9.2% DHU on Knit T-Shirts Three Days Before Shipment

1. Holes and ladders at shoulder and collar-attach seams. Mechanism: needle cutting. A size 11 sharp-point needle on a 180 gsm single jersey severs knitted loops instead of parting them. The damage is worst where several plies and a fusible or tape meet - shoulder and collar attach - because the needle is deflected and loads up. Contributing factors: maximum presser foot pressure (holds the loop in place so the needle cuts rather than pushes), an over-used or burred needle, and a needle plate hole too large, which lets the fabric be dragged down.

2. Skipped stitches that open when stretched. Mechanism: the needle thread is not being picked up cleanly by the looper. Causes to check in order: needle too fine for the thread size or wrong needle type for knits; incorrect needle-to-looper timing; thread path and tension on the needle thread too tight; needle deflection from the high presser pressure; and stitch length too long for the seam, so the seam has no reserve elasticity and 'grins' - the yarns separate rather than the thread breaking. A 4-thread overlock on a knit should have 10–15% stretch reserve and must stretch at least as far as the fabric.

3. Wavy, stretched neck opening. Mechanism: the neckline has been extended during sewing and then concealed by pressing. With differential feed at 1:1 and maximum presser pressure, the operator holds the neck back and stretches the ring while the machine feeds it flat, so the neckline grows. Ironing does not remove the stretch; it only hides it until the first wash, when the neck goes wavy again - which is exactly the condition the buyer's no-pressing rule is written to catch.

Machine and needle corrections

• Replace sharp points with ball-point needles: 75/11 or 80/12 for collar attach and 90/14 for shoulder seams on 180 gsm jersey. Change needles on a schedule - every shift or every 8 hours on knit seams - and discard any needle that has been deflected.
• Reduce presser foot pressure from maximum to the minimum that still feeds evenly; if the fabric slips, fix the feed, not the pressure. Check feed dog height and needle plate hole size and condition.
• Set differential feed to stretch the neckline slightly on the collar attach (roughly 1:1.2–1:1.3) so the ring is fed in rather than pulled out; keep 1:1 for side seams.
• Verify needle-to-looper timing, thread path and tensions; move to a suitable needle thread (textured polyester or cotton-wrapped polyester) matched to the needle size.
• Add a stay tape or a reinforcing strip at the shoulder seam where the defect concentrates.

Verification trial

1. Sew 30 pieces across all operations with the new set-up, alongside a 30-piece control from the old set-up, so the comparison is fair.
2. Test seam elasticity: a 100 mm seam sample should stretch to at least 80% of the fabric's own extension before the stitch cracks. Record the figures.
3. Test seam strength and inspect the needle line at 5× under a light to see loop damage, not just surface holes.
4. Wash three samples and measure the neckline flat before and after, after 24 hours' relaxation.
5. Only if DHU falls below 2.5 and the wash test holds does the line restart at full speed; keep the control chart running for the next three days.

Decision: do not ship the 800 packed pieces as they stand.

The reasoning runs from the buyer's own rule back to the shop floor:

• The buyer's manual forbids pressing to correct construction. The 800 packed pieces passed final inspection because the necklines were pressed back into shape. Their pass result is therefore not evidence of conforming quality - it is evidence of a masked defect. Flat-measured after 24 hours' relaxation, a proportion will fall outside the neck-drop tolerance and the buyer's incoming inspection will find them.
• Factory DHU of 9.2% against a 2.5% limit is nearly four times the allowance. Meeting 2.5% in the shipped lot is only possible by 100% sorting, not by sampling.

Action

1. De-press and relax: unpack, lay flat for 24 hours, then measure every neckline. Do not re-iron before measuring; measure first, then rework only the failures (re-set the collar with a fresh needle set-up or replace the ring).
2. 100% sort and rework the 800, and 100% inspect the 600 in line.
3. Re-measure the sorted lot against the 2.5% DHU target; ship only the sorted, re-inspected quantity and record the sort as a deviation.
4. Talk to the buyer now, not after shipment: offer a split shipment with an agreed revised date for the balance rather than letting the buyer discover a masked defect on arrival.
5. Cost the two options - a short delay with sorting labour versus a chargeback plus return freight plus the loss of the account - and put the comparison in front of the production head, since the argument is commercial as much as technical.