Injection Molding Case Study

Most enclosure articles stop at the molding. This one does not. A six-part housing is not a single part with a quote, it is a chain of decisions that runs from the first email to the last carton on the truck. Miss one step and the failure shows up three stages later, usually right before the ship date. Below is the whole program Xiamen Ruicheng ran for a Russian industrial-equipment maker, phase by phase, using the drawings, tools, trial shots and change records from the program itself.

The customer arrived with a complete 3D assembly, six parts that had to agree with one another as a single enclosure, and a surface-specification discipline sharper than most. Their DFM feedback was not a wish list, it was a set of working rules, and it shaped the entire program. In their own words, carried into the project as written requirements:

That last exchange is the whole thesis in miniature. The customer was not asking for a box, they were asking for a visible face that would not betray a gate scar. Xiamen Ruicheng answered it by locking one drawing convention before tooling:

The assembly resolved into six components, each quoted as its own mold. The shared baseline was consistent: an S50C mold base, P20 core and cavity, a single cavity per tool (1x1), and ABS in black. What changed per part was the surface map, and that is exactly where the earlier dialogue paid off. The part codes below are the ones stamped on the tools, so the article, the drawing and the mold all use the same names.

Mold design ran as 3D STEP models before any steel was cut: the parting line was fixed, every face was checked for release, and the red/blue surface map was carried into the tool itself. The customer's question about gate position and shrinkage on the thinner walls was answered on these drawings, not discovered on parts.

The six tools, as built

What follows is the tooling record itself. These are photographs of the steel molds, not studio renders: mold bases, cavities, guide pillars, slides and ejector springs, with each tool stamped with its own part code. On a six-part program you are not buying one mold, you are buying six that all have to hold the same surface language.

How the six parts become one enclosure

The six tools above each make one part. Together they build the enclosure carried through the rest of this program — a housing whose six components had to agree on surface finish and fit before it could function as one product. With the tooling proven, the next step is committing all six parts to a single production dispatch.

A formal production dispatch was issued to lock the program: one project name used identically across every mold, revision and production order, six parts listed with their mold base, core, cavity, material and surface requirement, and a hard internal deadline to deliver sample sets by late July. Three of the six parts were already flagged in the dispatch as needing mold revision, so the engineering team went in knowing where to look rather than finding out at first article.

The value of the dispatch is not paperwork. It is that the surface map and the part codes were fixed in one document before the tools ran, so a T1 finding could be traced straight back to a requirement instead of an argument about what the drawing meant.

This is the phase that decides whether a program ships clean or slips. The T1 samples were compared part-by-part to the 3D model and the surface map. A drawing can never show what a sample reveals, and on this program the sample revealed three things.

The reversed-face issue is the one worth sitting with. It originated in the part definition, not the tool, which means the only safe fix is a validated surface treatment on the now-visible face, confirmed on samples, not a silent mold change. That discipline is what kept the program from a late, deadline-breaking re-cut.

Trial runs, straight from the press

The clips below are the tools running on the machine during sampling, five of the six parts. There is no clip for the door because none was recorded; the stills of that tool stand in for it. These are the moments a part is first pulled, which is why the T1 review is tied to them.

Late in the program a consolidated change round folded the T1 findings and new feedback into one reviewed package. Nothing was applied to volume unverified. Three concrete change items were sampled and confirmed with the customer.

With the tooling corrected and every change confirmed on samples, the program entered mass production for 500 sets. The surface maps that started as colour-coded drawings now governed the floor: each part ran to its agreed red/blue finish, and 100% appearance inspection sat at the production gate rather than as a final spot-check. Because the deviations had been closed at T1, the inspection step verified conformance instead of discovering surprises.

The same six tools that had been photographed open and closed during tooling now ran the volume. Nothing about the part changed between T1 and the run except the corrections already confirmed on samples, which is the whole point of closing deviations at first article.

The enclosure left the building only after the packaging spec was locked from the transit-abrasion lesson: each part in its own individual box, larger bags giving full coverage, foam bags above the relevant volume, and shipping marks carrying both quantity and part number so the customer could receive and sort without opening every carton. The whole order was sequenced to ship 500 sets ahead of the Chinese New Year holiday, which meant the production order had to be confirmed before year-end to protect the window.