firing pillow, so that no air movement would falsify the measurement and vacuum could be used during the firing process. During all measurements the temperature at positions 1 and 10 were used as references to ensure comparability. Due to the measurement setup and the possible errors of temperature measurement in the furnace by using too many thermocouples, simultaneous measurement at all 10 points was not accomplished. Instead there were only 3 thermocouples in addition to the reference ones at positions 1 and 10 at the same time in the furnace. The temperature was recorded in total at five different measuring points during a firing process. Each sample was heated up and cooled down twice for the thermocouples placed at every measurement point. Altogether 48 measurements were accomplished per group.

Every sample was placed in the middle of the furnace chamber and the furnace was heated to the final temperature (840 °C) at a heating rate of 60 °C/min and held there for 10 min to achieve a homogeneous initial test temperature. The selected final temperature was 70 °C below the recommended dentin porcelain firing temperature to enable sintering and retention of the thermocouples into the soft porcelain on the surface as without this step the results would be inaccurate. After 10 min at the maximum temperature the chamber was opened directly and the furnace was switched off. The only difference between the slow and fast cooling was that for the former the furnace remained completely closed, until it reached 600 °C (the glass transition temperature of VITA VM9 is 603 °C), and thereafter was opened. The samples were not touched until they reached room temperature.

Polarimeter method

The uniform thickness framework samples for the polarimeter (StrainMatic ® M3/250, ilis GmbH, Germany) were produced following exactly the same firing steps and in the same sizes as for the thermo-couple measurements. After the firing, a thin slice 0.5 mm thick was cut using a water-cooled diamond saw from the mid-point of the coping and mounted in the polarimeter. With the polarimeter the turn value of the light through the crown was determined. The change of the angle of the linear polarized light was determined with passage through the samples .

This value along with the characteristic stress optical coefficient, which carried the designation optical activity, enabled determination of the principal residual stresses perpendicular to the optical path within the veneering porcelain. It could not determine the stresses within the zirconia framework . The differences between the rapid and slow cooled samples are observable in a color-coded residual stress distribution. The residual stresses revealed are those of a radial nature, as such the magnitude of these stresses at the external surface is minimal. In addition the act of sectioning the axial symmetric crowns would have modified the hoop stress components.

Only gold members can continue reading. Log In or Register to continue

Nov 28, 2017 | Posted by in Dental Materials | Comments Off on Thermal gradients and residual stresses in veneered Y-TZP frameworks
Premium Wordpress Themes by UFO Themes