Infrared thermography-made-practicalhighlights mold, part hot spots

Your mold has hot spots. Real hot. Parts are sticking on the core. Cooling takes forever. Warp is a four-letter word.

If only you could see those hot spots without climbing into the mold, or interrupting the cycle. If only you could measure temperature without tapping holes in the mold. If only you could just look at the mold and see the heat. If only your mold temperature controller saw the same thing and fixed it for you.

The nirvana of that last sentence is the goal of Scientific Systems Co. Inc., a Woburn, MA-based firm that’s developed an infrared-based mold temperature control system that checks for mold hot spots and interfaces with a controller to keep temperatures under control. Led by president and CEO Raman Mehra, SSCI’s experience is mostly in controls and guidance systems for DOD and aerospace applications. This infrared system for molding is part of the effort by both the company and the federal government effort to transfer emerging technologies to commercial applications.

What It Is

The concept of this system is easy. An array of infrared sensors is attached to the platen next to or above the mold and positioned to read the temperature of a section or region on the opposite half of the mold. The operator punches into the controller a variety of temperature setpoints, and when a region exceeds its setpoint, the mold temperature controller is alerted and cooling to the region in violation is enhanced. Perhaps more ideally, this system can be used in the same way to measure part temperature when the mold opens.

The IR sensors SSCI uses measure about 31/2 inches long and 11/2 inches in diameter, and use an integrated signal processor. Each sensor is tied to a 41/2-by-41/2-inch circuit board that can be mounted on the molding machine. The system also comes with a small display screen to show current temperature readings. The circuit board itself can be connected to an optional PC running SSCI software that provides control features and data logging, and converts the IR signals to a type that can be read by a standard mold temperature controller. In an ideal application, the PC-based control would be linked to the mold temperature control system, which would use the temperature data to adjust cooling tactics.

The number of IR sensors a mold uses depends on how large the mold is and how sensitive the part’s dimensions are to mold temperature changes. Mehra says a typical four-cavity mold might have four sensors, two on each mold half, each reading a region on the opposite half.

A sensor is capable of reading an area ranging from 3/4 inch to 5 inches in diameter. This means a molder cannot practically read the entire mold half; he must instead selectively read different regions that are determined to be at risk for overheating.

SSCI has also developed a pan-and-tilt IR system. This version, like the stationary system, is positioned on the platen, but can read more than one region of the mold after it opens. Mehra says his IR system is likely not for every molder. He suggests that molders producing high-end, dimensionally sensitive parts for automotive and medical applications would find this system the most beneficial.

The Real World

This system has been tested in a real world environment. Bob Relyea is a senior plastics engineer at Nalge International in Rochester, NY, a manufacturer of plastic labware. Relyea and Nalge agreed to try out the infrared system on a single-cavity mold that produces a medium-sized graduated cylinder molded of PP. The system was also tried on a four-cavity mold producing filter components molded of PS. The IR system was run for about 6 hours on the graduated cylinder mold, and about 6 hours on each of three occasions on the filter component mold.

On the graduated cylinder one sensor was positioned to read the core at about its midpoint. Another read the cavity at a similar point. Relyea says that dimensional accuracy on the cylinder is critical, given that it’s used to measure fluids, and although he was experiencing no major problems with the part, the SSCI infrared system pointed out some unknown hot spots and proved potentially valuable.

Number one, he says, is the ability to measure part or steel temperature in process, rather than measure water temperature. “Getting the temperature of the product right at the product rather than measuring something downstream or upstream certainly is a better method to do it,” he says. Relyea says his faith in the results was such that he would consider doing SPC at the press by measuring part temperature.

Startups, he points out, were also shorter; once an ideal operating temperature was identified, he simply waited for the mold to reach that point. “Now I have shorter startup times because I can dial in a temperature and look for that temperature,” he says. The result, in the tests, was 30 percent less scrap than a standard startup.

The Bottom Line

The SSCI infrared system is not without a few minor drawbacks. The largest is that IR sensors can be fooled by low-emissivity surfaces, such as you would find on a mold. Mehra says one way to cope with this problem is to spray parts of the mold (not the core or cavity) with carbon black to increase emissivity. If emissivity is still a problem, SSCI recommends molders measure part temperature instead of mold temperature. Also, when reading a transparent plastic, some IR sensors, if not calibrated correctly, can look “through” the part and read objects or surfaces in the background. Mehra reports that he’s not had much experience with transparent parts yet, but does not expect they will present a problem.

Right now a standard IR system, which includes four sensors and all necessary controls, ranges from $5000 to $7000. The pan-and-tilt robot-based system ranges from $7000 to $10,000. SSCI is talking to major OEMs about possibly selling the system as part of a machinery package, and is currently marketing the IR system in conjunction with pulsed cooling system manufacturer REPS.