How to Design Cold Runners: Sizing, Layout and Scrap Rules

Someone sized the main runner from an older tool because it looked close enough. At T1 trial, cavity 4 of the 8-cavity job stopped at 70 percent fill, a 40 mm gap at the far rib, while cavity 1 flashed at the parting line. The runner was the whole story: 6 mm in a layout that needed 8, no cold slug well at the sprue base, and a half-round channel fighting a thick, cold resin. The part never changed. The runner did.

A cold runner is the cheapest part of the mold and the most repeated cost in the shot. It fills every cycle, cools with the part, and leaves the mold as scrap. On a small connector part the runner can outweigh the part. On thin-wall parts the runner freeze time, not the part, sets the cycle length. Typical cold-runner tools generate 5 to 30 percent regrind, by part size.

This guide covers cold runner design for molds that gate at the parting line: system anatomy, cross-section choice, sizing starts, layout patterns, pressure drop drivers, finish and venting, common defects, and scrap rules. It is written for mold designers, DFM reviewers, and buyers who check mold drawings and quotes. Values are industry practice and rules of thumb. Verify each one against the resin datasheet and the machine before steel is cut.

The Snapshot

• A cold runner system is sprue, cold slug well, main runner, branch runners, and gates, and it repeats as scrap on every single cycle.
• Full-round is the pressure baseline. Half-round runs about 30 percent higher pressure drop at equal area, and trapezoidal sits between the two.
• Start runner diameter at part wall plus 1.5-2 mm, then check flow length and viscosity. Never cut steel on the start rule alone.
• Keep the sprue taper at 2-4 degrees per side. Cut cold slug wells at the sprue base and every turn, roughly runner diameter plus 1 mm wide and 1-1.5 times the diameter deep.
• Short, polished, vented, and naturally balanced runners fill evenly. Long, rough, unvented runners cause short shots, hesitation, and cold slug marks.

Cold Runner System Anatomy: Sprue to Gate

Melt travels from the machine nozzle through five elements before it reaches the cavity. The sprue bushing, the sprue itself, the cold slug well, the main runner, and the branch runners lead to the gate. Each element adds pressure drop and scrap, so each one earns its length.

The sprue bushing meets the machine nozzle. The bushing orifice should clear the nozzle tip by roughly 1 mm so no shoulder traps melt between shots. The sprue bore tapers 2-4 degrees per side, a typical practice that lets the frozen sprue pull free when the mold opens. Keep the sprue as short as the cavity layout allows, because every sprue millimeter adds pressure loss, regrind, and cooling time.

At the sprue base sits the cold slug well. The first melt to leave the nozzle each shot is the coldest and the slowest, and that plug would otherwise ride the runner and stall the far cavity. The well traps it and holds it out of the flow path. The same well logic repeats wherever the runner turns.

The main runner runs along the parting line from the sprue toward the cavity field. Branch runners split off to feed each gate. Branches typically step down in diameter toward the far cavities, because flow demand falls as each cavity is served. Stepping also keeps the last cavity from overpacking through an oversized near-side runner.

The gate is the final throttle between runner and cavity. Its depth, land, and location control fill, packing, and weld lines, which is why gate geometry follows its own rules. Our guide to 7 gate design best practices covers that territory in full. For the runner designer, the gate matters as the smallest cross-section in the system, and the runner must deliver enough pressure to the gate to force the melt through it.

Runner Cross-Sections: Full-Round vs Trapezoid vs Half-Round

Three cross-section families dominate cold runner design: full-round, trapezoidal, and half-round. The choice is a pressure, machining, and scrap trade, and it is made once in the drawing and paid for on every cycle.

Full-round is the flow benchmark. A full-round channel has the least wall contact per unit of cross-sectional area, so it delivers the lowest pressure drop per meter and the lowest freeze rate. The cost is machining: each half of the parting line carries a half-round groove, and the two halves must align so the joint does not create a step. Alignment error shows up as drag, streaks, and a weak shear line in the melt.

Half-round is the shallow-plate fallback. One half of the mold carries the full channel, which is easy to cut when cavity plates are thin or when the parting line is offset. The penalty is flow. At equal cross-sectional area, a half-round runs about 30 percent higher pressure drop than full-round, a typical figure that rises with wall contact and faster freeze. Drag and freeze climb together.

Trapezoidal sits between full-round and half-round on pressure drop, typically 10 to 20 percent above the full-round baseline. The channel is cut on one side only, with drafted sidewalls of roughly 10 to 15 degrees per side so the runner strip releases cleanly at ejection. Trapezoids are the standard compromise when the parting line cannot carry a matched full-round pair.

Machining economics push many shops toward trapezoids and half-rounds, because cutting one side costs less than cutting and aligning two. That saving is real once, but the pressure penalty repeats every shot. On long-flow or stiff-resin parts, pay for the full-round. On short, free-flowing parts with deep plates, the half-round is often good enough.

← swipe to scroll →

Table: Relative pressure drop figures are typical industry practice at equal cross-sectional area. Absolute numbers depend on resin viscosity, flow rate, and channel length.

Sizing Rules of Thumb: Diameter, Taper and Cold Slug Wells

Runner diameter starts from part wall. The common rule of thumb is part wall plus 1.5-2 mm for the branch runner feeding a cavity, then adjust for flow length and resin viscosity. A 2 mm wall starts near 3.5-4 mm. A 3 mm wall starts near 4.5-5 mm. Do not size the runner by part volume, because volume is a decoy.

The runner exists to move pressure, not material. A big part with a short flow path needs less runner than a thin part with a long path. Flow length and viscosity move the diameter more than part weight does. Stiff resins like PC, PSU, and glass-filled compounds start at the high end of the range or above it. Free-flowing resins like PP, PE, and POM sit at the low end.

Set a freeze floor. Below roughly 3 mm, a runner freezes off before the cavity finishes packing, which turns a fill problem into a sink-mark problem. Thin-wall tools below 1.5 mm wall still run runners in the 3-4 mm class for that reason. When the flow path is long, step the diameter up 0.5-1 mm per extra flow segment or add gates, and confirm with a flow-length check in the mold design review.

The main runner runs one size class above the largest branch, a typical step of about 1 mm or more. Branches step down toward the far cavities so the near cavities do not overpack through oversized channels. Every step should be smooth, because a sharp step at a branch joint traps melt and sheds cold streaks into the flow.

Cold slug wells follow a simple sizing practice. Diameter is runner diameter plus about 1 mm, and depth is about 1-1.5 times the runner diameter. Cut one at the sprue base and one at every runner turn. A turn is where the leading melt front stalls and cools first, and without a well that cold plug races down the next branch.

← swipe to scroll →

Table: Branch runner starts for typical unfilled resins, based on the wall plus 1.5-2 mm rule of thumb. Verify with a flow-length and pressure budget check before cutting steel.

Check the flow budget before the drawing is final. If the fill analysis shows cavity pressure above 70 to 80 percent of the machine rating, the runner is undersized or the layout is too long. Runners sized by feel pass the drawing review and fail at the press, where the fix costs steel and schedule.

Runner Layout Patterns: H, X and Radial for 2 to 16 Cavities

Layout pattern decides balance before any diameter is chosen. Symmetry is the cheapest balancing tool in mold design, because equal flow paths to every cavity fill evenly by construction. Asymmetric layouts buy balancing work that repeats on every resin change and every process shift.

Two cavities run as a straight split or a simple H from the sprue. Four cavities use an H pattern, where the main runner crosses the center and branches reach two cavities per side. The classic H gives equal path length to all four cavities when the geometry is symmetric.

Eight and sixteen cavities scale the same logic. An 8-cavity tool is often a doubled H, and a 16-cavity layout stacks four branch legs off a central cross. Each leg carries the same runner geometry so every cavity sees the same path. On round or radial parts like gears, rings, and round connectors, a radial spoke layout from a central sprue keeps paths identical and the plate compact.

Natural balance means every cavity sits at the same distance from the sprue, measured along the melt path. H, X, and radial layouts achieve it by construction when cavities are spaced symmetrically. Natural balance survives process changes far better than artificial fixes, because nothing about the geometry has to be tuned.

Artificial balance is the fallback when the part layout cannot be symmetric. The designer varies runner diameter, branch length, or gate size to equalize fill timing across unequal paths. It works on paper and drifts on the floor, because the compensation is tuned at one viscosity and one fill rate. Every resin change reopens the tuning. Our guide to balancing multi-cavity molds walks through the trial method and the numbers.

Verify balance with short shots at trial. Run the mold at 50, 80, and 95 percent fill and compare the flow front across cavities. Fronts that match at all three stages confirm the layout. Fronts that drift apart point at a runner diameter error, an alignment step, or an artificial-balance assumption that did not hold.

Pressure Drop Drivers: Length, Diameter and Viscosity

Runner sizing is really pressure accounting. In laminar pipe flow, pressure drop scales with flow length and viscosity and inversely with the fourth power of diameter. The practical version of that physics is the doubling rule: at the same flow rate, doubling the channel diameter cuts pressure drop by roughly 16 times.

Diameter errors are brutal because of that fourth power. Dropping a main runner from 8 mm to 6 mm at the same flow raises the pressure demand about 3 times, not 25 percent more. A 1 mm machining undersize on a 4 mm branch is proportionally worse than the same error on an 8 mm main. Measure the channel after machining, because cutter wear and hand polishing both move the true diameter.

Length is linear. Double the runner length and you double the pressure demand for the same flow, before shear heating complicates the estimate. Every bend, step, and alignment mismatch adds its own loss on top. That is why short, straight, naturally balanced layouts win: they remove length instead of compensating for it.

Viscosity multiplies the whole budget. Melt stiffness varies widely between resin families, and a stiff resin at half the flow length can demand the same pressure as an easy resin at full length. Glass-filled and high-temperature resins sit at the punishing end. When the resin is stiff, the runner cross-section and layout matter more than the machine's rated pressure.

Two effects soften the math. Shear heating thins pseudoplastic melts at high flow rates, which reduces effective viscosity as fill speed rises. And the gate, not the runner, is often the biggest single pressure loss in the system because it is the smallest cross-section. The runner must still deliver enough pressure to the gate, and the runner is where long thin channels quietly eat the machine's headroom.

Keep a pressure budget on the drawing. Assign fill pressure across sprue, main runner, branches, and gate, and confirm the total fits the machine with margin. If the runner consumes the budget, the cavity runs starved and every defect in the short-shot family appears at trial.

Runner Finish and Venting

Runner surface finish is flow insurance that costs polishing time once. A smooth channel lowers drag, keeps the melt core hot, and releases the runner strip cleanly at ejection. A rough channel adds pressure, freezes the skin faster, and makes the runner stick in the groove.

Watch for the three finish killers. EDM leaves a recast layer on runner walls that drags melt and sheds particles, so polish or grind the runner after EDM. Cutter scallops from rough machining act as thousands of tiny steps. Witness steps at the joint of a full-round pair catch melt and streak the flow, which is why matched half-channels need clean blend at the parting line.

Runner venting follows cavity logic. Air trapped in the runner must escape to the parting line, or it compresses and burns. Vent the dead ends of every runner leg, the cold slug wells, and the far end of the main runner past the last branch. Runner vents run in the same depth class as cavity vents, typically in the low hundredths of a millimeter, and the land stays short and polished.

An unvented runner fails in a signature way: burn marks and short shots at the end of the melt path while the near cavities fill clean. The burn appears dark and localized at the runner dead end, not at the weld line. Vent depth rules vary by resin, and the venting guide on this site lists the ranges by material family.

When the Runner Becomes the Defect

Most runner defects are not runner failures at all. They are the runner exposing a sizing, finish, or venting decision made too early in the program. Read the symptom and it names the fix.

Hesitation appears when melt stalls in a thin section while the thicker runner or cavity region keeps filling. The flow front stops, the skin freezes, and the stall line shows as a weak surface band or a short shot at low fill speed. Runner-driven hesitation follows a runner that is oversized relative to the gate and wall, so the melt prefers the runner until pressure forces it through the thin gate. Step the runner down, raise melt or mold temperature, or move the gate so the path to the thin wall stays short.

Cold slug marks are visual and near the gate. A silver streak, a rough patch, or a tiny cold knob appears where the leading plug of melt entered the cavity. The fix is in the wells: the sprue base and every turn need a cold slug well sized to trap the plug, and the well must be emptied cleanly each cycle. A shallow well lets the plug escape on the next shot.

Unbalanced fill shows up as weight and dimension spread across cavities. The near cavity overpacks and flashes while the far cavity short shots or sinks. Compare cavity weights at trial, because weight spread of 1 to 2 percent points at the runner before it points at the machine. Rebalance the layout or the diameters, then re-run the short-shot series.

Gate freeze timing sits on the runner boundary. If the gate freezes before the cavity finishes packing, the part sinks even though the runner stayed molten. If the runner freezes first, packing stops and the far cavities short shot. Both failures trace to the runner-gate diameter ratio, which is why the runner and the gate are designed as one system, never as two independent features.

Runner Release and Regrind Economics

How the runner leaves the mold sets the automation and the scrap stream. In a two-plate mold, the part and the runner eject together and the runner is degated afterward, by hand, by robot, or at a trim fixture. In a three-plate mold, the runner strip separates automatically when the plates open, which is why three-plate tools dominate pin-point-gated, fully automatic jobs. Our guide to two-plate vs three-plate molds compares the release mechanics and the cycle cost.

Runner scrap is the bill for keeping the melt hot to the gate. A typical cold-runner tool generates 5 to 30 percent regrind, and the percentage tracks part size. A heavy automotive part with a short runner lands near the low end. A small connector or thin-wall part with a long runner sits at the high end, where the runner may outweigh the part several times over.

Regrind reuse is a quality decision, not a cost reflex. Many molders blend regrind at 10 to 25 percent for structural and appearance-tolerant parts, and drop to zero for optical, medical, and property-critical grades. Color shifts, property drift, and contamination all track regrind content, so the blend cap belongs in the process spec, documented and measured like any other parameter.

Right-size the runner against the scrap bill. Oversizing buys fill margin you may not need and pays for it with extra regrind and a longer freeze every cycle for the life of the tool. Undersizing saves grams of scrap and pays with short shots at trial and a narrower process window forever. The economic optimum is the smallest diameter that fills and packs reliably at the target process settings, not the smallest that fills once at maximum pressure.

When the runner bill gets loud enough, the hot runner conversation starts. A hot runner removes regrind, shortens cycles, and opens gate placement, at the cost of a manifold, heat control, and maintenance discipline. The decision logic between the two systems is covered in our hot runner vs cold runner comparison.

FAQ: Cold Runner Design

Q1. What is the best runner cross-section?

Full-round delivers the lowest pressure drop and the slowest freeze, which is why it is the baseline for long-flow and stiff-resin parts. The cost is machining both halves of the parting line and aligning them. Trapezoidal is the standard one-side compromise, and half-round is the shallow-plate fallback at roughly 30 percent higher pressure drop.

Q2. How big should my cold runner be?

Start the branch runner at part wall plus 1.5-2 mm, then adjust for flow length and resin viscosity. Run the main runner about one size class above the largest branch. Keep a freeze floor near 3 mm so the runner stays molten through packing, and confirm the whole budget against machine pressure with a fill check before cutting steel.

Q3. Why do cold runners need cold slug wells?

The first melt to leave the nozzle each shot is the coldest and slowest plug in the system. Without a well at the sprue base, that plug rides the runner and stalls or streaks the far cavity. Every runner turn does the same, so wells go at the sprue base and every turn, at runner diameter plus about 1 mm wide and 1-1.5 times the diameter deep.

Q4. How much regrind is normal from a cold runner?

Typical cold-runner tools produce 5 to 30 percent regrind by part size, with small parts at the high end. Many molders blend regrind at 10 to 25 percent for structural parts and use none for optical or medical grades. Document the blend cap in the process spec and verify part properties at the chosen ratio.

Q5. Should I use a hot runner instead?

Hot runners remove regrind, cut cycle time, and enable automatic degating at high volumes. They add manifold cost, heat control, and maintenance, and they punish heat-sensitive resins. Cold runners stay simpler, cheaper, and easier to change, and they remain the right call for short runs, quick programs, and resins that cannot tolerate residence time.

The Bottom Line

Cold runner design is pressure and scrap accounting done once on the drawing. Start the diameter from part wall plus 1.5-2 mm, prefer full-round where the parting line allows, taper the sprue 2-4 degrees per side, and cut cold slug wells at the sprue and every turn. Lay the cavities out symmetrically so natural balance does the work, keep the runner short and polished, and vent every dead end.

Check the numbers before steel, not after trial. A diameter undersized by 2 mm can multiply the runner pressure demand by 3 times or more at the same flow, and an unvented dead end burns at the exact point the runner was supposed to protect. The tools that fill evenly on day one carry sized, vented, naturally balanced runners and a scrap plan that matches the part value.

DieStrike builds cold- and hot-runner tooling under IATF 16949 across 120+ machines, with cavities hardened to HRC 62 where abrasive flow demands it. Molds ship in 2 to 5 weeks with geometry at ±0.005 mm standard and ±0.002 mm precision, and standard mold components ship in 3 to 7 days. Connector OEMs including TE Connectivity, Amphenol, and Luxshare run programs on the same runner discipline.