Microbunch instability simulation

[simona.bettoni]

Hello,
I am looking to the function described in the "Specialized tools for Elegant" smoothDist6. I have other functions which are in the same package, like astra2elegant, so I suppose I downloaded the right group of files. Where can I download it?
Thanks.

[soliday]

The actual name of the program is smoothDist6s. It is included with the Linux RPM packages. Which operating system are you using?

[simona.bettoni]

Thanks a lot for your answer: now I have it running under Linux.

[simona.bettoni]

Hello,
I would like to reduce the noise in the Elegant runs. I am presently using the LSC_LOW_FREQUENCY_CUT_OFF0 and LSC_LOW_FREQUENCY_CUT_OFF1, which I am putiing at 0.25 and 0.20 respectively. Is this wrong? Because when I am using them the simulation crashes saying:

"tracking 200000 particles
*** glibc detected *** elegant: realloc(): invalid pointer: 0x000000001f652070 *** ======= Backtrace: ========= [0x6b97dd] [0x5e4f91] [0x567334] [0x50b11b] [0x50d401] [0x47d9c9] [0x431b68] [0x48ad61] [0x697809] [0x4001e9] ======= Memory map: ========
00400000-007d2000 r-xp 00000000 00:16 2132544 /gpfs/homefelsim/bettoni_s/elegant/aps_oag/elegant
009d1000-00a04000 rw-p 003d1000 00:16 2132544 /gpfs/homefelsim/bettoni_s/elegant/aps_oag/elegant
00a04000-00a70000 rw-p 00a04000 00:00 0
1f55a000-2067b000 rw-p 1f55a000 00:00 0 [heap]
2b98acfab000-2b98ae05b000 rw-p 2b98acfab000 00:00 0
7fffb94c3000-7fffb94e4000 rw-p 7ffffffdd000 00:00 0 [stack]
7fffb95fc000-7fffb9600000 r-xp 7fffb95fc000 00:00 0 [vdso]
ffffffffff600000-ffffffffffe00000 ---p 00000000 00:00 0 [vsyscall]

Terminated by SIGABRT
Program trace-back:
"
Thanks for any help or suggestions,
Simona.

[michael_borland]

Simona,

Sorry about the crash.

The LSC_LOW_FREQUENCY_CUTOFF0 value should be less than the LSC_LOW_FREQUENCY_CUTOFF1.

Also, LSC_HIGH_FREQUENCY_CUTOFF0 and LSC_HIGH_FREQUENCY_CUTOFF1 are more important for controlling noise.

--Michael

[simona.bettoni]

Thanks, now it's ok.
Simona.

[simona.bettoni]

Hello,
for the first part of the injector I am using Opal. I have 50 millions lines file in the h5 format from this. I wrote a small Matlab script to read, convert and write them in a file (binary) to use them as the input distribution for Elegant. I don't know how to produce a file which is usable by Elegant. Does Elegant accept only a ascii file?
Thanks for your comment,
Simona.

[michael_borland]

Simona,

Elegant accepts either binary or ASCII SDDS files for the particle input. With 50 million particles, binary is strongly recommended as ASCII will be very slow. A description of the input columns is available under the documentation for the &sdds_beam command:
http://www.aps.anl.gov/Accelerator_Syst ... ode64.html

You can write SDDS files from MATLAB following the instructions here:
http://www.aps.anl.gov/Accelerator_Syst ... tlab.shtml

Another alternative is to use MATLAB to make a flat binary file, then convert it using plaindata2sdds. See
http://www.aps.anl.gov/Accelerator_Syst ... ode34.html

Finally, you can write the binary SDDS data by custom coding it in MATLAB. The binary file protocol is described here
http://www.aps.anl.gov/Accelerator_Syst ... ode10.html

With so many particles, I assume you are using the parallel version of elegant. If not, it is strongly recommended.

--Michael

[simona.bettoni]

Thanks for the very quick answer.
I have another question... I run a scan over the wavelength and the amplitude for several number of particles. I put a the low pass filter in the simulation (0.20 and 0.25 parameters). For the output did you use a smoothing? sddsdigfilter? Do you suggest something better?
Thanks,
Simona.

[michael_borland]

Simona,

To get the gain curve, I used the procedure described in http://prst-ab.aps.org/onecol/PRSTAB/v11/i3/e030701, page 8. I'm attaching my script, which might be helpful.

--Michael