Material Recession in a High-temperature, High-velocity Water Vapor Environment

Transcription

1 Material Recession in a High-temperature, High-velocity Water Vapor Environment A Technical Report In STS 4600 Presented to The Faculty of the School of Engineering and Applied Science University of Virginia In Partial Fulfillment of the Requirements for the Degree Bachelor of Science in Engineering Science Jack Valentine April 11 th, 2012 On my honor as a University student, I have neither given nor received unauthorized aid on this assignment as defined by the Honor Guidelines for Thesis-Related Assignments Signed: Date: Approved: Elizabeth Opila, Materials Science and Engineering Date:

2 2 Abstract A laboratory testing device was created to simulate the environment of a combustion turbine engine. Liquid water was pumped into a furnace, moved through a small diameter capillary, vaporized and expelled as a gas at accelerated velocities. A SiO 2 sample was tested in the device for 170 hours at 1300 C and produced visible surface degradation due to a volatilization reaction which allowed estimations for the recession and weight loss rates. The rates were comparable to theoretical predictions. These results demonstrate the device s capacity for future testing of potential engine materials in a low-cost laboratory setting. Introduction For combustion turbine engines, increased efficiency requires higher operating temperatures. 1 Having reached the upper limits of current high-temperature metals, new ceramic materials, which offer the possibility of higher engine operation temperature, are required for further improvement in engine efficiencies. These materials still need significant characterization before they can be implemented effectively into designs. 2 The key problem with using ceramics is they react with water vapor (a product of combustion reactions) and volatilize rapidly. 3 This paper outlines a method for creating a laboratory testing device that replicates a combustion environment. Silicon carbide (SiC) and similar ceramic composites are leading candidates for engine application due to their high temperature capability. However, SiC is prone to rapid deterioration through the following reactions with water vapor. 4 ( ) ( ) ( ) (1) ( ) ( ) (2)

3 3 As shown in Eqn. 1, the SiC reacts with water vapor to form an oxide layer (SiO 2 ). This oxide layer then reacts with water vapor as shown in Eqn. 2 and volatilizes into Si(OH) 4. The high rate of these reactions can lead to severe material loss and has been well documented in the papers of Opila 7 and Robinson and Smialek. 8 This paper focuses on a method for observing high-temperature materials reaction with water vapor in an environment typical of combustion turbines. Based on the design of dos Santos et al. 2, Ferber and Lin 5 and Sudhir and Raj 6 the objective is to create observable rapid recession and compare the results with theoretical predictions. To observe these degradation mechanisms, there are several design objectives. The oxidation mechanisms described in Eqn. 1 and 2 have a strong dependence on a variety of factors including: temperature, water vapor partial pressure, and gas flow velocity. 7 It is crucial to replicate these factors of a combustion environment in the testing device to create meaningful results. This device will additionally act as a faster alternative to very expensive burner rig tests and allow for more precise control of independent variables. 2 It will also allow for observation of the reactions in a controlled environment. Experimental Procedure The testing device design shown below in Fig. 1 uses a horizontal tube furnace (CM Furnaces Inc., Bloomfield, NJ, Model #100803) with a 1m long alumina tube. Figure 1: Depiction of Testing Device

4 4 The stainless steel end cap has two inlet openings; the top one holds a fused quartz (SiO 2 ) capillary through which liquid water is injected and the bottom one holds a sheathed thermocouple (type R) which measures the temperature at the sample interaction site. The fitting that holds the capillary also has a T-fitting that can be used to introduce additional permanent gases (e.g. Ar, O 2 ) in future testing should it be necessary. The sheathed thermocouple (0.25 OD) runs through a hollow alumina tube (0.875 OD) that acts as a support for the fused quartz capillary. Without this support the fused quartz capillary bends and breaks under its own weight at high temperatures. The capillary used in this design has an outer diameter of 3mm and inner diameter of 1mm (Quartz Scientific, Inc., Fairport Harbor, OH). The capillary is cut to approximately 70cm so that when placed in the furnace, the exit opening is at the end of the hot zone of the furnace. Liquid water is pumped into the capillary using a peristaltic pump (IDEX Health & Science GmbH, Wertheim-Mondfeld, Germany, Ismatec Tubing Pump Model REGLO Analog) at a rate of 1.8mL/min. The water then travels through the capillary, enters the furnace, boils and accelerates toward the sample as a gas. The gas velocity (v g ) of the water vapor was determined using an adaptation of the ideal gas law in the same way as dos Santos et al. 2 shown below (3) Where ṁ is the mass flow rate, R is the universal gas constant, T is the temperature of the jet, A is the cross-sectional area of the capillary, MW is the molecular weight of water and P is the exit pressure at the end of the capillary. Using Eqn. 3 and assuming: ambient pressure = 1 atm, MW = g/mol; with test conditions of T = 1289 C, A = 3.141E-6 m 2, ṁ = 1.8 g/min; a value of v g = m/s is obtained.

5 5 The sample used in the test was a 1 x 1 SiO 2 slide (Quartz Scientific, Inc., Fairport Harbor, OH) that was chosen to simulate the volatilization of the oxide layer shown in Eqn. 2. The sample was placed in a slotted alumina boat as shown in Fig. 2 and positioned in the furnace so that the gas would impinge directly onto the sample. The positioning of the sample was determined after calibration of the tube furnace temperature profile shown in Fig. 3. Since the furnace is 22 long, it is expected that the hot zone would be in the middle at Figure 2: View down tube furnace (Digital Camera) the 11 mark. To give the water additional time to boil before exiting the capillary, the end of the capillary was positioned at the 13 mark and the slide was positioned ½ from the end of the capillary. With this current Figure 3: Furnace Calibration configuration, the water exits the capillary with an audible sputtering effect in which most of the water transforms into steam but there are sporadic bursts of liquid water through the furnace. Future tests will try to correct this issue by raising the liquid water temperature before entering the furnace.

6 6 Experimental Results The sample shown in these results was exposed for 170 hours at 1289 C with a constant water flow rate of 1.8 ml/min. Unless noted otherwise, these images were taken with a digital microscope (Hirox-USA, Inc., Hackensack, NJ, Model # KH 7700). The blue arrows indicate the direction of the water jet and the red-dashed lines draw attention to material loss by outlining the original sample boundaries. Fig. 3 on the right shows the orientation of the sample in the furnace from the perspective of looking down the tube. The quartz capillary rests on top of the alumina support cylinder and expels water vapor directly at the sample slide. As Fig. 4 shows, there is a distinct difference between the slide before and after the furnace test. The sample on the right has likely crystallized into cristobalite. The interaction site between the water vapor jet and sample is also evident in the top right corner of the sample. By weighing the sample before and after the furnace run, a weight loss equal to g was calculated. Figure 4: Comparison of SiO 2 slide before and after exposure (Digital Camera)

7 7 Fig. 5 gives a more detailed perspective of the interaction site of the water vapor and the sample. The sample recession is evident and the observable area of material loss on the front side of the slide (outlined in green) is approximately 0.5 cm 2. This sample also showed visible recession when viewed from a top perspective. As Fig. 6 Figure 5: Front view of SiO 2 Slide after furnace exposure depicts, the red dashed line is the approximate dimensions of the sample prior to the run and the change in the width of the slide from the leading edge backwards is evident. Figure 6: Top view of SiO 2 slide after furnace exposure The point of greatest recession is at the leading edge and shows a width change of 126 µm. Since this test was run for 170 hours, this produces a maximum recession rate of approximately 0.74 µm/hr. Discussion of Results To determine the accuracy of these results and the effectiveness of the device, it is necessary to compare the results to theoretical models and predictions of material loss. Using the equation developed by Robinson and Smialek 8 for SiC, which is shown below in Eqn. 4, an expected weight loss rate (k l ) is calculated for SiO 2 (neglecting the density difference between SiO 2 and

8 8 SiC) based on the velocity (v), temperature (T), and partial pressure (P) of the water vapor and other environmental conditions. ( ) ( ( ) ) (4) By using values of T=1562 K, v = m/s and assuming P=1 atm, this gives us a value of k 1 = mg/(cm 2 hr). For comparison, an experimental value determined from the observed weight loss and affected area on the sample is calculated with Eqn. 5. ( ) ( ) (5) Where m is mass loss, A is area of mass loss and t is time of exposure, Using input for m = g, A = 1 cm 2 (for the two sides of the slide) and t = 170 hours; it is found that k 1 = mg/(cm 2 hr). The calculated and measured linear weight loss rates, k, differ by a factor of 3.38 but the order of magnitude is correct. The discrepancy may be attributed to the rough estimation of the affected surface area and the density difference between SiO 2 and SiC which was neglected. This comparison gives the semi-quantitative confirmation that the testing device provides a realistic simulation of the water vapor flows common to combustion environments. Summary This furnace water jet provides an inexpensive testing device for characterizing materials interactions with high-temperature water vapor under controlled and independently variable temperatures and gas velocities. The results for SiO 2 show rapid and measurable material recession that agree with theoretical predictions. This device will enable a more comprehensive understanding of the basic science behind high-velocity, high-temperature water vapor

9 9 interaction with materials and the development of durable structural materials for combustion applications. 1 Perepezko, J. H. (2009). The hotter the engine, the better. Science (New York, N.Y.), 326(5956), dos Santos e Lucato, Sergio L., Sudre, O. H., & Marshall, D. B. (2011). A method for assessing reactions of water vapor with materials in high-speed, high-temperature flow. Journal of the American Ceramic Society, 94, s186-s195. doi: /j x 3 Lee, K. N. (2000). Current status of environmental barrier coatings for Si-based ceramics. Surface and Coatings Technology, (0), 1-7. doi: /s (00) Opila, E. (2011, August 17). Oxidation Mechanisms: Applications for Hypersonics Materials & Structures. Presented at Summer School on Materials & Structures for Hypersonic Flight, University of California, Santa Barbara. 5 Ferber, M., Lin, H. (2005). Environmental characterization of monolithic ceramics for gas turbine applications. Key Engineering Materials, 287, Sudhir, B., Raj, R. (2006). Effect of steam velocity on the hydrothermal Oxidation/Volatilization of silicon nitride. Journal of the American Ceramic Society, 89(4), Opila, E. J. (2003). Oxidation and volatilization of silica formers in water vapor. Journal of the American Ceramic Society, 86(8), doi: /j tb03459.x 8 Robinson, R. C., & Smialek, J. L. (1999). SiC recession caused by SiO2 scale volatility under combustion conditions: I, experimental results and empirical model. Journal of the American Ceramic Society,82(7), doi: /j tb02004.x