Manufacturing mistakes can be costly to both time and money but a large one can even injure or kill workers. The bets way to combat mistakes is to monitors every step of the process but when it comes to creating complex materials this isn't always so easy. Learn how the sensors used work and why they are so important to the process.
This episode is sponsored by Gefran, to learn more about their products and what they can do for you click the link here.
This episode is sponsored by Materials Today, an Elsevier community dedicated to the creation and sharing of materials science knowledge and experience through their peer-reviewed journals, academic conferences, educational webinars, and more.
Thanks to Kolobyte and Alphabot for letting us use their music in the show!
If you have questions or feedback please send us emails at [email protected] or connect with us on social media: Instagram, Twitter.
Materialism Team: Taylor Sparks (co-creator,co-host), Andrew Falkowski (co-creator,co-host,editing assistance), Jared Duffy (production, marketing, and editing).
I would like to describe a field in which little has been done, but in which an enormous amount can be done.
This field is not quite the same as the others in that it will tell us little of fundamental physics,
but it will tell us much about the strange phenomena that occur just below our perception.
In contrast to the natural philosophers of the past,
the scientists of this field delve into the recesses of nature and show how she works in her hiding places.
Their quest is to understand and create the imperceptible.
After all, there is plenty of room at the bottom.
Hello and welcome to the Materialism Podcast, an exploration of the past, present, and future of material science and engineering.
I'm joined, as always, by my co-host Taylor Sparks.
And today, we're going to be talking a little bit about sensors
and how we actually measure the things that we use to control the production of materials.
We always talk about, you know, temperature and pressure
and various other sorts of things that you can monitor and use to change the outcome of material.
A lot of thermodynamic quantities are going to be based on those.
A lot of kinetic processes are also controlled by temperature and pressure.
But we rarely talk about how you actually go about controlling those,
how you go about controlling them accurately,
as well as what are the consequences of not controlling them.
Yeah, you know, there's been some really high-profile failures
that have had to do with exactly this topic of temperature and pressure
not being monitored carefully and leading to either an explosion or something drastic happening.
In fact, there was just one that I was going to point to in 2017 that happened in St. Louis, Missouri.
This was at the Loy Lang Box Company.
The United States Chemical Safety Board has a series of really awesome explanations
of when disasters like this happen.
In this case, there was four people that actually died here.
And this company makes coriated cardboard.
And as a part of that process, they had this big pressure vessel
where they put superheated water.
And they had oxygen, and they oxygenated this water.
So they're actually sort of, just like you can have carbon dioxide in your soda pop that you drink,
you can do the same thing with oxygen.
You can dissolve oxygen into water.
And it's not maybe surprising that that makes for a really corrosive environment.
So they were putting that oxygenated water into a steel vessel,
and that is going to be corroding.
They're supposed to put oxygen scrubbers into it,
things that can pull that oxygen out.
And there's processes in place for venting that system,
but they just weren't being done correctly.
And so corrosion was happening.
And so this specific company,
there had actually been a number of indicators of leaking over time.
They saw actually some dripping coming through.
There was clearly like some rust forming.
So at one point, they'd actually cut the bottom off this big tank,
which was pretty large, like it's larger than a person.
It'd be sort of like, picture like a water heater in your home.
It was something akin to that size.
They had cut the bottom out of it and replaced the metal pan,
but they didn't replace all the portion that had been corroded,
just part of it.
Anyways, there's instances now where in hindsight,
they've done a really careful study,
and they've shown that clearly there was an uncontrolled corrosion reaction
happening in this pressure vessel.
There was an ineffective repair.
There were gaps in the inspection process,
and the process for actually managing the safety was totally inadequate as well.
All this led to this vessel with this superheated water inside of it.
When the rupture took place, now you have room temperature air
entering this atmosphere where you had superheated water.
So it's going to instantly convert to steam.
Now the pressure is no longer containing it.
In that process, there's a big volume expansion,
and you essentially created a missile.
It led to the death of these four people really tragically.
All this is to point out that just because we're material scientists,
and you think like, what can go wrong if I'm making materials,
and worst thing will happen is I end up with a product that I can't sell,
that's not quite right, right?
Because you can actually be putting yourself in a dangerous way
if you're not monitoring the actual process of making materials itself.
So to that end, we wanted to talk to a company,
Jefran, and you'll meet some of their employees in a moment,
who are making materials that allow you to not only make these measurements
of temperature and pressure, but make corrections in real time.
Because it'd be one thing if you were looking at a measurement on a screen
of temperatures rising, if you were going to look at that,
and then by the time that you started to observe that it's wrong,
and start to make the adjustment, well, how much are you going to know
to, say, reduce the electrical current that's causing the heating?
That's a really tricky thing.
I remember the first time I learned about PID loops.
If you're not familiar with PID, a PID controller,
well, the PID stands for Proportional Integral Derivative Controller,
and essentially, instead of just looking at the absolute value
of the signal you're measuring, whether that's temperature or pressure
or something else, it's actually looking at the derivatives.
It's looking at the rate of change of those things.
So really carefully, if you want to increase your temperature
from 50 degrees to 70 degrees, instead of just turning a knob
and sort of wildly, like you're flying an airplane,
swinging too high and then too low, it's actually really,
in a very accurate way, picking the exact signal that you need
to get the resulting process out of it.
So PID is still, I believe, a pretty common tool that's used
in most controllers. Is that right, Andrew?
Yeah. At the very least for temperature controllers,
most furnaces that are using those small, little, digital,
almost analog-looking controllers are still PID controllers.
Maybe that's changed, but I believe that is still
very much an industry standard method of controlling various processes.
Now when you get to things like safety sensors,
those can usually be a little bit different,
in the sense that they're not necessarily trying to control a process,
as they are just, they have a threshold,
and once that's surpassed, they send an electrical signal.
Yeah, or they can even be mechanical.
I remember one of the early problems I had on a homework
way back in the day when I was taking my materials science course
was a shear pin, right, where under a certain shear stress,
the pin would shear, it would break across and allow the pressure
to degas before it got to the catastrophic one
that would cause the vessel to rupture.
So you're intentionally, and this is why it's so important
to understand materials failure, right,
you're designing it to fail as a safety mechanism,
which I thought was pretty slick.
That said, this is all sort of depending, you know,
you've got these control systems that can tune your temperature
and pressure accordingly so you stay safe,
but that all depends on you being able to measure temperature and pressure.
Andrew, it's not the most straightforward thing in the world
to measure temperature and pressure. Are you familiar with how they work?
I think I have a nice conceptual understanding,
but could I design one myself?
It would take a little bit of work.
Yeah, let's start with temperature measurements.
There's different, obviously, ways to measure temperature,
and some of them are really sophisticated.
Let's just talk about the two most common ones that I'm aware of.
The first one is just going to be your thermocouple.
So these are not like the sort of thermometers that you're thinking of
if you cook at home, right, where you can dip it in the liquid
and you see, hopefully not the mercury,
but some other substance rise in the sort of glass beaker.
This isn't that.
A thermocouple, instead of relying on a fluid expansion
that's related to a temperature change,
and that fluid then goes across a scale
and you can read it like a ruler, as it were,
it does something different.
You're getting an electrical signal out of it.
So how do you get an electrical signal
from something like a thermal process?
Well, the name sort of gives it away.
You're going to use something like a thermoelectric, right?
There's actually devices called thermocouples,
and these are just two different types of metal.
And if you put them together,
in the presence of a temperature difference
across one end to the other,
in a metallurgical junction,
what you will observe is that you will have
a difference in temperature
across the other two ends of the metal.
So this is the exact same thing that happens
inside of a thermoelectric device.
If you listened to our previous episode
way back on thermoelectric materials,
this is how they work.
We harness that voltage that gets produced
to actually extract power from a system,
in the case of a thermoelectric generator,
but it's the exact same principle.
If one end is heated and the other end is not,
so they can use that voltage
to actually determine the temperature.
The only catch is that
knowing from first principles exactly
what the voltage will be
depending on the temperature range given
is not like an easy thing to know.
So instead what they typically do
is have an empirical relationship,
or essentially like a lookup table.
Somebody knew what the temperatures were,
you put your thermocouple in contact with it,
so now you know what the temperature difference is
as your ground truth,
and you observe the voltage
and you do this as you increase temperature,
and then you essentially fit a curve to it
and you use that fitted curve
as a lookup table for determining
what now the new voltage is
from your thermocouple reading.
Obviously there's different types of thermocouples,
maybe some of our listeners have bought these,
there's type J, there's type K, there's type R,
and typically what that means
is that it's telling you
what types of materials come together
to make that metallurgical junction.
Copper and something is iron and something,
that all has to do with what the nomenclature is,
and they're typically used in different temperature ranges.
Yeah, that was always my understanding
was that these define
the temperature range
at which you're going to be operating.
Now the second common way
that we measure temperature
is through infrared radiation emission,
so if you've taken
an electronic properties of solids class
you've probably learned about black body radiation
at some point.
If you've seen a lightbulb, an incandescent lightbulb,
then you have observed black body radiation
as you heat something up,
in this case the lightbulb filament,
as it gets hot enough it just glows.
Now lightbulbs in our modern area
are pretty simple, they just glow bright
at a perfect temperature that has already been set
so the brightness or the whiteness is sort of fixed,
but in the early days,
or if you've ever seen a piece of wire
get heated up slowly,
it doesn't go straight to white,
instead it doesn't start glowing at all
and then it eventually turns to that white color.
The wavelength
that's coming off is changing.
At first you don't see it at all
because our eyes,
we can only see a pretty minuscule section
of the electromagnetic spectrum,
but it's clear that
as you saw it go from an orange
to a red to a white,
that that's related to a temperature change
and you can see it with your own eye.
So the other way that they measure temperature
is with a more sophisticated version of an eye essentially.
It's measuring the wavelength
of energy coming off an object as it's heated up.
This is the infrared emission.
And again,
there's a lookup table that you can use
that will determine the temperature of the object
based off of the light that it's emitting
and that's your infrared measurement of temperature as well.
Oh yeah, and I mean,
once you get up into the temperature range
where thermocouples
can't really operate or you kind of risk
doing damage, then those are the only ways
you can get temperature readings.
We've got a
home-built SPS system, that's spark plasma
sintering. Again, we have a couple episodes.
Check those out if you want to learn more about that.
But in that type of device,
we actually rely on both thermocouples
and an optical measurement.
So we have a thermocouple that
we drilled a little tiny hole inside
of our material, our dye, and we put
the thermocouple in contact with the
innermost portion of it that we dare to get to.
And then we have another hole
on the other side of the device which we point our
infrarometer at, and it's able
to measure the light coming out of that little tiny hole.
And it's interesting, as you center this device,
the dyes actually are compressing
and moving a little bit, and so you actually
have to move the aim of that
optical measurement so that it stays
fixed on the tiny hole, right? Because if the dye
moves, and all of a sudden you're not pointing at the hole,
well the outside of the dye is going to be much
colder than the inside. And so sometimes you'll
see this huge jump in the temperature from
your optical measurement, and it's pretty clear
that what happened is that you stopped looking
at the little tiny hole that you were supposed to be looking at.
So do you have to adjust it mid-process?
Yeah. So normally
during heat-up, it helps if you've got an extra person
on hand that can be watching that and keeping
that dialed in, while the other
person is running the other parts of the
process. Gotcha.
And that's a big problem. Like we talk
about safety, even if you're measuring
it, if you have
some misalignment and all of a sudden the thing that's
controlling your PID controller is reading
a cold spot or something, and it just
starts amping up that current.
It's a big problem. Yeah, it is.
I've definitely had it happen before. We actually melted a sample
like crazy, and we thought that we
were just fine. We pull it out
and it has been completely obliterated,
totally melted it, and we realized, oh shoot,
it must have actually been looking at the felt lining,
which is meant to be a sort of protector.
It's meant to absorb this infrared
light coming off your sample so that your
vessel doesn't get too hot. And
our infrared camera must have been pointing at
that instead of at the hole, and it caused our sample
to get completely obliterated, and we didn't know
about it until afterwards. And that's
why a lot of industrial practices
use redundant sensors
to account
for this, to account for the fact that if there is
a miscalibration, not only
could that cost the company millions of dollars,
but it could cost people their lives.
Oh, totally. And at first I was annoyed with it
because the thermocouple didn't always
match the pyrometer measurement.
They would be off by a little bit, and so
it's just like, well, which one is it?
But in hindsight,
having that second sort of
foolproof, you know, double
check that things are going like you think they are
is really valuable, actually, I think.
So that's temperature measurement.
Now we have to think about pressure measurements.
So how are you going to measure pressure?
So again, just like there's two
main ways of measuring temperature, there's two
main ways of measuring pressure.
The first is with piezoresistive
strain gauges. If you've ever
heard of a Wheatstone bridge, which
maybe some of our listeners have heard of,
these are of that type.
And essentially what's going on here is
Andrew, if I took a wire
and I hooked up
a multimeter to the ends of a wire,
and if I pulled on that wire a little bit,
I strained it, technically the
resistivity of that sample would have changed
a little bit.
Now oftentimes the amount of deformation
needs to be really big
to get a big measurement that you can easily
read, or your wire needs to be
really, really long.
So that you can essentially, the resistance
is large over the length of a long wire.
And so the way that they get around
this with strain gauges is by
making, in pretty miniature scale,
a serpentine pattern
of a resistive material,
such that when this plate or
the sensor itself gets
strained, the overall plate
is going to change dimension, and you've printed
this wire along it in a long serpentine
path. So even though
it's occupying a pretty small footprint,
you have quite a bit of length of your wire
that's all going to be expanding or contracting
depending on what the pressure is doing.
So you can actually get a
measurable difference in the resistance from it
and relate that to
your pressure in just the same way that we do
with temperature, using essentially a lookup table
that you have defined ahead of time.
Another way that they do this, though, instead of
using piezoresistive strain gauges
is there are capacitance
pressure transducers.
We've talked a lot about capacitors on this show
before. I think we will on a future episode
specifically related to piezoelectric
materials, but if you've taken
an Intro to Physics course, you probably learned
about a parallel plate capacitor.
If you haven't, then picture two pieces of metal
and as you
bring them close together but not touching,
and you hook a battery up so
the positive and negative are on either side of those
plates of metal, if they're not
touching, then you can't pass current through
the system. There's no electrical current running
through this loop. But as they get closer
and closer, what you will do is you will accumulate
positive and negative charges
on the two different plates.
As you bring them closer together,
that amount of charge that you accumulate
will increase.
The way that these pressure transducers work is
they take advantage of the fact that as the plates
get closer and closer, you're going
to accumulate more charge along
these electrodes, and you can measure that
charge, that surface charge.
That becomes your signal that you're going to
use is essentially the voltage
between these two plates as they get squeezed
closer together or further apart as the pressure
changes.
I'm sure there's lots of other interesting
configurations that people have come up with
depending on the application
or the cost.
One of the ones I'm familiar with is actually
a guy in our department, Jeff Bates,
he and I were working on some water sensors
and we actually came up with a hydrogel,
which is a polymer that swells in the presence
of water in a fairly predictable way.
We took a steel mesh
on one side and put the pressure
so you start out with dirt, maybe you're trying
to measure the water content of dirt.
You take your dirt, you take a metal
mesh, you take your hydrogel
and put it on the other side, and then you put the whole thing up
against your pressure transducer such that when
the hydrogel swells,
it can't push into the steel
mesh any further, so it's going to have to push the other direction
into the pressure transducer.
You can actually get a water
hydration measurement
by using this in conjunction with a
pressure transducer.
That was a nice little introduction to sensors, but
now maybe we should go talk to some of the experts
in the field. To talk more about
how we can prevent these sorts of
accidents in the future, we need to
learn about the types of sensors that can be giving us
the information that engineers could
take action on, and to do that,
this episode is sponsored by Jefren,
and from Jefren we have Joe Vasco and
David Newell. Do you want to introduce yourselves, guys?
Sure. Joe, I'll go first.
My name is David Newell,
a plastics
industry participant,
for lack of a better way to put it, for about 40
years now.
Background is extrusion, injection
molding, compounding
of many kinds, including
rubber and
wet coatings
as well. Primarily
interested in
quality control and
safety mechanisms
throughout the career, and
it's primarily been focused on
assisting customers
in batch-to-batch
sustainability across
the board so that they can remain viable
suppliers to their customers.
Okay. And you, Joe?
My name is Joe Vasco. I'm an
application engineer with Jefren.
I've been with the company for 15 years, and
for 20 years prior, I worked for a machine
manufacturer. My job
is to work with customers to find solutions
to their problems and
help them make a better
produced product and work with
them to the best of their abilities
to solve problems.
Okay. So if our listeners have not heard
of Jefren before, clearly we've mentioned that you
make sensors, things like temperature sensors
and pressure sensors, but you do more
than that. You have systems actually for
taking information from those sensors and actually
providing feedback loops to
maybe you want to reduce the temperature, or you want to reduce
some pressure, or you want to increase those things.
So you've got control systems that you
offer, you have the systems themselves,
and you have the software that helps can run it.
Are there other things that you want to tell us about the company?
Joe, I'll defer to you.
You've been with the company longer, have a deeper understanding.
Jefren is a company
that's actually been around for close to
40 years. It was started by
two brothers who found a
need for
better control for the
injection molding industry. So on
the temperature control side of that was the
initial starting point.
From that point, Jefren grew
instead of just being a simple company
that made temperature controllers,
they grew to add pressure controls,
different types of sensors,
and they even added a whole separate
division for motion control.
So over time,
their task has been to
solve customer problems and
work with that and grow the business
together as a partner.
I was going to say, there's a lot of companies out there that make
sensors. What makes Jefren special?
David, do you want to address that?
Sure, I'll field that one.
And what I'm going to say, I think
Joe will agree,
pertains to all of our product lines
which are
different industries serving. But on the
sensors side, particularly on the
plastics melt sensor side,
Jefren
is primarily concerned with
producing a state-of-the-art product
from the outset. They don't
come in and try
and be a copycat
in any application whatsoever.
Their goal from
their inception that
I have learned is to be a
premier supplier from
the get-go. And in
that regard, they have succeeded.
They have an excellent name in
the
global marketplace in all their applications
and are probably considered
one of the top three
global suppliers in any of those
applications as a result of that successful
effort. So in material
science, we often interest ourselves
in the so-called tetrahedron
in material science, which is that this discipline
is defined by understanding relationships
and those relationships are between
the structure of material, the processing of
material, the properties of material, and its
application, which is going to be related to its
performance. So in
what parts of those
in what parts of that tetrahedron
do we see the need for sensors?
I'll field this
one. When it
first and foremost, and this is
strictly a personal observation after
40 years in the business, it's
all about safety first.
Many
of these products
inception came about
post-industrial
revolution, midway through the industrial
revolution in the 1900s
when there was a
high increasing demand for
rubber, plastics,
chemically put together products,
compound products of many kinds
and suppliers
would reverse engineer solutions as
to ways to bring these products to market
but in that process
safety was not a large
concern and while none of them invited
industrial accidents,
using power equipment
at high speed, high temperature
resulting in high friction and
high pressure, oftentimes
produced safety
concerns or worse
with being industrial accidents.
So initially, many
of these products were created to help
control the process, monitor the process
particularly at high levels
of attributes I talked about
before, pressure, temperature,
that kind of thing
because once you get up into that genre
you are inviting an
industrial accident of many kinds
so they would need to determine what was a
safety point, what was a danger point,
have the proper sensing equipment
to monitor it and keep that in place
from happening. So that's really
number one and throughout
the 1950s and 60s, post
World War II and that very strong
industrial environment,
these products were
developed. After that
they were learned
to be quality control
tools and you were able to
control the quality of
your batched, mixed,
compounded product
via the use of pressure sensing,
temperature sensing
and in very simply put,
if you under pressurize, under temperature
mix or compound
or to use a generic term, cook
a compound, it's going to come out
under mixed and it's going to
under perform. It's not going to mold properly
if it's coating, it's not going to dry properly
if it's a rubber, it's not going
to perform properly and conversely
if you over mix any of it,
it's not going to perform properly, it's going to
be too brittle, too hard,
it won't mold at all, it'll crack and break quickly
so it's very, there's a critical
low and a high
area of all of those
root applications
that is important
to learn and to know and to maintain
within to be able to put out
batch to batch consistent product
time after time.
A lot of industries that these sensors
are being used in have specific standards
for their products. I know the aerospace
industry has a number of standards that they
expect their materials to be at but also
things like the pharmaceutical industry as well.
Is that right? Correct.
Some of our controls meet the
different standards so we do have
controls specific to that as well
as the sensors that match that.
As you alluded to, the aerospace
industries uses the ASM2750
standard, the automotive
uses the CQI9
and then the FDA or pharmaceutical
uses
section 21 CFR
part 11. So all those different
industries have different specific
requirements and they
basically come down to the quality of
the control and the precision of the
control. And the Jeffran controls
and sensors are able to meet those requirements.
And how does that affect
maybe the sensor design or
how it's integrated into the
end application?
It's a great question.
Each sensor is
application specific. For instance,
we're manufacturing sensors not only
into plastics compounding
and or rubber compounding.
We're also manufacturing into
motion control applications or
linear position applications. Entirely
different, all three entirely
different industrial applications.
So the sensor
must be designed to
what has reverse
engineered been determined
to be the specific need
for the application. So a great
deal of scientific
approach goes into the research and development effort.
First, understanding what the need of the
customer is and then
determining what your sensor is going
to have to do to successfully
meet this need. And those
are the two criteria you meet and those are the
two goals that you keep
focusing on as you get through R&D
and get to final product.
So, you know, I've never worked with
sensors, not really anyways. And so I sort of just
took for granted that they would be available,
that they would work, they would survive the
environments in which they're going to be used.
But that's not always the case, right?
In chatting and learning about your company, we've learned
that there's actually quite a bit of customization
that you offer for customers
that have unique environments. Can you
describe some of the things that you've seen along the way?
Yes.
Many of the, we can
take polymer compounding, for instance.
Some customers
will be working with a
generic polymer that
yes, per an ASTM
standard, will have the same operating
temperature prescribed and
have the same operating pressure
prescribed. But
oftentimes, and really more often than
not, those customers will,
in order to develop a niche in the marketplace,
want to develop
a specialty for their
compound. And in doing so, they will incorporate
different additives
into their compound. And now
you are coming out of
anywhere where an ASTM standard exists
other than the base polymer that you're working
with. And through chemical
engineering, you need to
determine what those additives
you're adding in are going to do in terms
of increasing temperature,
pressure, creating
friction, mixing time, etc.
And you're going to need to develop new standards
that can only be done through
a thought process from a
CE side, a chemical engineering
side, but also through
trial and error. And once
you do that, you can
begin to have the right additive
products to go with it. What are some of them?
On a low side, you could be
introducing some pigments that
are slightly abrasive, such
as any whitening compound made of
a TiO2, which is actually a very small
sphere but very hard and can
be abrasive as it passes
by various metal parts, all
the way up to very
serious abrasive and even caustic
compounds where they have
incorporated them into the polymer
because it's what's required to
mold their finished product, but they
can be quite abrasive
and diminishing
on the parts inside of the
mixing equipment and
the barrel and the sensors. And
therefore, you must incorporate things
like specialized steel alloys to the tip
of the sensors and other parts in order
to stand up to the material that's now asked
to be mixed. Do you have any
specific case studies of where
you've had to adjust your
where you've had to adjust
your sensors or maybe
the design of them or the application
to meet some of these extreme conditions?
Sure. A
recent example of a
very extreme one
is a global electronics
manufacturer that
everyone would know of
needing, they were producing
their own packaging
foam and
their ability to be, their continuing
need to be self-integrated. They were
producing packaging foam for their finished electronics
to go into
to be shipped in cartons
in. And that foam
was a challenge
to be formulated in terms
of the way it needed to mold
and the way they needed it to dry
and in essence, through
their production line process
they needed it to finish
curing after
they had loaded the electronic
component into it. So in other words, they would
put the electronic component into it
and it would take the shape
of the outside electronic component
and finish curing right to it
so that it was a very secure package to ship.
Well, that foam
was a caustic
mix of
base polymer and wet chemicals
and one of those wet chemicals
being an acid-based product
and as most people know
any acid-based product that comes in contact
with any metal whatsoever
is going to
begin to rot that metal very
quickly because the two just don't mix.
They're like oil and water.
So we were able to
produce
an alloy tip
a Hastelloy C270 tip
something very common in the marketplace
but we were able to make our tips
out of that instead of the standard 15-5
stainless that would be degraded
quickly and the customer had
been using our sensors with the 15-5
stainless so they weren't lasting very long
because the caustic
solution was quite high in
acid content and when we introduced
the Hastelloy C270
roughly 8 months ago
9 months ago, they've yet to have
a problem with it since and it seems to be
working fine. So that's a good success story
and that's a typical success story
and the only way we would have known that
is to ask the customer
what's in your compound without
giving away any proprietary
information. What are you mixing
in? What's happening? Let's try to figure out
what's degrading that stainless steel
and once they started to talk about it in generic
terms, we knew pretty quickly
we need to go with a compounded
alloy tip that's going to
stand up to what they're doing because we can't go back
and ask them to change their compound
they've already found a successful
way to do what they want to do
so at that point it's our job to
adapt to their needs.
Case study
wise in a different application
similar we've had customers
that we'll call it part of the pressure
control and part of the safety
control. We have melt pressure systems
which are a combination of
using the sensor that's controlling
the process at a very
high speed for the extruder drive
that's driving a motor. So we had a customer
that was doing compounding so
they're taking base materials and partially
recycled materials
which can be somewhat abrasive, putting it through
an extruder and
which is then melting it, then coming out
of a die to make polymer tips
the problem they were having was
an inconsistent flow so
that would cause jamming further down the line
where the, if you want to think of it
like spaghetti noodles
coming out of the end of a machine, there's a
knife coming around cutting it off
and consistently
so it doesn't get gummed up on the
tip, the flow needs to be
controlled. So we used one
of our temperature controllers and
high speed controllers
to solve the problem. It actually did
two things. It controlled
the motor drive which was in this case
a big 200 horsepower drive
and within 5 milliseconds
it could change or adjust to pressure
changes. So if there was any problem with the
feed, the material
coming through, it would immediately adjust
to it and provide a more
consistent compound and more consistent
product. So in this
case it was a pelletizing line where a customer
was putting in
material that was inconsistent
and it was also part of its
dye color to match it
and we were able to solve the solution by providing
both a melt pressure sensor
and then this controller,
our 2500 unit
that controlled the high speed
process directly and was
able to solve the problem for the customer.
Okay, so clearly we
need to measure things like pressure
and temperature if we're going to understand process
control and materials, but we haven't
yet said how actually those work and
maybe some people are curious. Do you want to describe to us
how a thermocouple actually works
and how is pressure actually being measured in
your actual sensors themselves?
Joe, you want to go with this one?
Do you want to talk about the strain gauge on the pressure
sensors? Sure.
The way the pressure sensors
were designed and
successful design, when I talked
earlier about the
being produced as a safety mechanism
back in the early 1950s, they were
two professors
from MIT and Boston that
were
charged with trying to produce
some kind of a product to
monitor pressure in what those days
was rubber extrusion primarily because
the northeast
was very strong
in that industry and
they were
asked to try and come up with something to
solve this considerable safety
problem. And around 1950 and
1951, they came up with the first ones
and
through a design that still
exists today, there
is something called a strain gauge
that's made
of various materials, but it's
a soft cotton material with
some wiring and
it's embedded behind
at the very tip of the sensor
with a very thin 15-5 stainless
over the top of it and the strain gauge
swells and contracts and expands
and contracts with pressure and temperature
and that
reading is
sensed, that sensing is sensed
through the 15-5 stainless tip
which also kind of waffles,
expands and contracts a little bit
and then as that
reads that
sensing, as it senses and goes by
and does its
as the tip
swells and contracts, the strain gauge directly
behind it reads that
change in
feeling, if you will, as pressure
and then through
a series of algorithms
that information is
sent up using
a filler material,
usually mercury, but now these days
you can do it with oil and you can do it with sodium
potassium as well, and
it's sent up in a temperature
range and as
that temperature range through the mercury fill
a very small amount, reaches the
electronics at the top, it is
sensed and converted through a series of algorithms
to pressure
and we also are able to, in this
industry, embed a thermocouple
into the sensing
tip down near the strain gauge
and just below the surface and
a similar activity takes place
where the temperature is sensed
and sent up the
embedded thermocouple as well
and both readings can be taken at the surface.
And Joe, if you can improve
on that, please do.
Along the line of sensors,
thermocouple is basically
two dissimilar metals
that create a millivolt signal
when they're just joined mechanically.
The most common
industries, sorry
about that noise in the background, let's try it.
Thermocouples are another
type of sensor that are basically two
dissimilar metals.
The two dissimilar metals, when they're joined
mechanically, create a small millivolt signal.
So one of the most common thermocouple
types is J-type thermocouple
which is simply made up of iron
and constant or basically
a copper nickel alloy base.
When they're mechanically
joined or soldered
at the tip, they create a small
millivolt signal. And as the
temperature changes, that millivolt signal
changes. Now that signal is not
a linear signal. So
based on temperature, there
is a specific value
that corresponds to a temperature.
So temperature controllers then have
to adjust for the type of material
or the thermocouple type
which is one type of sensors
that's used and know
what that means as far as temperature.
Similarly, there's also
RTDs or resistance thermocouples
which are linear in their action
that also make a small
millivolt signal and
adjust for temperature.
The simple
process that we've
had thermocouples around for years,
for example, in
gas piloted operated furnaces
that many people have in their homes,
there's a mechanical thermocouple
in there that creates a millivolt signal that
keeps the gas valve basically operating
as long as it sees the right temperature
in the flame sensor.
So the controls take that
base signal and then
if we will, we say we look at it
and we improve the control
based on the
background information.
Our listeners will remember we did a previous episode
on thermoelectrics where we talked about
the Zebic effect where you get this voltage
produced by having junctions of
dissimilar metals.
It's interesting that you point out that these are non-linear
and so it's not just as simple as get one
or two data points. They really have to
empirically solve this, don't they?
You have to actually know very carefully what the actual
temperature is and I'm not totally sure how they're
doing that if it's with
measurements or something else
but you basically calibrate
your voltage response to what the
temperature difference is and then you just fit
it to some sort of curve empirically
going forward as your sort of ground
truth for that.
And then obviously that would depend on
different thermocouple junctions are made of different
materials so they're going to have their own different
empirical data that you're going to have to fit it to to
actually get a measurement out of it.
You're exactly correct on that.
What happens in the background of the
controller, there's basically
a lookup table.
For a J-type thermocouple like I alluded
to, there are specific values
that correspond to a specific temperature
and part of the control's
process, you have to be within a
certain point on the accuracy.
So we're talking very small
signal differences based on
a temperature difference that corresponds
to a chart
that's all built into the firmware
of the control. So all these things
are handled in the background and are
basically
almost commonplace and automatic
but there is a big strong background
of what's happened in the back
to make that work.
Talking about accuracy of the measurements
and the ability for there to be some
drift, how does Jefferin handle
I guess sensor redundancy?
Is that something that is
important or is that something that customers ask
for? As in having
multiple thermocouples to ensure that you're
getting, you know, one of them isn't deficient
for instance, or multiple pressure
sensors. Is that something that's very common or
not so much? Have sensors gotten good
enough that those sorts of issues
aren't as much of a problem these days?
Joe, I'll run with this one
if you'd like.
It's a great question and
the more
I would say the more premier
product the customer is making
the more apt you
are to find redundant
sensing and temperature sensing at
the same point in a
process. For instance
I was at a customer
earlier this year
that's manufacturing
product for a top tier
automotive supplier
with a
very
serious quality control
acceptance procedure
whose rejection
of which you do not want to be a part of
because it's very
strict and after two or three times you could lose
the account. And this
particular company
had five
sensors, both pressure and temperature
at the same point on the extrusion line
but not exactly the same spot
so if you could look straight at the barrel
and visualize it as a hollow
pipe or a clock, they had put
them at one, four
seven,
ten and noon or something like that
and they were getting five
temperature readings, five pressure readings
from exactly the same model
sensor coming in at the same
time to their control
and that was how they were sure
that they had
a decent uniform
homogenized mix into
360 degree circumference within
the barrel because they had realized over the
year, over the years
that taking it at one location
you can pick your spot, let's say it was at
8pm on a clock, if they had a
sensor installed there where they were taking pressure
and temperature readings, they
had figured out that on the other side
of exactly the same space
where melt flow was or in another
spot, when you're looking at it from a
360 degree standpoint, you
might not have the same homogenous mix
and you might not therefore get the same readings
so they realized over time
that they needed to install them on all sides
so that they knew they would get in the same
pressure, same temperature,
360 degree circumference, therefore
they could have with relative surety
know what was coming through the screen
into a finished product form.
Sensor wise, the other thing that's
done when there's a need for
precision or for quality, for
example in the FDA, requires
both the sensor and the control to
basically be certified every 6
months. So as I mentioned, our
controls look at a signal
and know what that corresponds to temperature.
So this control has to be
brought back to an NTIS
standard and make sure that it's
within its accuracy every 6 months
so basically an engineer
comes in, hooks up a special piece
of equipment to the control,
types in several test points
and confirms that the control is working.
So along with doing this
calibration, you know
that the control is working in the background
it's monitoring the data, the data
is being captured by a SCADA system
so customers are able to simply
see when something is not
correct, when it starts going out of
tolerance, before it starts making a
bad product, they can actually see a sensor
failing with our control.
Is it always clear when it's a sensor
failing versus when your product is failing
like there's something wrong with the system?
Because let's say you're doing a polymer processing
and so you're relying on monomers
that come from somebody and you're
trying to control what they ship to you
but that's a little bit out of your control, right?
And so what if they send you monomers that have changed
a little bit and so your process has now changed
leading to a difference in temperature pressure than you're used to
seeing. How do you know it's due to the monomer
and not due to your sensor failing?
That's a very good question.
The sensor failing is more
predictable in terms of
we'll consider it like noise.
Different materials
may shift
the process
point high or low.
When the sensor actually starts
going bad, you see random spikes
if you will, where a signal
is out of the range and then comes
back. So based on the
signal and the data that's being captured
you can determine whether it's a
sensor failure that's happening or starting
to happen or going to happen in the near future
or if it's just something
the process isn't
operating as would be expected because
of a difference in the base materials.
It does take a little bit of
work by a process engineer
who's familiar with the control system to understand
the difference but it is distinguishable
between the types of failure
or let's say quality of
what's happening in the process.
Yeah. If a
customer came to you and
it sounds like you use mostly for example thermocouples
for temperature measurements but if they wanted
an optical measurement using
infrared or something, are you
capable of adjusting your system to
rely on a different input like that?
Correct. Our temperature
controllers or any of our process
controllers all accept a wide range
of signals. So we have
thermocouples which are the non-linear.
We have PT100 which is a
linear type sensor but then
we also can work with
as you mentioned the infrared type sensors
which typically have a 4 to 20 milliamp
signal and that corresponds
to a chart as well.
So we can work with any of those
processes. For example
similar slightly out of this material
area but doing the
curing of solar cells
you're looking at a solar silicon
wafer and you're
controlling the temperature in the oven on that
point and you are using an
infrared thermocouple. You're not actually touching
the product but you're looking at the surface
of the product to see what the temperature
is. I feel like sensors
have been around for a while and I think they're going to play
an increasingly important role
especially as safety
is certainly a concern for industry
but data and data acquisition
and data analysis has also become
just an incredibly important
part of materials processing and it's going
to continue to be that. What do you
see as the future of sensors
in industry and how do you think they're
going to evolve?
We're actually kind of seeing part of that evolution
now. As you mentioned
everything is about the capturing of data
for quality control and
for verification of process
and for making sure the process
is working. So
since we're kind of seeing it in a
two-fold process. On our
melt pressure sensors we are actually
adding what's referred to
as one of the newer systems, IO-Link.
So that
brings along the data
if you will of actual the pressure
along with that pressure
data. It also captures in the
background of the sensor high peaks
how many times it's gone above
a certain threshold and
other material information
about the sensor, extremes
it's seen and it's captured that
and recorded in the background of the sensor
so you know how many hours the sensor's been
in use. You know it's peaks
and this becomes data
that is available to the PLC
or control system on an
as needed request. It's not data that
comes over every minute but it's
data that comes along. So
we're actually seeing the shift of
adding that type of system
to our pressure sensors.
Similarly on a control side
we see more and more requests
for data and how can we get
the data from the controller as fast as possible.
So we now have interfaces
to a lot of our controllers that work
with the different PLC systems whether
it's the Rockwell
system, you know using
the Ethernet IP, whether
it's PROFINET or
any of the other field busses that are out there
so those
are a need or a requirement that
are being brought upon or requested
by customers. So that's driving
the design and control
of both the sensors and the controls.
Gotcha. And David what's your perspective
on the future of sensors?
I would add to all of the
great technical side of
things that Joe is very proficient in.
From my side
batch to batch consistency
is key with
all suppliers today. For years
it was not.
They ignored
making scrap
for whatever reasons and
right around the market
crash in 2007 and then now
certainly post COVID
people have realized that they cannot afford
to produce nor are they in business to
produce scrap product
that either must be reworked or
scrapped altogether. They're in
business to produce batch
to batch quality product for the customer
that is how they remain a
viable profitable business
so they don't want their
extruders going down, they don't want their
injection molding machines going down
they don't want
any of their
industrial equipment with hydraulic
or linear position sensors in them going
down. They need them in service
doing what they were built to do
on whatever basis
they do it, 8 hours a day,
16 hours a day, 24 hours a day
and that is very
important today as supply
chain and so many other things have
put a crimp on costs.
Customers have realized that
their job is to remain up and running
at all times and producing the same
product for the customer at all times and
sensors and controllers
and indicators of all kinds
must be state of the
art and performing
in the way that they need to and the
suppliers of them need to be
savvy enough
to be able to ask the right questions
when a customer says we're having a problem
but we don't know what it is and they need to
be able to ask the right questions so that they can get involved
with them, reverse engineer
themselves with them to
the origin of the problem
and therefore craft a solution.
I think that's it
though. Anything else you guys wanted to
say that you didn't get a chance to?
Not me, I'm good. I think we're good.
I appreciate the effort and the work
of putting this together. Yeah, me too.
Okay, thanks guys.
This episode of the Materialism Podcast is
sponsored by Jefferin Incorporated
located in North Andover, Massachusetts
with over 5 decades
of experience and nearly 700 employees
worldwide, Jefferin designs,
engineers and manufactures sensors,
controls and components for industrial
equipment. OEMs, engineers
and end users rely on Jefferin
products for process control, automation
and equipment electrification
in applications related to plastics, metals,
glass, packaging, fluid
power, heat treatment and more.
In addition to standard product lines,
Jefferin's vertically integrated supply
chain allows the company to customize
products for its OEM customers
and 14% of the workforce
is dedicated to research and development.
You can visit the company website
at jefferin.com. Jefferin is
spelled G-E-F-R-A-N.
The Materialism
Podcast is sponsored by Materials
Today. You can visit materialstoday.com
to stay up to date on the latest happenings
in the materials science field and read
some of their fantastic articles they've published.
You can also head over to elsevier.com
to find out more about their journals, books,
conferences and related programs.
Thank you for listening to this
episode of the Materialism Podcast.
If you've got any questions or feedback,
send us an email. Reach out to us. We're easy to get a hold of.
You can find us at materialism.podcast
at gmail.com or
we're actually much more engaged on Instagram
and Twitter so our Instagram handle is
at materialism.podcast and you
can connect with us there. We also post lots
of fun pictures and additional stuff
about the show, some behind the scenes things
so check us out there.
It would really, really help us if you would
leave us a five star review on iTunes,
Spotify, Google Play, wherever you
listen to your podcast. Leave us a positive review
and it will help us expose
the show to new people. That would be really great.
Finally, we want to give a big shout out to
Alphabot and Colobite. They're the ones who make the
really cool music that we use in this podcast so
thanks to them. We think they make good stuff and we
think that you should support them. You can find their stuff on Spotify
and YouTube. So with that, we will
see you guys next time.
Red. Glass. Silk. Cotton.
The makers of tools.
The captors of lightning.
The architect. The engineer. The musician.
Are all beneficiaries
of the materials of this world
and are bound only by
their imaginations in manipulating those
materials.