TMT telescope structure system: seismic analysis and design

Abstract

This paper documents the methods used for the seismic design and analysis of the Thirty Meter Telescope (TMT)2. The seismic analysis includes response spectrum and nonlinear time history methods. Several seismic restraint design options are considered, both linear and nonlinear, and the seismic performance is presented for these options. The paper addresses several issues specific to large optical telescope seismic design and analysis: generation of appropriate response spectra and time histories; use of operational and survival level earthquakes; selection of damping coefficients; use of reduced degree of freedom models and their calibration with more detailed models; and local response spectra for telescope-mounted systems.

Full text

TMT Telescope structure system – seismic analysis and design
Dominic Tsang
a1
, Glenn Austin
a
, Mike Gedig
a
, Christie Lagally
a
, Kei Szeto
b
, Genady Sagals
c
,
Larry Stepp
d
a
Empire Dynamic Structures;
b
National Research Council Canada, Herzberg Institute of Astrophysics, Dominion Astrophysical
Observatory;
c
Association of Canadian Universities for Research in Astronomy;
d
Thirty Meter Telescope Project
ABSTRACT
This paper documents the methods used for the seismic design and analysis of the Thirty Meter Telescope (TMT)
2
. The
seismic analysis includes response spectrum and nonlinear time history methods. Several seismic restraint design options
are considered, both linear and nonlinear, and the seismic performance is presented for these options. The paper
addresses several issues specific to large optical telescope seismic design and analysis: generation of appropriate
response spectra and time histories; use of operational and survival level earthquakes; selection of damping coefficients;
use of reduced degree of freedom models and their calibration with more detailed models; and local response spectra for
telescope-mounted systems.
Keywords: Thirty Meter Telescope, TMT, seismic design, seismic restraint, transient analysis, nonlinear, isolation,
damping, response spectra
1. OVERVIEW OF TELESCOPE AND MODEL
Given the high-seismicity of the candidate Thirty-Meter-Telescope (TMT) sites
[1]
, it is paramount that sufficient
protection be provided for the telescope structure, optics, and instruments against earthquakes. This paper solely
discusses the design and analysis work for the seismic restraints. The overall telescope structure design is described in a
separate paper in this conference by Szeto et. al
[2]
The telescope is formed by two main structural parts, the elevation structure and the azimuth structure, which provide
support for the telescope-mounted systems (Figure 1). The elevation structure holds the segmented primary, secondary,
and tertiary mirror systems
3
(M1, M2, and M3, respectively) and can rotate about the horizontal elevation axis from
zenith to horizon pointing. It is supported by four hydrostatic shoe bearings (HSB) on the azimuth structure. Two
Nasmyth platforms located on either side of the elevation structure support various instruments and are tied to the cradle
of the azimuth structure. The latter is supported vertically on six HSB around the perimeter of an azimuth track of
17.6m radius, allowing the telescope to rotate about the vertical azimuth axis. Operational lateral loads are transmitted
to the ground through the pintle bearing located at the centre of the azimuth structure.
The current TMT finite element model (FEM) representation of the telescope is shown in Figure 2. It is composed of
beam, shell, and spring elements. The mass of the telescope is 1800 tonnes, including both structural and non-structural
masses. The fundamental mode frequency is 2 Hz. The structure is 55m wide, as measured from the ends of the
1 Further author information – D.T. (correspondence). Email: dominic.tsang@empireds.com; Address 1515 Kingsway Avenue, Port
Coquitlam, B.C, Canada V3C 1S2; Telephone: 604-468-7646; Fax 604-941-7447
2 The TMT project is a partnership between ACURA (Association of Canadian Universities for Research in Astronomy) in Canada,
the University of California and Caltech.
3
In addition of the mirror, each mirror system contains a mirror support system with actuators for figuring control and positioner for
alignment control and the associated control electronics.
Ground-based and Airborne Telescopes II, edited by Larry M. Stepp, Roberto Gilmozzi,
Proc. of SPIE Vol. 7012, 70124J, (2008) · 0277-786X/08/$18 · doi: 10.1117/12.790288
Proc. of SPIE Vol. 7012 70124J-1
2008 SPIE Digital Library -- Subscriber Archive Copy

A
Nasmyth platforms, and rises about 50m from the ground. Due to its height and the obscuration limit imposed for the
primary mirrors restricting the size of its support structures, the M2 is particularly sensitive to seismic loads.
In the FEM, lateral seismic loads are transmitted from the ground to the structure through springs representing soil
stiffness, which are connected to the concrete pintle bearing pier. The latter transfers the load to the pintle bearing
springs and seismic restraint springs, which finally pass it to the structure, resulting in inertial loads or structural
accelerations.
Figure 1: TMT CAD model. The structures supporting the science instruments on the Nasmyth platforms are not shown.
Proc. of SPIE Vol. 7012 70124J-2

AN
EW4ENTS
MAY 6 2008
REAL NUM
14: 31: 44
Figure 2: TMT finite element model.
2. DESCRIPTION OF SEISMIC RESTRAINT REQUIREMENTS
Two seismic performance criteria are specified, based on the severity of the earthquake
[3]
. For an Operational Basis
Survival Condition (OBSC) earthquake, which is defined as a seismic event with an average return period of 200 years,
the telescope system shall suffer no damage and astronomical observations and regular maintenance operations shall
resume after inspection lasting no longer than four hours. For a Maximum Likely Earthquake Condition (MLEC) event,
the return period becomes 500 years and operation resumption is five days for the telescope structure system.
Several restraint design criteria are specified in response to the seismic performance requirements:
¾ The restraints shall not interfere with normal telescope motions and operations
¾ The restraints shall be the primary lateral-motion resisting and load bearing devices during a MLEC earthquake
and protect the rest of the structure and telescope-mounted systems from damage
¾ The structure and restraints shall both behave elastically during an OBSC earthquake
¾ The restraints may behave inelastically during a MLEC earthquake to keep the structural elements within the
elastic level and the optics & instruments from being damaged
¾ Restraint against uplift shall be provided if necessary
¾ The restraints shall retain sufficient stiffness and strength to also protect the structure against aftershocks
Secondary mirror
support
Elevation
journal
Nasmyth
instrument
support
Nasmyth
platform
Azimuth
structure
pier/soil
Pintle
bearing
pier/soil
Primary
mirror cell
Elevation
structure
bearing
Azimuth structure
bearing
Azimuth
structure
cradle
Azimuth structure
center
Tertiary
mirror
Secondary mirror
Proc. of SPIE Vol. 7012 70124J-3

3. ANALYSIS
This section describes the analysis method implemented to support the seismic restraint design. It is important to choose
a suitable method to capture the dynamic and potentially nonlinear behavior of the seismic restraints.
3.1 Methods
There are several popular seismic analysis methods, namely: 1) equivalent static load procedure, 2) spectrum analysis,
and 3) time-history (transient) analysis. The first procedure will be used only as an order of magnitude level verification
since it lacks accuracy for complex, irregular structures. The spectrum analysis is appropriate only for structures
behaving linearly. Since it is anticipated that the restraints will behave in a nonlinear manner, time-history analysis is
selected as the primary analysis method. It is also the most flexible method in terms of the type of structural behaviors
and load scenarios allowed.
The downside of transient analyses is long computational time. With over 20,000 nodes and 35,000 elements, the
current TMT FEM requires over nine hours to perform one analysis over 1500 time points (15-sec total) on a high-end
PC workstation. The long run-time may be mitigated by the use of the substructuring technique, which effectively
allows the simplification of the linear-behaving portion of the FE model to a single “super-element” and thus
significantly cuts down run time. Although the stiffness of the simplified model would be maintained during this
process, the mass and mass moment of inertia of the original model need to be redistributed by assigning “master
nodes”, which are nodal locations where original structural masses are lumped.
There is no specific rule for master node assignment, but it is found that putting master nodes at locations of
concentrated masses and at the interface of the super-element and the original elements would preserve the dynamic
characteristics of the original structure. This is judged by comparing both the time-domain transient analysis and the
frequency-domain harmonic analysis results, the latter involving the evaluation of frequency responses of the two
models with unit seismic loads applied. Currently, about 250 master nodes are assigned to the simplified model. A
transient analysis of this model is drastically reduced to about one hour.
The accelerations of non-structural components (e.g. the secondary mirror subsystem) due to seismic loads are
dependent on their own natural frequencies. Specifically, if the natural frequencies of an instrument are similar to those
of the dominant modes of the telescope structure, the resulting resonance effect can amplify the instrument’s
accelerations. To examine this effect, a local response spectrum can be developed for any telescope-mounted system.
Using the secondary mirror system as an example, one can extract the time-history response at M2’s mounting point and
then convert these results into the frequency-domain to derive a local response spectrum. With this information, the M2
design team can then determine which frequencies are sensitive to seismic loads and modify their design accordingly, if
necessary. This also forms the basis of seismic loads interface requirements between the telescope structure and the
telescope-mounted systems.
3.2 Assumptions
Damping is a major source of uncertainty in seismic design due to lack of suitable experimental data. Table 1
[4]
lists
several damping sources relevant to telescope structure design. For transient analyses, Rayleigh damping will be
employed, which involves modification of the mass and stiffness matrices by factors alpha & beta, respectively, to
incorporate these damping sources. In relation to the more commonly used (but not applicable for transient analyses)
damping ratio concept, the two factors can be used to generate a V-shaped damping ratio vs. frequency graph. Initially,
alpha & beta damping coefficients will be conservatively chosen so that the constant damping ratio between 2 and 10
Hz, where the dominant vibration modes lie, is 1% or less (Figure 3). In addition, beta damping is defined locally for the
springs representing soil. With alpha equals 0, a linear damping ratio vs. frequency relationship is generated where the
soil damping ratio is 10% at 10 Hz. Further investigations will be conducted to determine whether higher damping
values for the overall structure or specific components (e.g. bearings) is warranted, which would reduce structural
accelerations and restraint load demand.
Proc. of SPIE Vol. 7012 70124J-4

Table 1: Damping mechanisms
Damping Ratio vs. Natural Frequency
0.0%
0.2%
0.4%
0.6%
0.8%
1.0%
1.2%
1.4%
1.6%
1.8%
2.0%
0 5 10 15 20
Natural Frequency, Fn, Hz
Damping Ratio, zeta, %
Damping <= 1%,
between 2 & 10 Hz
Figure 3: Damping ratio vs. frequency curve generated with the Rayleigh damping method.
Because a final observatory site has not been selected for TMT, site- and soil- specific ground motion time histories are
not available for analysis. Instead, synthesized ground acceleration time histories based on a 500-yr return period
earthquake in Mauna Loa, Hawaii are used
[5]
. This data set contains 40 seconds of motion in 0.01-second increments
for all three orthogonal directions. It was used for the seismic analysis of the Gemini Observatories. The peak ground
acceleration is 0.30g in the horizontal directions and 0.22g in the vertical direction.
3.3 Seismic loads
Seismic loads are applied simultaneously in three orthogonal directions as constrained displacements at the ground
nodes. These loads are transmitted upward to the rest of the structure as described previously. Displacement inputs are
constructed from the ground acceleration data by double-integration over time. Although 40 seconds’ worth of data is
available, peak structural responses can be captured with only 15 seconds of data and thus analysis times can be reduced
by about half.
Currently, only one set of ground motion data (Hawaii) is being used. In the final analysis, multiple sets of ground
motions will be employed. These ground motions will be generated by commercial software which incorporates local
soil and seismology information. This includes soil stiffness, soil type, seismic event probability, and fault types &
locations in the surrounding area.
3.4 Load cases
A large number of load cases need to be run to sufficiently understand the behavior of seismic restraints under various
operational scenarios and restraint parameters.
Damping Type Energy Absorption Mechanism
Base/soil damping Frictional interactions or movement between soil particles and/or the foundation
Frictional damping Friction between bolted joints, restraints, attached walkways, cables and hoses, etc.
Viscous damping Drag from air or wind as the structure vibrates in a medium
Control system damping Mechanical, magnetic or hydraulic damping mechanisms (active or passive)
Structural damping Inter-molecular interactions in the material from which the structure is made
Proc. of SPIE Vol. 7012 70124J-5

The telescope is a dynamic structure that operates in multiple configurations. The zenith angle of the elevation structure
is among the most important parameters, for several reasons: 1) the stiffness of the elevation structure journals (and thus
the overall elevation structure) varies with zenith angle, 2) the height of telescope-mounted systems relative to the
ground changes, and 3) the acceleration direction of telescope-mounted systems changes, which results in the need to
consider inertial loads on the telescope optics in multiple directions. Other loading parameters include: HSB on or off
oil (which alters bearing stiffness), elevation and azimuth structure brakes on or off (which affects rotational stiffness),
Nasmyth platform instruments installed or not (which alters system mass and mass eccentricity), etc.
A major restraint design choice is whether to utilize nonlinear material behavior. The primary benefits of designing for
elastic restraints, which deform proportionally to the applied forces, are quicker recovery from seismic events (since no
damage) and simpler analyses. However, because maintaining elasticity implies constant restraint stiffness during an
earthquake event, the resulting structural loads and instrument accelerations may become excessive, which would
disqualify the elastic restraints from being a feasible solution. On the other hand, by utilizing nonlinear behavior, the
restraints deflect linearly up to an elastic force limit, then displace with no (or low) stiffness afterwards while dissipating
the seismic energy. The restraints themselves may require repair or replacements after the earthquake, but their ability to
limit seismic loads on the structure helps maintain elasticity of the structural members. Examples of nonlinear devices
include friction dampers, buckling-restrained braces, and yielded members. To further reduce seismic load, the brakes
that resist elevation and azimuth structure rotations may be designed to allow slippage. The nonlinearity characteristics
of the restraints and brakes are important parameters to be investigated.
The seismic restraints are co-located with the pintle bearing at the central base of the azimuth structure. The selection of
load path is another design choice. The two lateral-load resisting devices (bearings and seismic restraints) can be loaded
in series or in parallel. In the case of serial loading, seismic load flows from the ground to the restraints, then to the
pintle hydrostatic bearing pads. In the case of parallel loading, seismic load is transmitted simultaneously to both
components. Generally, parallel load paths result in restraints with higher load-resisting capacity and stiffness since
stiffness of the components is additive. On the other hand, serially-loaded restraints are easier to install and align.
Tables 2 and 3 provide additional comparisons of linear vs. nonlinear restraints; and of restraints loaded in series vs. in
parallel.
Table 2: Comparison of linear vs. nonlinear seismic restraints
Linear Restraint Nonlinear Restraint
Force transmitted to structure Higher Lower, since seismic load is
limited by nonlinear behaviour
Required load capacity of the
lateral HSB
Higher Lower
Analysis complexity Lower Higher, requires use of time-
consuming transient analysis
Analysis accuracy Use standard analysis
methods with confidence
More work is needed to verify
result accuracy
Fabrication tolerance
requirements
Similar
Installation tolerance
requirements
Similar
Downtime Short, since no damage Longer, to repair/replace
components
Relative cost Lower Higher repair/replacement costs
4
4
The higher costs may be offset by the lower telescope-mounted system costs which may be designed for lower seismic loads.
Proc. of SPIE Vol. 7012 70124J-6

Table 3: Comparison of seismic restraints with serial and parallel load paths
Serial Parallel
Force transmitted to structure Same if linear behaviour
Required load capacity of the
lateral HSB
Higher, since lateral HSB
takes the same load as the
restraint
Lower, since the restraint can be
designed to take the majority of
loads
Analysis complexity Lower Higher; need to be concerned
about load sequence
Analysis accuracy Use standard analysis
methods with confidence
More work is needed to verify
result accuracy
Fabrication tolerance
requirements
Lower Greater precision is required
Installation tolerance
requirements
Lower Greater effort required to align
components so they are loaded
as intended
Downtime from seismic event Similar
Relative cost Lower Higher
4. RESULTS
With transient analyses, an extensive amount of result data is generally generated over the time-history. However, the
simplifications made by the substructuring technique can restrict the result extraction. Without resorting to time-
consuming “expansion pass” analysis which essentially expands the simplified model back to the full model, nodal
results are only available for nodes assigned as “master nodes” and elemental results can be extracted only from
elements that are not part of the “super-element”. However, with judicious selection of master nodes and exclusion of
certain elements from the super-element, this restriction is not a major impediment.
Table 4 lists the results to be extracted from the time-history analysis which provide quantitative comparisons to guide
the development and optimization of the seismic restraint design. Displacements, accelerations, and element forces are
reported in three orthogonal directions where applicable in the component’s local coordinate system. For member stress,
results at only selected time points will be determined using expansion pass analyses. The time points will be selected
based on the instances of high structural accelerations or restraint forces. For all other result types, time-history series of
outputs will be determined.
As the seismic restraint design and analysis work is in progress, only representative results are provided in this paper.
Figure 4 shows the Mauna Loa ground acceleration time history which constitutes the x-component of seismic load,
peaking at 0.3g as mentioned. (The y- and z-components of loads are also applied, but not shown) The next two figures
show the preliminary results for x-direction accelerations of M2 and M3 support points for both linear
5
and nonlinear
models
6
. Table 5 summarizes the peak results of both cases. It can be seen that, comparing to the linear model, utilizing
a nonlinear restraint in the nonlinear model results in lower peak accelerations. It should be understood that these results
are preliminary and the current seismic restraint parameters have not yet been optimized for seismic performance. The
following design and modeling strategies will be implemented to refine the telescope-mounted system seismic
acceleration predictions:
• Utilization of component-specific damping ratios for bearings, restraints, and soil, where warranted
5
For the linear FE model, the restraint is modeled as a linear spring with a stiffness value of 1x10
9
N/m.
6
For the nonlinear FE model, the restraint is modeled as a bi-linear spring with the same stiffness value of 1x10
9
N/m, but with an
elastic force limit of 2000 kN at which point stiffness becomes zero. Similarly, the brakes are modeled as bi-linear springs to allow
slippage upon exceedance of brake capacity.
Proc. of SPIE Vol. 7012 70124J-7

• Optimization of seismic restraint parameters, e.g. stiffness, elastic force limit, nonlinearity, load path, energy-
dissipation mechanism
• Optimization of the support towers for the telescope-mounted systems, e.g. stiffening of the M3 support
structure will reduce M3 accelerations
• Integration of the M2 & M3 system structural designs with the support towers to minimize the acceleration
amplification effect by structure
Table 4: Transient analysis results
Result type Item
Nodal displacement
Nodal acceleration
Both ends of seismic restraint elements
Base of segment handling cranes
Edges of Nasmyth platforms
Selected M1 segment support points
M2/M3 support points
Centre of gravity locations of:
Selected M1 segments
M2/M3
Nasmyth platform instruments
Laser guide star facility components
Element force Elevation and azimuth structure hydrostatic shoe bearings
Elevation and azimuth structure drives
Elevation structure guideshoes
Pintle bearing
Seismic restraints
Sliding distance Seismic restraints (if nonlinear)
Member stress All elements where applicable, at selected time points
Proc. of SPIE Vol. 7012 70124J-8

Mauna Loa Ground Acceleration
-0.40
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
0 5 10 15 20 25 30 35 40
Time, s
Ground Acceleration, g
Figure 4: Lateral ground acceleration time-history, applied as constrained displacement in x-direction after double-
integration over time. Only the first 15 seconds of data is used in the analysis. [Maximum = 0.30g]
Secondary Mirror Support Point Time-History X-Acceleration Results
-4
-3
-2
-1
0
1
2
3
4
0 2 4 6 8 10121416
Time, s
Acceleration, g
M2 Acceleration - Linear restraint
M2 Acceleration - Nonlinear restraint
Max x-acceleration:
M2 - linear: 3.2g
M2 - nonlinear: 2.3g
Figure 5: Comparison of M2 support point x-acceleration between models with linear vs. nonlinear restraints.
Proc. of SPIE Vol. 7012 70124J-9

Tertiary Mirror Support Point Time-History X-Acceleration Results
-3
-2
-1
0
1
2
3
0 2 4 6 8 10 12 14 16
Time, s
Acceleration, g
M3 Acceleration - Linear restraint
M3 Acceleration - Nonlinear restraint
Max x-acceleration:
M3 - linear: 2.4g
M3 - nonlinear: 1.5g
Figure 6: Comparison of M3 support point x-acceleration between models with linear vs. nonlinear restraints.
Table 5: Seismic accelerations of telescope-mounted systems, for linear and nonlinear FE models (in units of g)
Description Linear Model Nonlinear Model
x y z x y z
M1 support frame 1.7 1.9 1.9 0.9 0.7 1.4
Nasmyth platform 1.3 2.1 1.4 0.7 1.0 1.4
M3 support point 2.4 3.8 1.5 1.5 2.3 1.4
M3 centre of gravity 3.0 4.6 2.5 2.1 3.0 2.4
M2 support point 3.2 5.0 1.5 2.3 2.0 1.5
M2 centre of gravity 3.7 6.1 2.8 3.1 3.1 2.8
5. IMPLEMENTATION CONSIDERATIONS FOR THE SEISMIC RESTRAINT
CONCEPTS
Broadly speaking, there are two main methods of protecting structures from earthquakes and they may be termed
restraint and isolation. The restraint method uses a structural tie that reduces a degree of freedom between two otherwise
moving parts; for example, the pintle bearing assembly is a restraint between the azimuth structure and the azimuth
track. The isolation method uses a spring, damper, or gap device that reduces the transmission of ground acceleration
motion to a structure. Both types of devices may be sacrificial, i.e. damaged during a MLEC earthquake, provided the
time needed to replace them is within the recovery specification of five days.
5.1 Pintle bearing assembly
The baseline design places five hydrostatic shoe bearings mated to a cylindrical steel track at the top of a structural
column referred to as the pintle bearing assembly. By definition, the pintle bearing assembly provides the center of
rotation for the telescope and reacts to forces only in the horizontal plane. Thus it serves as the lateral seismic restraint
Proc. of SPIE Vol. 7012 70124J-10

for the telescope structure by preventing it from sliding off the azimuth track. In so doing, the ground motions and
seismic forces are transmitted through the bearings to the telescope. Currently, two implementations are under
consideration:
5.1.1 Restraint in parallel with pintle bearings
An alternate pintle bearing configuration with three pairs of two bearings with each pair mounted to a load dividing
bogie is shown in Figure 7. This concept, as shown in Figure 7, places a second structural load path beside the pintle
bearings to protect them from earthquake forces that would otherwise damage them.
5.1.2 Isolation in series with pintle bearings
This concept places springs or slip-able elements at any one of the interfaces along the horizontal seismic load path
between the pintle bearing pier support column and where the pintle bearing assembly contacts the azimuth structure.
Ideas under study include tailoring the spring rate of the pintle column itself, installing industrial slip devices between
the bearing assemble and the azimuth structure, and, mounting the pintle bearings on force-limiting bogies.
5.2 Azimuth bearings and azimuth track
Due to the 17.6m vertical offset between the telescope centre of mass and the pintle bearing assembly, the horizontal
ground acceleration will create reactions at the azimuth bearing locations which subtract from and add to the static
weight reactions. Upstops can be used to prevent uplift of the azimuth bearings if the compressive load falls below zero.
The FE simulation will provide the peak compressive reactions at each azimuth bearing during the earthquake and
determine whether the current size and number of bearings is sufficient to survive the MLEC earthquake. Alternate
configurations under study with greater load capacity for the azimuth bearings are six pairs of HSB600 size bearings
with each pair mounted to a load dividing bogie, or alternatively, four sets of 3 at four corners, where each set of 3
bearings is mounted on a load dividing bogie.
In this case seismic restraints are needed to prevent the telescope from lifting up off the azimuth track. A current concept
uses a mechanical upstop restraint shaped like a hook that reaches under the azimuth track plate’s edge and runs with a
small amount of clearance during normal operation (Figure 8). Again FE analysis will be used to establish the need for
and strength required of the upstops, their connections, and the azimuth track.
Figure 7: Plan view of azimuth structure seismic restraints and
pintle bearing pads in parallel load path.
Figure 8: Elevation view of an upstop
device.
Pintle
bearing pad
Seismic
restraint
Proc. of SPIE Vol. 7012 70124J-11

5.3 Elevation bearings and rockers
The current design mates 2 HSB600 hydrostatic shoe bearings to each of the two 10.75m radius elevation journals, with
each bearing placed 25° off of vertical to cradle the journal. This gravity-organized bearing system requires static and
seismic restraint in the direction of the elevation axis. The current design places additional hydrostatic shoe bearings
against the flat faces of the journals to provide this constraint. The baseline assumption is that the elevation seismic
restraint is expected to be linear system. FE simulation will be used to size these bearings and determine if additional
restraint or isolation is required and to determine if upstops are needed to prevent the elevation journals from lifting off
of the main elevation bearings.
6. CONCLUSION & FUTURE WORK
The seismic restraint design is in progress. Upon generation of the full first set of transient analysis results, the restraint
design space (stiffness, nonlinearity mechanism, load path, etc) can be considerably narrowed. Further analyses will
then be conducted with a point design to examine a comprehensive set of load cases to ensure that the telescope structure
and the telescope-mounted systems will be sufficiently protected under various telescope configurations and load
scenarios. For these analyses, site- and soil-specific ground motion data, and realistic global & component-specific
damping values will be defined to obtain accurate responses.
Finally, the seismic design and analysis methodology will enable an integrated telescope structure and seismic restraint
system to be optimized efficiently and to ensure that the telescope-mounted systems maintain safe acceleration values.
ACKNOWLEDGEMENT
The authors gratefully acknowledge the support of the TMT partner institutions. They are the Association of Canadian
Universities for Research in Astronomy (ACURA), the California Institute of Technology and the University of
California. This work was supported as well by the Gordon and Betty Moore Foundation, the Canada Foundation for
Innovation, the Ontario Ministry of Research and Innovation, the National Research Council of Canada, the Natural
Sciences and Engineering Research Council of Canada, the British Columbia Knowledge Development Fund, the
Association of Universities for Research in Astronomy (AURA) and the U.S. National Science Foundation.
REFERENCE
[1] “Thirty Meter Telescope Construction Proposal”, http://www.tmt.org/news/TMT-Construction%20Proposal-
Public.pdf (2007).
[2] Szeto, K., Roberts S., Gedig, M., Austin, G., Lagally, C., Patrick, S., Tsang, D., MacMynowski, D., Sirota, M.,
Stepp, L., Thompson, P., “TMT Telescope Structure System – Design and Development Progress Report,” Ground-
based and Airborne Telescopes, ed. L. Stepp and R. Gilmozzi, SPIE 7012, Marseille, France (2008) (this
conference).
[3] Angeli, G., “Observatory Requirements Document”, Thirty Meter Telescope Project, document
TMT.SEN.DRD.05.001.CCR18 (2008).
[4] Lagally, C., “Evaluation of Damping Ratios for the Design of TMT”, Thirty Meter Telescope Project (2007).
[5] Crouse, C. B., “Seismic Hazard Analysis for Gemini 8-M Telescopes Project” report, Dames & Moore, Inc. (1994).
Proc. of SPIE Vol. 7012 70124J-12