PSPAC: Apply space charge effect on particles at SPCH elements defined in the lattice. Please ask Ohmi-san for the usage and parameter settings. Tracking only.

(to be continued)

As described in the manual, three flags, SPAC, WSPAC and PSPAC, which will take space charge into account when do tracking. In my preliminary understanding, SPAC + SELFCOD could calculate space charge based on a cylindrical symmetry chamber, referring to the current center of beam but not orbit given by emit. Meanwhile, "WSPAC" could consider space charge with assuming Gaussian distribution in the all dimensions, corresponding to number of particles in beam (PBUNCH). As for "PSPAC" use the PIC method to decide space charge ( which should be relative to the number of macro-particles?).

However, I truly confuse of "SPAC" on the space charge is simulated according to the number of particles in beam or number of macro-particles?

Many thanks!

At least '/bin/csh' in init.n should be replaced with a more popular shell. It will be done in the next update.

Following the guide of SAD home page, I installed k64 version of SAD into my desktop (Ubuntu 14.04 LTS).

Then, I try to run SAD to calculate Twiss parameters and Emit (of ATF damping ring). It agree well with old result. Unfortunately, there are two errors occur when I run SAD:

*** Today: 12:04:04 Wednesday 01/18/2017 ***
OFF LOG ECHO;READ 77 ; 23
*** main *** syntax error: invalid input for toplvl ""
*** main *** syntax error: invalid input for toplvl ""
*** main *** syntax error: invalid input for toplvl ""
*** SADScript Initialization: /home/root/soft/oldsad/Packages/init.n ***
sh: 1: /bin/csh: not found
*** Run time Environment: /home/root/soft/oldsad/Packages/init.local.n ***
RFSW RADCOD RAD FLUC INTRA POL COD DAPER EMIOU CMPLO FOURI SMEAR
F F F T F F T F F F F T

Besides, when I try to tracking beam in the ring, beam loss quickly after several tens of turns (only 1% of all the particles loss in the old simulation). I wonder there is something wrong with the installation or the due to the difference between old and new version (2015) of SAD. Therefore, I write to ask some help/guide. Thank you in advance!

Dear Users,

1. A bug was found in an optics calculation with CAVI in Linac. The parameter HARM was used for PHI. This was since V1.1k64.

Dear Users,

1. A bug was found in THIN elements SEXT, OCT, DECA, DODECA since V1.1. K value was not accepted.

Dear Users,

0. The changes related to CALC6D should correspond to the new version number 1.1 according to the original roadmap of 30 years ago.

1. A wrong usage of mkstemp creating unnecessary files has been corrected.

2. The flag RADTAPER no longer needs CODPLOT.

3. Closed orbit finding with RADTAPER has been revised.

4. Accessing keys of elements with an internal table kytbl has been changed to use parameters if possible.

5. Fitting value specification with two locations treats the value as the ratio for BX, BY, BZ, and as the difference for other functions:
FIT P1 P2; BX 1.5 : BX(P2) / BX(P1) == 1.5
FIT P1 P2; EX 0.2 : EX(P2) - EX(P1) == 0.2

6. A bug was created in around Sept. 2016, and fixed in temit.f. It might have affected the value of emittances.

Dear Users,

1. The next version V1.0.10.11k64-pre3 will be able to calculate 6D optics for a linac (i.e. the design momentum depends on s).

2. The MATRIX(MAT) command and the TransferMatrix function will return 6x6 matrix when CALC6D is ON. TransferMatrix takes a new option Symplectic, which returns a symplectic matrix for a linac. The default is Symplectic->False, which means a physical transfer matrix, taking the chenge of the design momentum due to s.

3. A bug in calculation of dispersions for CALC4D is corrected. This bug was introduced in V1.0.10.11k64-pre to make the results unpredictable in a case of linac.

Dear Users,

1. A new flag CALC6D has been defined to specify the optics calculation in CALC and GO as 6D. The antonym is CALC4D. The default is CALC4D. CALC and GO with CALC6D should work similar to CALC4D even for matching of Z-parameters. The off-momentum matching and finite amplitude mathcing are not yet implemented. Also CALC6D yet works for a linac.

2. An obsolete flag CANON was removed.

3. The closed orbit finding for EMIT and CALC/GO with CALC6D was improved, esp with RADCOD.

Dear Users,

1. A bug was found and fixed on handling the Context of SetCoupledElements. It should be System`, but some routines used ` .

Dear Users,

1. The next version will be able to calculate a full-6D optics parameters by EMIT/Emittance[] with the flag CODPLOT:

New functions: AZ, BZ, NZ, ZX, ZPX, ZY, ZPY

- These are compatible with Extended Twiss Parameters, printed by EMIT.
- ZX and ZY have no dimension. ZPX and ZPY are in 1/m.
- So far limited with a periodic condition. Not usable for a transport line yet.
- Thus usable with RADCOD, RADTAPER, etc.
- If x-z or y-z coupling are too strong, the calculation will fail.
- These functions are accessible by Twiss, DISP Z, OpticsPlot, DRAW.
- As CALC and GO are still in 4D, they clears the results by EMIT/Emittance[].
- No direct matching is possible yet. Use FitFunction with Emittance[], in such a way as FitFunction:=(Emittance;Twiss["BZ","$$$"]=20).
- TransferMatrix or the MAT coomand does not support 6D yet.

As this change might affects parts, please be careful.

2. A bug related to a keyword which ahas an alias, such as ROTATE of MULT, has been fixed (already in 1.0.10.10.4k64).

Dear Users,

1. Another bug was found in Parallelize to leave some shared memory unclosed and accumulated.

Dear Users,

1. An issue of closed orbit finding in Emittance[] is fixed. It stopped finding when encountered NaN during the search. Now it tries to continue the search.

Dear Users,

1. ListDensityPlot takes new options:
ColorScale (default:True) display a color scale on the right of the plot.
Smoothing (default: 1) The number of linear interpolation between data points.

2. The way to determine the plot area according to AspectRatio is changed to make it more continuous.

Dear Users,

1. The next version 1.0.10.10.5k64 will fix a bug to cause segmentation fault with @ or @- when the fit point is $$$. This bug was introduced by 1.0.10.10k64.

2. More fixes may come. Please wait for a while.

Dear Users,

1. A bug to mishandle F1/F2 in MULT for nonzero SK1 was fixed. It was introduced by 1.0.10.10k64.

2. A few internal relocations of modules.

3. A small simplification of memory allocation routines.

Dear Users,

1. More bugs are found, esp. kyindex, a reverse table of kytbl introduced on 1.0.10.10k64.

2. The handling of overlapped frame tick labels is improved.

> Dear Users,
>
> 1. The handling of F*K1* for MULT in SOL was incorrect in 10.10k64.
>
> 2. Tick labels of graphics Frame now avoids overlapping to other labels.
>
> 3. Element labels in OpticsPlot now avoids overlapping.

Dear Users,

1. The handling of F*K1* for MULT in SOL was incorrect in 10.10k64.

2. Tick labels of graphics Frame now avoids overlapping to other labels.

3. Element labels in OpticsPlot now avoids overlapping.

Dear Users,

0. The next version, V1.0.10.10k64 will be a major chage to contain the following. Any bugs or inconsistencies can happen. Please report them here or email to me.

1. Removed almost all 32-bit memory allocations, except for a few obsolete routines.

2. The expansion of a beam line in MAIN level is rewrtten using a new module with new types sad_el and sad_comp, and module routines loc_comp, loc_el, idcomp, dircomp, idcompc,idelc, idtypec, idvalc, direlc, compelc, pnamce, lpnamec.

3. Common blocks are largely replaced with modules.

4. Physical constants are updated with PDG2014 data in all .f and .c routines.

5. MULT with nonzero ANGLE was reviewed to be more consistent with BEND or QUAD. The Maxwellian fringes of K1 and higher are not yet implemented though.

6. New keywords F1K1F, F2K1F, F1K1B, F2K1B are introduced to describe asymmetric fringes at the entrance and the exit of QUAD and MULT. The traditional F1 and F2 are still valid: F1 + F1K1F(B) F1 + F2K1F(B) work at the entrance(exit) if the orientation of the element is positive, and vice versa when negative.

7. A routine to find out keywords of an element is renewed for efficiency. A new key KEYWORDS_ALL is added in Element to return all keywords including voids ('-').

8. The position of a legend in graphics now searches to avoid an overlapping with the contents of the graph. Also the size and frame are modified.

9. FindRoot is modified to handle a function that becomes invalid with non-real argument.

バージョンV1.0.10.10k64で違いは小さくなります。まだMaxwellian fringeの有無などの違いは残ります。

> > 仕様として、MULTはBM、QMを再現しているわけではないということを
> > 教えて頂きました。現在、B、Q、SXの３成分が全部入った、放射光の
> > ブースターを設計しようとする場合、BにK1を入れ、薄切りにして
> > 薄レンズ6極を挟むのが最も良いやり方とのことです。また、BのK1も
> > バグを含むことがあり、オプションk64で計算しないとチューンがずれる
> > ことがあるそうです。(/SAD/bin/gs -k64)
> >
> 補足
> * サジッタが存在するので、Rectangle BENDのスライス分割は簡単ではありません
> ** Sector BENDであれば、端部を除き s方向の並進対称性を持つので均質なスライスへ分割可能です
> ** SOLに囲まれた MULTスライスに Knを展開し、SOLの幾何境界条件で外部のビームラインに対してアライメントするというモデリングが可能ですが、境界条件の設定難易度が高めです
> * K1<>0のBENDについては、MAIN trunk/k64-1-6 branchともにエラッタが存在します
> ** ANGLE → 0や(K0 + ANGLE) → 0の極限にエラッタ有り
> * gsスクリプトの -k64オプションは、KEKBのSAD cluster計算機環境固有の実装です

> 仕様として、MULTはBM、QMを再現しているわけではないということを
> 教えて頂きました。現在、B、Q、SXの３成分が全部入った、放射光の
> ブースターを設計しようとする場合、BにK1を入れ、薄切りにして
> 薄レンズ6極を挟むのが最も良いやり方とのことです。また、BのK1も
> バグを含むことがあり、オプションk64で計算しないとチューンがずれる
> ことがあるそうです。(/SAD/bin/gs -k64)
>
補足
* サジッタが存在するので、Rectangle BENDのスライス分割は簡単ではありません
** Sector BENDであれば、端部を除き s方向の並進対称性を持つので均質なスライスへ分割可能です
** SOLに囲まれた MULTスライスに Knを展開し、SOLの幾何境界条件で外部のビームラインに対してアライメントするというモデリングが可能ですが、境界条件の設定難易度が高めです
* K1<>0のBENDについては、MAIN trunk/k64-1-6 branchともにエラッタが存在します
** ANGLE → 0や(K0 + ANGLE) → 0の極限にエラッタ有り
* gsスクリプトの -k64オプションは、KEKBのSAD cluster計算機環境固有の実装です

> 　テラタームでKEKショナイネットからショナイ側サーバafsadに入ろうとすると、
> plain passwordではFreeBSDに比べてsshの規格が古い？のが原因で
> 入れないので、秘密鍵を用意しようと試みたり（失敗）、
> kekb-login経由にしたり苦労していたのですが、ログインの時
> 「チャレンジレスポンス認証を使う（キーボードインタラクティブ）」を
> 選ぶと、ログインできます。（詳しい方に教えて頂きました。）　今まで
> ２年近く苦労していたのは何だったのか、という大発見でした……
>
TeraTermの仕様として、login時の認証手段は手動選択・フォールバク無しだったはずです
(OpenSSHのclient実装は、自動ネゴシエーション・フォールバック有り)

また、SSH client側が古い場合、認証・暗号ネゴシエーション時の初期パラメータが
既に失効している形式を含んでおり、接続に失敗するケースも報告されています

兎に角、clientをSSHプロトコル的に最新のものにしてください
(開発が停止しているSSH clientを使い使い続けるのは、運用互換性・セキュリティー的に問題です)

SuperKEKB Phase-II開始までに、afsad系は FreeBSD11に移行する予定であり、
そのタイミングで OpenSSH 7.x標準に従い、非推奨の認証オプション・古い公開鍵の
サポートが終了します

仕様として、MULTはBM、QMを再現しているわけではないということを
教えて頂きました。現在、B、Q、SXの３成分が全部入った、放射光の
ブースターを設計しようとする場合、BにK1を入れ、薄切りにして
薄レンズ6極を挟むのが最も良いやり方とのことです。また、BのK1も
バグを含むことがあり、オプションk64で計算しないとチューンがずれる
ことがあるそうです。(/SAD/bin/gs -k64)

　テラタームでKEKショナイネットからショナイ側サーバafsadに入ろうとすると、
plain passwordではFreeBSDに比べてsshの規格が古い？のが原因で
入れないので、秘密鍵を用意しようと試みたり（失敗）、
kekb-login経由にしたり苦労していたのですが、ログインの時
「チャレンジレスポンス認証を使う（キーボードインタラクティブ）」を
選ぶと、ログインできます。（詳しい方に教えて頂きました。）　今まで
２年近く苦労していたのは何だったのか、という大発見でした……