Checking the Correctness of the Process of Brazing of the Honeycomb Seal to the Base by Ultrasonic Method
, , und | 28. Nov. 2023
Article Category: research article
Online veröffentlicht: 28. Nov. 2023
Seitenbereich: 47 - 68
DOI: https://doi.org/10.2478/fas-2022-0005
Schlüsselwörter
© 2022 Agata Świerek et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
In modern aircraft turbine engines, a very important role is played by the proper selection and fabrication of seals between the rotating parts of the compressor and the turbine mounted on the main engine shaft (e.g. blades) and the elements of the fixed casing (the stator of a turbine). One of the most modern solutions in this field are honeycomb seals, which have a ‘honeycomb’ structure and are soldered perpendicularly with the edges of individual cells to a cylindrical surface. This process consists in simultaneous soldering to the surface of a large number of thin-walled cells, wherein the contact line between the soldered elements is a hundredth part of a millimetre thick and the connection length reaches hundreds of metres.
In the process of soldering honeycomb seals in aircraft engine components, the most important technological aspects are preparation of the surface for soldering, positioning of the seal structure to the stator, selection of soldering process parameters and the method of checking the correctness of the process.
This paper is devoted to the last of these issues.
A commonly used method for testing the correctness of the soldering process of honeycomb seals is the visual method. In this method, the operator is most often supported by simple optical devices (magnifying glass, endoscope, borescope, etc.). The method requires the involvement of the operator during the entire process, and his/her decision regarding the correctness of the connection is made based on subjective observations.
During the analysis of other non-destructive testing methods, the ultrasonic method was selected as the one that can be used for this purpose.
The aim of the work was to check the possibility of using the ultrasonic method to test the correctness of the soldering process of honeycomb seals in aircraft engine components and to compare the tests carried out using this method with the currently used visual tests.
The motivation to take up this topic was to check whether there is a possibility to improve the checking process, which is currently the most labour-intensive and time-consuming method.
Section 2, ‘Honeycomb seals in aircraft turbine engines—analysis of literature and industrial documentation,’ presents the advantages of honeycomb seals compared to the previously used metal–ceramic seals and describes the technology of soldering honeycomb seals, paying attention to the most important aspects of the process, such as: preparation of the soldering surface, positioning of the seal before soldering and the currently used method of checking the correctness of the soldering process.
Section 3, ’Aim and methodology of research,’ presents the aim of the work. Two non-destructive testing methods are described, and test samples specially made by Aalberts Surface Technologies, Kalisz, are presented. The research methodology used for ultrasonic tests and the test stand are described.
Section 4, entitled ‘Test performance and result analysis,’ describes the tests performed using the ultrasonic method and presents the results of these tests. The section ends with a summary and relevant conclusions.
In modern turbine aircraft engines, a very important role is played by the proper selection and implementation of seals between the rotating parts of the compressor and turbine mounted on the engine shaft and the elements of the fixed housing.
In different places of the power unit, various types of seals are used.
The choice of the type of sealing depends on the working conditions of a given engine section, which is associated with the occurrence of significant, variable temperature and pressure values, as well as the presence of strongly erosive and corrosive influences of combustion gases.
The most important seals include those whose task is to reduce the gap between the rotating parts of the turbine (blades) and the elements of the stator of a turbine. With the reduction of this gap, the amount of gases flowing through the turbine increases, which in effect increases the efficiency and performance of the engine.
So far, metal–ceramic seals were used in turbines and the efficiency of the turbine was determined only by the size of the gap between the blades and the steering wheel. Currently, honeycomb seals made of nickel alloys are used.
The advantage of honeycomb seals over the previously used ones is that individual cells act like low-pressure gas-compressing elements between the blades and the steering wheel. There is a phenomenon of gas sealing, consisting in the fact that gases flowing through the gap with a seal are inhibited by the walls of the cells, reflect from them and serve as additional resistance for flowing gases. As a result, the efficiency of the turbine with such sealing is determined not only by the size of the gap but also by the effect of dynamic gas engine sealing obtained owing to the honeycomb structure.
An additional advantage of the honeycomb seal is its abrasion resulting from the low density of the structure. This abrasion radically reduces the possibility of damage to the rotor in the event of its contact with sealing.
The technology used to connect the honeycomb seals with the stator is soldering at high temperatures, carried out in vacuum furnaces. In this process, the most important technological aspects are preparing the surface for soldering, positioning the structure of sealing to the stator, selection of parameters of the soldering process and the method of checking the correctness of the process.
The honeycomb seals used for steering devices in turbine aircraft engines are made of nickel alloy based on Aerospace Materials Specifications (AMS)-5536 (Hastelloy X). It is an alloy with a high content of chromium, characterised by high resistance to high temperatures and oxidation. To produce honeycomb seals, hot-rolled sheets are used, which are then subjected to supersaturation at a temperature of 1,150–1,175°C and quickly cooled down.
The honeycomb seal is manufactured from a thin metallic foil, typically 75–150 µm thick, which is pleated and then laser-welded at the knots to hold the adjacent walls of the corrugated strips together to form hexagonal cells. The welded honeycomb is trimmed and shaped to fit the components to which the honeycomb seal is brazed.
The standard structure of the honeycomb seal is shown in Figure 1.
Figure 1.
Honeycomb seals.
Soldering of honeycomb seals is performed in vacuum furnaces. The most important technological elements of this process are selection of the appropriate solder, preparation of the substrate surface, positioning of the sealant to the apparatus body and selection of soldering process parameters.
Nickel-based alloys are used in the soldering of honeycomb seals to steering devices. Several types of solder are used, produced by specialised companies cooperating with the aviation industry. The chemical composition of these solders is strictly defined by American Welding Society (AWS) standards and AMS. These solders are used in the form of a powder combined with an organic binder and are supplied in the form of a double-sided adhesive tape with a thickness of one to several tenths of a millimetre and a width of several millimetres to tens of centimetres, depending on the requirements. The most used solder for connecting honeycomb seals with guiding devices is the solder marked BNi2/AMS 4777.
An important factor affecting the quality of soldering is the preparation of the surface to which the sealant is bonded. It should have a varied roughness, enabling proper distribution of the solder on the surface. This varied roughness is achieved by blasting with different process parameters, where aluminium oxide of various grit sizes is used as the abrasive.
The second important factor related to the surface is its cleanliness. Parts intended for soldering honeycomb seals on them must be very carefully washed and degreased. The bodies of the steering apparatus, although they are vacuum castings, have many microscopic impurities on their surfaces and in the near-surface layer. Surface contaminants are removed by abrasive blasting by the casting manufacturer, while contaminants in the near-surface layer come to the surface during vacuum brazing, preventing the solder from wetting the surface. Therefore, a vacuum cleaning of the surface is applied before the soldering operation. The most effective variant of vacuum cleaning is cleaning with the following parameters: temperature = 1,020°C, time = 2 h, vacuum = 10−5 Tr, furnace cooling to 800°C, followed by rapid cooling. Contaminants that come to the surface during vacuum cleaning are removed by abrasive blasting. The surfaces of the body of the steering apparatus, to which the honeycomb seals are soldered, are electroplated with nickel. The thickness of the nickel coating is 3–6 µm.
The next step in the process is to apply solder to the honeycomb capillaries. For soldering honeycomb seals, solder is used in the form of a double-sided adhesive tape made of metal powder with an appropriate composition (in accordance with the AMS 4777 standard) and a binder that burns out at a temperature of 982°C, leaving no undesirable product. Solder in the form of a tape is pressed into the capillaries of the honeycomb sealer using, for instance, a special roller press.
Another important technological element of the process, which has a very large impact on the quality of the connection, is the appropriate positioning of the capillary structure of the sealant to the body of the steering device. An additional difficulty is the requirement to maintain a gap not exceeding 0.1 mm between the sealant tape and the steering device. This positioning and pre-fixing is done by spot welding.
The honeycomb seal attached to the body of the steering device is brazed in a vacuum furnace according to the following parameters: temperature: 1,040±10°C; time: up to 6 min; vacuum: 10−5 Tr; cooling of the furnace to 800°C.
The basic, currently used method of controlling the correctness of the process of soldering of the honeycomb seals in the elements of air engines is visual control using a borescope and gravitational tightness test.
Visual control consists in checking if there is a meniscus on the edges connecting the honeycomb seal with the basic material and whether there are disqualifying damages in the structure of the seal.
Figure 2 schematically presents examples of correct and bad layouts of the solder in the honeycomb seal.
Figure 2.
Examples of the correct and bad layouts of the solder in the sealant cell.
The left cell in Figure 2 shows an example of correct soldering; a dark ring indicates that the solder swam on the walls of the cell to the correct height. The central cell shows an example of average soldering; the dark ring is much thinner. The illustration indicates that the solder swam on the walls of the cell to a low height. The right cell shows an example of bad soldering, as there is no solder on the cell walls.
Figure 3.
Borescopic test stand.
The gravitational tightness test is carried out in such a way that the soldered set with the capillary holes is immersed in penetrant on the section of the 90° arch (one-fourth of the length of the circuit) at time 5–10 s. Then, without taking the assembly out of the bath, it is rotated by 90° for the next 5–10 s. The operation is repeated until full coverage with the penetrant is obtained. The penetrant will not get into tightly soldered capillaries because the air in the capillary will not allow it. From the leaky capillaries, the air will be pushed out by the penetrant, which will fill the capillary. After immersing the entire assembly in the penetrant, it should be laid in such a way that the walls of the capillaries are arranged horizontally, and the penetrant can flow out of the leaking cells. From the leaky capillaries, the air will be pushed out by the penetrant, which will fill the capillary. After the assembly is dried and cooled, it is checked under ultraviolet light. Leaky capillaries with penetrant will glow, and tight capillaries without penetrant will not glow.
The aim of the work was to check the possibility of using the ultrasonic method to test the correctness of the soldering process of honeycomb seals in aircraft engine components and to compare the tests carried out using this method with the currently used visual tests.
For this purpose, we attempted the following:
a method for testing samples of sealant soldered to the base with the use of ultrasonic detection was developed,
ultrasonic testing of samples was carried out,
an analysis of the obtained test results was carried out, and
the summary compares both test methods—visual and ultrasonic—in the context of testing the correctness of the soldering process of the honeycomb seal.
Test samples were made at Aalberts Surface Technologies in Kalisz. This company carries out the process of soldering honeycomb seals to turbine elements of aircraft engines. As a result, we obtained samples made by the currently used vacuum brazing technology for testing.
This is a honeycomb sealer correctly soldered to a thin sheet. Testing of this sample was to check whether such a structure is suitable for ultrasonic testing (Figure 4).
Figure 4.
Sample no. 1.
This is a sealant soldered to a thin sheet with specially generated two areas where the sealant was not soldered. Examination of this sample was to check whether the existence of these defects could be observed using the ultrasonic method (Figures 5 and 6).
Figure 5.
Sample no. 2 – clearly visible unsoldered areas.
Figure 6.
Sample no. 2.
This is a cut fragment of the real structure with a soldered sealant. Testing this sample was to check whether the ultrasonic method would also work in the case of testing such elements (Figure 7).
Figure 7.
Sample no. 3.
This is a large fragment of the structure with a soldered sealant with unsoldered areas generated. Examination of this sample was aimed at checking whether the ultrasonic method could detect all defects in the connection of the sealant with the base. This sample was also used to compare both research methods discussed in the paper (Figures 8 and 9).
Figure 8.
Sample no. 4 – clearly visible unsoldered areas.
Figure 9.
Sample no. 4 – clearly visible unsoldered areas.
The tests were carried out using the ultrasonic method using the echo technique. The signal generated by the flaw detector transmitter is sent to the transducer located in the head. Stimulated by this signal, the transducer emits an acoustic wave. Then, this wave passes through the coupling medium—a liquid that covers the surface around its contact with the head—to the tested part of the object. At the substrate–sealer interface, the wave undergoes reflection and is received by the head transducer.
If there is a proper connection between the base and the sealant, then the wave is partly reflected and partly propagated in the walls of the sealant. The reflected wave, after processing, is visualised on the flaw detector screen in the form of pulses with specific amplitudes, located in specific places (Figures 10 and 11).
Figure 10.
The image of the reflected wave at the correct connection of the base and the sealant.
Figure 11.
The image of the reflected wave at the correct connection of the base and the sealant.
If the connection is incorrect, i.e. there is an air gap between the base and the seal, almost complete reflection of the waves occurs. This is visible on the flaw detector screen by the increased amplitude of the pulses (in relation to the proper connection) (Figures 12 and 13). Depending on the amount of ‘attached’ remnants of the filler cells, this amplitude fluctuates to some degree, but it always remains higher than in the places of proper connection.
Figure 12.
The image of the reflected wave at the wrong base–sealer connection.
Figure 13.
The image of the reflected wave at the wrong base–sealer connection.
A standard ultrasonic flaw detector Phasor by GE Inspection Technologies was used for the tests. It is a portable device that gives the opportunity to conduct research both in the laboratory and in the so-called ‘field,’ i.e. in places of production or exploitation.
The tests were carried out with a Krautkramer head with a transducer operating frequency of 5 MHz. This head is designed to detect small defects located close to the tested surface.
For the purposes of ultrasonic testing of the connections between the base and the honeycomb sealant, for determining the initial and target (when conducting tests of the facility) flaw detector settings and interpretation of ultrasonic signals, Samples 1 and 3 were taken as reference samples.
When making the samples, materials corresponding to those used in the target structure were used. The joining technology used was identical to the production technology of the target structure.
Areas with a solder joint defect consisting in the lack of this joint were modelled in the samples.
Vegetable oil was used as a coupling medium in the tests. Glycerine or a mixture thereof with water, or water, can also be used. What is important, especially in ‘field’ conditions, is that the degree of dilution of glycerine can be selected depending on the location of the surface from which the tests are carried out (smaller for vertical surfaces, greater for horizontal ones).
The test stand is shown in Figure 14.
Figure 14.
Test system for assessing the condition of base–cellular sealant connections. 1 – ultrasonic flaw detector, 2 – head, 3 – coupling medium, 4 – reference sample, 5 – tested object.
Setting the operating parameters of the flaw detector–head system.
Determination of the initial test parameters on the reference sample in the area with the proper state of the substrate–cellular sealant connection (Figure 15). Image of the signal on the flaw detector screen for the area with the correct state of the base–cellular sealant connection.Figure 15.
Verification of test parameter settings on a reference sample in the area with incorrect connection status (Figure 16). Image of the signal on the flaw detector screen for the area with the wrong base–cellular sealant connection.Figure 16.
Optimisation of test parameter settings on the reference sample for the images of signals corresponding to the areas with the correct (Figure 17) and incorrect (Figure 18) connection status. Signal image with proper connection status. Signal image with wrong connection status.Figure 17.
Figure 18.
Methodical examination and recording of results for the areas of the base–cellular sealant connection of the test object.
Sample no. 1 was a honeycomb sealant correctly soldered to a thin sheet; therefore, there were no areas with an incorrect connection state.
As a result of the test, a signal characterising the proper connection of the hive sealant with the substrate was obtained.
Sample no. 2 was a sealant soldered to a thin sheet with two specially generated areas where the sealant was unsoldered.
Depending on the tested place of the sample, signals characteristic of the correct and incorrect connection status were obtained.
In the case of a proper connection, the signals obtained were very similar to those obtained during the test of Sample no. 1, which was entirely made without defects. When examining areas where the sealant was not soldered, a response signal with a much larger amplitude was obtained.
To facilitate the examination, the flaw detector can be scaled so that the response signal is shown on the screen only when there is a connection fault (the last two screen images below).
Sample no. 3 was a cut fragment of the real structure with a sealant brazed on.
Ultrasonic testing identified a small area of poor connection, which is marked in blue. In this region, the signal on the flaw detector has a clearly greater amplitude.
Sample no. 4 was a large fragment of the structure with a soldered sealant with generated unsoldered areas.
While testing Sample no. 4, all areas of unsoldering generated in the soldering process were identified. In these areas, pulses of high amplitude appeared on the flaw detector screen, indicating the existence of an air gap between the base and the seal. Previously, the flaw detector was scaled so that no signals appeared on the screen in the areas of the correct connection.
The tests of four samples carried out using the ultrasonic flaw detection method showed very well that there are premises to use this method to test the correctness of the soldering process of honeycomb seals in aircraft engine components.
During the tests of the samples, the areas of proper connection of the sealant with the base were clearly indicated, when the wave is partially reflected and partially propagated in the walls of the sealant, as well as areas of improper connection, when there is an air gap between the base and the seal, and the waves are almost completely reflected.
Since Sample no. 4 was tested by both methods—visual and ultrasonic—the durations of these tests were compared. In both cases, the time to prepare the test stand was approximately the same, i.e. about 15 min. The time of the visual examination alone was 10 min, while the time for the ultrasonic examination with the marking of areas of incorrect connection was half as long. It should be noted that a specially cut part of the engine part was tested, and the increase in the area to be checked significantly extends the time of visual inspection than that of ultrasonic inspection.
In addition, performing a visual check of the joint of the sealant with the base requires the operator’s full involvement throughout the entire process and his/her decision regarding the correctness of the joint is based on subjective observations. In the case of using the ultrasonic method, it is possible to automate the inspection process, fully document it and objectively assess the correctness of the connection.
The research and analyses carried out in the work indicate the possibility of implementing the ultrasonic flaw detection method to test the correctness of the soldering process of honeycomb seals in aircraft engine components into practice in industrial conditions.