00:00:00
This is the one I made very soon after I got
the idea. It was back in 1997. So this is
00:00:06
the first prototype, the first of its kind. The first one that I made out of a child's toy Meccano. Today I'm at the
00:00:12
University of Cambridge to speak to a professor
who has invented what has been called a genius
00:00:17
device. We're going to hear the story behind
this device directly from its inventor. And
00:00:22
this is a fascinating story that has it all. The
biggest spy scandal in the history of Formula 1.
00:00:27
A humble professor who has proved that an
established engineering principle used
00:00:31
for more than 70 years was wrong. And
a secret device that some of the best
00:00:35
engineers in the world were not able to
understand despite looking right at it.
00:00:43
Before moving on, let me briefly recall some
elements of the spy scandal to give you a
00:00:48
clear sense of the impact that this professor has
had on Formula 1. In 2007, the Formula 1 world was
00:00:54
rocked by what became the biggest spy scandal
in the history of the sport. One of the elements
00:01:00
at the center of that scandal was a mysterious
device called J-Damper. Renault had obtained a
00:01:06
drawing of the never heard before J-Damper from an ex employee of McLaren and Renault tried to have
00:01:11
McLaren disqualified because they thought that the
J-Damper violated Formula 1 regulations. In turn,
00:01:16
McLaren tried to have Renault penalized because they
committed industrial espionage by getting hold of
00:01:22
the drawing of that device. When the proceedings
of the World Motorsport Council investigation
00:01:27
were made public, they kept secret the nature of
the device. The only information made public was
00:01:33
that McLaren was not disqualified because the J-Damper did not violate any rule and in fact the
00:01:38
engineers at Renault misunderstood what this device
did. Medias around the world tried to figure out
00:01:43
what this device was and eventually a connection
was made with Malcolm Smith, Professor of Control
00:01:48
Engineering at the University of Cambridge. Today
we have the opportunity to speak to Professor
00:01:52
Smith and have those events narrated by his
very protagonist. So let's go. Professor Smith,
00:01:58
your background is in mathematical control
theory and then later on you became known
00:02:03
for your work on the inerter, you invented this
device called the inerter. But in this interview,
00:02:09
I would like to go through the story of the
inerter and how you came up with the idea and maybe
00:02:14
we can start with your background. So you can tell
us a bit what you did before the inerter. Yes. Well,
00:02:21
Giordano, thank you and it's always a pleasure
to explain this and it connects with my
00:02:29
background as you say. I started out in control
theory. My first degree was in mathematics and
00:02:38
for my PhD I worked on multivariable robust
control, frequency response methods,
00:02:45
so how to design controllers when you've got many
inputs and many outputs how to generalize the
00:02:51
classical theory to that situation and later I got
involved in something called H-infinity control
00:02:58
partly through visiting George Zames at McGill
University in Montreal, and one of the key
00:03:05
ideas is stabilizing or looking at the class
of all controllers that stabilize a given plant.
00:03:15
So then you try to optimize over that class
to minimize whatever you want: a robustness measure or
00:03:24
sensitivity but that being able to solve over
a general set and to optimize something in the
00:03:30
frequency domain was a very important idea. Okay.
Thank you very much. And so then how did you get
00:03:36
involved in Formula 1? Yes, that was a chance
event. I had just joined this university as
00:03:44
a new lecturer in 1990 and the first summer
1991, August of that year. Everyone was on
00:03:53
holiday except myself really among the academic
staff and a phone call came through from Williams
00:03:59
Grand Prix Engineering, from Paddy Lowe who I didn't
know at the time. Those who know about
00:04:08
Formula 1 will know that he eventually became a
very famous Formula 1 designer and engineering
00:04:15
director at McLaren and then later at Mercedes. At
the time he was in charge of active suspension
00:04:25
development at Williams Grand Prix and they had a
stability problem. So he phoned the group. He
00:04:36
was wanting to speak to Professor Keith Glover.
He'd been taught by him at Cambridge as
00:04:44
an undergraduate, but Keith was on holiday for
another two weeks. And Paddy explained they had
00:04:51
a test at Silverstone the following week. And
so I was rather bold and in taking the job and
00:05:00
was hoping that Keith would not be annoyed.
Fortunately Keith was pleased that I took the job.
00:05:07
But that got me started in consulting
on this active suspension system. So you started
00:05:16
to be involved as a consultant. Then did this
idea started to have an influence on your
00:05:22
research. Yes. Well, not initially. Perhaps one
should say a little bit about active suspensions
00:05:27
first. A car, a normal car suspension has
springs and dampers and anti-roll bars. Those are
00:05:34
also springs. That's the basic normal passive
suspension. With an active suspension, you're
00:05:41
replacing the springs and dampers with powered
hydraulics. You could have a piston and a cylinder
00:05:47
and you'd have pressured hydraulic fluid that
can be directed with silver valves above and
00:05:56
below the piston to extend and contract. So that's
if you like the actuator in the control system.
00:06:03
Then you have measurements all around the car:
accelerometers, deflection sensors, pressure
00:06:09
gauges. You can have as many as 20 measurements
on the car being fed back into a computer that
00:06:17
implements the controller and then that drives
these four actuators that extends and contracts
00:06:22
the strut at each wheel station. I was brought in
as a control specialist knowing about
00:06:30
multivariable control design. And of course you
bring those methods in and you apply them.
00:06:39
As we made progress and I worked more on these
I became interested in the systems themselves and
00:06:48
I started to develop some research problems.
There are many different layouts, mechanical
00:06:56
layouts that are possible and different sensing
arrangements and then the question is which is
00:07:03
the best, and then how do you design the controller
to produce maximum improvement in mechanical grip?
00:07:13
That's how the tire deflects and contacts with
the road and how do you best control the ride
00:07:19
height and so on. So the engineering setup the
design of these systems becomes fused with the
00:07:32
control systems design in possibly a complicated
way. So that became a research question for me.
00:07:37
How do you simplify things down and find a
way to approach this, which could answer
00:07:45
some basic questions? So you started then to do
research on active suspension systems. Was this
00:07:51
then implemented by Williams? Well, there wasn't
really time. The consulting work and the
00:07:58
implementation of the control system happened
so fast from August 1991. The system was
00:08:05
actually raced for the first time with Nigel
Mansell at the South African Grand Prix in
00:08:12
1992. And Williams had extraordinary success
in the 92 season and the 93 season. They
00:08:22
won the championship easily by a big margin and
then active suspension was banned at the end of
00:08:29
the 93 season. Why was it banned? This is often a
complicated story, but technologies often come
00:08:38
and go and and get banned in Formula 1. And most
commonly a technology when it's successful and
00:08:45
is producing a big advantage for one team, it can
get banned at that stage because the races become
00:08:51
boring because a team is always winning. One team
is always winning. So people don't want to watch
00:08:57
the races or are not so interested and then there
is pressure on the authorities to equalize things
00:09:05
So that the sponsors are happy and so on.
Okay. So unfortunately what you were
00:09:09
developing couldn't be used anymore but then
did you manage to do some other progress? Yes
00:09:16
well all was not lost in the sense that I'd
started to think about the design of these
00:09:22
systems and to bring in some of the methods.
I mentioned H-infinity and one of the ideas
00:09:28
there is to parameterize all controllers that
can stabilize a given plant and to be able to
00:09:35
optimize a performance measure over that set.
But then that's the start of the inerter story
00:09:42
because you can take that same idea and instead
of optimizing over all stabilizing controllers,
00:09:50
you can try to optimize over all passive
suspensions. Now I knew some theory from my PhD
00:09:59
time about passivity of electrical circuits and
I knew it was possible to completely characterize
00:10:08
all the electrical circuits you can build out of
passive components. So then it seemed natural to
00:10:17
say well we should do the same with mechanical
mechanisms, passive mechanical mechanisms and
00:10:27
I think people who knew that theory would have
expected that it was a direct correspondence and
00:10:33
that would be a relatively simple thing to do.
It turned out not to be straightforward and
00:10:40
some of the difficulties illustrate why the inerter
hadn't been done before if I can put it that way.
00:10:47
So if I understand correctly, so given
the fact that you know how to realize,
00:10:52
you want to apply the same ideas on mechanical
systems and I actually studied something like
00:10:58
this in university and the idea is that there
is an analogy between electrical circuits and
00:11:06
mechanical systems, hydraulic systems, heat
systems, basically any linear system. And
00:11:12
the idea is that you need to identify what
behaves like a current. So this call it a through
00:11:18
variable and what behaves like a voltage, which
is an across variable and so for mechanical
00:11:23
systems you have the force that behaves like a
current and you have the... what do you have for the
00:11:29
voltage? The voltage becomes the velocity
and the force becomes the current
00:11:36
in the analogy you're speaking of, the force-
current analogy. I remember the book said that
00:11:41
and I mean it was very convincing in the fact
that the analogy then of the components were that
00:11:48
the damper is a resistor. So it dissipates energy.
The inductor behaves like a spring and then the
00:11:56
capacitor behaves like a mass, right? So
the problem is solved. You just have to use this
00:12:01
this mapping, or not? Yes, the mapping that
you've described is the one that we use
00:12:07
to understand this. And the first important
thing to say is that it's power preserving
00:12:15
because you're mapping voltage and current onto
velocity and force. So the product of voltage and
00:12:23
current is a power and the product of force and
velocity is a power. So anything that's passive in
00:12:30
the electrical domain, if you map it over exactly,
will be passive in the mechanical domain and
00:12:36
vice versa. And the analogy you've just
described is what you find in the textbooks
00:12:45
with the element correspondences. By the way,
just a brief aside, there are two analogies. But
00:12:51
the advantage of the one we've described that's
based on through and across variables is that
00:12:57
series connections in one domain become series
connections in the other, parallel become parallel.
00:13:03
So the circuits are topologically identical.
And the other thing that's very good about
00:13:09
that analogy is that ground, electrical ground,
which is a datum voltage, becomes a point
00:13:18
with constant velocity, which is a reference point
in the inertial frame. So that allows one to
00:13:30
go one step further to discuss the analogy
you've described. Spring and inductor, resistor
00:13:39
and damper, those are straightforward because those
are both two terminal devices. So let's think
00:13:48
what are terminals in the mechanical domain,
people often don't think about this. If you take a
00:13:54
spring, the terminals are the attachment points,
the two end points of the spring. And we're
00:14:01
looking at the force, the equal and opposite force
at the terminals produced by the spring from its
00:14:09
relative displacement. Right? You push the spring
and you receive a force. Yes. Yes. Yes. That's
00:14:14
right. But it's an equal and opposite
force applied at the terminals. Similarly with the
00:14:20
damper the terminals are the connection points.
One connection point will be the housing of
00:14:28
the damper and the other will be the piston
rod and those are the two connecting points,
00:14:36
and both the terminals of those of the spring
and the damper are freely movable in space.
00:14:44
Now let's go to the third element: the mass. So
what are the terminals on the mass element? Well
00:14:55
if you think about the way we do mechanical
modeling the mass is usually treated as a point
00:15:00
mass and its motion is governed by Newton's second
law and that's the acceleration in an inertial
00:15:09
frame. So if you're careful and you write the
circuit diagram of the element you'll see that the
00:15:15
mass element is actually analogous to a grounded
capacitor! So there's only one... you can't slide the
00:15:23
mass in and out with two ends that are freely
movable in space in the way you can do with a
00:15:28
damper. The mass is a point mass. So the mass
is analogous to a grounded capacitor and that limits
00:15:41
what you can do. Yes. That's not my observation. Okay. That's in a sense well known. But most of the
00:15:49
textbooks don't really explain it and also some
of the books that really appreciate the most try
00:15:55
to cover it up by drawing a symbol for the mass
that looks like a slider. So here it is, the
00:16:05
theory if you like, from electrical circuits
forces a certain question because to build our
00:16:14
electrical circuits you need the three elements,
resistor, capacitor, and inductor and they all need
00:16:20
to have two terminals. So the natural equivalent...
you have a problem with the mass because
00:16:30
it's a one terminal device and so the circuits
that you can realize with it are more limited.
00:16:38
So it's impossible then to go from the circuit
to the mechanical device? Well, yes. I became puzzled
00:16:46
at this point and the theoretical work forced
a question. It's easy to see that if one were
00:16:58
to have a device which is genuinely two terminals,
like the spring and damper, it has two attachment
00:17:04
points which are freely movable in space, with
the property that the equal and opposite force
00:17:10
at the attachment points is proportional to the
relative acceleration between those terminals.
00:17:17
So if one had such a device then the mapping
would be perfect for the synthesis theory and then
00:17:27
you could say: we can build any of these positive
real complex impedances. We can build them
00:17:35
mechanically. We can have a mechanism here
that we can design and build and hopefully put it
00:17:40
on the car if we can make it small enough and
light enough. So that was the point I was
00:17:47
stuck at that point for a while, and thinking
that "spring-mass-damper it's in all the
00:17:54
books", that's the standard and that's
how it must be and maybe there is a limitation
00:18:01
that mechanical circuits are not completely
equivalent to electrical ones. When did you
00:18:08
realize that actually something more could... Well
it was a thought experiment to...
00:18:17
I was trying to prove that you couldn't do it.
You couldn't make something like that. But if you
00:18:22
think positively, supposing you could build such
a device and you held it in your hands, what would
00:18:30
it feel like? And one of the properties would
be that once you've set it in motion, it keeps
00:18:38
going with constant velocity. Having realized that
it's fairly quick, it's almost immediate really
00:18:48
to think of a way of making such a thing. And I
sketched something like this that we have on
00:18:58
the table. And this is the first thing
I did after having thought of that idea that
00:19:11
we should try to build one and and play with it
and understand it. So this is the device. This was
00:19:17
the first one I made and it's made out of a Meccano,
which is a child's modeling, building kit that
00:19:24
I was familiar with from childhood. So you
can see it's a two terminal device. The housing
00:19:29
here is one terminal and the end of the rod here
that moves in and out is the other terminal. And
00:19:37
the rod itself is a rack connecting with a
pinion. So we have a rack and pinion mechanism.
00:19:44
And the pinion is on this first axle here. And
then there's a gear and a pinion connecting
00:19:52
to a second shaft. And a similar gear. They're
both 5:1 ratio, gear wheel pinion here, and on
00:20:00
the third shaft we have this flywheel. So as the
rod moves in and out the flywheel shaft and the
00:20:09
flywheel rotates in proportion. So we need to
think about inertance. Inertance is the constant
00:20:19
of proportionality between the force and the
relative acceleration and it's in kilograms.
00:20:27
That's different to the mass of the device.
This weighs less than a kilogram. That's
00:20:33
the mass of the device. The inertance is the
constant of proportionality between the relative
00:20:37
acceleration and the force. And typically
for inerters it is much larger than the mass.
00:20:46
So we can get an idea approximately of the value
of the inertance here. The two 5:1 ratios... that
00:20:55
gives a multiplier of 25. The other thing that's
significant is the pinion radius and
00:21:02
the radius of giration of the flywheel. It's
approximately 2:1. So that gives another factor of
00:21:07
two making 50. But the 50 is a ratio of forces but
also it's an inverse relationship of velocities.
00:21:16
So it actually squares up. You get a multiplier of
2500 times the mass of the flywheel. The flywheel
00:21:24
is 100 grams, 0.1 of a kilogram. So overall this device
has an inertance of 250 kg approximately. So
00:21:33
that's a quarter of a ton between the terminals.
You can, with a small force, you can move the rod,
00:21:41
but not very quickly. So what you have to think
here is resistance to acceleration or resistance
00:21:49
to relative acceleration is the characteristic.
So can you actually explain a bit better this
00:21:55
point? So here you have an inherance of 250 kg,
right? But you can move it with very little force.
00:22:04
Can you explain better why that's the case? I
mean, because if you imagine a mass that is 250 kg,
00:22:09
it's difficult to push it. So what is actually,
what does 250 kg mean? Yes. So if you imagine
00:22:15
if we had a 250 kg mass on this table and it was a
frictionless table. So like on ice. Yes. You would
00:22:23
be able to push it, right? With a small force
and it would start to accelerate and it would keep
00:22:28
going for some some period, that's just like the
inerter does. And so, what makes you think that
00:22:37
you can't move 250 kgs is first of all, if you had
to support that weight. Most people would not
00:22:45
be able to lift 250 kg. But if you're just pushing
it along a plane, you still wouldn't be able to do
00:22:53
that if there's a lot of friction. So a small
car for example, one person can't push a car
00:23:00
because of the friction in the bearings
and the tires. It's not because of the mass itself.
00:23:06
A small force will accelerate it a bit, so it
will move a little bit, and that's the same with
00:23:12
the inerter because these have rather low friction.
Okay. In a certain way this reminds a bit of a damper
00:23:19
in the sense where there is one terminal which
you push and there is some kind of resistance.
00:23:23
Maybe you can explain what the difference is. Alright, let's look at this device here. This is
00:23:29
a Formula 1 damper that was made
by Penske Racing Shocks. So,
00:23:37
this is just a rubber handle, so one can move
it by hand, but you would have a ball
00:23:45
joint here which would connect to some suspension
element and another connector at this end. But
00:23:51
it's a two terminal device and we have a rod
moving in and out of... we have a piston and cylinder
00:23:58
here and a floating piston and fluid inside.
But this is a totally different device than the
00:24:04
inerter because the force, the equal and opposite
force, is proportional to relative velocity. So
00:24:11
to keep it moving you have to apply a constant
force. To move it faster then the force increases.
00:24:17
So this is force proportional to relative velocity
whereas for the inerter is force proportional to
00:24:24
relative acceleration. So essentially
with the inerter you will feel a resistance when
00:24:29
you start to push... exactly it's changes in velocity...
changes in velocities. Yes. And okay so this is
00:24:35
one particular realization of the inerter, but is this what has been then implemented in formula one? Fairly
00:24:43
soon we started to think of different ways to
do it. What turns out to be a nice construction is
00:24:50
is to have the flywheel spin axially about
the rod and you can do that with a ball screw
00:24:58
mechanism. Do you have an example? Yes, I can show you one of these. Take a look at this. This is
00:25:06
a demonstrator that we made in the engineering
department and it has a perspect cover
00:25:16
so you can see the mechanism inside it. So you
see the threaded rod and inside of this
00:25:26
assembly here is the nut which spins when
there's relative velocity between the two ends
00:25:33
of the device and attached to the nut here is the
flywheel. This is the flywheel which rotates
00:25:42
as you get a relative velocity [sic] between the
terminals. So this is a compact simple way to
00:25:55
implement inerter without having multiple shafts
and racks and pinions and so on. So maybe can we
00:26:02
now try to understand how these are used in formula
one. So okay you come up with this new device.
00:26:08
Yes. Then what happened? Having got the idea and
having explored it with with Cambridge Enterprise
00:26:15
in terms of whether to protect it, we put it to
the test and we approached a team in Formula 1.
00:26:22
And by this time, Paddy Lowe that I'd worked with
at Williams, he'd moved to McLaren and I also knew
00:26:31
other people at McLaren, Dick Glover
for example, and we decided to contact
00:26:36
them and to take the inerter idea together with
some initial calculations I'd done on possible
00:26:43
benefits... and McLaren were interested and they
signed an agreement with the University to
00:26:51
develop it and we went through all the stages of
engineering design and simulation and whether
00:27:01
it can... what it can improve in the car in terms of
reduced oscillations or improving mechanical
00:27:08
grip and all the way through to testing and to
producing an actual lap time gain on the car.
00:27:17
And once that was achieved, it went on the car
really for the next race and that was an exciting
00:27:26
event, it was 2005. So time has elapsed.
So yeah, what is exactly the timeline? So when
00:27:33
is that you come up with the idea, the first
prototype and then eventually got it on the
00:27:38
car? My first Meccano model goes back to 1997. My approach to McLaren was around about 2002 and
00:27:48
the development happened well there's not a
lot I can say about the internal development but
00:27:56
let me just point out the known facts, that
it was raced for the first time in 2005 and it
00:28:04
produced a lap time gain and the first race was
with Kimi Räikkönen driving and it was at the
00:28:14
Spanish Grand Prix in 2005 and he won the race for
McLaren's first victory of that year, and in fact
00:28:24
it led to several victories, a number of
victories throughout the year. McLaren were
00:28:30
close to getting the championship but didn't quite
manage it that year. That was very exciting.
00:28:36
Yes, it was very exciting for me... thrilling
really that a theoretical concept is on the car
00:28:46
that you're watching on the Grand Prix at the
weekend. So at that stage it wasn't a story
00:28:56
I could I could relate to anyone because McLaren
went to fairly elaborate lengths to conceal the
00:29:05
nature of the device. In fact, this one is
an illustration of an early ball screw inerter
00:29:13
of roughly the size and packaging that
we used in Formula 1. And so it's again
00:29:22
a two terminal device and it's similar to the
one I showed you before with the the ball screw
00:29:28
and the flywheel rotating axially about the
line joining the two terminals. But otherwise
00:29:38
you might think this is some sort of exotic
damper. Indeed. And if you look at the
00:29:43
car from some distance you wouldn't really
think that there's anything unusual here.
00:29:51
So McLaren was able to conceal it for quite
a while. Can you tell us a bit how they did that? Yes. So first of all they
00:30:03
named it something different. They called it the
J-damper and that was a deliberate attempt to
00:30:11
decouple what they were doing with the technical
literature. We were publishing papers on the
00:30:15
inerter and describing the idea and its advantages.
So the idea was public? It was public domain and
00:30:21
that was part of the agreement with McLaren that
we would continue with the academic research and
00:30:27
that that wouldn't be held back and so their
strategy was to try to to cover it up, quite
00:30:36
legally in terms of the sport. You don't have to
tell your competitors exactly how you're doing
00:30:41
things. So so that was fine. Also they
did another clever thing. They stopped calling
00:30:51
the inertance... describing the units of
the inertance as kilograms. I went to a meeting
00:30:58
and suddenly found that the x-axis was Zogs rather
than kilograms. And so that was again a decoy,
00:31:06
in case a team member left McLaren and went
to another team and described this device,
00:31:13
knowing that one of the axes is kilograms
is a giveaway. It's a clue to what's
00:31:18
actually going on. So those two things
together, they succeeded really in keeping
00:31:26
their approach quiet for a couple of years. But
then what happened is exactly what you described
00:31:33
that an employee from McLaren moved to another
team and it leaked some information. Right.
00:31:41
That's correct. So McLaren found out about
this because they hired someone from that team.
00:31:48
It was Renault in fact and the person coming back
from Renault was able to say, well we have one of
00:31:54
your drawings of the J-damper and it's been
circulated in the company and examined and so on.
00:32:01
So McLaren at that stage knew that the device
had leaked out to another team. I should say
00:32:09
this happens all the time in Formula 1 because the
market for engineers is quite fluid.
00:32:19
People get hired from one team to another all the
time and you can take what's in your head. That's
00:32:25
allowed. You're not supposed to take drawings...
but that does happen as well. But the
00:32:33
story is that even if they had a drawing, actually
this was not enough, right? Yes. Well, there's
00:32:42
a very amusing story about how this eventually
came into the public domain and it was
00:32:50
through what is now called the for 2007 Formula 1
espionage controversy. And it started with a case
00:33:02
being brought against McLaren. It concerned
two of their... or one of their employees who had
00:33:12
drawings of the Ferrari. That became
known and Ferrari brought a case against McLaren
00:33:26
and they were found guilty of breaching the
sporting code at a a world council hearing
00:33:34
even though the actual impact on the design of
the McLaren it wasn't shown to be a significant
00:33:43
impact nevertheless they were found guilty
and they received the largest fine in sporting
00:33:50
history. It's still the largest fine. It was
a hundred million dollars that McLaren was fined.
00:33:58
And this point, and in fact, as part of the
defense, in anticipation of this, they were
00:34:04
expecting a large fine. McLaren tried to say,
"Well, actually, there's a lot of this going on,
00:34:10
and we know... there's a case that we can
prove." And so, McLaren took their case,
00:34:22
to the FIA, or as soon as McLaren blew the
whistle, Renault owned up and said that they
00:34:32
had various things. There were other things apart
from the J-damper, but the J-damper was one of
00:34:38
them. And there was another hearing. Renault
was found guilty of breaching the sporting code,
00:34:48
but they were given no fine, which is quite
quite a lot less than $100 million.
00:34:57
And in regard to the J-damper there
was again an interesting substory. Renault had
00:35:07
used to their advantage a device on their car
for a number of years called the mass damper, the
00:35:12
tuned mass damper. That's essentially a mass on
a spring. And it's a classical idea in mechanical
00:35:18
vibrations. But they managed to find a way to get
an improvement. Now, it turned out that
00:35:28
Renault's reaction to the drawing of the J-damper was to contact the FIA to try to get the
00:35:36
J-damper banned under the interpretation that
it was a mass damper, which was banned at that time?
00:35:42
which was banned. Renault had success with
the mass damper and it was banned sometime...
00:35:48
I forget the exact year but by that stage it
was banned. Okay. So Renault wanted to say, well,
00:35:55
this thing should be banned because this is a mass
damper. And McLaren were able to counter that it's
00:36:01
not a mass damper. It's something completely
different. And the mass damper is a mass
00:36:06
on a spring. So it has a one point connection to
the car. So it's not a two-terminal device
00:36:16
as the inert is, so that you connect between two
movable points in the suspension. So McLaren were
00:36:22
able to argue that the inerter shouldn't be banned
because it didn't come under the description of
00:36:32
the device that was relevant to the mass damper.
That was used at the World Council hearing and in
00:36:39
fact in the transcript of the hearing, the justification for the no fine being
00:36:46
applied to Renault is that "they had fundamental
misunderstandings of the nature of the device and
00:36:53
therefore it couldn't affect the championship". It
nevertheless as far as the inerter is concerned
00:36:58
it's an amusing story to tell. So then at this
point still no one knows what the J-damper
00:37:05
is, right? So when was the connection found between
the J-damper and the Inerter? Yes, there was a lot of
00:37:11
pit lane gossip, because information leaks out
and people leave one team and join another and
00:37:21
eventually people start to figure out what it is
in the pit lane and so on.
00:37:31
Still though it wasn't widely known and it
was an Autosport journalist Craig Scarborough
00:37:41
who published the scoop in Autosport in... I think
it was 2008, describing the connection between
00:37:51
the J-damper, the mysterious J-damper, and the
inerter and at that point the cat was out of the bag.
00:37:58
So everyone then in the Formula 1 world knew that
McLaren's device was the inerter and it connected
00:38:07
with the technical literature. Okay. And so since
then the inerter has been used in F1? The inerter very
00:38:14
quickly was used on most if not all cars within
Formula 1 and became a standard component and that
00:38:22
continued that way until 2022, until the inert
was banned! Oh, that's must be very disappointing
00:38:30
for you. Yes. In Formula 1 circles, it's
almost inevitable because this happens so often.
00:38:39
So, we had a long run with the inert up
to that point. But the reasoning in this
00:38:48
case didn't seem really as strong. It was
argued on the grounds of cost cutting. How much
00:38:55
is an inerter? Well, Penske Racing Shocks which
was licensed by the university to supply
00:39:04
them, and they're of the order of some thousands of
dollars, $5,000, perhaps $4,000, that kind of
00:39:13
range. That's like a rounding error in the budget...
One would think so, yes. But nevertheless,
00:39:20
the inerter became a target when they were
looking to reduce the cost of the sport.
00:39:26
The cost of the device is only one
aspect of it. The design and simulation
00:39:34
of the suspensions and the testing of course has
to be added to that as well. So if you go back
00:39:40
to a spring and damper suspension,
you can argue it's less expensive. Though in fact
00:39:48
the cars started to have stability problems
when the inerter was removed. In fact, the ban was
00:39:56
delayed by one year so they could dial out these
difficulties. So some of the stability problems
00:40:04
that we had with the the active... they commonly
occur with these types of racing cars which are
00:40:13
very very stiffly sprung and have ground effect
aerodynamics. Do you think that is possible that
00:40:19
in future it will be brought back or you think that's
not a possibility? It's hard to tell. Someone
00:40:26
within the sport would have to advocate that.
It would be nice if it did come back. It gives
00:40:33
more... from an engineering point of view it gives
more to explore for the engineering designer and
00:40:41
Formula 1 cars are supposed to be advanced
vehicles from an engineering design point of view.
00:40:46
Okay. So you described what the damper does and
it's quite clear, right? It dissipates energy.
00:40:52
Yes. So maybe can you tell us what actually
an inerter does? Yes. So that's always an
00:40:59
interesting question because you're faced with
the electrical engineers way of thinking about
00:41:05
things that you have a box of components and
you shape your impedance... and you don't ask
00:41:11
"this capacity here. What is it doing?" Whereas
the mechanical engineers they like to think "this
00:41:16
spring is supporting a load". "This damper is
reducing energy". What does the inerter do? Well,
00:41:21
"it's producing a force proportional to relative
acceleration". That doesn't always satisfy them.
00:41:27
So you can think analogously and get insights in
various ways, which is not necessarily the complete
00:41:35
story. But the insights are always useful. So if
we take the inerter again, one interesting
00:41:43
little experiment you can do with an
inert is to try to impose an oscillation and one
00:41:52
thing that is very striking is that the maximum
force occurs at the maximum displacement. That's
00:42:01
the same as if we had a spring, but the device
is pulling rather than pushing. So the electrical
00:42:10
engineers would say that the inerter is 180° out
of phase with the spring just as the capacitor
00:42:16
is 180° out of phase with the inductor. So,
some people like to think that way, that if you're
00:42:25
in a state of oscillation, and you have a
an inerter in parallel with a spring, then the
00:42:33
sum of those two forces in a way they partially
cancel out. And at resonance they would, they
00:42:40
would cancel out exactly. But the other thing
to note is how the force varies with frequency.
00:42:49
As frequency increases, the force in the
inerter increases much more rapidly than the damper
00:42:57
or the spring, and that's something again that has
to be brought into the picture. It's not just a
00:43:03
phase characteristic. So I would stress that (A)
the inerter is an energy storage device and (B) it's
00:43:13
the dual of the spring in a mechanical sense, just
as the capacitor is the dual of the inductor and
00:43:22
(C) that it is the third component that allows you
to produce this complex impedance of an arbitrary
00:43:34
shape. So you might need, for a very
complex frequency varying impedance, you might
00:43:44
need a large number of components. In the
electrical domain that's not a difficulty, but in the
00:43:50
mechanical domain you want to restrict the number
of components, so you perhaps can't shape the most
00:43:57
complicated impedance functions that you would
like to shape. Nevertheless the analogy I think
00:44:02
is is important to explain the device. The idea of
the inerter seems quite simple, right? It's about
00:44:09
analogies with the circuit theory. And so probably
also looking at the beginning when you started to
00:44:15
work on this, you probably wondered if this
was already in the literature for instance and
00:44:20
maybe later on when you come up with
the device and this was implemented, people
00:44:27
misunderstood how this was working. So in one
way the idea is very simple but in the other way
00:44:34
it escaped so many people. So why do you think
this is the case? Yes, I think that it has partly
00:44:42
to do with electrical engineering thinking versus
mechanical engineering thinking, systems thinking
00:44:51
versus a more reductionist way of thinking about
a mechanical system. The latter where you
00:45:00
want to ask "what does this element do on its own"
whereas the systems way is you're asking what is
00:45:09
happening to the whole system, the closed-loop
control systems terms, or the vehicle together
00:45:16
with the suspension system, and all the components
working together. And it's the richness of the
00:45:24
component set that one thinks about in terms of
widening the available behavior.
00:45:33
We saw that the inerter was banned in 2022. But
maybe there are other uses for the inerter, maybe
00:45:39
in other fields. Can you tell us a bit about
these? Yes. Well, it is still used in
00:45:45
some areas of motorsport, but perhaps the first
applied area for inerters was, apart from vehicle
00:45:54
suspensions, was building suspensions.
And one sees applications there in Japan with ball
00:46:03
screw inerters being applied in building suspensions
to improve the response when you have
00:46:10
earthquake excitations. You can think of the
vibration suppression in a helicopter for example
00:46:16
between the rotor and the fuselage trying to
damp out or even prevent vibrations
00:46:23
being transmitted, or motorcycles. We looked at
weave and wobble instabilities in motorcycles and
00:46:30
found that there could be an advantage of the
inerter in that context. I mean... I guess
00:46:36
that, you know, we had springs and dampers for
thousands of years while inerters just for 20
00:46:43
years. So probably we still are exploring all
the possibilities. I think it's still early
00:46:48
days, and for example we're looking at railway
suspensions and there's a possible advantage
00:46:54
there in terms of reducing track wear. So we
have some projects looking at that. Okay. Thank
00:47:01
you very much for your time. This was a wonderful
story. Pleasure talking to you. Thank you. [Music]
00:47:20
[Music]
