HPC surface roughness

The operation of an aircraft engine causes several different deterioration modes acting on the engine performance. One significant factor when focusing on the performance reduction of an HPC is described by surface roughness (Seehausen et al., 2020). Based on environmental factors such as aerosols, dust, and other particles, blade’s surface roughness is increased, resulting in performance losses (Tarabrin et al., 1998; Morini et al., 2009; Bons, 2010; Wensky et al., 2010). Gilge et al. optically measured the blade surfaces of HPCs which had completed 20,000 cycles (Gilge et al., 2019a, b). Additionally, the equivalent sand-grain roughness ks is calculated using the correlation of (Bons, 2005) in conjunction with the shape and density parameter of (Sigal and Danberg, 1990). The k_s values calculated were used by (Seehausen et al., 2020) to study the multistage effect of surface roughness. Based on previous studies, (Seehausen et al., 2020) could extrapolate the measurements of (Gilge et al., 2019a, b) to all blades of the HPC. Figure 2 presents the extrapolated average roughness height across the stages on blades and vanes, on both suction and pressure side. The surface roughness significantly decreases across the first half of the HPC, caused by deposits and particle impacts on the suction side. On the pressure side, a more homogeneous roughness distribution is formed. To illustrate the complexity of the surfaces examined, examples of characteristic surface structures depending on the position in the HPC are shown in Figure 2.

Figure 2.

Roughness distribution across the stages of the HPC after 20,000 cycles in service and examples of measured surfaces (Gilge et al., 2019a).

In (Goeing et al., 2020), the effect of HPC surface roughness on the transient engine performance was extensively investigated in ASTOR. To model the transient manoeuvre in ASTOR, performance maps at different rotational speeds and roughness levels are calculated using numerical solutions from TRACE. Following which, the performance maps are scaled using a method described by (Li et al., 2011) which results in lower computational effort. The individual scaling factors indicate a dependency between the Reynolds-number and the roughness effect, which is consistent with the findings of (Nikuradse, 1933; Moody, 1944). In the left plots of Figure 3, the HPC performance maps used in this study are illustrated. The green lines represent the map of the smooth blades and vanes, while the black lines correspond to scaled maps of average roughness height after 20,000 cycles in service.

Figure 3.

CFD performance maps. Left: Performance map of the HPC with smooth blades and vanes at different rotational speeds compared with the scaled map of average roughness height. Top: Pressure ratio πtt over mass flow m˙. Bottom: Polytropic efficiency ηpol over mass flow m˙. Right: Turbine performance maps with different tip gaps. Top: Mass flow m˙ over pressure ratio 1/πtt. Bottom: Polytropic efficiency ηpol over pressure ratio 1/πtt.

Surface roughness reduces the mass flow due to thicker boundary layers. This comes along with a decreased efficiency and increased pressure losses. At low rotational speeds, the effect of surface roughness increases due to smaller Reynolds numbers.

Reitz (Reitz et al., 2018) performed a scaling of the HPC performance map with a doubled tip gap size. As shown by (Seehausen et al., 2020), the overall effect on the performance is larger for surface roughness than for tip gaps. In contrast to surface roughness, doubled tip gaps mainly affect the polytropic efficiency and not the pressure rise. A CFD study by (Reitz and Friedrichs, 2018) on the effect of modified geometries in the front and rear stages, caused by deterioration, revealed that the performance could be increased by local geometry changes. The intensity of the effect strongly depends on the geometrical parameter modified and varies with the stage (Lange et al., 2012). Regardless of the deterioration type, i.e. surface roughness or geometrical deterioration, the performance is predominantly affected by the front stages resulting in multistage effects.

HPT tip clearance

HPT blades and vanes are exposed to high thermal and mechanical loads due to high gas temperatures and rotational speeds. This causes a range of different deterioration types, an increased blade tip gap being one of them. Typically, the gaps between the turbine blade tips and the shroud are designed to be as small as possible. As described by (Denton, 1993), the secondary flow in the tip gap contributes significantly to the losses in the turbine. Bindon showed that the losses increase with increasing radial gap (Bindon, 1989). To investigate the effect of the increased radial gap on the overall performance, the CFD simulations of the V2500-A1’s HPT are performed.

The V2500-A1 HPT consists of two stages and in this study the blade tip gaps of both stages are varied. The variation is set as a ratio of radial gap s and blade height h. Accordingly, two meshes with different s/h ratios are created. A ratio of s/h = 1% is used as a reference and the increased tip gap is represented by s/h = 3%. Table 2 summarises the meshing setup for tip gaps.

Results are evaluated in terms of corrected mass flow m˙corr and polytropic efficiency ηpol as a function of the total pressure ratio 1/πtt. The resulting HPT performance maps are shown in the right diagrams of Figure 3.

The lower plot of the right diagrams in Figure 3 shows the influence of the increased gap (blue lines) on the polytropic efficiency. The higher radial tip gap decreases polytropic efficiency by up to 3.8% across all speed lines, which corresponds to earlier studies (Hourmouziadis and Albrecht, 1987). This is caused by increased secondary flow losses in the tip gap. Hence, a negative effect of the increased tip gap on the overall HPT performance can be observed. The influence on the corrected mass flow is not as explicit and decreases by only 0.06%–0.35%

Thrust specific fuel consumption (TSFC)

In the upper left diagram of Figure 4 the thrust specific fuel consumption (TSFC) is shown over the time t for all engines. In general, the full TSFC range is between 6.3–15.5 g/(kN⋅s). During the slam acceleration, the TSFCs have overshoots, due to the spool’s inertia and the resulting attenuated mass flow. The same effects are responsible for the peaks during decelerations. It is recognisable, that the TSFC of the new engine is consistently the lowest. The mean TSFC of engine Θ2 is 5.1% higher and Θ3 is 4.9% higher compared to Θ1 with changing signs at Band B. Moreover, Θ2 has got an increased sensitivity towards transient operation, indicated by the higher peaks. The combined deteriorated engine Θ4 constantly shows the highest TSFC at each thrust level throughout the entire performance simulation. The predicted TSFC is 10.2% higher compared to the new engine. Therefore, an engine with combined degradation domains, thus a higher degree of overall deterioration, requires more fuel flow to achieve the same thrust levels. Additionally, the simplified combined engine Θ5shows an accurate prediction in Band C and D.

Figure 4.

Left: Identification of the degree of deterioration of the 5 engines. Upper left: Thrust specific fuel consumption (TSFC) over the time t during the pass off test. Bottom left: Exhaust gas temperature (EGT) over the time t. The dots represent the EGT of the steady operating points of the new engine. Right: Identification of the cause of deterioration. Upper right: High-pressure rotational speed (N2) over the time t during the pass off test. On the bottom right: Temperature downstream of the HPC (T_t3) over the time t during the pass off test.

Exhaust gas temperature (EGT)

Like the TSFC, the exhaust gas temperature (EGT) downstream of the low-pressure turbine is an important identifier for the degree of deterioration. The course of the EGT is plotted over time t in the diagram on the bottom left of Figure 4. Furthermore, the steady state EGTs are represented with dots for each thrust level for the new engine. In these curves, different transient effects are visible. During acceleration and deceleration, the impact of in-/decreasing fuel to air ratio is visible due to the shaft/gas inertia. In the constant fuel flow regions, temperature is increasing slowly due to the heat flow from gas to material. Subsequently, the steady performance is not reached after the 10 s duration of each Band. Like the TSFC, the highest EGT is attained by Θ4. The mean deviation from the new engine Θ1 is 12.5% (12.9% for the simplified combined engineΘ5). The deterioration of the HPC as well as the HPT results in a higher EGT. Both engines Θ2 and Θ3 are at a similar EGT level with changing sign in Band B. The mean EGT is up to 6.3% higher compared to the new engine. Furthermore, the transient peaks are more pronounced in Θ2 and Θ3. These peaks increase even more for the combined engine Θ5.

The TSFC and EGT are feasible to determine the degree of deterioration. Obviously, the combined engine has the highest level, but these identifiers are not sufficient to grade the single deteriorated engines.

HP rotational speed N2

In the upper right diagram of Figure 4 the HP system’s rotational speed N2 is plotted during the test. In contrast to TSFC and EGT, the order of the spool speed levels does not correspond to the level of deterioration. Here, the highest rotational speed is detected at engine Θ2. Hence, this engine achieves the target thrust levels with a higher speed compared to the new engine (1.4% on average).

In addition, engine Θ3 requires a significantly lower spool speed during the manoeuvre (−1% on average). In contrast to engines Θ2 andΘ3, engine Θ4is within the same range as the new engine. Only at Band D Θ3 is close to the new engine while, Θ4 is close to Θ2. In general, the pseudo engine Θ5is approximately coincident to Θ4. The same deviations are visible during Band A and Band D for these two engines.

Temperature Tt3

In contrast to pt5 and LP rotational speed N1, the temperature Tt3 shows a did not understand the meaning. This temperature is measured in the V2500 for burner flame out protection. Other engines (e.g. CF34) are not equipped to measure this value. The total temperature Tt3 is shown on the bottom right of diagram of Figure 4. Like the rotational speed, engine Θ2 has the highest Tt3 value at all thrust levels. Basically, the temperature is up to 2.9% higher compared to Θ1. On the other hand, a deteriorated HPT of Θ3 results in a lower temperature. The mean deviation from the new engine is −0.6%. In contrast to the spool speed, Θ4 is higher than the new engine (2.3% on average). The simplified combined engine Θ5approximates coincides with Θ4.

Surge margin SM

The interaction between deterioration and stability of the system is investigated using the HPC surge margin. The change in HPC surge margin compared to the steady state operating line is analysed over rotational speeds N2 during acceleration (Band D to Band A) in the right diagrams of Figure 5.

Figure 5.

System stability during the acceleration between 0 s and 8 s. Right: Surge margin of operating line in HPC performance map over the spool speed N₂ of the HPC. Transient operating line in solid lines and with marker, steady in dashed lines. Left: Relative deviations between transient and steady HPC operating line.

The definition of the surge margin SM is given in Equation 6:

The relative deviations between transient and steady operating line are calculated with

In the right diagram of Figure 5, the SM of the transient operating line (OL) is between 0.52 and 0.22 for the new engine. Compared to the steady state OL, the SM decreases up to −21% (see left diagram of Figure 5). A significant contrast to Θ1 is visible in the SM of Θ2. This SM is between 0.31 and 0.07. Moreover, the deviation from the steady operating line has a maximum of −35%. The stability of Θ3 is similar to Θ1. However, an offset is present and the SM of the transient OL is between 0.50 and 0.21 with a maximum of −22% to the steady state OL. Furthermore, the OL moves towards lower rotational speeds.

The impact of combined HPC/T deterioration is shown by Θ4. The SM has the largest offset to the new engine and is between 0.3 and 0.08. The max. deviation to the state steady OL is like Θ2 with −37%. The simplified combined engine Θ5 corresponds to Θ4. A significant difference in the performance between the engine Θ2 and Θ3 is visible. Therefore, the highest deterioration is found in engine Θ4, followed by engine Θ2.

Summary/Explanation

In Figure 4 different identifiers for the overall degree of engine deterioration as well as components affected by deterioration have been analysed. In general, two effects are responsible for the performance degradation. First, the reduced efficiency ηpol of HPT and HPC is comparable to a reduction of nozzle cross sectional area and a shift of the HPC OL above the OL of a new engine (throttle effect). Second, the deteriorated HPC changes the shape of the HPC performance map and the position of the stability line moves towards the OL (shift effect) (see upper left diagram of Figure 3).

The HPC of Θ2 is not able to provide enough pressure for the energy recovery in the HPT (shift effect). Hence, the fuel flow increases to achieve the same thrust level, which results in a higher EGT. Subsequently, the rotational speed increases the pressure level which delivers enough pressure (with the combination of fuel flow). Furthermore, a reduced HPC ηpol increases the compressor exit temperature Tt3 (see Equation 2).

The reduced HPT efficiency (see bottom right diagram of Figure 3) results in a decreased energy output from the turbine with a smaller pressure ratio 1/πtt. This results in a smaller compressor pressure ratio πtt and mass flow (throttle effect). Consequently, the rotational speed decreases and more fuel flow is necessary to achieve the same thrust level (EGT increases).

The combination of both degradations increases the EGT significantly compared to the other engines. However, the performance impact of deteriorated HPT dampens the increasing temperature Tt3 and spool speed N2 of the responsible deteriorated HPC, compared to the engine Θ2. Therefore, both curves move in the direction of the new engine performance.

In the next step of this analysis, the impact of transient loads and system stability is investigated. Especially during the acceleration, an increasing EGT peak occurs due to the system’s inertia (see Figure 4). The reduced mass flow has an impact on the fuel to air ratio and therefore on the transient loads. Hence, the reduced mass flow Δm˙decreases for both degradations, due to the shifted operating lines (throttle effect) and the shape of the characteristic curves.

Furthermore, both effects have a significant impact on the surge margin. The degree of HPC/T deterioration are at the same level (see TSFC and EGT), but the SM of Θ2 is significantly lower than that of Θ3. Therefore, the second effect (shift effect) has a higher influence on the system stability compared to the throttle effect.

Finally, the simplified combined engine model Θ5 can be used to represent this performance and is close to the values of Θ4.

The shift and throttle effect influence the engine performance in different ways. However, the modified compressor and turbine aerodynamics with further deteriorations such as a reduced leading/trailing edge, max. thickness and the resulting stagger angle are investigated in serval studies (Scharfenstein et al., 2013; Reitz et al., 2014; Ernst et al., 2016). These results exhibit that the positions of the characteristic curve (mass flow, pressure ratio and efficiency) are qualitatively and quantitatively dependents on the type of deterioration. Based on the results of engine Θ4, it can be concluded that combined degradation in one turbomachine increases the deterioration of engine performance and stability, but also dampen it.