Structural Dynamics Challenges in Launch | Wisconsin Public Television

Structural Dynamics Challenges in Launch

Structural Dynamics Challenges in Launch

Record date: Dec 12, 2017

Matt Allen, Associate Professor in Engineering Physics at UW-Madison, discusses the physics behind rocket design. Allen highlights the structural dynamics, the vibration limits, and the amount of engine thrust that is necessary to successfully launch a spacecraft into space.

University Place Campus: 

University Place Lecture Series: 

University Place Subjects: 

Episode Transcript

- Good evening and welcome

to our guest speaker

tonight here at Space Place.

And I'm really pleased

to introduce tonight

Professor Matt Allen

from the mechanical

engineering department

here on campus, and he does some

fascinating work on trying to

figure out how things hold

together or fall apart

when they're moving

at hypersonic speeds.

He's been doing this

kind of work for a while.

Actually was at Sandia National

Labs before coming to UW-Madison

a few years ago to

join the faculty.

And so he's going to talk to us

tonight about how to get things

into space without them

falling apart on us.

Professor Allen.

- All right. Well, thank you.


Thanks and thank you

for coming tonight.

I'm looking forward

to talking to you.

So the question that we want

to talk about today is, yes,

"How do desire structures

or aircraft, spacecraft

to get to space,

"and how is that different than

designing normal structures,

"bridges and the things that

we can do more intuitively?"

So, for example, if

you were tasked with

designing the Statue of Liberty,

say we know that we need to

make this arm strong enough

that it can hold up the

weight of the torch there.

And so we could

imagine how we might,

how that might put stress

on Lady Liberty's arm

and where we might

need to reinforce that

to make sure we had

enough strength

to be able to hold

that torch in the air.

It's not too counterintuitive.

Now, what would happen, though,

if instead we put a controlled

explosion underneath

the Statue of Liberty?

And we tried to launch

it into space, right?

Does any of what we just

talked about apply?

Well, I'm hoping I can give

you a little bit of an idea

as to what does

and what doesn't.

And it seems like a pure joke,

but, actually, if you put these two structures side to side,

so this is the

Statue of Liberty

and here is NASA's

space launch system.

Now this, graphics these days

are so good I have to warn you

this thing has never

been fully assembled.

Actually, all the pieces

are not even done yet.

So this is just an artist's

rendition of what this may look

like some day as it launches,

assuming everything goes well.

But, actually, if you compare

these two structures,

they're fairly similar in size.

The SLS, NASA's SLS will

be a little bit taller.

The Statue of Liberty is

only about 450,000 pounds

to five-and-a-half

million for the SLS.

And if you think about that,

they're both basically

hollow metal structures,

but the SLS is a

hollow metal structure

packed full of explosives

that are hopefully

going to propel it into space.

Okay, we'll talk about some

of the issues involved

in designing a

structure like this.

I'll focus on the structure,

not so much the propulsion

or the avionics or

these other things

but try to give you an appreciation of what it takes

to design this vehicle to survive that trip to space.

A little bit more about me.

What do I have, what business

do I have talking about this?

So I'm a professor in

the mechanics program

in the Engineering Physics

Department at UW-Madison.

I teach courses.

My degrees are in

mechanical engineering,

and I teach engineering

courses in that area.

Also courses in the

astronautics/aeronautics program.

And I have a research group

with about 10 graduate students

and a handful of undergraduates

who are working on

solving various problems

to do with dynamics,

vibration, aerospace,

non-aerospace applications.

Also, for the last few years,

a few years ago I was invited

to be a member of NASA's

Loads and Dynamics

technical discipline team.

This is part of NASA's

Engineering and Safety Center.

So I'll tell you

about that quickly.

NASA's Engineering and

Safety Center, NESC,

everything at NASA has an

acronym you learn quickly.

They have books just to

define all of these acronyms.

But this was set up after

the Challenger accident.

And it was set up as a

separate governing board,

not governing board,

but oversight board.

And so this is a group

of experts from industry,

some from academia, who

provide guidance and

input and oversight.

So within the NESC, there are

various discipline teams.

You know, aero

sciences, avionics,

you can see the list here.

And down on the list

is Loads and Dynamics,

the group that I'm part of.

This is a picture of our

team at a recent meeting.

We actually got to tour the

Virgin Galactic facility there.

And in the middle you may

recognize this young guy among

all of the famous and well-known

engineers around me there.

Okay, but at this point I

should also add a disclaimer.

Anything that I say is not

official statement of NASA.

I have no authority

to speak for them.

I just-- They pay me to give

them advice and that's it.

All right, so let's dive into

what we want to talk about.

Before we start, we need to

understand one concept.

Oh, and here's where the

computer conversion

messed up my equation slightly,

but we'll run with it.

So first we'll start with

Newton's second law.


This is the core of dynamics,

and it's fairly easy

for us to understand.

You could imagine that if

you were pushing on a car,

that that force that you

apply to the back of the car

will eventually cause

the car to accelerate

in that

same direction.

And the heavier the car is,

the less acceleration you get.

And that's where the mass times

acceleration product comes from.

So that's the

basics of dynamics.

The basic dynamic force

is the fact that we push

and something wants to

resist changing speed.

So this is not too big of a deal

if we're just trying to

get a car to take off,

but now imagine that the car

is going down the road

at 60 mils an hour and

suddenly hits a rock, right?

So now we have a

huge acceleration.

Now we're on the other

side of the equation.

The acceleration going from

60 miles an hour to zero

in maybe a fraction of a second

causes a huge force on the

other side of the equation.

And this is one

of the challenges

that we deal with in dynamics.

Here's an example

from the internet.

This is a Ferrari,

about a million-dollar car,

and the rumor has

it that this car

was not very

many miles on it

when the driver took it out

to see how fast it could go,

lost control of the car

and hit a telephone pole

at almost 200 miles an hour.

And these are the

pieces of the car

that were left

after that impact.

Now, let's back up a minute.

Imagine that we took this car

and we put a big

rope on the front

and all of us lined up

and we pulled

on the front of the car.

How many people, hundreds of

people, thousands of people,

do you think it would

take before we could

rip off the front of

the car like that?


I mean, you can't

even imagine getting

enough people together

to do that, right?

And, yet, dynamics,

Newton's laws,

did that in a

fraction of a second

just by getting it up to speed

and hitting a telephone pole.

So these are the kinds of

problems we deal with.

Now, if we think a Ferrari is

fast, though, this picture tries

to give you an idea of how

fast we're talking about

when we talk about space.

So a typical passenger

jet here on the bottom.

That goes three times as fast

as a Ferrari already.

And that's not the fastest

airplane we've ever built.

One of my favorites,

the SR71 Blackbird,

which would go Mach 3.3, or

3.3 times the speed of sound.

So that airplane, we're talking

about 2,000 miles an hour.

And we keep going up.

Up here in this range are

just experimental vehicles.

Most of them never went

very fast at that speed.

And now, if we took this scale

and we put another

one on top of it,

up into the top of the roof,

and another one above that,

we eventually would

get to Mach 25,

which is 25 times

the speed of sound.

That's the speed with which

the shuttle, the space shuttle,

would reenter the atmosphere.

And everything becomes a

little bit counterintuitive

at these kinds of speeds.

For example, you know that if

you put your hand out the window

on a hot day as you drive

down the road in your car,

your hand feels cooler.

The air blowing by

cools it off.

At these speeds, the

friction from the air

passing over the surface

of the vehicle is so great

that it actually wants

to melt the shuttle.

And so it's coated

with these tiles.

If you look closely,

this is a photo I took

at the museum in Washington.

You can go see these now and

get really close to them.

And you can see every single

tile has a serial number.

Every little piece had to

be positioned just right.

And every tile would be checked

after every flight to make sure

that this protective layer

of insulation was intact.

And some of you may have heard

of the Columbia accident.

Here's an example where

this didn't go very well.

So, as you may have heard,

during the launch,

some chunks of foam came off of

the main, the core stage here,

the main fuel tank,

and struck the wing

and did some damage to some

of those thermal tiles.

And, as a result, when the

shuttle was coming back

into the Earth's atmosphere,

some of the hot gas,

the hot air generated

due to the friction

between the aircraft

and the atmosphere

was able to leak into the

wing and actually melted

the aluminum structure

within the wing.

And the wing

eventually came off.

And if this was Hollywood, at

that point the shuttle would

explode and there would be

some great special effects.

Actually, that's not what

happened in real life.

The computers, right before

everything went south,

were showing that they were

using their full ability

to try to right the aircraft,

to try to make up for

this turning moment

that was applied to the

aircraft due to the extra drag

on the one wing

that still remained.

But they didn't have enough

control authority to do that,

and so the aircraft went

into an uncontrolled tumble.

And it was at that point

that the aerodynamic forces

tore the aircraft apart,

and these are the pieces

that were eventually recovered.

So this was the space

shuttle hitting air, right?

But hitting air at

such a high speed

that it generated enough force to tear the shuttle apart.

And unfortunately leading

to the loss of the crew

and those on board.

So this is why we do

engineering, why we do analysis,

and why we study these things to

hopefully ensure that these

types of accidents don't happen

again and that we design

a structure that can take

the loads that it will see.

Here, for example, there was

an adequate appreciation

of the amount of force that

those falling pieces of foam

could exert on the shuttle.

And so nobody thought, well it's

foam, it's spongy stuff, right?

What damage could

that possibly do?

But it did enough.

Okay, so that

covers the impacts.

That's just one type of load

and type of problem

that we might see

with a launch vehicle.

The other one I want to

introduce is vibration.

And you can get a sense for this

thinking about this

child on a swing.

So we all know that if we

walk up to a kid on a swing

and we just shake

them at random,

they won't be very

happy with us

and they won't really

go anywhere, right?

But if we time our pushes just

right so that we always push

whenever the child

is swinging away

and not when they're

coming back towards us

and we time

those pushes just right,

they go higher and higher

and higher, right?

And eventually they

soar into the air

and everyone has a good time.

That is called resonance,

and that's a fundamental property of any system.

And the simplest one is

this spring mass system

that we could represent here.

So you can imagine the child

represents the mass.

There's an inertia there.

The spring is a little

counterintuitive here,

but the gravity actually

provides the spring.

So hanging from that rope,

gravity wants to make the child

always go back to

the center position.

And so that's what we treat

mathematically as the spring.

And then the last property

here is this dashpot

that takes away

some of the energy.

So you know that if you stop

pushing, eventually the child

will come to rest, the

swing will stop again.

And that's because, with very

oscillation, this damping force,

this dashpot, takes

away some energy.

So that's a spring and a mass.

And it turns out any

structure can be represented

as more or less a bunch

of springs and masses.

And so we can

understand anything.

Before we go there, though,

I wanted to make an

analogy with music.

Music is also a

representation of vibration.

It's a manifestation of that.

And we measure vibration

in terms of frequency

or how many oscillations

we see in a second.

So a typical vibration in music

might be in the

hundreds of hertz range.

For example, if we

took this piano

and we go to the

middle of the scale

and we find middle C

and we hit that key...

[piano note]

We'd hear a sound

kind of like that.

[piano note]

And that sound that you hear

is oscillations of sound

of air molecules at

262 cycles per second.

And also there are

harmonics and other things.

But that is just a

vibration of the air

and a vibration of a

string within the piano

that's tuned

to 262 hertz,

transmits sound into the air,

and that eventually

makes it into your ears.

And so musical instruments

have many different keys

that they're able to play,

or many different notes.

Every one of the keys

hits a different string

that's tuned to vibrate

at a different frequency.

And by controlling that

you can make great music.

The same is true--

[piano note] Oops.

The same is true

for any structure.

For example, this is a bridge

that I worked on as part

of my doctoral research.

And this bridge was instrumented

with vibration sensors

all along the length

and on the piers.

And measurements

were then taken,

a large chunk of concrete

was dropped onto the bridge,

and then the vibration that

that caused was measured.

And you can see a

visualization of that here.

So the concrete was dropped

right in the center

of the bridge,

and you'll see this

restart in a minute.

And you can see

that initial impact.

Yep, there it is.

And you can also

see how that turns

into a vibration of the bridge.

Now, the real bridge

didn't bend this much.

I've amplified that

in the simulation.

The real bridge would break

long before moving that much.

The vibration was probably

on the order of millimeters

or maybe into an inch or so.

But even that can be enough for

a stiff structure like a bridge

to cause damage and to

cause an unsafe condition.

But anyway, the interesting

thing here is you can see

the musical notes that

this bridge is playing.

So this looks like

a complicated mix

of all kinds of notes

happening at once,

but that can actually be

decomposed into a bunch

of simpler patterns,

simpler deformation patterns.

And each of these has

a note or a frequency

that it will resonate at

and that it will oscillate

at when it's struck.

And so the bridge is

nothing more than a piano

with its notes tuned

to certain frequencies.

And, as engineers, we designed

the bridge to play the notes

that we want it to play so that

it will survive the loadings

that it will see as

cars drive over it

or as other things happen.

So the same is

true for a rocket.

Now, this is actually a

highly simplified model

of the space launch system.

The actual models,

I've seen them

but they're proprietary

and I wouldn't be able

to show them to you.

But so, actually, I had one of

my grad students just make a

very simple shell structure

that looks kind of like the SLS

so that we could see what the

vibration might look like.

And qualitatively these

will give you a good idea

of what might happen

during launch.

So these are some of the

different frequencies then

that the vehicle might exhibit.

And you can see the different

ways that it deforms.

So in these two here, we see

bending of the rocket.

As it's trying to ascend,

the whole vehicle is bending.

And those are

definitely problematic

because the guidance system

is trying to steer the rocket

and make small adjustments

to the engines

at the bottom to

keep a steady course.

And so as it detects

this bending motion,

it can send the

aircraft off course

and lead to instabilities

that can cause it to crash.

And so it's critical

to understand

these types of motions.

The other thing is that every

type of motion like this

puts stress on the structure

in a different place.

And if you can see in the video,

there's coloring to show you

the hot spots where the

stress is the greatest.

You can maybe see

that best over here.

And so engineers can decompose

a structure like this

into these fundamental motions

and look at where

the stresses happen.

And if the stresses

are too high,

we can redesign it to

try to move the stress

around and eventually

come up with a design

that can survive that

launch environment.

So the other thing I wanted

to point out here is that

NASA's currently under a mandate

to reuse as much as possible,

you know, things

from other programs.

So the main engines on the

SLS under the core stage

here in the middle are engines

from the space shuttle.

And so you might think

that this is pretty easy.

Those engines have

already flown to space,

we'll just slap them on a

new vehicle and away we go.

But hopefully this

kind of illustrates for you

that as we redesign the vehicle,

we changed the way it moves.

And so those engines,

which on the space shuttle

had a certain level of vibration and certain type of motion,

have a totally different

one on this vehicle.

And so a lot of engineering

has to go in to making sure

that the engines themselves

can survive this environment.

They were able to survive

the space shuttle

vibration environment

but they may not

survive this one.

And the same is true

for every little piece.

Every flight computer,

every nut and bolt

has to be positioned

and designed

such that it can take the

forces that it will see.

So let's go through in

a little more detail

and talk about the

different stages of launch

and the different

loadings that we see.

And I'll use the space shuttle

as an example because we have

some open public domain

information on this.

So this is a picture of the

space shuttle on the launchpad.

And the shuttle was tied to the

launchpad through four posts

under each one of the

solid rocket boosters.

I think these were on the

order of one-inch steel

or something like that.

And those posts

would hold it there,

and the main engines on the shuttle would be fired up.

And you can imagine that

would cause the whole vehicle

to want to kind of lean over.

It will also take

some of the weight.

Originally those eight posts

are holding all the weight

of that vehicle.

Well, that takes some

of the weight off

and it causes the vehicle

to kind of flex over.

But it doesn't stay there.

It, just like the

kid on the swing,

it swings a little

farther than it wants to

and starts to come back.

And they call that the twang.

The astronauts who

were right at the top,

they're seated up at the

very top of the shuttle,

would feel this swinging motion, this twang,

that happened at about one

cycle per second or so,

similar to the

kid on the swing.

And a little after

that twang was over,

some explosive bolts at the

bottom of those posts

would release the shuttle,

and it would begin to ascend.

And so we'd see the solid

rocket boosters going

and the main engines going.

And you can see the fire

coming out of the solid

rocket boosters.

They're a solid propellant.

The main engine burns

hydrogen and oxygen,

which makes an invisible gas.

But anyway, all of that thrust

would drive the vehicle

to higher and higher speeds.

And as the vehicle is ascending,

the air is getting

thinner and thinner

but its speed is getting

higher and higher.

And higher speed means

more force from the wind,

more air, more drag.

But the higher altitude

means less air, less drag.

And so if you compare

those two forces,

the space shuttle eventually

reaches a point called max Q,

or maximum aerodynamic pressure.

And that's one of

the critical places

during the launch profile.

The vehicle is under tremendous

pressure from the front.

This picture, this is

actually a picture

of a shockwave on the vehicle.

But it gives you an idea of the

pressure that it's feeling.

And so for the shuttle

this would happen

about one minute into flight, and it would--

So the vehicle would hit

this critical speed,

and to make sure

that the vehicle survived,

we'd do two things.

We'd throttle back

the main engines

to slow the acceleration

a little bit to reduce the load,

and the solid rocket boosters themselves were actually

designed so that at this

point of the flight

the thrust would

scale back to 71% or so

of the nominal maximum thrust.

So all of that would happen

and we'd get through this

point of maximum shaking,

maximum noise, and the

vehicle would eventually

break free of the air and

get to even higher speeds.

- What speed is it at

when it hits max Q?

- You know what?

I don't know off

the top of my head.

Yeah, good question.

I know the information I found

said it's one minute into flight, about 35,000 feet.

The other piece of

data I can give you

is within a few

minutes of launch--

we launch in Florida--

within a few minutes of

launch it's over Europe.

So you can do the math and

figure out how many thousands

of miles that is in a

couple of minutes, right?

But, yeah, good question.

Okay, so shows you

the full profile.

And I'll show this just to

mention one or two other events.

Once we get into space,

there's no more air,

things get much

easier to deal with.

The last main event that

we have to worry about

is the separation

of the boosters

and the main

tank from the vehicle.

And this picture

shows a test

that's done on the

noise of a rocket.

Here's an explosive going off

to simulate this stage separation event.

And, you know, we can't just

cut a string and let them go.

These, they're subjected

to huge forces,

so there's strong connections

holding them there.

And so it takes a

powerful explosive event

or something to break

those boosters free.

But we need to be sure that the

shockwaves that that explosion

sends into the rest of the

vehicle doesn't damage it.

And so a lot of testing goes in

to make sure the electronics

and the vehicle can

survive those events.

And you can also imagine

how the current SLS

is going to be more of a challenge in the shuttle.

The shuttle is kind

of riding here,

kind of loosely connected

to this main tank.

And these two boosters

are connected

to the main tank as well.

So those blast events

will put a lot of loading

onto that main fuel tank,

but we're going to

drop that in a while

so it's not as

critical as putting it

on the launch

vehicle itself.

Okay, so that's the launch

and the flight environment.

To put some numbers on

some of these things,

I've taken some numbers

from the Delta IV

rocket's launch handbook.

So this is a handbook

that's out there.

If you were designing a payload

to go up on the Delta IV,

this gives you all the

information that you would have

to design your satellite

or your payload

so that it could make

it safely to space.

So the first thing that

you would think about

is static acceleration.

Because the rocket is

accelerating at such a rate,

it's accelerating at

a rate of six times

or five times that of gravity,

so the total acceleration

that you would feel

in the rocket could be 6 Gs

or six times out of gravity.

So if we go back to our

Statue of Liberty example,

we'd have to design

Lady Liberty's arm

to be strong enough to hold six times the weight of that torch

if we want it to

get to space.

So that's what's

in consideration

as we design the SLS.

There's an acoustic environment,

which is measured in decibels.

So to give you an idea of

what that would be like,

normal conversation

here in this room

we have a level of

maybe 60 decibels.

If I fired up a lawnmower

in here, we'd get up to 90.

Every 20 decibels is 10 times

more pressure at sound.

And then a rock concert

is 1,000 times louder

than what we have in this

room right here at 120 dB.

And you go a little beyond that

and you find finally

get to the level

that's within the inside

of the Delta IV rocket.

Okay, so there's the acoustic

environment we worry about,

and then last is the

dynamic acceleration

or all of the shaking,

all of these resonant

forces that we talked about.

Those also can be measured in

Gs, acceleration due to gravity,

and this might be

on the order of

eight Gs root mean square, RMS.

So if we had an oscillation,

for example, here in blue,

the RMS is a measure of the

height or the amplitude

of that oscillation.

So we might have plus or

minus 14 times gravity

the dynamic motion that the

vehicle is going through.

So to put that into perspective,

a typical hard disk drive

within your computer,

if you still have one of these

older spinning disk models,

is designed to

operate at 0.4 Gs RMS

or 20 times less.

And it could survive

up to three Gs.

So even if you were to turn

it off and stow it away,

at eight Gs, the hard drive

wouldn't be operational

when it got to space.

And so you can't just go,

you know, slap an iPhone

or a MacBook on the SLS

as the guidance computer.

Every little bit of

electronics has to be designed

to be able to survive

these kinds of loads.

Not to mention the thermal and

the radiation effects in space.

Okay, so how do we do that?

Well, this is actually a picture

of an electrodynamic shaker about the size of me,

that's housed at UW-Madison.

What it is

is it's a huge speaker,

a big electromagnet here.

And there's a plate on top.

And on that plate we can mount

things like this hard drive.

And then we can

command that plate

to go through an

eight G environment.

And, actually, what's done is a

recording of a launch

environment is actually played

back, and our electronics ride

out that environment, sometimes

with some safety factor.

And then, if the electronics

survive, they're certified

to fly and they're allowed

to go up into space.

So I have a little video that I

can show you to try to give you

an idea of what we

might see here.

So this is a video of a

little wing structure

on top of the shaker.

And you can't really see

the motion of this plate.

This is moving with something

on the order of, you know,

a launch-type environment.

And as it does, you can

see the huge vibration

that's being generated due

to resonance of the wing.

We're driving it right

at a resonance mode.

And you wouldn't be able

to see that motion either

because it's happening

very quickly,

but there's a strobe light

that happens to be timed

with the vibration.

And that's what

allows you to see

what seems to be a slow

motion of the wing there.

So that gives you an idea,

and if I had the audio file,

I could play another video

that would actually show you

what it sounds like

to be in the room.

But it's actually a very

loud, unpleasant sound.


So we'll skip over that

and go back to

the presentation here.

Okay, so that's our wing demo.

And, actually, the

first time I saw this,

our technician was running it

and I saw the level

of vibration and I said,

"That doesn't really look safe.

"Are you sure that thing

isn't going to break?"

And he said, "Oh, no. I've done this demo, you know, for years.

It's fine."

He'd done the demo for a minute

or two once per year for years.

So, you know, we thought

it wasn't a big deal.

But sure enough,

on this rendition,

a few minutes into it,

we started to hear some funny

noises coming from the machine.

We took it apart and this is

the piece of the wing right

in the back there,

under the clamp.

And you might not be

able to see on the video

but it's actually

developed a crack,

and half of the wing

just about cracked off,

was ready to crack off.

So these are the things we're

trying to design against,

trying to make this

structure strong enough,

stiff enough, so that

this doesn't happen.

And that is easy if we're

building something small.

So this picture on the far

left shows you a cube sat.

This is a little tiny

satellite about this size.

And these are relatively

easy to design.

They're quite small and rigid

and so it's easy to make them

such that they can make it to

space without breaking.

So you can actually

design one of these,

and for relatively inexpensive

they can be added on to other

payloads and put into space.

Now, if we want to

make something larger,

like the Hubble Telescope there,

it's much larger,

much more flexible,

and it takes a lot of weight to

stiffen the structure enough

that it can survive

that environment.

And even more so, you know,

as we make the largest

and most complicated

vehicle ever to be built,

the flexibility is larger,

the size is larger,

and so it's even

more of a challenge

to get this to survive

the trip to space.

So how do we do that?

How do we make sure that

our rocket can make

it safely to space?

Well, what we do is

create simulation models.

These are called finite element

models because we take the part,

here's an example of

a two-cylinder engine,

and we break the part into

lots of little squares.

And for every little square

we can solve the

mathematical equations

for force and


And so the computer can help

us to go through millions

of squares and add all that up

and figure out where

the highest stresses are,

where the biggest

loads are.

So this method works great.

The challenge is that, you know,

you see to represent this

complicated geometry just of

this little two-cylinder engine,

I already have something like

a million squares, right?

And our computers can't handle

much more than, you know,

tens of millions of squares.

Even a supercomputer

might cap out

at a hundred million squares.

So even for a little engine

like this, engineers will spend

a lot of time

simplifying it down,

trying to figure out what

are the key elements,

and then meshing that with

larger squares that will be,

that can be run on

our modern computers.

Okay, so all of that work is

done, that modeling is done,

and we've made all kinds of

simplifications and assumptions

to make the model tractable

on our computer.

And so the next thing

we need to do is

do a test to see if

that model is correct.

And if the model is correct,

we go forward.

If not, we go back

and we fix things.

And in the aerospace industry

an expert told me recently in

30 years of launch vehicles,

they've had only one case where

the model was ever

correct the first time.

Every other case it

required some updating

and fixing after the testing.

Okay, so those tests

are done on the ground

on pieces of the vehicle.

And then that model is used

to estimate the forces

and the loading as the

vehicle goes into space.

And there might be

thousands of load cases

for each stage of flight.

You know, the initial launch,

10 feet up

when the weight of

the rocket is reduced

due to all of the propellant

that's burned already,

and at every stage in flight,

and especially the

critical ones like max Q.

And so all of those

cases will be analyzed

and will have key components,

we're looking at stress,

and we'll go back

to the drawing board

and redesign places where

the stress is too high

and run the simulation

again and keep doing that

until we have a vehicle we think

can survive launch to space.

So that work is underway

right now for the SLS.

So this shows you some of

the different components.

These are at various

stages of completion.

The boosters are actually

space shuttle boosters

with one additional segment.

So those are probably

farthest along.

And then there are various

stages of completion.

And so right now, as

each piece is built,

the actual hardware

that will one day fly

is being tested by applying

forces, by measuring vibration.

And that's being checked

against computer models

to see if those

models are accurate

and to fix them

and update them.

And all of this is being

done hopefully on schedule

to get this vehicle ready to

launch within a year or two.

Okay, so, you know, I showed

you earlier a computer model

that my student made in an hour.

Modeling a real

vehicle like this is

a much more complicated process.

Every nut and bolt, every place

you might have a failure

has to be checked.

And the models are

incredibly complicated.

One of the most challenging

things actually is modeling

the places where the different

components come together.

I don't have any examples of

the SLS that I can show you,

but I do have some pictures

from an F16 fighter engine.

So this is the engine of an F16,

and you can see the housing here

that holds it all together.

And this picture shows you what

the bolts look like that are

holding together the different

piece of the housing.

And you look at that and think,

well, that should be pretty easy

to make a computer model of.

I'll just glue

everything together

everywhere it comes together

and not worry about it.

But it turns out that these

joints are one of our biggest

challenges in modeling

vehicles like this

or like the SLS,

because those bolts

don't actually just

squeeze things together.

On a microscopic level,

when we tighten a bolt,

we actually squeeze the

material under the bolt

but it opens up in other places.

And there's a

complicated pattern

of loading and deformation.

And even more challenging is the

fact that this model is assuming

that everything is flat and

smooth, but, in reality,

there's some roughness

to the material.

And so there might be

a microscopic piece

of material and asperity.

And, you know, if we really

wanted to model it on detail,

we have to think about

the individual atoms that are

coming into contact and the

bonding forces between them.

That's where friction and

other forces come from.

So our computers

today can barely,

actually can barely model

this for one or two joints,

much less for an entire engine.

And so we have to use all

kinds of assumptions.

And certainly, everything

down here just has to be,

we have to do a test and

characterize a material.

None of that can

be modeled currently.

So a vehicle like this, each

component has to be modeled.

We have simplifications, and

everywhere there's a joint

we make some kind of assumption.

Like assuming that we can

connect all of the pieces here

to all of the pieces here

through one little element

that will represent our bolt.

And then, as the test data

comes in, we can figure out

what the stiffness and strength

of that element should be

to match what we saw in tests.

So I won't go into this here

in the interest of time,

but the other thing that's

important in all of this

is the fact that the

friction at these bolts

is what gives us

most of our damping.

Most of the energy

that's absorbed

is absorbed in friction

due to these joints.

And that's something

that we can't predict.

That has to do with

these microscale

and nanoscale effects.

So we rely on testing

today just as we did,

you know, even in the 1960s.

And why do we care so much

about this dissipation?

Well, just as the child on the

swing, if we apply a force,

the vibration amplitude grows.

And the only thing that can

limit the amplitude of vibration

is the damping or the friction

and dissipation

within the structure.

And if we get the damping wrong,

suppose our computer

model tells us

that this red curve is the

damping level that we have,

when in fact, we have less

damping, we have the blue curve.

Well that's twice as much

vibration, twice as much stress,

twice as much loading.

And on launch day that could

have disastrous effects

because if we haven't

designed for that,

components can fail and our vehicle goes up in flames.

Okay, so hopefully I've

given you some appreciation

for why this still is hard.

It's hard to imagine that, you

know, in the '60s we designed

a space shuttle using computers

like the ones you see there.

Now we have Google

data centers with--

You know, I don't even know

what the number is.

It's probably way more

than billions of times

the computational capacity

of the computers back there.

And yet our models are still

probably billions and billions

of times too coarse to

even begin to capture

all of the physics

that we might need.

So we still rely on testing

and updating to make sure

that our computational

models are correct.

This same process is used for

all kinds of things, for cars,

you know, motorcycles, musical

instruments and speakers,

even cellphones

to make sure that

they don't break when

they're dropped.

The one difference is

in all of these cases

they can build a prototype,

they can drive it,

and they can see whether

they got what they wanted

and whether this

structure is okay.

With the SLS, the first time

we will know whether we were

successful is when it

flies for the first time.

And so that's why so much work

is going into testing

each component

and checking each

step along the way

because we're going to put a

lot of confidence in that model

and we're going to believe

it when the model says,

yes, the vehicle

is ready to fly.

All right, so I thank you

for listening today.

I hope you've got a little bit

of an appreciation for some

of the aspects of launch and

the dynamic environments

that spacecraft and other

vehicles experience.

And we're excited to see what

happens with this vehicle.

The testing is well underway.

The first launch is expected in

early 2019 at this point.

And so if all of the different

pieces can come together in time

and all of those computer

models can be tuned

and everything comes together

and we come up with a design

that can survive

the trip to space,

we hope to see that launch

in about a year or two.

All right, and with that I'll

also just quickly acknowledge

all of my graduate or

some of my graduate

and undergraduate students

and the funding agencies

that have made our work possible

and allowed us to study

these types of problems.

And, with that, I'd be

happy to take any questions

that you may have.

Thanks for listening.


Share this page