Method of Cleaning Turbine Blades
AssigneeAirbus DS GmbH
InventorWalter SCHWARTING
Life of this patent
Estimating the term…
Loading patent…
AssigneeAirbus DS GmbH
InventorWalter SCHWARTING
Estimating the term…
In a process for cleaning contaminating silicon dioxide SiO2 off of turbine blades, silicon dioxide that is on the turbine blades is reduced selectively in a vacuum process that takes place in dry conditions and at elevated temperature inside a high vacuum furnace, by feeding gas having a reducing action, consisting of hydrogen gas or carbon monoxide, is introduced into the high vacuum furnace. The temperature inside the high vacuum furnace is at least about 1000 degrees Celsius, and the residence time of the turbine blades in the high vacuum furnace is determined with reference to the diminishment of the SiO concentration, as analysed with a mass spectrometer. The electrically heatable high vacuum furnace is evacuable by means of a vacuum system that consists of a backing pump and a turbomolecular pump with an adjustable throttle valve and a water-cooled baffle. Further, the gas with reducing action is fed into the furnace via a precision high vacuum regulating valve, and it is equipped with a mass spectrometer.
Process for cleaning contaminating silicon dioxide, SiO2, off of turbine blades, said process being carried out in a dry vacuum process that takes place at elevated temperature inside a high vacuum furnace, and in which at least some of the silicon dioxide on the turbine blades that are to be cleaned is removed by means of a gas with reducing action, characterised in that the silicon dioxide, SiO2, to be removed is reduced selectively to the silicon monoxide stage SiO by the gas with reductive action consisting of hydrogen gas H2 or carbon monoxide CO, which contains a minimum quantity of H2O or CO2 respectively, is then vaporised in a vacuum and pumped out in gas form, wherein the residence time of the turbine blades (2) in the high vacuum furnace (1) is selected until the time when the SiO concentration falls significantly.
Dependencies are inferred from the claim text.