Testing 1-2-3: Material of the Future: 3-D Printed Molds Improve Titanium Casting’s Potential

Researchers developed molding methods and materials to provide exceptional design flexibility while simplifying production.

Sairam Ravi and Jerry Theil, University of Northern Iowa, Cedar alls, Iowa

(Click here to see the story as it appears in the April issue of Modern Casting.)

Titanium alloys continue to remain some of the most versatile alloys available with high strength to weight ratios and excellent resistance to corrosion. Titanium’s use as an engineering material in cast components has been limited due to its high cost, some of which results from the casting process. Because of strong reactivity with oxygen, current methods for producing titanium castings include rammed graphite molding and investment casting. Both involve multiple steps and require extensive equipment.

Recently, three dimensional (3-D) printing equipment has become a readily available technology possessing the accuracy required to produce high quality molds and cores for castings. The University of Northern Iowa, Cedar Falls, Iowa, with funding provided by the U.S. Department of Defense, has developed molding methods and materials to enable casting titanium in 3-D printed (binder-jet printed) molds.

This system provides exceptional design flexibility while reducing the time, steps and complexity of producing titanium castings. It was designed to address the high reactivity of liquid titanium with molding materials and minimize alpha case depth. Although the scope of this research was limited to 3-D printed molds, the technology developed also could be used with conventional molds.

1. Background

Casting remains the single most versatile metal forming technology. No other process has the ability to form intricate internal geometric features at a fraction of the cost of competing technologies. It is commonplace for conventional ferrous and nonferrous materials to be cast in intricate shapes in sizes from ounces to hundreds of tons. But because of titanium’s reactivity with oxygen in the atmosphere or vessels that contain the liquid metal, its use as a casting material has been limited. Titanium is as strong as steel but more than 40 percent lighter. It is only 60 percent heavier but over four times stronger than aluminum and with high temperature characteristics not possible with other alloys.

Currently, titanium components are predominantly used in ultra-high performance applications including aerospace. Increasingly stringent regulations on fuel efficiency and lower greenhouse gas emissions has focused interest on weight reduction and improving performance. Titanium has great potential to be the designer’s material of the future, because of its unique combination of metallurgical properties such as high strength-to-weight ratio in the temperature range from sub-zero to 1,004F (540C).

Another remarkable property of titanium and its alloys is their excellent corrosion resistance to a range of acids, alkalis and chemicals. This attribute makes them ideal for applications in power plants and the chemical industry. Titanium is the only structural metal having corrosion fatigue behavior in saltwater that is practically identical to that in air, making it ideal for all sea water-based applications, such as pumps and valves. This technology will be critical in the next generation of nuclear power generation facilities.

Titanium melts between 3,000 to 3,050F (1,649 to 1,677C). Casting the alloy always has been problematic because of the reactivity of the liquid metal with materials that contain it. Any refractory used for titanium casting must have minimal reactivity. The more negative the free energy of formation, the greater the stability of the oxide and the less tendency of the oxide to give up its oxygen to the titanium. Oxygen obtained from molding materials combines with molten titanium to form a hard brittle surface that is termed alpha case. This alpha case must be removed by machining or chemical etching before the casting can be used. This additional process step can increase the cost of a cast titanium component significantly.

The most common refractory flours used in casting aluminum and steel, zircon (ZrSiO4) and fused silica (SiO2), are not used as the face coat refractory when casting titanium as these contain some amount of silica, which is readily reduced by the molten titanium. But, alumina exhibits a highly negative free energy of formation while providing a high melting point and good availability at a moderate cost.

2. Procedure

The experimental methodology in this research work was divided into three components.

First, the study investigated the formulation and characterization of a suitable nonreactive coating for titanium casting applications. The coating was evaluated with steel Gertsman castings, which are designed to test for metal penetration into the mold.

Trial cylindrical mold metal reactivity molds were produced using a ceramic bonded with the furan and sodium silicate binder systems to evaluate the two binder systems. Molds were also made using alumina and zirconia aggregates to compare the specialty aggregates.Preliminary titanium casting trials were held at Iowa State University, Ames, Iowa, where the molds were poured. The alpha case layer formed was measured.

The final component of the testing involved titanium casting trials at the Rock Island Arsenal, Rock Island, Ill. Silica sand molds and the newly developed refractory coating were used. The alpha case layer of titanium was correlated to the preliminary trials.

To evaluate the aggregates, binder systems and coating, a lab-scale glove-box-based melting system was designed and fabricated at Iowa State University. The system was used for autogenous weld passes for systematic investigation of cooling rate effects on solidification structure; cold-crucible melting with direct-chill solidification; and inert/vacuum arc-melting and tilt-pour casting to study solidification in shape castings and the performance of various mold materials’ chemical compatibility, cooling behavior and interface heat transfer.

The overall system is shown in Fig. 1.  A water-cooled copper hearth was integrated into a tilt-pour system to operate within the glove box containment system. The hearth and pouring assembly were then adapted to accommodate the molds. This is shown in Fig. 2. Vickers microhardness was measured for all castings starting from the alpha-case layer down to the base metal, where the hardness flattens out. The depth was recorded at the same time.

The titanium alloy used at Rock Island Arsenal was titanium with 6% aluminum and 4% vanadium (TiAl6V4). The solidification behavior of this alloy was provided by the University of Iowa. Based on the solidification results, a wedge casting was designed which enabled the evaluation of alpha-case layer defects with respect to different section thicknesses of the casting and different cooling rates. This is shown in Fig. 3.

An equipment vendor produced the test mold using a 3-D printer. Silica sand with a 105 AFS-GFN was the aggregate and a furan binder system was used. The mold was coated with the EAC. A Baumé of 80 was used. The test mold is shown in Fig. 4.

The molds were poured at Rock Island Arsenal using a vacuum induction melting furnace and vacuum pouring chamber. The resulting casting was sectioned according to predetermined thicknesses. The cross sections were then evaluated for its microhardness using a Buehler microhardness tester. A magnification of 500X was used with a load of 500 g (1.1 lb.), and the microhardness was measured across all cross sections in steps of 50-150 micrometers. Charts were plotted using the depth in micrometers for the X axis and the microhardness for the Y axis.

3. Results and Conclusions 

Zahn cup, viscosity and percent solid results from the coating characterization were plotted against the respective Baumé to determine a suitable Baumé for coating cores and molds. A target range of 10-15 seconds for zahn cup measurement and 75% solid fraction were set as optimal for coating applications.

A Baumé of 78 or 80 resulted in desired viscosity and percent solids results for the experimental alumina coating (EAC). It was decided to coat the test molds and cores at a Baumé of 80. Table 1 shows the detailed results obtained from the tests.

Ceramic and silica sand Gertzmann cores bonded with furan binder system were placed in the molds. Low alloy steel with a target chemistry of 0.3 % carbon, 0.6% silicon and 0.5% manganese was poured in these molds with a target pouring temperature of 2,912F (1,600C). Later, the castings were cleaned and photographed. Figures 5 and 6 show the castings from cores made of ceramic and silica sand. Both cores were coated with the experimental alumina coating (EAC). It can be seen that both the castings with the EAC have no visible defects except for a small area of sand burn on the bottom of the casting made of the ceramic sand.

From the steel casting results, it can be seen that the EAC is effective in forming a protective layer around the core. As a result, it was decided to use the EAC for the preliminary casting trials of titanium at Iowa State University and evaluate the performance of the coating in titanium castings.

The coupled expansion and viscosity results for uncoated silica sand bonded with the furan binder follow an expansion pattern typical of silica sand. An Alpha-Beta transition is seen at approximately 1,063F (573C). A peak at higher temperatures, which corresponds to the cristobalite transition in silica sand, also denotes the sintering temperature of the sample, when sand grains partially melt and start fusing to each other. The temporary increase in viscosity at this stage depicts an increase in the strength of the core or mold due to partial fusion of the sand grains. The viscosity measurement can be used as an indication to the extent of the same.

From the expansion and viscosity results for the coated silica sand sample, it was determined that the peak expansion at the alpha-beta phase transition was reduced. A peak expansion of 0.007 in./in. (0.018 cm/cm) was measured. Rather than a contraction after 1,112F (600C), a steady expansion was observed until the cristobalite phase transition. This behavior prevented the coated core or mold from cracking from high strain caused by contraction of the sample. An increase in viscosity vs. the uncoated sample lead to higher strength, enabling the core to withstand higher temperatures.

As mentioned earlier, four molds were cast with titanium at the Iowa State University. All molds were coated with the EAC. The resulting mold metal reactivity  castings were evaluated for alpha-case depth. Vickers hardness was measured for all castings and the extent and depth of the alpha layer was measured. A chart was plotted with the depth in microns on X axis and Vickers hardness on Y axis. In all cases, it was seen that the hardness was initially high at a lower depth and decreased with depth till the base layer of titanium was reached. At this point, the hardness stabilizes and is representative of the unaffected base metal.

The base layer of titanium is reached at approximately 350 VHN. With the exception of ceramic with sodium silicate binder system, the other three samples have similar results. Zirconia with sodium silicate, alumina with sodium silicate and ceramic with furan recorded an alpha case layer depth of 115 microns, 129 microns and 160 microns, respectively. Ceramic with sodium silicate recorded an alpha case layer depth of 237 microns. The results are in Table 2.

The performance of the ceramic with furan resin compares well with those of specialty aggregates such as Alumina or Zirconia. The alpha case layer depths recorded for all samples in the preliminary casting trials is comparable to investment casting of titanium. The EAC is believed to have played a major role in protecting the base layer of the core from being reduced by titanium. This can be explained from the similarity seen in the results for the four samples.

Titanium castings were poured using 3D printer molds at Rock Island Arsenal. The mold was coated with EAC and dried before pouring. The resulting castings were sectioned and evaluated for microhardness (Vickers).

The wedge casting had different section thicknesses to evaluate the effect of cooling rate on the formation of alpha case layer defects. The results were similar across the sections. This shows the different cooling rates in the casting did not play an active part in the formation of the alpha-case layer defect of titanium. The base layer was measured to be at approximately 350 VHN (see Table 3).

The results from the experiments at Rock Island Arsenal were compared with previous results from Iowa State University to check for variances. The results were similar between the pour at the two facilities. The silica sand casting has a maximum hardness of 544 VHN compared to a hardness of 587.8 VHN for the ceramic with sodium silicate.

For the same section thickness, a hardness of 510.89 was measured with ceramic sand and furan resin compared to 544 for the silica sand casting. Hence, similar alpha case layer depth was measured with the 3D printed silica sand mold. The results are comparable to the specialty aggregates tested in the preliminary trials.

In conclusion, the research indicated that by using a specially designed refractory coating, titanium castings could be produced to an acceptable quality level in silica sand molds. The refractory coating effectively prevented the molten titanium from reacting with the base molding material. This was evidenced by the depth of the oxygen rich alpha-case layer exhibited in the experimental castings. Large-scale experimental castings agreed well with laboratory scale castings.

From all the testing done and looking at the final casting results, it can be concluded that using rapid prototyping technology to produce silica sand molds and using an effective refractory coating such as a water-based alumina coating, the alpha-case layer defect in titanium castings can be reduced to a large extent in sand casting.

In addition, 3-D printed molds provided the quality and integrity required to pour titanium castings, as long as the surface was protected with a nonreactive refractory coating.

This technology has the potential to reduce the cost of titanium castings and increase their use in land-based applications.  

This article is based on a paper (14-029) that was presented at the 2014 AFS Metalcasting Congress. 

Thanks to streamlined simulation, tooling, casting and machining capabilities, an intricate water passage went from purchase order to prototype in just 17 days.

Most efficient cleaning with dry ice
UNSUPPORTED - please refresh your browser.