Sprue design plays a major role in casting any type of alloy for fixed and removable prosthesis. Preston and Berger1 stated that "spruing is an art which is not understood carefully". Many researchers examined the effect of different sprue designs on the porosity of fixed partial denture components.2-4 However, the influence of sprue design on soundness of RPD frameworks was not adequately investigated.
A removable partial denture (RPD) framework should be free from defects within its structure in order to fulfill its function. Yet, many RPD castings contain internal and external defects of varying size and shape which can cause failure of the prosthesis5 or result in an inferior appliance.6
Defects such as fissures and microporosities with surface connections may result in increased corrosion of the alloy.7 Large pores or many porosities when situated in critical regions of the RPD frame may promote fracture.8 The presence of defects in RPD castings is said to depend on many factors including the spruing method and the metal feeding direction. Evaluation of such defects can be achieved both visually and radiographically. Faults in a dental casting can result from one or more elements such as spruing,9-15 investing,16 casting,10,16,20 alloy melting,9,11,12,16  wax elimination, 10,16-20 and freezing of casting.9 Dental technicians have the habit of ignoring the effect of sprue design on the casting by producing different designs which may affect the quality of castings. This study investigated the effect of various sprue designs on soundness of RPD castings

 
A dental cast representing Kennedy Class I Modification I Removable Partial Denture (RPD) was chosen to serve as the master cast. The RPD framework design was decided utilizing three terminal abutments, teeth # 13 with a ledge resl and # 16 and # 24 with mesial rests. Circumferential clasps with retention bucaiiy and reciprocal arms lingually were designed on teeth # 16 and # 24. The RPD clasp with mid-buccal retention was designed on tooth # 13. Proximal plates were placed on the mesial of # 16 and distals of # 13 and # 24. Anterior and posterior palatal strap was chosen as the major connector. Undercuts on the master cast were blocked out. The cast was relieved, and external and internal finish lines were established. The master cast was then duplicated using Perplex* duplicating material to produce 15 identical refractory casts. The casts were divided to three groups. Each group of five casts were sprued with a different spruing arrangement using pre-formed wax**. Three different sprue design arrangements were tested.
The first spruing arrangement was designated as the tree sprue design. A tree sprue design consisted of four wax formers, each 3mm in diameter, going directly to four different spots on the major connector.
The second spruing arrangement was designated as the circular sprue design of which circular feeder was made from a 3mm diameter wax. Auxiliary sprues emanating from the circular feeder were six, each was 2mm in diameter and were attached directly to the major connector. The circular feeder was attached to the sprue button by four sprues, each 3mm in diameter.
The third sprue design was designated as the ball sprue design which consisted of four feeders, 3mm in diameter as with the tree sprue design except that a 5mm ball reservoir was placed on each feeder at 5mm away from the wax pattern.
Five refractory casts were thus sprued by one of each of the three spruing arrangements. All wax patterns were invested using the same batch of Wirovest*** investment material. Refractory casts were left to bench set for two hours before burnout.
A two-stage burn-out cycle was employed whereas the investment block was placed in a room temperature oven with the sprue base resting on the muffle floor. The oven was heated to 300°C and maintained at that temperature for 30 minutes. The ring was then rotated so that the sprue base is facing upward and the oven temperature was raised to 975°C in 3.5 hours. The investment was then heat soaked at 975°C for 45 minutes before it became ready for casting.
A centrifugal induction casting machine* was used and the metal was cast at 1200°C using Wironit** cobalt-chromium alloy. The centrifugal force used was 9 bars and is the same as is used with clinical RPD castings in the laboratories at King Saud University College of Dentistry. All RPD frameworks were cast using a full metal charge of 25-35 gms. Variation in weight of the casting metal was due to differences in spruing arrangement and size. The appropriate weight for each spruing arrangement was determined experimentally in a pilot study.
After recovering the castings from the investment, they were blasted with 50 micron aluminum oxide*** at 80 psi and were ultrasonicaliy cleaned. All 15 RPD frameworks were completed and void of any major deficiency. After photographing the cast RPD frameworks and marking each with a random number, all sprues were cut-off and all frameworks were finished and each was placed in an envelope carrying its random number. Examples of the laboratory procedure in producing the RPD frameworks are shown in Fig. 1 a and b. Two evaluation procedures were used to fully assess the absence of porosities from the cast RPD frameworks.

 
The visual rankings of clasps, major connectors and meshworks for the 15 frameworks are reported in Table 1. It was noted that there were no defects observed in any of the clasps. However, defects in major connectors were eight with as many as three (3) defects in one major connector. Defects in meshworks were eight (8), all of which occurred when a circular sprue arrangement was used with as many as two defects in one framework.
The radiographic rankings of the components examined are reported in Table 2. Again, there were no radiographic defects in any of the clasps. Defects in major connectors were 8. The three defects visually observed in one major connector were also seen radiographically. Defects in meshworks were 11. One of the meshwork that depicted two defects visually showed five defects radiographically. Otherwise, the visual and radiographic defects were identical.

 
The number of defects as a function of factors (spruing method, observation method and area examined) is reported in Table 3. Descriptive statistics (mean, standard deviations) for defects as a function of the three factors are shown in Fig. 2. Table 4 shows that the spruing method and area examined have no significant influence on defect occurrence. However, the occurrence of defects was significantly influenced by the method of observation. In other words, clasps, major connectors and meshworks were not evenly influenced by changing the sprue design; rather they were affected differently. To detect which of the three areas was most affected by changing sprue design, the multiple range analysis reported in Table 5 was conducted. The analysis showed that clasps were not affected, however major connectors and meshwork were affected. While meshworks seemed more affected by sprue design than the major connector, homogeneity test showed no significant difference between them.
The fact that clasps were less subjective to casting defects when compared to meshwork areas is likely due to the difference in their cross-sectional areas. Defects in the major connector areas
are likely due to the relatively large liquid metal volume in this area and its slower rate of solidification. When an area of the RPD frame remains liquified after the sprue feeders have solidified, this area becomes subjected to shrinkage porosity.

 
Results and statistical analysis of this study conclude that any of the three sprue designs used in this study is suitable for producing sound RPD frame castings. When the circular sprue design is used, allowances should be made to assure adequate meshwork thickness.

 
The author gratefully acknowledge the experimental assistance and guidance offered by Prof. H. Mohammed-Al Tahawi, Chairman Department of Removable Prosthetic Dental Sciences, KSU College of Dentistry. Acknowledgment is also due to Dr. Nazeer Khan, Biostatistician, Research Center for his valuable assistance in the statistical analysts.