Comparison of hydrocodes in Wengen test4

A collection of Wengen test4 results

E-PPM	Enzo, 1024×1024×104 (eff.), PPM with 1 level of refinement
E	Enzo, 1024×1024×104 (eff.), Zeus-mode with 2 levels of refinement
E-M2	Enzo, 1024×1024×104 (eff.), Zeus-mode with 2 levels of refinement, +2% mass
E-M5	Enzo, 1024×1024×104 (eff.), Zeus-mode with 2 levels of refinement, +5% mass
E-DP	Enzo, 1024×1024×104 (eff.), Zeus-mode, double precision
E-M2-HR	Enzo, 2048×2048×208 (eff.), Zeus-mode with 3 levels of refinement, +2% mass
F	FLASH, base resolution (J ≥ 8)
F-HR	FLASH, high resolution (J ≥ 12)
F-LR	FLASH, low resolution (J ≥ 4)
F-Z1-LR-M10	FLASH, low resolution (J ≥ 4), cube cells (dx = dy = dz, +10% mass
F-M2	FLASH, base resolution (J ≥ 8), +2% mass
F-LR-M10	FLASH, low resolution (J ≥ 4), +10% mass
F-Z1-X64	FLASH, cube cells (dx = dy = dz), dx ≥ 1/64 (initial conditions at best)
F-O-Z1-X128	FLASH, cube cells (dx = dy = dz), dx ≥ 1/128, primary rests at the origin
F-O-HR-Z1-X128	FLASH, cube cells (dx = dy = dz), dx ≥ 1/128, primary rests at the origin, high resolution (J ≥ 12)
F-O-HR-Z1-X128-M2	FLASH, cube cells (dx = dy = dz), dx ≥ 1/128, primary rests at the origin, high resolution (J ≥ 12), +2% mass
F-O	FLASH, initial conditions shifted, primary rests at the origin (FFT)
F-O-NEW	FLASH, initial conditions shifted, primary rests at the origin (tipgrid)
F-O-A	FLASH, initial conditions shifted, primary rests at the origin (interp2cart)
F-B2	FLASH, computational domain doubled in size
F-M2-HR	FLASH, high resolution (J ≥ 12), +2% mass
F-C2	FLASH, CFL number = 0.2
F-HR-I256	FLASH, high resolution (J ≥ 12), low resolution initial conditions
F-CV	FLASH, artificial viscosity on
G	Gasoline, base resolution (200k particles)
G-HR	Gasoline, high resolution (1M particles)
G-M2-HR	Gasoline, +2% mass, high resolution (1M particles)
G-M2	Gasoline, +2% mass (200k particles)
C	CHYMERA, base resolution (256×1024×52) 23 Jan 2008
C-HR	CHYMERA, high resolution (256×2048×52) 05 Feb 2008
C-M2	CHYMERA, +2% mass, base resolution (256×1024×52) 11 Feb 2008
C-Z2	CHYMERA, base resolution in (r, φ), doubled z-resolution (256×2048×104), 13 Feb 2008
C-LR-OLD	IU code, test resolution (256×512)
C-OLD2	IU code, base resolution (256×1024), old

Movies

Code	surface density	FFT of surface density	frames	time span
C	x-y map (3.3MB)	r-m map (2.9MB)	61	0.00 – 6.28
C-HR	x-y map (6.5MB)	r-m map (5.8MB)	120	0.03 – 5.59
C-LR-OLD	x-y map (4.3MB)	r-m map (3.8MB)	83	0.08 – 7.24
C-M2	x-y map (6.0MB)	r-m map (5.6MB)	100	0.07 – 10.23
C-OLD2	x-y map (6.7MB)	r-m map (5.9MB)	125	0.02 – 5.87
C-Z2	x-y map (5.8MB)	r-m map (5.1MB)	100	0.07 – 7.52
E	x-y map (996KB)	r-m map (848KB)	18	0.10 – 9.00
E-DP	x-y map (224KB)	r-m map (192KB)	4	0.50 – 3.00
E-M2	x-y map (1.2MB)	r-m map (988KB)	20	0.00 – 10.00
E-M2-HR	x-y map (620KB)	r-m map (520KB)	10	1.00 – 10.00
E-M5	x-y map (644KB)	r-m map (552KB)	11	0.00 – 10.00
E-PPM	x-y map (544KB)	r-m map (456KB)	9	1.00 – 5.00
F-B2	x-y map (3.4MB)	r-m map (2.9MB)	61	0.00 – 6.00
F-C2	x-y map (3.3MB)	r-m map (2.8MB)	61	0.00 – 6.00
F	x-y map (3.3MB)	r-m map (2.8MB)	61	0.00 – 6.00
F-CV	x-y map (3.3MB)	r-m map (2.8MB)	61	0.00 – 6.00
F-HR	x-y map (5.8MB)	r-m map (5.0MB)	101	0.00 – 10.00
F-HR-I256	x-y map (3.6MB)	r-m map (3.1MB)	61	0.00 – 6.00
F-LR	x-y map (3.0MB)	r-m map (2.5MB)	61	0.00 – 6.00
F-LR-M10	x-y map (3.3MB)	r-m map (2.9MB)	61	0.00 – 6.00
F-M2	x-y map (3.4MB)	r-m map (3.0MB)	61	0.00 – 6.00
F-M2-HR	x-y map (3.6MB)	r-m map (3.1MB)	61	0.00 – 6.00
F-O-A	x-y map (3.5MB)	r-m map (3.0MB)	61	0.00 – 6.00
F-O	x-y map (5.8MB)	r-m map (5.0MB)	101	0.00 – 10.00
F-O-HR-Z1-X128	x-y map (3.5MB)	r-m map (3.0MB)	60	0.00 – 6.00
F-O-HR-Z1-X128-M2	x-y map (3.7MB)	r-m map (3.2MB)	61	0.00 – 6.00
F-O-NEW	x-y map (3.5MB)	r-m map (3.0MB)	61	0.00 – 6.00
F-O-Z1-X128	x-y map (3.3MB)	r-m map (2.8MB)	61	0.00 – 6.00
F-Z1-LR-M10	x-y map (3.0MB)	r-m map (2.6MB)	61	0.00 – 6.00
F-Z1-X64	x-y map (14MB)	r-m map (12MB)	240	0.00 – 23.90
G	x-y map (2.5MB)	r-m map (2.2MB)	48	0.00 – 6.00
G-HR	x-y map (572KB)	r-m map (488KB)	11	0.00 – 10.00
G-M2	x-y map (2.4MB)	r-m map (2.2MB)	43	0.00 – 10.00
G-M2-HR	x-y map (756KB)	r-m map (660KB)	14	0.00 – 6.50

All movies are MPEG-1, set at 5 fps, consisting of "I" frames only so XAnim should be able to play them frame-by-frame. MPlayer also could be useful.

Images

Code	surface density map, t =										FFT of surface density, t =										Code
C	0.93	1.96	2.92	4.00	5.02	6.01					0.93	1.96	2.92	4.00	5.02	6.01					C
C-HR	0.97	2.01	3.02	4.01	4.99	5.59					0.97	2.01	3.02	4.01	4.99	5.59					C-HR
C-LR-OLD	0.94	1.95	3.00	4.06	5.02	5.99	6.94				0.94	1.95	3.00	4.06	5.02	5.99	6.94				C-LR-OLD
C-M2	1.02	1.97	2.91	3.95	4.99	6.04	6.98	8.02	9.01	10.00	1.02	1.97	2.91	3.95	4.99	6.04	6.98	8.02	9.01	10.00	C-M2
C-OLD2	1.01	1.98	3.01	3.98	5.02	5.87					1.01	1.98	3.01	3.98	5.02	5.87					C-OLD2
C-Z2	1.04	2.00	3.01	3.97	5.02	6.02	6.97	7.52			1.04	2.00	3.01	3.97	5.02	6.02	6.97	7.52			C-Z2
E	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00		1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00		E
E-DP	1.00	2.00	3.00								1.00	2.00	3.00								E-DP
E-M2	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	E-M2
E-M2-HR	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	E-M2-HR
E-M5	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	E-M5
E-PPM	1.00	2.00	3.00	4.00	5.00						1.00	2.00	3.00	4.00	5.00						E-PPM
F-B2	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-B2
F-C2	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-C2
F	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F
F-CV	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-CV
F-HR	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	F-HR
F-HR-I256	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-HR-I256
F-LR	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-LR
F-LR-M10	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-LR-M10
F-M2	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-M2
F-M2-HR	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-M2-HR
F-O-A	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-O-A
F-O	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	F-O
F-O-HR-Z1-X128	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-O-HR-Z1-X128
F-O-HR-Z1-X128-M2	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-O-HR-Z1-X128-M2
F-O-NEW	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-O-NEW
F-O-Z1-X128	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-O-Z1-X128
F-Z1-LR-M10	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					F-Z1-LR-M10
F-Z1-X64	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	F-Z1-X64
G	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					G
G-HR	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	G-HR
G-M2	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	1.00	2.00	3.00	4.00	5.00	6.00	7.00	8.00	9.00	10.00	G-M2
G-M2-HR	1.00	2.00	3.00	4.00	5.00	6.00					1.00	2.00	3.00	4.00	5.00	6.00					G-M2-HR

Code	FFT integrated over r			FFT integrated over r, averged over t					Code
C	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	C
C-HR	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	C-HR
C-LR-OLD	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	C-LR-OLD
C-M2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	C-M2
C-OLD2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	C-OLD2
C-Z2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	C-Z2
E	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	E
E-DP	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	E-DP
E-M2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	E-M2
E-M2-HR	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	E-M2-HR
E-M5	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	E-M5
E-PPM	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	E-PPM
F-B2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-B2
F-C2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-C2
F	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F
F-CV	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-CV
F-HR	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-HR
F-HR-I256	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-HR-I256
F-LR	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-LR
F-LR-M10	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-LR-M10
F-M2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-M2
F-M2-HR	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-M2-HR
F-O-A	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-O-A
F-O	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-O
F-O-HR-Z1-X128	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-O-HR-Z1-X128
F-O-HR-Z1-X128-M2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-O-HR-Z1-X128-M2
F-O-NEW	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-O-NEW
F-O-Z1-X128	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-O-Z1-X128
F-Z1-LR-M10	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-Z1-LR-M10
F-Z1-X64	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	F-Z1-X64
G	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	G
G-HR	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	G-HR
G-M2	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	G-M2
G-M2-HR	linear m	log m	A_m(t)	1 ≤ t < 2	2 ≤ t < 3	3 ≤ t < 4	4 ≤ t < 5	1 ≤ t ≤ 4	G-M2-HR

(the time labels above are in test4 units)

Other plots

Peak plot for some variations of "F" run.
Peak plot for some "F-*" and "G-*" runs.
Clump position and density profiles.

Comments

Base FLASH run – Model "F" uses J ≥ 8 as a refinement criterion, with halved z-cellspacing (dx = dy = 2dz), with best resolution dx = 1/256. Model "F-HR" differ only in threshold for refinement (J ≥ 12). Model "F-LR" uses dx = dy = dz and J ≥ 4 (marginally meets Truelove’s requirements)
Model "L" is supposed to have primary moving freely, but still requires some care.
There are a few more "F-*" models, I’ll add the most interesting ones.
Spikes for m divisible by 4 on "F-LR-*" FFT integrated over r images are due to poor resolution of the grid.
Data on pictures is now expressed in more classical units, like AU, M_☉ or g/cm².
Model "C-M2-ALT" uses alternative differencing scheme but it is practically indistinguishable from "C-M2".
Model "F-Z1-X64" uses J ≥ 8 as a refinement criterion, but does not go beyond resolution of IC file. This ensures that the whole disc is well-resolved all the time, but resolution at peaks may drop below the threshold. GI activity is severely suppresed and clumps never form. Highest peaks have reached density = 2.55 (in code units) which correspond to J = 5.5, still better than Truelove's condition. In outer parts of the disc this model has comparable resolution to the "C" model.
Primary’s rest-frame model "F-O" does not differ much visually from the "F" model, except that it is slightly offset i y-direction. Some differences are clearly visible in spectral representations, especially less power of m = 1 mode in "F-O" model. In both cases polar sampling was done with respect the origin of coordinate system, which means that it is slightly differently situated with respect to the disc in both cases. I haven’t yet checked whether it accounts for all the difference.
NEW: All postprocessing was done in a coordinate system centered on primary’s initial position and moving with its initial velocity. Thus there should be no differences in spectra due to different choice of the origin of coordinate system and m = 1 mode on spectral pictures is not that much artificially enhanced.

Initial conditions

A 512×512×52 ASCII file test4-512 (194MB)
To read it, the innermost loop has to go through x coordinate, outer ones are for y and z. The outermost loop should read density first, then x-velocity, y-velocity, z-velocity and internal energy. Data is cell-centered and grid spacing is exactly 1/64 (in code units). At t = 0 the primary’s position is (-0.00190265, 0.0379506, 0.00083884) and its velocity is (0.00172268, 0.00178423, -0.0000620918) (in code units).
An alternative version (194MB)
It is organized in the same way as the original, except that it was resampled in primary’s rest frame at t = 0, so:
- all data was shifted in space so the primary’s position corresponds to the origin.
- all velocity components were corrected to primary’s initial velocity.
Note that this is not a rest frame for the disc (and thus star + disc system). Resampling was done using tipgrid tool, i.e. the same way as the original IC file was created, except that input SPH particlle set was shifted as described above. The above file was tested with FLASH code, in a setup related to the "F" model. The results closely match the original "F" run. Movement of the primary in this coordinate frame is realy small (<0.003 AU), so in case of problems in implementation of a moving primary, the above file can be used as an easier alternative.

Problem units (or code units) vs. physical units

mass unit = 0.1 M_☉
length unit = 6.25 AU
Gravity constant G = 1 ⇒ time unit = 7.864 yr = 2.5^7/2 / π yr

Last update: Sat Mar 21 12:03:14 CET 2009

Alternative page at AstroSim Wiki.

All .png images in one tarball

Comments, questions? Mail me .