The following article appears in the journal JOM, 48 (9) (1996), pp. 29-31. JOM is a publication of The Minerals, Metals & Materials Society

An Ultrasonic Sensor for High-Temperature Materials Processing

This article details the development and preliminary testing of an ultrasonic displacement sensor that can be used to monitor materials processing in-situ at temperatures and pressures significantly higher than current ultrasonic displacement sensors. The target processing environment for this sensor is that typically found in a hot isostatic pressure (HIP) vessel (i.e., temperatures exceeding 1,000° and pressures above 150 MPa). A reliable sensor that could provide information on the actual consolidation of the material would save HIP time and cost. Further, HIP runs could be terminated as soon as consolidation is complete, reducing undesired effects due to holding the materials at elevated temperatures for extended periods of time. Recent progress has been made in adapting eddy-current sensor technology to HIP vessel consolidation.¹ However, this technology is limited in the range of displacement that can be measured (typically less than 20 mm) and the high cost of the probes.

Standard room-temperature ultrasonic technology can measure displacements on the order of 0.02 mm over ranges exceeding 100 mm. Successful implementation of the ultrasonic sensor will allow high-resolution, large-range (100 mm) displacement monitoring of the consolidation process in a HIP vessel, ultimately producing substantial savings in HIP costs and high reproducibility between components. Considering the extreme environment present in HIP vessels, successful testing suggests that application of the ultrasonic sensor to other types of high-temperature materials processing should be feasible. At present, commercially available ultrasonic transducers are limited to approximately 350° for sustained in-situ use because of transducer materials.

Extending ultrasonic displacement measurement to an elevated temperature environment creates many problems difficult to overcome. The piezoelectric materials used in typical ultrasonic transducers become inefficient and lose their piezoelectric properties if temperatures exceed the Curie point of the material. Typical efforts to adapt standard ultrasonic transducers to high temperatures use a buffer rod between the transducer and the hot material; for HIP-vessel use, a buffer rod is unacceptable due to the necessity of breaching the vessel wall.

Several methods of producing ultrasonic energy other than using piezoelectric materials have been developed. Electromagnetic acoustic transducers (EMATs) have been used successfully to detect defects and measure mechanical properties of metals at high temperatures.² One drawback to using EMATs is the requirement that the target material be electrically conductive. Also, EMATs are generally very inefficient producers of ultrasonic energy, so the resulting signal-to-noise ratios are often small. Recently, laser technology has been used to create and receive ultrasonic energy in a variety of materials.³ This technique shows considerable promise for producing ultrasonic energy in materials at elevated temperatures, but currently is very expensive and requires the use of moderate-power lasers. Both ultrasound-producing techniques described, while useful for defect detection and characterization of mechanical properties, are not easily adapted for measuring the displacement of a workpiece in a HIP environment.

Recognizing the potential applications in high-temperature materials processing for a sensor based on AlN, the University of Dayton Research Institute (UDRI) and the U.S. Air Force began investigations into the requirements for producing a sensor suitable for industrial use (Figure 1). The main tasks of the sensor development were identification of a substrate suitable for deposition of the AlN film; development of a rugged, high-temperature sensor housing; development of electrical-connection designs that would work at high temperatures; and research into methods of coupling the ultrasonic energy to the target object.

a	b
Figure 1.The high-temperature sensor: (a) schematic and (b) actual components.

A nickel-based superalloy was selected as the housing material because of its high-temperature stability and ease of machining characteristics. Careful design resulted in relatively straightforward machining of the housing at a reasonable cost. The design also allowed off-the-shelf components to be used for electrical connections and insulation purposes. The rhenium spring and Inconel^® push rod keep the substrate pushed against the front of the housing. Since the thermal expansion of the housing is matched by the expansion of the push rod, the spring force is constant regardless of temperature.

Achieving electrical connections to the AlN film that were stable at high temperatures and matched with the ultrasonic pulser-receiver instrumentation required considerable research and development. To produce ultrasonic energy, a potential difference of several hundred volts was required across the AlN film. Requiring the substrate material to be conductive provided an electrode on one side of the film. The substrate was insulated from the housing (which was designated as the electrical ground) by a ceramic collar that fit between the substrate and housing. Electrical connection to the other side of the AlN film was achieved by plating the film with platinum and mechanically forcing the platinum-covered surface against the inside of the sensor housing. Attaching the electrodes (sensor housing and substrate) to a cable connecting the sensor to the pulser receiver was achieved using a metallic-sheathed thermocouple wire fitted through a hole in the sensor. The sheath served at the ground wire from the pulser-receiver to the housing. The inner leads of the thermocouple wire were welded to the push rod fitted into a hole in the substrate. The thermocouple wire contained an insulating material between the inside signal wires and the metallic sheath.

Two approaches were taken to produce efficient ultrasonic coupling between the sensor and the workpiece in the HIP vessel. The first approach used the gas in the HIP vessel as the coupling medium, while the second used metals or glasses that melted at 700-800° as liquid couplants between the sensor and test object. The gas-propagation approach proved successful in preliminary tests and was used for all of the prototype sensor testing. Currently, tests are being conducted to assess the viability of using metal films and glass frits as couplants for high-temperature testing with the sensor in contact with the workpiece.

Considerable research was done on the propagation of ultrasound in high-pressure gases in the 1940s and 1950s.⁶ Although none of the research extended into the temperature and pressure ranges anticipated in the HIP vessel, enough data existed to imply that the HIP vessel gas could be used as an ultrasonic couplant. Preliminary tests conducted at low temperatures (less than 100°) over a pressure range from atmosphere to 200 MPa demonstrated that ultrasound between 15-25 MHz could be propagated through the gas at pressures above 20 MPa. Propagation distances of 100 mm were easily achieved.

Figure 2.The amplitude of the reflection from the back of the sensor sub strate as the HIP vessel temperature was increased to 1,000°C. The pressure in the HIP vessel ranged from 100 MPa to 150 MPa during the temperature increase.

Two different ultrasonic signals were monitored and acquired during the testing. One signal represented the ultrasonic reflection from the back of the substrate—the AlN film emits ultrasonic energy from both sides. The other signal represented the reflection from a nickel-based superalloy target 33 mm in front of the sensor. The ultrasonic energy reached the target by propagating through the argon gas. Two HIP-vessel tests were conducted using the high-temperature sensor. During the first test, the signal from the back of the substrate was monitored while the pressure and temperature were increased. In the second test, both the substrate reflection and a reflection from a target 33 mm away from the sensor were monitored.

During the first test, the pressure in the HIP vessel was ramped to 100 MPa while the temperature was held at 65°. After the pressure reached 100 MPa, the temperature was increased to 1,000° at a rate of 10° per minute. During the temperature ramp, the pressure increased to more than 155 MPa. Figure 2 shows the amplitude of the ultrasonic reflections from the back of the substrate as a function of temperature. The amplitude varied somewhat as the temperature increased but stayed at approximately 60% full-scale (150 digitizer counts) up to 850°. The last data were recorded at 880°, and the signal was detectable up to 940°. The data recorded at room temperature were obtained with the sensor outside of the HIP vessel and connected directly to the pulser-receiver. This was true for both data points acquired at room temperature—before and after the HIP run. All other data were acquired with the sensor inside the HIP vessel. The lower signal amplitudes for all data taken above room temperature were due to connecting the sensor to the pulser-receiver using thermocouple wires passing through the HIP vessel wall.

The second HIP run was conducted with the sensor positioned so that the ultrasound emitted from the front face of the AlN propagated through 33 mm of argon gas to a nickel-based superalloy target. The test started by ramping the pressure to 100 MPa, at which point a reflection from the target was recorded. Figure 3a shows the ultrasonic signal reflected from the target at a pressure of 150 MPa at 150°. The pressure was increased to 155 MPa before beginning the temperature ramp. The temperature in the HIP vessel was increased to 1,100° at a rate of 20° per minute while holding the pressure constant at 155 MPa. Ultrasonic data were recorded at approximately 100° intervals. Figure 3b shows the ultrasonic signal propagating through the argon gas to the target and back at 576°. At this temperature, the signal was becoming weaker, although it was still useful for detecting the presence of the target. The signal propagated through argon was detectable up to 676°, after which it was overwhelmed by electrical noise. The sensor continued to emit ultrasound up to a temperature of 940°, at which point electrical noise overwhelmed the substrate-reflection signal, making identification of the ultrasonic reflections impossible. The HIP run was continued to 1,100° and then cooled to room temperature. After removing the sensor from the vessel, it was determined that the electrical noise arose due to deterioration of the electrical connections to the AlN film. Subsequent testing of the AlN film and substrate and visual inspection of the housing and internal components revealed no damage from the 1,100° temperature.

a	b
Figure 3.The reflection from a metal target 33 mm from the front of the sensor is shown at 2.5 microseconds at (a) 150°C and (b) 576°C. The pressure was 150 MPa. The data acquisition was delayed for 83 microseconds, relative to the initial sensor excitation, before acquiring this signal.

References

For further information, contact D.A. Stubbs, University of Dayton Research Institute, 300 College Park Drive, Dayton, Ohio 45469; (513) 229-4756; e-mail STUBBSDA@ml.wpafb.af.mil