private:teleconsnotes20180627 [Probe Mission Study Wiki]

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Telecon Notes June 27, 2018  ======

Attendance: SH, Jamie, Jacques, Charles, Amy, Al

Notes by: 

=== Agenda ===
  * Signals and Foregrounds Figure (Jacques, Josquin)
    * {{:private:sensitivity_vs_frequency_jun27th_2018.pdf|Option 1}}
    * {{:private:sensitivity_vs_frequency_jun27th_2018_largest_scales_2_10.pdf|Option 2}}
  * Review of TeamX slides (Al, Jamie, Bill, Shaul)
  * Status of thermal model (Amy)
  * Decadal Panel Science, Projects, and complementarity papers (Shaul, Julian)

=== Notes ===

__Signals and Noise Figure__ \\
  * Paradigm: keep figure simple such that it can relay zero order information to the broad astrophysics public
  * Add r=1e-4
  * Use just one synchrotron line; Charles will provide model parameters
  * Change vertical label to RMS polarization intensity
  * We are not adding CO Lines

__Decadal Panel Papers__
  * Julian + SH report on plans to have near term science papers, and later project + complementarity paper.

__Review of TeamX Slides__
  * Comments from Al, Bill, Jamie, SH listed below. 
  * Certain commonalities detected among all comments. SH collected commonalities. They are listed below, including some of Amy's responses. 

=== Compilation of Common Comments (including SH) ===
 
  * Discussion of I+T
    * __Amy's response__: No I+T during instrument session. No I+T person during mission session. Amy met with him/her after session, but was too late to change slides. Issue known, to be fixed next round. **Shaul will get Bill to describe I+T again.** 
  * Detectors and Focal Plane 
    * additional options (mu-MUX, KIDS) - confusing, not useful; SH: need to separate the message to the decadal "need to fund technologies", which will be included in the report, from the message to SOMA, which needs to be simple and clean. 
    * "Design Weakness: TES not operated in space", misleading
    * Cosmic ray mitigation - confusing, incomplete; SH: slides do not capture the legwork we have done to address the issue.
    * __Amy's response__: Day's first TeamX. Acknowledges that current set of slides need to be improved along the lines described above.
  * Discussion of Cryocooler power (Kogut): Accounting of margin upon margin is silly and technically not feasible. 
    * __Amy's response__: Agree, but the experience at JPL is that this is what the cost estimators expect. 
    * __SH comment__: Agree that this is what was discussed at JPL, but I am not sure that the point was (at the time) that accounting this way violates physical principles.  
  * Discussion of cryocooler heritage (Shaul): (one mention of Sumitomo option: what is LB assuming? What did CORE assume?)
    * __Amy's response__ : 
  * Super-expensive data collection and compressing computer ($12.5M) and motor controller in CDS: cost seems far too excessive
    * __Amy's response__ : Will be costed on the instrument side and should come out cheaper. SH points out that we expect these costs to 'come out cheaper' by factors of >10, not ~2, 3. 

=== Discussion of Other Items ===
  * Jamie has reservations about the spinning platform. Response: All detector and cooler electronics are on the spinning part, not stationary. Amy acknowledges that this was not captured properly in the slides. 
  * There will be heat dissipation on the rotating part. Yes, radiators need to be added; already in the work package.
  * Deployment of TM antenna removed. Gimbaled, permanently deployed antenna at center. 

----

__Shaul's Comments__

  * Replace all references from 'mirrors' (or most with 'telescope') to 'reflectors'

  * Instrument Slides
    * slide 7: total instrument mass is vastly overestimated. (see e-mails)
    * slide 9: ADCS comment vague, not clear, open to interpretation of 'inconclusivity'
    * slide 11: Decon should mimic Planck. I sent an e-mail to Jan Tauber
    * slide 15: fix spin rate. Make comments regarding Planck only to the extent they support heritage; otherwise they imply conflicts.
    * slide 16: replace with SH corrected figure
    * slide 19: focal plane mass needs to be updated. UMN working on focal plane model; 
    * slide 19: correct 1K filter mass from 10 kg to 1 kg CBE. 
    * slide 19: correct 18K box, and v-groove mass
    * slide 20: correct ADR power values
    * slide 32: 70% cost is $80,874. Is this inconsistent with slide 31 $87,823?
    * slide 33: given the importance of the cost of cooling we need to augment the words about the Sumitomo cooler with more evidence. Shaul will supply references, but someone may need to do some research on cost. 
    * slide 59: masses need to be revised
    * slide 62: comments about cross-pol are about how to handle systematic effects in data analysis and calibration of the instrument. They are not part of the technical implementation of the focal plane. 
    * Slide 62: the comment about cosmic rays implies that pico won't work. We have done a lot of work to show that we have a good handle on this. This should show up better. 
    * slide 64: requirement is *not* 3xCBE. We are planning for 1.5xCBE
    * slide 66: what table 8? where is table 8?
    * slide 67: revise mass estimate. Quote data rates more accurately. We know them. 
    * slide 69, 71: too many options for focal plane. Confusing and not useful. 
    * slide 72: clean up the message and combine with comment on slide 62. 
    * slide 79: wrong on many levels. 
    * slide 80: sometime we call it MIRI, sometime NGAS. Call it the same name throughout. 
    * slide 82: why no reference to flown Sumitomo coolers?
    * slide 83: what is POD? 


  * Mission Slides
    * slide 12: what is the reference to "2020 PICO Instrument .."
    * slide 56: isn't control supposed to be 1'? It say 0.01 deg = 0.6'. Change to 0.02? 
    * slide 56: no flight spares for flywheels. Is that commensurate with Class B? 
    * slide 80: first bullet 'Large telescope'. Why 'large'? It isn't large and the word large implies technical challenges. 
    * slide 81: second bullet implies additional cost that haven't yet been included. 
    * slide 83: not *large* telescope. 
    * slide 88: there is a comment on 'mass savings'. See my earlier comments.
    * slide 88: comment on potential mass upper invites questions about cost estimate.
    * slide 92: comment on mass savings that can reduce cost. See my earlier comments. 
    * slide 126: the costs quoted are absolutely ridiculous relative to the tasks needed to be carried out, both for the compression computer and for the spin motor control box. By the way, do we need a redundant box? I'd say 'yes'. Is that factored in? 
    * slide 127: all of SVIT slides are entirely opaque.
    * slide 162: cost drivers: what is ATLO? 

----


__Kogut Comments (June 25)__ 

Hi Shaul,

I went through the TeamX instrument and mission reports.  I agreed with most of it, but found some problems and some areas that need clarification to avoid mis-interpretation.  Starting with the big items, we have:

1) Discussion of I&T is inadequate for a cryogenic instrument.  The instrument report estimates $6.1M for I&T  at 50% confidence (slides 31 & 32) but gives no detail.  Is this a wrap?  Is the percentage for I&T in family with previous cryogenic missions?    The mission report says $1.9M for observatory-level integration and test (slide 131) and shows only 20 days for observatory thermal vacuum testing.  Both are gross under-estimates. 60 days for observatory thermal vacuum testing would be more appropriate. Adding some discussion of chamber mods for sub-K instrument would also be a good idea.

2) The estimated heat loads for the cryogenic stages seem low to me. The instrument report slide 81 says 3.8 mW CBE for the load at 4.5 K. Does this include the heat rejected from colder stages by the ADR? I note that PIXIE has much larger heat leak at 4.5K despite a much smaller cold volume.   There's also a discrepancy between the heat loads reported in slides 79 and 81 of the instrument report: Slide 79 shows 14 uW load on the 100 mK stage while slide 81 says 4.8 uW.
PIXIE carried 20 uW at 100 mK, again with a much smaller cold volume.
Do we have any info on the estimates from LiteBIRD or CORE?
It would help to show PICO estimates as in-family with other mission
concepts.

3) The instrument report was not clear about how the cryocoolers reject
heat.  The compressors need to be mounted on the rotating stage,
and take a fair amount of power (339W CBE), but I couldn't find any mention
about radiators for the cooler.  Is the cooler radiator in the outermost
V-groove?

4) I don't see the value in including detector design option 2 (mixed
TES / MKID focal plane) in the TeamX instrument report.  MKIDs are
considerably  lower TRL and there's no real discussion of the (limited)
benefits.  If we can show we meet the science requirements within the
cost cap using the baseline design, I would leave it at that and not open
potential cans of worm for lower-TRL designs.

5) I don't see any clear description of the various observatory modes.
In particular, what is safe mode?  The instrument report slide 53 discusses
using thrusters to slew during safe mode.  Why do we want to slew during
safe mode?  Keeping the spin seems reasonable, but I don't see any
justification for slewing...

6) The mission wraps for systems engineering (mission report slide 36, WBS 2)
seem a little light, particularly for a cryogenic mission.  It's cited
at $17.0M, or 4% of the total hardware costs in WBS 5+6.  5% would be more
appropriate.  Granted, it's only a $2M difference, but why skimp here?

7) There needs to be a better discussion of the contingency for cryo-cooler
power.  A given cooler design (e.g. a MIRI cooler) has a fixed max power.
You can't run one at 110%.  The margin and contingency for the cooler
are kept on the thermal side: typically want 100% at this early stage.
Thus, the cooler will show 100%  margin/contingency vs the expected thermal
load, but a fixed max power with no added contingency beyond the contingency
already built in to account for the thermal load.  This typically need
to be spelled out to the cost folks.  The instrument report slide 83
spells out 100% margin in thermal load PLUS an additional 70% contingency
in power. This is too conservative, so perhaps we can use this to cut costs
(in the cost model, anyway).

8) PICO plans to field a large cryogenic focal plane with lots of detectors.
Can we provide any comparisons with current state of the art?  For
example, the instrument report slide 67 says there will be 21 detector
modules at 0.3 kg/module, for a total of 6.3 kg at 100 mK.  What's the
heaviest existing focal plane assembly?  At what temperature?
Similarly, Slide 62 says the detector yield will be above 90%.  What is
the demonstrated yield for existing focal plane arrays?

9) The Cost rationale for WBS 7 (Mission Ops) in the mission report slide 41
should cite the simplicity of PICO ops: Single instrument with a single
science  observing mode. Science mode is simple spin/precession with no
targets of  opportunity  or guest investigator program.  No part of the sky
is more or less  interesting than any other.

There are also some relatively minor points:

10) The instrument executive summary (slide 9) should include the CBE
(or maybe MEV) thermal loads at the critical 4K and 100 mK stages.
It's hard to think about the coolers without knowing the loads.

11) Change "Mirror" to "Reflector", throughout.

12) The instrument report slide 35 says that the cost driver for the cooler
is the minimum required temperature.  Nothing is said about total heat lift
at minimum temperature.  Should we say that the PICO expected heat lift
is in family for the various coolers cited?

13) What is meant by the "negligible (1 nW)" power per detector in the
instrument report Slide 67? The detector readout is considerably bigger
than that ...

14) the discussion of cosmic ray hits (instrument report slide 72) is
confusing for anyone not already an expert.  Need to state that the
vast majority of Planck CR hits were to the detector frame,
not the detectors per se.  The PICO design eliminates such frame hits
by heat-sinking frame to bath, as validated by Suzaku, Keck and SPIDER.

15) Three reaction wheels is OK, but provides no margin against
single wheel failure.  Four wheels is better (robust against single failure).
Mission report slide 58 says design is tolerant to single wheel failure.
It's not immediately clear to me how this happens.

16) The mission report slide 144 says that the instrument continues to
observe during Ka-band downlinks.  Is there a risk for science data loss
at Ka band or harmonics?  Given the 4 hr/day download, this could be
significant.  Can we provide an estimate of the leakage from the Ka-band
transmitter to the Ka-band detectors?

Not all of these need to be addressed via the cumbersome TeamX
response/rebuttal process, but we should fix what we can now and
keep the rest in mind for the eventual PICO report.

Cheers, Al 


----

__Jamie's Comments__

Hi Shaul,

Reading through the report, here are some suggestions.  I read the main report first, the instrument report second, as indicated by the dividers.

__Major - Main__

1.  Analogies to justify the schedule on slide 34.  Because Planck and Herschel were much longer, one may argue the schedule represents a major cost risk, and it seems very aggressive to me.  I think the integration and testing of a large 100 mK instrument especially needs to be carefully justified and costed.

2.  Along these lines, I did not see a development plan for instrument through to S/C level testing.  Will it go through the usual practice of EM and qualification units like Planck and Herschel?  That is arguably appropriate but adds time to the schedule.  Not having a schedule or at least an integration plan is a problem and will cause reviewer confusion.  For example, the V-groove test may be justified (slide 113), but how does this fit with the general demonstration of the cryo instrument?  What are the steps?

3.  There is little mentioned about spinning satellite platforms to justify the costing.  I have no idea how comparable this is to SMAP, though this was not a cheap mission.  Providing analogies to existing systems to justify the cost estimate would strengthen the case.  Is the SMAP pointing performance suitable or are we in a different regime?

4.  Where is the boundary with the payload?  I see the v-grooves are part of the S/C thermal system.  I would say the V-grooves are integral to the instrument performance and would put them in the payload WBS.  Certainly the bottom panel is a much simpler interface than the cold side of the V-grooves.  Likewise I am unclear about the cables, but would make the boundary after the instrument electronics.  I would generally make the interface to the payload the bottom of the spin platform (except for the ACS components on the spinning deck). An easily defined and simple interface boundary will help.

5.  For the SVIT, Planck ran a TVAC test with the spacecraft with the payload cooling to 100 mK.  Is such a test planned?  I can't imagine it fitting in 20 days (slide 139).  I think one should look hard at this phase of the project for Planck to see if it is underscoped.

============================

__Major - Instrument__

1.  Need explicit cost analogies for major parts:  detectors, readout electronics, telescope, v-grooves, I&T and science.  Assume the cost modelers don't know what these units even are.  Provide sufficient cost and schedule information from Planck and WMAP to justify these estimates.  E.g. explain how you scale from single-detector fabrication and test activities with Planck to fabrication and test of arrays.  Likewise need cost analogies for the 4 K cooler and ADR from other missions.  Some of these items are reasonably well covered (telescope), some are open (e.g. science).  Science certainly will cost more than the standard team-X wrap.

2.  Can the instrument electronics, compressors and cooler electronics really reside in the non-spinning section?  That requires a huge number of low-signal wires and pressure lines through a rotating connection.  Scares me!  I would put these units on the spinning platform and have rotating connections for data and power.

__Minor - Main__

1.  It would be helpful to clearly call out the single operating mode.  That's a plus obviously but you have to dig through the report to find it.

2.  Team-X generally does a good job on the standard spacecraft systems and uses similar cost models, so I am only picking out the unusual aspect of the spacecraft.

3.  The report states 3 wheels (slides 23, 53).  Are these redundant (slide 54, 58) for both parallel and perpendicular?  For a 5-year mission they should be.

4.  Is the harness mass just for the cryo instrument?  100 kg or 63 kg both seem high for cryo cables.  The $10M for harness design and implementation (slide 91) seems high.  Cables above the spinning section interface should be carried in the payload WBS.

5.  Surely one can find analogies for bipods (e.g. Spitzer) and v-groove Planck costs (slide 89) and performance (slide 115).  The Planck V-groove design and performance is a reasonable starting point for heritage.

6.  One can cite experience with Planck, Herschel and Spitzer in avoiding contamination problems from the thrusters (slide 105) on both the panels and the cryogenic optics.  Worth mentioning, since contamination is a frequent topic for reviewers.

============================
__Minor - Instrument__

1.  Comment about contamination venting from detectors on to optics is amusing (slide 11), but I think Planck decontamination process is good heritage.

2.  I see no benefit to mentioning MKIDs, only downside.

Jamie


----

__Bill's Comments__

Hi Shaul,

I looked through slides 127-142, doing my best to discern content hidden behind the jargon and acronyms.  As far as I can tell, there is probably a sensible costing of the I&T of the spacecraft based on considerable heritage of this relatively simple platform.  It didn't seem to clearly indicate the cost and schedule drivers for the science instrument.  Is that to be found elsewhere?    Best I could find was slide 31 of the second document, WBS 5.04.12 I&T budget of 6M$

There are some inconsistencies on slide 34, including the headline cost of the telescope (6.47 or 6.07 M$?).  When I run the model using the same inputs I get something higher than either. Maybe they are using the projected diameter?  This cost claims to include the I&T?  That includes cold metrology?  I think the cost of this will be significantly higher.

The sum of these two items is *less* than what they have budgeted for the computer to do the data compression.  I cannot imagine how the computer would be this expensive (it is trivial), nor can i imagine how the former would fit in that budget.

Is the science instrument (cryo+optics+detectors+readout) I&T budgeted elsewhere, that I have missed?

Slide 75 of the draft report cites a design weakness that TES detectors have not been operated in space.  I think this is not fair - the TES and both the tMUX and fMUX have been tested on autonomous balloons.  This is the same level of technical readiness as the detector technology flown on Planck and Herschel.  PICO does not have less heritage than these missions did, at least in this regard.

Bill

----
User Tools

Page Tools
