is part of the Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Taking the heat (away) with pulsed cooling

You wouldn't guess it based on the relatively few molders who actually use it, but pulsed cooling has been around for more than 20 years. However, it's been just in the last few years that technology has made this method of mold temperature control more attractive to molders.

Look on the Internet at injection molding forums, and you often see a molder asking about the fundamentals of pulsed cooling. Most have heard of it, a few have seen it, several are curious: What is it? How does it work? Does it work? Is it worth it? How much does it cost? What do I do to get into it?

Pulsed cooling is one of those technologies that requires you embrace a relatively different concept, in the process setting aside some of the mold temperature control tactics you've used for so long. The physics that describes pulsed cooling is probably voluminous enough to make for a good master's thesis. But even the basics are worth exploring.

Figure 1. Conventional continuous flow mold temperature control (top), according to pulsed cooling proponents, generates temperature gradients called isotherms (contour lines in illustration). The isotherm segregates the two halves of the mold. Although heat travels from the polymer to the metal to the coolant quickly, the mass of steel between the channels and the mold shop goes unused. Often the largest part of the mold, this section increases heat flow from the shop air to the mold metal to the coolant, which can lead to sweating in extreme cases. Pulsed cooling (bottom), because it's not continuous, eliminates the isotherms that segregate a conventionally cooled mold. Heat from the part is absorbed not only by the cooling channels, but also by the large mass of steel on the shop-side of the mold. When the fill stage is complete, coolant circulates quickly to whisk away excess heat and quickly bring the mold and part back to minimum temperature.
Conventional Control
Conventional mold temperature control, say pulsed cooling advocates, is inefficient because it attempts to keep the mold at a set temperature throughout the cycle, whether the cycle is in the fill or pack and hold stage. This creates what's called a thermal barrier. It's a gradient band of temperate mold steel that starts at the cooling channels and emanates outward. Eventually these gradients merge, segregating the core and cavity from the rest of the mold (Figure 1). As a result, the steel on the shop-side of the mold is unused for cooling and can sweat if its temperature falls below the dew point.

Further, while coolant enters the mold at a set temperature, it absorbs heat as it travels through the mold. Therefore, cavities near the inlet of circulating fluid tend to run cooler than the cavities near the outlet. The proponents of pulsed cooling believe a mold kept at one temperature throughout the cycle can restrict the flow of resin entering the mold; as soon as melt hits the cavity, it starts to cool and solidify, even before the part is full. This, combined with cool and warm spots, can lead to molded-in stress, uneven cooling, warpage, and out-of-spec parts.

Figure 2. Like molding itself, pulsed cooling is a cyclic process. In a conventional continuous flow water system (top), coolant forces a heat gradient around cooling channels, resulting in molded-in stress in parts. With a pulsed cooling system (bottom), melt entering the mold is not resisted by a continuously cooling mold. Instead, melt is allowed to flow freely to fill the mold. After fill, a pulse of coolant quickly removes the heat gained, and the mold seeks thermal equilibrium as the heat gradient dissipates. Key to Symbols: Q+ = BTU melt heat input, Q- = BTU cooling pulse input, ÿt1 = warm water cooling temperature, ÿt2 = cold water cooling temperature difference, St = surface temperature, Rt = reaction time, Vot = valve open time, Vct = valve closed time, X = reduction in cycle time.
Complicating this configuration is that most mold temperature controllers measure the temperature of the circulating fluid, not the temperature of the mold steel itself. While the fluid in the cooling channels absorbs heat from the mold steel, the temperature of the fluid is always "behind" the temperature of the mold and will rise and fall only after the temperature of the mold steel rises and falls. This delay makes it difficult for mold temperature control systems to adjust quickly to changing temperature conditions.

Pulsed Cooling Control
Simply put, pulsed cooling is a closed loop system that measures mold surface temperatures and pulses cool or ambient water to the mold to quickly dissipate heat created by the addition of hot melted resin. The idea is to add and remove heat from the mold in natural peaks and valleys in the molding cycle (Figure 2). "It's a cyclic action," says Horst Wieder, president of Cito Products (Watertown, WI), a pulsed cooling systems manufacturer. "It's not extrusion. It should be treated cyclically."

First, pulsed cooling, despite the relatively sophisticated name, requires little in the way of additional hardware. To make it work, all you need are at least two mold temperature sensors, a regulating valve between the press and the coolant source, and a controller. Beyond that, it draws from the same chillers and coolant systems you have right now. That's the easy part. Where most people get hung up is in understanding just how and why it works.

Pulsed cooling relies on three basic principles that make it different from traditional cooling:

  1. Measure and control the temperature of the mold steel.
  2. Don't inhibit the flow of resin during fill by circulating coolant.
  3. After fill, pulse cool water through the mold to remove heat.
Measure and control the temperature of the mold steel. This concept is at the core of pulsed cooling. Instead of measuring the temperature of the coolant, molders install temperature sensors in the mold itself, at least one for each half of the mold (Figure 3). Holes are drilled between the cavity and the cooling channels and any competent molder can do the job himself. The sensors are then connected to a pulsed cooling controller, a relatively simple device that tracks temperatures and modulates the flow of coolant.

Each sensor in the mold is called a zone. At least two zones are recommended for effective cooling. You can go up to four, using eight or more zones for larger and multi-cavity molds. You can use pulsed cooling without temperature sensors, but such a configuration strays greatly from the fundamentals of the process.

Don't inhibit the flow of resin during fill by circulating coolant. This is where most molders start to balk at pulsed cooling. The idea is, during the fill stage of the cycle, to cease the flow of coolant, letting it sit in the channels. This is done by the pulsed cooling controller, which shuts off the valve that regulates the flow of coolant from the chiller.

Figure 3. Effective cooling is best achieved by measuring the mold steel temperature--not the coolant temperature. For efficient cooling, pulsed systems require the installation of at least two temperature sensors, one in each half of the mold. They should be installed between the cooling channels and the part. These sensors are wired to a controller which governs the operation of a pulsed temperature control system.
When the hot resin hits the mold, it flows freely into the mold. Rather than immediately trying to lower the temperature in the cavity, heat is allowed to build naturally, unabated. Flow, in theory, is relatively even and uninhibited, producing a stronger, more robust part. Heat generated, meanwhile, transfers from the part to the mold steel to the static coolant in the channels. The heat can also move past the channels to the steel on the shop side of the mold, warming it and preventing sweating.

After fill, pulse cool water through the mold to remove heat. Let's say you've installed the temperature sensors, hooked up the controller, and connected the regulating valve between the press and the coolant source. You decide that at ejection, you want the mold steel to be 150F--the temperature that produces good parts at the best cycle. So, you start a cycle, coolant stops flowing, resin is injected, and heat starts building. Then, at the end of fill, or at a set maximum temperature--say 220F in this case--the controller opens the regulating valve and cool water pulses through the mold, removing heat and quickly bringing the mold back down to 150F before part ejection.

How does this work? Letting heat build naturally is not as drastic as it might sound. Remember, the fill stage for most cycles doesn't last more than a second or so, depending on the size of the part, melt pressure, and rate of injection. And even on large parts with multi-second fills, it's still a minority fraction of the cycle.

So if the mold is 150F at the start of the cycle, a pulsed cooling system might let the temperature go up to 220F, as opposed to a continuously circulating system that might hold the temperature to 180F. The question is how can pulsed cooling bring the mold back down 70 deg F (from 220F to 150F) faster than the continuously circulating system, which has only 30 deg F to go?

It relies on a basic premise of thermal conductivity: As the temperature difference increases between the two media (coolant and mold steel), the rate of heat transfer also increases. Remember, you're pulsing cool water, maybe 60F or so. That means the temperature difference (ÿt) is 160 deg F. This difference means heat quickly and readily transfers from the part to the mold steel to the coolant. The ÿt in a conventional system, where the coolant might be 130F, is not as high. The rate of cooling, therefore, is reduced. The result, most often, is a faster fill (no resistance) and faster pack with repeatable hold.

Who Supplies It
Currently, there are two primary suppliers of pulsed cooling systems. One is Wieder's Cito Products. The other is RE Promotional Services (REPS), based in Cannock, England. Systems from each suppliefollow the same cooling philosophy, but there are some differences imechanics and execution.

Temperature control fundamentals
Rowland Evans, owner of REPS, has assembled this list of mold temperature control fundamentals. These basics of cooling are the guts of his pulsed cooling philosophy. However, this list bears studying and saving whether or not you use pulsed cooling.
  1. Injection molding is a cyclical process; therefore, the heat input from the polymer is also cyclical, reaching its maximum temperature just after the part is filled.
  2. Hot polymer contacts mold surfaces for only part of the molding cycle. This varies for the fixed and moving half of the mold.
  3. Heat can be removed from the molded part only while it's in contact with the mold surface.
  4. Coolant cools the metal of the mold. It's the metal of the mold that cools the plastic part.
  5. Heat flows from regions of higher temperature to regions of lower temperature. The greater the temperature difference between the two regions, the faster the heat will flow. The only limit is the thermal conductivity and diffusivity of the heat transfer medium (mold steel).
  6. The amount of heat it takes to raise the temperature of the medium (steel) depends on its specific heat and mass--for practical purposes, its volume and density.
  7. The heat required to take a part from melt temperature to ejection temperature must be transferred to the mold metal and to the coolant in less than one molding cycle time--between part fill and ejection.
  8. It's impossible to design too much cooling into a mold if it is controlled. It is easy to design too little cooling into a mold, which limits the part quality and cycle time.
  9. The mold cooling channel system is as important as the gate position, part venting, or ejector pins.
  10. All molds have a minimum of three different cooling requirements:
    • a) The moving half.
    • b) The fixed half.
    • c) The polymer feed system (sprue, runner, gate, or hot runner).
      It is rare that all three have the same cooling requirements. For all practical purposes, assume they never do.
  11. To produce good quality parts, from a cooling point of view, there are two considerations:
    • a) The heat characteristics of the part design, mold design, and mold cooling system design.
    • b) The mold temperature control system.
  12. There are no magic black boxes. Any control system can only make the mold perform as well as its design allows. A control cannot "fix" a poorly designed mold, it just helps it run as well as possible.
  13. Coolant is only useful when it removes heat from the metal that is in proximity to the part. Everywhere else in the mold it only cools the molding shop.

For instance, Rowland Evans, owner of REPS, interfaces his controller with the machine's and starts coolant pulse when it reads the end-of-fill signal.

The coolant then runs for a calculated time to remove the heat input from the polymer and bring the mold back to ejection temperature. Wieder says his systems circulate the coolant when the sensors reach a certain maximum temperature, usually at the end of fill. Then, at a set minimum temperature, circulation is stopped.

Evans says most molders can grasp the concept of pulsed cooling. What many don't like, he notes, is the thought of putting a temperature sensor in the mold. "I think it's a lack of understanding of how very important the sensor is," he says. "It's just a reluctance to do something they're not used to doing." For many molders and shop floor managers, he says, drilling a hole in the mold falls outside of standard operating procedure. It contradicts what they've done and have known to work for many years. Still, knowing the temperature of the mold metal is the centerpiece of pulsed cooling. "Molders talk about mold temperature, but most of the time they're talking about water temperature," Evans says. "And it's fatal."

Both Evans and Wieder say pulsed cooling can be used on any injection molding application, regardless of material, cycle length, or heating and cooling requirements. The two also admit most of their new business comes from molders who've hit the wall with a particularly troublesome high-heat material that's chronically out-of-spec (see sidebar) but emphasize that any molding application can benefit from the technology. Evans says pulsed cooling is widely used on molds with molding surface temperatures between 50F and 140F. Pulsed heating and cooling is used between 86F and 380F without the use of heated fluids. Wieder says he has a system installed on a 5-second cycle mold as well as on molds with cycles that are minutes long. "There really is no limit," he says.


GE Plastics tested the effectiveness of pulsed cooling with a laptop computer bezel mold, processing these three materials at different temperatures. The cycle time and mold surface temperature were measured to assess the performance of the cooling systems.
  Melt Mold
Material temperature, F temperature, F
PC/ABS 525 175
PBT 490 110
PC 620 180

Experimental Results
Theory and philosophy aside, there is empirical evidence demonstrating the effectiveness of pulsed cooling. Most notable is a paper authored by Amanda Yelsh and Peter Zuber, technical development engineers at GE Plastics. The paper documents the results of a series of tests Yelsh and Zuber conducted while molding with pulsed cooling systems. The part used in the tests was a laptop computer bezel. Yelsh and Zuber installed four temperature sensors, one in the A half and three in the B half of the mold.

Three different materials were processed in this mold with different mold-temperature-at-ejection and melt-temperature setpoints (see Table 1). GE used a Cito Products pulsed cooling system for the tests and also installed strip heaters in the mold to help it reach operating temperatures faster on cold starts. All three materials were processed with a traditional circulating mold temperature control system and the pulsed cooling system. Yelsh and Zuber measured mold surface temperatures at part ejection with a pyrometer and recorded cycle times for each material and cooling configuration to obtain results.


At part ejection, the mold surface temperature was measured with a pyrometer to determine which process--conventional cooling or pulsed cooling--best maintained the setpoint. On average, pulsed cooling varied ±3.2 deg F. Conventional cooling varied ±23.7 deg F on average.
  PBT   PC/ABS   PC  
  110F setpoint   175F setpoint   180F setpoint  
  Actual Deviation Actual Deviation Actual Deviation
Zone 1 110 0 173 -2 178 -2
Zone 6 111 1 174 -1 177 -3
Zone 8 121 11 173 -2 173 -7
Zone 1 148 38 184 9 191 11
Zone 6 140 30 174 -1 195 15
Zone 8 167 57 188 13 219 39

For cycle time, Yelsh and Zuber measured a 14 percent reduction between conventional and pulsed cooling systems for the PC/ABS and a 10 percent cycle time reduction for the PC. The PBT could not be measured because significant shrinkage in the part prevented the discovery of a good benchmark to identify minimum cycle time.

Next, they measured the mold surface temperature upon part ejection. On average, Yelsh and Zuber discovered that pulsed cooling held average surface temperatures at ejection to within ±3.2 deg F in the three zones measured. Conversely, conventional cooling, on average, held surface temperatures to within ±23.7 deg F (see Table 2). Not only did the pulsed cooling system hold temperature better, but in the PC/ABS test, each pulse averaged .2 gal of water per cycle. This compares to the 3.8 gal average used in a continuous-flow system.

Anecdotal evidence
Link Plastics, a 55-press custom molder in Telford, England bought its first pulsed cooling system from REPS about four years ago. John Wagstaff, product development manager at Link, says, at the time, Link had just won a job to mold electric drill housings from glass-filled polycarbonate for Makita. He said the mold had to be kept at 250F but ran consistently too warm. "Dimensionally, the parts were changing from room to room and from batch to batch," he says.

Makita gave Wagstaff permission to modify the mold, and he put in the pulsed cooling equipment. Cartridge heaters in the original mold were replaced with induction heaters for preheating, temperature sensors were installed, and pulsed cooling was used to bring the mold down to temperature after fill. "There's no installation required, really," he says.

After the system was installed, "the hassle factor completely disappeared," he says, "and the customer was happy." Dimensions stabilized, and scrap was reduced. Link's work for Makita went from four molds to 130, thanks largely to the reliability of the process and consistency of parts. Wagstaff says Link installed pulsed cooling on other glass-filled PC molds and uses it where cooling requirements are crucial.

For those considering pulsed cooling, Wagstaff says molders who've already invested in traditional mold temperature control systems should not bail out completely. He recommends a switch to pulsed cooling gradually as new molds and programs are introduced and as old equipment fails. For molders building a new facility or starting from ground zero, he sees no reason why pulsed cooling couldn't or shouldn't be implemented plantwide.

The Cost
As with most systems designed for use with molding machines, the cost of a pulsed cooling system is dependent on a lot of things. For a ballpark estimate, Evans says he sells a basic four-zone system for as little as $6000, but a typical system runs about $10,000. For an eight-zone configuration, you can expect to spend $18,000 to $20,000. Wieder says his systems generally run about $2000 a zone, including sensors, valves, controllers, and other ancillary equipment. A basic two-zone configuration will run $3600 to $4000, he says.

Return on investment also "depends," but among savings and improvements Evans and Wieder cite a 6 to 10 percent cycle time reduction, reduced water use, less molded-in stress, reduced scrap, and reduced energy use.

Wieder and Evans both say installation and setup is relatively simple and can be done by the molder.

Contact information
Orion, MI
Don Seres
Phone: (248) 391-1405
Fax: (248) 391-1409

Cito Products Inc.
Watertown, WI
Horst Wieder
Phone: (920) 261-5799
Fax: (920) 261-1350

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.