Scramjet Intake Analysis

akshaydeshpande · October 24, 2014, 16:35

Hello everyone,
I am doing a 3d scramjet analysis on ansys fluent but am not obtaining the desired shock pattern after ~12000 iterations. I have used the following conditions:
-Density based solver with density of air mentioned as "ideal gas."
-Energy equation-on; k-omega SST turbulence model.
-The operating pressure is 0 Pa.
-The boundary conditions I have used are:
1. Inlet: Pressure far field - Supersonic gauge pressure: 12350 Pa; M=3.2; Turbulent intensity and viscosity ratio=2%, 10.
2. Top, Bottom: Wall; No-slip conditions.
3. I have not modified the outlet b.c since fluent extrapolates the outlet conditions as the simulation proceeds, from the inlet.
-Solution methods: Implicit Roe-FDS; Discretization: Second order (started initially from first order).
-I have reduced the under relaxation factors initially and went on increasing gradually as the solution started to become stable.
-All residuals: 1e-05; Surface monitor: Mass flow rate at the outlet.
-To capture the shocks, I have followed the method as per the ANSYS tutorial: Grid Adaption using pressure gradient; Coarsen threshold:0.3; Refine threshold:0.7; Max no of Cells:20000.
I have attached the geometry, mesh (with mesh quality histogram) and some of the results (at the mid plane) after the analysis. As observed in the pressure contours (Fig 4), the pressure should gradually increase from the inlet instead of decreasing while in the velocity contours (Fig 3), the velocity should gradually decrease instead of increasing. Also, no oblique shocks are observed.
Please suggest any changes or improvements required.

vineeth_tom · November 18, 2014, 07:32

Check with M=7, you'll get it.....and use T=5 % and length ratio...and instead of giving pressure inlet ,outlet,,give pressure far-field for the outlet part as well as you dont need the exit things..try ,you'll get the result..this is because the inlet would slow down m=3.5 to subsonic speeds and ur starting mach no is obviously not 3.5...and what is that pointed extension..not that clear...

akshaydeshpande · January 21, 2015, 01:20

Sorry for a belated reply! I discussed this issue with my professors and they told me that my domain definition is incorrect. I have been working on their suggestions and will post as soon as I get the result.

vineeth_tom · January 21, 2015, 06:17

i am working on scramjets too........can u help me with 3d meshing prefarably in hypermesh or any other meshing software for the domain if i give u a pic.??Mapped or manual meshing ,,,not adaptive

akshaydeshpande · January 27, 2015, 05:04

I myself am learning ICEM CFD for structured meshing. But you can send me your domain. I'll try my best.

October 24, 2014, 16:35	Scramjet Intake Analysis	#1
akshaydeshpande New Member Akshay Deshpande Join Date: Sep 2014 Posts: 19 Rep Power: 12	Hello everyone, I am doing a 3d scramjet analysis on ansys fluent but am not obtaining the desired shock pattern after ~12000 iterations. I have used the following conditions: -Density based solver with density of air mentioned as "ideal gas." -Energy equation-on; k-omega SST turbulence model. -The operating pressure is 0 Pa. -The boundary conditions I have used are: 1. Inlet: Pressure far field - Supersonic gauge pressure: 12350 Pa; M=3.2; Turbulent intensity and viscosity ratio=2%, 10. 2. Top, Bottom: Wall; No-slip conditions. 3. I have not modified the outlet b.c since fluent extrapolates the outlet conditions as the simulation proceeds, from the inlet. -Solution methods: Implicit Roe-FDS; Discretization: Second order (started initially from first order). -I have reduced the under relaxation factors initially and went on increasing gradually as the solution started to become stable. -All residuals: 1e-05; Surface monitor: Mass flow rate at the outlet. -To capture the shocks, I have followed the method as per the ANSYS tutorial: Grid Adaption using pressure gradient; Coarsen threshold:0.3; Refine threshold:0.7; Max no of Cells:20000. I have attached the geometry, mesh (with mesh quality histogram) and some of the results (at the mid plane) after the analysis. As observed in the pressure contours (Fig 4), the pressure should gradually increase from the inlet instead of decreasing while in the velocity contours (Fig 3), the velocity should gradually decrease instead of increasing. Also, no oblique shocks are observed. Please suggest any changes or improvements required.

November 18, 2014, 07:32	:)`	#2
vineeth_tom New Member vineeth tom Join Date: Sep 2014 Posts: 7 Rep Power: 12	Check with M=7, you'll get it.....and use T=5 % and length ratio...and instead of giving pressure inlet ,outlet,,give pressure far-field for the outlet part as well as you dont need the exit things..try ,you'll get the result..this is because the inlet would slow down m=3.5 to subsonic speeds and ur starting mach no is obviously not 3.5...and what is that pointed extension..not that clear... Last edited by vineeth_tom; November 18, 2014 at 11:33.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to implement a sensitivity analysis for certain items	Mason liu	CFX	10	August 17, 2017 09:46
analysis of wall injector based scramjet combustion	samironsaikia	FLUENT	1	April 18, 2011 01:34
Intake manifold analysis	alonzoo	STAR-CCM+	2	July 6, 2010 05:00
Intake manifold analysis	mehndan	Siemens	3	August 22, 2008 05:24
Intake manifold analysis	mehndan	Main CFD Forum	6	July 3, 2008 06:28

January 21, 2015, 01:20		#3
akshaydeshpande New Member Akshay Deshpande Join Date: Sep 2014 Posts: 19 Rep Power: 12	Sorry for a belated reply! I discussed this issue with my professors and they told me that my domain definition is incorrect. I have been working on their suggestions and will post as soon as I get the result.

January 21, 2015, 06:17		#4
vineeth_tom New Member vineeth tom Join Date: Sep 2014 Posts: 7 Rep Power: 12	i am working on scramjets too........can u help me with 3d meshing prefarably in hypermesh or any other meshing software for the domain if i give u a pic.??Mapped or manual meshing ,,,not adaptive

January 27, 2015, 05:04		#5
akshaydeshpande New Member Akshay Deshpande Join Date: Sep 2014 Posts: 19 Rep Power: 12	I myself am learning ICEM CFD for structured meshing. But you can send me your domain. I'll try my best.