[Miguelgh24]

Hi all,

I´m working in the design of a ramjet intake as my final thesis for my bachelor. The ramjet is going to work with mixed compression as it makes it more stable for different situations (Picture attached)

The problem that I´m facing is that I cannot get the normal shock formed. As you can see in the picture, the ramjet consists in a compression ramp, a constant section zone (isolator) and it ends with a divergent zone. I´m trying to get the normal shock wave in this last divergent section, so once the flow gets subsonic it can continue being decelerated ( working as a subsonic diffuser).

To get that normal shock in the divergent section, I must impose a backpressure at the outlet. As you can see in the picture attached, I´ve only used farfield pressure and outlet pressure boundaries. When I define pressure at the outlet, it is not considered when simulating, so the last section reaccelerate the flow (working as a supersonic nozzle).

I understand that in a supersonic flow, the outlet conditions have no influence upstream, but in this case, imposing backpressure is the only way to obtain the normal shock.

I´ve seen this same doubt in this and other forums, but I didn´t see any solution.These are the most relevant answers that I have seen:

1. In supersonic flow downstream properties are only defined by upstream flow. So imposing pressure at the outlet has no effect

Although it is not considered, the ramjet intake ends in a combustion chamber, so it must be a specific pressure in the outlet. It is also necessary to the normal shock formation.

2. Initialization importance. Once the freeflow simulation has converged, and the shock waves are established, impose the backpressure and simulate (without initializing)

I´ve tried this process and I get same results, it converge again, and shows the same as before.

3. Use pressure based solver and define the outlet as pressure outlet

I get same solution as using density based, so it has no effect. Furthermore, I thought density based solver is more suitable for supersonic simulations.

I haven´t seen seen any posts where the author gets the solution, however, I´ve seen different research articles with different ramjet geometries, achieving the normal shock with that back pressure imposing, so I know its possible. I´m new to CFD and Ansys Fluent (the software that I´m using), so maybe I´m making some other mistakes. Are the boundary conditions well suited?

I really appreciate all kind of help and discussion. It is my first post and I hope it is a good way of starting in this CFD world and forum.

Thank you,

Miguel

PD:

I´ve tried other geometries as a full external compression ramjet with constant section throat (picture attached), but neither could I get a neat normal shock.

[LuckyTran]

Use a pressure outlet.


What's your next question?

[Miguelgh24]

Originally Posted by LuckyTran

Use a pressure outlet.


What's your next question?

I am using pressure outlet, but changing the pressure has no effect and it is not considered. The outlet is determined by the upstream flow, so increasing pressure at the outlet has no effect. That's what I don't understand. I need that pressure to get the normal shock.

[JBeilke]

Your inlet on the left side should be defined as stagnation inlet.

[FMDenaro]

Originally Posted by Miguelgh24

Hi all,

I´m working in the design of a ramjet intake as my final thesis for my bachelor. The ramjet is going to work with mixed compression as it makes it more stable for different situations (Picture attached)

The problem that I´m facing is that I cannot get the normal shock formed. As you can see in the picture, the ramjet consists in a compression ramp, a constant section zone (isolator) and it ends with a divergent zone. I´m trying to get the normal shock wave in this last divergent section, so once the flow gets subsonic it can continue being decelerated ( working as a subsonic diffuser).

To get that normal shock in the divergent section, I must impose a backpressure at the outlet. As you can see in the picture attached, I´ve only used farfield pressure and outlet pressure boundaries. When I define pressure at the outlet, it is not considered when simulating, so the last section reaccelerate the flow (working as a supersonic nozzle).

I understand that in a supersonic flow, the outlet conditions have no influence upstream, but in this case, imposing backpressure is the only way to obtain the normal shock.

I´ve seen this same doubt in this and other forums, but I didn´t see any solution.These are the most relevant answers that I have seen:

1. In supersonic flow downstream properties are only defined by upstream flow. So imposing pressure at the outlet has no effect

Although it is not considered, the ramjet intake ends in a combustion chamber, so it must be a specific pressure in the outlet. It is also necessary to the normal shock formation.

2. Initialization importance. Once the freeflow simulation has converged, and the shock waves are established, impose the backpressure and simulate (without initializing)

I´ve tried this process and I get same results, it converge again, and shows the same as before.

3. Use pressure based solver and define the outlet as pressure outlet

I get same solution as before, so it has no effect.

I haven´t seen seen any posts where the author gets the solution, however, I´ve seen different research articles with different ramjet geometries, achieving the normal shock with that back pressure imposing, so I know its possible. I´m new to CFD and Ansys Fluent (the software that I´m using), so maybe I´m making some other mistakes. Are the boundary conditions well suited?

I really appreciate all kind of help and discussion. It is my first post and I hope it is a good way of starting in this CFD world and forum.

Thank you,

Miguel

PD:

I´ve tried other geometries as a full external compression ramjet with constant section throat (picture attached), but neither could I get a neat normal shock.

For such type of problem, I would use a density-based formulation. Clearly, for a supersonic outflow you have no effect in prescribing any Dirichlet BC, you must focus on the inlet condition. IF the inflow has fully supersonic conditions, you can prescribe the total enthalpy and the stagnation temperature T0. Then the velocity conditions and density.

The problem is different if you have subsonic inflow.

[Miguelgh24]

Originally Posted by JBeilke

Your inlet on the left side should be defined as stagnation inlet.

I´ve used it as velocity-inlet, and I get the same solution as pressure-farfield boundary. The outlet pressure is not considered, so there is no normal shock formed. Defining the inlet as pressure-inlet doesn´t make much sense, as the inlet velocity is known.

The inlet properties are:

Mach 3
Static Pressure 18800 Pa
Temperature 217K

[Miguelgh24]

Originally Posted by FMDenaro

For such type of problem, I would use a density-based formulation. Clearly, for a supersonic outflow you have no effect in prescribing any Dirichlet BC, you must focus on the inlet condition. IF the inflow has fully supersonic conditions, you can prescribe the total enthalpy and the stagnation temperature T0. Then the velocity conditions and density.

The problem is different if you have subsonic inflow.

Density based solver is used. The speed and altitude operation of the Ramjet are known, so I can define Temperature, Static Pressure, Mach number, Density.. at the inlet. However, what I can´t achieve is imposing pressure at the outlet. When I define the outlet as pressure outlet, the specified pressure is not considered.

The Ramjet operates at Mach 3, 18800 Pa and 217K (Pressure and Temperature defined by atmosphere altitude).

[FMDenaro]

Originally Posted by Miguelgh24

Density based solver is used. The speed and height operation of the Ramjet are known, so I can define Temperature, Static Pressure, Mach number, Density.. at the inlet. However, what I can´t achieve is imposing pressure at the outlet. When I define the outlet as pressure outlet, the specified pressure is not considered.

The Ramjet operates at Mach 3, 18800 Pa and 217K (Pressure and Temperature defined by atmosphere height).

You cannot prescribe pressure outlet (and any of the other variables) for supersonic outlet. What is the issue? This is what your flow problem requires.

[Miguelgh24]

Originally Posted by FMDenaro

You cannot prescribe pressure outlet (and any of the other variables) for supersonic outlet. What is the issue? This is what your flow problem requires.

Thats what I don´t understand. The only way to get a normal shock in a divergent section (which accelerates supersonic flow), is by imposing pressure at the outlet, otherwise the normal shock is not going to be formed.

I´ve seen different research articles that impose that backpressure, so they can get a normal shock in the duct.

I´ve also tried to get a normal shock in a constant section intake, but the shock wave reflections never end, and the normal shock is not formed.

[FMDenaro]

Originally Posted by Miguelgh24

Thats what I don´t understand. The only way to get a normal shock in a divergent section (which accelerates supersonic flow), is by imposing pressure at the outlet, otherwise the normal shock is not going to be formed.

I´ve seen different research articles that impose that backpressure, so they can get a normal shock in the duct.

I´ve also tried to get a normal shock in a constant section intake, but the shock wave reflections never end, and the normal shock is not formed.

But the presence of a normal shock in the divergent section will produce subsonic flows behind the shock and the outlet is, thus, subsonic. In such a case the pressure condition is well posed.

[Miguelgh24]

Originally Posted by FMDenaro

But the presence of a normal shock in the divergent section will produce subsonic flows behind the shock and the outlet is, thus, subsonic. In such a case the pressure condition is well posed.

Exactly, that´s what I´m trying to get. The normal shock will decelerate the flow to subsonic and pressure outlet boundary will be correctly defined.

It is not posible to get that normal shock without imposing pressure at the outlet. So its kind of a loop, I need the pressure to get the normal shock but I can´t define pressure outlet, until I get the normal shock.


At least, that´s what I understand.

[FMDenaro]

Originally Posted by Miguelgh24

Exactly, that´s what I´m trying to get. The normal shock will decelerate the flow to subsonic and pressure outlet boundary will be correctly defined.

It is not posible to get that normal shock without imposing pressure at the outlet. So its kind of a loop, I need the pressure to get the normal shock but I can´t define pressure outlet, until I get the normal shock.


At least, that´s what I understand.

Depending on the operating conditions at inflow, the shock can appear at the outlet section to satisfy the pressure condition.

Of course, the quasi one-dimensional theory is not exact for a two dimensional flow.

[JBeilke]

Originally Posted by Miguelgh24

I´ve used it as velocity-inlet, and I get the same solution as pressure-farfield boundary. The outlet pressure is not considered, so there is no normal shock formed. Defining the inlet as pressure-inlet doesn´t make much sense, as the inlet velocity is known.

The inlet properties are:

Mach 3
Static Pressure 18800 Pa
Temperature 217K

With this information known you can calculate the total pressure and total temperature and use a stagnation inlet.

[LuckyTran]

Originally Posted by Miguelgh24

Exactly, that´s what I´m trying to get. The normal shock will decelerate the flow to subsonic and pressure outlet boundary will be correctly defined.

It is not posible to get that normal shock without imposing pressure at the outlet. So its kind of a loop, I need the pressure to get the normal shock but I can´t define pressure outlet, until I get the normal shock.


At least, that´s what I understand.

What is the issue? Do you have some miracle disease that puts you into cardiac arrest if you use a pressure outlet? Just do it.

A pressure outlet doesn't affect the characteristics of the upstream supersonic flow but that doesn't mean that it doesn't affect the flow. Consider what happens when you set the pressure outlet to be 10000 bar for example... And then consider what happens if you set the outlet pressure to 0 Pa. Imagine what happens... It will not always be ignored. You still must specify the correct static pressure condition since you expect the flow to eventually become subsonic after the normal shock.


If you don't get the flow you expect then either the value of the outlet pressure is wrong or your inlet BC is wrong.

[Adithya Ajay]

Originally Posted by Miguelgh24

Hi all,

I´m working in the design of a ramjet intake as my final thesis for my bachelor. The ramjet is going to work with mixed compression as it makes it more stable for different situations (Picture attached)

The problem that I´m facing is that I cannot get the normal shock formed. As you can see in the picture, the ramjet consists in a compression ramp, a constant section zone (isolator) and it ends with a divergent zone. I´m trying to get the normal shock wave in this last divergent section, so once the flow gets subsonic it can continue being decelerated ( working as a subsonic diffuser).

To get that normal shock in the divergent section, I must impose a backpressure at the outlet. As you can see in the picture attached, I´ve only used farfield pressure and outlet pressure boundaries. When I define pressure at the outlet, it is not considered when simulating, so the last section reaccelerate the flow (working as a supersonic nozzle).

I understand that in a supersonic flow, the outlet conditions have no influence upstream, but in this case, imposing backpressure is the only way to obtain the normal shock.

I´ve seen this same doubt in this and other forums, but I didn´t see any solution.These are the most relevant answers that I have seen:

1. In supersonic flow downstream properties are only defined by upstream flow. So imposing pressure at the outlet has no effect

Although it is not considered, the ramjet intake ends in a combustion chamber, so it must be a specific pressure in the outlet. It is also necessary to the normal shock formation.

2. Initialization importance. Once the freeflow simulation has converged, and the shock waves are established, impose the backpressure and simulate (without initializing)

I´ve tried this process and I get same results, it converge again, and shows the same as before.

3. Use pressure based solver and define the outlet as pressure outlet

I get same solution as using density based, so it has no effect. Furthermore, I thought density based solver is more suitable for supersonic simulations.

I haven´t seen seen any posts where the author gets the solution, however, I´ve seen different research articles with different ramjet geometries, achieving the normal shock with that back pressure imposing, so I know its possible. I´m new to CFD and Ansys Fluent (the software that I´m using), so maybe I´m making some other mistakes. Are the boundary conditions well suited?

I really appreciate all kind of help and discussion. It is my first post and I hope it is a good way of starting in this CFD world and forum.

Thank you,

Miguel

PD:

I´ve tried other geometries as a full external compression ramjet with constant section throat (picture attached), but neither could I get a neat normal shock.

1. Instead of farfield conditions, try Pressure inlet and at outlet pressure outlet
2. Experimentally how they impose the back pressure is by using a cone like structure at the exit to choke the flow . Cone can be moved along the aix left and right to control the pressure. You can make such a structure at the exit. I have seen some researchers simulating like this. This way you can avoid defining back pressure.

[sahebbedi]

Are you using a steady state solver or transient sovler ?
Try using pseduo-time method under steady state and get an initial solution first with 1-3 bar and then amp the backpressure up.