22-Mec-A4 Design and Manufacture of Machine Elements · December 2017

Question 3 of 6: Weld-Line Defect and Feature Mouldability in Injection Moulding

Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)

Question 3: Weld-Line Defect and Feature Mouldability in Injection Moulding (25 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

(a) Why the weld line forms

A weld line, or knit line, appears wherever two advancing melt fronts meet inside the cavity and fail to fuse completely. There are two ways to create the meeting: the melt stream divides around an obstruction — a core pin forming a hole, a boss, an insert, or a sudden thick-to-thin transition that the flow prefers to run round — and rejoins on the far side; or the cavity is filled from more than one gate and the streams collide somewhere between them. The part in Figure 1 shows the first case, with the line running away from a moulded feature.

The reason the joint is weak is that by the time the two fronts touch, neither is a free liquid surface any more. Injection moulding fills by fountain flow: the hot core of the stream rolls forward, spreads outward at the front and is deposited against the cold cavity wall as a frozen skin. Each front therefore arrives carrying a skin that has cooled well below the melt temperature, has been in contact with air, mould release and outgassed volatiles, and is oxidised. When the two fronts press together, three things prevent a true weld. First, the polymer chains must interdiffuse and re-entangle across the interface, and that requires time above the glass-transition or melt temperature; the fronts are at their coldest and slowest at the moment of meeting, and the pressure that could hold them together is at its lowest because it has been consumed by flow. Second, the air and volatiles trapped at the meeting point have nowhere to go unless the mould is vented exactly there, and a gas cushion physically holds the fronts apart while it burns or compresses. Third, fountain flow orients the molecules, and in a fibre-filled or highly oriented resin the chains and fibres at the front lie parallel to the weld interface rather than crossing it, so almost nothing bridges the joint.

The result is a notch, both geometric and molecular. Visually it is a visible line or groove, sometimes with a colour or gloss difference where pigment and filler orientation change. Mechanically the local strength is typically only 40 to 90 % of the bulk resin for an unfilled polymer, and can fall below 30 % for a glass-filled grade, so the weld line becomes the initiation site for the crack seen in the defective part.

(b) Remedies

Raise the temperature of the meeting fronts. The single most effective change is to increase melt temperature and mould (tool) temperature, which keeps the skins molten longer, allows interdiffusion across the interface and reduces the frozen-layer thickness. Raising injection speed has the same effect for a different reason — less time in the cavity means less heat lost before the fronts meet, and shear heating adds to the melt temperature — and increasing the holding pressure and the packing time presses the interface together while it is still capable of knitting.

Move the weld line rather than remove it. Where the flow must divide, a weld line is unavoidable, so relocate it to a region that is neither cosmetic nor structurally loaded. Repositioning the gate, changing the number of gates, or adding flow leaders and thicker flow channels shifts where the fronts meet. Modern practice is to settle this with mould-filling simulation before the tool is cut, because moving a gate afterwards is expensive.

Give the trapped gas somewhere to go, and the poor material somewhere to end up. Vent the cavity at the predicted knit point, since an unvented weld line will not knit at any temperature. Better still, add an overflow well immediately beyond the weld line so the cold, contaminated front material flows out of the part and is trimmed off, leaving hot melt at the joint.

Reduce what interferes with fusion. Minimise mould release, which is dragged to the front by the fountain flow and forms a barrier film; reduce filler and pigment loading where the specification permits, and be aware that a glass-filled grade will never develop full strength across a weld line; and dry hygroscopic resins properly, since moisture flashes to steam precisely at the front.

Design around it. Where the line cannot be avoided or strengthened enough, thicken the section locally so the reduced strength acts on a larger area, or eliminate the obstruction — for example, drill the hole as a secondary operation instead of moulding it around a core pin.

(c) Which feature is easier to mould

Feature A, the wider slot, is easier to produce.

(d) Why

The decisive point is that neither slot is machined into the moulding — each is produced by a blade of mould steel standing proud of the parting line, and the moulding is simply the space around it. So the question is really which core is easier to build and run, and the governing parameter is the core's depth-to-width ratio. Feature A gives a short, thick core; feature B gives a tall, thin one. Four independent consequences follow, and all of them point the same way.

Core deflection under melt pressure. A core blade is a cantilever loaded by the injection pressure, which is of the order of 50 to 150 MPa. Its tip deflection scales as $PL^4/EI$ with $I$ proportional to the cube of the blade thickness, so halving the width of the blade increases the deflection roughly eightfold at the same depth — and feature B is both thinner and deeper. A deflected core walks off centre, producing a slot with one thick wall and one thin one, out-of-tolerance geometry, and in the worst case a core that touches the opposite cavity face and leaves a hole in the part.

Cooling and cycle time. A thick core has room inside it for a cooling channel or a bubbler, as the sketch shows. A slender blade does not: there is no space to drill, so the only heat path is conduction down the blade into the bolster. The blade therefore runs far hotter than the rest of the cavity, becomes a local hot spot, and the plastic around it is the last material in the part to freeze. That extends the cycle time for the entire part, which is exactly the opposite of what a high production rate needs, and it invites sink marks and dimensional drift around the feature.

Filling and venting. A narrow deep slot is a high-resistance flow path with a large surface-to-volume ratio, so the melt loses heat fast and may hesitate or freeze off before it reaches the blind bottom, giving a short shot. The bottom of a blind slot is also a gas trap, and venting it requires a vent at the very end of the blade — awkward to build and quick to clog. Feature A, being wider, fills at lower pressure with far less risk.

Tool life and ejection. The slender blade is fragile in service, is easily bent or snapped by an ejection mishap or a piece of trapped material, and wears rapidly where the melt scours it. Repairing or replacing it means work on the cavity insert. It also needs draft, and on a deep feature the draft angle noticeably thins the blade towards its base or its tip, worsening every effect above.

The general design rule that this question is testing is to keep moulded ribs, slots and bosses shallow and generously proportioned: hold the core depth to no more than about three times its width where possible, put at least half a degree to one degree of draft per side on the feature, keep rib thickness to roughly 50 to 60 % of the adjacent wall to avoid sink marks, and radius the base of the feature to ease flow and reduce the stress concentration in the moulding.