For most injection molding jobs, cooling is 95% of the cycle time. Cooling starts at the end of the first stage of injection, the rest of the molding cycle—pack, hold, cooling, mold open, and mold close—is cooling time. So, if cooling is the greatest portion of the cycle, why do most processors overlook this critical aspect of molding and lose an enormous percentage of profit as a result? Cooling rate also has a significant effect on part dimensions. To keep your dimensions on target you must cool the part identically on every cycle.
You have to take the heat out of the plastic that’s filling the mold. For heat transfer to take place, there has to be a difference in temperature between the two objects, a delta T. You have two heat-transfer paths, plastic to steel and steel to the coolant, usually water. The greater the difference in temperature, the faster the heat transfer.
The first ΔT, between plastic and mold temperature, is usually large, which is the good news. The bad news is that plastics shrink upon cooling and the part pulls away from the cavity wall before you go into second-stage pack and hold. If the plastic isn’t touching the steel, heat transfer drops astronomically. Make sure the mold parting line is tight so you can use appropriate pack pressure to push the plastic up against the cooled steel for as long as possible.
The second ΔT, between the mold and coolant temperature, is also large, and again this is the good news. The bad news is how poorly we work at maximizing this ΔT. Transferring heat from the steel to the coolant is not as simple as you would like.
But thanks to Oswald Reynolds, we have a handle on this and how it relates to ΔT. Reynolds studied heat transfer in pipes and found that you need a certain minimum water flow through a given size pipe to obtain the optimum ΔT. If water travels too slowly through the pipe you get laminar flow, which means the water flowing near the sides of the pipe gets hot and the ΔT goes down. Reynolds taught us how to get the maximum ΔT: Make sure you have turbulent flow and do not allow more than a 4° F change between the delivery and return coolant temperature.
How do you tell if you have laminar or turbulent flow? Turbulent flow is defined as a Reynolds number of ~5000 or higher. The higher the Reynolds number, the better the heat transfer. The Reynolds number equation is:
Re = 3160 x Q ÷ D x n
Re = Reynolds number (measure of turbulent flow)
Q = gallons/minute (GPM) of coolant flow
D = cooling channel diam., in.
n = kinematic viscosity, centistokes (cs), which varies with temperature
This equation isn’t difficult, but it’s not going to be used out on the production floor, where nobody has time to make such calculations. So make it painless and easy. Develop a spreadsheet that allows you to plug in the temperature of the water and the diameter of the cooling channel in order to read out the GPM needed for good cooling (turbulent flow). You can copy my spreadsheet (scientificmolding.com/reynolds.asp). Just remember, you have to do this for each cooling line in the tool and then add it all up to find the minimum GPM needed for the complete mold. You then will know how to specify the chiller or temperature-control unit (TCU) you need for this mold.
Assume, for example, that a mold has four 1/4-in. NPT lines and two 1/8-in. NPT lines. At a water temperature of 100 F, each 1/4-in. line needs 0.5 GPM and each 1/8-in. line needs 0.3 GPM. Add them all up (4 x 0.5 + 2 x 0.3) and you know the cooling unit has to provide at least 2.6 GPM to provide a Re of 5000 or higher for each line.
Once the GPM demand is established, we have to plumb the mold to make sure each line gets what it needs. Mother Nature is against us, as water will take the path of least resistance. In the example above, using a parallel-flow hookup (a line to each channel from a manifold), even if we had 3 GPM, the 1/4-in. lines would get most of the flow, so the 1/8-in. lines would not have turbulent flow. You can buy water regulators, but due to factors such as cost, poor water quality, temperature range, pressures required, and the number of water regulators needed, they are not practical for most plants. Other issues with parallel-flow plumbing are detecting blocked or partially blocked lines and a greater chance for incorrect water-line hookups.
There is the alternative of series plumbing for the mold, where one loops the inlets to the outlets. This makes for faster plumbing connections, as you only have to hook up one delivery and one return coolant line if you permanently install the jumpers. However, you have other issues, like violating the rule for maximum temperature increase from delivery to return lines and the need for higher pressure to pump the coolant, as the flow path can be long.
To attenuate these factors, use series plumbing together with a positive-displacement pump on your cooling unit rather than the typical centrifugal pump. The benefits are many, and there are no worries about water-flow regulation, as the pump will always put out its rated GPM.
If there is a blockage, it shuts down and alarms. Use a 5-GPM pump and the odds of the temperature rise being above 4° F become low. Monitor delivery vs. return temperatures with either a surface-contact thermocouple, infrared non-contact temperature sensor, or infrared thermography. Accurate temperature measurements are ideal, but you only need to know the temperature difference. My bet is you will be surprised at how many molds you can plumb in series with acceptable ΔT using positive-displacement pumps at 5-10 GPM.
If heat transfer depends on turbulent flow and the temperature difference between the steel and the water in the line, you now have to make sure the water lines are clean. (I can hear the groans now about rust, calcium buildup, fishbones and other trash in the water. ) Use filters and water softening or another technique to reduce calcium content to parts/million. And have a protocol in place to clean your cooling lines in all tools at least once/year.
Molding is a thermal process: You melt the plastic, force it into a mold, and cool it. If you want consistent parts, the thermal cycle must be consistent from shot to shot and run to run. Keeping the mold at the right temperature and paying attention to consistent GPM and turbulent flow in each channel are essential. But you need one more thing for all of the above to work: consistent cycle times.
John Bozzelli is the founder of Injection Molding Solutions (Scientific Molding) in Midland, Mich., a provider of training and consulting services to injection molders, including LIMS, and other specialties. E-mail firstname.lastname@example.org or visit scientificmolding.com.