HUANG Jian
1 Shanghai Institute of Space Propulsion,Shanghai 201112
2 Shanghai Engineering Research Center of Space Engine,Shanghai 201112
Abstract:Liquid bipropellant attitude control rocket engines are widely used in satellites,manned spaceships,deep space probes and other spacecraft.The performance of an attitude control engine is directly related to the lifetime,control precision and safety of a spacecraft.The study of flow characteristics of an engine transient process is important to improve its performance.In this paper,the transient flow test of a transparent test piece was carried out during the starting process of the attitude control engine.Then the transient process of the test piece was simulated and compared with the test results to verify the rationality of the simulation model.Transient flow simulation was carried out for the starting process of the real engine injector.The results show that the filling of the outer ring of the oxidant circuit is slower than that of the central collecting cavity,and the filling of the second layer of the outer ring is slower than that of the first layer.The filling process in the fuel path starts from the cooling hole near the inlet side and the fuel flows out in the circumferential direction.Installation direction has little influence on engine starting flow process in the ground state.The filling time of the engine in its vacuum state is longer than that in the ground state,the filling time of oxidizer is 31%longer than that in ground state,and the filling time of fuel is 57% longer than that in ground state.
Key words:attitude control engine,starting process,transient,flow characteristics
The liquid propellant attitude control rocket engines can be divided into monopropellant engines and bipropellant engines according to the number of components.The characteristics of the bipropellant engine are as follows: 1) It has the ability to start many times;2) Long working hours;3) Long service life in orbit;4) Fast response time.
Liquid bipropellant engines are widely used in satellites,manned spaceships,deep space probes and other spacecraft for attitude control.They have a wide range of applications in the aerospace field,with many varieties,large quantity and high requirements.The performance of the attitude control engine is directly related to the life,control precision and safety of the spacecraft.
A lot of research has been done on the steady-state characteristics of the engine at home and abroad,but research on the transient operating characteristics is not sufficient[1-3].Gauffre and Ansart[4]conducted an experimental and numerical study on the filling of the collecting cavity,comparing the experimental results,with the simulation results showing that the pressure in the collecting cavity going through a wide range of changes and fluctuations.Ruth and Ahn[5]used the transient model LRTC to simulate the engine of Titan launch vehicle,and the results showed that the simulation results in the early ignition period are different from experiment,because the LRTC approach is difficult to simulate the nonuniform fuel injection into the chamber caused by the complex geometry between the thrust chamber valve and the combustion chamber.LI and ZHANG[6]proposed to estimate the engine response time by measuring the filling time of each engine component.YANG and SHEN[7]analyzed the influence of operation timing on the response characteristics of multiple thrusters,and the results showed that a reasonable selection of starting interval time can effectively reduce the coupling effect of multiple thrusters and improve the quick response ability of the system during start process.TAO[8]proposed to use the liquid volume fraction to visually describe the filling process during the engine filling process,which complements the relevant phenomena in the flow diagram.ZHENG and WANG[9]carried out the simulation analysis of engine dynamic characteristics,which showed that the fault simulation results were in good agreement with the ground test fault results,but the established analysis model still had some shortcomings in accurately describing the real physical and chemical process.
In order to meet the demands of great agility and precise control for spacecraft,it is necessary to shorten the start-up time of the engine.Through experiments and simulation research,this paper identifies the transient flow characteristics of the engine,and explores the influence of installation direction,back pressure and other factors on the transient flow of the engine,which is of great significance to improve its response speed and control accuracy.
The parts of the transparent injector test piece were designed and a system test set for the study of the transient flow characteristics of the engine was established (see Figure 1),thus enabling the visualization of the filling process of the oxidizer path and the fuel path in the transparent injector test piece.
Figure 1 Transient flow test
As shown in Figure 2,the test piece of engine injector is composed of flange,distribution plate and injection core,forming a complex internal flow channel.
Figure 2 Test piece of sprayer
High speed cameras were used to observe the filling process of oxidant and fuel channels respectively,and the filling position of water in the injector channel at typical times was marked,which provided an experimental basis for subsequent simulation calculations.
The flow channel of the collecting cavity of the engine injector is annular.As the renormalization group k-εmodel is more suitable for calculating the flow characteristics in a bend than the standard k-εmodel,and can better simulate the transient flow[10],so the renormalization group k-εmodel was adopted for the turbulence model.
In the transient flow process of the engine,there is gas-liquid two-phase flow in the injector channel,and the interface between gas and liquid is clear.In this work,a three-dimensional numerical simulation based on the volume of fluid (VOF)was performed to investigate the transient multiphase flow.The VOF model tracks the interphase interface by solving the volume fraction continuity equation of a certain phase (or multiphase),so the equation for the volume fraction of the q phase can be expressed as Equation (1).
For multiphase flow,phase transition shows mass transfer between different phases.When the propellant evaporates or condenses,the phase transition model needs to be considered.The mass transfer in the evaporation condensation process is determined by the vapor transport equation,which is expressed as Equation (3).where subscriptsvandlare vapor phase and liquid phase respectively,andare the corresponding mass transfer rates during evaporation and condensation.
The calculation model includes the experimental calculation model and the real calculation model.The test calculation model is shown in Figure 3,which is used to simulate the flow characteristics of the transparent injector in the transient test.
Figure 3 Simulation calculation area and grid of test model
The real calculation model is shown in Figure 4,which is composed of the real oxidant channel,the actual fuel channel and the fluid region of the combustion chamber,and is used to simulate the transient flow characteristics of the real engine.
Figure 4 Fluid model simulation calculation area and grid
Figure 5 shows the comparison diagram of simulation and experiment at different times in the filling process of the fuel path.The simulation calculation shows that about 60% of the volume of the first layer of liquid collecting cavity outside is filled within 10 ms after startup,which is close to the experimental results,verifying the accuracy of the simulation calculation method.The simulation results of the oxidant filling process are also roughly similar to the test results,which are not be repeated here.
Figure 5 Simulation and test comparison of fuel path filling process at different times
The ground test of the engine is generally a horizontal installation,so the oxidant path and fuel path in the horizontal installation direction are simulated.Figure 6 shows a dynamic diagram of oxidant path filling for the engine injector.The oxidizer flowed out of the injection hole of the inner ring 1.1 ms after startup,and the first layer of liquid collecting ring and the liquid collecting cavity in the central area of the second layer were filled 1.6 ms after startup.The oxidizer began to flow out of the injection holes,in part of the outer ring,2.1 ms after startup,and oxidizer flowed out of all injection holes 3.1 ms after startup,while the upper part of the second liquid collecting ring of the outer ring near the inlet direction was still not filled with oxidizer.The filling of the engine injector was completed 3.6 ms after startup.The liquid collecting ring of the outer ring of the oxidant path was filled more slowly than the liquid collecting cavity in the central area,and the filling time of the second layer of the outer ring was about 2 ms,which is slower than the first layer of the outer ring.During the engine design process,the volume of the second layer of the outer ring of the oxidant path can be appropriately reduced,so that the filling time of the outer ring is shortened so the filling time of the upper and lower layers of the liquid collecting ring are close,so as to improve the engine response speed.
Figure 6 Dynamic diagram of oxidant filling at typical time
According to the traditional trickling method,the mass of oxidant required to fill the oxidant channel is 3.67 g.Assuming that all injection holes are filled at the same time,and the inlet flow is at the working rated value,the filling time of the oxidant channel is 9.8 ms.Through simulation calculation,the filling time of the oxidant path needs 3.6 ms,which is 63% shorter than that calculated by the trickle method.In the actual filling process,the inlet flow is not at the working rated value,so compared with the filling time obtained by the traditional trickling method there is a large error.
Figure 7 is a dynamic diagram of the fuel path filling of the engine injector showing typical time.Fuel has flowed out of the injection hole in the inner ring 1.4 ms after startup,and fuel began to flow out of the injection hole in part of the outer ring 2 ms after startup.The inner ring of the engine injector was filled 3.6 ms after startup,and the injector was filled 7.2 ms after startup.Comparing the filling states at 2 ms,2.2 ms and 2.6 ms after startup,it can be seen that the filling process of the cooling hole shows fuel flow out in a circumferential direction from the cooling hole near the inlet side.In the process of engine design,the volume of the inner and outer ring liquid collecting ring on the second layer of the fuel path can be appropriately reduced.The inner ring liquid collecting ring needs reduced volume at both ends of the slow filling,and the outer ring liquid collecting ring should have a gradual reduction of the volume at the corresponding position along the circumference from the inlet side,so that the fuel outflow time of the injection hole is relatively uniform hence the engine response time is improved.According to the traditional trickle method,the fuel mass required for filling the fuel passage was calculated as 2.576 g,and the filling time of the fuel path was 10.8 ms.The simulation result was 33% shorter than that of the trickle method.
Figure 7 Dynamic diagram of fuel path filling of engine injector at typical time
Figure 8 shows the velocity nephogram of oxidant circuit and fuel circuit at 4 ms and 8 ms after startup,at which time the air in the channel is displaced.The flow rate of oxidant at the bottom of the first layer of liquid collecting ring directly opposite to the inlet reduces to zero,and a reflux zone is formed around the “stagnation” zone and at the connecting channel between the first layer of liquid collecting ring and the second layer of central liquid collecting cavity,so the velocity in the reflux zone is low.The flow speed is high at the inlet of the injection hole.When the oxidant flows into the injection hole,due to the large flow around the corner and fast flow speed,flow separation occurs near the inlet of the injection hole.The flow velocity of the fuel at the bottom of the first liquid collecting ring directly opposite the inlet reduced to zero or stagnates,and a reflux zone was formed around the stagnation zone.The velocity and pressure in the reflux zone were low.Flow separation also occurred at the inlet of the fuel injection hole.
Figure 8 Velocity nephogram of oxidizer and fuel circuit for 4 ms and 8 ms after startup
Select an injection hole (A1) in the inner ring and an injection hole (D1) in the outer ring of the fuel path.Figure 9 shows the flow change curve at the outlet of A1 and D1.The flow change trend at hole A1 and hole D1 was similar,and the flow increased rapidly from zero after startup,and tended to be flat after filling.After the propellant enters the combustion chamber and reacts,the increase of back pressure causes the outlet flow to decrease.The flow tends to be flat 25 ms after startup,continued to decrease at about 35 ms after startup,and approached the rated flow 38 ms after startup.The flow change gap between A1 and D1 is related to the structure of the channel itself.The inner ring injection hole at A1 completed the filling faster,the initial flow increased faster,and the flow became stable faster,while the outer ring injection hole at D1 needed more time to complete the filling.
Figure 9 change curve of A1 and D1 outlet flow after startup
Under different installation directions of the engine,the flow difference will be impacted due to the action of gravity.Taking the oxidant flow channel as an example,the flow situation of the starting process under different installation directions was simulated.The filling time of the vertical upward installation state was 3.5 ms,while the horizontal installation state was 3.6 ms,and in the vertical downward installation state is was 3.5 ms.The filling time difference was not more than 0.1 ms.It can be seen that the installation direction in the ground state has little influence on the filling process for the engine starting process.
Figure 10 shows the dynamic diagram of oxidant filling at a typical time.The oxidant enters the first layer of the liquid collecting chamber 0.6 ms after startup,and the central area was filled 2.2 ms after startup.The outer ring of the second layer flows relatively slowly,and only one injection hole did not flow oxidant 4.4 ms after startup.The oxidant circuit was full of oxidizer 4.7 ms after startup.The completion time of oxidant filling in a vacuum state was 31% longer than that in the ground test state of 3.6 ms.Due to the phase change of some liquid oxidants to form gas,the content of liquid oxidants is reduced,and the completion time for filling is correspondingly longer.The simulation result of the filling time of the oxidant circuit in a vacuum state was still 9.8 ms less than that of the traditional dripping method.
Figure 11 shows the dynamic diagram of fuel path filling at typical times.The fuel enters the first layer of liquid collecting cavity 0.6 ms after startup,and reached the second layer of liquid collecting cavity at 1.9 ms.Fuel flowed out of the inner ring 3.5 ms after startup,and the inner ring was basically filled 6.3 ms after startup.10.7 ms after startup,the last cooling hole in the outer ring still did not show fuel flow.11.3 ms after startup,the fuel path was filled with liquid fuel,so the filling completion time was 57% longer than the 7.2 ms on the ground.The simulation result of filling time in a vacuum state was 10.8 ms higher than that of traditional dripping method,with a close difference of 0.5 ms.
Figure 10 Dynamic diagram of oxidant filling at typical time
Figure 11 Dynamic diagram of fuel path filling at typical time
Figure 12 shows the dynamic distribution of gaseous oxidant in the filling process.0.6 ms after startup,part of the liquid oxidant evaporated into gas in the reflux low-pressure area.There was a small amount of gaseous oxidant near the injector outlet,and the volume fraction of the maximum gaseous oxidant content was less than 0.01%.The central area was filled 2.2 ms after startup,and a small amount of gaseous oxidant was distributed in the combustion chamber,with a volume fraction of less than 0.4%.Filling was completed 4.7 ms after startup,and the content of gaseous oxidant in the combustion chamber increased,and the volume fraction was still less than 1%.
Figure 12 Dynamic distribution diagram of gaseous oxidant at typical time
Figure 13 Dynamic diagram of local distribution of gaseous and liquid oxidants 10 ms after startup
Observing the propellant return cavity near the injection hole of the fuel path.Figure 13 shows the local distribution dynamic diagram of gaseous and liquid oxidants.10 ms after startup,a small amount of gaseous oxidant accumulated near the injection hole of the fuel path,and some of it entered the liquid collecting cavity,with a volume fraction of about 1%.The gas in the fuel injection hole and the nearby liquid collecting cavity condensed,and the condensed oxidant was mainly concentrated at the outlet of the injection hole,with a maximum volume fraction of less than 2%.
In this paper,the transient flow characteristics of an engine starting process were studied,and the main conclusions are as follows:
1) Through the transient flow simulation of engine startup,the improvement direction for the injector channel is deduced.The oxidant circuit can appropriately reduce the volume of the second layer of the liquid collecting ring in the outer ring,while the fuel circuit can appropriately reduce the volume of the second layer of the liquid collecting ring in the inner and outer rings,the inner ring liquid collecting ring reduces the volume at both ends of the slow filling,and the outer ring liquid collecting ring gradually reduces the volume at the corresponding position along the circumference from the inlet side,hence that the filling time of the two layers of liquid collecting ring is optimized to be similar,so the propellant outflow time at the injection hole is relatively consistent,thus improving engine response time.
2) The influence of the installation direction on the engine start-up flow process was studied.Different installation directions have little influence on the start-up filling process,and the filling time difference was not more than 0.1 ms.
3) The starting flow process of the engine in a vacuum environment was studied.The oxidant filling time in a vacuum state was 31% longer than that in the ground state,and the fuel filling time was 57% longer than that in the ground test state.
4) Comparing the filling time of the engine obtained by the two calculation methods,the calculation result based on the VOF model was closer to the real process than that of trickle method.Under the ground test condition,the filling time of the oxidant path was 63%shorter than that calculated by drip method,and the filling time of fuel path was 33% shorter than that calculated by drip method;In a vacuum environment,the filling time of oxidant circuit was 41% shorter than that of trickle method,and the filling time of fuel circuit was 5% longer than that of trickle method.
5) The propellant phase change phenomenon caused by low back pressure during the engine starting flow in a vacuum environment was studied.The simulation results show that the phenomenon of vapor returning to the cavity during the starting process ointment the oxidant circuit.The content of vapor returning to the cavity was very small,and the volume fraction was about 1%.Some of the vapor returning to the cavity condensed,and the maximum volume fraction was less than 2%.
扩展阅读文章
推荐阅读文章
明翰范文网 www.tealighting.com
Copyright © 2016-2024 . 明翰范文网 版权所有
Powered by 明翰范文网 © All Rights Reserved. 备案号：浙ICP备16031184号-2