Temperature Uniformity on a 24-Cavity Valve-Gate Hot Runner Manifold
A naturally balanced runner tree gives every cavity the same flow length. That is necessary on a 24-cavity valve-gate PET preform manifold, but it is not sufficient. Steel temperature still varies from drop to drop, and melt viscosity follows that temperature. Cooler corners fill a little short; a hotter centre fills a little heavy.
What the original layout was doing
OSLON reviewed such a manifold with an in-house three-dimensional thermal-conduction model coupled to a flow network. Every path from the sprue to a gate already had the same length, yet the drop-to-drop steel temperature spread was still 14.6 °C, with cold corners and a hot centre row.
After a set of heat-path changes — not a change to the balanced channel tree — the calculated spread fell to 2.8 °C. Peak steel temperature fell by about 9 °C, and the estimated flow imbalance, including nozzle resistance, was below 1%. These figures are comparative model results. They have not yet been confirmed on a mould trial, and they should be read that way.
Heat was not following the same symmetry as the melt. Tubular heaters in a conventional quadrant layout left the four corners under-heated relative to the mid-plate, where channels, steel mass and heat sources concentrate. Zone thermocouples sitting inward of the outer drops allowed the centre to run above the setpoint while the corners lagged. Valve-bushing contact also leaked heat, unequally, into the cooled plates around the manifold.
The channel tree itself was not the problem. The issue was how heat was added, how it was measured, and how it left the plate.
How the heat paths were changed
The redesign kept uniform-wattage tubular heaters on both faces of the manifold. Segmented-wattage heaters were not used: they are harder to manufacture, and they were not required once the groove route and the heat leaks were corrected.
Heater-groove rerouting
The grooves were shifted so the same uniform heaters put more length around the outer and corner drops, and less at the mid-plate. Left- and right-hand grooves remain mirror images, so the heater types stay interchangeable.
Titanium alloy (TC4) insulator rings at the valve bushings
A titanium ring under each valve-bushing flange raises the thermal resistance of that contact. Less heat is pulled out of the drops, and the remaining loss is more even from cavity to cavity.
Thermocouple relocation
The control thermocouples were moved closer to the outer drops, on the nozzle side of the manifold and between adjacent cavities. The controller then sees the region that had been running cold, instead of holding the centre too high.
Titanium heat-sink pads into the cooled cylinder plate
The mid-plate still collects more heat than the corners. A small number of titanium pads conduct that excess into the water-cooled cylinder plate, so the zone controllers deliver more power to the outer drops. Pad locations were chosen so they do not over-cool a nearby channel.
Removal of non-essential fastener holes
Through-holes that were not required for clamping or location were taken out of the outer row. That improved conduction in the outer steel and brought the spread down from an intermediate 3.9 °C to the final 2.8 °C.
Alternatives that were evaluated and not adopted
A serpentine heater route around the outer row was attractive on paper. Many variants were tried, including a route on one face only and on both faces, with the pads and thermocouples re-tuned. The best serpentine result was a 4.2 °C spread — weaker than the adopted layout, and more demanding to machine. Uniform-wattage heaters following a serpentine path concentrated heat at the mid-outer drops and left the centre row relatively cool. Routing both faces that way made the imbalance worse.
Lightening holes in regions of spare wall thickness produced no meaningful improvement. Unequal power on the two faces of the manifold, with a higher rating on the nozzle side, likewise left the spread essentially unchanged.
On this plate, uniformity improved when heat was redirected and when unnecessary leaks were controlled, not when heater power was simply redistributed by rating.
Results, and how to read them
| Design stage | Spread |
|---|---|
| Original layout | 14.6 °C |
| Heater-groove reroute and insulator rings | 9.6 °C |
| Thermocouples moved toward the outer drops | 8.6 °C |
| Titanium heat-sink pads added | 3.9 °C |
| Final layout (non-essential holes removed) | 2.8 °C |
| Serpentine heater route (evaluated, not adopted) | 4.2 °C |
Peak steel temperature was about 9 °C lower in the adopted layout than in the original. Estimated cavity-to-cavity flow imbalance, including nozzle resistance, fell below 1%.
The analysis used OSLON's in-house three-dimensional thermal-conduction model and a companion flow network. It is not a Moldflow or Moldex3D study. Heat-loss boundary conditions have not been calibrated against measured plate temperatures, so absolute values are not predictions. The useful result is the comparison in the same model, before and after the changes. The layout will be confirmed at trial by measuring temperatures near the corner drops and the centre row.
What this means for a PET preform hot runner
A balanced manifold is a geometric statement. Production consistency is a thermal one. On valve-gate PET preform tools, a few degrees at the drop are enough to move fill, gate vestige and colour. The work above is one example of why a layout that looks symmetrical on a drawing still needs a thermal check.
Discuss a manifold or a PET preform mould with OSLON
To review a valve-gate manifold or a preform mould, send the preform drawing, neck finish, cavity count and machine data to frank@oslonmold.com. We reply to drawings within eight hours.
Hot runner systems for PET preform moulds
PET preform moulds machined in Taizhou
Questions buyers ask
Why is a naturally balanced runner tree not enough on a 24-cavity valve-gate manifold?
Steel temperature still varies from drop to drop, and melt viscosity follows that temperature. Cooler corners fill a little short; a hotter centre fills a little heavy.
What heat-path changes reduced the drop-to-drop steel temperature spread?
The redesign kept uniform-wattage tubular heaters on both faces of the manifold. The heat-path changes were heater-groove rerouting, titanium alloy (TC4) insulator rings at the valve bushings, thermocouple relocation, titanium heat-sink pads into the cooled cylinder plate, and removal of non-essential fastener holes. The balanced channel tree was not changed.
How should the 2.8 °C spread be read?
These figures are comparative model results. They have not yet been confirmed on a mould trial, and they should be read that way. Heat-loss boundary conditions have not been calibrated against measured plate temperatures, so absolute values are not predictions. The useful result is the comparison in the same model, before and after the changes.
Why was a serpentine heater route not adopted?
The best serpentine result was a 4.2 °C spread — weaker than the adopted layout, and more demanding to machine.