ADR Valves: Old vs. New

This entry has some prerequisite reading: my entries of July 3, 19 and 22. Those posts will give you the necessary background to understand the following:

So… what are the downsides to ADR-First? In other words, why has that approach not (yet) gained more broad appeal historically?

First, and most broadly, the perception/reputation of the early ADRs still is the common perception today, and the biggest issues were reliability and price. Let’s take a look at both:

Older ADRs used water valves to actuate the knives of the cutterwheel. This led to two issues: First, they could not be actuated crisply enough to extend one knife blade alone, the minimum was two. In principle, the system needs to throw only one knife blade when cutting a defect that is near the end of a strip (the majority of defects fit this description). To throw that extra blade where there is no strip is to invite extraneous cuts, increasing white cube rates and decreasing product length. And that was the norm for ADRs built before 1999.

Further, since it was not possible to throw a single knife, it was therefore not economical to cut good strips to length (the processors/sensors didn’t provide for that anyway, but that is another matter). If for every cut a cube is created, the yield loss does not justify the other benefits.

Further, water valves had short lives. When actuated, they created “water hammer” (that banging noise in your house water pipes that happens when you shut off a valve quickly, and no hammer arrestor is installed). Water hammer did just what its name implies: hit the valve with a large force each time it was actuated; basically the force was the result of the water’s momentum being stopped suddenly. Remember from physics, mV=FΔt? V is high due to the high impact required to crisply drive the knives, and Δt is short, again to keep actuation crisp. The result? Force applied by the momentum shift was VERY high.

And the valves beat themselves to death, usually in a few months.

Further, the failure mode of the valves was relatively catastrophic in terms of yield: When a valve failed, it would fail “open”, directing a continuous stream of water to the knives of a row, extending all blades continuously. So, for that one lane of the system (1/44th of the total flow), all strips were turned into cubes and sent to co-product, until someone noticed the problem and replaced the valve. That is a significant yield loss in most applications.

The conclusion of the industry at the time? “Send only the strips necessary for cutting (defectives). Do not send any good strips to the ADR, if possible. The chances are too high that good strips will become by-product.” That perception largely holds today.

Even though ADRs since 1999 employ air valves, which both practically eliminate the catastrophic failures and enable single knife throws. Not to mention the savings in maintenance labor and parts.

ADR is a different solution today, folks. It truly is a system worthy of consideration of being used as a primary defect control device. More on price and payback next time.

Tim

Stumble It!

Bookmark on del.icio.us