Where can I reliably check an aircraft turbine? Only in Russia…
Experience is a matter of gain
With all the constant complaints about the problems of the Russian aviation industry, no one seems to have noticed how domestic technologies for diagnosing turbojet engines have become the most advanced in the world. aviation long lasting story, and gas turbines have even more. The turbine was invented at the end of the 18th century to be used in metallurgy instead of blacksmith bellows, but it never found practical use.
At that time, many designers, due to the novelty of the technology, did not yet understand what they were up against. But even today, many diagnostic technologies make your hair stand on end due to the fact that even in factory, not pre-flight tests, extremely dangerous methods were and are still used.
For example, in the spring of 1937, engineer Hans-Joachim Pabst von Ohain began conducting stationary tests of the HeS 1 turbojet engine. Instead of kerosene, hydrogen was used as fuel, which has increased stability during combustion, but is highly explosive when combined with air.
For a long time, control and measuring devices have been used to evaluate the performance of turbojet engines. Thus, when testing the TR-1 engine, the "pulsation" of the rotating blades was evaluated in more than five hundred different positions. Data was monitored from a large number of sensors, pyrometric devices, thermocouples and other equipment.
There was something quite exotic by today's standards. In May 1939, the HeS3 engine was installed on a propeller-driven aircraft with another He-118V2 engine and suspended under its fuselage for in-flight testing directly during flight.
But the main problem was the need to disassemble turbojet engines into parts already during operation. This entailed an increased role of the human factor, someone somewhere overlooked something, as a result, a turbojet engine malfunction is always a disaster.
Russian technology companies have been working hard in recent decades to reduce human errors through automation. And their successes by global standards are not just noticeable, but significant. Another thing is that not all of these technologies are fully automated.
Assembly not on a conveyor belt
Among other tests, scanning of the technical condition of the turbine has long been carried out without disassembling it into parts. For example, the company SHINING 3D uses three-dimensional scanning, but the disadvantage is that the FreeScan UE Pro handheld scanner, the accuracy of which is up to 0,02 mm, must be held by the operator in his hands, which is why it is made so light, weighing less than a kilogram.
Further, the data is indeed processed by artificial intelligence, but such a diagnostic tool cannot be considered fully automatic, especially since the AI data must still be checked by an operator. And such technologies, although they use AI, clearly need to be modernized.
An example of this: the following photographs. The first one is a 2D section. It is clear that this is already the last century, but let's still remember our youth.
The area of the scanning defect is very clearly visible, two blades are either burnt out or broken, two more give the impression of being corroded. Instead of a nozzle, there is a solid surface.
The second picture shows that 3D is not much better. If not worse. It seems that instead of blades in the area highlighted by the red square there is some kind of shapeless mass.
Such distortions can be both fatal and, at best, costly.
A more modern and progressive technology is tomography, not to be confused with SHINING 3D technologies, which are related to photogrammetry. The tomogram displays the internal structure of an object as accurately and clearly as possible. Various materials, characterized by different density and chemical composition, are easily identified from each other.
The geometric parameters of any hidden area or partition can be determined with an error of no more than 50 micrometers. Imperfections such as pores and foreign particles become visible. Even microcracks of 50 micrometers in width are reliably detected, regardless of their direction and location.
The only issue is that the specifics of tomographic studies of jet engines are associated with fundamentally different requirements for scanning materials used to create parts of the “hot” and “cold” parts of the engine.
This is not a hospital for you.
For the "hot" part of the engine, it is important that the parts withstand high temperatures and overloads. The "cold" part - the housing and compressor - is made of lightweight materials to reduce the overall weight of the engine as much as possible. The catch is that dense heat-resistant materials can absorb all radiation, and lightweight materials are almost transparent.
Often, intermediate, "compromise" radiation energies are used, which allow for the reconstruction of both denser and less dense parts. But this also has a negative effect: some of the real defects are not visible on the engine tomography, and false defects appear in places where there is not enough information. A repetition of the same photographs that we saw above.
The company "Promintro" once worked in this way, but unfortunately it is no longer in operation. Probably, the reason for the company's disappearance from the turbojet engine diagnostics market is its too low level of automation and AI use.
This year, however, Russia has made a breakthrough in the field of interaction between AI and tomography technologies for diagnostics of turbojet engines in assembly.
Scientists at Smart Engines have proposed a new combination of tomography and AI to scan a jet engine without disassembling it into parts and in a single study on a serial industrial installation. The technology creates a digital twin of the engine and allows for the reliable detection of defects - cracks, voids, delaminations - and foreign objects, such as metal shavings, in a single measurement.
The breakthrough of the scientists at Smart Engines is the creation of special high-performance computer tomography algorithms for the correction of distortions that occur when using intermediate radiation energies. The algorithms developed by Smart Engines allow for reliable compensation of the effects of radiation scattering by an object, radiation scattering by a detector, radiation polychromaticity when studying multi-material products, as well as noise associated with photon starvation.
Leafing through the calendar of tragedies
AI enables diagnostic functions to be performed even in the case of complex measurements performed at the limit of the device's sensitivity, making it possible to carry out flaw detection of aircraft jet engines without compromising the integrity of the structure.
History teaches us from the mistakes of the past. Oversight in the maintenance of turbojet engines has caused many tragedies. In 1989, at the Nasosnaya airfield near Sumgait, an Il-76MD transport aircraft, several minutes after taking off from the runway following a turbojet engine fire and an emergency approach planned in connection with this, suddenly began to rapidly lose altitude and fell into the water.
Everyone died then. The investigation showed that the cause of the accident was damage to the low-pressure turbine shaft of the D-30KP engine, caused by the failure of the bearing between the shafts. A more recent incident that occurred with a turbojet engine, when high technology was supposedly already in full swing, also led to human casualties.
On May 5, 2019, an Aeroflot SSJ-100 aircraft flying from Moscow to Murmansk returned to Sheremetyevo Airport 37 minutes after takeoff due to a malfunction of one of these devices. As a result, a fire broke out, which led to the death of 41 people out of 78 on board.
On November 1, 2019, a Sukhoi Superjet 100 of Yamal Airlines, flying from Tyumen to St. Petersburg, landed at the departure airport an hour and a half after takeoff due to the failure of the left engine. During the climb, the flow through the turbine stalled. There were no casualties.
Therefore, until high technologies are implemented that use AI to check turbojet engines, it will definitely not be possible to achieve 100% safety (it is clear that this is a utopia, but we must strive for something). Everywhere someone has failed to do something, has overlooked something, and people are dying. Maybe AI will do it better?
Information