Minimally invasive restorative treatments that preserve tooth structure are increasingly popular due to reduced biological impact. Successful outcomes depend not only on dentists' clinical skills but also on the fabrication of ultra-thin dental prostheses with material properties that meet clinical requirements. Computer Aided Manufacturing (CAM) such as Computerized Numerical Control (CNC) milling of lithium silicate or zirconia blocks, improve fabrication efficiency; however material constraints persist. Zirconia milling can induce micro-defects [1], [2], while lithium silicate requires wet grinding with diamond burs. According to the user guides for Zenostar and IPS e.max CAD, both materials require a minimum thickness of ≥ 0.4 mm for reliable machining.

Additive manufacturing (AM) enables the fabrication of ultrathin geometries (0.1–0.2 mm) [3], [4], [5], but glass-ceramics (e.g., lithium disilicate, ∼300–500 MPa strength) are at risk of long-term fracture [6]. Although AM zirconia offers superior mechanical properties (flexural strength ≥800 MPa, fracture toughness 5–10 MPa·m1/2) [7], it faces challenges related to dimensional accuracy for restorations thinner than 0.3 mm [5]. A new hybrid CAM product, the self-glazed zirconia ultra-thin veneer has also reported successful clinical applications; however its thickness is still limited to approximately 0.2–0.3 mm [8], [9].

Despite advances in Digital Light Processing (DLP) 3D printing enables possible fabrication of ultra-thin zirconia restorations (∼0.1 mm), clinical translation remains limited by a combination of interrelated accuracy factors across the entire process. These include:1.

Light-scattering artifacts: Particle-induced scattering during photopolymerization broadens cured features, particularly in sub-300 μm structures [10];

2.

Warpage during printing: The conversion of photosensitive monomers during polymerization leads to warpage ranging from 0.075 mm to 0.225 mm [11];

3.

Support structure limitations: Gravity-induced deformation of unsupported overhangs increases marginal gaps by 33 % (from 33.2 μm to 44.2μm) [12].

4.

Equipment constraints: High-resolution DLP systems (≤25μm pixels) achieve submillimeter precision [10], but the printing area is limited to 38.4 mm × 21.6 mm (for 1920 × 1080 pixels), restricting capacity to just 5–8 dental crowns per build.

Current research primarily focuses on isolated process parameters, such as optimizing slurry formulations or exposure settings to mitigate broadening effects and warpage during printing [10], [11], [13], [14], [15], refining support structure angles [16], or controlling sintering shrinkage through solid loading [17]. In addition, some studies investigate the placement angle of printed objects to reduce deformation [12]. However, few studies address the synergistic effects of full-process variables—from light-curing kinetics to post-sintering dimensional stability—on final prosthesis accuracy. This fragmented approach fails to establish clinically actionable standards, as evidenced by the wide variability in reported marginal gaps (22.8–140.4 μm) across DLP-printed zirconia crowns [18], [19], [20], [21], [22], highlighting the need for a comprehensive accuracy optimization framework.

To address this gap, this study systematically examines four critical yet understudied process control parameters that influence the dimensional fidelity of ultra-thin DLP printed zirconia dental prosthesis:1.

Light curing performance: Quantifying resolution limits via cure depth (Cd) and lateral broadening (Wexc) models to optimize photopolymerization accuracy.

2.

Support structure design: Establishing maximum allowable overhang angles through deformation mechanics to minimize unsupported structural distortion.

3.

Sintering shrinkage control: Investigating asymmetric shrinkage variations between inner and outer surfaces to refine post-sintering precision.

4.

Whole process-induced deformation: Controlling warpage from green-body formation to final sintered prostheses.

By systematically optimizing these process parameters, we demonstrate that DLP systems with a 70 μm resolution can successfully fabricate ultra-thin zirconia veneers (0.1 mm thickness) achieving < 60 μm root mean square (RMS) deviations in the intaglio area - equivalent to the precision of milled glass ceramics. These findings establish a validated framework for advancing AM zirconia restorations from laboratory prototypes to clinical-grade production.