Transcript Slide 1

Discussion Topics for Workshop on Seismic Design for New Observatories
December 3, 2007
Michael Sheehan <[email protected]>
Kei Szeto [email protected]
Jeff Barr [email protected]
Suggested topics in black
Notes from discussion in red
• Which elements of the facility are appropriate to design to code standards, and which require the
more rigorous dynamic calculation (typically using the a full spectra and FEA).
Some are obvious but others are somewhere between "civil” and "telescope" structures:
dome structure, dome mechanisms, telescope pier, platform lifts, lower enclosure structure, mechanical
equipment supports in the facility, coating plant infrastructure.
Moving parts of the telescope & dome were generally addressed as requiring dynamic analysis.
All others were discussed in the context of code-based criteria, although for some observatories more
conservative load factors are applied to these building and equipment items as well.
• How best to identify the appropriate survival & operational events to design for.
•Is lifespan of the observatory a relevant factor?
Not discussed, except to note that the % likelihood in a 50-year time span, used as the usual
seismic event definition, corresponds closely to typical telescope useful lifetime
•Actuarial type analysis for liability?
Unlikely as an issue, although noted that VLT (ESO) has insurance that covers a % of earthquake
damage. Insurance companies or observatory Risk Managers could weigh in.
•Code standards?
Codes generally considered inadequate, as they insure only life safety, not survival of building &
other structures. ALMA considered code standards as only adequate for temporary structures.
EURO Code 8 & US- ASCE-7 noted as comprehensive guides for Seismic Design. The Chilean code
also has special engineering sections for seismic.
• Should there be another lesser level of event defined for design of transitory conditions like
instruments hanging on hoists, mirror being removed for coating, or should these be fully safe for an
"operational event“?
Antenna movement at ALMA, mirror transporter use at VLT, and camera removal at for LSST were raised
as examples. Consensus seemed to be that the level of risk during such activities is unique to the
particular observatory and component, and that general conclusions are hard to draw. The sense was
that rational criteria should be based on the % likelihood of a given seismic event occurring during the
limited duration of the task, and the criticality of the at-risk elements.
• How to consistently and logically include a factor of safety for observatory design – as an increased
importance factor ?, lower response factor for the proposed type of structure than would be code
allowed? Does it make more sense to include a completely separately factor, applied independently of
the code calculation?
It was noted that many approaches have been taken (AEOS-Maui – importance factor of 1.25, ALMA –
reduced response factors of structural system, TMT – reduced damping factors) and in most cases a
combination. The resultant applied load is the main point, so it didn’t seem to matter too much how it
is developed – as long as the factor(s) used are traceable.
Lots of side discussion of damping factors. A wide range from .5% to 9% can be considered
appropriately conservative depending on the component it is applied to.
A suggestion was made that an overall Factor of Safety would be a useful baseline recommendation to
come out of this conference, but no specifics resulted from discussion.
• Is there any expectation of improved sensing in the foreseeable future that would provide a warning
period immediately prior to a significant seismic event? Would it be useful for observatories to
collaborate on some monitoring system and program. If so, how would this impact design?
There is a good prospect of cooperating on a system based on sensing of initial P-waves that would
provide ~10s warning of major damaging shock waves to all Chilean observatories. What to do with
that in regard to design is less clear. VLT is clearly ahead of others on this, with on-site accelerometer
sensing and automatic clamping mechanisms for mirror and telescope systems. It was triggered in the
‘97 quake – before operations – and never since. General sense that any realistic response action
would have to be automatic (people are likely to freeze), although the VLT protocol calls for some
discretionary control by on-site operators based on the warning level display in their control room.
• On the practical end, incorporating into the design aspects of:
• design-for-inspection
This was recognized as critical and often poorly planned aspect of observatory design. Damaged
areas – especially drive surfaces, are often difficult or impossible to effectively inspect. As with
other issues, appropriate design is very specific to the individual observatory in question.
Difficult to derive general recommendations.
•design-for-alignment,
Recognized as a common problem from recent events. More information on this to come during
the observatory-specific presentations the following day
• design-for-replacement after seismic events
Stocking appropriate spare parts briefly addressed. Often dictated by budget more than
operational design concerns. With regard to specific part survivability, actual vibration testing of
critical components was suggested. It was noted that this could require costly sacrificial test
pieces - as the vibration from the testing itself would reduce the part life.
• Considerations in the design stage criteria that will help the observatory operation determine when
it is safe to replace or repair the observatory after an seismic event.
This led to discussion of design for personnel safety: safe egress capacity, people-safe working areas
and other issues. Other than code-based criteria – which were recognized as often unrealistic to
apply to such areas as mirror cells, instrument enclosure/platforms, and other telescope specific
areas. Further discussion of these issues was encouraged during the next days safety-response
portion of the conference.
Suggestions for approaching the subject of New Observatory Design:
Not directly discussed but raised in the context of some of the other questions
•Differentiate and separate design requirements (loads, code standard etc.) between the facility
building, enclosure and telescope structure. Different philosophies and strategies may apply even
though the observatory cannot operate if any one is damaged.
• Start with methods to define the load requirements (input),
Analysis methods to calculate the design loads (including discussion on available methods to model
soil, foundation, soil and structural damping, damping and stiffness of non-structural components
such as bearing and drive etc.),
• Strategies to mitigate and reduce seismic loads and dynamic amplification etc.
• What level damage is tolerable given cost and downtime considerations,
• Other non-telescope systems that required considerations such as encoder and brake systems on the
telescope structure, other systems in the facility and enclosure.