Scientific Molding: Part III - Infrared thermography

Injection molding is, after all, a thermal process. Despite the "mechanics" of the molding process - the rotating screw, the injecting ram, the moving platen - the ultimate function of the press is to control temperature. Friction due to screw rotation, aided by heater bands on the barrel, melts resin pellets, while turbulent water flowing through the mold removes that heat to produce a hardened part - quickly. Too much or too little heat at the wrong place at the wrong time can cause a myriad of headaches: short shots, galling, splay, stuck parts, shear, material degradation, and brittleness, to name a few.

The machine's attempt to maintain and control this precarious thermal system is an ongoing struggle. Unfortunately, you can't see thermal energy, only its effects. Thermal energy radiates in the infrared spectrum, outside the spectrum of visible light on which we humans rely so heavily to assess our environment, including the molding environment.

It was this conundrum that drove John Bozzelli to start using infrared thermography to measure temperature on a molding machine. Bozzelli, the principal of Midland, MI-based IM Solutions - a one-man firm advocating and teaching the ways of scientific molding - often works with Ashland Chemical Co.'s General Polymers Div., helping molders improve and optimize processes. His work with infrared thermography has become a de facto extension of the scientific molding process, bringing to light an important world that is usually unseen by most molders.

What the Heck Is It?

Along the electromagnetic spectrum - which includes everything from gamma rays and x-rays to FM and AM radio waves - visible light occupies only a fraction, in fact the smallest fraction. The wavelength of visible light ranges from 4 x 10^-7m on the violet side to 7 x 10^-7m on the red side. Just past red visible light is the infrared range, running from about 10^-7m to 10^-3m. Energy with such a wavelength is radiant or thermal energy, not visible to the human eye. In the summer when you barbecue, hot coals glow red, but the heat you feel from them is given off in the infrared region of the spectrum.

So as you can imagine, the injection molding process is bustling with infrared radiation. Sources include the barrel, the nozzle, cavities and cores, fresh parts, water lines, heat exchangers, electrical cabinets, dryers, hoppers, and hydraulic lines. Through the eyes of an infrared video camera - designed specifically to detect energy in the infrared range - the injection molding machine, mold, and auxiliary equipment take on a startling and enlightening appearance.

Cooler temperatures, presented by the camera, appear blue. Warmer temperatures start red and move to yellow. The very hottest temperatures appear white on the camera's display. This camera can show you things you might never have noticed, or might not have noticed until it was too late.

Case in Point - Again

Such was the case at the South Carolina molder visited by Bozzelli and General Polymers' technical service representative, John McDonnell. After analyzing and optimizing filling and cooling, they brought in an infrared camera to look at the hidden thermal world. Bozzelli says the goal of the infrared analysis is to look for relative temperature discrepancies. That is, an infrared camera can be fooled by energy reflected off a shiny surface - such as mold steel - making a surface appear warmer than it really is. What the camera does do well is show hot spots, or dramatic and unwanted changes in temperature. For example, an infrared camera trained on mold cores can show if one is running particularly hot, relative to the other cores. Bozzelli says he typically looks for large temperature differences, usually anything +/-20 deg F. "That usually means the system's not stable," he says.

At the South Carolina molder, Bozzelli and McDonnell shot infrared images of the barrel, the nozzle, major hydraulic lines, cores and cavities, parts fresh from a mold just started, parts fresh from the mold after running for 30 minutes, water lines in and out, dryer outlets, and heat exchangers. The most startling and revealing discovery with the infrared camera came when it was trained on the electrical cabinet of the machine.

Although it was difficult to see with the naked eye, on the camera's display one particular wire leapt out as bright white. Wires, because they reflect little energy, show their actual temperature reliably. In this case, the camera indicated the wire was 116F (Figure 1, right). Upon closer inspection with the naked eye, it was obvious that a short in the wire had caused almost all of the cover to burn off. "Looking at it, it seems so obvious. You wonder why no one noticed it before," Bozzelli remarks, "but hindsight is 20/20. We just didn't pick it up until we saw it with the camera." The molder had the frayed wire and short fixed within a day.

In the mold, Bozzelli looks for "a variance in mold temperature from one side to the other or hot spots in the mold." The challenge presented by the mold is that the shiny tool steel easily reflects heat from other sources. Because of this, it's sometimes more accurate to shoot parts fresh out of the mold, instead of the mold itself. Typically not as reflective, parts betray hot spots in the cores and cavities. To establish a benchmark, Bozzelli likes to shoot parts emerging from a cold mold first. Then, 30 minutes later, he'll return to shoot the parts after the mold and machine have warmed to operating temperature. At the South Carolina molder, the parts - gunshot primer trays - appear consistent in temperature; each exhibits particularly warm centers (Figure 2, right).

When he can shoot the cores and cavities without too much reflection, Bozzelli says he pays particular attention to the temperature difference between the core and the rest of the mold (Figure 3, right). Because a core typically removes 60 to 70 percent of the BTUs in the part, most parts shrink away from the cavity and onto the core. A temperature difference of 20 to 100 deg F here can cause parts to stick and induce warp or bowing. This can result in longer cycles as cooling times are increased to bring the part back within spec.

Bozzelli also likes to use the camera to check water lines and heat exchangers. Turbulent water flow, even if the water is cool, should emit some radiant energy. A water line that looks particularly cold on camera may actually be blocked, which, in the case of a mold cooling system, could account for hot spots in the cores and cavities. At the South Carolina molder, an apparently blocked line at the heat exchanger appeared as light red to the infrared camera, where it should have been dark blue (Figure 4, right).

At the dryer outlet, Bozzelli likes to check the temperature to verify that the hot air emerging approximates the setpoint on the dryer control panel (Figure 5, right). He also points out that hot air traveling through an uninsulated hose can lose as much as 10 deg F/ft en route to the hopper; as the temperature drops, moisture can sneak back into the system. Because of this, he likes to get an infrared image of the hot air inlet at the hopper, to see what energy was lost.

The Goal

What Bozzelli is ultimately aiming for with infrared thermography is a way to accurately and consistently measure temperature on the press, not just in the mold or on the dryer, but in the barrel as well. "Can we put a man on the moon?" he asks. "Yes. But can we correctly measure temperature on an injection molding machine? No." He says the infrared technology to accurately measure temperature exists, it's just a matter of properly applying that technology to the process.

The problem infrared thermography presents is that every material has its own infrared "signature," or wavelength at which it is best read by a camera. At certain wavelengths, an infrared camera may shoot right "through" a clear polycarbonate part, while at its signature wavelength, it will read the part directly.

Over the last several years Bozzelli has tested several different cameras that shoot at different wavelengths. He has settled on what he believes is a camera with a wavelength that most accurately detects and measures temperature on and around the press, including most plastic parts. The wavelength Bozzelli chose is proprietary, but he is expecting a new camera using this wavelength to arrive sometime in January 1998 and will be shooting willing molders soon after.