A Lesson On Impact Of Variation In Yarn Quality

What The Eye Couldn’t Miss—The Data Did

“Can you spot the difference between these two fabrics and say which one is not good in appearance. Both were made from the same 60s combed compact yarn,” said Kato-san, spreading the samples under the focused glow of overhead lights.

Three of us—the Factory Manager (me), the QAD In-charge, and the Production Head—gathered around the table. We adjusted our positions and even switched sides, trying to get the best possible view.

Eventually, all three of us pointed to the same fabric sample. It had a faintly cloudy appearance, lacking the crisp brightness we expected. Though both samples were bleached white, this one looked slightly dull under the lights. There was a subtle inconsistency in the way it reflected light, almost as if a gentle haze clung to its surface.

Kato-san smiled. “That,” he announced, “is the fabric made with your yarn.”

We stared, momentarily stunned. How could this be? With our top-tier machinery and processes, we were certain our yarn should have looked superior. The silence between us was heavy with disbelief.

But Kato-san wasn’t done.

He gave us further shock!

“For your information,” he added, “the IPI of the better-looking fabric’s yarn is 175—yours is 110.”

It took us a moment to register the implication. Higher imperfections, yet better appearance? That went against everything we believed.

“Did the other mill use a different mixing?” our QAD ventured.

“No. Both fabrics used MCU-5 cotton,” Kato-san replied. “Same count. Same cotton.”

We couldn’t help but question if there had been a mix-up—labels switched, cones swapped?

Kato-san looked taken aback. He looked mildly offended as if we had questioned his integrity.

“Absolutely not,” he said, his tone clipped.

Then, he asked us to bring our lot reports. Side by side, we examined our quality data next to that of the other mill. The culprit was clear: our coefficient of variation (CV%) values were significantly higher.

“It’s not just the average imperfections that affect the fabric’s look,” Kato-san said. “It’s the variation that matters. That’s what the eye picks up.”

Despite our machinery and management systems, this was a painful truth to confront. I admitted, “Even with our modern setup… we clearly missed something.”

Kato-san nodded. “You’ve got the tools. But maybe there’s still more to optimize. I believe there’s still room for improvement.”

He began outlining his observations. We leaned in to listen carefully.

“Start with your mixing strategy,” he said. “You’re blending 25 bales. Why not aim for more, say 50? And in the drawing frame, make sure every feed creel is as diverse as possible, by representing cans from the maximum number of cards. Label the card cans clearly. Avoid repeating cards in the same batch at the draw frame feed. It’s the same principle for other back processes. Machines differ in age, condition, and settings—averaging across them helps reduce variation.”

Our production head cut in. “What about spinning?”

“That’s where it often gets overlooked,” said Kato-san. “Spindle-to-spindle variation matters—even more than we realize.”

We tried to remain neutral, but he read our expressions. Or perhaps, our poker faces weren’t quite as unreadable as we thought. He didn’t waste a single moment.

“Okay! Let’s check,” he said, briskly heading to the spinning department.

Down in the spinning department, he noticed a machine nearing doff. He asked us to stop it, mark the spindle numbers on each cop, then restart for automatic doffing.

Later, at cone winding, we adjusted the clearer to a tighter clearer setting. Of the 1,008 cops fed in, 28 were rejected. Kato-san reviewed the spindle IDs and pulled rejection data from the CPU. We checked those spindles and uncovered a range of overlooked issues to our shock.

The culprits are:

• Damaged top roller cots
• Missing or wrong spacers
• Off-centre spindles
• Worn-out travellers and aprons
• Clogged and loose lattice aprons of the compact system
• Lapped bottom rollers

Some issues were embarrassingly basic. A dozen defective cops remained unexplained—possibly due to roving faults.

It was humbling.

Kato-san wrapped it up perfectly: “Don’t let confidence in your systems stop you from questioning them. Revisit. Revalidate. Regularly.”

P.S.: This incident happened some 25 years ago. We then followed by checking the ring frames by marking the spindle numbers, once in a month. Nowadays, we have an automatic spindle monitoring system in the linked winders.
 
(Murugan Santhanam is Managing Director of Texdoc Online Solution Pvt. Ltd.)