CW beam and bunch Freq. (MHz)
-
FrancescoS - Initiated
- Posts: 30
- Joined: Thu 11 Feb 2021 16:45
- Country:
Czech Republic (cz)
Re: CW beam and bunch Freq. (MHz)
Post
by FrancescoS »
ok, so the frequency value here is not relevant, as long as the beam has the correct distribution, am I correct?
Re: CW beam and bunch Freq. (MHz)
When you have built a dst file, as the data is defined using the beam phase instead of z, the frequency from the file is no longer used, but the one defined in TraceWin is used.
-
FrancescoS
- Initiated
- Posts: 30
- Joined: Thu 11 Feb 2021 16:45
- Country:
Czech Republic (cz)
-
FrancescoS
- Initiated
- Posts: 30
- Joined: Thu 11 Feb 2021 16:45
- Country:
Czech Republic (cz)
Re: CW beam and bunch Freq. (MHz)
Post
by FrancescoS »
Goodmorning, I'm back here again!
:=)
I did the test again, now using the new dst that should be CW.
I have made 3 plots as a resume, where the beam parameters are shown how the change with the Bunch Frequency (I used, as before, 1, 33, 45 and 1000MHz, in the dst case I added 39MHz). In the plots x is the frequency axis in log scale and y is the beam parameter axis; according to the plot it can be emit[rms], alfa or beta.
In general, what I see looking at the plots, is that the beam parameters change value for every frequency and, in my opinion, it's not so straight forward to say which is the frequency associated with the correct calculation. It's only partially true that 1 MHz conresponds to uncorrect values: generally at this frequency the parameters are kind of different from what is calculated for 33 and bigger frequency values, but if we only look at the alpha twiss for the beam in the real space it's not like that.
The longitudinal dynamics is noisy for low frequency, that's true, but, as far as I can see, the noice is reduced by a factor of 2 from the 1MHz case to the 33-45MHz range case (the noise is +/-5e-5 in the first case and +/-2e-5 in the second). How can I say the first case is more noisy? I don't see a big difference and also che longitudinal phases plots do not visually help.
What is actaully looking as the best solution is the high frequency, but I need to perform the test with the parallel beam as you suggested.
Considering this missing test, I would be glad to know your comments.
To me, it's not so straight forward to decide which is the best value providing the most accurate result.
Again, I quickly tried to put some optics there (one or two quads), and I had the feeling the situation becames more complicated, but I honestly need to look at this case more carefully
Thanks
F.
:=)
I did the test again, now using the new dst that should be CW.
I have made 3 plots as a resume, where the beam parameters are shown how the change with the Bunch Frequency (I used, as before, 1, 33, 45 and 1000MHz, in the dst case I added 39MHz). In the plots x is the frequency axis in log scale and y is the beam parameter axis; according to the plot it can be emit[rms], alfa or beta.
In general, what I see looking at the plots, is that the beam parameters change value for every frequency and, in my opinion, it's not so straight forward to say which is the frequency associated with the correct calculation. It's only partially true that 1 MHz conresponds to uncorrect values: generally at this frequency the parameters are kind of different from what is calculated for 33 and bigger frequency values, but if we only look at the alpha twiss for the beam in the real space it's not like that.
The longitudinal dynamics is noisy for low frequency, that's true, but, as far as I can see, the noice is reduced by a factor of 2 from the 1MHz case to the 33-45MHz range case (the noise is +/-5e-5 in the first case and +/-2e-5 in the second). How can I say the first case is more noisy? I don't see a big difference and also che longitudinal phases plots do not visually help.
What is actaully looking as the best solution is the high frequency, but I need to perform the test with the parallel beam as you suggested.
Considering this missing test, I would be glad to know your comments.
To me, it's not so straight forward to decide which is the best value providing the most accurate result.
Again, I quickly tried to put some optics there (one or two quads), and I had the feeling the situation becames more complicated, but I honestly need to look at this case more carefully
Thanks
F.
- Attachments
-
- Test - H.pdf
- (972.87 KiB) Downloaded 73 times
Re: CW beam and bunch Freq. (MHz)
Dear Francesco,
Honestly, I have a lot of difficulty in seeing significant differences between the different frequencies except for the longitudinal noise.
I think Nicolas has already explained to you what is recommended for the frequency from a mesh point of view for the space-charge. So I must admit that I don't really understand why we should go and see what happens if we move away from these recommendations. Indeed the code will end up giving different or even incoherent results.
In addition, an essential parameter here is the number of space-charge kicks in tracking mode and it must be defined by meter for this type of comparison, is this the case in your example (you didn't show it) ?
Regards,
Didier
Honestly, I have a lot of difficulty in seeing significant differences between the different frequencies except for the longitudinal noise.
I think Nicolas has already explained to you what is recommended for the frequency from a mesh point of view for the space-charge. So I must admit that I don't really understand why we should go and see what happens if we move away from these recommendations. Indeed the code will end up giving different or even incoherent results.
In addition, an essential parameter here is the number of space-charge kicks in tracking mode and it must be defined by meter for this type of comparison, is this the case in your example (you didn't show it) ?
Regards,
Didier