2 Digital Control and Software Compensation Capabilities - MAST ...

amaranthgymnophoriaElectronics - Devices

Nov 15, 2013 (3 years and 11 months ago)

143 views

PROJECT SUMMARY
______________________________________________________________________________

This proposal presents an integrated approach to the simulation of structural response of buildings and bridges in
moderate- and large-scale earthquakes, linking large-scale testing of structures with three-dimensional nonlinear
analyses of structural components and systems. The proposal focuses on the creation of a new large-scale
earthquake experimentation facility within the United States. The primary equipment installation will comprise a
Multi-Axial Subassemblage Testing (MAST) system. The MAST apparatus enables multi-axial cyclic static and
pseudo-dynamic tests of large-scale structural subassemblages including portions of beam-column frame systems,
walls, and bridge piers. The specific equipment category of this proposal, as per the NSF NEES categories listed in
Table 1 of the program solicitation, is D, Large-Scale Laboratory Experimentation Systems.

Large-scale structural testing of structures or components of structures can deliver engineering insight into structural
behavior that cannot be realized by any other means. Often, the boundary effects where the specimen couples to the
reaction structure are reduced to simple uniaxial loading configurations not necessarily representing the physical
boundary conditions experienced in practice. Furthermore, the difficulty of imposing multiple-degree-of-freedom
states of deformation and load using conventional structural testing means can be expensive, time-consuming, and
difficult to achieve.

The MAST system concept, as conceived by the University of Minnesota, employing an MTS six-degree-of-
freedom controller to position a crosshead using eight actuators (four vertical and two pairs of actuators oriented
horizontally, orthogonal to each other), can be used to apply realistic states of deformations and loading in a
straightforward and reproducible manner. The six-DOF control software employs advanced control technology to
locate the position of the crosshead through real-time simultaneous control of the eight crosshead actuators. Another
advantage of the MAST system is that it not only enables control of the position of a point in space, it enables
control of a plane in space. This feature makes it possible to apply biaxial control of structures such as multi-bay
subassemblages or walls. It also enables application of pure planar translations, as well as the possibility of applying
gradients to simulate overturning (e.g. axial load gradient in the columns of a multi-bay frame, or wall rocking).
With this system, a full six degree-of-freedom loading conditions can be imposed on the test structure, thus
providing a new capability that will enhance the earthquake engineering community’s understanding of complex
failure states.

With the MAST system, it will be possible to program any displacement history, and at the same time, it is also
possible to control the axial load on the test specimen using mixed mode control capabilities. As an example, the
vertical load on the test structure may be controlled as a function of the lateral position of the test specimen which
might be used to control moment-to-shear ratios on wall elements. The MAST system is unique and will greatly
expand the large-scale earthquake experimentation capabilities both nationally and internationally. Due to the nature
of the design, the system enables a high degree of flexibility for configuring various test programs. With the inherent
ease of test setup, overall flexibility, and the incorporation of teleobservation and teleoperation, the testing
community at large will greatly benefit from this system.

To achieve a successful program in the most cost effective manner, the University of Minnesota and MTS Systems
propose a collaborative solution. MTS will provide the actuators, hydraulic distribution system (including pump),
digital controller with six-degree-of-freedom software (control system), and the University will provide the lateral
load resistance for the horizontal actuators through the development of an L-shaped reaction wall system. The
University of Minnesota will also design and oversee the fabrication of the large steel weldments (crossheads) for
the top and bottom reaction surfaces.
1
A. RESULTS OF PRIOR NATIONAL SCIENCE FOUNDATION SUPPORT
Grant: NSF/GER 9023596
Title: NSF/PRESSS: Ductile Connections between Precast Beam-Column Frame Structures
PI: Catherine W. French
Amount: $140,000 Total
Period: 1991-1995
The purpose of the NSF PRESSS project on ductile frame connections was to develop new connection
concepts to effectively utilize precast concrete in seismic regions. In the project, four types of ductile
connection systems were identified: tension-compression yielding, nonlinear-elastic, friction-type, and
shear yielding. Quasi-static component tests were conducted at the University of Minnesota on four
details representative of the first two connection categories. The nonlinear-elastic systems showed the
most promise by limiting the damage to the connection region, and exhibiting very low residual
deformations after being subjected to large lateral drifts. Numerical models were developed to
characterize the observed moment-rotation behavior of the hinging regions. Model-based simulations,
incorporating the numerical models, were subsequently run to compare the response of entire structural
systems incorporating these concepts subject to a variety of ground motions. One Ph.D. student, L.
Palmieri, and one M.S. student, J. King, were co-funded on this project with the NSF FAW award
described below. The following is a list of selected publications:
1. Palmieri, L., Saqan, E., French, C., and Kreger, M., "Ductile Connections for Precast Frame Systems,"
ACI
SP162
, Mete A. Sozen Symposium, 1996, pp. 313-356.
2. Palmieri, L., and French, C., “Ductile Frame Connections,” Proc., 2nd Int. Scientific Conference on Analytical
Models and New Concepts in Mechanics of Structures, Keynote Paper, Lodz, Poland, June 1996.
3. Palmieri, L., French, C. and King, J., "Evaluation of Ductile Connections in Precast Frames subjected to
Earthquake Loading," Proc., XII Fed. Internationale de la Precontrainte, Washington, D.C., May 1994.
4. French, C. and Kreger, M., "Ductile Connections for Precast Concrete Frame Systems," Proc., ASCE Structures
Congress XI, Irvine, CA, April 1993.
5. Palmieri, L. and French, C., "Ductile Connections for Precast Frame Systems," Proc., Eleventh World
Conference on Earthquake Engineering, Acapulco, Mexico, June 1996.

Grant: NSF/GER 9023596
Title: Faculty Award for Women Scientists and Engineers
PI: Catherine W. French
Amount: $50,000/year for five years
Period: January 1992 to April 1998
The funds from this program were used to support several research projects including: a pilot study on
the Effective Force Testing method, research on high-strength concrete and its applications to seismic
regions, and partial support of an NSF PRESSS project on ductile frame connections described above.
The first project involved the development and implementation of a new experimental testing technique
for real-time earthquake simulation studies of large-scale structures. The technique, called “Effective
Force Testing” (EFT), complements existing forms of earthquake simulation techniques which include
shake table studies, quasi-static cyclic tests on structural components, and pseudo-dynamic tests. EFT is a
force-control method which enables real-time testing because the effective forces are known a priori. The
effective forces are simply a function of the mass of the structure at each floor level multiplied by the
ground acceleration. The effective forces are applied to the lumped masses of the structural model which
is attached to a fixed reference frame at the base. The response of the structure subjected to the effective
forces is identical to that of a structure subjected to a ground acceleration history on a shake table. In the
case of the Effective Force Testing method, the size of the structure is not limited by the shake table
(Dimig 1999). Successful implementation of EFT required a modification within the feedback loop to
correct for “velocity feedback” associated with the inability of the actuators to apply force near the
C-
2

structural resonance. The pilot study incorporating the velocity feedback correction led to successful
implementation of the method on a linear elastic single-degree-of-freedom system.
The high-strength concrete studies included investigations of the flexural, shear and bond characteristics
of high-strength concrete. A number of M.S. students were also supported on this project including J. M.
Dimig, J. Timm, C. Fasching, and D. Cumming. The following are some of the refereed publications
which have resulted from this project:
1. Fasching, C., and French, C., “Flexural Behavior of High Strength Concrete,”
ACI SP-176
Applications of
High Strength Concrete to Seismic Regions, 1998, pp. 137-178.
2. Dimig, J., Shield, C., French, C., Bailey, F., and Clark, A., “Effective Force Testing: A Method of Seismic
Simulation for Structural Testing,” J. Struc. Engrg., ASCE, 125(9), pp. 1028-1037.
3. Murcek, J., Shield, C. K., French, C. W., and Clark, A., “Effective Force Seismic Simulation for the
Earthquake Engineering Laboratory,” Proc., 11th WCEE, Acapulco, Mexico, June 1996.

Grant: CMS-9904110
Title: Influence of Lateral Load on the Stability of Masonry Walls
PI’s: Arturo E. Schultz and Henryk K. Stolarski
Organization: University of Minnesota
Amount: $276,006
Project Period: September 1, 1999 to August 31, 2002
This project comprises an integrated experimental and analytical program for the study of instability in
unreinforced and prestressed masonry walls subjected to out-of-plane lateral loads. Transverse loading
has the tendency to produce instability in unreinforced masonry because 1) it reduces wall cross-section
through flexural tension cracking, and 2) it produces lateral that gives rise to second-order moments (i.e.,
P-∆). Six stability tests of unreinforced masonry walls, and six bending tests of prestressed masonry
walls are planned, in addition to computational modelling using nonlinear finite element analysis tools.
The specimens for the stability tests are under construction, and the test setup is being fabricated. The
following publication has resulted from this project:
1. Schultz, A. E., Mueffelman, J. M., and Ojard, N. J. (2000), “Critical Axial Loads for Transversely Loaded
Masonry Walls,” Proc., 12
th
Int. Brick/Block Masonry Conf., Madrid, Spain, June 2000 (paper accepted).
Grant: CMS- 9810005
Title: Composite Interaction of Steel Frame Members and RC Walls under Seismic Loading
P.I.’s: Jerome F. Hajjar, Arturo E. Schultz, and Carol K. Shield
Amount: $109,935 + $5,000 for Res. Exp. for Undergraduates
Period: 15 September 1998 to 31 August 2000
This two-year project, to be completed in August 2000, is funded as part of the U.S.-Japan Research
Program on Composite and Hybrid Structures. This research includes the development of appropriate
finite element formulations to simulate the nonlinear transient dynamic behavior of 3D steel frames with
composite concrete walls, with a focus on the cyclic behavior of the composite interface. Parametric
analyses of this structural system will also be conducted. One Ph.D. student is working on this research.
Grant: CMS-9632506
Title: Seismic Behavior of Steel Moment-Resisting Frames with Composite RC Infill Walls
P.I.’s: Arturo E. Schultz, Jerome F. Hajjar, and Carol K. Shield
Amount: $243,896 + $10,000 for Res. Exp. for Undergraduates
Period: 15 September 1996 to 31 August 2000
This project, to be completed in August 2000 and funded as part of the U.S.-Japan Research Program on
Composite and Hybrid Structures, includes an experimental and analytical research program for the study
of the design and behavior of steel moment-resisting frames with composite reinforced concrete (RC)
C-
3

infill walls subjected to seismic excitation. Two one-third scale, two-story, one-bay frames are being
tested to determine the cyclic behavior of steel frame-RC infill composite wall systems. The two
specimens will differ in the detailing of the composite interaction. In addition, twelve full-scale cyclic
shear specimens, comprised of two steel wide-flange sections connected with two shear studs each on
either side of a concrete panel, were tested to quantify the strength of shear studs under cyclic shear and
axial tension. Principal variables included shear stress history, axial tension magnitude, and the
configuration of confining reinforcement (which mitigates cracking in the infill). Data analysis coupled
with linear and nonlinear system analyses, along with a prototype structure design, will serve to establish
preliminary analysis and design recommendations. To date, one M.S. student completed his thesis in
1999 and one Ph.D. student remains working on this project. In addition to internal reports, the following
publications are completed:
1. Schultz, A. E., Hajjar, J. F., Shield, C. K., Saari, W. K., and Tong, X. (2000). “Study of the Cyclic Interaction In
Steel Frames with Composite RC Infill Walls,” Pap. No. 2727, Proc., 12
th
WCEE, Auck., New Zealand.
2. Schultz, A., Hajjar, J., Shield, C., Saari, and Tong. (1998). “RC Infills in Steel Frames as Composite Systems for
Seismic Resistance,” Paper No. T186-2, Proc. of the 1st SEWC, San Francisco June 1998.
Grant: CMS-9416363
Title: The Effect of Composite Floor Behavior on the Failure of Steel Moment-
Resisting Connections
P.I.'s: Roberto T. Leon (Georgia Institute of Technology) and Jerome F. Hajjar (U of MN)
Amount:$65,000
Period: 1 September 1994 to 31 August 1996
This project, conducted at the University of Minnesota and directed by Hajjar, sought to determine the
effect of the concrete floor slab on the behavior of steel moment frame connections during the Northridge
earthquake. The experimental portion of this research included tests of three full-scale beam-to-column
subassemblages, including two with partially composite floor slabs. The corroborating computational
research included conducting detailed 3D inelastic continuum finite element analyses of the specimens
using ABAQUS to correlate to strains in the experiments, and conducting a complete set of 2D and 3D
static and transient dynamic second-order inelastic frame analyses of an existing four-story SMR frame
which suffered severe damage during the earthquake. This last research was partially funded by the SAC
Joint Venture, Phase 1 (PI: Hajjar). Primary findings include: (1) A comparison of the experimental
results of the bare steel specimen and the specimens having a composite floor slab clearly shows strains at
the bottom of the bottom flange, near the beam-to-column full penetration weld, that are generally two to
three times larger in the composite specimens than in the bare steel specimen. (2) The 3D continuum
analyses clearly show stress and strain concentrations at the bottom of the bottom flange near the beam-
to-column weld, and at the top of the bottom flange, at the center of the flange adjacent to the access hole.
(3) The frame analyses of the four-story structure clearly show that the results of 3D transient dynamic
inelastic frame analyses using a site-specific ground motion correlated closely with the detailed damage
documented in a forensic investigation of the structure, while results from linear elastic analyses, static
“pushover” studies, and 2D transient dynamic inelastic analyses did not correlate well with the damage.
One post-doctoral fellow, G. Forcier, and one masters-level research assistant, B. Gourley, were funded
by this project during its first year. Two additional M.S. students completed this project for their theses.
In addition to internal reports, the following refereed publications were produced:
1. Hajjar, J. F., Leon, R. T., Gustafson, M. A., and Shield, C. K. (1998). “Seismic Response of Composite
Moment-Resisting Connections. I. Performance. II. Behavior,” J. Struc. Engrg., ASCE, 124(8), 868-885.
2. Hajjar, J., Gourley, B., O'Sullivan, D., and Leon, R. (1998). “Analysis of Mid-Rise Steel Frame Damaged in
Northridge Earthquake,” Journal of Performance of Constructed Facilities, ASCE, 12(4), 221-231.
3. O'Sullivan, D. P., Hajjar, J. F., and Leon, R. T. (1998). “Repairs to Mid-Rise Steel Frame Damaged in
Northridge Earthquake,” Journal of Performance of Constructed Facilities, ASCE, 12(4), 213-220.

C-
4

Grant: EEC-9619750
Title: Undergraduate Experiences in Electrical Engineering at the University of Minnesota
PI Name: Douglas W. Ernie, with Co-PI: Lori E Lucke
Award Amount: $531,142 (Estimated) with additional $600,000 matching from U of MN
Period Covered: 1/1/97-12/31/01
As we enter the 21st century, it is essential that a more culturally diverse group of individuals be
encouraged to pursue careers in science and engineering. Consequently, the primary objective of this
REU site program is to provide women and minority undergraduates, and undergraduates from non-
research universities, with a challenging and rewarding introduction to research in electrical and computer
engineering. The program centers around a ten week intensive summer research experience in which
students are paired with faculty mentors, along with complementary educational activities. As a capstone
to their REU experience, each student prepares a final written report and gives a presentation at a
concluding poster session. Program participation to date includes 10 students in summer 1997, 12 in
summer 1998, 17 in summer 1999, and 15 for summer 2000. The results of several of the participant's
research have been submitted for presentation at professional conferences and, in a few instances, for
publication. This program has clearly enhanced the participants' interest in an engineering research career
and in attending graduate school, with our follow-up surveys indicating that over 90% of our participants
have gone on to graduate school even though only about 50% expressed such a strong interest before
attending our program.
Grant: CDA-9502979
Title: Applications over High-Speed Networks: A Pilot Project for the NII
PI Name: David H.C. Du with Co-PIs: Woodward, Yew and Kumar
Period Covered By this Report: 8/1/96-4/30/99
Award Amount: $1.5M with additional $600k matching from U of MN and $800k from IBM
The University of Minnesota has been awarded a pilot project for the development of a prototype
distributed computing, storage, and scientific visualization facility for the National Information
Infrastructure (NII). Key elements of the facility include (i) a distributed parallel storage system
consisting of several physically distributed RAID systems with high speed ATM connections and a high-
bandwidth disk system with direct connection to a Fibre Channel switch, (ii) a distributed computing
server consisting of clusters of high-end workstations (IBM RS/6000, SGIs, and SUN Sparcs),
multiprocessor(s) (IBM Scalable PowerParallel SP2), (iii) a high resolution display unit, and (iv) a high-
speed local, metropolitan, and wide area network infrastructure consisting of HIPPI, Fibre Channel and
ATM network switches. The two main objectives of this pilot project are the following: (i) supporting
existing and future NII applications over the aforementioned facility, and (ii) studying enabling
technologies which range from distributed application development and systems integration to high-speed
network inter-operability and connectivity. Both of these objectives will be achieved through the
construction of the prototype distributed, scalable, storage, computing, and visualization environment
over high-speed local and wide area networks.
C-
5

B RESEARCH VISION AND DESIGN OF THE NEES EQUIPMENT
B.1 Introduction
This proposal presents an integrated approach to the simulation of structural response of buildings and
bridges in moderate- and large-scale earthquakes, linking large-scale testing of structures with three-
dimensional nonlinear analyses of structural components and systems. The proposal focuses on the
creation of a new large-scale earthquake experimentation facility within the United States. The primary
equipment installation will comprise a Multi-Axial Subassemblage Testing (MAST) system. The MAST
apparatus enables multi-axial cyclic static and pseudo-dynamic tests of large-scale structural
subassemblages including portions of beam-column frame systems, walls, and bridge piers. The specific
equipment category of this proposal, as per the NSF NEES categories listed in Table 1 of the program
solicitation, is D, Large-Scale Laboratory Experimentation Systems.
Large-scale structural testing of structures or components of structures can deliver engineering insight
into structural behavior that cannot be realized by any other means. Often, the boundary effects where the
specimen couples to the reaction structure are reduced to simple uniaxial loading configurations not
necessarily representing the physical boundary conditions experienced in practice. Furthermore, the
difficulty of imposing multiple-degree-of-freedom states of deformation and load using conventional
structural testing means can be expensive, time-consuming, and difficult to achieve.
The MAST system concept as conceived by the University of Minnesota, employing an MTS six-degree-
of-freedom controller, can be used to apply realistic states of deformations and loading in a
straightforward and reproducible manner. With this system, a full six degree-of-freedom loading
conditions can be imposed on the test structure, thus providing a new capability that will enhance the
earthquake engineering community’s understanding of complex failure states. This system is unique and
will greatly expand the large-scale earthquake experimentation capabilities both nationally and
internationally. Due to the nature of the design, the system enables a high degree of flexibility for
configuring various test programs. With the inherent ease of test setup, overall flexibility, and the
incorporation of teleobservation and teleoperation, the testing community at large will greatly benefit
from this system.
To achieve a successful program in the most cost effective manner, the University of Minnesota and MTS
Systems propose a collaborative solution. MTS will provide the actuators, hydraulic distribution system
(including pump), digital controller with six-degree-of-freedom software (control system), and the
University will provide the lateral load resistance for the horizontal actuators through the development of
an L-shaped reaction wall system. The University of Minnesota will also design and oversee the
fabrication of the large steel weldments (crossheads) for the top and bottom reaction surfaces.
The main features of the MAST system and degree-of-freedom control are described in the following
sections.
B.2 Six-Degree-of-Freedom Control Concept and Technology
In most engineering applications, one thinks in terms of an absolute Cartesian coordinate system as
shown in the inset to Figure B.1 that can be described in terms of six degrees of freedom (6-DOF). That
is, three translational DOF (longitudinal in x, lateral/transverse in y, and vertical in z) and three rotational
DOF (rotations about the two horizontal axes, termed “pitch” and “roll,” and rotation about the vertical
axis, termed “twist” or “yaw”).
When using multiple actuators to position a platform or crosshead, it is impractical to control the system
by individually controlling each actuator. The required combination of actuator commands needed to
position the platform in space is complex. However, because the position and orientation of the platform
can be defined in terms of its six degrees-of-freedom (6-DOF), it is possible to program the desired
motion in DOF terms. The proposed MAST controller provides simultaneous DOF control in all six
degrees of freedom. It uses a proven real-time MTS controller that has been successfully installed at
C-
6

dozens of sites all over the world. The initial implementation of this technology was applied to all MTS
multi-degree-of-freedom Seismic Table Systems. It has also been used in a 6-DOF Hexapod testing
system for the automobile and defense industry. The 6-DOF control concept is directly applicable to the
proposed Multi-Axial-Subassemblage Testing (MAST) system to be operated statically or pseudo-
dynamically at the University of Minnesota.
To illustrate the concept, the seismic platform system shown in Figure B.1 will be used. From the figure,
it is evident that each degree-of-freedom is affected by one or more actuators depending upon actuator
locations (i.e., the actuators are coupled through the changes in geometry of the platform brought about
by displacement). Each actuator directly affects one or more DOF's. For controlled motion in all six
degrees of freedom, the individual actuators must each be coordinated with all other actuators affecting
the same degree of freedom.



















Figure B.1 Typical Actuator – DOF Coordinate System for Table Application
Consider a simplified 3-DOF system represented by only the four vertical actuators in Figure B.1. All
four actuators affect the vertical position equally and in the same direction, so in DOF control, vertical
feedback for this platform is the average of the vertical feedbacks of all four actuators. Pitch is affected by
actuators #1 and #2 in one direction and by #3 and #4 equally but in the opposite direction. So pitch
feedback is the difference between the average feedbacks of these two pairs of actuators. Roll is similar to
pitch, but with actuators #1 and #3 paired against #2 and #4. In DOF control, the feedbacks for each loop
are determined by summing together all individual feedbacks that contribute to that DOF, and each
actuator drive signal is determined by summing together all individual DOF error signals that are affected
by that actuator.
The aforementioned discussion is a simplified explanation of the 6-DOF control concept. Appendix 1
contains more detail of 6-DOF controller features required to operate the MAST system, including servo
compensation techniques for geometric cross coupling compensation and force balance compensation
which are briefly described here. Geometric cross coupling can be illustrated using Figure B.1. Lateral
movement of the platform controlled by the lateral actuators also requires deformations of the actuators in
the longitudinal and vertical directions to enable the lateral movement. Geometric cross coupling
compensation is used to control those required movements. The second type of compensation, force
balance compensation, may be illustrated considering only the vertical actuators in Figure B.1. Only three
C-
7

vertical actuators are required to control the position of the planar platform. Additional actuators, are said
to "over-constrain" the system. As such, the force control in the vertical actuators may be accomplished
with a variety of individual actuator loading combinations. Force balance compensation ensures that the
force is distributed equally among the driving actuators so that force is not wasted by the actuators
"fighting" each other or through the actuators trying to distort the platform.
The next section, describes the proposed MAST concept which features the 6-DOF controller.
B.3 Multi-Axial Subassemblage Testing (MAST) Concept
In the University of Minnesota Multi-Axial Subassemblage Testing (MAST) system installation, shown
in Figure B.2, the platform described above is replaced with a stiff steel crosshead in the shape of a
cruciform. Two sets of 200 ton actuator pairs, capable of applying a total force of 880 kips each pair with
strokes of +/- 16 in. will control the displacement of the crosshead in the longitudinal and lateral
(transverse) directions. The longitudinal and lateral (transverse) actuator pairs will be secured to an L-
shaped strong wall in the University of Minnesota Structural Engineering Laboratory described in
Sections C and E, and Appendix 1. Four 150 ton vertical actuators, capable of applying a total force of
1320 kips with strokes of +/- 20 in. would mount on large standoff tubes anchored to the strong floor with
universal swivels.




























C-
8

Figure B.2 MAST Pictured with Two-Story Two-Bay Frame Subassemblage Test
A combination of specially fitted back-flow vales, as well as a system of removable support struts will be
used to prevent the top crosshead from dropping vertically onto a test specimen when the MAST is not
being operated, or in case of power outages. Horizontal clear distance between the vertical actuators can
accommodate specimens up to 20 ft. in length in the two primary orthogonal directions. In addition, it is
possible to orient test specimens along a diagonal, which would increase the horizontal clearance
limitations by several feet in all directions. The vertical clearance is only limited by the height of the L-
shaped reaction wall to which the longitudinal and lateral crosshead actuators are attached. The vertical
clearance for the test specimens can be varied by repositioning the lateral and longitudinal actuator
attachments to the reaction wall. Figure B.2 shows the MAST system configured with the maximum
vertical clearance envisioned on the order of 25 ft.
The full six degrees-of-freedom would enable application of complex biaxial load histories on
subassemblages via control of the crosshead. Currently most biaxial load tests on subassemblages are
accomplished in one of two ways:
(1) Apply complete load history in one direction, then
rotate the test specimen in the test frame, and
apply complete load history in the other direction
(e.g., French 1989, Bugeja 2000).
Drift %, N-S
(2) Apply a clover-leaf type pattern similar to that
shown in Figure B.3, in which case the structure is
loaded in the longitudinal direction to a peak
displacement, then while maintaining that stroke,
the specimen is then loaded to a peak displacement
in the lateral direction. The next step is to bring
the specimen back to the original position in the
longitudinal direction, followed by bringing the
structure back to the original position in the lateral
direction. This displacement history forms one of
the “leaves” in a “four-leaf clover” pattern
(Cheung 1991, Kurose 1991, Bolong 1991).
Drift %, E-W
Figure B.3 Clover-leaf biaxial load history
The latter approach applies simultaneous biaxial loading, but because it has not been possible to couple
the control of the actuators in the two respective directions simultaneously, the loading history is limited
to changing displacement in one orthogonal direction at a time.
With the MAST system, it will be possible to program any displacement history, and at the same time, it
is also possible to control the axial load on the test specimen using mixed mode control capabilities. The
6-DOF control software employs advanced control technology to locate the position of the crosshead
through real-time simultaneous control of the eight crosshead actuators.
Another advantage of the MAST system is that it not only enables control of the position of a point in
space, it enables control of a plane in space. This feature makes it possible to apply biaxial control of
structures such as multi-bay subassemblages or walls. It also enables application of pure planar
translations, as well as the possibility of applying gradients to simulate overturning (e.g. axial load
gradient in the columns of a multi-bay frame, or wall rocking).
The system is also equipped with four 220 kip ancillary actuators with strokes of +/- 16 in. Each of the
ancillary actuators will have the option of independent master/slave control combinations with the
flexibility of slaving the actuators to scaled master control signals off of the 6-DOF controller. Using the
ancillary actuators to apply simulated gravity loading to test structures might be an example of a situation
when one would employ independent control of the ancillary actuators. An example of slaving the
C-
9

ancillary actuator(s) to scaled master signals off of the 6-DOF would be the case of using the ancillary
actuators to apply lateral displacements (or loads) to intermediate stories of multi-story subassemblages
tested in the MAST system. Another example of this control combination would be the case of employing
the ancillary actuators to control the beam end boundary conditions at assumed inflection points. In this
case, the ancillary actuators would be programed to maintain constant elevation based on the translational
DOFs of the MAST system.
B. 4 Teleobservation and Teleoperation of the MAST System
The proposed MAST System includes teleobservation and teleoperation capabilities to provide both on-
site and remote investigators with the same type and quality of information. The
teleobservation/teleoperation infrastructure will provide all relevant information needed for both
monitoring and interpretation of the experiments. As such, this facility will incorporate real-time
teleobservation of all visual monitoring information during an experimental run, real-time transmittal of
all acquired sensor data, and limited real-time teleoperation of hydraulic equipment and teleobservation
monitoring equipment. Note that real-time implementation, as referred to in this proposal, will for some
components be near-real-time, with a typical latency constraint of 2-3 seconds imposed by processing,
serving, and networking infrastructure, as discussed in Sections G and H. It is envisioned that an on-site
research fellow will participate in the execution of all experiments to ensure safety and accuracy in
execution of a remotely-operated experiment.
For the purposes of this proposal, a teleobservation and teleoperation system is proposed that is indicative
of the requirements needed for the MAST system; however, it is anticipated that the specific hardware
and software (e.g., graphical user interface) specifications and components for teleobservation and
teleoperation will be determined in conjunction with the NEES System Integrator. The plan, described in
Sections G and H, identifies key features of the teleobservation and teleoperation system envisioned by
the investigators, and highlights the specific qualifications of investigators Du and Ernie to implement
these capabilities.
It is proposed that teleobservation in the MAST system will be achieved primarily through a set of eight
high resolution digital audio/video cameras and eight digital still cameras spaced uniformly around the
perimeter of the three-dimensional specimen. Most importantly, to enable reconstruction of the
experiment for later analysis, and to enable subsequent coupling of experimental response with computer
simulations to facilitate model-based simulation, a complete archive of all data (including video, audio,
still images, and measured response) will be stored in a readily available format on a Visualization and
Archiving Server (described in Section H) for subsequent on-site and remote access.
The investigators are prepared to work with the NEES System Integrator to establish model-based
simulation, visualization, teleobservation, teleoperation, and data storage interfaces that allow integration
of the proposed MAST system into the complete NEES system.
B.5 NEES Vision
The vision of the NEES initiative is to advance seismic design and performance of the national
infrastructure through a networked national resource of shared-use next-generation research equipment.
This unprecedented investment in integrated facilities for earthquake engineering research has facilitated
the development of well-conceived experimental systems such as MAST that will enable the development
and validation of complex numerical and analytical models of structural behavior.
The process of accomplishing the NEES vision will be finalized in conjunction with the NEES
Collaboratory. What follows, is a conceptual example of how the MAST facility fits into an integrated
data-centric approach for experimentation, computation, theory, databases, and model-based simulation.
It is envisioned that the development of three-dimensional geometric models of structural systems will be
at the core of the integrated approach. These models may be used during experimental tests, conducted
with MAST, to display the measured response of the test structure obtained from sensors. The geometric
C-
10

model could illustrate the global view of the structure (e.g. deformed shape) as the test progresses. In
addition, the model will enable the observer to “click on” regions of interest to view local measurements
either numerically or visually (e.g. strains). Through the use of color, the geometric model might indicate
regions that may have been damaged. These features, desirable for an observer, are essential for
teleoperation. The features enable the remote operator to maintain an overall perspective of the global
structural behavior with the flexibility of focusing on regions of interest. The teleoperator will also have
control of the position of the video and still cameras through zoom, pan, and tilt. The control of these
cameras could be linked to the geometric model. For example, the teleoperator could select or “box” a
region on the geometric model to control the camera position and zoom.
Another feature of the geometric model is that it may be used to generate an analytical model of the
structural system that may contain combinations of force- and stress-based numerical models. These
numerical models would be used in three primary ways: (1) prediction of structural response; (2) control
of tests (e.g. by updating the structural model based on measured behavior); and (3) characterization of
the observed structural response.
The MAST facility is particularly well-suited to this integrated approach. It may be used to conduct cyclic
static tests to “fingerprint” the characteristics of structural subassemblages subjected to complex loading
histories. This information may then be mathematically modeled and incorporated into three-dimensional
nonlinear analyses to investigate the effect of different types of ground motions on structural systems
incorporating these details. At the present time, this has been realized for tests incorporating relatively
simple load histories. The MAST system provides flexibility to investigate more realistic structural
boundary conditions and enables an infinite choice of controlled test history that may be operated in
mixed modes. It will be possible to apply combinations of proportional loading (e.g. constant moment-to-
shear ratios for structural walls) or to use a feedback signal or combination of feedback signals to define
the input signal for another degree of freedom (e.g. axial loading on column or wall as a function of the
force in a lateral DOF that is specified as a drift history). The MAST facility may also be operated in a
pseudo-dynamic mode, as described above, either independently or in consort with other NEES
experimental facilities running tests in parallel on subassemblages representing other critical regions of
the same overall structural model (Liu, 2000a).
One of the most powerful features of the integrated approach to model-based simulation is the
accumulated database, which will feature the ability to “replay” tests. In this mode, the “observer” has
access to the wealth of data to develop and calibrate analytical and numerical models for model-based
simulation. One of the keys to the recreation of test results through model-based simulation is the
knowledge of the measured material properties of the components of the test structure and the tests
(typically ASTM) used to measure those properties. Other important database information includes the
structural dimensions, location of instruments, and identification of boundary conditions. As described
earlier, the geometrical model of the structural system can serve as a platform to access the overall
measured response of the structural system as well as local detailed measurements. In addition, the
geometrical model could serve as a platform to access other key information such as the structural
dimensions and measured material properties to be used in evaluating the performance of analytical and
numerical models. If a common platform is used for the storage of all experimental results from the NEES
Collaboratory this would be a powerful means to maximize the accessibility and usefulness of the data.
The geometric model in addition to “storing” and mapping the data and replay features could serve as the
basis for the development of the numerical model of the test structure. Users would have the option to
incorporate their own force- or stress-based models readily into the platform for calibration.
Even as model-based simulation becomes a reality through NEES, facilities such as MAST will continue
to be important in the validation of new analytical and numerical models to characterize the behavior of
structural components and subassemblages as new structural materials and concepts, not yet imagined, are
realized. As indicated above, researchers at the MAST facility will work to be consistent with the
direction of the NEES Collaboratory.
C-
11

B.6 Three Examples of Research Projects to Achieve Research Vision
The possibilities for structural testing with the MAST system are limitless. The following represent three
examples of types of structural configurations and variations on those concepts that could be tested with
NEES equipment to achieve the research vision. The loading history, described below as “user-defined,”
represents a multitude of options including using an input from a multidirectional pseudo-dynamic test
input, in which case, the tests described below might represent one component of a structural system
which is tested simultaneously at a number of NEES sites as described in Section B.5.
EXAMPLE 1- Beam-column subassemblage (multidirectional test).
The concept of multidirectional beam-column subassemblage tests was described in Section B.3. A
typical subassemblage is shown in Figure B.4, in which case the test specimen represents a portion of a
structure modeled between inflection points assumed to occur at midheight of the columns and midspan
of the floors. To represent the boundary conditions, the column ends would be attached to the MAST
crossheads with universal joints. Four ancillary actuators with universal joints would maintain the story
elevation of the beam ends as the column is subjected to the user-defined displacement history. The
MAST 6-DOF controller enables complex biaxial
displacement histories, while through mixed-mode control,
the axial load on the column can be controlled as well. Table
B.1 shows the required load and displacement demands to
test a biaxially loaded subassemblage that would be similar
in concept to one of the peripheral beam-column connections
in the NSF PRESSS project in which case the lateral load
resistance was provided by perimeter frames (Palmieri
1996). As a consequence, it is envisioned that the demands
listed in Table 1 are near the
maximum limit one would anticipate for a large-scale biaxial
subassemblage (assuming this connection is taking the
maximum loads in each direction). The lateral/longitudinal
loads and ancillary actuator loads are based on the
assumption that a beam-hinging mechanism develops. The
maximum displacements are associated with extreme drifts
(up to 8%, assuming 13ft. columns and 20 ft. beam spans).
The axial load represents the gravity load on a lower story
subassemblage in a 15-story building.
Figure B.4 Beam-Column Subassemblage
Example 2 – Multi-story, multi-bay frame (unidirectional test)
C-
12

The next example is a two-bay, two-story steel structure that was shown configured in the MAST Facility
in Figure B.2. Figure B.5 gives examples of possible variations in multi-story, multi-bay frame









Figure B.5 Examples of Possible Frame Configurations



alternatives that might be tested in the MAST Facility. These concepts could be expanded to include
multidirectional testing capabilities. The figures also introduce different concepts for using the ancillary
actuators (e.g. to provide simulated gravity loads to the floor systems, or to provide lateral load at
intermediate story heights). With the 20ft. clear distance between the vertical actuators of the MAST
system, the multi-bay structure as shown is limited to ½ scale (1:1 scale would correlate with
W14x311(50) columns and W33x150(50) girders). The scale of the test structure could be increased by
orienting the structure on a diagonal in the MAST Facility. This would not alter the means for controlling
the position and load on the test structure, because of the DOF control capabilities, and expands the
options for testing within the MAST facility.

Example 3 – Flanged wall (multidirectional test)
Post-earthquake reconnaissance often identifies building
corners formed by intersecting concrete or masonry walls as
vulnerable to seismic damage, and biaxial load effects are often
cited as the reason for the damage (Arnold 1989; Schultz 1994).
However, little has been done to quantify the biaxial loading
and wall resistance characteristics at these corners, nor to
systematically verify details to mitigate this damage. The
MAST system would enable full biaxial testing of full-scale or
near full-scale subassemblage tests of wall cross sections. The
sample reinforced concrete core wall shown in Figure B.6,
typical of an elevator core, represents the lower two stories of a
multi-story wall. To control the loading imposed on the wall
through the boundaries, it is envisioned that rigid concrete
blocks would be cast on the top and bottom of the wall to
transfer the load from the cruciform-shaped crossheads to the
flanged wall cross section. The 3:4 scale demands listed in
Table B.1 were scaled from a 10-story prototype with 12 ft.
stories. Concrete strengths of 4000 psi were assumed and a
vertical reinforcement ratio of 2% uniformly distributed
throughout the cross section.
Figure B.6 Flanged Wall with
Biaxial Moment Gradient
Assuming a moderate amount of inelastic behavior in the prototype, the centroid of the total lateral force
distribution at maximum base shear is assumed at mid-height of the wall, resulting in a moment-to-shear
(M/V) ratio of H/2, where H is the height of the prototype. Testing of this system would proceed with a
prescribed lateral drift applied along each horizontal direction to define the biaxial load pattern (e.g. user-
defined source, or pseudo-dynamic input). At the same, mixed-mode (multi-mode) control could be
employed to impose the desired moment to shear ratio at the boundaries in the two orthogonal directions.
The procedure might be as follows, a prescribed lateral drift is imposed simultaneously in the two
orthogonal directions. The resulting moments in the two orthogonal directions, caused by the actuators
applying longitudinal and lateral loads, apply a moment to the structure. The distribution of lateral,
longitudinal and vertical loads would be controlled via the 6-DOF controller to ensure that the desired
M/V ratio is maintained. The two remaining world DOF’s (e.g. twisting moment about vertical axis,
termed “yaw,” and the resultant vertical force or axial load) may be suppressed or controlled as well. As
an example, the vertical force might be specified as a constant or cyclic either independent or
synchronous with the longitudinal/lateral drift histories.
C-
13

It should also be noted that MAST is not limited to particular types of building materials. The three
example specimens described above might feature reinforced concrete, precast concrete, steel, and
masonry or masonry infills. There may also be new building materials or structural configurations, not yet
envisioned that could be tested within the MAST facility.
Table B.1 Anticipated Specimen Load and Stroke Demands
Dimensions
Longitudinal
Lateral
Axial
Ancillary
Specimen
Type
Column
(in)
Beam
(in)
Load
(k)
Stroke
(in)
Load
(k)
Stroke
(in)
Load
(k)
Load
(k)
Stroke
(in)
EX. #1
scale 1:1
40x42
18x40
±340
±13
±340
±13
1200
±220
±8
EX. #2
Scale 1:2
W8x67
F
y
=50ksi
W16x31
±150
±15
optional

1000
optional
EX. #3
Scale 3:4
15x15ft. in plan
9” thick
1

±650
// to web

±150
⊥ to web

900
optional
MAST
Capacity
20x20 ft. in plan
vertical 25 ft. (var.)
±880
±16
±880
±16
1320
+/- 20”
±220
±16
1
Flanged wall dimensions 15x15ft. in plan, 9 in. thick, with longitudinal loading parallel to the web, and lateral
loading normal to the web.
These three examples illustrate typical structural components and the appropriate scaling of a few
example structures for testing in the MAST facility. Although these are but three examples, the
possibilities are limitless. To explore this further, consider the third example structure, the flanged wall.
This specimen is one of many core wall sections used regularly in practice, but which are seldom studied
under multi-axial cyclic or monotonic loading due to 1) the complexity of the actuator control system, 2)
the large forces needed for large-scale specimens, and 3) the inability to control the test by maintaining
user-prescribed M/V ratios. As a consequence of the latter issue, to invoke realistic loading conditions,
full-height wall sections must be tested to supply the required loading to the heavily loaded bottom story
or two. The proposed MAST facility is innovative because it allows the full 6-DOF control, enabling the
use of economical subassemblage tests without sacrificing realistic loading conditions.
B.7 Equipment Acquisition Process
It is anticipated that most of the equipment to be purchased through this grant will go through the bid
procurement process. The proposers will provide detailed specifications on the requisition to ensure that
the equipment to be purchased will meet the necessary performance requirements. The following is a brief
summary of the University of Minnesota purchasing procedures.
To order goods, the department must submit a requisition to Purchasing Services. For purchases up to
$10,000, the department may request bids from vendors or may request that Purchasing Services take
bids, or may place orders immediately depending on the competitive nature of the items. For purchases in
excess of $10,000, Purchasing Services will attempt to obtain at least three bids in writing. If it is
determined that bids will be solicited, the requisitions should include detailed specifications for the
product. Every feature or condition that will be used in making the selection must be included on the
requisition. These will become the bid specifications.
B.8 Equipment Design Team Qualifications and Past Experience
MTS Corporation specializes in high performance loading systems including seismic tables and multi-
axial loading systems such as the proposed MAST. MTS has a proven track record for delivering high-
value custom solutions and design custom systems for many customers every year, such as the more than
50 large-scale shake tables in 17 different countries, vehicle simulation and testing apparatus, and high
performance, multi-DOF loading systems. There is a description of a number of recent MTS projects in
these fields in Appendix 2.
ESI Engineering is a consulting engineering firm providing specialized services on design for high force,
vibration control and isolation of buildings, equipment and unique structures. ESI has a broad range of
C-
14

qualifications in structural design and foundation engineering, and they have designed a number of highly
specialized laboratory facilities during the past five years (Appendix 2).
CNA Engineers offers professional engineering services to the construction and mining industries with
highly specialized expertise in underground construction. During the past five years, CNA has contracted
with the University of Minnesota to provide technical expertise on the construction and excavation for the
MINOS Project in the Soudan Mine (Section E).
B.9 Background of the Principal Investigators and Senior Associates
Five of the core structural engineering faculty and one of the geomechanical engineering faculty in the
Department of Civil Engineering at the University of Minnesota are involved in earthquake engineering
research. Two of the faculty members (French and Schultz) have conducted earthquake simulation
research on a shake table at the University of Illinois (French 1985; Schultz 1990). Two of the faculty
(French and Shield) have also been developing and working on the implementation of a new test method,
effective force testing (EFT), for real-time earthquake simulation of large-scale structures. All five of the
core structural engineering experimental researchers at the University of Minnesota have experience with
structural subassemblage tests, which are an economical means of characterizing the behavior of critical
regions in structural systems.
As has been typically done in other structural testing laboratories throughout the world, subassemblage
tests conducted at the University of Minnesota have been executed using a number of independently
controlled (or master/slave control) actuator combinations with a reaction wall or erector-type reaction
system. These “subassemblage testing systems” are limited to testing unidirectional behavior or very
simple biaxial load histories (such as those illustrated in Figure B.3).
It is from experience of the proposers with these systems, and the experience of one of the faculty
members (Schultz) with the testing limitations of the NIST TTF facility, that the proposed MAST system
using proven MTS 6-DOF control technology was developed. Furthermore, with the participation of
faculty from the Departments of Electrical and Computer Engineering (Ernie) and Computer Science and
Engineering (Du), the teleobservation and teleoperation capabilities can be developed seamlessly to
provide maximum benefit to the Earthquake Engineering research community in the U.S.
B.10 Role of NEES Equipment to Meet Goals of the Institution and the Earthquake
Engineering Community
The equipment described in this proposal, the Multi-Axial Subassemblage Testing (MAST) system, will
advance the state-of-the-art of the subassemblage test. As mentioned above, all five of the core
experimental researchers at the University of Minnesota have experience with subassemblage tests, but
the current means to test subassemblages is usually limited in scale (~ 1/3 to 1/2 scale) due to the capacity
of the actuator systems employed at most laboratories around the United States, and is typically limited to
the investigation of the unidirectional response of components. The proposed MAST system incorporates
large-capacity actuators to facilitate large-scale subassemblage tests. The system also features six degree-
of-freedom control of the crosshead to enable multidirectional testing.
There are currently two systems in the world which come close to meeting the requirements of the
proposed MAST system. One is the Tri-directional Test Facility in operation at the National Institute of
Standards and Technology (NIST) in Gaithersburg, MD. The second is the Test Facility for 3-D Beam-
to-Column Connections in operation at the Building Research Institute (BRI) in Tsukuba, Japan
(Nishiyama, 1998). One of the researchers at the University of Minnesota, Dr. Schultz, had worked for
several years at NIST where he had experience using the NIST TTF for structural concrete and masonry
wall tests.
In many ways, the NIST TTF is a unique piece of equipment because it anticipated trends in earthquake
engineering research (Woodward 1984). Like the MAST, the TTF utilizes two large steel crossheads, one
fixed and the other movable, and a system of hydraulic force actuators. A computer-based controller
C-
15

determines the necessary stroke of the actuators based upon the target global positions. Unfortunately,
adequate six-DOF digital controllers were not available commercially at the time of its construction for
static force testing applications. Consequently, the TTF cannot function in load control, and operates very
slowly. Furthermore, due to budgetary constraints and space limitations, the clearances (10 ft square
horizontally and 12 ft vertically), force capacities (150 kips vertically, and 220 and 150 kips horizontally,
in the major and minor horizontal directions) and stroke limits (±3in. or ±6 in.) are too small for testing at
full-scale or near full-scale.
The Building Research Institute (BRI) beam-to-column connection tester is a multi-DOF structural testing
system designed exclusively for beam-column subassemblages. The system comprises a fixed, self-
restraining steel frame with eight electro-hydraulic actuators which control forces and displacements at
the top of the column, as well as at beam ends. Ample horizontal and vertical clearances enable testing of
full-scale beam-column subassenblages. However, load capacities are too low for full-scale testing (440
kips of column axial load, 176 and 88 kips of column lateral force, and 88 kips of beam shear), and the
system is restricted to beam-column subassemblages.
B.11 Location, Safety and Access to the NEES Facility
The MAST system will be housed in the Structural Engineering Laboratory in the Department of Civil
Engineering at the University of Minnesota. The equipment will be available nationally to users as well as
to the faculty and students at the University of Minnesota. It is envisioned that tests using the MAST
system will be coordinated by the NEES System Collaboratory, in conjunction with Dr. Shield, the NEES
MAST Facility Coordinator and two Research Fellows, as described in Section F. It is envisioned that
external access to the MAST System may be accomplished in one of three ways: (1) The national NEES
system researchers may be housed on site in offices provided by the Department of Civil Engineering;
(2) Research students and staff at the University of Minnesota will be available to facilitate the
construction, instrumentation, and testing of the subassemblages on behalf of the external users; and (3)
In conjunction with the second option, the external users may conduct the test via teleoperation. Safety
issues for the MAST system have been addressed. Whenever the system is in use, an on-site research
fellow will supervise operations. In addition, a system of restraints (one permanent, one temporary and
adjustable), not pictured in Figure B.2 for clarity, will support the crosshead when the hydraulic supply is
off, both in the case of specimen setup and hydraulic shut off.
MAST as a system is not easily severable from the University of Minnesota because of the integration of
the MTS equipment and the reaction system. Individual components comprising MAST may be
severable; they include: 6-DOFcontroller, individual actuators, hydraulic power supply (pump), data
acquisition system, video cameras, and still cameras.
As indicated by the letter provided in App. 1, the University of Minnesota commits to the allocation of
space and infrastructure for housing and operating the NEES equipment and to providing national,
shared-use access to the NEES equipment by the earthquake engineering research community through
September 30, 2014. In addition, the Institute of Technology at the University of Minnesota commits
$875,000 in matching funds to support this endeavor.
C DESCRIPTION OF CURRENT MAJOR EQUIPMENT FOR EARTHQUAKE ENGINEERING
RESEARCH AND EDUCATION
Completed in 1982, the Structural Engineering Laboratory (SEL) is the focal point for earthquake
engineering related experimental research programs in the Department of Civil Engineering at the
University of Minnesota. A significant component of the experimental research has concentrated on
quasi-static cyclic testing of building subassemblages to investigate the behavior of structural systems
located in seismic regions. Additional research has focused on the development and implementation of a
new real-time seismic simulation test method for large-scale structures termed Effective Force Testing
(EFT) (Dimig et al. 1999).
C-
16

C.1 Description of Existing Major Equipment
The SEL houses a reaction system (i.e. strong wall and strong floor), numerous hydraulic actuators and
controllers, and two data acquisition systems. A summary of the major equipment housed in the
laboratory is provided in Table C.1.
C-
17

Table C.1 Description of Existing Major Equipment
Item
Model
Year Acquired
Reaction Systems


40ft x 80 ft. strong floor

1982
40 ft. x 40 ft. strong wall

1982
Hydraulic Loading Equipment


MTS Hydraulic Power Supply (22 GPM)
510.21
1985
Central MTS Hydraulic Power Supply (150 GPM)
506.82
1994
Central Hydraulic Distribution system (hardlines)

1994
4 Service Hydraulic Manifolds
293
1996-98
One MTS 220 kip +/- 5 in. stroke actuator
244
2000
Two MTS 110 kip +/- 10 in. stroke actuators
244
1998
Four MTS 77 kip +/- 6 in. stroke actuators
244
1983
Two MTS 35 kip +/- 10 in. stroke actuators
244
1983
MTS 600 Kip Universal Testing Machine

1989
MTS 100T/200C Kip Universal Testing Machine

1998
Controllers


Flex Test IIm Digital Controller

2000
Five MTS 407 Controllers

1996-99
Three MTS 406 Controllers

1983-85
Two MTS 458 Controllers

1989
Data Acquisition Systems


OPTIM Data Acquisition System
3008
1993
National Instruments Data Acquisition System

1998

C.1.1 Reaction Systems - The current structures laboratory has a 40 ft. by 80 ft. strong floor and a 40
ft. high reaction wall system. The current strong wall has tie downs at 40 in. horizontal and 48 in. vertical
spacings with the capability of sustaining 80 kips over 50 sq.ft. Currently, the base shear and moment
capacity of the wall is 450k and 1670k-ft, respectively. The strong floor has a tie-down pattern spaced at
40 in. centers with each hole set having a capacity of 100kips.
C.1.2 Hydraulic Loading Equipment - The hydraulic loading equipment in the structures laboratory
has been upgraded several times during the past 18 years. Original purchases of five fatigue-rated MTS
actuators were funded by an NSF equipment grant in 1983. Since then an additional four fatigue rated
MTS actuators have been purchased with funds from the University of Minnesota and funds generated by
contract testing. The capacity of the existing nine actuators are two 35 kip, four 77 kip, two 110 kip, one
220 kip with servovalves ranging in size from 10 GPM to 90 GPM. The laboratory also houses two
universal testing machines: a large 600 kip compression/tension universal testing machine with 15 ft.
vertical clearance and 6 ft. horizontal clearance purchased in 1989 through an NSF instructional
equipment grant, and a 200 kip compression/100 kip tension universal testing machine which was
donated to the University when the Bureau of Mines in Minneapolis was decommissioned. This machine
is currently undergoing substantial retrofit and is expected to become functional in the next 6 months.
Originally the hydraulic actuators were powered by a 21 GPM pump. In 1994 a 150 GPM pump was
installed with hard lines running from the pump to numerous stations underneath the laboratory floor to
power the hydraulic equipment in the structural engineering, rock mechanics, and pavements laboratories.
The pump and hardlines were funded by the Minnesota Department of Transportation in conjunction with
the University of Minnesota. Sharing of the hydraulic system in the SEL is accomplished through the use
of four service hydraulic manifolds, which allow for hydraulic shut-off (emergency or interlock) at the
manifold, instead of the pump, allowing multiple users of the pump without any possibility of unwanted
shutdown.
C-
18

C.1.3 Controllers - The laboratory has a variety of controllers to operate the hydraulic actuators and
universal testing machines. They include three MTS 406 controllers, five MTS 407 controllers, two MTS
458 controllers, and one four channel MTS FlexTest II digital controller. When needed, external analog
control signals for the 406 and 407 controllers can be produced by a six channel D/A card housed in a
Pentium computer. This system allows for simultaneous independent control of up to six actuators. The
current software (written in-house) is capable of producing ramp, stepped ramp and hold, sinewave, and
user defined control signals with a user-friendly interface.
C.1.4 Data Acquisition Systems - The laboratory has two data acquisition systems. The first
system, an OPTIM 3008 purchased in 1993, has the capability to monitor 56 strain channels, and 24
voltage channels at very high rates (up to 300 Hz when all channels are being used) with 16 bit resolution.
In 1997 a SCXI based National Instruments high data channel count data acquisition system was
purchased as part of an NSF Equipment Grant. As currently configured, the system is capable of
monitoring 96 strain channels (16 bit resolution), and 64 voltage channels (12 bit resolution) with a
scanning rate of 1 Hz. The user-friendly windows based software for the national instruments system was
written in-house, and provides significant flexibility with real-time access to the data being collected with
multiple options for processing and display. This system is modular, and the channel count can be easily
increased. Software enables the two data acquisition systems to be used simultaneously to collect the data
for a single test in a master-slave configuration (although access to the data in real-time is only available
on the National Instruments channels).
C.2 Physical Access to the Laboratory
Access to the laboratory is provided through a loading dock. The 30kip overhead crane can be used to
unload trucks in the loading dock area, and the crane has full access to the entire strong floor area.
C.3 Equipment Administration and Access to External Users
All of the equipment in the shared facility is administrated by a team comprising the Structures
Laboratory Coordinator (a faculty member, currently Dr. Carol K. Shield) and a full-time Research
Fellow (currently Paul M. Bergson). This team is responsible for assignment of space and resources
(equipment, undergraduate research assistant time, and Research Fellow time) to the projects utilizing the
laboratory. Prior to assignment of resources, all projects are reviewed by the Structures Laboratory
Coordinator, the Research Fellow and PI for schedule, space, equipment, undergraduate student support,
and Research Fellow support. After this review, space and resources for the projects are scheduled. The
Structures Laboratory Coordinators update resource and schedule information for ongoing projects
quarterly. Use of the laboratory equipment is extended to the community through external sales.
Engineers from 3M, Xerxes Composites, Lull/Omni Corp and MTS have used the equipment to improve
their products. Access to the equipment and laboratory personnel (undergraduate assistants, and research
fellow) is also provided to industry for contract testing.
C.4 Annual Equipment Usage and Downtime
The research fellow keeps logs for all of the hydraulic equipment in the laboratory indicating the use of
the equipment by project (number of cycles, load ranges, stroke ranges, hours on the pump). Table C.2
shows an estimate of the annual usage and downtime for the major equipment. Four of the six actuators
were sent for remanufacture when they were not being actively used on projects. The remanufacture of
the two 110 kip actuators was expedited with a downtime of approximately four weeks. The only other
equipment downtime over the past five years was during an oil change, when the entire hydraulic system
was offline for three days. Typically, at any given time, the strong floor area is 90% occupied, and 50% of
the actuators are in service.
C-
19

Table C.2 Annual Equipment Usage (U) and Downtime (D), %

600K
110K
77k
*

35k
*

Hydraulic system

U
D
U
D
U
D
U
D
U
D
1996
100
0
NA
NA
75
0
50
0
95
5
1997
100
0
NA
NA
50
0
50
0
95
5
1998
100
0
100
0
50
0
0
0
95
5
1999
98
2
100
0
50
50
0
0
90
10
2000
100
0
90
10
50
0
75
0
95
5
Avg.
99.6
0.4
96.7
3.3
55.0
10.0
35.0
0.0
94.0
6.0
*
If the Usage (U) + Downtime (D) does not total 100%, the difference indicates the percentage of time the
equipment was available but not in use.
C.5 Income and Expenses Associated with Equipment Services
The Structures Laboratory Coordinator manages the laboratory equipment, and supply and maintenance
accounts. A breakdown of maintenance costs and sources of income for the last two years is given in
Table C.3. Funds for maintenance, miscellaneous laboratory supplies, and new equipment are generated
by contract testing, set-up funds, and laboratory user fees. The major source of funding for laboratory
maintenance and supplies comes from the contract testing performed in the laboratory. Currently there are
no user charges for research related use of the equipment in the SEL. As shown in Table C.3, income for
the past two years was $87,755 for FY99 and $82,264 for FY00. The major maintenance costs over the
last two years were for an oil change for the hydraulic system and remanufacture of six of the hydraulic
actuators. The costs were $29,423 for FY99 and $15,327 for FY00. In scheduling maintenance, care is
taken to minimize equipment downtime.
Table C.3 Maintenance Costs and Sources of Support
Maintenance Costs
FY99
FY00
Oil Change
$4910

Actuator Rebuild
$22,013
$12,765
salaries/support personnel
$785
$785
Supplies/miscellaneous
$1715
$1777
Totals
$29,423
$15,327
Sources of Support
FY99
FY00
Contract Testing
$63,262
$82,265
Setup funds
$22,013

Other
$2480


The Department of Civil Engineering Shop is situated directly off the SEL and provides shop services for
the laboratory at a rate of $23/hr. The research fellow is funded 50% by hard funds from the Department
of Civil Engineering, and 50% by project related charges at a rate of $250/day including fringe.
C.6 Research and Educational Use of Existing Equipment and Its Impact
The SEL is the focal point for all experimental research in structural engineering at the University of
Minnesota, including infrastructure engineering, as well as earthquake engineering. The SEL supports the
activities of seven faculty members, a full-time research fellow, 10-15 graduate students, and 5-10
undergraduate students on a yearly basis. During the past five years, this equipment has been used in over
18 research projects listed in Table C.4, eight of the projects were directly related to earthquake
engineering research.
The impact of these projects on the earthquake engineering community has been substantial. As
examples, the impact of three projects will be described. The first project is the development and
implementation of a new test method for real-time earthquake simulation of large-scale structures, called
Effective Force Testing (EFT). This test method offers a fourth alternative to researchers conducting
C-
20

Table C.4 Research Projects During the Past Five Years
Project Title
Sponsor
Dates
Funds
Earthquake Related Projects



Development of Effective Force Testing: A Method of Seismic Simulation
*

NSF
1999-02
$342k
Influence of Lateral Load on the Stability of Masonry Walls
NSF
1999-02
$275K
Reassessment of Design Criteria and New Alternatives for Column Transverse
Stiffeners (Continuity Plates) and Web Doubler Plates
AISC
1999-01
$250k
Seismic Behavior of Steel Moment-Resisting Frames with Composite
Reinforced Concrete Infill Walls
NSF
1996-00
$254k
The Effect of Floor Slabs on Steel Moment Frame Connections
AISC
1995-96
$10k
The Effect of Composite Floor Behavior on the Failure of Steel Moment-
Resisting Connections
*

NSF
1994-96
$65k
Dynamic Testing of Viscoelastic Dampers
3M
1996
$20k
Faculty Award for Women
NSF
1991-96
$250k
Ductile Connections between Precast Beam-Column Frame Structures
*

NSF
1991-95
$140k
Non-earthquake related projects



Experimental Investigation of the Effect of Vertical Pre-Release Cracks
Mn/DOT
1999-02
$150k
Use of FRP Sheets to Retrofit Pier Caps
Mn/DOT
1999-02
$180k
Repair of Fatigued Steel Girder Bridges with Composite Fiber Strips
Mn/DOT
1999-01
$100k
Effect of Welded Stiffeners on Crack Growth Rate
Navy
1997-00
$128k
Performance Tests for Modular Bridge Joints
NCHRP
1997-00
$350k
Acoustic Emission Monitoring of Fatigue Cracks in Steel Bridge Girders
Mn/DOT
1997-99
$50k
Variability of GRFP-Concrete Bond
Mn/DOT
1996-98
$70k
Use of Non-Metallic Reinforcement in Concrete Structures
Mn/DOT
1994-96
$35k
High-Strength Concrete Applications to the Prestressed Bridge Girder Industry
Mn/DOT
1991-96
$195k
Tests of Two High-Strength Prestressed Bridge Girders
MNPa
1992-95
$20k
*
Due to space limitations impact of only these projects is included in this section
experimental studies on the behavior of structures under seismic loading (the other three test methods
being shake table, subassemblage, and pseudo-dynamic). The primary advantage of EFT is that it enables
real-time testing of large-scale structures using standard laboratory actuators (with high flow servovalves)
to further the understanding of structural behavior under dynamic seismic loads. EFT is an ideal testing
platform for researching the application of passive, active, and semi-active structural control schemes on
large-scale specimen. The second project, "Effect of Composite Floor Behavior on the Failure of Steel
Moment-Resisting Connections," represents an example of the exploration of the behavior of existing
structural details. Although the connection failures during the Northridge Earthquake were largely
attributed to material issues, structural considerations also may have contributed to the failures. The
project results indicated that the asymmetry inherent in the connections should be considered in design.
The project entitled, “Ductile Connections between Precast Beam-Column Frame Structures,” included
the development of new ductile connection concepts to enable the construction of precast concrete frame
systems in seismic regions. The behavior of the new connection concepts was investigated by conducting
tests on beam-column subassemblages incorporating the details. Numerical models were subsequently
developed to characterize the observed experimental results. By incorporating these numerical connection
models into nonlinear structural analyses, model-based simulation studies were conducted on a variety of
structural configurations, subjected to a variety of ground motions to broaden the impact of the results.
In the course of conducting this experimental research, the equipment described in this section was used
to fulfill the requirements of three PhD dissertations (one of which was earthquake engineering related),
and twelve MS theses (five of which were earthquake engineering related). Over the same time period,
these projects led to twelve refereed archival journal publications. Three of these publications were
earthquake engineering related. As the list of projects reported above indicates, the equipment described
in Table C.1 has been used extensively over the past five years.
C-
21

D PROJECT MANAGEMENT AND SCHEDULE (10/1/2000-9/30/2004)
Successful installation of the MAST Facility at the University of Minnesota will require a well managed,
integrated approach. The four main components of site installation include design and construction of a
new reaction wall, strengthening of an existing reaction wall and floor, installation of the hydraulic,
control, and structural testing system (the MAST system), and installation of the
teleoperation/teleobservation/networking system.
D.1 Project Management Team
Figure D.1 shows an organizational chart of the University of Minnesota management team for the
installation of the MAST System. The overall effort will be coordinated by the PI, French. Dr. French has
over 20 years of experience in experimental earthquake engineering research. In addition to being
responsible for overseeing the installation at the University of Minnesota, including tracking costs,
schedule and quality control, she will also be responsible for coordination with the NEES System
Integrator. Much of the day-to-day project management will be performed by Anna McDonagh, the
Assistant Director for Design and Construction Services at the University of Minnesota’s Department of
Facilities Management who will act as the general contractor for the installation. In this role, she will be
responsible for costs, schedule, and quality control of the subcontractors' work. Ms. McDonagh, who has
extensive experience in similar-scale multi-component, multi-million dollar equipment installations at the

NEES MAST Facility
Installation Coordinator:
C. W. French
Principal Investigator
Investigator
A
dvisory Group:
C. K. Shield
J. F. Hajjar
R. J. Dexter
Teleobservation /
Teleoperation
Installation:
D. Ernie
D. Du
MAST
Installation:
A. E. Schultz
Teleobservation /
Teleoperation
Installation
Research Fellow
#1 (50% time)
MAST Control
and Data
A
cquisition
Research Fellow
#1 (50% time)
MAST Structural /
Hydraulic
Installation
Research Fellow
#2 (100% time)
CNA, Inc. and
ESI, Inc.
Design/Build:
Construction of
strong wall/floor
MTS Systems
Corporation:
MAST installation
NEES
System Integrator
NEES MAST Facility
Project Management:
A. McDonagh
A
sst. Director, Design and
Construction Services
Facilities Management
U. of Minnesota
Infrastructure
Management:
Network / Utilities

Figure D.1 NEES MAST Facility Organizational Chart
University of Minnesota (Section E), will be directly responsible for coordinating the design and
construction of the reaction walls with the contractors. She will also coordinate with MTS for a timely
C-
22

installation of the MAST system after completion of the wall and wall retrofit, including all needed
modifications to the existing facility for water, electrical power, hydraulic power, and network.
Technical coordination for the design of the wall and final design of the MAST system will be provided
by Co-PI Schultz. Dr. Schultz has more than 15 years of experience in earthquake engineering
experimental research including the operation of a smaller version of the MAST System at NIST (the
NIST TTF) for several years. His responsibilities include overseeing the technical specifications of the
control and data acquisition systems, as well as the hydraulic and structural systems. To perform this task,
Schultz will have one full-time research fellow and one half-time research fellow working under him. The
full-time research fellow will be responsible for the structural and hydraulic systems, and the half-time
research fellow will be responsible for the control and data acquisition systems.
The teleobservation, teleoperation, and network installation will be overseen by Ernie and Du. Dr. Ernie
is a member of the Department of Electrical and Computer Engineering and has extensive experience in
broadcast television and internet streaming video technologies. He currently manages the four-channel
ITFS educational broadcast system (UNITE Instructional Television) providing over 180 hour per week
of graduate credit level programming to distance learners. Dr. Du is a member of the Department of
Computer Science and Engineering. His research is in the area of video and audio streaming. He has
implemented a controllable software architecture for a Video-on-Demand (VOD) server (Liu, 2000b).
Directly under Ernie and Du’s supervision will be a half-time research fellow responsible for the
development and installation of the teleobservation and teleoperation systems.
In addition to the core management team, three other Co-PI’s (Dexter, Hajjar, and Shield) will act as an
advisory group. Dexter has 19 years experience in the design of steel structures and structural research,
including large-scale fatigue and fracture tests. He will be responsible for providing input on the design
of the structural steel (e.g. crosshead) in the MAST system. Dr. Hajjar has 13 years experience in
visualization and model-based simulation. He will be providing input on the design of the teleobservation
and teleoperation systems, and the integration of these systems with model-based simulation. Dr. Shield
has five years of experience as the Structures Laboratory Coordinator, and was responsible for the
hardware and software development of the existing National Instruments D/A system.
The University of Minnesota project management team will cooperate with the NEES System Integration
awardee through September 2004 to facilitate the interconnection of all NEES equipment installations, the
development of the NEES Collaboratory, and the development and implementation of the NEES data
protocols. The team is committed to NEES Collaboratory participation through at least September 2014 to
facilitate integrated management, shared-use access, and remote operation of the facility, and will
participate in all outreach and training activities developed by the Collaboratory.
D.2 Anticipated Problems and Solutions
Because there are several subcontractors involved in the design and installation of the MAST system,
potential problems in the installation revolve around schedule and budget. To alleviate potential
problems, the University of Minnesota will be responsible for ensuring the quality control of the facility
and equipment. The project management team will also be responsible for monitoring the costs and
schedule, conducting NSF review meetings, and preparation of the annual reports. Plots of cumulative
expenditures in manhours and cost will be made each month and compared to the planned level and rate
of expenditures. Any significant deviation from the schedule or from the planned expenditures will be
investigated and resolved. Tight specifications for all major equipment will be developed to ensure that
equipment that is purchased or built in-house meets the performance requirements. Costs and progress on
award performance milestones will be reported annually to the NSF through September 2004 at least 90
days prior to the end of the budget period.
With respect to management of risk, one of the risks for the MAST Facility is the schedule for feasible
teleoperation. It is anticipated that teleoperation capabilities will be implemented in a staged fashion.
Operation of the MAST Facility will be always possible through direct control of the MAST system by
C-
23

the on-site research fellow. No matter what level of teleoperation is utilized, the research fellow will be
on-site for all experiments. The next level of teleoperation will be by via the Control and Data
Acquisition Server (described in Section G). This level of operation is anticipated soon after
commissioning of the facility. The third level of teleoperation involves controlling the experiment from
an on-site Client Server (described in Section G). This level of operation will minimize latency problems
associated with the network. The final level of teleoperation will implement control of the experiment
from an off-site Client Server. Assessment of feasibility of the different levels of teleoperation will be
made by Du, Ernie, and the MAST Facility Coordinator.
A second risk associated with the MAST Facility is possible compliance effects of the crosshead and
structural walls on the control of the actuators. The design of the crosshead and structural walls is
primarily stiffness-controlled to ensure that displacements of these elements resulting from actuator loads
will be much smaller (order of 5%) than the displacements of the test specimen. If there is substantial
movement of the reaction system, then the correct displacements will not be imparted to the structure at
the DOFs that are being used in displacement control, as the controller is currently configured. This
problem exists when the LVDTs that are internal to the actuators are used for displacement control
feedback. Sydeski, of MTS Systems Inc., indicates that the controller can be modified to use external
LVDTs referenced to a fixed frame for displacement feedback, should compliance of the reaction system
be found problematic. The University of Minnesota has a tradition of working with MTS to overcome
control problems in structural testing. As an example, MTS worked with Civil Engineering Faculty to
overcome control problems in an accelerated pavement tester. Changes in spreader compliance during
the application of a moving load caused unanticipated problems with the control of the test. After
consultation, MTS personnel developed a new control algorithm, coupled with a minor modification to
the loading fixture, which resolved the problem.
D.3 Schedule
Figure D.2 shows a chart of the schedule of the major tasks as well as principal milestones associated
with the proposed project. Sufficient contingencies have been built into the schedule and budget to
handle the typical technical problems and delays that will inevitably occur in a project this complex.
There should therefore be no problem meeting the schedule and level of effort of the project as specified
in this proposal.
C-
24


Figure D.2 Gantt Chart
E CONSTRUCTION, INSTALLATION AND COMMISSIONING
E.1 Plans
Construction and installation activities include: reaction wall and floor strengthening and construction,
and installation of MTS equipment and the network. Milestones are identified in Section D.
E.1.1 Construction and Strengthening of Reaction Wall - Strong wall and floor construction
includes the retrofit of the existing (East) wall, the retrofit of the floor below the MAST setup, and the
construction of the new (South) wall. Retrofit of the existing wall includes excavation and installation of
rock anchors, removal of concrete block partitions behind the wall, relocation of mechanical and electrical
conduit, construction of new post-tensioned concrete buttresses, electrical and mechanical upgrades for
the MTS pump, and strengthening of front and back walls. The retrofit of the strong floor includes
excavation below the existing cellular floor, installation of rock anchors, and construction of additional
wall footings. Construction of the new wall includes the removal of concrete floor, excavation to
bedrock, installation of rock anchors, and construction of the post-tensioned concrete buttressed wall.
More detail on these construction activities is provided in Appendix 1.
C-
25

The construction activities associated with this item require access to and use of the space behind the
existing strong wall, and the space adjacent to the reaction floor in the south corner of the Structural
Engineering Laboratory (SEL). Once this activity is completed, the space behind the East strong wall can
be returned to its previous use. Unit cost and estimated quantities are used to define total cost.
The design, construction and commissioning of these items will follow standard procedures developed by
the University of Minnesota for capital projects. This procedure includes a predesign phase (define
requirements, explore alternatives, select project team), a design phase (establish contract with design
firm, review and approve final design and cost estimates), a pre-construction phase (development of
construction documents, bid process), and a construction phase (building permits, relocation, construction
and commissioning). Throughout this process, an “owner’s representative” from the Facilities
Management Division of the University of Minnesota (A. McDonagh) will oversee the process, including
commissioning of the facility. The required systematic testing of the constructed systems to ensure
compliance with the specifications for this facility will be established in the Predesign and Design Phases,
and will be carried out by the PI and co-PI’s with supervision from the owner’s representative.
Potential problems with this activity include the relocation of current activities (primarily the graduate
student offices behind the existing strong wall). However, with advanced planning, students can be
assigned to offices at other locations in the building prior to the beginning of this activity. Another
potential problem concerns soil instability in the excavation work below the laboratory strong floor. One
of the consultants who assisted in the development of this proposal (CNA Engineers) has ample
experience in this area and has been able to control vertical soil faces using a combination of groundwater
drains, grout injection and reinforcing sheets. A third potential problem is access to the SEL for
construction equipment and materials. The existing truck loading dock and overhead crane are adequate
for this purpose. All of the planned activities are commonly done to renovate and make building additions
on the University of Minnesota campus.
E.1.2 MTS Equipment Installation - Installation of the MTS equipment includes delivery and
temporary storage of the actuators, crossheads, pump, and electronic equipment. Once the strong floor
and wall construction has been completed, the MTS pump, hydraulic service manifolds and pressurized
oil distribution lines (i.e., hardlines) will be installed. This will be followed by attachment of the lower
crosshead to the laboratory floor, installation of the vertical actuators, installation of the top crosshead
(with temporary supports), installation of the horizontal actuators, and installation of the MTS electronic
control equipment.
This activity requires access to and use of the space at the East end of the SEL, which will be dedicated to
the proposed MAST facility. Other experimental research activities using that space which have not
reached completion by the time the proposed construction begins, will be relocated to other parts of the
Structures Laboratory, other rooms in the Civil Engineering Building, other buildings in the University of
Minnesota campus, or other space leased by the University of Minnesota.
Cost basis for the MTS equipment is lump-sum estimate included in the price quotation from the MTS
Systems Corporation. Plans for the installation and commissioning of the MTS equipment will be
developed by the PI and co-PI’s in conjunction with the MTS design team. Supervisory oversight from
the Facilities Management owner’s representative is expected.
E.1.3 Construction Needed for the Network - Construction needed for the network consists of
three components. The first component is the installation of a 150-foot conduit run from the control room
to a network access panel outside of the NEES laboratory. The second component is the installation of
the fiber into the conduit and the termination of the fiber into fiber termination boxes. The third is the
installation of an additional fiber run in existing conduit from the network access panel to the MDF (Main
Distribution Frame) in the basement of the Civil Engineering building. There are no known problems to
be overcome with this construction/installation. Cost basis for this work is unit cost and estimated
quantities.
C-
26

E.2 Experience
Three projects involving similar-scale research equipment, have been undertaken in recent years by
various units of the University of Minnesota. These projects have involved substantial technical and
managerial expertise, as well as funding, and are described below.
E.2.1 Recent Projects
a)
“MINOS” Project
- The MINOS (Main Injector Neutrino Oscillation Search) project is designed to
provide confirmation that neutrinos change type as a result of oscillations accompanying their
propagation through space. In this four-year project, which is presently in its second year, neutrinos
are emitted from Fermilab (Illinois) towards the Soudan Mine in northern Minnesota (750 miles
away), where an extremely sensitive and unique detector is located. The project involves 21
institutions and more then 100 physicists from the US, UK, Greece, Russia, China. The University
of Minnesota, in cooperation with Fermilab, manages the entire project. Fermilab operates the
neutrino gun, while the University of Minnesota is responsible for excavation in the Soudan mine,
construction of the detector and equipping the site with the necessary electronics. The detector
consists of 243 octagons which are 8 m (26 ft) in size and comprise two layers of 25-mm (1-inch)
thick steel plate, for a total of 5,400 metric tons. Earl Peterson, Professor of Physics at the
University of Minnesota manages the project, Anna McDonagh, Assistant Director, Facilities
Management, is the owner’s representative, and CNA Engineers, located in Twin Cities, is the main
contractor on the project. The total cost for the Minnesota portion of the project exceeds $56M
($10.81M for magnets and steel, $23.325 for the Scintillator detector, $5.273M for electronics and
data acquisition, $6.691M for the ‘far detector’ installation, $2.937M for the ‘near detector’
installation, $1.547M for project management, and $6M for excavation and site preparation).
b)
Institute of Technology (IT) Characterization Facility
- The IT characterization facility provides
researchers at the University of Minnesota with state-of-the-art material research testing equipment.
It was originally part of the Center for Interfacial Engineering (CIE), one of the NSF Engineering
Research Centers affiliated with the Department of Chemical Engineering and Material Science at
the University of Minnesota. With the end of the CIE in the Fall 1999, the facility became the IT
Characterization Facility, which is managed by Dr. Stuart McKernan. Construction and equipment
installation are overseen by the Facilities Management Division of the University of Minnesota.
During the past five years, the IT Characterization Facility has acquired an x-ray diffractometer
($0.60M), and x-ray analyzer ($0.20M), a scanning atomic force microscope ($0.15M), and a new
analytical scanning electron microscope ($0.60M). Installation of these items has required $0.20M
in construction costs, for a total of $2.75M during the past five years.
c)
Center for Magnetic Resonance Research
- The Center for Magnetic Resonance Research,
directed by Dr Kamil Ugurbil, was conceived as state-of-the-art magnetic resonance research
facility, serving the entire community of the University of Minnesota. It began in 1996 when a grant
of $1.190M was awarded by NIH to the University of Minnesota and the $8M capital project was
approved and constructed. In 1997, a 25-ton magnet, 7-Tesla, was purchased ($2.745M) and
installed ($8M). The National Science Foundation also provides support for this center.
E.2.2 Facilities Management (FM) - The University of Minnesota Facilities Management Division
has ample recent experience in the management of capital projects. In 1998, the University began a
multi-year capital campaign effort to revitalize and modernize the Twin Cities campus. More than
$400M has been earmarked to build, expand, and/or renovate numerous University facilities over a four-
year period. A total of eleven capital projects have been completed, including the construction of 6 new
buildings, and the renovation of 5 existing buildings. Another nine capital projects are underway,
including four new buildings and five renovated buildings. Throughout this process, the Office of
Facilities Management of the University of Minnesota has overseen a perfect record of successful project
completion (on-time, within budget and accident-free).
C-
27

E.2.3 Network and Telecommunications Services (NTS) - The University of Minnesota Office
of Networking and Telecommunications Services has provided networking installation and support on
campus for over 10 years. NTS has designed, planned and installed networks ranging in size from the
campus backbone to whole buildings down to single rooms. NTS supports the campus phone system and
provides direct service to over 20,000 networking customers and ISP service to another 500 customers on
campus. NTS has successfully supported a high speed ATM OC3c backbone for the past 4 years, which
includes an OC3c to Chicago for access to the vBNS.
E.2.4 ESI Engineering - ESI Engineering is a Minnesota based consulting engineering firm providing
specialized services on design for high force, vibration control and isolation of buildings, equipment and
unique structures. ESI has a broad range of qualifications in structural design and foundation
engineering. Services in structural and foundation design include: strong floors, reaction masses and
foundations to resist high forces; preparation of construction drawings of isolation mass and foundation
systems for equipment that produces shock and vibration; analyzing shock and vibration problems in
structures and equipment; and measuring vibration levels.
E.2.5 CNA Engineers - CNA Engineers offer professional engineering services to the construction and
mining industry in the areas of geotechnical, civil, structural and construction engineering. Most, but not
all, of the company projects are constructed underground. These projects include tunnels by open-cut and
mined methods for utilities, highway, storage, water, sanitary and storm sewers; underground space for
industrial and office use, laboratories, and storage; and foundations design for dynamic loadings. CNA
Engineers was the designer of the deep excavation for the Civil Engineering Building.
E.2.6 MTS Corporation -

MTS specializes in high performance loading systems such as the proposed
MAST, and has a proven track record for delivering high-value custom solutions. The University of
Minnesota has worked closely with MTS Corporation on many projects over the past 25 years, including
partnering to jointly design the MTS 815 Rock Mechanics Testing System. These systems are still
supplied after 25 years and have had a significant impact on the understanding of rock mechanics.
F EQUIPMENT MANAGEMENT AND OPERATION (Post-Construction)
This section describes the management plan for the maintenance and operation of the MAST Facility
from the time it becomes operational until September 30, 2004. It is envisioned that a similar
management plan and organizational structure will continue through September 30, 2014.
F.1 Equipment management and operation
The MAST Facility organizational structure is shown in Figure F.1. The management and operation team
for the MAST Facility will consist of a MAST Facility Coordinator, one full-time research fellow, one
part-time research fellow, and undergraduate research assistants as needed for operation of the equipment.
In addition to the management and operation team, the facility will provide Technical Liaisons. It is
envisioned that the MAST Technical Liaisons (French and Schultz) will interface with the NEES
Consortium, the MAST Management and Operation Team, and the System Users, to provide the technical
expertise to communicate the range of testing capabilities available through the MAST Facility, in
addition to implementation detail associated with testing within the MAST facility. With their intimate
involvement in the specifications of the MAST system, extensive past experience in subassemblage
testing and experimental techniques, as well as, Dr. Schultz's previous experience with the NIST TTF
facility, they will make an ideal team for Technical Liaison.
The management and operation of the MAST Facility will be conducted by the MAST Facility
Coordinator, Shield, who currently serves as the Structural Engineering Laboratory Coordinator. The
MAST Facility Coordinator will review those projects sent forward by the NEES Consortium for
schedule and resources, and will report back to the Consortium and System User (Project PI) on the
availability of resources (space, equipment, undergraduate research assistants, and research fellow time)
in order to schedule the selected projects. The MAST Facility Coordinator will also have responsibility
C-
28

NEES MAST Facility
Technical Liaisons for
Experimental Testing:
C. W. French
A. E. Schultz
NEES MAST
Facility Coordinator:
C. K. Shield
Director, Structural
Engineering Laboratory
MAST Structural /
Hydraulic Facility:
Specimen Testing and
Facility Maintenance
Research Fellow #1:
Paul M. Bergson
Undergraduate
Research
A
ssistants
NEES Consortium
(after 9/30/04)
Undergraduate
Research
A
ssistants
Project Principal
Investigator at
University of
Minnesota or
Remote Location
Graduate Research
A
ssistants
at University of
Minnesota,
Remote Location,
or Both
Undergraduate
Research Assistants
at University of
Minnesota,
Remote Location,
or Both
System Users
Model-Based Simulation
Teleobservation / Teleoperation
Network: Development and
Maintenance
Research Fellow #2

Figure F.1 NEES MAST Facility Organizational Chart
for maintaining all accounts for the use and maintenance of the MAST Facility and overseeing all system
maintenance. It is estimated that the MAST Facility Coordinator will spend 10% of her time annually on
the management of the MAST Facility.
The two research fellows will report directly to the MAST Facility Coordinator. They will be responsible
for the daily operation of the system, interacting with the visiting research teams, training, performing
routine maintenance of the system, and supervision of the undergraduate research assistants. The full-time
research fellow will be responsible for the use and maintenance of the structural testing system, including
coordination of delivery of test specimen and supplies, installation of the test specimen into the MAST
Frame, and maintenance of the hydraulic, structural and control systems. It is estimated that 80% of this
research fellow's time will be spent on specific MAST project related activities (to be funded by the
projects after September 2004), and 20% of his time will be spent on coordination, maintenance, and
safety activities (to be funded by user fees after September 30, 2004). The part-time research fellow will
be responsible for the use and maintenance of the teleobservation, teleoperation, networking, and data
systems. This Research Fellow will also likely work closely with the NEES System Integrator to ensure
seamless access over the network to data from the MAST Facility. It is estimated that the part-time
research fellow will spend an average of 15 hours per week maintaining the teleobservation,
teleoperation, networking and data systems (funded by user fees after September 30, 2004), and 5 hours a
week on project specific tasks such as training (funded directly by projects after September 30, 2004).
Funding for both of these research fellows prior to September 30, 2004 is included in this proposal
because it is difficult to estimate the exact date that the facility will be operational.
F.2 Plan for Assessing Equipment Performance and Management
Assessment of initial hardware performance will be accomplished by conducting shakedown tests on an
elastic system (e.g. steel column). The column will be subjected to combinations of force and
displacement control, eventually testing all combinations of the six DOFs. In addition to using this test
C-
29

for the shakedown of the hardware system, it will also be used to ensure that the network infrastructure
described in Section G is operational.
The two research fellows will be responsible for keeping logs of equipment usage and performance.
These logs will be reviewed quarterly by the MAST Facility Coordinator. The information in these logs
will be used to ensure that recommended maintenance is performed on the system. Funds for
recommended maintenance will be generated through user fees.
Network performance will be assessed by running applications locally and recording network usage
requirements. This data will then be used as a benchmark with which to compare the information
recorded during remote client sessions to gage network performance. If local network congestion is
found, steps will be taken to alleviate the problem by provisioning more bandwidth. If network
congestion is found in the wide area network, appropriate parties will be contacted for problem resolution.
The plan for assessing management of the operation of the MAST Facility includes two key components.
The first component will be the establishment of an advisory group to assess the performance. This group
will consist of personnel from the University of Minnesota Vice President for Research’s office,
University of Minnesota Network and Telecommunications Services, MTS Systems Corporation, and a
faculty member at the University of Minnesota associated with the MINOS Project (a similar-scale
research project at the University of Minnesota). The second means of assessment will be through
documentation of any deviations from project schedule. All PI’s will be required to submit a project
schedule to the MAST Facility Coordinator. The Coordinator will evaluate the cause for deviations from
schedule. If delays were due to management, the sources of these delays will be corrected.
F.2 Facility management
Because the MAST Facility will be located in an on-campus University of Minnesota building, all
building management will be provided by the University of Minnesota Department of Facilities
Management. This organization is already responsible for maintenance of the loading dock facility
(including loading dock doors), overhead crane in the existing laboratory, and power, water and other
services coming into the building.
F.3 Schedule through September 30, 2004
The system installation and performance testing is scheduled to be completed during the fourth year.
After this time, the facility will be available for use by other University of Minnesota researchers, or for
shared use. Determination of the schedule for operation will be made in consultation with the NEES
System Integrator.
G SHARED-USE, TELEOBSERVATION, AND TELEOPERATION
Real-time teleobservation of all visual monitoring and sensor measurement information during an
experimental run, along with real-time teleoperation of the monitoring equipment, is an integral part of
the proposed NEES facility. The investigators anticipate identifying appropriate specifications for all
hardware and software for the teleobservation and teleoperation system for the MAST Facility, and to
work with the NEES System Integrator to implement and develop this system. The proposed equipment
is based on hardware and software components readily available or which will be available within the
next 12 months, although upon ordering in 2002 or 2003, it will be possible to take advantage of the
considerable enhancement in capabilities. Where applicable, proposed specific models of equipment are
listed in Appendix 3. Specific equipment that is remotely-operable has been identified for all
teleobservation capabilities discussed below. Note that real-time implementation, as referred to in this
proposal, will for some components be near-real-time, with a typical latency constraint of 2-3 seconds
imposed by processing, serving, and networking infrastructure.
For the purposes of this proposal, the following hardware and interface described below and shown in
Figure A4.1 (Appendix 4) are proposed to identify the specific characteristics necessary for
C-
30

teleobservation and teleoperation of the MAST Facility. Teleoperation of the MAST Facility will be
handled through a Control and Data Acquisition Server, which will consist of a high-end PC. A
description of the control interface provided with the MAST Facility is included in Appendix 1. An
intelligent web browser on the Client Server will include the graphical user interface for teleoperation of
the MAST Facility. As discussed in Section C, the investigators have substantial experience at
developing and implementing a Control and Data Acquisition Server for the current testing that takes
place in the Structural Engineering Laboratory. This past experience will facilitate robust and safe
implementation of teleoperation. It is also envisioned that an on-site research fellow will participate in
the execution of all experiments to ensure safety and accuracy in execution of a remotely-operated
experiment.
The teleobservation needs for this NEES facility can be divided into two broad categories: data
measurements from sensors (e.g., strain gauges, position sensors) and visual information (e.g., digital
video and still cameras). Teleobservation of the first category will be handled through the Control and
Data Acquisition Server. A list of the sensor data acquisition equipment is provided in Appendix 3. The
data acquisition scan rate is expected to be on the order of 1 Hz. Consequently the maximum data
acquisition rate for several hundred sensors is approximately 32 Kbps, a rate easily managed and delivered
over a network connection by a single high-end PC system. With a minimal 20 GB hard drive, this PC
will be able to capture the data from experimental runs lasting up to several days. All incoming data will
be time stamped for subsequent correlation with video and still image data. This server will also deliver
the data in real time via network connections (described in Appendix 4) to the active Client Server, which
may be either on-site or remote. An intelligent web browser on the Client Server will enable numerical
display and plotting of data selected by the user. At the conclusion of a run, this time-stamped data will be
uploaded to the Visualization and Archiving Server.
Teleobservation of the second category of data, visual (and related audio) information, arises from the
need to continually monitor and document progressive behavior of the specimen during the entire
experimental run. It is proposed that remote-control digital video cameras (e.g., Canon GL1) with 20x
optical zoom will be utilized for overall, real time, full motion video monitoring of the specimen (with
525 effective scan lines at 30 frames per second). Audio will accompany the video feed. For the range of
possible specimens to be tested in the MAST Facility, up to eight video cameras will be required to
monitor all surfaces of the test structure. These cameras will each be mounted on floor-standing movable
support posts, repositioned as needed for different test configurations. Each camera mount will
incorporate a motorized remote-control pan/tilt unit so that the cameras can zoom in on selected regions
of the test structure for closer observation. The video signals from these cameras will feed via S-video (or
IEEE 1394 as the appropriate MPEG encoders become available) connection into two Video Servers,
with four cameras per server, for encoding, storage, and network delivery. The Video Servers will
include 512 Mbytes of RAM, 126 Gbytes of disk capacity, and four video encoder cards for video
compression and time stamping the video into compressed MPEG-1 (and either MPEG-2 or MPEG-4 as
they become available) format for storage and network delivery. MPEG-1 compression will provide high
resolution video at a data stream rate of 1.5 Mbps supportable by the planned network infrastructure
(described in Appendix 4), both on-site and at remote facilities. Each Video Server will have the
capability of IP (Internet Protocol) multicasting of four real-time video streams to remote viewers over
Internet. It is envisioned that the Video Servers will deliver the video in real time via network
connections utilizing streaming video technology integrated into a Client Server intelligent web browser.
Each of these servers will require 120 GB hard drives to store up to 48 hours of video from the four
attached cameras. At the conclusion of a run, this time stamped video data will be uploaded to the
Visualization and Archiving Server. The Video Servers will also support teleoperation of the camera
functions (e.g., zoom, color balance) and the pan/tilt units through appropriate digital/analog control cards
within the Video Servers. Utilizing a Client Server intelligent web browser, remote users can have
control over video camera operation and pan/tilt positioning through a simple, integrated interface.
C-
31

While the digital video camera system allows overall monitoring of the structure during the test, higher
resolution visual images of specimen details are required for effective experimental analysis. Through the
use of high resolution digital still cameras, this capability will be available to remote users. It is proposed
that remote-control digital still cameras (e.g., Olympus C-3030) with 3x optical zoom will be used to
acquire high-resolution (2048 x 1536 pixels), close-up images of the test structure. For the range of
possible specimens to be tested in the MAST Facility, up to eight still cameras will be required to monitor
all surfaces of the test structure, with each camera mounted on a movable robotic manipulator. The
manipulator will provide for horizontal movement of the camera, both across the surface of the structure
and in and out from the surface. Each camera mount will also incorporate a motorized remote-controlled
pan/tilt unit so that the cameras can be tilted up and down. These functions, along with the camera's
zoom capability, will allow the cameras to take high-resolution photographs of a comprehensive range of
locations on the specimen. The digital still images (in compressed JPEG format) from these cameras will
feed via USB and Ethernet to one of two image acquisition PCs (Camera Servers) for storage and network
delivery. Due to the distance limits of USB, the USB signal from the cameras will need to be converted
near the cameras to Ethernet for delivery to the PC. The Camera Server will acquire, time-stamp, store,
and deliver each still image in real time via network connections to remote users through the intelligent
web browser. This server will have the same hardware configuration as the Video Server to minimize
complexity within the system, and will be able to store thousands of still images. At the conclusion of a
test, this time stamped image data will be uploaded to the Visualization and Archiving Server.
Teleoperation of the still camera remote functions (e.g., shutter trigger, zoom), robotic manipulators, and
the pan/tilt units will be through an appropriate digital/analog control card within the Camera Servers.
For targeting the still cameras, the low resolution video feed also available from these cameras will be
provided to the user. These NTSC video signals will be fed through a video switcher to a low resolution
video capture card in the Camera Servers. Utilizing webcam software integrated into an intelligent web
browser, the user will be able to view this video information as the cameras are manipulated into place
with the robotic manipulator and pan/tilt unit.
The specification, installation, and integration of the teleobservation and teleoperation equipment will be
performed by a team led by Ernie. Dr. Ernie has extensive experience in broadcast television and internet
streaming video technologies, network infrastructure, PC/workstation/server architecture, and data
acquisition subsystems. He manages a four-channel ITFS educational broadcast system (UNITE
Instructional Television) providing over 180 hours per week of graduate credit level programming to
distance learners (Ernie, 1998). This system operates several state-of-the-art broadcast/technology
equipped classrooms, videoconferencing facilities, and a master control facility. It relies heavily on the
use of teleobservation and teleoperation for its implementation. In addition, Doug Ernie currently leads
an interdisciplinary team in delivering over 50 hours per week of these courses to on-campus and distance
learners by streaming video, both by live webcast and video-on-demand. System integration of encoder
and editing workstations, video-on-demand servers, high bandwidth network infrastructure, and a simple,
consistent web browser interface are hallmarks of this effort (Jorn et al., 1999). Internet-2 issues
pertaining to high bandwidth video (MPEG-1 and MPEG-2), multicasting, and videoconferencing are
being explored as ongoing technology improvements to these programs (Ernie, 1999). System quality,
reliability, and ease-of-use have been key to the success of both these efforts, and will be critical in the
implementation of the teleobservation and teleoperation infrastructure proposed here.
It is estimated that the percentage of time through September 30, 2014, that this equipment will be
available for research by the host institution investigators and shared-use access by researchers outside the
institution is 85%, to be divided based upon access to the facility as established by the NEES Consortium.
The estimated percentage of time required for training users on equipment usage is 10%, and for
maintenance is 5%.
The fourth year of the grant is designated for shakedown testing of the complete MAST Facility. Because
it is difficult to estimate the date when the system will become completely operational, funding for
C-
32

personnel to develop and operate the system through September 30, 2004 has been included in the budget.
User fees will not be required for testing within the MAST Facility before September 30, 2004. After
September 30, 2004, user fees will be required from investigators at the University of Minnesota and from
investigators located outside the university. User fees will be established based upon the time allocated
by the NEES Consortium, in conjunction with the NEES MAST Facility Coordinator (see Section F), to a
specific project or investigator for use of the MAST Facility. Based upon the allocated time, pro-rated
user fees will be set to offset costs associated with the management and operation of the system. This
includes funding of the two MAST research fellows (see Section F) for training, test set-up, and test
execution; the Mast Facility Coordinator; the Mast Facility Technical Liaisons; the network maintenance
costs (through user fees to non-university users, and through indirect costs to the University of Minnesota
users); and scheduled maintenance for all hydraulic and electronic equipment associated with use the of
the MAST Facility.
H INFORMATION MANAGEMENT
The information gathered during an experiment within the proposed MAST Facility will include sensor
data through the data acquisition system, streaming video and associated audio data, and still images. It is
proposed that all of this data will be collected on associated servers and fed to a Client Server for real-
time teleobservation and teleoperation, as desribed in Section G. In addition, the proposed system will
include a Visualization and Archiving Server as the key component of the information management
system for the experiments conducted within the MAST Facility.
The Visualization and Archiving Server consists of a powerful high-end server with 512 Mbytes of RAM
and sufficient storage capacity for fast access to at least three complete cyclic experimental runs. This
storage will include approximately 1 Terabyte of hard disk storage. It is envisioned that the Visualization
and Archiving Server will operate the same graphical user interface software as the Client Server for
visualization, and will use software identified by the NEES Systems Integrator, or other appropriate third-
party software, for archiving. While full remote access to the locally archived files of experiments
conducted with the MAST Facility is a service which is envisioned at the proposed NEES facility,
permanent archiving operations, and the implementation of archiving capabilities, will also be
coordinated fully with the NEES System Integrator. An intelligent web browser, or similar software
specified by the System Integrator, will be implemented by a team led by Du. It should be noted that
alternatives to an intelligent web browser may be desirable if the full scope of applications within the
NEES system warrant software solutions different from an intelligent web browser. The investigators are
prepared to work with the NEES System Integrator to establish model-based simulation, visualization,
teleobservation, teleoperation, and data storage interfaces that allow integration of the proposed MAST
Facility into the complete NEES system. Du has expertise in multimedia authoring and synchronization
tools, streaming video, video server architecture, Internet proxy server, data storage interface, and high
speed networks.
Within the context of the MAST Facility, the advantages of using an intelligent web browser as the
primary graphical user interface to all data includes:
• When an experiment is in progress, all the captured data, video, and still images can be streamed to
interested parties to observe. A browser user interface is required to allow observers to switch
between videos and images. All videos and images have to be synchronized to reflect the process of
the experiments. These multiple streams of video also need to be synchronized with the data recorded
by various sensors.
• After the experiment, the captured data, including sensor data, videos, and still images, will be posted
to the web. Through the use of this intelligent web browser, a complete playback of the experiment
will be made available. Therefore, this experiment can be “repeated” as needed. This browser may
thus serve as a collaborative tool to allow researchers to visualize the experiment from several sites at
the same time so as to facilitate an informative discussion of the results.
C-
33

• After receiving approval from the principal investigator of the research, the final research results will
be posted to the intelligent web browser to allow other researchers to access the data. This is a quick
way to disseminate the research results.
From the aforementioned uses of such an intelligent web browser, a set of requirements that will be
developed in conjunction with the NEES System Integrator includes:
• Video and audio streaming capability: Researchers should be able to view all data from a remote site
without downloading the data first. Therefore, video and audio streaming capability will be enabled.
Different compression formats may need to be handled, depending upon the protocols established by
the NEES System Integrator.
• Multimedia authoring capability: Because the data posted will include text, still images, and video, a
user-friendly interface capable of manipulating multimedia information together is essential.
• Remote Accessibility Enabled: Researchers will be able to remotely control the playback of the
experiment.
• Multimedia synchronization capability: In a user-interactive environment, the captured data can be
played back in an interactive manner. That is, the playback of the experiment can be paused, resumed,
forwarded, rewinded, and repeated many times. An intelligent web browser may support flexible
presentations that incorporate variations in the way they are viewed.
The exponential growth of the Internet has resulted in a large and growing number of people with access
to the information services available on the Internet. The convergence on the Internet as a standard
vehicle for delivering information and the global nature of the Internet makes it an ideal way to
disseminate research results. However, implementation of an intelligent web browser or related software
in conjunction with the NEES System Integrator will require development because the protocols and
existing browsers for the World-Wide Web are designed to deliver static documents rather than full-
motion video. In the past, Du has worked on a video-based Internet browser called Networked
HyperQuick Time (Ma et al., 1998). This Networked HyperQuick Time follows the Client Server model
proposed herein. In that system, the information is stored on a server and many clients through the
Internet can access the desired information. A simple authoring tool is provided to allow hypermedia links
to be easily attached to different segments of a given video. Text annotation can also be provided. The
proposed intelligent web browser or related software developed in conjunction with the NEES System
Integrator would include a number of enhancements to this system.
Du has also been working on video and audio streaming, a key component of teleobservation and
information management. He has implemented a controllable software architecture for a Video-on
Demand (VOD) Server (Liu, 2000b). This software architecture uses user-level processes to provide
scheduling schemes, manage the VOD server and regulate the usage of system resources. Du has
implemented and investigated several different disk striping schemes and their impact on the utilization of
system resources of VOD server (Du et al., 1995; Du et al., 1996). In addition, he has developed a video
file allocation scheme to accommodate different user demands for various video files (Wang and Du,
1997). He has also done extensive work on a VOD prototype server which is based on an SGI shared-
memory multiprocessor with a mass storage system consisting of RAID-3 and RAID-5 disk arrays (Hsieh
et al., 1998).
The inherent nature of continuous media (e.g., video and audio), as well as its manipulation, imply the
need for temporal synchronization. The use of an appropriate multimedia synchronization model is
crucial to an intelligent web browser. The synchronization requirements that must be addressed to support
this environment are as follows:
• Fine-grained synchronization between media streams: matched video and audio require fine-grained
synchronization, generally within 100 ms with low slip. Multiple audio streams require even closer
synchronization (approximately 1 ms in the most demanding case). Because of these varying
tolerance requirements, a means of specifying the tolerance of a synchronization relation is needed for
C-
34

better resource scheduling. In addition, a mechanism is needed to maintain the synchronization
within the tolerance, and to control the rate of adjustment (second order effects).
• Synchronization of media segments of variable durations: visualizations and animations that are part
of the proposed web browser may not have a well defined duration which can be used for scheduling
the time of playback of associated media.
• User interaction: user interaction may affect the playout of some media segments but not others.
Flexible synchronization is needed that can adapt to these changes while accessing data from the web
browser.
There have been many approaches to modeling multimedia synchronization. Timeline-based systems are
common and generally map each object to a timeline. Removing one object does not affect the
synchronization of the others. Hierarchical specifications such as a Synchronization Relation Tree
provide a graphical model of a presentation using two main composition operations, parallel and
sequential. Both timeline and hierarchical specifications support fine-grained synchronization but have
difficulties with asynchronous events and media segments of unknown duration. Du has worked on and
proposed an event-based synchronization scheme for flexible presentation (Schnepf et al., 1996). Event-
based systems can be used to specify the loose relationship of coordinated but independent media and can
handle media segments of unknown duration. However, it is harder to identify inconsistencies in a set of
event specifications, and it is more difficult to handle fine-grained synchronization. In conjunction with
the requirements put forward by the NEES System Integrator, it is proposed to investigate a hybrid
specification model that blends the hierarchical and event-based models to provide a system with the
advantages of tight synchronization of the former, and the flexibility of specification of connected (but
independent) media segments of the latter.
The above requirements highlight the proposed visualization features of the Visualization and Archiving
Server. With respect to data archiving, a 100 Gbyte tape backup system will be implemented and
maintained by the on-site personnel. Requests for playback of specific past experiments will be accepted
either directly or through the NEES System Integrator. Uploading from tape of the requested
experimental data, including sensor and visual/audio data, will be conducted in a timely fashion.
The implementation of the servers outlined in Section G will each establish the data formats required for
storage of all experimental data. It is anticipated that much of these required data formats will be
transparent to the users. In addition, because the Visualization and Archiving Server is proposed to utilize
the same intelligent web browser as the Client Servers used for real-time visualization during a test, it is
anticipated that the researchers data will be available for the researchers use on the Visualization and
Archiving server and available for playback from remote sites shortly after the end of the test. However,
data privacy of the investigator’s research data will be considered, and the University of Minnesota will
work with the NEES System Integrator to establish standards for the processing and communication of
NSF-sponsored data sets.
The information management system and standards in current use in earthquake engineering research at
the University of Minnesota are centered on the control and data acquisition system developed by the
investigators and discussed in Section C. This past experience provides the stepping-stone for the more
comprehensive visualization and model-based simulation capabilities proposed herein.
I EDUCATION AND TRAINING
I.1 Education
The development of the MAST facility at the University of Minnesota will contribute to engineering
education in a broad, interdisciplinary way. This facility will be one of the focal points of the Structural
Engineering Laboratory (SEL) and will significantly contribute to cultivation of the excellent tradition of
engineering training at the University of Minnesota. Traditionally, the majority of the work in the SEL
has been performed by Graduate Research Assistants teamed with Undergraduate Research Assistants.
C-
35

The students have been allowed sufficient autonomy to develop their skills in design, construction,
instrumentation and testing of structural specimens. Their laboratory work is prefaced with introductory
training, followed by continued consultation with the faculty and full-time laboratory research fellow who
are committed to teaching and research. The PI’s anticipate that this division of work would continue
with the MAST facility, with teams of Graduate Research Assistants and Undergraduate Research
Assistants performing the majority of the work involved with testing structures in the MAST facility.
Every effort will be made to hire undergraduate and graduate students from underrepresented groups. PI
French, and Co-PI Shield both serve as mentors in the “Presidents Distinguished Faculty Mentor Program,”
an on-campus program that matches high quality freshman from underrepresented groups with faculty
mentors in their areas of interest. Through these interactions, students from underrepresented groups
interested in structural engineering are identified early in their academic career, and every effort is made to
involve them in activities of the SEL as Undergraduate Research Assistants. Typically 25% of the
undergraduate research assistants are female, and their performance has been outstanding. The faculty plan
continued involvement in these programs in the future.
Graduate students at the University of Minnesota will develop their theses based on testing and model-
based simulation performed using the MAST facility. Original contributions are expected in the fields of
structural engineering and computational simulation. The MAST facility will also be available for
research by undergraduate students at the University of Minnesota and elsewhere through shared use of
the facility.
In addition to the impact on those students directly involved in experiments using MAST, the facility will
also affect the curriculum. The proposed research will contribute to the knowledge of the three-
dimensional behavior of structures that faculty in the Department of Civil Engineering teach in steel,
concrete, and masonry design classes. Knowledge directly relevant to performance-based seismic design
will be used in a course that Co-PI Schultz teaches in a yearly graduate course on earthquake engineering.
When possible, demonstrations of the system will be made for Graduate Students enrolled in CE 8421,
Structural Dynamics, and for undergraduate students enrolled in CE3402, Construction Materials. The
teleobservation, teleoperation, and information/visualization management capabilities of this facility will
provide a unique educational opportunity for our undergraduate and graduate students. The delivery of
extensive video data by streaming video technology and the need for the visualization of complex
information from large data files is increasing rapidly in many areas; from research, to distance learning,
to commercial web-based broadcasting, to e-commerce. Consequently, incorporation of these topics into
modern engineering and computer science curriculum is pervasive. Utilizing this facility as an
example/case study for these courses will provide an enriched learning environment for our students. As a
consequence, all students enrolled in the Department of Civil Engineering, and possibly students in other
departments, will be impacted by the facility.
I.2 Training
Equipment training will be provided for users both from the University of Minnesota and from outside the
University (shared-access users). The level of training will depend on the users’ knowledge of the system
and on the scope of the users’ investigations. In each case, the training will consist of two parts, one
related to the technical and operational aspects of the system and the other related to safety rules. The two
research fellows will be responsible for providing training on all of the MAST components, including: the
structural, hydraulic, control, data acquisition, teleoperation, teleobservation, and networking systems.
The training model, employed by the Structural Engineering Laboratory at the University of Minnesota,
has been used with great success over the past five years to ensure safe, well-conducted experiments, with
successful data acquisition and control. In addition to providing user training through the research
fellows, Schultz and French will serve as Technical Liaisons to describe the technical capabilities of the
MAST facility for potential users, to help them to develop proposals to the NEES Consortium for use of
the system.
C-
36

J TECHNOLOGY TRANSFER
This section describes the past and current mechanisms used by the PI’s for effective technology transfer,
as well as describing future plans for technology transfer.
J.1 Past and Current Mechanisms of Technology Transfer
The PI’s have a proven track record of technology transfer to the international, national, and local
communities. On the international level French, Hajjar, Schultz, and Shield have all participated in the
planning of international conferences. French has co-organized several conferences including one on high
strength concrete in seismic regions. All four of these PI’s have participated in US-Japan programs
overseen by NSF. Dr. French is a member of FIB-CEB Working Group 3 developing a state-of-the-art
report for precast concrete in seismic regions, and recently spent a sabbatical in New Zealand at the
University of Canterbury. All of the PI’s publish regularly in premier archival journals, and present
findings at international conferences.
On the national level, the PI’s have been active in conference planning and participate on several national
technical and code committees of ASCE, ACI, AISC, and TMS. Hajjar is co-organizing the Composite
Construction VI conference to be held this summer, Shield has served on the planning committee for the
organization of an international workshop on bridge engineering, and Schultz has organized three
symposia of the ASCE Committee on Concrete and Masonry Structures (complete tracks in ASCE
Structures Congresses). Some of the committee assignments include chairing committees including the
Committee on Concrete and Masonry Structures (Schultz), ASCE-ACI 445 Shear and Torsion (French),
and ASCE committee on Load and Resistance Factor Design (Hajjar); and participating on committees
including ACI 318 Standard Building Code Committee (French), Masonry Standards Joint Committee
(Schultz), ACI 440 Fiber Reinforced Polymer Reinforcement (Shield), BSCC Provisions Update
Committee-Task Subcommittee 11 on Composite Construction (Hajjar), AISC Specifications Task
Committee 3 on Loads, Analysis and Systems (Hajjar), and AISC Specification Task Committee 10 on
Stability (Hajjar), and ACI 374 Performance Based Seismic Design of Concrete Buildings (Schultz and
French). French has also recently served on the ACI International Board of Directors. Refer to
biographies for other committee work.
On a local level, several of the PI’s are active in the local ASCE and ACI organizations. Shield chairs the
local ASCE Structures Committee, Hajjar serves on the board of direction of the Minnesota Section
ASCE Chapter, French is a Past President of both the Minnesota Section ASCE Chapter and the Iowa-
Minnesota ACI Section. French and Shield have been active participants in the organization of the
University of Minnesota Annual Concrete Conference, and the University of Minnesota Center for
Transportation Studies Annual Conference. Through all of these local activities, the PI’s have actively
sought out relationships with local engineers for the exchange of ideas and technology. In addition,
French, Schultz and Dexter have offered continuing education seminars for local engineers.
J.2 Future Plans for Technology Transfer
Future plans for technology transfer are based on the successful current and past programs. The results of
research obtained through use of the MAST facility will be disseminated through several channels:
• the undergraduate and graduate educational program at the University of Minnesota
• collaborations with researchers and students at other Universities (especially when they are users of
the MAST facility or other NEES Collaboratory facility studying similar structural behaviors)
• the continuing education program at the University of Minnesota, which is responsible for the Annual
Concrete Conference, the Annual CTS Transportation Conference, as well as an Annual Structural
Engineering Seminar Series
• Presentations at seminars, short courses, workshops and symposia
• Local, national, and international committee activities
• Journal publications and technical conference proceedings
• Specifications for the design and assessment of buildings and other structures
C-
37