A Pile of Money Sitting in the Yard Behind the Plant

The Problem

We were only just beginning to commission our new 8000 pound per hour mineral filled EVA sound deadening sheet line. A Farrel continuous mixer feeding a 10 inch hot melt extruder with hot oil temperature control in the feed of the screw, then out through a 110 inch sheet die — a significant investment that from day one refused to run correctly.

The thickness variation was our immediate headache. On thinner gauge product you could see it with your own eyes but regardless of target thickness we simply could not hold spec consistently.

The sheet scrap was every department's problem too. What we didn't fully appreciate at first was what that variation was doing downstream. In the thermoforming operation the inconsistency was creating real problems — thin sections would tear, thick sections wouldn't draw properly due to insufficient heat absorption and finished parts were failing to fit their assemblies correctly.

We were granulating the rejects and blending them back into the recipe at 25 to 50% but the reclaim granulator couldn't keep up with the volume we were generating. At 8000 pounds per hour the pile behind the plant was growing faster than we could deal with it.

The Investigation

Under pressure we did what untrained panicking teams do — we jumped around. We worked the recipe hard, adjusting EVA acetate content, resin melt flow and lubricant levels. We tried different screw designs, temperature profiles, melt temperatures, die adjustments and screen packs. We looked at the feeders, the Farrel mixer, the process oil delivery rate, the oil quality and the consistency of the mix. One by one we worked through everything we could think of, eliminating variables as we went.

Progress was slow. As we began to run out of obvious candidates I decided to hook up a strip chart recorder to the extruder head pressure. It was the one thing that would give us a continuous picture of what was actually happening inside the process. A more experienced engineer would have done this much earlier in the investigation. It would have saved us considerable time and a very large pile of scrap.

What we saw on that strip chart lets us know we were on to something. A sine wave pattern on the head pressure trace — rhythmic, repeating, consistent. The compound level in the feed section was rising and falling in time with it.

The Analysis

We now had something concrete to work with. The pressure was cycling which meant something in the process was cycling. We began to focus on the feed section of the extruder screw. The section was being heated with hot oil and with a mineral filled EVA compound — a material that becomes tacky at relatively low temperatures — we started to form a hypothesis. The screw wasn't consistently conveying compound out of the feed section. It was gripping, conveying, softening in the heat, gripping the root of the screw and slipping on the feed section /barrel feedsection wall. Then recovering and repeating. Stick slip — and that sine wave was its signature.

The Trials

We worked through combinations of screw feed section and feed throat/body temperatures, pushing the hot oil circuit on the screw as far as it could go. It wasn't enough. The system simply couldn't deliver the temperature drop we needed to break the cycle.

The Fix

We switched from hot oil to 60 degree city water on the feed section for a quick evaluation.

The sine wave on the strip chart flattened out. Head pressure stabilized. The sheet came into spec and the pile behind the plant stopped growing.

Were we completely certain of the stick slip theory? Honestly no. But the result spoke for itself.

The lesson we carried from this one had less to do with feed section/screw temperatures and more to do with how we approached the problem. The data was available to us far earlier than we chose to look at it. Hook up your instrumentation early, read what it's telling you and work backwards from there. Measurement minimizes guesswork.

Got a surging story of your own? Different material, different cause, different fix — drop it in the comments. Let's learn from our on the shop floor experiences.

The Fix — One real problem. One real solution. Every week on Plastic Extrusion Problem Solving.