Engineering Solutions for a Seismically Resilient Seattle

Thanks for coming. My name is Mike Bragg. I’m the Frank and Julie
Jungers Dean of Engineering here at the University
of Washington. I’d like to thank
you for joining us on this first presentation of
the 2016 engineering lecture series, titled City Smarts,
Engineering Resilient Communities. This series is
presented in partnership with the College of Engineering
and the UW Alumni Association. So if you’re not a
member, I encourage you and I’ve seen it up on the
screen to go to the website and consider joining the Alumni
Association, where you’ll find out about more about
these kind of events at the University. So I think the topic
of tonight’s lecture is very timely for
those of us who live in Seattle and this region. Many think and there
have been articles recently that Seattle is overdue
for a major earthquake event, due to its location
on the Cascadia fault. We, as engineers and
scientists at the University of Washington– and I’m sure
many others– are working to better understand the
impact of a major earthquake would have on the community,
how we would be better design buildings, how we
would impact public policy to make sure that those
designs are implemented, and that the public is all
safer in case something like that does happen. Our speaker tonight
is Jeff Berman. Jeff is the Thomas and Marilyn
Nielson Associate Professor in Structural
Engineering Mechanics in the Department of Civil
and Environmental Engineering here at the University
of Washington. Jeff does research
developing tools and methods to design structures more
resilient to earthquakes, right? Hence the topic. He’s been very involved, besides
being an outstanding teacher. He’s a very active
researcher and leader of two of the big projects on
campus in this area. One, he’s the
co-principal investigator of what we call the
M9 Project, which is an interdisciplinary
research center addressing the scientific and engineering
challenges that we face in the event of a Cascadian
subduction zone magnitude 9 earthquake. He’s also the
co-principal investigator on a recently awarded NSF
natural hazard engineering research infrastructure center. That was just announced. We’re just ramping
that up and we’re very excited about that center. This center will work
to collect information from all kinds of
different natural disasters so they can be studied and we
can learn from those and then better design buildings and
communities as a result. So I’m very privileged
to invite to the stage Jeff Berman who’s going
to give our lecture. JEFFREY BERMAN: Hi. Thank you very much. And my goal here is
actually, hopefully, not to scare everyone. No, actually, so
what I’d like to do is actually start the talk by
kind of looking at a photo. This photo is just
kind of I think illustrates the
power of earthquakes and what we’re talking about as
far as the potential dangers. It’s a photo that
was posted to Reddit. It was taken minutes
after the February, 2011– I think it actually say 2010
Christchurch earthquake in New Zealand. And you see the debris? This is kind of downtown
Christchurch here. I’ll actually use the cursor. This is downtown
Christchurch here, what is called the
central business district. And you see the dust and
debris being generated both from liquefaction
of the soil as well as collapse of
building structures. If we zoom in a
little bit, I think we get a even better
picture of what’s going on in downtown
Christchurch following that earthquake. We see collapsed buildings
and extensive damage to the infrastructure. Should note here that a lot
of these buildings that you see that are collapsed
along the streets here are what are known as
unreinforced masonry buildings. And we do have a lot
of those in Seattle. And I’ll talk a little bit
about those as we go along. So I think when you look
at a photo like this, one of the primary takeaways
has to be that the adage, really, that earthquakes
don’t kill people, it’s the buildings
that kill people and the response
of those buildings during those earthquakes
that we have to worry about. So in today’s talk, I want to
try to accomplish a few things. You know, I think
you should go away actually learning something
from a talk like this. So I would like to talk about
what our seismic hazard is in the Pacific
Northwest, what makes it different from the
seismic hazard in other parts of the world, and
then talk about why that matters, how that
difference actually effects infrastructure. And my focus mostly
will be on buildings, although a lot of
the lessons could be applied to other components
of the infrastructure. Then I’m going to highlight
some of the work we’re doing to develop
some engineering solutions to these problems. And then I’ll touch a little
bit on our new facility, our new center, where we’re
trying to collect data from the laboratory
of the real world to better understand the impacts
of natural hazards on the built environment. And so that’s what
I’ll take you through. I’ve kind of peppered
the talk with a bunch of photos of damaged
buildings from earthquakes. And so I won’t necessarily
introduce all of them. They’re there more
for your enjoyment. I do attribute them
wherever possible. This one is from my
colleague, Assistant Professor in Structural Engineering,
Paolo Calvi, who actually did a reconnaissance
mission following the earthquake in Italy. He’s actually Italian. And so he did a
reconnaissance mission and collected a bunch of
information after that event. So I will at least
talk about that one. OK, so let’s get started
by kind of understanding the background,
the seismic hazard that we live with here
in the Pacific Northwest. And you can’t zoom into
the Pacific Northwest before you look at a little bit
of high school earth science and understand the kind
of tectonic environment of the crust of the earth. So we’re floating around on
a bunch of tectonic plates that intersect
each other and move against each other
in different ways. And kind of this is an
illustration from USGS. And I’ve added these
arrows to kind of indicate how most of the plates are
moving against each other. And where the arrows,
the little red arrows, are actually pointed
against each other, those are places
where we have what are called subduction zones. And we live on top
of one of those. Those are two plates
colliding with each other. And what happens
is one generally moves underneath the other one. And they lock up and
occasionally they slip. And that creates the movement
of the subduction zone. If we kind of zoom
in here around both what is called the Pacific plate
and out to the Nazca plate, here we see a lot
of subduction zones. And up into the Juan de Fuca
plate and where that intersects the North American plate, we
see a lot of subduction zones. And it turns out
that actually if you were to make a map from the
USGS website of the magnitude 8 plus earthquakes that
have occurred since 1800, all of those magnitude
8 plus earthquakes occur on subduction zones. Or not all, but many,
many, many of those– the vast majority of those
occur along subduction zones where those plates
are– one plate is subducting another one. Now one highlight here is that
there were subduction zones all the way around here. And there’s the Nazca plate
subducting the South American plate here. But the Juan de Fuca
plate, where we are, is subducting the
North American plate and there hasn’t actually been
a large earthquake since 1800 in our region. But we know that
there was one in 1700. And so, let’s zoom
in a little bit– now that we understand the
kind of tectonic environment– zoom in a little bit
and use this figure, again from the USGS, to look
more locally at our region. So what we have is the
Juan de Fuca plate here. It is moving underneath
the North American plate. That creates what we call
the Cascadia subduction zone at this interface. And it’s locked right now. It’s actually not moving. But the pressure is building up. And at some point,
it’s going to move and that will create the
Cascadia subduction zone earthquake that we’re
all concerned with. So there’s a few
things– there’s a few additional seismic hazards
that we should at least note, even though I’m not going
to talk much about them, a few additional seismic hazards
in the Pacific Northwest. We have a series of what
are called crustal faults. These are kind of surface
level faults in the upper crust of the North American plate. And we’re really
just kind of getting a handle on where those
are and what those are capable of kind of now. We also have some very
deep source earthquakes that occur kind of where the
Juan de Fuca plate is really, really, really far below
the North American plate. And this is the
type of earthquake that many of you in the
room probably felt in 2001. That’s the Nisqually earthquake,
was one of these deep events. And those can be scary. They are pretty intense shaking. But they’re so deep
that they don’t have the same capacity
for causing damage as some of the surface
faults and potentially the Cascadia subduction zone. So what’s interesting about
the region that we live in, and if we contrast it to the
experience in California, is we just don’t have
a lot of earthquakes. OK? We don’t have frequent,
moderate earthquakes. We have the potential for
infrequent but very large earthquakes. And I think, and I’ll try to
hit this theme as I go through, I think that can create a little
bit of a preparedness problem. Complacency, if you will,
in our preparedness. The other thing that
I think is problematic for the Pacific Northwest
is kind of written here, which is that we
really didn’t even know the Cascadia subduction
zone existed and was a seismic threat
until the 1980s. OK? And people understand the
seismology in California in the 80s a lot better than and
then, oh, just discovering it. So I think that’s a big deal. And the other note
here is that, when did we build so much
of our infrastructure around the Pacific Northwest? Just as an example, I-5
opened through Seattle with all of the
bridges, including the ship canal, in 1967. OK. So that infrastructure is old. It hasn’t been designed
for earthquakes. It was designed
well before we even had a real good idea of what
the seismic hazard was here. OK? So that’s kind of both
the tectonic setting and a little bit about
what Cascadia is all about. So we know that there was
an earthquake in 1700. And this is not my work. I should acknowledge the
work of seismologists, folks at the U-Dub, folks at USGS. This is a photo
from Brian Atwater, a very famous seismologist
who we’re lucky enough to have actually as an affiliate
faculty member here in earth and space sciences. And he works for,
primarily, for USGS. And he made this
discovery of ghost forests along the Olympic Coast,
and was able to date those in their present
to the 1700 event. And an illustration of how
those ghost forests formed is kind of over here on the
left, where we have these two plates intersecting each other. And what’s happening is
the North American plate is actually buckling. It’s actually bowing
up a little bit from the force of that impact. And when it slips it goes down. OK? And when it goes down, that
creates coastal subsidence. The seawater comes in
and kills the forest that was there before. And that’s how they were able
to actually date the 1700 event, at least initially,
at least some of the work. More recently,
what’s gone on are a series of boring
investigations by Chris Goldfinger. So they’ve actually
cored on the ocean floor and gathered technical data
on what the underlying soil deposits look like
under the water. And from this, they’re
actually able to determine that there are a series of
landslides that have occurred underwater along the
Cascadia subduction zone kind of all at the same
times through history. And they’re able to go
back thousands of years and discover that these
landslides occurred by dating the sedimentary
material that was deposited. And so what Chris Goldfinger
and his colleagues were able to put
together is essentially a timeline of the
historical earthquakes that have occurred along
Cascadia in the subduction zone. And so this is what the
timeline looks like. It goes back about 10,000 years. And they think that
there have been about nine– or sorry,
about 20 magnitude 9 Cascadia subduction
zone earthquakes. And a magnitude 9 earthquake
occurs when the entire Cascadia subduction zone ruptures, from
northern California all the way to Vancouver Island
in British Columbia. So it’s a fault
rupture of thousands of– 1,500 or so kilometers. It’s really long. So that generates
the magnitude 9s. The magnitude 8, they
also have records or think there were about
20 magnitude 8 to magnitude 8 and 1/2 earthquakes. And those occur only on the
southern portion of Cascadia, of the subduction zone from
about Northern California to kind of the middle
of Oregon or so. And so if you map
all this out and you do some probabilistic
calculations, my seismologist
colleagues tell me that there’s
probably a 10% to 20% probability of a magnitude
9 Cascadia subduction zone earthquake in the next 50 years. And there’s approximately
double that for a magnitude 8 plus on the southern
portion of Cascadia. And if you look, on average,
and took all of these events, the magnitude 9 event occurs
roughly every 500 years, but we see that that timing
is not consistent at all. OK? I think the windows are as
narrow as a couple or as 50, 65 years, something like that. And as long as 1,000 years. OK? So there’s a lot of
uncertainty in this. And hopefully you can
take that away from this. OK? So given all of
that information, we end up with articles
like this in The New Yorker. And I’m sure lots of
people read this article and maybe that’s
even why you’re here. So in the news, the New
Yorker, “The Really Big One.” So if you read it, it
paints a really dire picture of what might happen
in the Pacific Northwest in a magnitude
9 Cascadia subduction zone earthquake. The quote lots of
people take away from this is that everything
west of I-5 will be toast. And you know, people’s
definition of toast is different depending on
who you are and what you do. So let’s dig a little
deeper and figure out what toast really means
and what we should really be concerned with. And so, I’d like to talk
now a little bit about what we call the M9 Project. And I think Dean Bragg
spoke a little bit about it. It’s a large,
interdisciplinary project, because these are really
interdisciplinary problems. So it’s a large
interdisciplinary project with a whole bunch of
people from UW and USGS. We’re doing work on the project
and a bunch of UW departments involved as well, lots
of graduate students and all of that. But if I can take you through
the workflow at least a little bit. The whole project starts with
computational models developed by colleagues at the USGS
and in seismology here that are actually able to
simulate computationally the subduction zone rupture. OK? And I’ll actually go through
a little bit about what those models look like, a
little bit about how they work, and some of the data that
we’re able to retrieve. These are huge, very
complex numerical models that are actually trying to
simulate Cascadia subduction zone earthquakes. And my talk is going to focus
a lot on the things that are kind of
highlighted dark purple here, the response
given our ability to predict what the ground
shaking might look like, what is the corresponding
response of the built environment for that
predicted shaking, and then what we can
do to kind of improve our resilience to that shaking. But other aspects of the project
are investigating the impacts– the generation and impact of
tsunamis generated by Cascadia, what happens with landslides,
earthquake induced landslides in a Cascadia event, soil
liquefaction and the risk that that might pose
the port facilities and things like that, how
we might utilize earthquake early warning to actually
help reduce risk, and then some community
engagement which I’m doing actually right now. And so, so my talk is going
to focus mostly on this. We could have hour long talks
on all aspects of this project. OK? So it’s really complicated. There’s a lot of
people involved. It’s a really
interdisciplinary problem, especially when
you get all the way down to trying to do something
about it, which is generally a social question. OK? Not necessarily an
engineering question. But I’m going to
focus mostly up here, and we’ll talk a little bit
about some of this other stuff too. OK so let me start with how
we– how not me personally, but how my colleagues
on the project are actually
attempting to simulate computationally Cascadia
subduction zone earthquakes. So we’re using– and I’m
using the royal we here, because it’s my
seismologist colleagues– we’re using what’s called
physics-based earthquake simulation. OK? And so the idea is developing
a computational model of the actual earth. OK? And these are really big
computational models. And we’re going to
simulate the rupture along the locked
portion of the fault, and then we’re actually going
to have the waves propagate through the adjacent earth. And we’re then able to record
the shaking of the ground that is predicted from
that numerical model. Then we’ll be able to use that
recorded shaking of the ground to assess buildings
and infrastructure, and that’s kind of my end. So to tell you a little bit
about the computational model, it’s hard to look at
here, but these distances are measured in meters to
give you an idea, you know, so you’re talking
50 kilometers deep the model goes into the earth. OK? And this is northern California,
kind of the southern edge of the model. And way up here is
Vancouver, British Columbia. And what’s plotted– what
this heat map is showing– is essentially the
stiffness, or you can think of it a little bit
like the hardness of the rock. OK? And so it’s that stiffness of
the underlying rock that has to do with wave propagation. And it’s important that we get
that right to understand how the seismic waves
generated on the fault actually propagate
through the model. And we don’t need to know
the details of all of that. And we don’t want
to know the details of how it’s constructed. There’s a few
references down there, if you’re really,
really interested. But that’s the general idea. And you can see here the dip
of the Juan de Fuca plate. So that actually is
the subduction zone. And we’ll simulate slip
and rupture along fault and watch the waves
propagate through. Now one of the very interesting
things about this model is up here in Seattle. OK? And so if this heat
map is measuring the stiffness of the
rock– redder is stiffer, purple is less stiff– we notice
that actually up in Seattle, we sit on a whole bunch
of softer material. We sit on top of what is
called a geological basin, a sedimentary basin. Because of the geological
formations here, there were sedimentary
deposits for a long– kind of in the prehistoric past. And a glacier came around
and sat on those deposits for a long time. And it’s really hard stuff. Anyone heard of glacial till? OK. So this is glacial till
that we’re talking about. A glacier sat on
it for a long time, on top of a sedimentary
deposit, stiffened it all up. But it’s nowhere near as
stiff as the surrounding rock. OK? If you try to plant a
garden in glacial till, you’re going to be really upset
because it’s really hard to dig through, almost impossible. But it’s– and it’s great
for founding buildings on. It serves as a really
nice foundation material. But it does create
problems in seismic events. And so we’ll talk a little
bit about what that is. So if we kind of look down
on that map a little bit, it’s kind of interesting because
it ends up as a bull’s eye. So this is a map. You can interpret it as
the depth to real rock. That’s not exactly
what it’s measuring, but for our discussions I
think that that’s probably appropriate. So it’s the depth to real rock. And it identifies what we
call the Seattle basin, which is pretty much directly–
the midpoint is directly under Lake Washington. And this is a very unique
geological feature. It’s one of the
largest sedimentary basins in the world. One of the deepest. And there are very few of these
that are near subduction zones. So actually there’s
very little data on the interaction between
these deep sedimentary basins and subduction zone earthquakes. And so, that’s one
of the big things that the M9 Project
is trying to uncover, is how these subduction
zone earthquakes interact with these
deep sedimentary basins. OK. So, and the one other
thing I wanted to point out is Seattle is here. And I’m going to show some
recordings of ground motions from our simulations
right in Seattle. And I’ll use Snoqualmie
as a reference point that is outside
the basin, outside this geological structure. So why basins matter. These deep sedimentary
basins do a couple things. There’s actually a third
one that was too complicated that I threw out for this talk. But I’ll talk about
the first two. So they amplify ground motion. So what I’m showing here is a
ground motion source somewhere. This is a rupture in
a fault. And the waves are traveling through the
soil– or through the bedrock, and they’re going to get
to the edge of the basin. And they hit this
material that actually has a very different stiffnesses
than the material they’ve been traveling through. And that can create what
is called impedance. And it can amplify the
ground shaking quite a bit, at least at specific frequencies
which I’ll talk about. It’s kind of like– and
Art Frankel my colleague who generates all
this stuff, all these seismological
models– hates this analogy, but I’m going to say anyway. It’s kind of the bowl
of jell-o effect. OK? You can shake a bowl
of jell-o, and you can get the jell-o to shake
a lot more than the bowl. Right? If you shake it right. OK? And so that’s somewhat
similar to the impedance that’s happening here. The other effect it has is
it can focus seismic waves to a points within the basin. And this depends a lot on
the kind exact character, the exact structure,
of the basin. And kind of the attack angle of
the seismic waves, if you will, as they travel through. And that focusing can
actually cause severe ground shaking in some areas. And so a way to think about
this is actually a prism. OK? It’s kind of prism effect. The same way that
a prism can focus light, the geometry of the
basin can focus seismic waves into some specific locations. And if you were around in
the Nisqually earthquake, there’s a neighborhood in West
Seattle where a lot of chimneys fell down. And then we actually
think that that was likely due to
focusing from the geometry of the basin and
the attack angle from the Nisqually earthquake. OK? And so those things
have real effects. So we’ll look at
some videos now. I know this is pretty technical,
but hopefully everyone’s staying with me. So let’s look at a
couple videos of what an M9 computational
simulation looks like. I should note that
we’re going to watch it. It’s going to run
for like 30 seconds. It takes a week on 1,000
processors at the Pacific Northwest National Lab to run
one single scenario for the M9 Cascadia event. OK? So that computational
model is huge. It takes a lot of
effort to run it. But I will show the results of
two realizations in 30 seconds. So what these are showing
is the initial rupture, and then that rupture
propagating up to the North and down to the South. You can see the seismic waves. And if I pause it at
just the right time, you see the amplification
here in this particular video in the Seattle basin. So this is essentially
just a heat map of how fast the
ground is shaking. OK? And out here by the subduction
zone, it’s shaking a lot. But those waves
kind of dissipate as you get over the Olympics. But then as you get into
Seattle, they intensify again. OK? And that’s the effect of the
basin that sits underneath us. And similarly with
this realization, you can see a little
bit of that as well. So these two realizations,
you can’t tell yet but I’m going to
use them later on, they are kind of what we feel
is like a worst case and a best case scenario. OK? And we’ll talk about
them a little bit. And there’s a lot of
variability in between these. There’s a lot of
assumptions that go into how this rupture
starts, where it starts, how it propagates,
and all of that. And that creates a
lot of uncertainty. So the type of data we’re able
to extract from that model is– from those types of models are
actual ground acceleration time histories. And we’re able to extract
these on a 1 kilometer grid for the entire
Pacific Northwest. So it’s just a mountain of data. So if you want to know what
the expected shaking is at your house, we could
probably tell you. And so that’s– so these
traces are essentially the acceleration of the ground,
just like a seismometer would record, versus time. And you can see the
differences as you move in the Seattle basin,
outside the Seattle basin, down in Portland, and
then down in Medford. And these ground motions
take– the ground motions, the kind of significant
shaking might be as long as a couple
minutes in these recordings. And so, if I go back to those
things I wanted to talk about, we wanted to talk about the
seismic hazard and kind of what is unique about the region. So we have the unique
subduction zone. We’ve got the deep sedimentary
basin that’s unique. And so now I’d like to spend
a little bit of time talking about how that’s actually
different from the experiences from earthquakes in
different parts of the world. OK? So here I take one of
these time history, or this is kind of the shaking
of the ground, the acceleration of the ground,
from our simulation and compare it to scale– these
are both in the same scale– to the acceleration
of the ground recorded from a station in the
1994 Northridge earthquake outside California–
or outside Los Angeles. And you see that the ground
doesn’t shake quite as hard in the M9 simulations. The intensity of
that shaking is less. But it lasts a lot longer. OK? That record from Northridge
might be 20 seconds or so. And the duration
of strong shaking for these subduction zone
motions is very, very long. And so duration
matters because we’re asking buildings
and infrastructure to sustain– to
essentially sway back and forth– for a long time
in these ground motions. And that could be
significantly more damaging. The other way that
it’s different requires us to think just a
little bit about seismology and the fact that these
earthquakes really are a bunch of waves
traveling through the ground. And those waves have a
certain period, so essentially the time from crest to crest. OK? And so if we were to
decompose the wave form and study what period
waves are contributing a lot to this ground motion,
we can generate heat maps that kind of look like this. And essentially the more intense
or the more darker the colors, the more that wavelength
or that period of wave is contributing to
the ground motion. And what we see– this is just
comparing inside the basin to outside the
basin– what we see is that there’s a lot more
intensity for a longer period waves, that the longer
period waves inside the basin have a lot more intensity
inside the basin than outside the basin. And that’s going to
impact structural response when I get to actually talking
about structural response. So now we know a few
of the differences. OK? So we have the subduction zone. So it’s really long. It creates very long
duration ground motions because the rupture happens
over such a long length. Those ground motions
then reach Seattle and are amplified by
the geological basin that we sit underneath. That amplification intensifies
the shaking associated with some longer periods. OK? So that’s essentially
the takeaways from the seismology
work, all the work that my seismology
colleagues have done. And so now we know everything
that we could possibly know about Cascadia. Just kidding. So the next– so now
what I want to focus on is now that we know a
little bit about Cascadia and we know a little bit about
the ground motions, why this matters for the response
of the structures that we build on top of
this sedimentary basin here in Seattle. So the workflow
for an earthquake. This is the workflow
for an earthquake. So it starts to the left. The fault ruptures. The ground moves
around and shakes. And we’ve talked about that. And the ground really
moves in all directions. And then at some
point, it reaches– that shaking reaches buildings. And those buildings kind
of sway back and forth. OK? So the ground is
shaking underneath them and the buildings
sway back and forth. So that’s great in a
picture, but we can probably do a little better. This is a video from some
of my colleagues in Japan. It’s actually up on YouTube. So it’s an 18 story
steel building. It’s about 1/3 scale. This is the largest
shake table in the world. You see people on the balcony in
the back for a sense of scale. And so this is what
buildings do in earthquakes. And actually we’re going to see
it collapse right about there. So that’s what structural
collapse might look like. And it’s on a catch frame
there so it actually doesn’t damage the shake
table that it’s sitting on. And so when we talk
about earthquake shaking and buildings, that’s
what it actually looks like when we get there. OK? So it’s a fun video to watch. Not often you get to see these
kinds of structural collapse videos. If you want, you
can go to YouTube and there’s a whole
lot of great examples. OK. So I have one equation. Actually I have two equations,
but they’re all in this slide. OK? And there are actually
Newton’s second law, so maybe we can handle it. So if I take that building
responding to an earthquake– OK, that was on a shake
table and the building was swaying back
and forth– and I think about how I might do
a simple mathematical model, numerical model to represent
the response of that building to that earthquake. It might look like this. OK? It might be a lollipop
structure that has the entire weight or the
entire mass of the building lumped somewhere. OK? And then that lollipop is going
to have a certain stiffness. It’s going to sway back and
forth as the earthquake happens underneath it. OK? So if I were to
have a giant finger and be able to push on
the top of that building, I would be applying a force. That force in this figure
is F. And that building, if I move that building
a displacement, the top of the building, it displaces–
that funny Greek character, delta. What we call the
stiffness of the building helps resist that force. And so the total force
would be the displacement of the building times the
stiffness of the building. OK? Now that’s great,
but earthquakes don’t have giant fingers that
actually push on buildings. They accelerate the ground
underneath our buildings. And so what happens
there is the ground accelerates underneath
the building. And the mass accelerates
as the building vibrates back and forth. OK? And we can apply
Newton’s second law, which is that force is equal
to mass times acceleration. And so if we’re able to quantify
the acceleration of the mass, we can then get the
total acceleration, which is the
acceleration of the mass plus the acceleration
of the ground. And we can then get the
total acceleration actually multiplied by the
mass of the building, and determine a force. OK? And that force is
the force that we need to design our buildings
for for earthquakes. OK? It’s really just an application
of Newton’s second law. OK? So the question really
that we have to ask is, what is the acceleration. OK? If we can pin down
that acceleration, we can understand how buildings
respond to earthquakes. We can figure out how damaging
Cascadia may really be. And hopefully actually
improve the infrastructure. OK. So that acceleration is
a function of the input, so what the ground
acceleration looks like, but also the natural
period of the structure. And so I have a little animation
here of a weight on a spring. So that’s kind of
the natural period of vibration of that spring. When I looked at
it on my computer, it was moving a lot
faster than that. That’s actually not two
seconds from peak to peak. That would be the natural
period of that oscillator. It’s moving a
little slower here. So if we were to look
at the natural periods for some different structures–
that’s actually my house. Maybe it has a period
of 0.1 to 0.4 seconds. A four story building,
here represented by the public library,
maybe has a natural period of about 5 seconds. So that’s how it vibrates back
and forth, maybe about a half a second. As we move to taller
and taller buildings, we maybe have natural periods
of 1 to 2 seconds for a 10 to 20 story building. And, I don’t know. Jon Magnuson was here. So maybe I got
that number wrong. Maybe approximately 7 seconds
for the Union Square Building. So we’re interested
in how we get a. Because if we can get
that acceleration, we can determine
the force that we need to design buildings for. So this is probably the most
complicated and technical part of the talk. And hopefully we can all
get through it together. So if I take a
bunch of buildings, I have five buildings here. OK? And they all are represented
just by little sticks and masses up on top. OK? These five buildings all have
five different natural periods. OK? And I develop a numerical model
for all five of those buildings and I subject those buildings
to the same earthquake ground shaking. OK? The same ground acceleration
in all five of those. Then from my little
models, I can actually determine what the maximum
acceleration of the mass is. And if I can determine what
the maximum acceleration of the mass is, I’m golden. Right? So I can do a little
plot here that for each of the response of each
of these buildings, I can pluck off the
maximum acceleration of the building itself. And plot it here versus the
natural period of the building. This is the key to
earthquake engineering and how we design
structures for earthquakes all around the world. OK? This is called a
response spectrum. I said this was the most
technical part of the talk. Hopefully folks
are still with me. And it gives us that
maximum acceleration that we expect from a
particular earthquake for a variety of buildings
with different natural periods. OK? So then, if I’m interested
in designing the UW Tower, I need two things. I actually need three things. I need to estimate the natural
period of the U-Dub Tower. I need to determine the
mass of the U-Dub Tower. And I need to determine
that maximum acceleration. And if I have those things,
I can look up the maximum, I can use the
building’s period, I can find the maximum
expected acceleration, and I can determine my
seismic design force. OK? Now, this acceleration response
spectrum that I am showing here is smoothed out. This is actually kind
of the shape of what we use in the building codes. It’s actually specified
by the building codes. And it’s supposed to reflect
the local seismicity. So it’s kind of on average. Someone took a look at
a lot of earthquakes and that’s effectively what
those accelerations look like. And now we can design kind
of all these buildings uniformly for what we expect
to be the seismic hazard. And so we do get to
reduce this a little bit for systems that behave
well versus systems that behave poor. OK. So what is a system– a
structural system that behaves well in an earthquake
actually look like? OK. Hopefully it doesn’t
collapse like that building that we saw on the shake table. Instead, what we try to do
when we design buildings for earthquakes is we try
to pick out a few components that we understand really well. And we expect those
components to be damaged. And hopefully those
components are damaged but the rest of the
building is actually still standing after the event. OK? So it’s the idea of just
controlling where the damage is and understanding how
that damage occurs. And so this is a braced frame. And that seismic
force may act up here. And those braces help
resists that seismic force. And this is what it might
look like in an earthquake. Those braces are
certainly damaged. This building may have to be
torn down after an earthquake. But it hasn’t collapsed like the
one we saw on the shake table. It can still hold all of us up. OK? And we can walk out
of that structure after a really big earthquake. OK? So a system that behaves poorly. When those elements that resist
that seismic load get damaged, it also loses the ability to
carry our weight– the gravity loads, the weight of the
building occupancies. And so this is another example
of an unreinforced masonry building. This one’s again from the
Christchurch earthquake. And we see the damage
from the horizontal forces from the earthquake has
caused sufficient damage to the components that
actually carry the gravity loads of the structure. And that building has collapsed. OK? And so the difference
in these two is really all about life safety. Right? It’s all about
collapse prevention. And that’s what we’re trying
to do in modern building design for earthquakes. We’re just trying to save lives,
not necessarily think about damage necessarily
to the building. So with that, I’ll
tell you a little bit about what we’re trying to
do in modern building codes. We’re trying to
prevent collapse. OK? We don’t want buildings
to look like this. We don’t want buildings to
kill people in earthquakes. Instead, we consider a
very, very large earthquakes and we design for what is
called collapse prevention. But we do expect
significant damage. OK? Modern building
codes today do not give any kind of indication. We do really no design
for smaller, more frequent earthquakes. OK? We design for a really,
really, really big earthquake that has a 2% probability
of happening in 50 years. OK. That’s a really big earthquake. And all we’re trying to do is
design for collapse prevention. OK. And by doing that
we kind of hope that in smaller
earthquakes the damage is not so severe
that we would have to tear the buildings down. But we don’t actually do
anything to ensure that. And we really don’t consider
repair or replacement at all. So I think that’s always a
little eye opening for folks that aren’t
structural engineers, is that this is the
performance, this is what we’re doing in
structural engineering. And, you know, the
reason that we do that is that it’s really
expensive to design buildings to be undamaged in these
very, very large earthquakes. If we tried to do
that, everything, all of our buildings would
look like nuclear power plants. And so that’s not
appealing really at all. OK. So what does the expected
performance look like? And this isn’t intended to
say that this reinforced concrete shear wall– so you see
reinforced concrete here shear walls going up all over Seattle. They’re popular seismic
load resisting system’s for buildings. I snapped this photo of
myself from downtown. It’s a reinforced
concrete shear wall. It’s carrying the seismic loads
here, or at least part of it. And so what may happen? We don’t know what’s going
to happen to this wall. But what may happen to this
wall is we may have cracking. We may have some fairly
significant damage in very large earthquakes. And that building may
need extensive repair. It may not be usable
for a long time after this kind of
really large earthquake. OK? So that’s kind of what our
intentions for the building codes. So given– now we
have two pieces. OK? We have Cascadia. And we have– we
understand Cascadia. And we understand
building response. And now the next question
is, how do our buildings perform in Cascadia itself. OK? And so if I were to look at
the acceleration response spectra for Cascadia here
for two of our realizations– those two realizations
that I showed before– and compare it to see
the design spectrum, this is what we’re using
today for building codes, the story isn’t all that great. So we’re seeing
accelerations that are considerably larger than
what we’re designing for. And we expect some
significant damage. And we also see that
kind of the shape of this doesn’t really match the shape
of what we’re designing for. And that can cause
some issues as well. So I’ll skip ahead. I think this really shows kind
of some of the stuff that we’re finding from the
Cascadia simulations is that they can
be quite damaging. So let me skip ahead
just a little bit and tell you about your house. Because I think that’s
actually something that you’re probably
concerned with. So here I have these two
realizations for the Cascadia subduction zone earthquake
compared to what we’re actually designing for. But you should know that
your house lives down here. OK. It has a pretty short period. And it’s probably going to do
OK in a Cascadia subduction zone earthquake. Cascadia, of course, is
not the only natural hazard that we have in the region. We have the Seattle fault. And
so I would encourage everybody to go here and
take a look at what you can do to prepare your
own home for the Cascadia subduction zone earthquake. If you just Google Seattle
earthquake preparedness, that page will
actually come right up. And they give recommendations
on how to retrofit your house, how to prepare the
contents of your house for large earthquakes. So everyone should
go home and do that. So you know I don’t want
to be all doom and gloom. I do want to just take
three more minutes or so and talk about what we’re doing
to help improve the problem, help improve the performance
of infrastructure to this Cascadia subduction
zone or other earthquake events. And so, you know, the
whole notion of do we need to design structures
to be damaged in earthquakes, you know, these two
little structures go through very
large deformations. You know, that might
have been an earthquake. It was a finger, but it
could have been an earthquake as we saw it before. And they’re totally
undamaged, right? They bounce back and are
completely undamaged. So can we use some
of the technology, you know great technology
that went into constructing those little bumblebees,
and actually apply it to our infrastructure? And it turns out that you can. We’ve been developing,
my colleagues and I have been developing, various
structural systems that do indeed use that technology. So here’s an
illustration of– a video of a bridge on a shake table. And what you’re seeing
is damage occurring. This is a conventional kind
of modernly designed bridge. It doesn’t collapse. It actually performs kind of
as intended for modern bridges. But we do see a significant
amount of damage up here at the
top of the column. If we’re to employ some of
these new kind of rubber band technologies, the
displacements of that bridge are actually maybe
a little bit larger, but there’s
essentially no damage. And so that can actually be
a really cool technology. If we zoom in on what those
structural components look like after this event
on the shake table, we see all kinds of concrete
fracture, we’ve exposed rebar, we’ve got rebar
that’s fractured. But in the kind of new system
on the right, innovative system, we have actually
no damage at all. And so we’re taking– I’m
working with some folks– Lever Architecture and
KPFF engineers– and we’re actually
taking this technology and using it to design a
building system for a 12 story building that’s under
design in Portland, Oregon that uses actually wood frame
shear walls– or timber shear walls. But it has these big
giant rubber bands that extend from the
base of the building all the way up to
the top of the roof. And the idea is very similar
to that rubber band guy. OK? The wall actually rocks
back and forth, but then after the event, it
re-centers itself. OK? And we need to control
that rocking a little bit. And we add some steel pieces in
here to control that rocking. But those steel pieces
are designed to be easily accessed and replaced. And after an event, the
idea after a big earthquake, the idea is that we can actually
walk in and use this building after an event. So we don’t need to
design for damage. And in the last
kind of two minutes here, I will talk
about what we’re doing to learn about–
to use the real world, the data collected
from the real world– to learn about the
response of infrastructure, in this case to both earthquakes
as well as wind hazards in tornadoes and hurricanes. So, this is our new rapid
post-disaster reconnaissance facility. It’s a collaborative effort
between a bunch of places. But it’s headquartered
here at the UW. And what we’re going to have
is a whole bunch of kind of high tech toys to
take out into the field after an earthquake,
hurricane, or tornado and do reconnaissance, so make
very detailed measurements of what condition buildings
and infrastructure are in following an earthquake. We’ll actually even have
some social science tools to understand what
the social impacts of these natural disasters are. So we’re going to have
drones with fancy cameras. We’re going to have really
high tech measurement devices. And what all of that
enables us to do is to gather all that data and
then do– digitally reconstruct the damaged world, the
damaged environment both at the community
scale– and I’ll show a little animation here. So this is a 3D
reconstruction done after the Nepal earthquake. And it’s using
drone-based digital images that were collected. And we use a
computational algorithm to reassemble those individual
images to get a 3D view of the damaged environment. So now that damaged
environment, we’re actually able to take
back here to Seattle and study it in greater
detail because we were able to take
precision measurements. We can not only collect
community scale data like that, but we can also collect
very detailed data at a kind of a component
level and actually kind of explore these things in
a virtual reality environment. So here he’s wearing actually
virtual reality glasses, and we’re manipulating
this environment in 3D. And this is all data collected
from a reconnaissance mission following the
Christchurch earthquake. And he’s measuring those
rocks because those rocks were part of a big
rock slide that actually damaged this building. And so we have very detailed
measurements of the rocks that damaged this building, we have
very detailed measurements of the building itself. We can come back to the
kind of virtual laboratory here and explore that building
and make measurements. And in the end, actually,
develop computational models that actually simulate the
response of buildings to rock falls, to earthquakes, to
wind, and all kinds of things. So this is a
computational simulation of a similar rock going
through a masonry building, using the data
that was collected from that
reconnaissance mission. So that’s about it. So I have a few
closing thoughts. OK? I tried to– we tried
to learn a little bit about the seismology, a little
bit about how we engineer buildings for earthquakes, and
some of the new technologies that we’re able to put
to bear on the problem from an engineering standpoint. But I think this
table illustrates that the problem is not
just an engineering problem. It’s a social problem. It’s a political problem. And it requires a will to
do things in a better way. So this is a table showing in
San Francisco, in Los Angeles, and for actually some of these
are state laws, for where they require retrofit and/or
evaluation of different types of buildings. And what we require
here in Seattle. And it’s a pretty
stark contrast. And if you don’t know what
an unreinforced masonry building is, well
that’s what one looks like after the
earthquake in Italy. If you don’t know what a
nonductile concrete building is, that’s what one looks
like after the Christchurch earthquake. Kind of before and
after photos here. And if you don’t know
what a soft story timber building looks like,
that’s what one looks like after the Northridge
earthquake in California. OK? So we understand. We can develop
engineering solutions to help solve these problems. I think that is very possible. But I think coming up with
a community level solution requires community level and
interdisciplinary approach. So I’m going to leave you
with kind of what we know. We have a significant
seismic hazard in the Pacific Northwest. We have infrequent earthquakes,
but they have the potential for very high consequences. The Cascadia
subduction zone event is affected by very
unique regional geology that we are just
beginning to understand. Our infrastructure is old. OK? And we’re asking it
to go through and be subjected to what is very severe
ground shaking potentially. And in order for us to really
develop these solutions, we can develop
engineering solutions, but it takes an
interdisciplinary approach to actually solve this problem. And so, you know, it’s
not all doom and gloom. I think we are
making some progress. I still live here. I have two children and I
haven’t taken them on a plane as far away as possible. So I think we can address these
problems, but they are serious. And we should be doing
something about it. OK. So with that I’ll acknowledge
my collaborators, colleagues, graduate students, the National
Science Foundation, USGS, and thanks to UW for giving me
the chance to talk about this. And thank you all
for being patient. It seems like I did
go a bit over time. And I very much appreciate it. No one left. So you’re all really either
interested or asleep. OK. So thank you very much. And, yeah. So one final note on this slide
is we are making progress. This is the Cypress Viaduct
that collapsed in the 1989 Loma Prieta earthquake. The Alaska Way Viaduct is almost
identical to this structure, and we’re finally
taking it down. So progress is possible
and forward momentum.

Leave a Reply

Your email address will not be published. Required fields are marked *