In this podcast, Tom Rohlfs talks with John Maher about multi-layered bonded manifolds. He explains how they can route fluids and gasses through very complex channel patterns, and he talks about the advantages of this option over drilled manifolds.
John Maher: Hi, I’m John Maher. I’m here today with Tom Rohlfs, President and Principal Engineer at Controlled Fluidics, a plastics machining company specializing in precision manifolds. Our topic today is multi-layered bonded manifolds. Welcome, Tom.
Tom Rohlfs: Alright, John. Thanks for having me.
John: Tom, what are multi-layered bonded manifolds?
Tom: Multi-layered manifolds came into existence from expectations in the life sciences industry. They required more complex channel routing in a plastic block. Drilled manifolds are possible to be quite complex. However, oftentimes it requires plugging in order to produce a complete circuit, and the plugging operation would allow entrapment points, possible contamination from one run to the next.
The bonded manifold concept was one, allowing a very complicated internal fluidic channel pattern without those nooks and crannies that a drilled manifold might have. The solution to this problem was to take a manifold and cut it into two pieces, in effect.
You start with two layers. Each layer has a channel pattern machined into it, mirrors of each other, then those two pieces are then bonded together to create one complete manifold. This is a method that was first started with acrylic. Acrylic is very easy to fuse together, bond together, and it’s grown from there to polycarbonate, Ultem, and other materials.
John: You mentioned the life sciences industry. What other industries do you use multi-layered manifolds?
Tom: Pretty much every industry can utilize a manifold. Anyone who needs to control either liquid or gas pneumatic type flows. We see applications in the food industry, aerospace, a little military use, space applications, as well as medical devices, IVD. Any place anyone needs to control a fluid, there’s opportunity for a manifold.
John: Talk a little bit more about how multi-layered manifolds are made and how you cut those channels, then fuse the two parts together.
Tom: In a CNC machining process, typically, machine tools work in a 3-axis manner. We can simply drill straight holes, we can cut straight lines, things along that way. From a manifold perspective, a drill manifold is limited. Every internal channel has to be straight. When you want something to be squiggly or you want it to do round corners, you need to approach the process of fabricating that manifold in a different way.
A bonded manifold takes the traditional drill manifold and slices it, as if it’s a loaf of bread, that right down the middle. We can then machine all sorts of intricate channels between the two and fuse them together, so you can create configurations and features that wouldn’t be possible for straight linear drill.
The method to do it is, as I mentioned, you take two halves of the manifold. You machine both halves, add the channels, then you put them together. The bonding process is a combination of heat, time and temperature. You put it in an oven, you put pressure on that oven for a matter of time and a certain temperature, and those layers will fuse together to become one.
John: That’s the advantage of plastic. It would be, if you were using some other material like metal, fusing those two pieces together would be harder. But with plastic, like you said, you can heat them up, apply some pressure, then the plastic will bond to the other side.
Tom: The amorphous plastics are very bondable and they’re the ones that we mostly focus on. There are some plastics like metals that are difficult to bond. For instance, Teflon, you can’t make a bonded manifold out of Teflon, because it’s difficult to adhere together.
John: Talk a little bit about tolerances and if these channels have to, obviously, you have two sides and you’re machining one side separately from the other side, then they have to match up exactly in order to create these channels. How close can you make that, so that they actually create these round or different shape channels inside the plastic?
Tom: Getting a little bit into channel shape, there’s three options for channel shapes. You can have single-sided channels and double-sided channels. On the single-sided channel front, I’ll explain a little bit more about what that means, you can use either a square end mill, which will cut a square channel or a ball end mill, which will cut more something that we call D-shaped channels or a U-shaped channel.
Those are the simplest manifolds where we machine the channels just in one side, then we’ll put a flat cover plate over the top. That’s two styles of channels. From that perspective, there’s no alignment concerns, because the cover plate is simply flat and it goes across the top of the channel, so you’ve got it perfectly aligned.
John: You’re only machining on one side, then it’s just a flat piece on the other side.
Tom: That’s correct. With a full round channel, you’re right, there’s some alignment concerns there, so you do half a circle in one plate, half a circle on the other plate, then you blend them together. That fusion process, we’re accurate within about 2000. Fairly accurate on the size.
The nice thing about our bonding process is that, it’s a very low pressure process. Some difficulties in bond manifolds is, as you can imagine, you’re applying quite a bit of heat to the manifold itself, and that manifold gets hot, it’s plastic, it’s soft, it tends to move.
If you have a well controlled process, you control not only the size of the feature when you’re done, but you’re also controlling the translation of it. The channels tend to bow out, the channels tend to turn into things that might look like an hourglass from a square.
With good control, which we have, you’re able to control your channel shape after bonding. Our accuracy post-bonding is within a few thousandths, both on alignment, size consistency and position for that matter as well, within the manifold itself.
John: I imagine for different industries, they might have different tolerances for that, so you’d just have to… You’ll know going into it, what the tolerance is for how close those have to be.
Tom: Usually that’s a customer defined parameter. Generally speaking, the nice thing about manifold work is that, as long as the channel starts where they want it to and ends where they want it to, in-between doesn’t have to be super precise.
These aren’t devices that need to be normally held within plus or minus 1,000ths, which is very close in machining perspective, certainly in plastic. A plus or minus 5,000ths tolerance of manifolds is common and relatively easy for us to produce.
John: What are some of the benefits of multi-layered bonded manifolds versus other types of manifold manufacturing processes?
Tom: I’m going back to the drilled manifold side, because that’s where the advantage is. The nice thing about a bonded manifold, it allows a customer to include a lot more components on one device, on one assembly. What’s interesting is, when I first started this, we’ve been bonding manifolds for a long time now, I always expected everything to move towards a microfluidic trip.
Microfluidics is great, because they use very little reagent. Reagents are expensive. What’s interesting is, there’s certainly the world of microfluidics and we can talk about that later. But what I’ve found is, it’s gone the other way, manifolds are increasingly getting larger. What I’m seeing is, customers once had five, 10 disparate components. They had a pump here, they had a valve bank here, they had pressure sensors there, and they connected them altogether with tubing.
What a bonded manifold does, it takes all those components, puts it together on one block. If you have to field service that block, you can take it out of your device and put it back in. It’s very simple, it’s very clean. It cuts down on leak points, it allows troubleshooting, it allows good visualization of what’s going on in the process, if that’s needed.
It has a nice advantage by taking all these discrete components and replacing it with one monolith, with all the things in place. What has happened when we first started this years ago, most manifolds were four by six inches, but what I’m finding is, they’re slowly but surely expanding, because if I can make one manifold with these number of discrete parts on it, why don’t I take all my other components and put it on it too?
We’re seeing that devices, especially when they go to revisions two and three, people are trying to reduce those tubing and they just want to put it on one block. These manifolds are getting bigger and bigger as everybody keeps adding components to them.
It has a lot of advantages. When you look at the total cost of ownership, I bought a manifold because of its processing. It’s an expensive type of endeavor. However, when you look at it from a total cost, from field servicing, from ability to assemble it, troubleshoot, it really is quite advantageous in relationship to having a lot of discrete components connected together with tubing.
It solves a lot of issues. It makes the devices smaller, more compact, and everyone’s headed that direction.
John: You’re right.
Tom: Everyone wants a smaller device, a more compact device.
John: Even though the manifolds themselves are getting bigger in order to be useful for different types of processes, it overall, makes a product smaller, because you’re taking, instead of having, like you said, multiple tubes, you’re just putting it all on one manifold, so that just keeps it more compact and also lessens the failure points, I would imagine.
Every place where you have a tube, you have a place where it’s connected at one end, it’s connected at another end, the tube itself could split or break. You’re fixing a lot of those issues and failure points, so that if something does happen, like you said, you can just pop in another manifold and take care of the issue. But overall, you’re going to have less failures, because of using a bond to manifold.
Tom: It eliminates the leak points from component to component, from pump to valve, from valve to sensor. All those channels are now pushed inside a plastic block. Whereas, I’ve been to customers and they’ve opened up their device and it’s a massive spaghetti tubing. You can see the components, but it’s a massive spaghetti tubing, and oftentimes, it’s like, “Oh, where do I start with a bonded manifold?” It’s all right there in front of you.
As I said, you can even swap out the entire… A manifold is, all this tubing combined into one hard block that you can swap out as just one unit. It works really well for those customers who are trying to clean up the insides of their device, reduce the size of it, improve the packaging, improve the fitness of it. As you mentioned, you’re eliminating all those junctions and possible lead points over time.
John: What’s the biggest bonded manifold that you’ve made and how big are the channels?
Tom: Bonded manifolds, we can get up to 12 by 18 inches in size. That would be the largest we’ve made. Typical channel size, very common is between 20,000, half a millimeter up to a millimeter and a half, 60,000. That’s the meso-scale common range. We’re capable of microfluidic bonding down to 100 by 100 micron we’ve done, which is 4,000 by 4,000.
With microfluidics, because of the size of the channel, you tend to just do single-sided channels, then a cover plate over the top. It’s very common in configuration for them. In the bigger manifolds, we’ll go to double-sided channels. As we talked about, usually round configuration. Round configuration has a real advantage, and I didn’t touch on this before, so I should probably jump back to it.
Round channel skip are much more appropriate for liquid type applications. Reason being is, they have a full sweat volume. Every nook and cranny, we can essentially create a smooth channel that runs end-to-end without any square entrapment points, so that’s a real advantage. The flow profile, it’s got the lowest flow resistance, it’s got a better flow profile than a square channel or a D-shape channel might be.
There’s a real advantage on the liquid side to having full round channels. In fact, we recommend it. We consider it aesthetically the most pleasing. It’s also easiest to bond. It’s two arches that you’re pushing together versus just a flat plate over the top of a single-sided channel.
Oftentimes, my customers might say, “Oh, but double-sided channel, that costs me a lot more.” Not functionally. Generally speaking, we machine every side of our place when we’re bonding, precision has to be quite good, so you have great bonding opportunity.
It doesn’t really take us a lot of time to machine channels. It’s either we’re machining double deep on a single side or two halves on either side, so the machine time is only slightly more expensive than would be, single-sided. It really is relatively insignificant from a cost perspective, and we feel the advantages are substantial.
John: That’s really great information, Tom. Thanks again for speaking with me today.
Tom: You’re welcome.John: For more information, you can visit the website at controlledfluidics.com or call 603-673-4323.