J. Dent. 1992;
108
20: 108-l
14
Laboratory assessment of impression accuracy by clinical simulation R. W. Wassell and H. A. Abuasi Department of Operative Dentistry, The Dental School, Newcastle upon Tyne,
UK
ABSTRACT Some laboratory tests of impression material accuracy mimic the clinical situation (simulatory) while others attempt to quantify a materials individual properties. This review concentrates on simulatory testing and aims to give a classification of the numerous tests available. Measurements can be made of the impression itself or the resulting cast. Cast measurements are divided into those made of individual dies and those made of interdie relations. Contact measurement techniques have the advantage of simplicity but are potentially inaccurate because of die abrasion. Non-contact techniques can overcome the abrasion problem but the measurements, especially those made in three dimensions, may be difficult to interpret. Nevertheless, providing that care is taken to avoid parallax error non-contact methods are preferable as experimental variables are easier to control. Where measurements are made of individual dies these should include the die width across the
finishing line, as occlusal width measurements provide only limited information. A new concept of ‘differential die distortion’ (dimensional difference from the master model in one plane minus the dimensional difference in the perpendicular plane) provides a clinically relevant method of interpreting dimensional changes. Where measurements are made between dies movement of the individual dies within the master model must be prevented. Many of the test methods can be criticized as providing clinically unrealistic master models/dies or impression trays. Phantom head typodonts form a useful basis for the morphology of master models providing that undercuts are standardized and the master model temperature adequately controlled. KEY WORDS: J. Dent. 1992;
Impression materials, Accuracy 20: 108-l
14 (Received 5 July 1991;
accepted 2 October 1991)
Correspondenceshould be addressed to: Dr I?. W. Wassell, Department of Operative Dentistry, Framlington Place, Newcastle upon Tyne, NE2 4BW, UK.
INTRODUCTION Research into impression accuracy has relied heavily on in vitro tests rather than clinical in service evaluation. One reason for this is that clinically many variables conspire to produce inaccuracies. Such variables include tray flexion/ recoil and faulty impression technique. Another reason is that it is easier to make measurements in the laboratory than in the mouth. Two approaches have been taken in the in vitro assessment of impression materials. On the one hand properties thought to influence impression accuracy can be singled out and measured. On the other hand simulatory methods can be used where impressions are recorded of a master model and the resulting dimensional changes measured. The specification tests of the American Dental Association (ADA) (19) and British Standards Institution (BSI) (4269) are designed to consider individual properties @ 1992 Butterworth-Heinemann 0300-5712/92/020108-07
Ltd.
Table 1. BSI and ADA tests used to characterize some of the individual properties of impression materials. These tests do not reliably predict a material’s clinical accuracy Handling properties Mixing time Working time
Accuracy tests Compression test Strain in compression Detail reproduction Dimensional stability Compatibility with gypsum Metallizing bath compatibility Flow
(Table I). The main thrust of these tests is to ensure that impression materials meet minimum requirements before marketing and to provide continuing quality control. Although this approach has much to commend it, the specification tests cannot reliably predict a material’s accuracy under clinical conditions. It may be argued that the tests are insufficiently comprehensive to give a full
Wassell and Abuasi:
description of a material’s properties. For instance, the rate of development of a material’s elastic properties is not tested. A distorted impression may result from seating an impression into the patient’s mouth at a time when the elastic properties have reached a certain level (McCabe and Cat-rick, 1989). The viscoelastic properties of impression materials are clearly important and further work is being carried out to clarify their role. There are many types of simulatory tests available. Their main drawback is that none of them provides an absolute simulation of the mouth. Some tests provide realistic tooth morphology, arch form and temperature control, but none of them have mimicked soft-tissue consistency or the surface characteristics imparted by the oral fluids. Nevertheless, these simulatory tests can provide much useful information. A good example would be the study of undersized die dimensions resulting from tray wall or impression material recoil (Rehberg, 1977; Tjan et al., 1981; Wassell, 1984). Since there are so many simulatory tests available the aim of this literature review is to give a broad classification and to consider some of the more important drawbacks, particularly in respect of measurement technique. The simulatory tests can broadly be divided into two main groups depending on whether measurements are made of the impression itselfor of the resulting casts. Most workers have preferred to measure the casts. Measurements may either be made of the individual dies or the interdie relationships or both. Examples have been chosen to illustrate the various measuring techniques. The final section will consider the importance of using a master model with undercuts and temperature control to obtain adequate simulation.
DIRECT MEASUREMENTS IMPRESSION MATERIAL Direct within
In
Vito impression
recorded about the finishing preparation. The finishing critical region to get accurate paint the rest of the die with
Measurements
accuracy
109
line width or height of the line is probably the most since it is normal practice to die spacer.
of individual
dies
The measurement of dies poured from impressions is clinically a more realistic method of assessing impression accuracy than direct measurement of impression shrinkage. This is because the accuracy of the die will to a large extent determine the final tit of the restoration. In addition, it would be difficult to view microscopically the critical cervical part of a preparation within an impression. It must be recognized that the dimensions of the resulting die will be influenced by several factors relating to each of the following: the master die, the impression material, impression technique, the adhesive, the tray, time before pouring, dimensional changes of the die material and die material compatibility. In addition to standardizing laboratory variables such as temperature and humidity, it is clear that the above-mentioned variables also need to be carefully controlled. Type IV die stone is generally chosen because it has a minimal linear setting expansion of 0.1 per cent. Die measurements have been made in three main ways and the results compared with the master die: 1. Assessment of how well castings, made on each of the poured dies, fit the master die (Fig 2). 2. Assessment of how well a single master casting, made on the master die, tits each of the poured dies (Fig. 3). 3. Linear methods.
measurement
with contact
or non-contact
OF THE
measurement of an occlusal an impression
1. Individual surface
The ADA and BSI bodies both employ a scribed block which is used to form a disc of impression material (Fig. I). Measurements are made between the scribed lines on the block and the resulting lines on the impression discs to give an indication of time-dependent dimensional changes. In addition, the scribed lines provide a measure of surface reproduction. Light- and medium-bodied materials should be able to reproduce a 20 urn wide line. A logical progression of the Standards block method is to inscribe the occlusal surface of a crown preparation with fiducial marks and measure the resulting marks within the impression. Eames et al. (1979a) used such a technique to evaluate the effect on accuracy of different thicknesses of impression material contained within the tray. The major disadvantage of this method is that no information is
castings
from poured dies
Although this method best simulates the clinical situation it is used rarely. One reason for this is that a large number
12.5mm
12.5 mm
d
Fig. 1. Plan view of the BSI test block. The distance d-d is measured and compared with that on the impression. The vertical lines used for detail reproduction are ‘v’ shaped in section. They are 50, 20, and 75 pm wide respectively.
110
J. Dent. 1992; 20: No. 2
Lift + VT Master
Stone
die
Master
die
fig. 2. Test of casting fit (lift) simulating the clinical situation. The crowns are made on stone dies and then tried on the metal master die. An oversized stone die will result in a loose crown with no measurable lift.
I
J-7 Master
A T
Stone
die
A:
Difference
die
in seating
Fig. 4. Steel ring machined to give intimate fit with the axial walls of the master die. The position of the ring is measured on the stone die. If the ring seats further on the master die the stone die is undersized. If it seats less the stone die is oversized.
of castings has to be made. This is extremely time consuming and the results may be influenced by investing and casting variables which are difficult to standardize between laboratories. Accuracy is quantified by the ‘lift’ of the casting. This is the gap between the margin of the casting and finishing line of the preparation which may be measured with a travelling microscope or feeler gauge. The ‘lift’ may be influenced by the roughness of the fit surface of the casting, casting orientation and seating pressure. If the stone die is oversized the resulting casting will be a loose lit on the master die; looseness is difficult to measure. This problem is partly overcome by the ADA standard dies which are in the form of a full crown preparation and an inlay. It is assumed that if the crown is a loose fit then the inlay may produce a tight fit with measurable ‘lift’.
casting
made on the master die
a. Full coverage master casting This method differs from the previous one in that a single master casting is waxed and cast to tit the master die. The master casting is then tried on each of the poured dies. In this case if a poured die is undersized the casting will be loose so that the ‘lift’ cannot be measured. Several investigators (e.g. Eames et al., 1979b) have used this method and as only one casting needs to be made its popularity is easily explained. Casting variables and the problem of quantification of looseness of fit are compounded by a further difficulty: the rough tit surface of the master casting has the potential to abrade the stone die and affect the amount of ‘lift’. The stone dies may in turn
die
3. Non-clinical simulation where a single casting is made on the metal master die and tried on the stone dies. In this case an undersized stone die will not produce any measurable lift. Fig.
abrade the lit surface of the casting looser with repeated use.
b. Steel ring machined
so that it becomes
to fit master die
This method was used by Finger and Ohsawa (1983). A steel ring was machined to give an intimate fit to the axial walls of a tapered master die (Fig 4). The height of the ring from the top of the master die was measured with a toolmaker’s microscope. The ring was then tried on each of the resulting stone dies and its position remeasured. The axial discrepancy between the two positions was expressed as a percentage of the base diameter of the master die. Surprisingly, the results did not correlate with measurements made of free curing contraction which suggests that polymerization contraction is of minor importance in determining the accuracy of addition silicone impressions. The disadvantages of the technique include lack of standardization of ring alignment and seating pressure which could result in die abrasion. In addition the method relies on the production of a die symmetrical about its long axis. This means that a cylindrical tray, rarely used clinically, must be employed to obtain an even thickness of impression material. Clinically, the shape and rigidity of the impression tray can affect impression accuracy (Rehberg, 1977; Wassell, 1984) but these factors cannot be studied by this method. The effect of adjacent teeth on impression accuracy cannot be studied either.
3. Direct measurement 2. Master
Stone
die
of the die
Linear measurements may be made by contact or noncontact methods. Contact methods include the use of vernier callipers, micrometer, dial gauge or linear variable differential transformer (LVDT). Non-contact methods generally involve a travelling microscope, toolmaker’s microscope or the sophisticated reflex microscope (Setchell, 1984) which is capable of making measurements in three dimensions. Non-contact measurements are preferable as they avoid the risk of die abrasion by the measuring instrument. The master die must however have clear fiducial points which can be recorded in the impression and reproduced in the poured die. In both contact and non-contact methods the standardization of die orientation is mandatory to avoid parallax errors produced by viewing the master die and stone dies from
Wassell
a
and Abuasi:
In vita impression
111
accuracy
b
Fig. 5. a, A micrometer may be used to measure the width of a parallel-sided die. b, If used with a tapered die the orientation is clearly difficult to standardize and the stone may easily be gouged.
slightly different angles. The measurements made are usually compared with those of the master die and expressed as the amount under- or oversized. Alternatively, the differences are recorded as a percentage of the original dimension. Examples of the various methods will be discussed below:
a b Fig. 6. Differential die distortion. a, The width of the master die at the cervical margin is measured in two planes, A and B. b, Distortion of the die, typically the result of impression tray recoil, has resulted in the dimensions A and B. The dotted line represented the master die. This distortion is given by (A - A’) - (B - B’). Concentrically undersized or oversized dies would have a differential distortion of 0.
a. Micrometer The jaws of a micrometer consist of two parallel hardened steel plates which are applied to either side of the article to be measured under a force fixed by a spring-loaded ratchet. Brown (1980) used a parallel-sided cylindrical master die in combination with micrometer measurements of die diameter. No measurements were made of die height. The combination of cylindrical preparation and a cylindrical impression tray was a poor representation of the clinical situation. A tapered die however would have posed problems with orientation of the jaws of the micrometer. Moreover, the jaws would have gouged the stone of a tapered die (Fig. 5).
b. Simple non-contact measurement dies
of stone
Stone dies can be poured from impressions recorded of the ADA block. While an accurate assessment may be made of detail reproduction and surface finish (Morford et al., 1986) the simple linear measurements can provide only limited information for clinical dentistry.
.. .
III
ii
i
iv
a
i
ii
... III
iv
b Fig. 7. a, A concentrically undersized die may be compensated for by investment expansion: (i) horizontal section through preparation at the finishing line; (ii) concentrically undersized die; (iii) expanded casting (dotted line); (iv) acceptable fit of casting on preparation. b, A differentially distorted die cannot be compensated for: (i) horizontal section through preparation at the finishing line; (ii) differentially distorted die; (iii) casting on die; (iv) the casting will either catch on the axial walls resulting in ‘lift’or will have open margins if expanded.
c. Profile measurements Profile measurements involve lighting the die from below and viewing the resulting silhouette from above with a travelling or toolmaker’s microscope. If the die is oriented with its long axis perpendicular to the line of vision, width measurements can be made near the occlusal surface and at the level of the finishing line. It is crucial that the finishing line be measured as it is here that distortions due to impression tray recoil will be reflected most. Fiducial points scribed on the axial surface can facilitate measurements of height. This approach was used by Wassell (1984) and Johnson and Craig (1985, 1986). It must be remembered that the measurements will be only in one plane. A more complete assessment can be made by rotating the die through 90 degrees and remeasuring in a perpendicular
plane. If, for example, tray wall recoil has produced a decrease in buccolingual diameter but little change in mesiodistal diameter such distortions may then be detected. A concept of ‘differential die distortion’ has been evolved from a consideration of the distortion in two planes (Abuasi, 1990). Differential die distortion is simply the dimensional difference from the master die in one plane subtracted from the difference in the perpendicular plane (Fig. 6). If a die is slightly undersized this may be less important than if it is differentially distorted. This is because concentrically undersized dimensions may be compensated for during casting by the expansion of the investment (Fig. 7). Abuasi (1990) considered that a
112
J. Dent.
1992;
20:
No. 2
differential distortion of 60 pm or more would be clinically unacceptable while a differential distortion of 30 nm was considered acceptable. A grade of ‘possibly acceptable’ was assigned to a differential distortion between 30 and 60 urn. These figures were derived from the work of Christensen (1966) who studied the marginal discrepancies of cast metal restorations. In his study the least acceptable, visually accessible margin was open 39 l_trn although some barely acceptable margins were open by as much as 51 nm. In setting the three bands for differential die distortion it was assumed that the least marginal opening resulting from a differentially distorted die would be similar to the amount of differential distortion. The amount of potential marginal opening could of course be larger if the investment were to give less than ideal expansion.
INTERDIE
RELATIONSHIPS
If a bridge does not fit, the problem may be caused by distortion ofthe interdie relationship as well as inaccuracies of the individual dies. An important prerequisite of all interdie relationship studies is that there is no possibility of the individual dies moving slightly within the master model. If a phantom head typodont is used as a basis for a master model special precautions must be taken to prevent the teeth moving within their sockets. One-piece master models are clearly to be preferred.
Master
model and cast master bridge
This technique is similar to that used for individual dies except linked master castings are used to lit the abutments. Stauffer et al. (1976) used a machined stainless-steel master model with four turned abutments of 5 degree taper representing two canines and two molars. A surveyor was used to align the linked castings to the four dies during seating under a standard force. An average value of the ‘lift’ of the four castings was used to assess impression accuracy. Unfortunately, no distinction could be made between the amount of lift due to local die inaccuracies and that due to interdie inaccuracies. If the individual stone dies were undersized the fit of the master bridge would have been enhanced. In this case the method would not give a good indication of how well a bridge might tit the abutments clinically.
Master
model and machined
template
Reisbick and Matyas (1975) used an Invar steel master model which simulated two bridge abutments separated by an edentulous space. An Invar steel template was machined to tit the master die. Vertical slots in the template allowed optical measurements to be made of the degree of seating. After impressions had been recorded the template was placed on the resulting casts and the distance between the top of the die and the upper surface
Y
1
1 :
+X
0
rCX
Fig, 8. Stauffer’s method of measuring changes in interdie relationship using five dial gauges (arrows) to gauge shifts along the x and y axes. Measurements are first made of the master model and then the casts. The results are summated and expressed as a single vector of distortion for that impression material. of the template measured with a toolmaker’s microscope.
The problem of reliably seating the template and no distinction can be made between inaccuracies and interdie inaccuracies.
Contact measurement relationships
is obvious local die
of interdie
Stauffer et al. (1976) described a technique of measuring the interdie relationships of four master dies placed in the canine and molar regions of a master model. A jig consisting of a horizontal table, an L-shaped locating stop and five dial gauges as illustrated in Fig. 8 were used to measure the x and y coordinates of three of the dies in relation to the fourth die. The differences in the readings between the poured stone casts and the master model were expressed as a single vector of distortion which had both magnitude and direction. Unfortunately, the results of this method are difficult to rationalize clinically.
Non-contact relationship
measurement of interdie (two-dimensional)
Simple two-dimensional measurements may be made between the centre of the occlusal surfaces of two or more abutments with a travelling microscope. Valderhaug and Floystrand (1985) and other workers have used this approach, but it suffers the shortcoming that vertical changes in die relationship remain undetected. A simple method which could be used to detect vertical changes in die relationships would be useful.
Three-dimensional relationships
measurements
of interdie
Measurements may be made in three dimensions by using a travelling microscope to take x and y readings while a dial gauge or LVDT (Linkeetal., 1985) is used in the z axis.
Wassell and Abuasi:
More recently a totally non-contact method has been reported. Sim (1984) used a computer-assisted reflex microscope and elegantly described changes in interdie distances in relation to three external reference points. Measurements were made in the z axis by viewing a small dot of green light through the eyepiece. The apparent vertical position of the dot could be altered so that it appeared to land on the fiducial mark. The x, y and z coordinates were recorded and stored electronically. Despite the sophistication the clinical significance of the results was unclear.
THE EFFECT OF UNDERCUTS When an impression is withdrawn over an undercut, parts of the impression material will be in compression and other parts in tension. BSI 4269 and ADA specification No. 19 (American Dental Association, 1977; British Standards Institution, 1987) describe similar methods to test the recovery of a cylinder of impression material after being subjected to compressive stresses. This test is called the ‘compression set’ or ‘resistance of deformation’ and allows impression materials to be ranked according to their ease of distortion. For example, the addition silicones are the most resistant of the elastomers to permanent deformation while the polysulphides are the most susceptible. Mansfield and Wilson (1973) have devised a ‘tension set’ test to measure deformation under tensile forces. They found that the values of tension set and compression set did not correlate and thus recommended that both tests need to be done to evaluate a material properly. With both the ‘compression set’ and ‘tension set’ the time at which the measurement is made is important as the viscoelastic recovery of impression materials is time dependent. Undercuts have been incorporated in master dies and models to assess their effect on the accuracy of the resultant poured dies. Undercuts may arise in a number of clinical situations, detailed below.
in the die
In a supragingival crown preparation an undercut is often present below the cervical finishing line. A number of studies have incorporated such undercuts into their and master models (e.g. Eames et al., 1979a; Johnson Craig, 1985, 1986). This is a desirable feature of a clinical simulation but the amount of undercut should be quantified.
Undercut
between
the dies
Impression material may become distorted or torn where the gingival papilla is absent between adjacent preparations. Hanson and Eklund (1988) investigated the accuracy of different impression materials reproducing narrow spaces between dies with cervical undercuts. Their results showed that under these conditions a heavy-light bodied
accuracy
113
addition silicone impression material 15 min setting time could be removed or tearing.
with an extended without distortion
Undercut teeth
and
between
the
die
adjacent
The undercut associated with teeth adjacent to the preparation may influence the accuracy of the impression. Phantom head typodonts have been used in a number of studies to simulate this situation (e.g. Lacy et al., 1981). Ideally, two typodonts are needed. One to simulate a complete arch with master dies interposed between undercut standing teeth and the other with no teeth intervening between the dies. In the latter case when a stock tray is used the saddle areas are filled with a relatively large bulk of impression material. Any dimensional change of this intervening material will affect interdie accuracy. This approach was taken by Valderhaug and Floystrand (1984) but unfortunately only limited and inconclusive comparisons were made.
THERMAL
CHANGES
Impressions reach a temperature of approximately 33 “C after being in the mouth for 5 min (Jamani et al., 1988). On cooling to room temperature measurable dimensional changes occur as impression materials have a relatively high coefficient of thermal expansion. A similar rate and amount of impression temperature rise should be incorporated in laboratory tests. At present there is no agreement over the best way to do this. The standards organizations use a water-bath at 32°C while other workers prefer 35°C. Another approach is to use a heat source within the master model but the characteristics of heat flow into the impression have never been specified. Clearly, further work is needed in this area.
CONCLUSIONS Undercut
In vita impression
AND RECOMMENDATIONS
The various simulatory methods of assessing impression accuracy have been considered, and it is clear that while some methods are better than others no one technique is entirely satisfactory. Simple measurements made directly on the impression material are of limited relevance, especially if a clinically unrealistic die and impression tray have been used. When castings are used to evaluate accuracy other variables are introduced relating to the production of the casting. In addition, the risk of die or casting abrasion is particularly difficult to eliminate and the fit of a loose casting cannot be quantified easily. Contact measurements of the die also carry a risk of die abrasion and in common with noncontact measurements the master and stone dies must be precisely aligned to prevent parallax error. If the die can be reproducibly located and a clinically acceptable tolerance defined, then profile measurements offer a good
114
J. Dent.
1992;
20:
No. 2
alternative. To be comprehensive measurements should include the finishing line width, occlusal diameter and die height and should be made in two planes. Such measurements can give a useful indication of differential die distortion. The question of interdie relationships is more difficult to resolve. Two-dimensional measurements are of limited value as a distortion may occur in the third dimension and remain undetected. Three-dimensional measurements should be ideal to characterize interdie distortions, but require sophisticated instrumentation and the clinical implications of the results are unclear. To give the best simulation master dies and models should be constructed to copy oral morphology and should be temperature controlled. The commercially available phantom head typodonts provide teeth of standard shape so that undercuts of known dimensions can be incorporated. Ideally, both dentate and partially dentate master models should be used to provide clinically realistic results. The undercuts of the adjacent teeth give a good test for single dies while a one-piece partially dentate arch is a better test for interdie relationships. Whatever the means of measurement, statistical evidence should be given to support the reproducibility of the technique. It is hoped that the observations made in this article will stimulate the development of a generally accepted simulatory test for impression material accuracy. Impression material usage studies would also benefit from properly controlled clinical studies as it is in this environment that the materials will ultimately have to perform. These ‘in service’ tests could be of tremendous use in deciding how a material’s individual properties determine accuracy.
Acknowledgements The authors are extremely grateful to Dr J. F. McCabe, Department of Dental Materials Science, University of Newcastle upon Tyne for his constructive criticism of the paper.
References Abuasi H. A. (1990) A Comparison of a Range of Addition Silicone Putty-wash Impression Material in the One Stage Technique. MSc Dissertation, University of Newcastle upon Tyne. American Dental Association (1977) Revised American Dental Association Specification No. 19 for non aqueous elastomeric dental impression material. J. Am. Dent Assoc. 94,733-741. British Standards Institution (BS 4269: Part 1: 1987) Dental elastic impression material Part 1. Specification for elastomeric
impression
materials.
Brown D. (1980) Dental materials 1978 literature review, part II. J. Dent. 8, 222-248. Christensen G. J. (1966) Marginal fit of gold inlay castings. J. Prostbet. Dent. 16, 297-305. Eames W. B., Sieweke S. W., Wallace W. et al. (1979a) Elastomeric impression materials: effect of bulk on Dent 41, 304-307. accuracy. J. Pro&et. Eames W. B., Wallace S. W., Suway N. et al. (1979b) Accuracy and dimensional stability of elastomeric impression materials. J. Prosthet. Dent. 42, 159-162. Finger W. and Ohsawa M. (1983) Accuracy of stone-casts produced from selected additional-type silicone impressions. Stand. J. Dent Res. 91, 61-65. Hanson 0. and Eklund J. (1988) Impression for prosthodontic restorations reproducing narrow spaces and severe undercuts. Acta Odontol. Stand. 46, 199-206. Jamani K, Fayyad M., Harrington E. et al. (1988) Temperature changes of materials during impression taking. Br. Dent. J. 165, 129-132. Johnson G. and Craig R. (1985) Accuracy of four types of rubber impression materials compared with time of pour and repeat pour of models. J. Prostbet. Dent. 53, 484-490. Johnson G. and Craig R. (1986) Accuracy of addition silicone as a function of technique. J. Prosthet. Dent. 55, 197-203. Lacy A. M., Fukui H., Bellman T. et al. (1981) Time dependent accuracy of elastomer impression materials. Part II polyether and polyvinylsiloxane. J. Prostbet. Dent 45, 329-333. Linke B. A., Nicholls J. I. and Faucher R. R. (1985) Distortion analysis of stone casts made from impression materials. J. Prosthet. Dent. 54, 794-802. McCabe J. F. and Carrick T. E. (1989) Rheological properties of elastomers during setting. J. Dent. Res. 68, 1218-1222. Mansfield M. A. and Wilson H. J. (1973) A new method for determining the tension set of elastomeric impression materials. Br. Dent. J. 135, 101-105. Morford H., Tames R., Zardiackas L. et al. (1986) Effect of vacuum and pressure on accuracy reproducibility, and surface finish of stone casts made from polyvinylsiloxane. J. Prosthet Dent 55, 466-469. Rehberg H. J. (1977) The impression tray-an important factor in impression precision. Int. Dent. J. 27, 146-153. Reisbick M. and Matyas J. (1975) The accuracy of highly filled elastomeric impression materials. J. Prosthet. Dent. 33, 67-72. Setchell D. J. (1984) The reflex microscope-an assessment of the accuracy of 3 dimensional measurements using a new metrological instrument. J. Dent. Res. 63, British division 493 (abstr. 32). Sim K. Y. (1984) The Accuracy of Die Stone Casts from Impressions Made with Two Improved Alginates. MSc Dissertation, University of London. Stauffer J., Mayer J. and Nally J. (1976) Accuracy of six elastic impression materials used for complete-arch fixed partial dentures. J. Prosthet. Dent. 35, 407-415. Tjan A. H., Whang S. B. and Miller G. D. (1981) Why a rigid tray is important to the putty wash silicone impression method. J. Calif: Dent Assoc. 9, 53-58. Valderhaug J. and Floystrand F. (1984) The dimensional stability of elastomeric impression materials in custommade trays. J. Pro&et. Dent. 52, 514-517. Wassell R. W. (1984) Addition Silicone Impression Taken in Stock Trays. MSc Dissertation, University of London.
108
20: 108-l
14
Laboratory assessment of impression accuracy by clinical simulation R. W. Wassell and H. A. Abuasi Department of Operative Dentistry, The Dental School, Newcastle upon Tyne,
UK
ABSTRACT Some laboratory tests of impression material accuracy mimic the clinical situation (simulatory) while others attempt to quantify a materials individual properties. This review concentrates on simulatory testing and aims to give a classification of the numerous tests available. Measurements can be made of the impression itself or the resulting cast. Cast measurements are divided into those made of individual dies and those made of interdie relations. Contact measurement techniques have the advantage of simplicity but are potentially inaccurate because of die abrasion. Non-contact techniques can overcome the abrasion problem but the measurements, especially those made in three dimensions, may be difficult to interpret. Nevertheless, providing that care is taken to avoid parallax error non-contact methods are preferable as experimental variables are easier to control. Where measurements are made of individual dies these should include the die width across the
finishing line, as occlusal width measurements provide only limited information. A new concept of ‘differential die distortion’ (dimensional difference from the master model in one plane minus the dimensional difference in the perpendicular plane) provides a clinically relevant method of interpreting dimensional changes. Where measurements are made between dies movement of the individual dies within the master model must be prevented. Many of the test methods can be criticized as providing clinically unrealistic master models/dies or impression trays. Phantom head typodonts form a useful basis for the morphology of master models providing that undercuts are standardized and the master model temperature adequately controlled. KEY WORDS: J. Dent. 1992;
Impression materials, Accuracy 20: 108-l
14 (Received 5 July 1991;
accepted 2 October 1991)
Correspondenceshould be addressed to: Dr I?. W. Wassell, Department of Operative Dentistry, Framlington Place, Newcastle upon Tyne, NE2 4BW, UK.
INTRODUCTION Research into impression accuracy has relied heavily on in vitro tests rather than clinical in service evaluation. One reason for this is that clinically many variables conspire to produce inaccuracies. Such variables include tray flexion/ recoil and faulty impression technique. Another reason is that it is easier to make measurements in the laboratory than in the mouth. Two approaches have been taken in the in vitro assessment of impression materials. On the one hand properties thought to influence impression accuracy can be singled out and measured. On the other hand simulatory methods can be used where impressions are recorded of a master model and the resulting dimensional changes measured. The specification tests of the American Dental Association (ADA) (19) and British Standards Institution (BSI) (4269) are designed to consider individual properties @ 1992 Butterworth-Heinemann 0300-5712/92/020108-07
Ltd.
Table 1. BSI and ADA tests used to characterize some of the individual properties of impression materials. These tests do not reliably predict a material’s clinical accuracy Handling properties Mixing time Working time
Accuracy tests Compression test Strain in compression Detail reproduction Dimensional stability Compatibility with gypsum Metallizing bath compatibility Flow
(Table I). The main thrust of these tests is to ensure that impression materials meet minimum requirements before marketing and to provide continuing quality control. Although this approach has much to commend it, the specification tests cannot reliably predict a material’s accuracy under clinical conditions. It may be argued that the tests are insufficiently comprehensive to give a full
Wassell and Abuasi:
description of a material’s properties. For instance, the rate of development of a material’s elastic properties is not tested. A distorted impression may result from seating an impression into the patient’s mouth at a time when the elastic properties have reached a certain level (McCabe and Cat-rick, 1989). The viscoelastic properties of impression materials are clearly important and further work is being carried out to clarify their role. There are many types of simulatory tests available. Their main drawback is that none of them provides an absolute simulation of the mouth. Some tests provide realistic tooth morphology, arch form and temperature control, but none of them have mimicked soft-tissue consistency or the surface characteristics imparted by the oral fluids. Nevertheless, these simulatory tests can provide much useful information. A good example would be the study of undersized die dimensions resulting from tray wall or impression material recoil (Rehberg, 1977; Tjan et al., 1981; Wassell, 1984). Since there are so many simulatory tests available the aim of this literature review is to give a broad classification and to consider some of the more important drawbacks, particularly in respect of measurement technique. The simulatory tests can broadly be divided into two main groups depending on whether measurements are made of the impression itselfor of the resulting casts. Most workers have preferred to measure the casts. Measurements may either be made of the individual dies or the interdie relationships or both. Examples have been chosen to illustrate the various measuring techniques. The final section will consider the importance of using a master model with undercuts and temperature control to obtain adequate simulation.
DIRECT MEASUREMENTS IMPRESSION MATERIAL Direct within
In
Vito impression
recorded about the finishing preparation. The finishing critical region to get accurate paint the rest of the die with
Measurements
accuracy
109
line width or height of the line is probably the most since it is normal practice to die spacer.
of individual
dies
The measurement of dies poured from impressions is clinically a more realistic method of assessing impression accuracy than direct measurement of impression shrinkage. This is because the accuracy of the die will to a large extent determine the final tit of the restoration. In addition, it would be difficult to view microscopically the critical cervical part of a preparation within an impression. It must be recognized that the dimensions of the resulting die will be influenced by several factors relating to each of the following: the master die, the impression material, impression technique, the adhesive, the tray, time before pouring, dimensional changes of the die material and die material compatibility. In addition to standardizing laboratory variables such as temperature and humidity, it is clear that the above-mentioned variables also need to be carefully controlled. Type IV die stone is generally chosen because it has a minimal linear setting expansion of 0.1 per cent. Die measurements have been made in three main ways and the results compared with the master die: 1. Assessment of how well castings, made on each of the poured dies, fit the master die (Fig 2). 2. Assessment of how well a single master casting, made on the master die, tits each of the poured dies (Fig. 3). 3. Linear methods.
measurement
with contact
or non-contact
OF THE
measurement of an occlusal an impression
1. Individual surface
The ADA and BSI bodies both employ a scribed block which is used to form a disc of impression material (Fig. I). Measurements are made between the scribed lines on the block and the resulting lines on the impression discs to give an indication of time-dependent dimensional changes. In addition, the scribed lines provide a measure of surface reproduction. Light- and medium-bodied materials should be able to reproduce a 20 urn wide line. A logical progression of the Standards block method is to inscribe the occlusal surface of a crown preparation with fiducial marks and measure the resulting marks within the impression. Eames et al. (1979a) used such a technique to evaluate the effect on accuracy of different thicknesses of impression material contained within the tray. The major disadvantage of this method is that no information is
castings
from poured dies
Although this method best simulates the clinical situation it is used rarely. One reason for this is that a large number
12.5mm
12.5 mm
d
Fig. 1. Plan view of the BSI test block. The distance d-d is measured and compared with that on the impression. The vertical lines used for detail reproduction are ‘v’ shaped in section. They are 50, 20, and 75 pm wide respectively.
110
J. Dent. 1992; 20: No. 2
Lift + VT Master
Stone
die
Master
die
fig. 2. Test of casting fit (lift) simulating the clinical situation. The crowns are made on stone dies and then tried on the metal master die. An oversized stone die will result in a loose crown with no measurable lift.
I
J-7 Master
A T
Stone
die
A:
Difference
die
in seating
Fig. 4. Steel ring machined to give intimate fit with the axial walls of the master die. The position of the ring is measured on the stone die. If the ring seats further on the master die the stone die is undersized. If it seats less the stone die is oversized.
of castings has to be made. This is extremely time consuming and the results may be influenced by investing and casting variables which are difficult to standardize between laboratories. Accuracy is quantified by the ‘lift’ of the casting. This is the gap between the margin of the casting and finishing line of the preparation which may be measured with a travelling microscope or feeler gauge. The ‘lift’ may be influenced by the roughness of the fit surface of the casting, casting orientation and seating pressure. If the stone die is oversized the resulting casting will be a loose lit on the master die; looseness is difficult to measure. This problem is partly overcome by the ADA standard dies which are in the form of a full crown preparation and an inlay. It is assumed that if the crown is a loose fit then the inlay may produce a tight fit with measurable ‘lift’.
casting
made on the master die
a. Full coverage master casting This method differs from the previous one in that a single master casting is waxed and cast to tit the master die. The master casting is then tried on each of the poured dies. In this case if a poured die is undersized the casting will be loose so that the ‘lift’ cannot be measured. Several investigators (e.g. Eames et al., 1979b) have used this method and as only one casting needs to be made its popularity is easily explained. Casting variables and the problem of quantification of looseness of fit are compounded by a further difficulty: the rough tit surface of the master casting has the potential to abrade the stone die and affect the amount of ‘lift’. The stone dies may in turn
die
3. Non-clinical simulation where a single casting is made on the metal master die and tried on the stone dies. In this case an undersized stone die will not produce any measurable lift. Fig.
abrade the lit surface of the casting looser with repeated use.
b. Steel ring machined
so that it becomes
to fit master die
This method was used by Finger and Ohsawa (1983). A steel ring was machined to give an intimate fit to the axial walls of a tapered master die (Fig 4). The height of the ring from the top of the master die was measured with a toolmaker’s microscope. The ring was then tried on each of the resulting stone dies and its position remeasured. The axial discrepancy between the two positions was expressed as a percentage of the base diameter of the master die. Surprisingly, the results did not correlate with measurements made of free curing contraction which suggests that polymerization contraction is of minor importance in determining the accuracy of addition silicone impressions. The disadvantages of the technique include lack of standardization of ring alignment and seating pressure which could result in die abrasion. In addition the method relies on the production of a die symmetrical about its long axis. This means that a cylindrical tray, rarely used clinically, must be employed to obtain an even thickness of impression material. Clinically, the shape and rigidity of the impression tray can affect impression accuracy (Rehberg, 1977; Wassell, 1984) but these factors cannot be studied by this method. The effect of adjacent teeth on impression accuracy cannot be studied either.
3. Direct measurement 2. Master
Stone
die
of the die
Linear measurements may be made by contact or noncontact methods. Contact methods include the use of vernier callipers, micrometer, dial gauge or linear variable differential transformer (LVDT). Non-contact methods generally involve a travelling microscope, toolmaker’s microscope or the sophisticated reflex microscope (Setchell, 1984) which is capable of making measurements in three dimensions. Non-contact measurements are preferable as they avoid the risk of die abrasion by the measuring instrument. The master die must however have clear fiducial points which can be recorded in the impression and reproduced in the poured die. In both contact and non-contact methods the standardization of die orientation is mandatory to avoid parallax errors produced by viewing the master die and stone dies from
Wassell
a
and Abuasi:
In vita impression
111
accuracy
b
Fig. 5. a, A micrometer may be used to measure the width of a parallel-sided die. b, If used with a tapered die the orientation is clearly difficult to standardize and the stone may easily be gouged.
slightly different angles. The measurements made are usually compared with those of the master die and expressed as the amount under- or oversized. Alternatively, the differences are recorded as a percentage of the original dimension. Examples of the various methods will be discussed below:
a b Fig. 6. Differential die distortion. a, The width of the master die at the cervical margin is measured in two planes, A and B. b, Distortion of the die, typically the result of impression tray recoil, has resulted in the dimensions A and B. The dotted line represented the master die. This distortion is given by (A - A’) - (B - B’). Concentrically undersized or oversized dies would have a differential distortion of 0.
a. Micrometer The jaws of a micrometer consist of two parallel hardened steel plates which are applied to either side of the article to be measured under a force fixed by a spring-loaded ratchet. Brown (1980) used a parallel-sided cylindrical master die in combination with micrometer measurements of die diameter. No measurements were made of die height. The combination of cylindrical preparation and a cylindrical impression tray was a poor representation of the clinical situation. A tapered die however would have posed problems with orientation of the jaws of the micrometer. Moreover, the jaws would have gouged the stone of a tapered die (Fig. 5).
b. Simple non-contact measurement dies
of stone
Stone dies can be poured from impressions recorded of the ADA block. While an accurate assessment may be made of detail reproduction and surface finish (Morford et al., 1986) the simple linear measurements can provide only limited information for clinical dentistry.
.. .
III
ii
i
iv
a
i
ii
... III
iv
b Fig. 7. a, A concentrically undersized die may be compensated for by investment expansion: (i) horizontal section through preparation at the finishing line; (ii) concentrically undersized die; (iii) expanded casting (dotted line); (iv) acceptable fit of casting on preparation. b, A differentially distorted die cannot be compensated for: (i) horizontal section through preparation at the finishing line; (ii) differentially distorted die; (iii) casting on die; (iv) the casting will either catch on the axial walls resulting in ‘lift’or will have open margins if expanded.
c. Profile measurements Profile measurements involve lighting the die from below and viewing the resulting silhouette from above with a travelling or toolmaker’s microscope. If the die is oriented with its long axis perpendicular to the line of vision, width measurements can be made near the occlusal surface and at the level of the finishing line. It is crucial that the finishing line be measured as it is here that distortions due to impression tray recoil will be reflected most. Fiducial points scribed on the axial surface can facilitate measurements of height. This approach was used by Wassell (1984) and Johnson and Craig (1985, 1986). It must be remembered that the measurements will be only in one plane. A more complete assessment can be made by rotating the die through 90 degrees and remeasuring in a perpendicular
plane. If, for example, tray wall recoil has produced a decrease in buccolingual diameter but little change in mesiodistal diameter such distortions may then be detected. A concept of ‘differential die distortion’ has been evolved from a consideration of the distortion in two planes (Abuasi, 1990). Differential die distortion is simply the dimensional difference from the master die in one plane subtracted from the difference in the perpendicular plane (Fig. 6). If a die is slightly undersized this may be less important than if it is differentially distorted. This is because concentrically undersized dimensions may be compensated for during casting by the expansion of the investment (Fig. 7). Abuasi (1990) considered that a
112
J. Dent.
1992;
20:
No. 2
differential distortion of 60 pm or more would be clinically unacceptable while a differential distortion of 30 nm was considered acceptable. A grade of ‘possibly acceptable’ was assigned to a differential distortion between 30 and 60 urn. These figures were derived from the work of Christensen (1966) who studied the marginal discrepancies of cast metal restorations. In his study the least acceptable, visually accessible margin was open 39 l_trn although some barely acceptable margins were open by as much as 51 nm. In setting the three bands for differential die distortion it was assumed that the least marginal opening resulting from a differentially distorted die would be similar to the amount of differential distortion. The amount of potential marginal opening could of course be larger if the investment were to give less than ideal expansion.
INTERDIE
RELATIONSHIPS
If a bridge does not fit, the problem may be caused by distortion ofthe interdie relationship as well as inaccuracies of the individual dies. An important prerequisite of all interdie relationship studies is that there is no possibility of the individual dies moving slightly within the master model. If a phantom head typodont is used as a basis for a master model special precautions must be taken to prevent the teeth moving within their sockets. One-piece master models are clearly to be preferred.
Master
model and cast master bridge
This technique is similar to that used for individual dies except linked master castings are used to lit the abutments. Stauffer et al. (1976) used a machined stainless-steel master model with four turned abutments of 5 degree taper representing two canines and two molars. A surveyor was used to align the linked castings to the four dies during seating under a standard force. An average value of the ‘lift’ of the four castings was used to assess impression accuracy. Unfortunately, no distinction could be made between the amount of lift due to local die inaccuracies and that due to interdie inaccuracies. If the individual stone dies were undersized the fit of the master bridge would have been enhanced. In this case the method would not give a good indication of how well a bridge might tit the abutments clinically.
Master
model and machined
template
Reisbick and Matyas (1975) used an Invar steel master model which simulated two bridge abutments separated by an edentulous space. An Invar steel template was machined to tit the master die. Vertical slots in the template allowed optical measurements to be made of the degree of seating. After impressions had been recorded the template was placed on the resulting casts and the distance between the top of the die and the upper surface
Y
1
1 :
+X
0
rCX
Fig, 8. Stauffer’s method of measuring changes in interdie relationship using five dial gauges (arrows) to gauge shifts along the x and y axes. Measurements are first made of the master model and then the casts. The results are summated and expressed as a single vector of distortion for that impression material. of the template measured with a toolmaker’s microscope.
The problem of reliably seating the template and no distinction can be made between inaccuracies and interdie inaccuracies.
Contact measurement relationships
is obvious local die
of interdie
Stauffer et al. (1976) described a technique of measuring the interdie relationships of four master dies placed in the canine and molar regions of a master model. A jig consisting of a horizontal table, an L-shaped locating stop and five dial gauges as illustrated in Fig. 8 were used to measure the x and y coordinates of three of the dies in relation to the fourth die. The differences in the readings between the poured stone casts and the master model were expressed as a single vector of distortion which had both magnitude and direction. Unfortunately, the results of this method are difficult to rationalize clinically.
Non-contact relationship
measurement of interdie (two-dimensional)
Simple two-dimensional measurements may be made between the centre of the occlusal surfaces of two or more abutments with a travelling microscope. Valderhaug and Floystrand (1985) and other workers have used this approach, but it suffers the shortcoming that vertical changes in die relationship remain undetected. A simple method which could be used to detect vertical changes in die relationships would be useful.
Three-dimensional relationships
measurements
of interdie
Measurements may be made in three dimensions by using a travelling microscope to take x and y readings while a dial gauge or LVDT (Linkeetal., 1985) is used in the z axis.
Wassell and Abuasi:
More recently a totally non-contact method has been reported. Sim (1984) used a computer-assisted reflex microscope and elegantly described changes in interdie distances in relation to three external reference points. Measurements were made in the z axis by viewing a small dot of green light through the eyepiece. The apparent vertical position of the dot could be altered so that it appeared to land on the fiducial mark. The x, y and z coordinates were recorded and stored electronically. Despite the sophistication the clinical significance of the results was unclear.
THE EFFECT OF UNDERCUTS When an impression is withdrawn over an undercut, parts of the impression material will be in compression and other parts in tension. BSI 4269 and ADA specification No. 19 (American Dental Association, 1977; British Standards Institution, 1987) describe similar methods to test the recovery of a cylinder of impression material after being subjected to compressive stresses. This test is called the ‘compression set’ or ‘resistance of deformation’ and allows impression materials to be ranked according to their ease of distortion. For example, the addition silicones are the most resistant of the elastomers to permanent deformation while the polysulphides are the most susceptible. Mansfield and Wilson (1973) have devised a ‘tension set’ test to measure deformation under tensile forces. They found that the values of tension set and compression set did not correlate and thus recommended that both tests need to be done to evaluate a material properly. With both the ‘compression set’ and ‘tension set’ the time at which the measurement is made is important as the viscoelastic recovery of impression materials is time dependent. Undercuts have been incorporated in master dies and models to assess their effect on the accuracy of the resultant poured dies. Undercuts may arise in a number of clinical situations, detailed below.
in the die
In a supragingival crown preparation an undercut is often present below the cervical finishing line. A number of studies have incorporated such undercuts into their and master models (e.g. Eames et al., 1979a; Johnson Craig, 1985, 1986). This is a desirable feature of a clinical simulation but the amount of undercut should be quantified.
Undercut
between
the dies
Impression material may become distorted or torn where the gingival papilla is absent between adjacent preparations. Hanson and Eklund (1988) investigated the accuracy of different impression materials reproducing narrow spaces between dies with cervical undercuts. Their results showed that under these conditions a heavy-light bodied
accuracy
113
addition silicone impression material 15 min setting time could be removed or tearing.
with an extended without distortion
Undercut teeth
and
between
the
die
adjacent
The undercut associated with teeth adjacent to the preparation may influence the accuracy of the impression. Phantom head typodonts have been used in a number of studies to simulate this situation (e.g. Lacy et al., 1981). Ideally, two typodonts are needed. One to simulate a complete arch with master dies interposed between undercut standing teeth and the other with no teeth intervening between the dies. In the latter case when a stock tray is used the saddle areas are filled with a relatively large bulk of impression material. Any dimensional change of this intervening material will affect interdie accuracy. This approach was taken by Valderhaug and Floystrand (1984) but unfortunately only limited and inconclusive comparisons were made.
THERMAL
CHANGES
Impressions reach a temperature of approximately 33 “C after being in the mouth for 5 min (Jamani et al., 1988). On cooling to room temperature measurable dimensional changes occur as impression materials have a relatively high coefficient of thermal expansion. A similar rate and amount of impression temperature rise should be incorporated in laboratory tests. At present there is no agreement over the best way to do this. The standards organizations use a water-bath at 32°C while other workers prefer 35°C. Another approach is to use a heat source within the master model but the characteristics of heat flow into the impression have never been specified. Clearly, further work is needed in this area.
CONCLUSIONS Undercut
In vita impression
AND RECOMMENDATIONS
The various simulatory methods of assessing impression accuracy have been considered, and it is clear that while some methods are better than others no one technique is entirely satisfactory. Simple measurements made directly on the impression material are of limited relevance, especially if a clinically unrealistic die and impression tray have been used. When castings are used to evaluate accuracy other variables are introduced relating to the production of the casting. In addition, the risk of die or casting abrasion is particularly difficult to eliminate and the fit of a loose casting cannot be quantified easily. Contact measurements of the die also carry a risk of die abrasion and in common with noncontact measurements the master and stone dies must be precisely aligned to prevent parallax error. If the die can be reproducibly located and a clinically acceptable tolerance defined, then profile measurements offer a good
114
J. Dent.
1992;
20:
No. 2
alternative. To be comprehensive measurements should include the finishing line width, occlusal diameter and die height and should be made in two planes. Such measurements can give a useful indication of differential die distortion. The question of interdie relationships is more difficult to resolve. Two-dimensional measurements are of limited value as a distortion may occur in the third dimension and remain undetected. Three-dimensional measurements should be ideal to characterize interdie distortions, but require sophisticated instrumentation and the clinical implications of the results are unclear. To give the best simulation master dies and models should be constructed to copy oral morphology and should be temperature controlled. The commercially available phantom head typodonts provide teeth of standard shape so that undercuts of known dimensions can be incorporated. Ideally, both dentate and partially dentate master models should be used to provide clinically realistic results. The undercuts of the adjacent teeth give a good test for single dies while a one-piece partially dentate arch is a better test for interdie relationships. Whatever the means of measurement, statistical evidence should be given to support the reproducibility of the technique. It is hoped that the observations made in this article will stimulate the development of a generally accepted simulatory test for impression material accuracy. Impression material usage studies would also benefit from properly controlled clinical studies as it is in this environment that the materials will ultimately have to perform. These ‘in service’ tests could be of tremendous use in deciding how a material’s individual properties determine accuracy.
Acknowledgements The authors are extremely grateful to Dr J. F. McCabe, Department of Dental Materials Science, University of Newcastle upon Tyne for his constructive criticism of the paper.
References Abuasi H. A. (1990) A Comparison of a Range of Addition Silicone Putty-wash Impression Material in the One Stage Technique. MSc Dissertation, University of Newcastle upon Tyne. American Dental Association (1977) Revised American Dental Association Specification No. 19 for non aqueous elastomeric dental impression material. J. Am. Dent Assoc. 94,733-741. British Standards Institution (BS 4269: Part 1: 1987) Dental elastic impression material Part 1. Specification for elastomeric
impression
materials.
Brown D. (1980) Dental materials 1978 literature review, part II. J. Dent. 8, 222-248. Christensen G. J. (1966) Marginal fit of gold inlay castings. J. Prostbet. Dent. 16, 297-305. Eames W. B., Sieweke S. W., Wallace W. et al. (1979a) Elastomeric impression materials: effect of bulk on Dent 41, 304-307. accuracy. J. Pro&et. Eames W. B., Wallace S. W., Suway N. et al. (1979b) Accuracy and dimensional stability of elastomeric impression materials. J. Prosthet. Dent. 42, 159-162. Finger W. and Ohsawa M. (1983) Accuracy of stone-casts produced from selected additional-type silicone impressions. Stand. J. Dent Res. 91, 61-65. Hanson 0. and Eklund J. (1988) Impression for prosthodontic restorations reproducing narrow spaces and severe undercuts. Acta Odontol. Stand. 46, 199-206. Jamani K, Fayyad M., Harrington E. et al. (1988) Temperature changes of materials during impression taking. Br. Dent. J. 165, 129-132. Johnson G. and Craig R. (1985) Accuracy of four types of rubber impression materials compared with time of pour and repeat pour of models. J. Prostbet. Dent. 53, 484-490. Johnson G. and Craig R. (1986) Accuracy of addition silicone as a function of technique. J. Prosthet. Dent. 55, 197-203. Lacy A. M., Fukui H., Bellman T. et al. (1981) Time dependent accuracy of elastomer impression materials. Part II polyether and polyvinylsiloxane. J. Prostbet. Dent 45, 329-333. Linke B. A., Nicholls J. I. and Faucher R. R. (1985) Distortion analysis of stone casts made from impression materials. J. Prosthet. Dent. 54, 794-802. McCabe J. F. and Carrick T. E. (1989) Rheological properties of elastomers during setting. J. Dent. Res. 68, 1218-1222. Mansfield M. A. and Wilson H. J. (1973) A new method for determining the tension set of elastomeric impression materials. Br. Dent. J. 135, 101-105. Morford H., Tames R., Zardiackas L. et al. (1986) Effect of vacuum and pressure on accuracy reproducibility, and surface finish of stone casts made from polyvinylsiloxane. J. Prosthet Dent 55, 466-469. Rehberg H. J. (1977) The impression tray-an important factor in impression precision. Int. Dent. J. 27, 146-153. Reisbick M. and Matyas J. (1975) The accuracy of highly filled elastomeric impression materials. J. Prosthet. Dent. 33, 67-72. Setchell D. J. (1984) The reflex microscope-an assessment of the accuracy of 3 dimensional measurements using a new metrological instrument. J. Dent. Res. 63, British division 493 (abstr. 32). Sim K. Y. (1984) The Accuracy of Die Stone Casts from Impressions Made with Two Improved Alginates. MSc Dissertation, University of London. Stauffer J., Mayer J. and Nally J. (1976) Accuracy of six elastic impression materials used for complete-arch fixed partial dentures. J. Prosthet. Dent. 35, 407-415. Tjan A. H., Whang S. B. and Miller G. D. (1981) Why a rigid tray is important to the putty wash silicone impression method. J. Calif: Dent Assoc. 9, 53-58. Valderhaug J. and Floystrand F. (1984) The dimensional stability of elastomeric impression materials in custommade trays. J. Pro&et. Dent. 52, 514-517. Wassell R. W. (1984) Addition Silicone Impression Taken in Stock Trays. MSc Dissertation, University of London.
