Breaking Through Production Challenges to Create High-Quality LED Lamps: A Success Story

The bubbles that threatened this order were not a moulding fault. They were a drying fault: the press was behaving exactly as instructed, with resin that still carried moisture into the barrel. The deformation followed from the same source, because moisture and uneven cooling both pull a part out of shape after it leaves the tool.

Injection defects carry their cause in their appearance, and the order matters because the cheap corrections come first. A void sits inside the wall and comes from trapped gas or shrinkage; a sink mark is drawn into the surface over a thick section; a short shot never filled. Identifying which one is in front of you is what stops a shop from adjusting parameters at random.

The job was an LED lamp for a European customer with a strong position in high-quality office supplies, and it arrived with an unusually tight production cycle. Once production started, the problems came in a sequence: bubbles in the moulded parts, material deformation, and a defect rate that stayed high run after run. Left alone, the combination meant a missed delivery date as well as a quality failure, and the customer’s reputation in its own market was part of what was at stake.

The first correction was at the material, not the machine. The plastic was given a deep drying treatment and screened for impurities before it reached the barrel, with the moisture content in the raw material controlled closely enough to remove moisture-driven bubbles at their source. Material inspection was tightened at the same time, so that the incoming quality stayed uniform and the risk of deformation from inconsistent stock came down.

With the material under control, the process window was rebuilt. Injection temperature, pressure and speed were adjusted through a series of trials and a review of the recorded data, tuned to the characteristics of the specific material and the specific part rather than to a generic starting sheet. That produced more uniform flow during injection, fewer bubbles, and much less deformation from parameters that had been fighting the material.

The last change was to the tool. The mould structure was redesigned to add efficient exhaust grooves, so that gas could leave the cavity quickly during injection instead of being trapped in the melt. The cooling system was reworked so the material cooled uniformly and deformation from a temperature gradient came down, and the mould itself was built to a higher standard of precision and durability to reduce the defect rate that came from the tool rather than the process.

The result held. With the team working through the corrections in sequence, the defect rate fell sharply, the products shipped on time, and the customer received them at the quality level the order had specified. The delivery also settled the working relationship for later programmes.

Corrections are not equally expensive, and the sequence above follows the cost order that applies to most moulding problems. Process changes cost machine time and nothing else. Ejection and handling fixes cost fixturing or robot programming. Wall balance and ribs are a CAD change that has to happen before the tool is cut. Gate relocation and cooling-channel redesign are mould modifications. Working from the cheapest layer upward is what keeps a fix affordable, and it is worth agreeing the order with the moulder before authorising any tool work.

The sequence, and the distinction between uneven shrinkage and a high shrinkage rate, come from the shop’s notes on injection moulding warpage.

A defect rate looks like a quality number, but it is also a delivery number, because every rejected part is machine time spent on something that will be thrown away. Model it with stated assumptions: 5,000 good parts are needed, the press runs a 20-second cycle, machine time is charged at $45 per hour, and each part uses 18 g of resin at $3.20 per kilogram.

Halving the defect rate releases about $110 on that single order and roughly two hours of press time. On a tight cycle, those two hours are usually the difference between shipping on the promised date and negotiating one, which is why the drying and venting work behind this page was worth doing even though it was not the cheapest line on the job.

The records that make a moulding job repeatable are specific and cheap to keep: the drying log naming the resin lot, the dryer temperature, the time and the dew-point reading; the approved process parameter sheet that the first good part was run to; a first-article inspection report; the defect or scrap rate for each production run; and the mould maintenance record. Where a moulder has not kept a record, the honest answer is that it is available on request for the next run.

Two distinct causes look similar. Moisture carried in the resin turns to steam in the barrel and leaves bubbles in the wall, which is a drying failure rather than a machine failure. Gas trapped as the melt decompresses, or venting that cannot clear the last point to fill, leaves voids. Drying the material to the supplier’s specification comes first because it is the cheaper correction.

Dry and screen the resin first, then check the venting: efficient exhaust grooves at the last points to fill let gas leave the cavity instead of being trapped in the melt. Only after the material and the vents are right is it worth adjusting injection speed and back pressure, because parameters cannot remove moisture that is still in the pellet.

A sink mark is drawn into the visible surface over a thick section, so it shows on the outside of the part. A void sits inside the wall and comes from trapped gas or internal shrinkage, and it is found by sectioning or by X-ray rather than by looking at the surface. They are corrected by different levers, which is why the distinction matters before any parameter is changed.

Work upward from the cheapest correction: packing and cooling parameters first, then ejection and handling, then wall balance and ribs, then gate relocation, and only then cooling-channel redesign. If the bow follows the thick sections it is a wall problem; if it follows the flow direction it is a gate problem; if it drifts as the run goes on, the cooling circuit is the place to look.

The drying log with the resin lot, dryer temperature, time and dew point; the approved process parameter sheet; a first-article inspection report; the scrap rate for each run; and the mould maintenance record. These are what make it possible to reproduce a good run instead of rediscovering it.