_________________________________ Corrected Notes from Lattice database meeting - Oct 28, 2004 Present: Valeri, Vaia, Mike Martens, Paul Lebrun, Nathan Eddy, Tim B., Jim Patrick, Elliott Mccrory, Jean Slaughter (We covered a lot of topics in this meeting and I've taken the liberty of trying to organize the information in a logical way. Purpose: How do we obtain and use the appropriate lattice parameters in the instrumentation front ends, OACs, supertable, OSDA, etc.? Is there commonality with the online lattice used by TOP, etc. Action Items: 1. Suggest a database schema for the coupled measured lattices ( Tim, Elliott) 2. Send information about the Top(ul_model) database (Mike Martens - done - see message at the end of these notes. 3. meet again in 2-3 weeks. Wish List: 1. Develop/revive the ability to update the TOP database lattice. Figure something out for its continued maintenance. Discussion: 1. The Tev lattice is really 16 lattices. 2. Reliable measurements of the Tev lattice Much progress has been made by Valeri and Sasha to produce reliable measurements of the lattice in the Tevatron. To be more precise, they fit the measured data to a model which can then give lattice parameters at any point in the Tevatron. Note that this model is coupled, but algorithms that can mimic an uncoupled approximation exist. The procedure for building/updating of the model from the measurements is time consuming. With current tools, 1-2 updates per year is about all they can do. The new BPMs and some new code being developed in conjunction with Argonne will likely make things easier. It's not the taking of the data that is slow, it's the analysis. Also, they determine one of the 16 Tev lattices, not the whole set. The process of analysis takes about 1-2 weeks. The term is "the measured coupled lattices". 3. Changes in the lattice The lattice changes for identifiable reasons, such as new magnets, new low beta optics, re-shimming. etc. Also there is strong evidence that the lattice drifts or changes with time. Quantifying the later won't be possible until the process in point 2 is speeded up. ( Don't we have some measure of this at a few points in the ring from K. Genser's work with the Flying wires and synclite?) 4. Lattice repository The lattice repository contains the design descriptions for the various machines. In the case of the Tev, there are several of these, for example current conditions, BTEV, etc. There are about 10,000 elements. Lattices can be produced from the descriptions. By definition these are uncoupled, but I would think they could be if coupling were introduced into the elements? The term for these lattices are "the design lattices" 5. ul_model library lattice database used by TOP. There is an outdated but still probably adequate for smoothing lattice description in SYBASE that TOP uses. Mike Martens has sent email which I attach below with the information. The system for updating this database is unmaintained and broken. (Elliott, Leo, Mike have all worked on it in the past). Everyone agreed that a maintainable system with a responsible entity is needed here, but is not the top priority as the current database seems adequate for smoothing. There is no provision for history -just one set of 16 lattices. The term is "ul_model lattice database" 6. BLASTMAN database Just for completeness and because it is probably relevant for the future and for naming conventions, The BLASTMAN database contains the ring description and magnetic characteristics of SC dipoles and quads. The link is http://odets3.fnal.gov/~dbox/php/ and is under "databases" on the AD home page. The term is "BLASTMAN database". 7. OSDAPhysics and Emittance OAC code Right now the lattice parameters for emittance calculations in OSDAPhysics are hardwired in the code. They need to be put in Constants.java, with the store dependence. beta* is in Constants.java and is 35 cm. One of the motivations for this discussion is to be able to go back and recalculate the supertable or do analyses using the best lattices after they are determined. 8. Proposed Scheme - what we drew on the board a. Ignore the TOP(ul_model) database for now, except some attention for future compatibility, like naming conventions, would be useful. b. a new SYBASE database(table(s)) for the "measured coupled lattices" that has history. (In defining this, we should be aware of the ul-model, BLASTMAN tables and naming conventions and even the lattice repository.) All The 16 lattices won't be measured at one time. However, the model interpolated from existing (and future) measurements at the injection and collisions will represent machine significantly better than present design model. Nevertheless it is a complication and it needs to considered. One could 1) keep the 16 separately and just update the one that is new. This will require some intelligence by the code that fetches the parameters to get a compatible set 2) Always make a complete set, even if only one is changed. The p and pbar lattices might are not be the same but the difference is comparatively small. Therefore presently we plan to have Twiss parameters at central orbit only. But means to introduce additional 2*16 lattices for both helices need to be anticipated. What locations should be included? We all agreed on all instrumentation including BPMs plus all dipole correctors. They have stable names and it well be easy to look for history of changes at these elements. It is unclear about what to do with the other elements. Although their inclusion increases the number of elements from ~400 to ~10,000 it is not the issue that will affect the choice. The major reason is that there is high probability that the number or names of other elements will be changed in the future, but the history of Twiss parameters updates is desired to be kept. c. An OAC that reads the database and keeps the current best parameters for SBD, Flying wire and synclite locations available for instrumentation front ends. This will need to use the coupled to uncoupled approximations. d. OSDAPhysics routines that can read the database and apply it intelligently. e. utility for filling the database tables f. Of course a big caveat here - people power and relative priority with other projects. 9. Future extensions 1. Mechanism for ul-model database can be filled from the measured database (and/or from the design lattices?) Message from Mike Martens on the TOP lattice database ------------------------------------------------------- Information about using the ul_model routines to access the Tevatron design lattice parameters can be found at: http://www-ap.fnal.gov/~martens/tev_lattice/ul_model.html More information on the ul_model routines can be found at: http://adcon.fnal.gov/userb/www/controls/user_libraries/ul_model/ul_mode l.html Examples for retrieving data directly from the database tables are given below. The relevant tables are named lattice_slot_to_name and lattice_twiss under the user "olm" in the ADBS database server. . Joins of these two tables need to be performed in order to extract data correctly. (There are also many other tables as part of the ul_model library, but they are not relevant. ------------------------------------------------------------------ Information on the lattice can be found directly from the database ------------------------------------------------------------------ The Tevatron lattices for Collider Run II are stored in a sybase table on the ADBS server. The name of the lattices are Tev01 through Tev15 for the 15 steps in the low beta squeeze. There is also a lattice Tev16 which is low beta lattice step 15 with the phase bumps for the beam halo collimators in place. 1) log on to a console 2) enter sql with $ isql/server=ADBS/user="olm"/pass="olmpass" If needed switch to the development data base 1> use olmdb 2> go 3) in SQL you can a) get names of horizontal bpms by 1> select element_name from lattice_slot_to_name 2> where lattice_name = 'Tev01' 3> and element_type = 'hmonitor' 4> go b) get beta functions at horizontal bpm locations 1> select t.element_name, d.betax, d.alphax, d.psix, d.dispx, d.dprimex 2> from lattice_slot_to_name t, lattice_twiss d 3> where t.slot_number = d.slot_number 4> and d.lattice_name = 'Tev01' 5> and t.lattice_name = 'Tev01' 6> and t.element_type = 'hmonitor' 7> go c) List the names of the lattices 1> select distinct lattice_name from lattice_slot_to_name 2> go _________________________________ -----Original Message----- From: jean slaughter [mailto:il1of12@yahoo.com] Sent: Monday, November 08, 2004 5:25 PM To: val@fnal.gov Subject: from jean Slaughter - lattice notes Hi Valeri, Unforunately, my email didn't show the red/blue colors that you used to indicate changes. I think I deduced them, and I've attached my corrected version of the notes from the lattice meeting from 2 weeks ago. Would you please check to be sure i did it correctly? If you make corrections could you please indicate with "*" or something? If you send to this email (il1of12@yahoo.com) I might get the colors. thanks - jean ===== Jean Slaughter Please reply to slaughter@fnal.gov __________________________________ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.com