PhD Work– 12/09/07

09:27

RS simulations (A01, 2 & 3) are running on Altix & results seem to be ok. MT01-02 simulations are also still running on Solar and they seem ok too. Not sure quite what I should do for the moment. I could always look at the Re=100 results for MT02-01 but they’re at home. I will do that until lunch.

: 12 Sep 2007 09:32 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 11/09/07

09:55

Got in and got on. The MT01-02 jobs on Solar are still running so will keep an eye for them. The RS01 job has been set running on Altix again but I’m going to carefully stack those successive jobs now that I have got the code to run on.

The job now is to list the jobs to be done in the relevant tables. I will do this until 1100 then finish them after lunch.

: 11 Sep 2007 09:56 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 10/09/07

08:48

Got in 20 minutes late (forgot memory stick) so got to work.

I’ve officially started my final year now as of the 1st September and I have until the 31st August 2008 to get finished. However, my goal is to finish (I.e. submit) before my 27th birthday.

To do this, the goal is as follows:

Simulations
September 2007: RS01 (Lille bump without flow control)
October 2007: RS02 (Lille bump with a skewed, steady jet)
November 2007: RS03 (Lille bump with skewed, pulsed jet)
December 2007: RS04 (Lille bump with synthetic jet)
January-March 2008: Finish running of the above

Other simulations to be completed include:
- MT01-01 & -03: Flat plate validation of S-A model (inc. wall functions)
- MT01-02: Backward facing step validation of the DES formulation
- MT02: Circular cylinder validation of the Immersed Boundary Method
- FT01: Synthetic Jet simulations (I.e. from NASA CFDVAL2004)

These will be completed as and when I can get onto Solar.

On that, the resources to be used will be as follows:

Skua: RS Simulations (As this is the fastest)
Solar(when it works): MT Simulations
Firefly: FT simulations

I also intend to run a single, test simulation to test the speeds of each of these systems.

Writing
Unfortunately, there isn’t much to write until results start coming in and then I can add to the relevant chapters. As it stands:

Chapter1: PWC has checked this over and I simply need to apply changes
Chapter 2: ”
Chapter 3: Pretty much complete
Chapter 4: Finish Re=100 simulations (I.e. post process and write-up)
Chapter 5: Finish BF step simulations and write-up DES formulation
Chapter 6: Finish all SJ simulations and write-up
Chapter 7: Write-up as RS results finish
Chapter 8: Write at very end

I think that now, I will look at the necessary RS01 simulations and make a plan for getting these running.

15:37

Altix is down for maintenance at present so am working on other sims on Solar (which is actually working)

- MT01-02-A03 & B03 are running (I.e. BF step with fine grid)
- Need to get running MT01-02-C01 to C03 after this (I.e. BF step with 3D DES)
- Need to take a look at MT02-01 Re=100 case results, post-process & check they’re ok (Will do this tomorrow)
- Will head home now and get some FT01 simulations running (and list those needed)

After that, I will be visiting Wallace at GE and then finish that lecture for the ATC exped. training syllabus.

: 10 Sep 2007 09:07 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 06/09/07

10:03

I’ve been on holiday, I’ve pretty much wrapped up gliding for the season and so will am going to get back to work and triathlon training in preparation for the new term.

Here is a list of the things I need to do:

- File away all notes to date
- Read the paper on IBM for Yongmann
- List all simulations that need to be completed (I think I started this in May 07)
- Get bump simulation working fully including output of file
- Start getting a 3D bump simulation without flow control running on Altix (should take about a week)

: 06 Sep 2007 10:04 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 06/08/07

09:40

Apparently, Solar is down AGAIN so no dice there. What I will now do is go home and start making plans to run jobs on my home PC. It will take bloody ages but at least I will be getting closer to my goal.

The only other option is to use a CSC machine but serially if Yongmann can get clearance for me to do that. I will email him to discuss this.

11:03

Right, no joy with Solar so have emailed Yongmann Re. options. Until then, I am limited to using my own PC to at least finish my validation simulations. I will get on with this, this afternoon.

: 06 Aug 2007 09:41 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 16/07/07

08:40

Am back in after my holiday (although I will be aiming to go into SoAGC to catch up on the days I missed last week). I think the first thing to do, above all, is to sort out the plan of finishing. I aim to finish (at the latest) by the end of March 2008 (a good birthday present!) and so have 6 months left to prepare for the viva, complete corrections etc. This will also allow me to start a job at the beginning of the new academic year.

However, first I must go and see Terry as she’s sent some tuition work my way…

When I get back, I will get on with the planning and write it up here when I’m done.

17:11

This morning I finished the plan, met up with Peter and went shopping. I want to back-up all final results onto DVD but my main PC is playing up. Should be sorted this evening though so will do that then…

The PhD submission goal is 31st March 2008 and the plan for completion is as follows:

August 2007

• Complete all MT simulations
• Complete Chapters 4 & 5

September 2007

• Set-up and set running the RS01 simulations
• Complete Chapter 2

October 2007

• Set-up and set running the RS02 simulations
• Complete Chapters 1 & 3

November 2007

• Set-up and set running the RS03 simulations
• Complete Chapter 6

December 2007

• Set-up and set running the RS04 simulations
• Complete Chapter 7 (relevant sections)

January-March 2008

• Complete the simulations that were set running above
• Complete Chapter 7 as the relevant results are complete

: 16 Jul 2007 08:42 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 22/05/07

08:49

Have set running:

rs01-03-03: Full Re RANS sim but with CFL=0.5 to check stability
rs01-03-05: Halved dx to try and reduce instability
mt03-01-28: Have set running a new simulation to try and get stable. (Worth a try!)

Will also be setting up mt02-01-B01,2,3 & 4…

: 22 May 2007 08:50 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 18/05/07

11:04

Simulations currently running:

rs01-02-13: 3D bump using 2D DES initial conditions
rs01-03-02: Full Re using RANS. UNSTABLE at 10 so…
rs01-03-04: Same but using upwinding
rs01-03-03: Full Re (as 02) but using wall functions. Seems stable but will run for another 5

mt03-01-28: Running RANS sim with rescaling
mt03-01-29: Setting up 2D DES sim for rescaling later

11:13

In error, rs01-03-02 was already upwinding so have set rs01-03-04 to use Std CD. Will head home for lunch but will spend the rest of the time checking the simulations I’ve already done and rationalising where we’re going.

: 18 May 2007 11:07 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 14/05/07

08:29

PLAN FOR THE WEEK’S WORK

Simulations

• RS01: Lille Bump without flow control
  The primary goal is to get some reasonable full Re RANS results to compare with the work of Bernard (2003). I also need to get on with the 3D results which will include looking at the 2D DES simulation then extending to 3D.

• MT03: Turbulent Inflow Rescaling
  The subroutine is written and implemented and the simulation appears to be working ok but it does crash around 80 timesteps in. There appears to be an oscillating part at the edge of the boundary layer which probably seeds the instability. Will need to look into this

• RS02,03&04: Lille bump with flow control
  Maybe get on with writing the part in the ‘boundw3’ subroutine which will apply the jets to the flow.

• MT02-01: IBM Testing (Re=100)
  Carry on running this case until it has settled down into a nice periodic vortex shedding mode.

Writing
The results chapter could do with being updated as Lille results become available. To this end I should probably also get on with the subroutine which will assess the pressure at the boundary points and so work out the forces on the bump

Reading
Look for the recent work of Stanislas (in the last year or so in the AIAA Journal). Also, if there is no other reading to do, get into Peter’s book to build up some extra general fluids knowledge in preparation for the viva.

10:00

Have set running:

rs01-03-01: Needs to run ‘run10’ so have set running for that only
rs01-02-12: Have set a shell to run ‘run04’ and ‘run05’
mt02-01-B01: Have set running ‘run05’ only

20:13

Since this morning I’ve managed to solve the MT03-01 problem with the kink. The weighting function was being applied in reverse. At present, the following simulations are running.

• MT02-01: IBM Testing
  mt02-01-B01: As above

• MT03-01: TI Rescaling
  mt03-01-27: Turbulent inflow rescaling testing

• RS01-02: Lille bump without flow control
  rs01-02-12: Running on for 06-10 (2D DES)
  rs01-02-13: 3D version of 12

• RS01-03: Lille bump at full Reynolds number
  rs01-03-01: Running on for 11-15
  rs01-03-02: Same as 01 but with dx*2 so time saving of x4

Will head home now and spend the rest of the evening reading up from Peter’s book.

: 14 May 2007 08:39 |  Comments (0) |  Close comments |  Report a problem

PhD Work– 03/05/07

09:19

• MT03-01: Turbulent Inflow Rescaling

Have checked mt03-01-25c and it’s fine so have re-run mt03-01-26 using this. However, even before this mt03-01-26 was fine so these improved starting conditions should improve the issue. That can continue ticking away over the next few days and should be ready to go on Monday.

• RS01: Bump without flow control

rs01-02-11 seems to be completely settled down and a lot quicker than rs01-02-10 (that gives me a little something to write about once I’ve assessed the time savings)

I need to start getting the 3D simulations done. They will follow the convention below:

rs01-02-12: Without WF, Without IBM, DES, random flucts (do by taking rs01-02-11 and resetting the eddy viscosity)

Then, once this is finished:

rs01-02-13: With WF, Without IBM, DES, random flucts
rs01-02-14: Without WF, With IBM, DES, random flucts
rs01-02-15: With WF, With IBM, DES, random flucts
rs01-02-16: With WF, With IBM, DES, use turbulent inflow

: 03 May 2007 09:33 |  Comments (0) |  Close comments |  Report a problem