Testing of Abradable Coatings
Dr. Thomas Taylor, Praxair Surface TechnologiesThis talk introduces the newly completed "rub test rig", which is now undergoing trial runs.
Traditionally we have developed surface coating which are not intended to fall off such as MCrAlY type coatings. Tonight, however, we will be considering how to evaluate coatings which are specifically designed to wear away. These coating are used to help minimize the compression losses in turbine engines and specifically help maintain the seal between blade tips and their enclosing shroud. It is particularly noted that during an engine rev (from cold start, and in an aircraft take-off), the blades expand at a higher rate and sooner than the encasing shroud, and since it is very undesirable to have blade tips wear, an abradable coating is put on the shroud to allow for wear with minimal drag so that the seal may be maintained and the engine continue to perform. The need for such coatings is also driven by the fact that more and more performance is demanded from todays jet engines, and as the temperature limits of known material have been met, the only way to get further improvements is through less clearances and tighter tolerances in seal areas.
There are basically four current abradable coating used for this application. Ultra Low density YSZ and Low density CoNiCrAlY materials are applied by Thermal Spray. NiCrAl-Bentonite and Ni-Graphite are applied by electrodeposition. In some cases the abrasive resistant tip coating are formulated specifically to cut the abradable easily (reducing drag forces on the engine).
All of this seal technology has produced a need for a testing device to try to optimize the coatings, application techniques, and clearances toward some specific measurable parameters with the goal of long-life sealing and minimal power loss in the engine due to contact and abrasion. To meet this need Praxair has developed the test rig we are describing tonight.
Early on, some parameters were established for the test rig. A tip speed of up to 1300 ft/sec (12500 rpm on our rotation arm radius), and infeed (forced interference) of up to .300" with precision control over infeed rates. In our design tips are mounted on 1 to 4 symmetrical rotating arms on the main rotating shaft. These can be precision fed to hit a stationary strike plate for which two dimensional stresses (tangential and Radial) are measured remotely via metal arm linkage.
In addition to a super heavy construction (for safety), there is a precision balancing mechanism (to reduce false data due to vibration), and a heating system for the shroud material strike plates to produce a testing environment similar to various engine stages. Many thermocouple as well as optical devices monitor and record precise temperatures of test material components as a test is run. As many as 10,000 lines of data (temperature, force, velocity, displacement (infeed), etc) are recorded each second the rig is in operation)
An early demonstration test of the rig demonstrated of both "broad (condensed) views" and "detailed views" of the force data (as plots) can be related to the abrasion of the coating, and even to exact strikes which produced a spall off.
A problem with the four arm design of this rig is that it consumes a lot of power pumping air. The current motor driving the unit is only able to get the arms moving to 4500 rpm which is far short of the desired 12500 rpm capability. It has also been determined that "tip temperature" (which can vary significantly from "should temperature") would be a very desirable property to monitor, and the present unit does not have a means to track this. Finally a more direct method (as opposed to sensing the two dimensional forces remotely though measuring arms) would be desirable for measurement of strike forces.