TRANSCRIPT:
0:13
Good afternoon. Thank you for joining our webinar today. Our topic is going to be thermal considerations for medical devices. My name is Devon Pellicone. I’m the lead engineer at advanced cooling technologies (ACT) in our product development group.
0:27
I have more than 10 years of experience designing different thermal solutions for a wide range of applications.
0:34
At the end of this webinar, if you need more information, you can always visit our website at 1-act.com. We have a number of resources available there where you can find brochures on our products or reach out to a salesperson for more specific inquiries.
0:50
A quick note on the webinar format, you can enter your chat into the dialog box, shown on the right there. So, just click on the chat icon, type your chat, your question, into the chat box there, and we’ll receive that e-mail and respond to you with an answer to your question.
1:05
Today, we’re going to go through a number of things.
1:07
The overall theme of this webinar is to give you an idea of what different cooling solutions are available for medical device applications and others.
1:16
So we’ll talk a little bit about the difference between spot cooling and remote cooling, that’s cooling locally at your heat source or moving that heat to another location and a little bit about the difference between transient and steady-state applications.
1:31
First, a little bit about ACT. We are a thermal management solution provider. We’re founded in 2003 with 200+ employees and now we have over 120,000 square feet of manufacturing and office space.
1:44 Our core values are innovation, teamwork, and customer care, and we really do carry those core values through everything that we do.
1:53
We have a number of awards listed here.Most recently, we were awarded one of the top 10, fastest-growing companies in Central Pennsylvania.
2:04
We went to work with our customers, really, from the beginning all the way to the end, so we’d like to start with inquiries from our customers. Some challenges that you may be experiencing, work through those concepts with you, come up with a solution, and then ultimately a product that can be manufactured in medium to high volume.
2:22
We work in a wide range of different industries, power electronics, medical, energy, recovery. We also have a military and aerospace division of our company. And we work with some of the largest names in, in these industries.
2:36
So enough about ACT, we’ll talk specifically about medical devices now.
2:42
Medical devices cover a wide range of different products and applications. There’s no one specific solution for a medical device. The systems come in very small sizes, like surgical devices, to very large machines like MRI machines.
2:56
The packaging challenges are really what’s unique about medical devices. Sometimes there’s not enough space immediately around where the heat-generating component is to be able to dissipate the energy that’s required. So, you need to move that heat to a location where cooling is available.
3:11
That’s what we talk about when we say spot cooling versus remote cooling.
3:15
Temperature stability challenges are unique in medical devices, oftentimes precise temperature control is required.
3:22
For surgical devices, very precise temperature for cutting tissue is required and not to exceed certain values, the damaged tissue, there’s touch-safe temperatures that we need to maintain to make sure we don’t damage skin.
3:36
And then I start with the reality requirements for things like diagnostic equipment and making sure that all the samples within a PCR machine are within the same temperature. So, they cycle at the same rate.
3:48
Lots of different stability challenges and unique temperature requirements for medical devices that we’ll talk about ways to address.
3:56
The operating temperature requirements can range from really below ambient, meaning it needs refrigeration to be cooled or above ambient cooling, which is often available for things like computer chips and CPUs, things like that.
4:11
The last thing we’ll talk about is a little bit about what we call duty cycle challenges.
4:15
So, not all operation of electronics is steady state. Sometimes, the operation of the electronics is on and off with a relatively long off period in turn, on cycles.
4:26
If you try to cool the steady state max peak power, your cooling solution may be way oversized for what the maybe load shapes power of your application is. So we’ll talk about a solution that’s available for load shaping, any duty cycle operation for post applications in ways you can address those transient thermal challenges in your application.
4:49
So, first, we’ll talk about Spot cooling, spot cooing refers to the direct cooling of a heat-generating component within the immediate proximity. So you’re seeing on the right here.
4:59
A number of examples you may be familiar with, these are mostly CPU coolers, some thermoelectric based devices That components that are coming out today tend to be high enough hour, where there’s not enough space, immediately available around the heat-generating component to get enough heat sink volume where you can dissipate the energy with inappropriate temperature.
5:22
A lot of these applications we’re showing on the right here at relatively low power, where a heat sink fits within the footprint of the heat-generating component itself.
5:31
Then we can dissipate that energy by mounting a fan to it or ducting air directly to where that heat-generating component is.
5:39
If this type of situation is required, sometimes we can spread the heat. So, we have a computer chip that may be 10 mm by 10 mm. We can spread that to 50 mm by 50 to reduce the load on the heat sink.
5:52
So, you may not have more than 50 available then we need to move to something like a remote cooling application where we move that heat from your chip to where you do have the space available to dissipate.
6:05
And then the last note for spot cooling, there is a sub-ambient option where you can use a thermoelectric device. They are solid-state passive cooling devices where you can get local cooling below your ambient temperature on your chip directly.
6:20
They’re mostly suitable for applications where you have tens of watts. They’re not great for high power stations.
6:27
So first, we’ll talk about heat pipes, which is one way of utilizing all of the space you have in your immediate area. So if you’re not familiar, heat pipes are passive, two phase devices. We’re showing a schematic on the top right here.
6:41
They consist of an outer envelope, which is usually copper, and there are hollow tube on the inside there’s a wick structure and a working fluid.
6:49
These systems are hermetically sealed so they operate in a vacuum.
6:53
And at one end of the heat pipe, you input your heat that boils the fluid out of the structure. The vapor travels down the hollo intersection of the two, where it condenses wherever it’s called on the heat pipe, and then that condensed, Now, liquid, travels back to the heat generating portion of the heat pipe through the extraction, the capillary action.
7:14
This happens passively, and it will happen forever, as long as the temperature difference between the input and the rejection side of the heat pipe is maintained.
7:23
Heat pipes are great for situations where you need to spread that heat away from the immediate cross-section of your regenerating component.
7:31
So this allows us to get more surface area above the heat-generating component to dissipate higher powers.
7:38
Typically heat pipes will see anywhere between 2 to 5 degrees C temperature difference from one end to the other so they are very efficient conductors.
7:50
A vapor chamber is another way of spreading heat. So you can think of a heat pipe as a one dimensional heat spreader of vapor chamber would be a two-dimensional case.
7:59
So, it’s the same principle, where we’re boiling and condensing liquid inside of a form to geometry, with a weak structure inside of it, but in this case we can spread in two directions. So it’s a planar heat pipe. These are really great for higher heat flux applications.
8:15
We’ve demonstrated over 700 watts per centimeter squared. And they’re also really good for applications where you need very precise, very uniform temperature all the way across your equipment.
8:26
one application and medical devices is for a PCR read.
8:29
So in a PCR machine, you have a number of individual thermo wells inside of there with your DNA material in it and you want to cycle them at a very precise temperature profile. And you want every one of your samples to be at that same temperature all the way throughout the cycle.
8:45
And a vapor chamber is a really great way of maintaining up to 386 of those individual wells at the same temperature within two-tenths of a degree, all the way throughout that cycle.
8:57
So that’s a really great application for vapor chambers and medical devices. Custom geometry is very custom geometries are possible. We can really form of vapor chamber to meet the needs of your specific application.
9:12
Somewhere in between heat pipes and vapor chambers or what we call high K plates, this is basically a heat pipe embedded material. Typically, it’s aluminum.
9:22
And basically, we can embed heat pipes cleverly inside of a plate of material such that we can spread or move the heat from one location to another very efficiently. So, if we take a base material like aluminum, its thermal conductivity may be 170 watts per meter kelvin. Once we embed heat pipes in it, we can get that effective thermal conductivity, upwards of one thousand watts per meter kelvin.
9:45
What you’re seeing in the top right, there are just an example of one high K plane, where we took the max component temperature that was on that play down from 91 degrees, all the way down to 69 degrees C in this application. So they can be very effective in spreading hotspots that may be in your system, spreading that heat out to a larger area where it can be dissipated or again, moving it from point A to point B very effectively so that you can dissipate it more easily.
10:14
This is a more cost-effective version of a vapor chambers. They are not as great in terms of their isothermalization but they’re significantly cheaper than vapor chambers when heat spreading is required.
10:26
When we’re done, they do wind up still having the structural strength of aluminum and the weight of aluminum if that’s important for your application.
10:37
So here’s a quick summary of the Spot Cooling opportunities: Cooling technology. So, heat pipes, we talked about as being sort of a one D heat spreading typical temperature difference from one end of a heat pipe to the other is about two degrees C, these are good for CPU, cooling, LED cooling, maybe some surgical devices.
10:56
If you have a pair of scissors, surgical, scissors, you can take the heat out of that work or tip the surgical device and move it a little further down the handle to dissipate the energy.
11:07
So heat pipes are really good for moving from point A to Point B vapor chamber is the same thing. But in two directions, like we talked about, and they have much higher heat flux capability, so, power per unit area, capacity, than a heat pipe wood and they can be very customized to whatever geometry is required. So, PCR machines, laser modules, and diodes really benefit the most from a vapor chamber type of application.
11:33
Then, like I said, a HiK™ plate is somewhere in between those two, in terms of its use and cost, really.
11:39
Again, you get two D sort of pseudo two D heat spreading. It really conducts the heat along the path of the heat pipe which can be bent into weird geometries inside of your plate.
11:49
So you get this sort of two D heat spreading with the high gateway.
11:54
So when we cross into vapor chamber and again, it’s custom-designed so that we can dodge mounting holes inside of your application by embedding heat pipes in the plate. And this is good for all of the above plus maybe MRI machines and diagnostic equipment.
12:07
In medical device applications, then thermoelectric cores is really the best opportunity for spot cooling and getting sub-ambient cooling.
12:16
Again, these are only really good. For maybe less than 50 watts, they tend to be inefficient.
12:21
So, they require a significant amount of energy input in order to do the cooling of your local component.
12:26
But if you do need, let’s say, a laser module to be maintained at a temperature below the ambient air temperature in your room, you can use the thermoelectric module to get that local temperature down so that you can do spot pooling of your laser module is probably a good example for these types of technologies.
12:46
So, now, we’ll move on to what I’m calling remote cooling. Remote cooling is really one of the more common forms of thermal management or removing heat from the immediate area, to some distance away, it could be inches. It could be meters.
12:58
It could be tens of meters away from your heat-generating source.
13:05
There’s a number of technologies we use. We use heat pipes again, for transporting heat from one location to another. Then, we have other technologies that are more liquid-based going solutions with thermosyphon standard, liquid cooling, in that, what we call pump to phase.
13:20
And each of these technologies has an appropriate range of applications, and widgets, most suitable, so we’ll go through those now.
13:29
So, back to heat pipes. Now, we’re using heat pipes, instead of spreading heat.
13:33
Along a one-dimensional plane, we’re going to move the heat from point A to point B, So, you’re seeing a couple of examples on the right there where we have a heat-generating component, maybe several inches away from where we could potentially dissipate that energy.
13:46
And so, we’re using heat pipes in very complex geometries to take the heat out of the local heat generating component and move it over to offend stack that’s significantly bigger than what was available at the heat-generating source, where we can dissipate that energy much more effectively.
14:02
And these can operate in against gravity if designed properly so they can operate in any orientation, depending on what your packaging restrictions require.
14:11
The heat flux is here tends to be a little bit better than 40 watts per centimeter squared of heat input into the heat pipe itself.
14:19
If you have a hierarchy flux and that, then we’ll talk about our other solutions in a minute.
14:24
Typically, they’re air cooled.
14:25
So, the heat sink, The ultimate sink I’m calling here is the air in your room typically provided, via fans. We can either mount the fans directly to the heat sink like the image on the top left there or we can duct air from another portion of the system to the heat sink, so that we can dissipate the energy without having the fans taking up volume in and around your heat generating.
14:49
The envelope materials for heat pipes are most commonly copper they can be aluminum and they can be stainless steel.
14:56
The working fluid inside of the heat pipe is most commonly water. We also use methanol for lower temperature operations, sometimes refrigerants.
15:08
The next step up from a heat pipe in terms of being able to handle more power and possibly moving further distances is what we call loop thermos, right?
15:15
So, very similar to a heat pipe. A loop thermosyphon. Operates by boiling and condensing a working fluid, but in this case, we’re relying on gravity to return the working fluid to the heat-generating section of the device.
15:29
And so we have a liquid column, and in the picture we’re showing on the right here, we have a liquid column that is providing our gravity head or our pumping power.
15:37
That pushes the liquid into where the heat-generating components would be mounted. The heat-generating components boil the working fluid creating bubbles and voids inside of the evaporator, which creates an effectively less dense column of fluid than the liquid column.
15:53
So we get this net positive pressure gradient that drives fluid around the room, and if designed properly, we can push that two phase mixture tens of meters away from where the heat generating component may be.
16:08
Now that requires a certain amount of gravity head in order to do it, and so the restriction on these is that the condenser the portion where we’re rejecting the heat be physically above where we’re putting the heat in.
16:19
So there is a gravity orientation required for being able to use with …. If your application can support that, you can dissipate tens of kilowatts passively with no pump.
16:31
I’m using these devices with a very effective thermal resistance.
16:35
Another advantage to these systems, because we tend to use refrigerants is that they can be dielectric.
16:41
We can put dielectric lines on them, and we can use a dielectric fluid.
16:45
So if for whatever reason you need your evaporator or heat-generating components to float at some voltage potential and have the condenser grounded to the local ground of your system, we can do that and standoff tens of kilovolts. The voltage between the two.
17:02
That’s pretty useful for applications like power electronics, where you need to float the voltage of your modules for whatever reason and ground the rest of the system.
17:13
These systems can support sub-ambient cooling if you put a liquid-cooled condenser.
17:18
On the top of it then you can run chilled water through it and be able to get your heat-generating components down closer to the temperature of whatever chilled water you have available.
17:27
That does require the addition of a chiller in your system or some source of chilled fluid, and we’re able to do that. But it does provide you the option of doing something.
17:37
We can make these things in a wide variety of geometries. In the bottom right, you’re seeing a loop thermosyhon for a medical diagnostic piece of equipment, where it’s kind of doughnut-shaped.
17:45
So, the heat was actually local to the inside of the donut. So it created a radiative heat load onto this system, and then we dissipate that energy, move the heat up, and away from the evaporator to the condenser like you see it there.
18:00
Passively, again, no moving parts, high reliability, and lots of heat transport.
18:06
The envelope materials tend to be aluminum and copper standard cold plate materials, and the working fluids are most commonly refrigerants like R 134 A or the local whoa global warming potential equivalents of those types of infections.
18:22
The next step up from Luke Thermosyphon, or if you can’t maintain that gravity advantage would be a standard liquid cold plate or liquid cooling solution.
18:32
This can do tens of kilowatts, maybe hundreds of kilowatts if you have enough flow.
18:36
There’s a lot of flexibility in how you install liquid cold plates. They can be custom designed to really, any geometry that you have. Cooling channels can either be formed in the way of tubes like we’re showing on the screen here, so a certain team of tubing on the backside of your heat-generating components, or we can make a custom vacuum Brazed assembly where we duct the coolant through the cold plate. And through fins in a very custom way to dodge any restrictions you have in your packaging.
19:05
So, you can get very long transport lines if they’re pumpkin supported, it does require support equipment here.
19:11
So the previous two examples heat pipes and with thermosyphon, they’re sort of contained self-contained in and around where the heat generated components are.
19:20
This system requires what we call balance of plant combines pumps, accumulators, heat exchangers, kind of like we’re showing in the bottom right here, a point distribution unit that supplies the fluid to the cold plate, and then receives the hot fluid and removes the energy from it.
19:38
So there needs to be some space in and around your components where we can put the rest of the balance of plant components to support the liquid coolant.
19:47
Again, this can perform sub-ambient cooling. If you have a chiller available or chilled water available, we can do liquid, liquid heat exchange and be able to reject high amounts of energy below the local temperature in your room.
20:02
Materials, copper, aluminum stainless, most common, other materials are possible. and working towards tend to be glycol water mixtures. So that they suppress the freezing point.
20:13
Ethylene, glycol, propylene glycol tends to be better for most medical applications and then deionized water or really stay lean are also options for maybe surgical devices where you don’t want the risk of a toxic fluid entering the bottle.
20:31
The last remote cooling technology we’ll talk about here is our most advanced cooling technology to phase it’s basically a pumped version of the thermosyphon.
20:41
So, again, we’re relying on the boiling and condensing of the fluid to remove high amounts of energy, but instead of relying on gravity to supply the fluid here, we’re relying on a pump, and so we can do tens to hundreds of kilowatts.
20:54
This is our most capable thermal solution that’s available.
20:58
And the benefit of using pump to phase is that you get a very effective heat transfer coefficient at a very uniform temperature. So the boiling process happens along a line of constant temperature. And so if you have multiple components mounted on one evaporator, we can maintain all of them within a degree C of each other, using a pump two-phase solution. We can also maintain multiple parallel evaporators at the same temperature. This is good for applications, like laser modules or power electronics things, where you want them to all be at the same temperature to perform properly, or a precise temperature.
21:35
Again, because it’s a saturated system, we can control the temperature very accurately by controlling the pressure in this.
21:42
And so, it can respond very quickly to changes in pressure that we initiate with a control system, and so, we can do very precise temperature control possibly for laser applications or high power laser applications.
21:55
Again, we use refrigerants most commonly for these systems. So, they can be dielectric. We can use diametric lines and so we can again, again, standoff tens of kilovolts of voltage between multiple components in the system. if necessary.
22:09
And we already talked about the dependence on high efficiency, but that also translates into a high heat flux capability.
22:16
For very high power density components, this solution offers the most heat flux capability that’s currently available for thermal solutions.
22:27
It does still require the balance of plant components a pump reservoir condenser, which needs to be located somewhere. It can be outside of your immediate equipment.
22:37
But it needs to be in the relative proximity of your systems.
22:42
And again we can support subbed ambient cooling with the addition of a chiller or chilled water that may be available in your facility.
22:49
Jamba materials copper and aluminum again.
22:52
And then, we tend to use refrigerants for these systems. They have the best properties that we’re looking for most electronics cooling applications.
23:02
So, a summary of those technologies, heat pipes, very similar to spot cooling. But in this case, we’re using them to transport relatively short distances away, I’d say, less than a meter.
23:13
Even a meter may be able to aggressive, for a pipe, depending on the orientation, they can operate against gravity, but you do take a power hit. So, you can’t do as much power against gravity as you can when you have gravity to your favor with the pipes.
23:27
So that’s something to keep in mind and they’re really best suited for relatively low power CPU.
23:32
Laser applications, MRI machines, get a lot of new set of heat pipes and some diagnostic equipment or you’re trying to move the heat out of the immediate proximity of, um, where are the heat-generating components?
23:46
Whoop, dermis Siphons can transport a little bit longer, well further distances than heat pipes can maybe tens of meters less than 10 meters.
23:55
They must be gravity aided though so that distance needs to be vertical.
23:59
So if you’re going to be transporting energy from your local heat drying component, a meter away, it needs to be above, physically above the heat-generating component.
24:09
They can do watts of power, tens of kilowatts.
24:12
We’ve demonstrated systems up to 100 kilowatts, with the proper amount of heat, available.
24:18
And again, they have dielectric fluids available to provide some dielectric strength that’s required for your system.
24:25
So it’s good, really good for power electronics if you have those in your system laser modules, higher-powered, laser modules or diodes. Some MRI pieces of equipment can benefit from this and diagnostic equipment.
24:38
Like, we, again, we’re moving down the list in terms of how much distance away from the region and components we can move, Liquid cooling gives you some of the highest distances you can transport tens of meters away. If necessary, it’s got a higher power capability than dermis savings do. It eliminates that constraint of the need to be gravity aided.
25:00
And it does require some balance of plant components, Thomas accumulators, things like that, It’s good for all the same things that are with thermosyphons, but much higher power capability and much further distances.
25:11
Then Pumped two-phase is then a step further above liquid cooling, in terms of its power capability in terms of iso-formality, and then it adds the ability to use dielectric fluids and a little more easily than standard.
25:25
And then all the last three with thermosyphon liquid cooling and pumped to face are all capable of sub-ambient cooling if required but you do need the addition of a chiller or some source of chilled water.
25:41
The last thing we’ll talk about here today is the difference between a transient and a steady-state solution.
25:47
What I mean by that is if you have a situation where your heat-generating components have a duty cycle to them, so you have a high heat input, or high heat rejection requirement over a relatively short period of time. And then you’re off or idle for a period of time. After that, we can take advantage of that duty cycle and damp it out with a thermal battery or thermal capacity.
26:11
So, we typically call these phase change materials and that instead of changing phase from liquid to vapor like we’ve been talking about, this is most commonly a phase change from solid to liquid.
26:22
and in that melting and freezing process is a high amount of energy storage capability in the latent heat of fusion of the material.
26:30
And again, it happens at a constant temperature, which is a real benefit in terms of trying to maintain components at their proper temperature.
26:38
So what we’re showing on the bottom right here, is sort of A canonical profile temperature profile of a system and utilizing a phase change material.
26:48
So at the beginning, we start ramping up in temperature. So, this is be a solid material to start with.
26:54
As you input power, you raise the temperature of the solid up close to its melt point.
27:00
Once it gets to the now point, it operates at a constant temperature while that entire material melts. And so you basically think of a wax, you have a solid wax, and as you melt that wax, it maintains a constant temperature until all of the solid has become liquid.
27:16
Once that happens, then you start superheating your liquid from that point on.
27:20
So you can maintain your heat-sharing components at a temperature below this melt temperature, all the way up until you utilize all of your phase change material.
27:30
Which is a real benefit for applications, where you just need to store energy during a high period of pulse load. And then you have a period of time where you can reject that energy, which is longer than the amount of time that you spend putting that energy in.
27:46
So there’s a number of ways that we do this. There’s a lot of benefits to using a phase change material.
27:52
There’s a lot of materials available, typically, facing materials, really, you can get them in almost every one degree C anchorman from cryogenic temperatures, all the way up to molten salt type of temperatures. So what we’re showing on the right here is sort of a high-level comparison of different temperature operating types of phase change materials.
28:12
Paraffin wax is, are the most common, because they have temperatures that are in most electronics, cooling temperature ranges.
28:21
If paraffin wax is not suitable for your application being a medical device, there are vegetable-based and sugar alcohol-based non-paraffin, organic types of phase change materials that have similar melt temperatures.
28:35
The challenge with those for medical device applications is they tend to be very expensive being pure substances. But they are very effective. And they have a great weight and eat it.
28:47
The envelope materials that we often use are copper aluminum, stainless steel. And we already talked about the phase change materials that are available.
28:56
So, if you have an application where you need to be able to store energy for a period of time, just the survivor of a level of High Pulse Load. And then, you have a number of minutes of off time where you can re freeze that PCM and basically, recharge the thermal battery for the next pulse.
29:12
This is a great solution for being able to sort of load shape your heat input. And this will require a much smaller cooling system, then you would need otherwise to try and handle the peaking.
29:28
So what we want you to take away from this, really is there is no one size fits all cooling solution for medical devices, They come in all shapes, sizes, and power levels. Really, what you need to do is consider packaging considerations, do you need to cool the heat-generating component locally right around its immediate proximity?
29:46
Or can you afford to move that heater location, where it may be more conducive, dissipating more energy?
29:53
The power requirements really determine where you are in that list of technologies that we’ve shown you today then if you have higher power, that typically requires a more exotic or a higher capable cooling solution.
30:07
If you have a transient thermal load, you may want to consider a transient thermal solution to that problem.
30:15
If you are able to do that, then you can use a phase change material in order to dampen out that load so you can have a much slower cooling system.
30:24
Then consider whether you need above ambient or sub-ambient cooling for your application.
30:29
That’ll help you decide whether you need something like a thermoelectric module for your sparkling application, that you need to add a chiller to your remote, which the last node is to consider thermal management solutions early on in your development cycle. It’s a lot harder to retrofit a cooling solution into a finished package than it is to design it upfront. And so, if you’re able to think about thermal management earlier on in your design cycle, you get a more efficient cooling stage.
30:57
I want to thank you all for joining our webinar today. And, again, if you have any questions, put them into the chat box or reach out to us at 717-295-6061.
31:02
Thank you.