A trial that starts badly is rarely rescued later. The first thirty minutes set the operating window, and the operators’ confidence in the new material, for the whole run.
Before the run starts
Three preparations separate a clean trial from a confused one:
- Establish a baseline. Run the incumbent pulp for at least 30 minutes and record hammer-mill amps, forming vacuum, line speed, dust level and core weight.
- Condition the pulp. Bring the trial bales into the converting hall 24 hours early so moisture equilibrates with the room, not with the container.
- Fix one variable. Change the pulp, keep everything else constant. If you also adjust debonding or line speed, you will not know what caused the result.
Minutes 0–10: defibration
Watch the hammer mill current first. A rise of more than 8% against baseline within the first ten minutes usually means the pulp is drier or has a tighter fibre network than expected. A drop means the opposite — often a wetter bale.
Do not adjust the mill immediately. Let it stabilise for five minutes; fluff pulp takes time to reach steady state in the mill.
Minutes 10–20: forming
This is where the trial is usually decided. Look at the forming head vacuum and the visual appearance of the fluff stream.
- Vacuum rising steadily → the fibre mat is denser than baseline; check fibre length distribution.
- Visible lumps in the stream → moisture or insufficient defibration; slow the feed rate before changing anything else.
- Excess dust → too dry or too much debonding; check both before touching the mill settings.
Minutes 20–30: the core
Cut a core sample and check three things against baseline: weight per unit area, thickness under load, and absorption time.
If weight is within 2%, thickness within 5% and absorption within 10%, the pulp is behaving as specified and you can extend the run. If any of those three is off by more, stop and diagnose before running longer — a longer run with a drifting core produces scrap, not data.
The mistake most trials make
Teams often adjust three settings at once when the first core looks wrong. The result is a line that eventually runs well but nobody knows why, which means the knowledge cannot be transferred to the next line or the next shift.
Change one thing. Record what happened. That record is worth more than the trial itself.