
This article is sponsored by Atalys. In this Voices interview, Medical Design & Outsourcing spoke with Dr. David Crispino, Vice President of R&D at Atalys, to discuss design considerations when it comes to optimizing injection mold cooling. The advanced engineering group at Atalys addresses problems associated with mold design, build, and debug as well as tool validation – particularly when it comes to thermal management, simulation, and troubleshooting.
Editor’s note: This interview has been edited for length and clarity.
MassDevice: Walk us through the critical steps of injection molding.
Dr. David Crispino: Let’s define the start of the process when the mold is closed and ready for the fill phase. We inject the molten polymer into the mold to fill the cavity and create the geometry. That’s usually a quick portion of the overall cycle time. For the kind and size of parts that we do, you’re talking two seconds or less.
Then we go into a pack and hold phase which starts at about 95% of the cavity being full. During this phase, you go into a hold pressure to offset the volumetric change of the material as it goes from a molten state to a solid state. The density of the material goes up while the mass remains the same, so you’re going to have voids. The pack phase compensates for this and adds more material while you’re cooling the polymer. Once the gate freezes off, or if you’re in a hot runner tool and you close the gate, you cannot influence the fill anymore.
Then you go into the cure time, where the remainder of the polymer solidifies to where the part’s rigid and robust enough to survive mold open and ejection. Typically, you either have ejector pins or sleeves push the part out, or you have a robotic arm take it out. Then you close the mold and start again.
Why does the cure time account for such a large portion of the cycle time?
Cure time makes up the bulk of any cycle time. Cure time is when you’re solidifying the polymer. To improve your cure time, a lot can be done in the design of the tool and thermal system in that tool to cool the polymer.
There are aspects of the cycle time that could be machine dependent. How fast can you plasticate the material? How fast you can open and close the mold? Do you have side actions that require the mold to be opened gingerly? But the cure time mostly comes down to your cooling design. Did you do it efficiently?
The thicker the nominal wall of the part, the longer it’s going to take to set that part up and make it rigid. So, as a rule of thumb, the amount of time it takes for the center of the part to reach a certain temperature varies by the thickness of the part squared. That is, if you double the part thickness, it’s going to take four times longer to cure that center.
There are also certain material properties that affect cure time such as conductivity, density and specific heat. They combine to produce a constant called thermal diffusivity — how fast energy can move through the system. This value tells you how rapidly a material’s temperature will change when you change the temperature of its boundaries. Plastics have very low thermal diffusivity because they’re such great insulators.
Looking at thermal conductivity in SI units (W/m/K) for comparison, metals would have a conductivity anywhere from 10 to maybe up to 50. Plastics, on the other hand, would have a thermal conductivity of 0.2 to perhaps 0.35. If you consider heat capacity in joules per kilogram per degree Kelvin, on the other hand, metals would be at 400 to 500 while plastics would be 1,800 to maybe 3,500. So, plastics have more heat capacity but are terrible at moving it. Consequently, plastics are great insulators and don’t want to give their heat up and, therefore, take a long time to cure.
How do the material properties of medical plastics influence cooling?
Medical plastics may have some designations but these don’t affect the thermal properties of the material in a way that we’d design the thermal system differently. There are some high-performance polymers like polyether ketone, however, that have hot mold conditions when you’re molding them. Suppliers will recommend mold temperatures from 275 degrees all the way to almost 400 for these types of resins. That takes special considerations like how am I going to maintain the heat in that tool? All heat transfer is driven by a temperature difference and the greater that temperature difference, the more heat wants to move. When I start increasing mold temperatures to that range, the heat might want to go everywhere but into the tool. So, there’s some math modeling that we do to guarantee that we can bring those temperatures up to the recommended value without leaking it to the rest of the mold.
What are the foundational principles to design an efficient cooling system?
You need to quantify what you’re going after. Too many times people design a tool and say, “I’ll put some holes here, I’ll flow some fluid through there, and that’s it.” I can’t go forward with a terrible design. It’s going to equate to a very long cycle time or deficiencies in the part from a cosmetic, dimensional or functional aspect. You must quantify it. We quantify things two ways: either measure something or create a math model and simulate it.
We use a cure time calculator that we developed. You can find similar ones online. Cure time calculations are based on the one-dimensional heat equation: the transient heat equation for a planar wall. When we look at a part, we find its heaviest wall section and that becomes the thickness of our model. We find the material properties and then define our two fixed boundary conditions, which are representative of the mold temperature. The initial condition is the melt temperature we injected the plastic at. We can also easily figure out the amount of energy that’s left in the part as it exits the mold so we know how much heat needs to come out. There are some approximations here, but it’s a good model to get you started. The bottom line is the calculator gets us a range of what that cure time can be.
If there is some target cure time of, say 10 seconds, that this energy needs to be taken out, we can arrive at a rate at which this energy needs to be removed. But how does the heat get out? It does so through conduction, wherever the plastic is touching steel, which is the entire surface area of the part. The part’s surface area is an easy number to get from a CAD model and if we divide that into the rate of heat transfer, we get the amount of heat we need to remove per unit of time per unit area. We call that number the average heat flux.
Based on that number, I can tell you how big of a challenge you’re going to have designing a thermal system. Working in SI units, watts per meter squared for the heat flux, if you’re at 15,000 and below, you’re going to have a very wide design window for your thermal system to hit your target cure time.
Scale that to 40,000 or 60,000 and you’ll have a challenge. You’re going to have to do more analysis to ensure that you have a system that can handle that kind of heat load. If you get to 80,000 or above, you better start looking at longer cure times.
What mold cooling methods exist?
Traditional cooling lines are a circular diameter channel that goes through your mold. There are standard sizes you can use. There’s no problem with them except that you must be careful about how close you get to your cooling surfaces because there might be structural concerns.
Sometimes you can’t always get these cooling lines where you need them, because you must supply them from somewhere. If they get too sophisticated, there’s a lot of pressure drop. You must look at that to supply the coolant flow rate that you need.
Some of these fill pressures and pack pressures are on the order of 20,000 to 40,000 PSI and above. When you start putting loads like that onto anything, you start to cause significant displacement within the mold. So, you must worry about cyclic stress. Because an injection molding machine is cycling continuously, fatigue can happen.
Then there are baffles. If you have cylindrical cores, or you’ve got standing projections, you want to try to cool those. You may also have ejector pins and things going through them, which limits where you can put cooling lines. Here, you might want to put vertical channels and a baffle in. The only problem is, there is extra tooling difficulty to make them fit and have clearance. Otherwise, you’ll get blowby around the baffle rather than up and over the baffle.
Bubblers have the same principle. They’re a little more difficult to plumb into a tool, but we like them because they’re more controllable and I don’t have to worry about blowby. The only problem is that as you shrink the core size, the bubbler diameter gets small. And if you look at the pressure drop in any type of closed channel, and if you’re in the turbulent range, your resistance to flow scales.
Heat pipes are phase change cooling devices. But the problem here is you can only remove as much energy as you take out. And the way you take the energy out is to cool the base of that pin. Sometimes the length of the base isn’t going to fit into your tool very nicely.
Then there’s conformal cooling, which is very powerful. You’re able to put cooling lines into areas that you normally couldn’t, but with a caveat. Some of these channels can look like the capillary system of the human body, and that’s how small they are. So, how much fluid are you really pushing through them? It’s great to have all these cooling lines, but if I’m only dripping water through them, what am I accomplishing?
I could also do something with a high-performance, high-conductivity alloy, like a beryllium copper alloy, or even pure copper. You can engineer some great cooling elements based on the way you marry traditional tool steels with a high-performance alloy.
What is a steady-state, thermal model and how does it improve cooling?
When you cool a part in a mold, you have a three-dimensional heat transfer problem with material properties that vary with temperature, geometries, and all kinds of boundary conditions. Now your heat transfer problem becomes non-linear. You can’t get a nice solution for anything like that. The differential equations are horrendous to solve.
Now mold filling simulation software can be used to model these. They’re very powerful, but that’s not our first line of defense. What’s more advantageous is a steady-state model. It doesn’t look at the variation with time. You’re only looking at the steady state response of your cavity core, mold and heat flux. Why do we do that? Number one, it’s easy to solve. Number two, you get very meaningful results that you can use during the design phase. Number three, it’s quick. I can do this in an afternoon sitting with a tool designer and design a thermal system.
Those other methods, they’re great to validate a design once you have it. But they don’t tell you how many cooling lines are needed and how they should be spaced. They don’t tell you what the flow rates must be. They don’t tell you the coolant temperatures required.
They’ll analyze the system you’ve designed, but they don’t tell you how to design the system in the first place. These simulation software are very powerful: we have them and use them. But what you need is a tool that the engineers and designers can use to help avoid problems.
What the steady-state thermal models do is characterize the thermal resistance of your entire mold. I’m trying to get the profile temperatures of the surface to be where my cure time calculator needs them to be so that I hit my target cure time. That’s the goal.
It also points out hotspots. We can run a steady-state analysis which takes minutes — not half a day — and we get a nice temperature profile then iterate from there.
Want to learn more about practical solutions for accelerating cycle times in medical device manufacturing? Join Dr. David Crispino live at SPE/MPD’s Minitec at MD&M West in Anaheim, California, on February 4, 2025, where he’ll present “Enhancing Injection Mold Cooling Efficiency with Steady-State Thermal Models and Composite Core Technology.” Don’t miss this opportunity! More details are available on the Atalys website or sign up on LinkedIn.
About Atalys
Atalys is a vertically integrated manufacturing platform for the medical device and life sciences industry specializing in the delivery of highly engineered components and high-precision life-saving parts through its R&D, engineering, mold-making, and production expertise across the entire manufacturing process. Visit atalys.com for more information.