Injection Mould Runner and Ejection Design for 3D Printed Prototype Moulds | Precise3D

Why a Steel-Tool Rulebook Fails on a Printed Cavity

Every runner, gate and ejection rule in a standard moulding handbook was derived on hardened tool steel. Substrate strength was never the constraint, so the rules optimise other things: cooling uniformity, cycle time, cosmetic quality, tool life in the hundreds of thousands of shots. Put those rules on a printed cavity and you are optimising the wrong variable. On a printed tool the binding constraint is almost always the stiffness and strength of the cavity material itself, and everything else is secondary.

The numbers are not marginal. A filled tooling resin commonly lands between 2.0 and 3.5 GPa in flexural modulus, and an unfilled one lower still. P1 mould steel sits around 200 GPa. That is a factor of sixty to a hundred, and it shows up at three specific moments in the moulding cycle: when the melt pressure deflects the cavity wall, when the ejector system pushes the part off a core, and when the tool is clamped. Thermal conductivity is the other half of the story, at roughly 3 to 5 W/m·K for a printed tool against about 30 W/m·K for steel, which is what makes the cooling behaviour covered in our conformal cooling guide so valuable in the first place. This page takes the other half of the problem: the internal tooling geometry that decides whether the tool survives the shoots.

The practical consequence is that a printed prototype mould is a low-pressure, low-volume, low-stiffness device, and it should be designed as one from the gating plan outward rather than inheriting a steel plan and hoping. The make-or-buy crossover between printing and moulding in the first place is a separate decision, covered in our print versus mould framework.

Runner and Gate Design, Derived From Printed-Part Weld Lines

The single most important difference between a steel cavity and a printed cavity is not stiffness. It is that the part coming out of a printed mould is anisotropic in a way that a steel-moulded part is not, because the printed tool surface transfers its own texture and because the gate location determines where the flow fronts meet.

Where two flow fronts converge, they form a weld line. In a steel tool with full packing and a homogeneous melt, a weld line typically retains 80 to 95 percent of bulk tensile strength. In a printed-tool scenario the part is usually moulded at reduced pressure with less packing, so a weld line retains closer to 40 to 70 percent of bulk strength. That range is the governing number for gate placement.

The rule that follows is counterintuitive if you learned gating on steel: place the gate so the weld line lands in a thick, lightly loaded region, not in a thin cosmetic or structural face. A gate facing a thin rib, a living hinge, a snap-fit undercut or the sealing face of a bore will put the weakest material exactly where the part needs to be strongest. A gate facing a thick boss lets the weld line form where there is more material to carry it.

Gate type for printed tooling

Edge gates are the default for printed prototype tools because they are cut into the parting plane with a simple taper, they are easy to re-cut after a trial, and they do not require the tool to be split in a way that a printed cavity handles poorly. A direct or sprue gate is acceptable on a single-cavity tool in a thick section and is the most forgiving to cut. Fan gates reduce jetting on a cosmetic face but demand more flow length, which is a real cost at reduced injection pressure. Tunnel and hot-tip gates should generally be avoided on a printed tool: they concentrate heat and wear in a small printed feature, and the detail that makes them work in steel is exactly the detail a printed tool cannot hold.

Sizing the Runner: The Shot-Weight Waste Table

On a production steel tool, runner volume is a rounding error against a million parts. On a fifty-part bridge-tool run, runner volume is a direct cost multiplier on every shot, and it is the number that decides whether the printed tool is economic at all. A runner that is 40 percent of shot weight means the customer is paying for 1.7 times the material they receive.

The tension is that a small runner raises pressure. Because a printed tool tolerates less cavity pressure than steel, there is a floor on runner diameter below which the tool will flash, deflect or fail before it fills. The workable band is narrow, and it is set by part volume rather than by tradition.

Use the ratio of runner cross-section to part volume as the control number, and treat the runner as a pressure budget rather than a waste budget. Where a steel tool might run a 4 mm full-round runner feeding a 60 cm³ part, a printed tool feeding the same part wants a runner sized for the lower available pressure, commonly 5 to 6 mm, at the cost of a higher waste fraction that the fifty-part run simply has to absorb.

The reclaim decision follows from the same table. A 45 percent runner fraction on a 60-part run is worth regrinding if the material tolerates it and the colour does not matter. On a cosmetic part in a coloured engineering resin it is not, and the honest answer is that the runner is scrap and belongs in the cost of the bridge run. Building that into the quote up front is what keeps the per-part economics from collapsing at the invoice.

Ejection Force Budgeting on a Soft Tool

Ejection is where a printed tool dies. The part grips the core with a force proportional to the projected area in contact, the shrinkage, and the coefficient of friction, and the ejector system has to overcome that force by pushing on the part through a printed tool whose own bending stiffness is a fraction of steel. Get the budget wrong and the failure mode is not a sticking part. It is a cracked cavity wall or an ejected boss punched through a thin face.

The governing relationship is simple enough to use at the bench: required ejection force scales with total contact area and with the material's mould shrinkage, and available force scales with the projected area of the ejector pins and the compressive strength of the part at the moment of ejection. In practice, the design rule that matters is to spread ejector contact and never eject off a thin wall. A single ejector pin centred on a 250 mm² boss will punch through a hot PC-ABS part at the same force that four pins around its perimeter will transfer cleanly.

Draft angle derived from printed-layer texture

Draft is the cheapest ejection-force reduction available, and it has to be set against the surface finish of the printed tool rather than against a polished steel reference. An as-printed cavity face carries layer texture that increases friction substantially; a sanded or vapour-smoothed face reduces it. The practical baseline is 0.5 to 1.5 degrees on smooth printed surfaces and an additional 0.5 to 1.0 degree on as-printed surfaces, with texturing or deep layer lines pushing toward the top of the range. Every half degree of draft removed from a deep core multiplies the ejection force, and the multiplication is not linear once friction dominates.

Venting, Because There Is No Pressure Headroom

Steel tools get away with minimal venting because they are run at high pressure, which compresses the trapped gas into the part or drives it into a vent the operator never had to think about. A printed tool runs at roughly a third to a half of that pressure, so the same gas pocket produces a short shot, a burn mark or a diesel effect at the last point to fill, and it does so at a pressure the machine cannot push through.

Venting on a printed tool is therefore designed rather than inherited. The parting-line vent is the primary path: a 0.02 to 0.05 mm deep, 6 to 12 mm wide channel machined or printed across the parting plane at the last point to fill, ending in a relief that is not blocked by the clamp. Deeper than about 0.05 mm and the vent flashes, which is a permanent defect in a printed tool because there is no grinding it out. For deep ribs and blind bosses, add a printed vent pin or a porous insert rather than hoping the trapped volume is small.

The reason this matters more than it does on steel is not just pressure. It is that the printed tool's low thermal conductivity means the last-filled region stays molten longer, so a gas pocket there has a longer window in which to burn the material and scar the tool face. Venting is a thermal problem wearing a pressure problem's clothes.

The Five-Part Trial: What to Measure Before the Second Shot

Most printed protoype tools are commissioned by feel. A far better practice is a fixed five-part sequence in which each shot answers one question, so that a tool that is going to fail fails on part three and not on part forty.

Shot one, short-shot progression. Fill at 40, 60 and 80 percent of the estimated shot size. If the last region to fill is not the region the flow analysis predicted, the gate is in the wrong place and no amount of pressure tuning will fix the weld-line location.

Shot two, part weight against shot weight. Weigh both. The difference is your true scrap, not the runner estimate from the table above. A cached sprue that holds material will show up here as a weight the runner calculation did not predict.

Shot three, dimensional gradient. Measure the same feature at the gate end and at the far end. Printed tools show a wider shrinkage gradient than steel because the cooling is less even. If the gradient exceeds the drawing tolerance, the fix is a cooling change or a printed steel insert at the hot end, not a gate change.

Shot four, ejection signature. Run at the lowest ejection pressure that releases the part and photograph the surface. Drag marks now become a draft-and-finish decision made deliberately rather than a defect discovered at part forty.

Shot five, cycle stability. With no changes, run the same cycle five times and record cycle time. A cycle that lengthens across five identical shots means the tool is storing heat it cannot shed, and the run length is limited by thermal saturation rather than by tool wear. That number tells you whether the printed tool can make fifty parts or can only make twenty before it must rest.

Printed-Mould Failure Modes and Their Signatures

A printed tool fails in a small number of recognisable ways, and each one points at a different design decision. Reading the signature instead of replacing the tool is what separates a bridge tool that paid for itself from one that did not.

Flashing at the parting line. The clamp force exceeded what the printed material can carry, or the vent is too deep. If it is localised to the thickest region, it is cavity deflection under melt pressure and the fix is a steel bolster behind the printed cavity rather than more clamp.

Burn marks at the last point to fill. Insufficient venting, almost always. Add the parting-line vent before touching the shot size.

Short shots that pressure cannot cure. The runner is undersized for the available pressure, or the mould is too cold. Check cycle stability first, because a thermally saturated tool often presents as a short shot.

Cracking at a corner or a thin wall. Ejection force concentrated on too small an area, or a wall that was too thin to print reliably in the first place. This is the one failure mode that is usually a design error rather than a process one.

Wear on the gate after 100 to 400 shots. Expected and manageable. The gate is the highest-velocity, highest-temperature point in the tool, and on a printed tool it is the fastest-wearing feature. A gate designed to be re-cut is a gate that extends the tool's life, which is one more reason the edge gate is the default.

What to Specify When You Order a Printed Prototype Tool

A printed prototype mould is a tooling purchase, and specifying it like one is what keeps the bridge run economic. The specification that follows is short, and every line on it maps to a decision in this article.

Specify the part volume and material grade, because those set the runner and the pressure budget. Specify which face is cosmetic and which is structural, because that is what fixes gate placement and the weld-line target. Specify the expected run length, fifty parts or two thousand, because that decides whether the cavity is printed solid or needs a steel bolster behind it. Specify the ejector-pin layout and the drafted surfaces explicitly rather than leaving them to the toolmaker's default. And specify which dimensions are controlled, so that the trial protocol has something to check against and the printed tool is not measured against a tolerance nobody actually needs.

The printed prototype tool sits between a cast or printed master used for small-batch silicone casting and a fully machined steel tool. Where it wins is in the window where the geometry is final enough to mould but the volume is too low to justify cutting steel, and the design decisions above are what make it win. The rest of the tooling portfolio, from bridge tooling routes through manufacturing tooling and fixture libraries, is grouped in the design guides, and the approval paperwork that a moulded part still needs is covered in the PPAP guide.

← Back to Blog