Permanent Crown Resins: What Thermal Aging Tests Tell Dental Labs
A strength value alone cannot tell a dental lab which permanent crown resin fits its workflow. Check what was tested, how the specimen was made, and what happened after aging. In the studies reviewed here, thermal cycling repeatedly moved specimens between water baths at 5 and 55 °C to examine their response to a defined laboratory exposure. [1] [2] [3]
The three direct studies reviewed here illustrate why: one reported greater external-surface deviation alongside smaller measured marginal gaps; another detected no significant thermal-aging effect on crown fracture load or roughness; a third found differences between post-curing groups after aging, without measuring an unaged strength baseline. [1] [2] [3]
- Match the measurement and test setup: MPa, crown failure load and marginal gaps answer different questions.
- Check whether the evidence covers the resin, printer, post-curing and finishing sequence your lab plans to use.
- Read cycle counts as laboratory exposure conditions, not a verified number of service years.
Use the supplier checklist to turn these checks into specific documentation requests.
What evidence are we comparing?
These are laboratory studies of named commercial materials, prepared using specific printers and post-processing procedures. The background paper by Melc and colleagues helps distinguish test methods; it does not supply direct evidence for printed permanent crown resin. [4]
Table 1. What the evidence covers [1] [2] [3] [4]
| Study | Material and specimen | Exposure and measured outcomes | Main boundary |
|---|---|---|---|
| Peng et al. (2025) [1] | FREEPRINT crown; 20 DLP-fabricated crowns, 1.5 mm wall thickness | 10,000 cycles, 5–55 °C; 3D accuracy and marginal/internal gaps | One resin and printing workflow; no fracture test. |
| Sahin and Tokar (2026) [2] | Crowntec, VarseoSmile Crown Plus and milled VITA Enamic; 180 crowns on titanium abutments | 10,000 cycles, 5–55 °C; surface roughness and fracture load at three wall thicknesses | A particular crown–cement–abutment assembly; no cyclic mechanical loading. |
| Baytur and Diken Turksayar (2025; online 2024) [3] | Formlabs Permanent Crown; 90 discs, 10 mm diameter × 2 mm thick; nine post-curing groups | 5,000 cycles, 5–55 °C; optical and roughness measurements before/after aging; strength after aging | No unaged flexural-strength control; one resin and curing unit. |
| Melc et al. (2024), background only [4] | Milled CAD/CAM restoratives and direct composite bars | Combined thermal and mechanical aging; three-point bending | A methods comparison, not direct evidence for printed permanent crown resin. |
These studies answer different questions. Their values should not be combined into a universal ranking of materials.
First separate the test metrics
A value in MPa describes a different quantity from a crown failure load in N. Even measurements reported in the same unit can refer to different outcomes: a surface scan deviation and a marginal gap are both measured in micrometres, but use different reference points and methods. [1] [2] [3]
Table 2. Read the measurement before comparing the number [1] [2] [3] [4]
| Metric | What it measures | What must match for a useful comparison |
|---|---|---|
| Flexural strength — MPa | Stress calculated from a bending test on a defined specimen; Baytur and Diken Turksayar used biaxial loading of discs. [3] | Test method, specimen dimensions, preparation and conditioning. Biaxial discs and three-point bars are different test geometries. [4] |
| Crown fracture load — N | Force at failure of a crown within the tested support and loading assembly. [2] | Crown shape and thickness, abutment, cementation, load position and aging protocol. This is not the same quantity as MPa. |
| Surface trueness — RMS, μm | Root mean square surface deviation between the scanned crown and its CAD reference. [1] | Scanner, alignment method and measured region. RMS deviation is not a crown-to-abutment gap. |
| Marginal/internal gap — μm | Space at selected sites between a seated crown and its supporting form, estimated using a silicone replica. [1] | Designed cement space, seating method, measurement sites and replica procedure. Smaller is not automatically better. |
| Surface roughness — Ra, μm | Average profile roughness measured on a prepared surface. [2] [3] | Finishing sequence, measurement location, instrument settings, material and conditioning. |
| Color change — ΔE00 | CIEDE2000 color difference calculated from measurements before and after the specified exposure. [3] | Measurement conditions, specimen thickness, finishing, curing and aging medium. It is not a strength measurement. |
Use the same units and a comparable protocol. Matching units alone does not make two studies comparable.
For a lab comparing supplier reports, the practical first step is to place each result in the correct category. Then compare specimen preparation and test conditions. A higher number from a different bending method or a thicker crown does not by itself identify a better material.
Can dimensional deviation increase while a marginal gap gets smaller?
Yes. Peng and colleagues tested DLP-fabricated FREEPRINT crown restorations with 1.5 mm walls and a designed cement space of 50 μm. They assessed surface accuracy by scanning and fit using silicone replicas. The two reported outcomes below changed in opposite directions. [1]
Table 3. Two geometric outcomes in the same study [1]
| Reported measurement | Without thermal aging | After thermal aging | Interpretation |
|---|---|---|---|
| External-surface trueness, RMS (μm) | 60.8 ± 6.6 | 84.4 ± 4.6 | Greater deviation from the CAD reference. |
| Coronal-plane marginal gap (μm) | 70.8 ± 5.1 | 46.9 ± 6.9 | A smaller measured replica gap; not a general improvement claim. |
Mean ± standard deviation, from Peng et al., Results. The protocol used 10,000 cycles between 5 and 55 °C. [1]
The external-surface RMS increase was statistically significant (p < 0.001). The authors also reported reduced marginal and occlusal gaps at specified sites, while inter-sample precision was not significantly affected. Describing the result simply as “fit worsened” would lose these distinctions. [1]
The authors proposed water sorption and swelling as a possible explanation for the dimensional changes. That interpretation should remain separate from the measured results: this experiment did not directly establish the proposed mechanism. A smaller replica gap is also not, on its own, proof of better seating or clinical performance. [1]
Reporting note: the paper describes 20 crowns divided between unaged and aged groups, while its statistical section reports paired tests. Table 3 therefore follows the authors’ condition labels without assuming additional specimen-tracking details. [1]
What does “no significant aging effect” mean for crown fracture load?
Sahin and Tokar compared two printed composite resins with milled VITA Enamic using crowns cemented to titanium abutments. They tested three wall thicknesses. The fracture-load results were obtained from separate water-stored control and thermally aged groups. [2]
What to look for: compare the control and aged results within the same material and thickness before comparing brands. In this experiment, thickness affected fracture load, while the reported thermal-aging effect was not significant. [2]
Figure 1. Crown fracture load and surface roughness
Source: Sahin and Tokar, original Tables 2 and 3. [2]
View complete fracture-load data: means, ranges and statistical groups
Table 4. Crown fracture load under the conditions tested by Sahin and Tokar [2]
Scroll horizontally within the table to view all 7 columns.
| Material | Wall thickness (mm) | Control: water storage at 37 °C, n = 10 Mean ± SD (N) | Control Median (min–max), N | Thermal cycling: 10,000 cycles, n = 10 Mean ± SD (N) | Thermally aged Median (min–max), N | Pooled comparison Mean ± SD (N), Duncan group |
|---|---|---|---|---|---|---|
| Crowntec | 1.0 | 425.2 ± 66.3 | 444.1 (326.1–542.6) | 413.8 ± 76.7 | 395.6 (321.3–553.5) | 419.50 ± 70.00 d |
| Crowntec | 1.5 | 670.7 ± 74.8 | 676.7 (563.0–791.8) | 657.1 ± 70.8 | 665.9 (513.3–736.9) | 663.90 ± 71.21 c |
| Crowntec | 2.0 | 909.7 ± 97.4 | 892.5 (781.6–1063.2) | 902.4 ± 100.3 | 912.0 (756.5–1051.8) | 906.00 ± 96.30 b |
| VarseoSmile Crown Plus | 1.0 | 444.1 ± 82.9 | 420.5 (349.8–622.8) | 434.3 ± 72 | 423.1 (328.5–592.2) | 439.20 ± 75.76 d |
| VarseoSmile Crown Plus | 1.5 | 688.3 ± 70.8 | 687.3 (584.0–841.7) | 676.4 ± 69.4 | 679.4 (578.8–794.3) | 682.30 ± 68.51 c |
| VarseoSmile Crown Plus | 2.0 | 953.1 ± 89.8 | 939.5 (824.5–1109.1) | 935.2 ± 92.9 | 939.2 (782.3–1059.2) | 944.10 ± 89.40 b |
| VITA Enamic | 1.0 | 503.4 ± 51.5 | 506.0 (425.6–613.4) | 480.5 ± 65.1 | 459.3 (413.6–574.7) | 491.90 ± 58.29 d |
| VITA Enamic | 1.5 | 930.7 ± 83.6 | 937.6 (806.1–1063.1) | 898.4 ± 92.5 | 914.0 (782.8–1012.7) | 914.50 ± 87.37 b |
| VITA Enamic | 2.0 | 1291.3 ± 127.8 | 1296.4 (1042.7–1503.2) | 1245.2 ± 144.9 | 1229.2 (968.5–1447.1) | 1268.00 ± 135.07 a |
Data from the original Table 3; the final column separately reproduces the pooled control/aged comparison in original Table 4. Crowns were cemented to titanium abutments. Each material–thickness subgroup contained 10 control crowns stored in distilled water at 37 °C and 10 separate crowns exposed to 10,000 cycles between 5 and 55 °C, with 30 s intervals. These are independent fracture groups, not repeated fracture tests on the same crown. Shared Duncan letters indicate no statistically significant difference within that reported pooled comparison; they do not demonstrate equivalence. [2]
The authors reported no statistically significant effect of thermal aging on fracture load under these conditions (p > 0.05). Wall thickness and material affected the results: at 1.0 mm, the material comparison was not significant; at 1.5 and 2.0 mm, VITA Enamic had higher fracture loads than the two printed materials in this test assembly. [2]
These findings are useful within the experiment. They do not prove equivalence after aging, establish a clinical minimum thickness, or predict the result on a different supporting die. Thickness, crown geometry, cementation and loading position must accompany the numbers when a lab compares reports.
Surface roughness followed a different measurement design: the thermally aged crowns were measured before and after cycling. No significant pre-to-post change was detected in any material–thickness subgroup. This should not be confused with the independent groups used for the destructive fracture test. [2]
What can post-curing comparisons tell a lab?
Baytur and Diken Turksayar tested one Formlabs Permanent Crown resin in nine time–temperature groups. Each group received two post-curing cycles, followed by 5,000 thermal cycles. Color and roughness were measured before and after aging; biaxial flexural strength was measured only after aging. [3]
What to look for: the panels compare curing schedules after the same aging exposure. They help identify workflow sensitivity, but do not measure a before-and-after loss of strength. [3]
Figure 2. Post-curing conditions: strength, Weibull modulus and color change
Source: Baytur and Diken Turksayar, original Tables 1, 4 and 5. [3]
View complete post-curing data: strength, color change and statistical groups
Table 5. Post-aging strength and color change across the tested post-curing groups [3]
Scroll horizontally within the table to view all 5 columns.
| Time and temperature per curing cycle | Post-aging biaxial flexural strength Mean ± SD (MPa), Tukey group | Post-aging biaxial flexural strength Median (min–max), MPa | Color change after aging, ΔE00 Mean ± SD | Color change after aging, ΔE00 Median (min–max), Dunn group |
|---|---|---|---|---|
| 15 min at 40 °C | 225.14 ± 34.82 a | 232.37 (156.76–267.03) | 1.14 ± 0.34 | 1.13 (0.6–1.65) a,b,d |
| 15 min at 60 °C | 315.31 ± 27.27 b,c | 309.77 (272.93–368.55) | 0.83 ± 0.26 | 0.82 (0.32–1.16) a,c |
| 15 min at 80 °C | 315.51 ± 59.93 b,c | 290.12 (244.62–407.41) | 0.32 ± 0.18 | 0.27 (0.11–0.65) c |
| 20 min at 40 °C | 279.06 ± 57.84 a,b | 279.02 (172.04–362.39) | 1.44 ± 0.41 | 1.56 (0.47–1.91) a,b |
| 20 min at 60 °C | 293.41 ± 41.61 b | 283.77 (224.92–360.23) | 1.07 ± 0.44 | 1.09 (0.23–1.55) a,b,c |
| 20 min at 80 °C | 304.68 ± 43.98 b,c | 308.06 (250.55–380.92) | 0.46 ± 0.18 | 0.43 (0.23–0.8) c,d |
| 30 min at 40 °C | 320.31 ± 41.11 b,c | 319.55 (261.97–378.17) | 1.93 ± 0.44 | 1.96 (1.29–2.56) b |
| 30 min at 60 °C | 426.01 ± 39.51 d | 417.2 (381.05–508.45) | 1.47 ± 0.37 | 1.49 (1.03–2.13) a,b |
| 30 min at 80 °C | 368.33 ± 47.51 c,d | 367.02 (306.82–435.21) | 0.7 ± 0.44 | 0.58 (0.23–1.53) a,c |
Data reorganized from Baytur and Diken Turksayar, Tables 1 and 4. Each group contained 10 Formlabs Permanent Crown discs. The specified post-curing cycle was performed twice in Form Cure, with support removal between cycles. All groups underwent 5,000 thermal cycles. Strength was measured only after aging; there was no unaged strength baseline. Superscript letters reproduce the original post-hoc groups: groups sharing a letter were not significantly different within the corresponding outcome. Do not compare letters across strength and color tests. These schedules are experimental conditions, not instructions for another material. [3]
The strength results differed across groups overall (p < 0.001). The highest mean was recorded for 30 min at 60 °C, but this group was not significantly different from 30 min at 80 °C under the reported post-hoc grouping. A numerical maximum therefore should not be presented as a uniquely proven optimum. [3]
Most importantly, this is a comparison of post-aging outcomes between curing groups. Without an unaged strength control, it cannot quantify strength lost or gained because of thermal cycling. The study also did not measure degree of monomer conversion, so explanations involving conversion or cross-linking remain interpretations rather than directly measured mechanisms. [3]
For production review, the useful question is whether the available evidence covers the exact material, curing unit and full processing sequence. The table does not justify transferring a time–temperature schedule to another resin.
The other measured outcomes also varied by condition. Roughness decreased from 0.85 ± 0.23 to 0.64 ± 0.15 μm in the 20 min/60 °C group (p = 0.037), but increased from 0.58 ± 0.19 to 0.73 ± 0.20 μm in the 30 min/80 °C group (p = 0.032). The other seven roughness groups did not show significant pre-to-post changes. Relative translucency decreased significantly in only the 15 min/60 °C and 20 min/40 °C groups. These results do not support a blanket claim that every surface or optical property deteriorated. [3]
Why do results from different studies vary?
Several variables change together across these papers. Sahin and Tokar used complete crowns finished with rotary polishing kits, while Baytur and Diken Turksayar used discs prepared with silicon carbide papers. Resin, printing technology, geometry and curing equipment also differed. Their roughness results cannot isolate the effect of polishing alone. [2] [3]
Background example: combined thermal and mechanical aging
Likewise, water-bath thermal cycling and combined thermal–mechanical aging are different challenges. The Melc background study used bars and included mechanical loading; its results cannot be substituted for those of a cemented printed crown exposed only to thermal cycling. [2] [4]
Figure 3. Thermomechanical aging in background materials
Source: Melc et al., original Tables 2 and 3 (PDF pp. 6–7). [4]
Eight of the eleven material/curing groups showed a significant reduction in flexural strength; Lava Ultimate, Straumann n!ce and GC Initial LRF did not. For example, VITA Enamic changed from 183.94 ± 13.11 to 137.26 ± 35.40 MPa (p < 0.001). These bar-specimen results illustrate why the loading protocol belongs beside the result, while remaining separate from printed-crown evidence. [4]
For a lab comparison: record specimen geometry, support and loading conditions together. If several conditions differ, the studies cannot establish which single factor explains the difference in results.
What remains outside the evidence?
- Clinical lifespan. The cycle count describes the laboratory exposure. Authors’ time-equivalence assumptions are not validation of how many years a restoration will function. [1] [2] [3]
- Universal stability. A non-significant result under one test configuration does not prove that a material is unaffected by other exposures. [2]
- Every operating condition. Thermal cycling alone does not include cyclic mechanical loading, and the reviewed experiments do not cover every printer, curing unit, finishing sequence or supporting structure. [1] [2] [3]
- RayForm product performance. The data above belong to the third-party materials and experimental workflows named in the papers. They are not RayForm test results, product specifications or evidence of clinical suitability.
A supplier checklist a dental lab can use
Ask for the conditions behind a result before using it in a purchasing or production decision. Where comparable information is unavailable, document the gap rather than filling it with a result from a different material.
Table 6. Questions to ask before using aging data in a material comparison [1] [2] [3] [4]
| Check | Ask the supplier or testing laboratory | How to use the answer |
|---|---|---|
| Material and version | Which resin, shade, batch or formulation version was tested? | Match the report to the material being evaluated; record any mismatch. |
| Manufacturing workflow | Which printer, build orientation, layer thickness, washing, drying and post-curing procedure was used? Were curing cycles repeated? | Identify the workflow covered by the evidence, including the curing unit and settings. |
| Specimen and support | Was the test performed on discs, bars or complete crowns? What were the dimensions, support and cementation conditions? | Separate specimen strength from crown-assembly load; compare equivalent geometries where possible. |
| Finishing | How were surfaces finished or polished, and where was roughness measured? | Avoid attributing a surface result to the resin alone. |
| Aging and control | What were the temperatures, dwell times, cycle count and medium? Was mechanical fatigue included? How was the control stored? | Compare the actual exposure rather than a claimed number of clinical years. |
| Sample size and statistics | How many specimens were tested? Are means, variability, individual ranges and statistical comparisons available? | Read a non-significant finding as a result under the tested conditions, not proof of equivalence. |
| Available evidence and gaps | Which of these reports and workflow documents can be supplied, and which combinations or outcomes remain untested? | Record the gaps before deciding whether additional internal evaluation is needed. |
The checklist follows the test variables and limitations in the reviewed studies. Ask which information is available and record any gaps. [1] [2] [3] [4]
Discuss the documentation for your workflow
When contacting RayForm, include three details: the resin or material you are evaluating, your printer model, and your post-curing unit. Add the washing and finishing sequence if available, then ask which technical documents cover that combination. Visit Technical Support or send your workflow details to RayForm.
References
- [1] Peng PW, Chang WJ, Chou JS, Chen LX, Chuang CC, Lu YJ, Hsieh MC, Lee WF. Accuracy and fitness of digital light processing-fabricated resin-based permanent crowns after thermal aging. Journal of Dental Sciences. 2025. doi:10.1016/j.jds.2025.07.007. Full text.
- [2] Sahin NI, Tokar E. Mechanical Properties of Additively Manufactured Composite Resin vs. Subtractively Manufactured Hybrid Ceramic Implant-Supported Permanent Crowns Before and After Thermal Aging. Micromachines. 2026;17(1):116. doi:10.3390/mi17010116. Full text.
- [3] Baytur S, Diken Turksayar AA. Effects of post-polymerization conditions on color properties, surface roughness, and flexural strength of 3D-printed permanent resin material after thermal aging. Journal of Prosthodontics. 2025;34(3):298–307; first published online 2024. doi:10.1111/jopr.13818. Full text.
- [4] Melc T, Attin T, Eyüboğlu TF, Özcan M. Flexural Strength of Modern CAD/CAM Restoratives After Artificial Aging. Materials. 2024;17(21):5178. doi:10.3390/ma17215178. Used as methodological background only.
