[Hallb-engineering] Engineering Meeting Tomorrow -- Apr. 20

[A.M. Sandorfi]

Dave,

To be clear, when the router was reset and signals returned to IOCptarg, the
level in the IBC bath was 38% and the temperature of the nose was 5K and
rising. At that point the magnet was completely uncovered. Evidently, the
hour time without feedback was too much already. The rush of liquid into the
transfer line, with the subsequent boiling, blew the bath down further to
30%. The magnet quenched just as the level was turning around.

Andy




On 4/20/12 7:14 AM, "Dave Kashy" <kashy at jlab.org> wrote:

> When I setup pid loops in epics one critical parameter is the valve minimum
> position
> For liquid levels I always set it up so the flow is just enough to keep the
> u-tube etc cold.
> 
> So even on a loss of signal the u-tube maintains flow. The level will slowly
> fall but this will give some time to respond and allow much easier and
> smoother re-establishment of operation
> 
> I was not asked to look at the controls but suggest that someone implement
> this change.
> 
> There.may be more to it and I can go over it with someone if they call me on
> my cell phone 
> Dave
> 
> Sent from my iPhone
> 
> On Apr 19, 2012, at 2:09 PM, Saptarshi Mandal <mandal at jlab.org> wrote:
> 
>> Hi All,
>> 
>> We will have an engineering meeting at the regular time tomorrow. People
>> familiar with the control systems used in the hall should be present. Please
>> read the email from Andy below.
>> 
>> Saptarshi
>> 
>> -------- Original Message --------
>> Subject: Eng mtg tomorrow-re epics failure and IBC magnet quench
>> Date: Thu, 19 Apr 2012 12:31:03 -0400
>> From: A.M.  Sandorfi<sandorfi at jlab.org>
>> To:<miller at jlab.org>
>> CC: Krister Bruhwel<bruhwel at jlab.org>, burkert<burkert at jlab.org>,
>>    David Kashy<kashy at jlab.org>, Christopher Bass<cbass at jlab.org>,
>>     "mlowry at jlab.org"  <mlowry at jlab.org>
>> 
>> Hi Bob,
>> I gather Dave is away tomorrow and in such circumstances the Engineering
>> meeting is often cancelled. Volker has asked that we hold at least a
>> mini-meeting, particularly with the people familiar with the control systems
>> used in the hall, to discuss what can be done to avoid another event like
>> last night when the epics signals froze and the helium supply was cut to the
>> IBC. A brief history of last night's event is below.
>> 
>> I've copied Krister; please pass this on to whomever is appropriate.
>> 
>> Thanks,...Andy
>> 
>> 
>> 
>> 
>> _______________
>>  at about 17:25 EPICS communications stopped;
>> - epics alarm on alarm handler eventually went off, but probably not until
>> epics came back.
>> 
>> - the supply valve on the buffer dewar, EV8215, regulates LHe flow to
>> the IBC by looking at the liquid level in the cryostat. The last good value
>> read by epics happened to be high, so it began to close EV8215. Since the
>> level value remained constant, it continued to close until 18:45 when epics
>> signals returned.
>> - while the supply valve was closing, the level in the IBC LHe bath was
>> dropping, although the apparent epics value was a flat line at the last
>> 17:25 value.
>> - at 18:45 epics signals came alive again, the IBC LHe bath level
>> registered only 38%, and the EV8215 valve immediately opened. Since the
>> transfer line from the buffer dewar to the IBC had by now warmed, the LHe
>> flow from the opening of EV8215 boiled in the transfer line and the
>> expanding gas blew the IBC bath level down further, reaching 30% at 18:58.
>> - when epics came alive, the nose temperature was 5K and rising.
>> - by this time the solenoid magnet in the IBC nose, which was at 0.9 tesla
>> and 46 amps, was completely uncovered and no longer cooled. This caused a
>> quench, at about 18:55.
>> - mix1 =>  spiked to about 3 K
>> 
>> - polarizations dropped by about 40% !
>> 
>> 
>> <mandal.vcf>
>> _______________________________________________
>> Hallb-engineering mailing list
>> Hallb-engineering at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/hallb-engineering