Comparison of the Accuracy of 3 D Printed Prototypes Using the Stereolithography ( SLA ) Method with the Digital CAD Models

Each printing method and the materials used for prototype construction have distinct advantages and drawbacks. A large part of the 3D printer producers indicates high printing accuracy (positioning accuracies from 100 μm up to 1 μm and layer thickness from 100 μm tо 5 μm). In the process of layers deposition, the material is softened, melted or irradiated with a certain light source to polymerize. Thus, during the printing process, the volume of the materials is changing causing shrinking or expansion which often leads to the occurrence of precision errors in the prototype as opposed to the digital model. This phenomenon leads to distortions of the printed object known as a curling effect. Certain chemical compounds are added by the producers to multiple materials in order to enhance their mechanical and technological properties. However, this often results in changes of some printing parameters influencing the precision and quality of the manufactured object. The aim of the particular study is to establish the allowances with regard to the accuracy during manufacturing precise objects used in micro technologies as well as the necessary adjustments in dimensions because of the shrinkage and the positioning of the SLA 3D printed parts from photopolymer material.


Introduction
In recent years the development of electronics, machine building and micro-technologies has imposed extreme progress of different kinds of 3D printers (FDM, SLS, SLA, SLM, DLP, DMLS, EBM, etc.) working with wide variation of materials (PLA, ABS, PETG, PVA, photopolymers, ceramics, metals, etc.).The stereolithography is an additive process of 3D printing with liquid photopolymer that polymerizes under the laser irradiation.The work table is moving up-and downwards and after immersing in the resin the photopolymer hardens by selective irradiation of the laser beam.In that way, the printed object is formed as a bottom-up physical model with clearly discernible details.
In recent years 3D models of various anatomical regions have been presented in the international "3D Printing for professional -London 2016".The scientists from Washington State University are printing bone-like structures ("Inspiring ingenuity -Printing new bone") that could be used in orthopedic and dental procedures [1,2].This Corresponding author: eyankov@uni-ruse.bgbone-like material is acting as a scaffold for the new bone growth and after that, it dissolves with no side effects or needs for surgical interventions.Additionally, the complex anatomical shapes of the craniofacial or ear prosthesis, for example, are hard to be manufactured by conventional methods.The 3D printing can cope with this challenges of today more quickly and at less cost.After scanning the object with 3D scanners or cameras "https://digitalscan3d.com/" and thus generating the high-quality mesh, the digital model for the 3D printing is acquired.Simultaneously, the 3D solid model accuracy depends on the physical properties of the material during heating and cooling that results in changes of the object dimensions.With the increase of the model size, the errors accumulate for each single layer.These errors show positive or negative absolute value that could affect the durability and exploitation of the object afterward.The ultraviolet oven treatment changed also the solid model dimensions because of the residual stresses locked in between the layered structure.That's why the aim of the particular study is to establish the allowances with regard to the accuracy during SLA 3D printing as well as the necessary adjustments in dimensions because of the shrinkage and the positioning of the printed parts from photopolymer material.The measurement of the coordinates of the micro-squares after the 3D printing was carried out by a "Carl Zeiss" optical microscope at 12.5× magnification.The digital micrograph acquired by 5 Mpix camera and S-EYE software (Fig. 2) was used to measure the micro-grid deviations of the printed model.

Experimental procedures
As shown in Fig. 3, the coordinate grid method [3-5] that consists of measuring the relative "distortions" in different directions (εх and εу) was used to calculate the deviations during the photopolymerization of the printed model.The following equations were used: Where ΔX was the drift value of the X axis and ΔY was the drift value of the Y axis.The Xi and Yi were the original (initial) dimensions in the model and Xn,i and Yn,i was the changed values in dimension (Fig. 3).

Experimental results
After the mathematical processing of the results obtained (Fig. 4), the data are presented in graphical form (Fig. 5) showing the relative deviations of each cell against the reference digital model.Fig. 4

Fig. 2 .
Fig. 2. Digital image and measured values of the 3D printed model.

Fig. 3 .
Fig. 3.The setting of the deviations by the coordinate grid method.The coordinates of one of the corner point were used as reference fiducial marks.These values were entered into Excel tables in order to determine the differences between the coordinates of the specified digital model and those of the 3D printed prototype.The measured X and Y values were split into two separate tables for the purpose of defining appropriate tolerances between the prototype and the digital model of both axes.