Internal:Metric Adaptive CDT3D: Difference between revisions

From crtc.cs.odu.edu
Jump to navigation Jump to search
← Older edit
Ctsolakis (talk | contribs)
BrainMapping, CNF, HPC, I2MConversion, NDAM, Segmentation, UltraSound3D, Bureaucrats, external, internal, Administrators
2,152 edits
Ctsolakis (talk | contribs)
BrainMapping, CNF, HPC, I2MConversion, NDAM, Segmentation, UltraSound3D, Bureaucrats, external, internal, Administrators
2,152 edits
 
(169 intermediate revisions by 2 users not shown)
Line 1: Line 1:
= Delta Wing  (Scitech 2018) =
<span id="Contents"></span>
= Data =
[[Internal:Adaptive CDT3D Results | Adaptive CDT3D Results ]]


[[File:Delta_wing_50k_parallel_scitech.png| 500px]]
= Integration with SU2 =
* SU2 + refine + CDT3D pipeline [http://www.cs.odu.edu/crtc/SU2/SU2_pipeline.svg]
* SU2 + refine + CDT3D, ONERAM6 invicid [http://www.cs.odu.edu/crtc/SU2/onera_invicid.pdf]


==50k Complexity ==
* SU2 + refine + CDT3D, steady laminar flow low mach number over delta wing
===Starting from a coarse surface mesh  ===
** meshes [http://www.cs.odu.edu/crtc/SU2/delta_wing.pdf]
{| class="wikitable" style="text-align: right;"
** solution cuts of the vortex [http://www.cs.odu.edu/crtc/SU2/delta_vortex.pdf]
|+ All time data are given in seconds.
**  video [http://www.cs.odu.edu/crtc/SU2/delta_wing_005.ogv]
! rowspan="2" | # Threads
! rowspan="2" | # Tetrahedra
! rowspan="2" | # Vertices
! rowspan="2" | Intitialize Bg Mesh
! rowspan="2" | Boundary Adaptation (MMGS)
! rowspan="2" | Boundary Recovery
! colspan="6" style="text-align: center;" | Mesh Refinement
! rowspan="2" | Total Time
(without MMGS)
! rowspan="2" | Memory (GB)
|-
! # Iterations
! Par. Deactivation
! Par. Vertex Creation
! Vertex Insertion
! Partition
! Par. Reconnection
|-
| 1
| 357,165
| 62,177
| 1.44   
| 6.32   
| 0.36   
| 20   
| 2.09   
| 11.92 
| 0.09   
| 1.26   
| 57.35 
| 80.97 
| 0.44
|-
| 6
| 361,037
| 62,795
| 1.49   
| 6.23   
| 0.35   
| 21   
| 0.51   
| 2.78   
| 0.11   
| 1.45   
| 14.06 
| 27.12 
| 0.63
|-
| 12
| 363,278                                                                                                                             
| 63,198                                                                                                                               
| 1.41                                                                                                                                 
| 6.27                                                                                                                                 
| 0.33                                                                                                                                 
| 22                                                                                                                                 
| 0.32                                                                                                                                 
| 1.55                                                                                                                                 
| 0.11                                                                                                                                                                                                                                                                   
| 1.62                                                                                                                                 
| 8.24                                                                                                                                 
| 19.99                                                                                                                                 
| 0.86
|-
| 24
| 363,756
| 63,270
| 1.49   
| 6.26   
| 0.33   
| 22   
| 0.2   
| 0.88   
| 0.11       
| 1.6   
| 5.13   
| 16.15 
| 1.31
|}


=== Notes 1===  
= Improvements =
* according to Mike's email mesh should be around 100k vertices
[[Internal:Adaptive CDT3D - Smoothing | Metric Adaptive Smoothing ]]
* Metric Input Data available at [https://github.com/UGAWG/parallel-adapt-results/tree/master/delta-wing/fun3d-fv-lp2 https://github.com/UGAWG/parallel-adapt-results ]
* Input surface is the coase mesh available at [https://github.com/UGAWG/solution-adapt-cases/delta-wing/initial-grid/delta01.meshb https://github.com/UGAWG/solution-adapt-cases]
* The experiments performed on a DELL workstation with Linux Red Hat Enterprise, 24 hardware cores Intel® Xeon® CPU E5-2697v2@2.70 GHz, and 757 GB RAM
* Memory is measured at the end of Mesh refinement and thus it may reach a slightly higher value before the point of measurement
* parameters <code> --brec=2 --met=1 --meval=3 --mrecm2=1 --mrec4=1  --nadapt=2 --malign=0 --mpp=2  --ndfs=0 --nthreads=$THREADS --nbuckets=20*$THREADS --nthreadsaf=20*$THREADS --nbucketsaf=20*${THREADS} --blockmul=500000 --nqual=0 --nminpts=30 --ordt=0 --pardeact=1 --mdse=0  </code>
* using commit <code> bb7819619eb9b8831ac1d4515b9605869df72b47</code> or tag <code> Delta-wiki </code>
 
==500k Complexity ==
=== Starting from a 50K Complexity volume mesh ===
Size of intial mesh
Vertices : 104,305
Tetrahedra: 581,196
{| class="wikitable" style="text-align: right;"
|+ All time data are given in seconds.
! rowspan="2" | # Threads
! rowspan="2" | # Tetrahedra
! rowspan="2" | # Vertices
! rowspan="2" | Intitialize Bg Mesh
! rowspan="2" | Reconstruct Mesh
! colspan="6" style="text-align: center;" | Mesh Refinement
! rowspan="2" | Total Time
! rowspan="2" | Memory (GB)
|-
! # Iterations
! Par. Deactivation
! Par. Vertex Creation
! Vertex Insertion
! Partition
! Par. Reconnection
|-
| 1
| 3,079,286
| 521,152
| 1.18
| 2.02
| 31
| 34.66
| 138.72
| 0.82
| 26.6
| 1115.09
| 1320.18
| 1.34
|-
| 6
| 3,124,862
| 528,775
| 1.48
| 2.03
| 34
| 9.44
| 36.18
| 0.82
| 33.69
| 184.91
| 269.65
| 1.35
|-
| 12
| 3,136,586
| 530,696
| 1.15
| 2.03
| 32
| 4.92
| 19.09
| 0.82
| 33.1
| 101.82
| 164.04
| 1.42
|-
| 24
| 3,147,672
| 532,467
| 1.17
| 2.03
| 30
| 2.63
| 10.34
| 1.03
| 37.51
| 56.45
| 112.26
| 1.76
|}
 
This approach offers no boundary adaptation.
 
===Starting for a coarse surface mesh===
{| class="wikitable" style="text-align: right;"
|+ All time data are given in seconds.
! rowspan="2" | # Threads
! rowspan="2" | # Tetrahedra
! rowspan="2" | # Vertices
! rowspan="2" | Intitialize Bg Mesh
! rowspan="2" | Boundary Adaptation (MMGS)
! rowspan="2" | Boundary Recovery
! colspan="6" style="text-align: center;" | Mesh Refinement
! rowspan="2" | Total Time
(without MMGS)
! rowspan="2" | Memory (GB)
|-
! # Iterations
! Par. Deactivation
! Par. Vertex Creation
! Vertex Insertion
! Partition
! Par. Reconnection
|-
|1
| 3,301,987 
| 564,041 
| 1.3       
| 29.83             
| 6.39       
| 26       
| 26.59     
| 112.14     
| 1.0               
| 19.41     
| 858.58     
| 1,055.77   
| 1.34       
|-   
| 6       
| 3,327,519
| 568,583 
| 1.48       
| 29.64             
| 5.3       
| 28       
| 7.37       
| 28.32     
| 1.08             
| 26.08     
| 165.42     
| 265.23     
| 1.25       
|- 
| 12         
| 3,333,780 
| 569,607 
| 1.45       
| 30.8             
| 6.87       
| 29     
| 4.18       
| 16.91     
| 1.23             
| 31.54     
| 95.14     
| 188.78     
| 1.41       
|-
| 24         
| 3,333,430 
| 569,592 
| 1.33       
| 29.93             
| 5.96       
| 30       
| 2.47       
| 9.33       
| 0.99               
| 28.92     
| 53.59     
| 133.08
| 1.75
|}   
===Notes 2===
* '''TODO compare quality of two methods'''
* according to Mike's email mesh should be around 1M vertices
* in some cases boundary recovery reached a point of non-developed code. Rerunning created a different surface mesh and CDT3D was able to continue.
* commit <code> 912b8d5ec5b171ebf20bc6a61b9f7fe572f3d3d5 </code>
* parameters <code> --met=1 --meval=3 --mrecm2=1 --mrec4=1  --nadapt=1 --reconstruct=1 --malign=0 --mpp=2  --qstatr=1 --outmesh=${NAME}.meshb --ndfs=0 --nthreads=$THREADS --nbuckets=20*NTHREADS --nthreadsaf=20*NTHREADS --nbucketsaf=20*NTHREADS --blockmul=500000 --nqual=0 --nminpts=30 --ordt=0 --pardeact=1 --mdse=0 </code>
 
==2M Complexity ==
=== Starting from a 50K Complexity volume mesh ===
Size of intial mesh
Vertices : 104,305
Tetrahedra: 581,196
{| class="wikitable" style="text-align: right;"
|+ All time data are given in seconds.
! rowspan="2" | # Threads
! rowspan="2" | # Tetrahedra
! rowspan="2" | # Vertices
! rowspan="2" | Intitialize Bg Mesh
! rowspan="2" | Reconstruct Mesh
! colspan="6" style="text-align: center;" | Mesh Refinement
! rowspan="2" | Total Time
! rowspan="2" | Memory (GB)
|-
! # Iterations
! Par. Deactivation
! Par. Vertex Creation
! Vertex Insertion
! Partition
! Par. Reconnection
|-
| 1
| 12,765,609
| 2,146,330
| 1.19
| 2.07
| 27
| 123.99
| 555.07
| 3.98
| 90.67
| 5,101.49
| 5,879.56
| 4.27
|-
| 6
| 12,890,958
| 2,167,610
| 1.20
| 2.02
| 35
| 39.51
| 143.19
| 4.13
| 149.45
| 962.48
| 1,303.11
| 3.94
|-
| 12
| 12,894,271
| 2,168,319
| 1.29
| 2.06
| 32
| 19.59
| 73.35
| 3.93
| 137.36
| 474.90
| 713.58
| 3.95
|-
| 24
| 12,896,370
| 2,168,800
| 1.15
| 2.06
| 36
| 11.83
| 47.80
| 4.05
| 162.56
| 263.62
| 494.48
| 4.02
|}


= General Notes =
= General Notes =
* smoothing : [https://onlinelibrary.wiley.com/doi/abs/10.1002/%28SICI%291097-0207%2819971115%2940%3A21%3C3979%3A%3AAID-NME251%3E3.0.CO%3B2-9 Tetrahedral Mesh Improvement Using Swapping and Smoothing, Lori A. Freitag and Carl Ollivier-Gooch ] [http://www.cs.odu.edu/crtc/internal/Christos/Freitag_et_al-1997-International_Journal_for_Numerical_Methods_in_Engineering.pdf PDF]
* [[Internal:ACDT3D_Quality_Criteria | Comparison for quality criteria for ACDT3D]]
* [https://epubs.siam.org/doi/abs/10.1137/0715011 An Efficient Method to Solve the Minmax Problem Directly C. Charalambous and A. R. Conn ] [http://www.cs.odu.edu/crtc/internal/Christos/Charalambous_An_Efficient_Method_To_Solve_The_Minmax_problem_directly.pdf PDF]
* [[Internal:ACDT3D_(Scitech2018)_stability_data | Stability Data for 500k]]
 
[[#Contents | Back To Top]]
= Stability for 500k case=
== Comparison of Quality Metric using different amount of cores and surface adaptation ==
[[File:Histogram_mean_adapt.png|800px]]
[[File:Histogram_ratio_adapt.png|800px]]
== Comparison of Quality Metric using different amount of cores and mesh reconstruction ==
[[File:Histogram_mean_reconstruct.png|800px]]
[[File:Histogram_ratio_reconstruct.png|800px]]
 
==Comparison Surface Adaptation vs Mesh Reconstruction (reproducibility) ==
[[File:Histogram_mean_compare.png|800px]]
[[File:Histogram_ratio_compare.png|800px]]
 
Surface adaptation offer better edge lengths but mesh reconstruction slightly better mean ratio quality

Latest revision as of 12:42, 22 November 2018

Data

Adaptive CDT3D Results

Integration with SU2

  • SU2 + refine + CDT3D pipeline [1]
  • SU2 + refine + CDT3D, ONERAM6 invicid [2]
  • SU2 + refine + CDT3D, steady laminar flow low mach number over delta wing
    • meshes [3]
    • solution cuts of the vortex [4]
    • video [5]

Improvements

Metric Adaptive Smoothing

General Notes

Back To Top

Retrieved from "https://crtc.cs.odu.edu/index.php?title=Internal:Metric_Adaptive_CDT3D&oldid=2590"