Evaluation of Flow Structure in Gas Turbine Combustor Models by PIV

Transcription

1 Evaluation of Flow Structure in Gas Turbine Combustor Models b PIV Yuto ENDO 1,*, Yoshihiro YARITA 1, Tomohiro IDE 2, Nagaoshi HIROMITU 2, Hisanobu KAWASHIMA 3, Tsuneaki ISHIMA 3 1: Graduate School of Engineering, Gunma Universit, Gunma, Japan 2: IHI Corporation, Japan 3: Gunma Universit, Gunma, Japan * correspondent author: t @gunma-u.ac.jp Abstract For the gas turbine combustor, regulation of emissions is becoming severe ever ear. In order to clear the regulation, new combustors are suggested. One of the solutions for realizing the new combustor is to appl the Rich burn Quick quench Lean burn (RQL) sstem. The RQL sstem can realize low emissions b miing rapidl fuel and air. The combustion test results of RQL model indicates that emission is changed b changing alignment of the primar dilution holes. Changing the shape of the primar combustion zone is suggested to reduce more emissions. A flat plate named flow guide is placed in the primar combustion zone. As a result, better results are observed b changing the shape of the primar combustion zone. However, the flow characteristics are still unclear. In this stud, the flow structure and the turbulent intensit of the model with flow guide are observed b using PIV. Comparison between the results in both of the models without and with the flow guide is carried out. In the primar combustion zone, the flow structure and the turbulent intensit are different b using flow guide. Flow along the flow guide eists in the primar combustion zone. The swirl flow is nearl homogeneous for the model without flow guide. The swirl flow is not observed in the flow in the model with the flow guide. The turbulence intensit of model with flow guide is larger than that of the model without flow guide. 1. Introduction The gas turbine is necessar for man industrial works, for eample aircraft engine, power plants, and so on. The environmental problem has become serious over the recent ears. In an aircraft gas turbine engine, regulation of emission is becoming strict. In Japan, the project Environmentall compatible aircraft engine for small aircraft (ECO engine) has been conducted b NEDO. The ECO engine project is focused on environment responsive technolog required to reduce particularl NO. The new combustor is suggested to achieve target of the ECO engine project, and this research is progressing [1-2]. One of the solutions for realizing the new combustor is to appl the Rich burn Quick quench Lean burn (RQL) sstem. For the RQL sstem, the fuel is burnt b the fuel-rich condition in the primar combustion zone. The combustion air is diluted and cooled rapidl with a large amount of air. It is burnt b the fuel-lean condition in the secondar combustion zone. As a result, it becomes low NO b shortening the combustion time in the area where NO is rapidl generated. In the previous stud of the combustion test with the RQL model, the emission gas is changed b the alignment of the primar dilution holes [3]. The RQL sstem indicates the possibilit to reduce the emission from the combustor. To reach the final goal of emission level, it is required to improve the detail of the RQL sstem. The combustion performance of the RQL sstem depends on the flow characteristics in the combustor. The authors have performed the flow measurement b using PIV (particle image velocimetr) in the combustor models with different alignment of dilution holes [4]. To improve the sstem, one trial test has been performed b using a flat plate, which is called flow guide [3]. The flow measurement of the combustor model with flow guide has not been clarified. In the stud, the difference - 1 -

2 in the flow structures and the turbulent intensit b the flow guide is discussed with the eperimental results b using PIV. 2. Eperimental Setup 2.1 Combustor model The measurements are carried out in the combustor models which are without and with flow guide. Figure 1 shows the combustor model. The combustor models are applied the Rich burn Quick quench Lean burn (RQL) sstem. The swirling flow is generated b swirler. The back and side of the model are made b acrlic plates for visualization. Primar dilution holes are located at the upstream side in the top and bottom of the model. Primar dilution holes are four holes, and the are located in the staggered alignment. Secondar dilution holes are on the downstream side in the top and bottom of combustor. Secondar dilution holes are si holes which are located in the opposed alignment. Primar combustion zone is defined as the area from fuel nozzle to primar dilution holes. Secondar combustion zone is defined as the area from primar dilution holes to secondar dilution holes. A flat plate, named flow guide, is placed in the primar combustion zone. Fig.1 Combustor model specification Fig.2 Eperiment setup Fig.3 Measurement section and position of dilution holes 2.2 Measurement setup and methods Figure 2 shows eperimental setup. The eperimental stud has been carried out b using PIV (particle image velocimetr). The PIV is used to measure the mean velocit and turbulent intensit in the flow. The PIV (TSI: INSIGHT Version3.0) consists of Nd:YAG Laser (New Wave Research: Minilase III 50mJ) as a source of light, and CCD camera (TSI : PIVCAM TSI Model pi. 1018pi.). The laser sheet is generated using a clindrical lens with thickness of approimatel 1mm. Oil particle with a mean diameter of 2.0 µm is used for the tracer particle. A final interrogation area size is piels and 50% overlap. The pulse interval is set from 5µs to 10µs. The flow rate is set to 0.075m 3 /s. The flow rate is calculated from the orifice and manometer installed in the duct of the combustor model upstream. The fuel is not injected from the fuel nozzle. Figure 3 shows the measurement sections and the dilution holes position. The measurement sections are three planes along direction of the flow and a vertical plane to - 2 -

3 the direction of the flow. The measurements planes are represented as A-A section, B-B section, C-C section, and D-D section. The ais is set to a direction of the main flow. The ais is set to horizontal direction. The z ais is set to vertical direction. The origin is set on bottom of the primar dilution hole. 3. Results and discussions In the stud, vector maps are obtained b averaging from 200 instantaneous velocit vector maps. The flow patterns in both of combustors are compared with each other. All of the coordinates in the vector maps are translated to dimensionless length b using length from edge of primar dilution hole to edge of secondar dilution hole. The mean flow velocit is also normalized b using inflow velocit U 0. Turbulent intensit is calculated b the following epression. Turb = u u ' ' = u = u ( u ) + N ( u ' 2 ' 2 2 { ( )} u, 1 2 u u N ) guide, the swirl flow which is observed in Fig. 4(a) cannot be observed. The velocit vectors are directed in a radial direction in the region of 1 z/l 3 and 2 /l 3. The organizing motion cannot be appeared in the region of 1 z/l 3 and 1.5 /l 2. The flow outside of these regions has weak swirl motion. The turbulence intensit is large at center region of figure 4 (b). 2 2, = u u u + Here, u and u are instantaneous velocities, u and u are mean velocities. In the following sections, the mean flow velocit distribution is shown b the vector, and the turbulence intensit is shown b the contour. 3.1 A-A section Figure 4 shows the result of a vector map and the turbulent intensit in the A-A section. Figure 4(a) is the results in the combustor model without the flow guide, and figure 4(b) is those with the flow guide. The A-A section is -z plane near the swirler. guide, the swirl flow with an anti-clockwise eists. The turbulence intensit is large at 1.5 z/l 2.5 and 1.5 /l 3. Fig.4 Vector maps and turbulence intensit in A-A section - 3 -

4 The mean flow velocit distribution and the turbulence intensit are compared between the combustor models without and with the flow guide. The combustor model without flow guide is observed homogeneous swirl flow. On the other hand, the flow of the combustor model with flow guide is not homogeneous and the flow of radial direction in the center area can be observed as shown in figure 4 (b). On the whole eperimental region, the turbulence intensit of the combustor model with flow guide is larger than that of the combustor model without flow guide. Fig.5 Vector maps and turbulence intensit in B-B section 3.2 B-B section Figure 5 shows the results of a vector map and the turbulent intensit in the B-B section. Figure 5 (a) is for the result without the flow guide and Fig. 5 (b) is that with the flow guide. guide, an inclined jet from the bottom side to the top side eists in the secondar combustion zone. The jet is formed b the flow from the primar dilution hole. The inclined angle of the jet becomes large toward downstream. In the primar combustion zone, the downward flow is confirmed. This flow is formed b the swirl flow. This swirl flow interacts with the jet from the primar dilution hole of bottom in 1.5 z/l 3.5 and = 0. In the bottom of the primar combustion zone, a vorte is confirmed. This vorte is generated b the interaction between the swirl flow and the jet. The mean flow velocit of the primar combustion zone is smaller than that of the secondar combustion zone. The turbulence intensit is large at the center of the vorte and between the primar dilution holes of top and bottom. guide, the jet is observed in the secondar combustion zone. The jet is the same as the above result. For the primar combustion zone, the flow is directed to the upstream direction in 1 z/l 3. The top and bottom areas of the primar combustion zone have flows along the flow guides near the flow guides. For the primar combustion zone, the downward flow is observed on /l = 0. The turbulence intensit is large between the primar dilution holes. This turbulence intensit on 0 z/l 4 and /l = 0 is caused b miing the flows. For the primar combustion zone, the turbulence intensit becomes large when the back flow and the flow along the flow guide are combined. The differences in the mean flow velocit distribution and the turbulence intensit without and with flow guide are discussed. The flow in the primar combustion zone is different when the flow guide condition is changed. The counter flow is formed in the region for the model with flow guide. In - 4 -

5 corner of the primar combustion zone, the flow along the flow guide eists in the result with the flow guide. The flow in the region without the flow guide becomes weak. In the area between the top primar dilution hole and bottom one, the large turbulence intensit area in the model with flow guide is larger than the model without flow guide. On the whole measurement region, the turbulence intensit of the combustor model with flow guide is larger than that of the model without flow guide. 3.3 C-C section Figure 6 shows the results of a vector map and the turbulent intensit of the combustor model in the C-C section. Figure 6(a) is for the result without the flow guide and figure 6 (b) is for that with the flow guide. guide, the downward flow is observed in 1.5 z/l 3.5 and 0 /l 1. In the center of the primar combustion zone, the flow directed to the direction eists. In z/l 1.5 and 0.5 /l 1, the upward flow is observed. In the secondar combustion zone, the flow with positive direction is observed. The turbulent intensit is large at 0.5 z/l 2 and 0 /l 1. This large turbulence intensit is caused b the miing of the flows which have both of positive and negative flow directions in z direction. The turbulent intensit is also large at primar combustion around the nozzle. guide, the flow has positive direction in the primar combustion zone. The flow along the flow guide eists for top and bottom of the primar combustion zone. In the area of 0.5 z/l 3.5 and 0 /l 1, the flows both with upward and downward eist. The impinge each other in the region. In the secondar combustion zone, the flow with positive direction is observed. The turbulent intensit is large at 0.5 z/l 3.5 and 0 /l 1. The large turbulent intensit is caused b miing flows with both positive and negative direction in z ais. The turbulent intensit is also large at primar combustion around the nozzle. Differences in the results with and without the flow guide indicate the difference in the mean velocit near the flow guide in the corner of the primar combustion zone. In the region, the flow is ver weak in the model without the flow guide but the airflow eists along the guide in the model with the flow guide. The turbulent intensit distribution is different between the flow guide conditions. Fig.6 Vector maps and turbulence intensit in C-C section - 5 -

6 3.4 D-D section Figure 7 shows the results of a vector map and the turbulent intensit in the D-D section. Figure 7(a) is for the results without flow guide and 7(b) is for that with flow guide. guide, the jet from top to bottom is observed in the secondar combustion zone. The weak downward flow is observed in the primar combustion zone. The flow is affected b the jet from the primar dilution hole. In the region near the nozzle, the upward flow eists. This flow is caused b the swirl flow as shown in Fig. 4. The turbulence intensit is large at the primar combustion zone. This large turbulence intensit is affected b the miing between the swirl flow and the jet from the primar dilution hole. guide, the jet eists in the secondar combustion zone. The direction of the jet is from top to bottom. The jet is formed b the airflow of the primar dilution hole of the top. The primar combustion zone has the flow along the flow guide near the top flow guide. The mean flow velocit along the flow guide of bottom is smaller than that of top side. In the center of the primar combustion zone, the flow with positive direction eists. The turbulence intensit is large at primar combustion zone and secondar combustion zone. The mean flow velocit distribution and the turbulence intensit in the combustor model without the flow guide are compared with that of the combustor model with the flow guide. The difference b the flow guide is appeared in the mean flow vectors at the primar combustion zone. The flow is ver weak and it is mainl directed to the downward. However, the flow of the combustor model with flow guide, the positive direction flow in the primar combustion zone is appeared. The turbulent intensit distribution is different between the models with and without flow guide in the primar combustion zone. In the secondar combustion zone, the turbulent intensit of the model with flow guide is larger than that of the model without flow guide. Fig.7 Vector map and turbulence intensit in D-D section - 6 -

7 4. Conclusion The differences in the flow characteristics in the combustor models are discussed with the eperimental results. The combustor models without and with flow guide are tested. For both of the models, the flow is measured b using PIV. The flow structures in both of the model are compared with each other. The conclusions are followings: [3]Tomohiro I. et al. (2011) Simple Low NO Combustor Technolog for Environmentall Compatible Engine(ECO Engine). 10 th International Gas Turbine Congress 0232:1-4 [4]Yarita Y. et al. (2011) PIV Measurement in Gas Turbine Combustor Model. 10 th International Gas Turbine Congress 0224:1-7 [1] The flow guide can affect the formation of the swirl flow near the nozzle eit. The swirl motion is formed in the model without the flow guide. The radial flow is formed in the model with the flow guide. [2] The flow along the flow guide is observed in the result with the flow guide. [3] The flow in the primar combustion zone is different with the flow guide conditions. [4] The turbulence intensit of model with flow guide becomes larger than that of the model without flow guide. 5. Acknowledgements This stud is conducted under the contract with New Energ and Industrial Technolog Development Organization (NEDO) as a part of aircraft and space industr innovation program and energ innovation program of Ministr of Econom, Trade and Industr (METI). Here, the authors epress their gratitude to the NEDO staff. The authors would like to thank to Prof. Saito and Prof. Sanada, Shizuoka Universit, for their help to use their PIV sstem. 6. References [1]Nagaoshi H. et al. (2007) Simple Low NO Combustor Technolog. IHI engineering report Vol.47 No.3 [2]Hideki M. et al. (2008) Research and Development of the Combustor for the Environmentall Compatible Small Aero Engine. MHI engineering report Vol.45 No.4-7 -