The Method for Fabricating Proppant and Cenosphere Sand-Based Casting Molds Involving the Use of Binder Jetting 3D Printing with Furan Binder and Impregnation with Colloidal Silica Binder

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The Authors present an impressive work that deals with the possibility of using sand molds (3D-printed molds made from proppant or cenosphere refractory particles and furan resin binder and then impregnated with a silicate binder) for the manufacturing of nickel-based superalloy castings. The method itself was shown to be able to replace the investment casting method for applications, which are not sensitive to surface quality. The manuscript is well-written: the introduction provides all the information needed for a better understanding of the topic, the methods and materials are well-described. The discussion of the results is detailed and all the conclusions are supported by the results. Further research may aim at improving surface quality of castings and reducing linear shrinkage of the molds. The manuscript can be accepted without any changes. 

Author Response

The Authors present an impressive work that deals with the possibility of using sand molds (3D-printed molds made from proppant or cenosphere refractory particles and furan resin binder and then impregnated with a silicate binder) for the manufacturing of nickel-based superalloy castings. The method itself was shown to be able to replace the investment casting method for applications, which are not sensitive to surface quality. The manuscript is well-written: the introduction provides all the information needed for a better understanding of the topic, the methods and materials are well-described. The discussion of the results is detailed and all the conclusions are supported by the results. Further research may aim at improving surface quality of castings and reducing linear shrinkage of the molds. The manuscript can be accepted without any changes. 

Answer: The comments are not provided.

Reviewer 2 Report

Comments and Suggestions for Authors

This study investigates the use of two refractory materials with distinct thermal properties, fabricated via binder jetting process with a furan resin binder system. The printed molds were subsequently impregnated with colloidal silica, sintered, and used to cast nickel-based superalloy components. The approach combines hollow cenosphere sand (with high insulation properties) for the gating system and proppant sand (with high thermal conductivity) for the mold body, aiming to balance rapid cooling of the casting (to minimize shrinkage defects) and effective feeding via the insulated riser. While the methodology demonstrates innovation, several critical issues require clarification and improvement:

1. The solid content and colloidal particle size of the silica binder are not provided. These parameters significantly influence impregnation efficiency, binder distribution, and final mold strength.
2. The proppant sand (median size: 254 μm) and cenosphere sand (70 μm) were printed with layer thicknesses of 0.4 mm and 0.3 mm, respectively. However, empirical guidelines suggest layer thickness should be 2–5 times the particle size. The rationale for deviating from this standard—particularly for the cenosphere sand (layer thickness ≈4× particle size)—needs justification.
3. While the binder jetting printer (SP500) is mentioned, critical details such as the printhead type (e.g., piezoelectric or thermal) and nozzle diameter are omitted. These parameters directly affect binder droplet size, resolution, and bonding quality.
4. The cenosphere sand required 4 wt.% curing agent, far exceeding the 0.5 wt.% for proppant sand and typical binder jetting ranges (0.3–0.6 wt.%). This raises concerns:
   • Is this due to improper layer thickness or particle size selection?
   • Does such a high curing agent content compromise the flowability of the sand mixture during printing?
5. The 1-minute immersion time for colloidal silica impregnation appears insufficient. High-viscosity colloidal silica (common at high solid content) may fail to penetrate the mold’s internal pores under atmospheric conditions, potentially weakening the structure. Prior studies suggest vacuum-assisted impregnation or longer durations are necessary for uniform infiltration.
6. The alloy composition (e.g., Inconel 718, CMSX-4) must be explicitly stated, as its melting behavior, reactivity, and thermal expansion directly affect mold performance and casting quality.
7. The "pink regions"in attributed to silica binder leakage (lines 386–387) require further analysis: Is this due to gravitational drainage or insufficient infiltration caused by capillary resistance? Micro-CT data should quantify binder distribution gradients.
8. The study only reports room-temperature bending strength (3–4 MPa). For high-temperature alloy casting (e.g., 1200–1500°C), the mold must retain sufficient strength at elevated temperatures to prevent collapse during pouring. High-temperature mechanical testing is essential to validate feasibility.

Author Response

This study investigates the use of two refractory materials with distinct thermal properties, fabricated via binder jetting process with a furan resin binder system. The printed molds were subsequently impregnated with colloidal silica, sintered, and used to cast nickel-based superalloy components. The approach combines hollow cenosphere sand (with high insulation properties) for the gating system and proppant sand (with high thermal conductivity) for the mold body, aiming to balance rapid cooling of the casting (to minimize shrinkage defects) and effective feeding via the insulated riser. While the methodology demonstrates innovation, several critical issues require clarification and improvement:

1. The solid content and colloidal particle size of the silica binder are not provided. These parameters significantly influence impregnation efficiency, binder distribution, and final mold strength.

Answer: This information was provided: “The UltraCast One+ (Technopark, Moscow, Russia) colloidal silica binder, which contains 25.5–27.5 wt.% SiO2 solid content and 8–10 nm particle size, was utilized for the impregnation process.”

2. The proppant sand (median size: 254 μm) and cenosphere sand (70 μm) were printed with layer thicknesses of 0.4 mm and 0.3 mm, respectively. However, empirical guidelines suggest layer thickness should be 2–5 times the particle size. The rationale for deviating from this standard—particularly for the cenosphere sand (layer thickness ≈4× particle size)—needs justification.

Answer: An experimental investigation was conducted to assess the impact of varying printing regimes on the thickness of printing layers and the consumption of furan binders. This information is not included in this paper, as it is part of a separate, unpublished study focused on the use of multi-material sand molds, comprising silica, proppant, chromite, and cenosphere sands. However, the findings of this study demonstrate that the printing regimes employed result in sufficient strength of the mold parts for further impregnation.

3. While the binder jetting printer (SP500) is mentioned, critical details such as the printhead type (e.g., piezoelectric or thermal) and nozzle diameter are omitted. These parameters directly affect binder droplet size, resolution, and bonding quality.

Answer: The printing parameters are added and this part was rewritten:

“In this study, the ID50-K proppant sand (Carbo Ceramics Inc., Houston, USA) and cenosphere sand (Tekhnokeramika, Obninsk, Russia) were utilized. The printing process was performed using a SP500 binder jetting printer (Additive Technologies, Saint Petersburg, Russia). The binder and curing agent were composed of the DF-400 furan resin (Suzhou Xingye Materials Technology Co. Ltd., Xushuguan, China) and the DFG-30A acid curing agent (Suzhou Xingye Materials Technology Co. Ltd., Xushuguan, China), respectively. The commercial MH2820 piezoelectric printhead (Ricoh Company, Ltd., Ōta, Tokyo, Japan) was utilized, with a printing resolution of 1370×150 dpi and 1370×300 dpi for the proppant and cenosphere specimens, respectively. The volume of each drop was approximately 40 pL. Prior to the initiation of the printing process, the refractory sand was blended with the curing agent at a proportion of 0.5 wt.% for the proppant sand and 4 wt.% for the cenosphere sand. The layer thickness of the proppant and cenosphere specimens printing was set as 0.4 and 0.3 mm, respectively. Following the completion of the printing process, the specimens were maintained in the printer camera for a period of 24 hours, after which the unbound sand was removed using a brush.”

4. The cenosphere sand required 4 wt.% curing agent, far exceeding the 0.5 wt.% for proppant sand and typical binder jetting ranges (0.3–0.6 wt.%). This raises concerns:

• • Is this due to improper layer thickness or particle size selection?
  • Does such a high curing agent content compromise the flowability of the sand mixture during printing?

Answer: The rationale underlying this phenomenon pertains to the low density of the cenosphere sand (0.4 g/cm³). A non-significant disparity in volume percentage is observed when comparing the typical binder jetting ranges. Consequently, the flowability of the curing agent-cenosphere sand mixture is analogous to that of the conventional curing agent-silica sand mixture.

5. The 1-minute immersion time for colloidal silica impregnation appears insufficient. High-viscosity colloidal silica (common at high solid content) may fail to penetrate the mold’s internal pores under atmospheric conditions, potentially weakening the structure. Prior studies suggest vacuum-assisted impregnation or longer durations are necessary for uniform infiltration.

Answer: It is our mistake. The 1 min impregnation time was only for the 40x20x8 mm samples. For the mold parts the impregnation time was increased for 5 min. This information was added to the article text. During impregnation, air bubbles floating to the surface of the binder. The 1 min for samples and 5 min for the mold parts was sufficient for the bubbles to stop floating

6. The alloy composition (e.g., Inconel 718, CMSX-4) must be explicitly stated, as its melting behavior, reactivity, and thermal expansion directly affect mold performance and casting quality.

Answer: The alloy composition was provided but without the grade because it is available only in Russian state standard, that not available in English.

7. The "pink regions"in attributed to silica binder leakage (lines 386–387) require further analysis: Is this due to gravitational drainage or insufficient infiltration caused by capillary resistance? Micro-CT data should quantify binder distribution gradients.

Answer: The fundamental basis of this phenomenon can be ascribed to gravitational drainage. During the fabrication of the mold components, it was observed that the upper surfaces of the mold demonstrated diminished binder content. The capillary resistance is not the primary factor, as in this scenario, the defects must be equivalent in the upper and lower portions of the mold.

8. The study only reports room-temperature bending strength (3–4 MPa). For high-temperature alloy casting (e.g., 1200–1500°C), the mold must retain sufficient strength at elevated temperatures to prevent collapse during pouring. High-temperature mechanical testing is essential to validate feasibility.

Answer: We are in partial agreement with this comment, and in subsequent research, we endeavor to measure the high-temperature bending strength. In this study, the molds were fabricated, and the casting was obtained. However, no de-hardening of the mold was observed after casting the alloy. It was necessary to exert effort to break out the mold.