Injection Molding Shrinkage and Warpage – Complete Engineering Control Guide
Complete engineering guide to injection molding shrinkage and warpage. How to measure, calculate and control shrinkage and warpage in amorphous and semi-crystalline thermoplastics to ensure dimensional accuracy of injection-moulded parts.
TEDESolutions
Команда експертів
Introduction: Shrinkage and Warpage in Injection Molding
Shrinkage and warpage are the two most consequential physical phenomena in injection moulding, directly determining whether a finished part meets its dimensional specification or becomes scrap. Shrinkage is the volumetric contraction that occurs as a polymer melt cools from its processing temperature and solidifies within the mould cavity. When molten plastic — typically at 200–300 °C depending on the material — is injected into a mould held at 20–100 °C, it undergoes a significant reduction in specific volume as it transitions from a low-density melt to a high-density solid. This contraction is unavoidable and inherent to the physics of thermoplastic polymers; the mould designer's task is not to eliminate it, but to predict it accurately and compensate for it through oversized cavity dimensions.
Warpage is a distinct but related phenomenon: it is the permanent, irreversible out-of-plane or in-plane distortion of the part geometry after ejection from the mould. Whereas shrinkage is a uniform contraction that can be compensated by a simple scaling of cavity dimensions, warpage arises from non-uniform shrinkage — different regions or directions of the part contracting by different amounts. This differential creates internal stress gradients that, once the constraining force of the mould is removed at ejection, cause the part to bow, twist, cup or otherwise deviate from its intended flat or profiled geometry. A flat panel that exits the mould looking like a potato chip is a classic warpage failure.
The industrial significance of these phenomena cannot be overstated. Dimensional rejections attributable to shrinkage miscalculation or warpage are consistently ranked as the number one quality problem in injection moulding production. Modern automotive, medical device, and electronics applications routinely demand dimensional tolerances of ±0.05 to ±0.10 mm on moulded features. A PP automotive clip at 50 mm nominal length with a 1.8% actual shrinkage instead of the designed 1.5% will be 0.15 mm undersized — potentially outside tolerance without any process fault. In a precision gear for an office machine, a warpage of 0.3 mm across a 150 mm diameter can render the entire assembly non-functional. At scale, these failures represent enormous financial losses: rework, scrap, mould modifications, production delays, and warranty claims.
Both phenomena share a common root cause — the thermodynamic and rheological behaviour of molten polymer inside a constrained mould — but they require fundamentally different engineering countermeasures. Shrinkage control is primarily a mould design problem: cavity dimensions must be calculated to compensate for the expected contraction of the chosen material under the intended process conditions. Warpage control is a systems problem that spans material selection, part design, gate location, cooling circuit design, and process parameter optimisation. This engineering guide covers all of these dimensions systematically. It provides the underlying physical mechanisms, comprehensive shrinkage coefficient data tables for over 25 commercial polymers, dimensional compensation formulas with fully worked numerical examples, process parameter optimisation guidance, mould design and cooling system principles, measurement methodology, a comprehensive troubleshooting table, and guidance on CAE simulation tools. Whether you are a mould designer sizing a new tool, a process engineer troubleshooting dimensional drift in production, or a quality engineer establishing a measurement protocol, this guide provides the engineering foundation you need.
Mechanism of Plastic Shrinkage
To control shrinkage effectively, a process engineer must understand the physical mechanisms that produce it. Shrinkage is not a single phenomenon but a superposition of at least two physically distinct mechanisms: thermal contraction and, in semi-crystalline polymers, crystallisation-induced volume reduction. Both operate simultaneously during the cooling phase and contribute to the total linear shrinkage observed in the finished part.
Thermal Contraction
The most fundamental source of shrinkage is thermal contraction: as temperature decreases, molecular motion decreases, intermolecular distances shorten, and the material occupies less volume. This is quantified by the volumetric thermal expansion coefficient (β), which for typical thermoplastic melts is in the range of 5–8 × 10⁻⁴ °C⁻¹, roughly three to five times higher than for metals. The temperature drop experienced by a polymer inside the mould is enormous: melt temperatures for common polymers are typically 200–300 °C, while mould temperatures are typically 20–100 °C, meaning the material must traverse a temperature difference of 150–250 °C in the solid state after passing through the glass transition (amorphous polymers) or crystallisation temperature (semi-crystalline polymers).
The consequence of this large temperature excursion is significant volume reduction. The melt density of common polymers at processing temperature is typically 0.80–1.00 g/cm³, while the solid density at room temperature ranges from 0.90 g/cm³ (low-density polyolefins) to 1.40 g/cm³ (POM, PET, PA66). This density increase — the inverse of volume decrease — directly corresponds to linear shrinkage. For an isotropic, unfilled amorphous polymer, approximately two thirds of the total linear shrinkage arises from thermal contraction between the melt and solid states, making it the dominant mechanism in materials such as ABS, PC, and PMMA.
Thermal contraction is predictable, relatively isotropic (equal in all directions), and well described by published material data. However, it cannot be considered in isolation from the pressure history of the polymer inside the mould cavity, since the specific volume of a polymer is a function of both temperature and pressure — a relationship captured by the Pressure-volume-Temperature (PvT) diagram.
The PvT Diagram and In-Mould Behaviour
The specific volume–temperature (PvT) diagram is the single most important material characterisation tool for understanding and predicting shrinkage. It plots specific volume (cm³/g) against temperature (°C) at multiple constant pressure levels, typically from 0 MPa (ambient) up to 200 MPa (representative of in-mould holding pressures). For any given combination of melt temperature, mould temperature, and holding pressure, the PvT diagram allows the engineer to determine the specific volume change the material will undergo — and thus the expected linear shrinkage.
The key insight from the PvT diagram is that higher holding pressure "shifts" the solidification curve downward, meaning the polymer solidifies at a higher density and hence a smaller specific volume. A part moulded under 100 MPa holding pressure will exhibit significantly less shrinkage than the same part moulded under 50 MPa, because more material has been packed into the cavity to compensate for the thermal contraction. This is why holding pressure is the most powerful single process lever for shrinkage control. For semi-crystalline polymers, the PvT diagram shows a characteristic "kink" at the crystallisation temperature — a discontinuous volume decrease that is absent in amorphous polymer diagrams and that explains why semi-crystalline materials shrink so much more than amorphous ones.
The injection moulding process subjects the polymer to a complex PvT path: injection at high pressure and high temperature, followed by packing (high pressure, decreasing temperature), followed by cooling (low or zero pressure, continuing temperature decrease). The final specific volume — and hence shrinkage — depends on where in the PvT space the material solidifies at the gate seal point, and what happens during subsequent free cooling after gate freeze. Understanding this path is essential for predicting how changes in process parameters will affect shrinkage.
Two-Stage Shrinkage: In-Mould and Post-Mould
Shrinkage occurs in two distinct temporal stages that engineers must account for separately. In-mould shrinkage occurs while the part is still inside the mould cavity and the polymer is cooling from melt to solid. During this stage, the mould walls physically constrain the polymer, and the holding pressure compensates for some of the volume reduction by pushing additional melt into the cavity. In-mould shrinkage is partially — but never fully — compensated by the holding pressure during the packing phase.
Post-mould shrinkage begins at ejection and continues for 24–48 hours after the part leaves the mould. During this period, the part continues to cool from its ejection temperature (typically 50–90% of the glass transition temperature or heat distortion temperature) to room temperature, and semi-crystalline polymers may continue to develop additional crystallinity if their ejection temperature was above the glass transition of the amorphous phase. Post-mould shrinkage is typically 10–30% of total shrinkage for semi-crystalline polymers and 5–15% for amorphous polymers. For dimensional measurements intended to verify compliance with a nominal specification, ASTM D955 specifies that specimens must be conditioned at 23 °C ± 2 °C and 50% relative humidity for 40 hours before measurement, precisely to allow post-mould shrinkage to stabilise.
Amorphous Polymers
In amorphous thermoplastics — including ABS, PC, PMMA, PS, SAN, and rigid PVC — the polymer chains do not form ordered crystalline regions upon cooling. Instead, the material passes through the glass transition temperature (Tg), below which molecular motion is essentially frozen and the material behaves as a rigid solid. Because there is no discontinuous crystallisation event, the shrinkage of amorphous polymers is relatively modest (typically 0.3–0.8%), predominantly thermal in origin, and largely isotropic in unfilled grades. The absence of crystallisation means that the PvT curve for amorphous polymers is a smooth, continuous function without a kink — shrinkage is predictable and consistent across a wide range of process conditions, which makes amorphous polymers considerably easier to mould to tight dimensional tolerances than their semi-crystalline counterparts.
The primary remaining source of dimensional variability in amorphous polymers is residual stress from rapid cooling and high-velocity injection, which can manifest as warpage or post-mould relaxation (creep) over time, particularly if the part is exposed to elevated temperatures in service. For precision optical components such as PMMA lenses or PC light guides, residual stress from processing is a primary design consideration beyond simple shrinkage compensation.
Semi-Crystalline Polymers
Semi-crystalline thermoplastics — including PP, PE, PA6, PA66, POM, PET, PBT, and LCP — have a distinct crystallisation temperature (Tc) below which the polymer chains spontaneously arrange into ordered, tightly packed crystalline lamellae. The formation of these crystalline regions involves a significant additional volume reduction beyond thermal contraction, because crystalline domains are considerably denser than the surrounding amorphous matrix. This crystallisation shrinkage can add 5–15% additional volumetric contraction on top of thermal shrinkage, which is why semi-crystalline polymers exhibit dramatically higher total shrinkage values than amorphous materials of similar chemical backbone. Polypropylene, for example, shrinks 1.5–2.5% linearly, while chemically related amorphous polystyrene shrinks only 0.4–0.7%.
The degree of crystallinity achieved in a moulded part — and hence the magnitude of shrinkage — depends critically on the cooling rate. Rapid cooling (low mould temperature, thin walls) suppresses crystallisation, produces a lower degree of crystallinity, and results in lower shrinkage. Slow cooling (high mould temperature, thick walls) promotes crystallinity and increases shrinkage. This cooling-rate dependence of crystallinity is the fundamental reason why wall thickness and mould temperature have such a large effect on the shrinkage of semi-crystalline polymers, and why thick-section parts in PP or PA are particularly susceptible to dimensional problems. Additionally, the crystallisation process in semi-crystalline polymers may continue slowly at room temperature for hours or days after moulding (post-crystallisation), contributing to post-mould dimensional change and requiring the conditioning protocols specified in ASTM D955.
Types of Shrinkage: Thermal, Crystallization, Differential
For engineering purposes, it is useful to classify shrinkage into three categories that differ in their physical origin, magnitude, predictability, and the engineering countermeasures they require. Understanding which category dominates for a given material and part geometry is the first step in selecting the correct design and process approach.
Thermal Shrinkage
Thermal shrinkage is the component of total shrinkage attributable to the reduction in molecular spacing as temperature decreases — the classical thermal expansion effect operating in reverse. It is characterised by the volumetric coefficient of thermal expansion (CTE) of the polymer in both the melt state and the solid state, and is present to some degree in all thermoplastic materials regardless of whether they are amorphous or semi-crystalline. For amorphous polymers, thermal shrinkage accounts for essentially all of the total observed shrinkage, because there is no crystallisation event to add a discontinuous volume reduction component.
Thermal shrinkage is relatively isotropic in unfilled polymers — it acts approximately equally in all three spatial directions — and it is relatively predictable, because it depends mainly on well-characterised material properties (CTE) and the temperature differential between processing and ambient conditions. The main variability in thermal shrinkage arises from the fact that the effective "solidification temperature" is not a fixed material constant but depends on the pressure history inside the mould cavity: higher holding pressure means the polymer freezes at a higher density, effectively reducing the temperature range over which free thermal contraction occurs after solidification. This is the physical basis for the widely observed relationship between holding pressure and shrinkage.
In practical mould design for amorphous unfilled polymers, thermal shrinkage values from material data sheets are generally reliable to within ±20% of the actual measured value under nominal process conditions, which is sufficient for most industrial tolerance requirements. For precision applications, process trials on the actual mould with production material are always recommended to verify the actual shrinkage before committing to final cavity dimensions.
Crystallisation Shrinkage
Crystallisation shrinkage is the additional volume reduction that occurs specifically in semi-crystalline polymers when polymer chains fold into ordered crystalline lamellae during cooling. Unlike thermal shrinkage, which is a continuous function of temperature, crystallisation shrinkage is a discontinuous event that occurs primarily within a relatively narrow temperature range around the crystallisation temperature (Tc). For polypropylene, Tc is approximately 130–140 °C; for PA66, approximately 250 °C; for POM, approximately 160–175 °C.
The magnitude of crystallisation shrinkage depends on the degree of crystallinity achieved, which in turn depends on the cooling rate, nucleating agents in the formulation, and the molecular weight distribution of the polymer. Typical crystallinity values in injection-moulded parts range from 30–50% for PP, 30–45% for PA66, and 55–75% for POM — significantly lower than the theoretical maximum crystallinity of 70–90%, because rapid cooling during injection moulding suppresses full crystallisation. The crystallisation shrinkage component for PP is approximately 1.5–2.0% linear, which combined with thermal shrinkage gives the total 1.5–2.5% typically observed.
Because crystallisation rate is highly sensitive to mould temperature and cooling rate, the shrinkage of semi-crystalline polymers is inherently more variable and less predictable than for amorphous materials. A change of 10 °C in mould temperature for a PP part can shift shrinkage by 0.3–0.5 percentage points. Furthermore, the spatial variation in cooling rate across a complex part geometry produces spatial variation in the degree of crystallinity — and hence spatial variation in shrinkage — which is a primary contributor to warpage in semi-crystalline moulded components.
Differential Shrinkage and Anisotropy
Differential shrinkage is perhaps the most practically important category for warpage control: it refers to the difference in shrinkage magnitude between different directions within the same part, or between different locations within the same part. When shrinkage is not uniform throughout a part, the regions that shrink more are restrained by the regions that shrink less, generating internal stress that relaxes as geometric distortion after ejection — which is warpage.
The most significant source of differential shrinkage in commercial injection moulding is anisotropy in fibre-reinforced materials. When short glass fibres (GF) or carbon fibres (CF) are incorporated into a polymer matrix, the injection flow aligns the fibres preferentially in the flow direction. In the flow direction, the fibres — which have a CTE approximately ten times lower than the polymer matrix — effectively constrain thermal contraction along their length, dramatically reducing flow-direction shrinkage. In the transverse direction, where fibre reinforcement is minimal, the polymer matrix shrinks freely at its full rate. The ratio of transverse to flow-direction shrinkage in GF-reinforced polymers typically ranges from 2:1 to 4:1. For PA66-GF30, for example, flow-direction shrinkage may be 0.3%, while transverse shrinkage is 0.7–1.2% — a factor of 2.5 to 4.0 difference. This extreme anisotropy means that even if the average shrinkage is correctly compensated in the mould, the differential will inevitably produce warpage unless the part geometry, gate location, and cooling system are all carefully optimised to minimise the stress gradient. CAE simulation with accurate fibre orientation prediction is effectively mandatory for tight-tolerance parts in GF-reinforced materials.
Differential shrinkage also arises from geometric factors in unfilled polymers: thick sections cool more slowly than thin sections, achieving higher crystallinity (in semi-crystalline polymers) and greater total shrinkage; asymmetric cooling (different mould temperatures on core and cavity sides) produces differential shrinkage between the two surfaces; and gate-proximal regions, which receive higher effective packing pressure, shrink less than gate-distal regions. All of these effects produce internal stress gradients that manifest as warpage.
Shrinkage Coefficient Table: 25+ Polymers
The following table provides reference shrinkage values for over 25 commercially important thermoplastic materials. Values are expressed as linear shrinkage percentage under standard injection moulding conditions as defined in ASTM D955 and ISO 294. Note that shrinkage values from material supplier data sheets should always be used in preference to these reference values for a specific commercial grade, as formulation differences (molecular weight, nucleating agents, plasticisers, fillers) can cause significant variation from the tabulated mid-range values.
| Material | Shrinkage [%] | Structure Type | Notes |
|---|---|---|---|
| PP homopolymer | 1.5–2.5 | Semi-crystalline | High crystallinity; mould temperature has large effect |
| PP copolymer | 1.2–2.0 | Semi-crystalline | Lower crystallinity than homopolymer; better impact |
| PP-GF20 | 0.5–1.0 | GF-reinforced | Significant anisotropy; flow < transverse |
| PP-GF30 | 0.4–0.8 | GF-reinforced | High anisotropy; warpage simulation recommended |
| ABS | 0.4–0.7 | Amorphous | Consistent, isotropic; good dimensional stability |
| ABS-GF20 | 0.2–0.4 | GF-reinforced | Reduced shrinkage; moderate anisotropy |
| PC (polycarbonate) | 0.5–0.7 | Amorphous | Low shrinkage; high mould temperature required |
| PC/ABS blend | 0.5–0.7 | Amorphous | Similar to PC; good flow/property balance |
| PA6 (Nylon 6) | 0.8–1.5 | Semi-crystalline | Moisture absorption causes post-mould growth; specify conditioned state |
| PA6-GF30 | 0.3–0.6 | GF-reinforced | Dry-as-moulded values; conditioned dimensions larger |
| PA66 (Nylon 66) | 1.0–2.0 | Semi-crystalline | Higher Tc than PA6; faster crystallisation |
| PA66-GF30 | 0.3–0.7 | GF-reinforced | Flow direction: ~0.3%; transverse direction: ~0.7%; severe anisotropy |
| POM (acetal/Delrin) | 1.8–3.0 | Semi-crystalline | Very high crystallinity; precise mould temperature control essential |
| POM-GF20 | 0.8–1.5 | GF-reinforced | Significant reduction vs. unfilled; anisotropy moderate |
| PMMA (acrylic) | 0.4–0.8 | Amorphous | Optical grade; residual stress affects birefringence |
| PET | 2.0–3.0 | Semi-crystalline | High shrinkage; hot mould required for crystallisation |
| PET-GF30 | 0.5–1.0 | GF-reinforced | Major shrinkage reduction; anisotropy significant |
| PBT (polybutylene terephthalate) | 1.5–2.5 | Semi-crystalline | Faster crystallisation than PET; good electrical properties |
| PBT-GF30 | 0.4–0.8 | GF-reinforced | Widely used in electrical connectors; warpage simulation advised |
| HDPE | 2.0–4.0 | Semi-crystalline | Highest shrinkage among common thermoplastics; slow cooling required |
| LDPE | 2.0–3.0 | Semi-crystalline | Lower crystallinity than HDPE; flexible parts |
| PS (polystyrene) | 0.4–0.7 | Amorphous | Consistent shrinkage; brittle; low mould temperature |
| SAN (styrene-acrylonitrile) | 0.3–0.7 | Amorphous | Improved chemical resistance vs. PS; optical clarity |
| PVC rigid | 0.2–0.5 | Amorphous | Lowest shrinkage among common thermoplastics; heat-sensitive |
| TPE/TPU | 1.0–3.0 | Variable | Wide range depending on hardness and formulation |
| LCP (liquid crystal polymer) | 0.1–1.5 | Semi-crystalline | Highly anisotropic; flow direction may be <0.1%, transverse up to 1.5% |
The data in this table reveals a clear pattern when materials are grouped by structure type. Amorphous polymers — ABS, PC, PC/ABS, PMMA, PS, SAN, rigid PVC — consistently exhibit shrinkage in the range of 0.2–0.8%. This modest shrinkage arises solely from thermal contraction, is relatively isotropic, and is tolerably predictable from published data. Mould designers can typically use the midpoint of the published range as a starting estimate and refine through trial moulding. The most important source of variability for amorphous materials is the effective holding pressure, which shifts the solidification point on the PvT curve.
Semi-crystalline polymers — PP, PA6, PA66, POM, PET, PBT, HDPE, LDPE — exhibit markedly higher shrinkage values, typically 1.0–4.0%, because crystallisation contributes an additional volume reduction beyond thermal contraction. The shrinkage of these materials is more sensitive to process conditions (particularly mould temperature and cooling rate, which control the degree of crystallinity achieved) and is inherently less predictable. For production tooling in semi-crystalline materials, process trials at multiple mould temperatures are strongly recommended before finalising cavity dimensions, as a 10 °C change in mould temperature can shift shrinkage by 0.3–0.5 percentage points in a material like PP or POM.
Glass fibre reinforcement reduces shrinkage by 50–70% across all polymer types, because the rigid fibres mechanically resist the thermal contraction of the surrounding polymer matrix. However, this benefit is accompanied by a serious liability: the fibre alignment produced by injection flow creates dramatic anisotropy between flow-direction and transverse-direction shrinkage, with ratios typically ranging from 2:1 to 4:1 and occasionally as high as 6:1 in LCP materials. This anisotropic shrinkage is the dominant driver of warpage in fibre-reinforced injection-moulded parts, and controlling it requires careful attention to gate location (to manage flow patterns), cooling circuit design, and — for tight-tolerance parts — mandatory CAE simulation with fibre orientation prediction.
Effect of Process Parameters on Shrinkage
The injection moulding process offers the engineer several adjustable parameters that directly influence the shrinkage of the finished part. Understanding the mechanism and magnitude of each effect allows the engineer to use process optimisation as a tool for dimensional correction in production, and to establish appropriate process windows during development.
| Parameter | Effect on Shrinkage | Mechanism | Recommendation |
|---|---|---|---|
| Melt temperature (Tm) | Higher Tm → greater shrinkage | Lower melt density at higher temperature; more volume to contract on cooling | Use mid-range of supplier window; avoid upper limit to control shrinkage |
| Mould temperature (Tf) | Higher Tf → greater shrinkage | Slower cooling → higher crystallinity in semi-crystalline; more free contraction | Control tightly (±2 °C); lower for less shrinkage, balance against surface quality |
| Holding pressure | Higher pressure → lower shrinkage | More material packed into cavity; compensates for thermal contraction | Set at 60–80% of peak injection pressure; most powerful shrinkage lever |
| Hold time | Longer hold → lower shrinkage (until gate seal) | Extended packing compensates more shrinkage; no effect after gate freeze | Verify gate seal by weight method; extend to gate seal + 1 s safety margin |
| Injection speed | Faster injection → lower shrinkage | Higher dynamic pressure; higher effective packing during fill; less gate cooling | Optimise for fill balance; excessive speed causes surface defects |
| Wall thickness | Greater thickness → more shrinkage | Longer cooling time → higher crystallinity; weaker packing pressure gradient | Design uniform wall thickness; use coring to reduce thick sections |
Melt temperature affects shrinkage through its effect on melt density. A polymer melt at 260 °C has a significantly lower density than at 220 °C — the same mass occupies a larger volume, and therefore must contract more to reach the same final solid density. For a typical amorphous polymer, increasing melt temperature by 30 °C increases linear shrinkage by approximately 0.05–0.10 percentage points — a moderate effect. For semi-crystalline polymers, higher melt temperatures also produce a more homogeneous melt with higher crystallisation potential, which can paradoxically increase or decrease shrinkage depending on whether the dominant effect is the melt density change or the change in crystallisation kinetics. In practice, melt temperature is rarely used as a primary shrinkage control tool, because changing melt temperature has broad effects on material properties, cycle time, and surface quality that constrain its use as a single-variable adjustment. The supplier's recommended melt temperature window should be followed, with adjustments at the lower end of the window if shrinkage reduction is needed and other levers have been exhausted.
Mould temperature has a particularly strong effect on the shrinkage of semi-crystalline polymers because it controls the cooling rate, which determines the degree of crystallinity achieved. A higher mould temperature slows cooling, allows more time for crystalline structure development, and increases the final crystallinity — and hence shrinkage. For PP, the shrinkage difference between moulding at 20 °C and 60 °C mould temperature can be 0.5–0.8 percentage points, which is a very large effect for dimensional control. Mould temperature also affects surface quality (higher mould temperature generally gives better gloss and replication of mould surface texture), so there is often a conflict between cosmetic and dimensional requirements that must be resolved by other means such as controlled cooling rate or annealing. For precision dimensional work, mould temperature should be controlled to ±2 °C or better using dedicated Temperature Control Units (TCUs) with flow meters and return temperature monitoring.
Holding pressure is the single most effective process parameter for shrinkage control, and it is also the parameter with the clearest physical mechanism. During the packing phase, the injection unit continues to apply pressure to the screw, forcing additional melt into the cavity to compensate for the volume reduction as the polymer cools and contracts. Higher holding pressure drives more material into the cavity, raising the in-cavity pressure and density, and resulting in a part that exits the mould at a higher density — meaning it has less remaining shrinkage to occur after ejection. The relationship between holding pressure and shrinkage is approximately linear over the typical process window (40–120 MPa), with each 10 MPa increase in holding pressure reducing linear shrinkage by approximately 0.05–0.15% depending on the material and gate geometry. However, excessively high holding pressure creates its own problems: excessive cavity pressure can cause flash on parting lines, overpacking near the gate (leading to part sticking and ejection difficulty), and high residual compressive stress near the gate that can cause warpage or distortion in thin-walled parts. The optimal holding pressure must be determined empirically for each mould.
Hold time is effective only until the gate freezes. Before gate freeze, extending hold time allows the packing phase to continue compensating for thermal contraction as the part cools inside the cavity. After gate freeze, the gate solidifies and no further material can be pushed into or pulled from the cavity regardless of how long hold time is maintained — the part weight stabilises and hold time beyond this point is pure cycle time waste. The gate seal point can be determined by the "weight method": mould parts with progressively longer hold times (e.g., 3s, 5s, 7s, 9s) and weigh each part; when part weight stops increasing, the gate is sealed. Setting hold time to gate seal time plus a 1-second safety margin is standard practice. For thin gates (<1 mm), gate freeze may occur within 1–3 seconds; for thick gates or sprue gates, it may take 10–20 seconds.
Injection speed influences shrinkage through its effect on both dynamic pressure and gate cooling. Faster injection speed creates higher hydrodynamic pressure at the flow front, which effectively contributes to packing — particularly in the early phase of filling before the material has begun to cool significantly. Faster injection also reduces the time available for gate cooling before the switchover to hold phase, effectively delivering more melt into the cavity before the gate begins to restrict flow. The net effect is that faster injection typically produces slightly lower shrinkage for a given set of temperature and pressure conditions. However, injection speed is primarily determined by filling requirements (fill time, weld line positions, surface quality) rather than shrinkage, and the shrinkage effect of speed variation within the practical range is relatively modest compared to the effect of holding pressure changes.
Wall thickness is a design parameter rather than a process parameter, but it has a profound effect on shrinkage, particularly in semi-crystalline materials. Thicker walls cool more slowly — in fact, cooling time scales approximately with the square of wall thickness — allowing more time for crystallisation and hence greater shrinkage. Additionally, the packing pressure gradient from gate to cavity end is steeper in thick sections, meaning that gate-distal regions of thick parts receive less effective packing compensation. The combination of higher crystallinity and less effective packing in thick sections produces significantly higher shrinkage in thick areas relative to thin areas of the same part. This differential shrinkage within the same part is a primary source of warpage in components with non-uniform wall thickness. Good part design practice specifies a uniform wall thickness wherever possible, with transitions between different thicknesses accomplished by gradual tapering rather than abrupt steps.
Warpage: Mechanism and Root Causes
Warpage is the permanent geometric distortion of an injection-moulded part from its intended design geometry, occurring after ejection from the mould. Unlike shrinkage, which is a volumetric phenomenon that reduces the overall size of the part in a predictable way, warpage is a shape change — a flat panel becomes bowed, a rectangular frame twists, a cylindrical housing goes oval. Warpage is the result of internal stress gradients that develop during the moulding process and then relax once the constraining influence of the mould walls is removed at ejection. Understanding the mechanism requires understanding how these internal stresses develop.
The fundamental cause of warpage is non-uniform shrinkage: the fact that different regions of the part, or different directions within the same region, attempt to contract by different amounts as the part cools. If the part could deform freely to accommodate these differential contractions, the result would simply be a shape change with no internal stress. But during cooling inside the mould, the part is constrained by the mould walls, which prevent the regions that want to shrink more from doing so freely. The regions that are constrained from shrinking develop tensile stress (they were prevented from contracting), while adjacent regions that could shrink more freely develop compressive stress. At ejection, these stress gradients are suddenly released, and the part deforms elastically and then plastically in response — which is warpage.
The warpage deflection in a simple flat plate can be estimated using the following engineering formula:
Warpage deflection: δ = ΔS × L / (2 × t)
Where: δ = deflection [mm], ΔS = differential shrinkage [decimal fraction — the difference between shrinkage in two directions or two regions], L = part length in the direction of warpage [mm], t = part wall thickness [mm].
As a worked example: consider a PP flat plate 200 mm long and 3 mm thick. The PP is unreinforced, and due to anisotropic flow patterns, the flow-direction shrinkage is 1.0% (0.010) while the transverse shrinkage is 2.0% (0.020). The differential shrinkage ΔS = 0.020 − 0.010 = 0.010. Applying the formula: δ = 0.010 × 200 / (2 × 3) = 2.0 / 6 = 0.33 mm. This 0.33 mm warpage on a 200 × 3 mm plate may or may not be acceptable depending on the application — it would be unacceptable for a precision instrument cover but acceptable for a disposable consumer product.
Non-Uniform Thermal Contraction
The most fundamental source of non-uniform shrinkage is non-uniform temperature distribution within the part at the point of solidification. If the mould cavity side is cooler than the mould core side — as frequently occurs when cooling circuit design is asymmetric or cooling circuit placement is non-uniform — then the surface adjacent to the cooler mould wall solidifies first and at a lower temperature than the surface adjacent to the warmer wall. The surface adjacent to the warmer wall solidifies later, at a higher temperature, and therefore has a larger temperature drop available for subsequent thermal contraction. The result is that the surface adjacent to the warmer mould wall shrinks more than the opposite surface — a through-thickness differential shrinkage that causes the part to bow, with the concave side facing the warmer mould wall (toward the side with more shrinkage). This is the dominant mechanism of warpage in flat or gently curved parts made from unfilled polymers.
The magnitude of this effect depends on the temperature difference between core and cavity walls, the part thickness, and the material's CTE and elastic modulus. For tight-tolerance flat parts in amorphous polymers, a temperature difference of even 5 °C between core and cavity mould walls can produce measurable warpage. This is why precision-moulded flat parts — display panels, optical covers, precision instrument housings — require careful cooling circuit design to maintain temperature uniformity of ±2 °C across the mould surface.
Molecular and Fibre Orientation
During injection, the polymer melt flows at high velocity from the gate toward the far end of the cavity, with the polymer chains and any dispersed fibres being aligned by the shear flow in the direction of fill. This orientation is partially "frozen in" when the material cools below Tg (amorphous) or Tc (semi-crystalline), because the cooling rate is too fast to allow the chains to randomise fully. The frozen-in orientation creates mechanical anisotropy in the solid part: the mechanical properties — including thermal expansion — are different in the flow direction versus the transverse direction.
In unfilled polymers, the orientation-related anisotropy in shrinkage is moderate — typically <0.5 percentage points difference between flow and transverse directions — and produces proportionally modest warpage. In glass fibre or carbon fibre reinforced grades, however, the fibre orientation anisotropy dominates over the polymer chain orientation effect. As discussed in the shrinkage types section, the ratio of transverse to flow-direction shrinkage in GF30-filled grades can be 2:1 to 4:1, creating enormous internal stress gradients and severe warpage if not controlled. For these materials, warpage is essentially certain without deliberate design measures to mitigate it: centre gating, balanced multi-gate systems, and carefully designed cooling circuits are all necessary tools.
Residual Stress from Rapid Cooling and High Pressure
The third category of internal stress that contributes to warpage is "residual stress" — frozen-in mechanical stress arising from the combination of high injection pressures and rapid cooling. When the polymer near the mould wall solidifies rapidly under high pressure, the molecular chains are frozen in an oriented, high-stress state. These residual stresses can be tensile or compressive depending on location and history, and they contribute to the total stress state driving warpage. Residual stress from processing is distinguishable from differential shrinkage stress because it can be partially relaxed by annealing at elevated temperature without changing part dimensions, while differential shrinkage stress relaxes as dimensional change (warpage). In practice, both contributions are present simultaneously and their effects superpose.
High injection speed and high injection pressure increase residual stress by creating steeper velocity gradients in the melt, which produce more intense chain orientation at the frozen layer adjacent to the mould wall. This is one reason why excessive injection speed can increase warpage even when it reduces average shrinkage. The optimal injection speed for warpage-critical parts is the slowest speed consistent with filling the cavity completely before premature freeze-off — not the fastest speed achievable by the machine.
Factors Contributing to Warpage
Warpage in injection-moulded parts rarely has a single cause. In most production problems, two or more of the following contributing factors act simultaneously. Effective troubleshooting requires identifying all contributing factors and addressing the dominant ones first.
Non-Uniform Cooling
Unequal temperature between the core and cavity mould walls is among the most common and most controllable contributors to warpage. When cooling circuits are positioned closer to one face of the part than the other, or when the core circuit operates at a different temperature than the cavity circuit, a systematic temperature gradient develops across the part thickness. The hotter side solidifies later and contracts more during subsequent free cooling, producing a through-thickness shrinkage differential that causes consistent, reproducible warpage in the same direction for every part. This type of warpage is characterised by a consistent bowing direction across the production run, which helps distinguish it from warpage caused by residual stress (which tends to be more variable).
Fibre Orientation in Reinforced Materials
As detailed above, glass fibre and carbon fibre reinforcement creates strong anisotropy between flow-direction and transverse-direction shrinkage. The severity of this anisotropy depends on the fibre content (more fibres = more anisotropy), the fibre aspect ratio (longer fibres = more anisotropy), the degree of flow alignment (which depends on gate location, fill speed, and part geometry), and the stiffness of the part (stiffer parts resist warpage from internal stress more effectively). For GF-reinforced parts with tight dimensional requirements, warpage prediction by CAE simulation with fibre orientation analysis is effectively mandatory before mould fabrication.
Gate Location and Size
The position of the gate fundamentally determines the flow pattern within the cavity, which in turn determines both the fibre orientation distribution and the residual stress pattern. An off-centre gate on a symmetric part creates an asymmetric flow front that produces different levels of orientation and residual stress on the two halves of the part, inevitably resulting in warpage. A gate that is too small creates excessive shear at the gate, which intensifies orientation effects immediately downstream and can create a high-stress zone near the gate. For flat parts, a fan gate or film gate that delivers melt across a wide front produces the most uniform flow pattern and typically the least warpage. For parts with complex geometry, gate location optimisation using CAE flow simulation is the most reliable approach.
Wall Thickness Variation
Abrupt changes in wall thickness are a particularly problematic source of warpage because they create local shrinkage differentials at the transition zone. A 5 mm thick section shrinks significantly more than an adjacent 2 mm section, and the constraint between them creates high stress at the junction that relaxes as warpage or cracking. Good part design practice specifies gradual thickness transitions — typically a taper ratio no steeper than 3:1 over a transition length of at least 3× the thickness difference — and avoids abrupt steps wherever possible. Where thick bosses or ribs are unavoidable, their base width should be limited to 60% of the adjacent wall thickness to minimise sink marks and differential shrinkage.
Residual Stress from Process
Excessive injection speed, very high injection pressure, and insufficient cooling time all increase the residual stress frozen into the moulded part. As discussed above, these stresses add to the stress state driving warpage and can convert a borderline acceptable part into one with unacceptable distortion. Process optimisation for warpage-critical parts typically involves deliberately slower injection speeds (longer fill times) and moderately lower injection pressures than the machine can deliver, compensating for the lower fill pressure with optimised gate design that reduces flow resistance.
Asymmetric Mould Temperature
Asymmetric mould temperature — different temperatures on the core and cavity halves — is a design and maintenance issue that consistently produces systematic warpage. It arises from poorly balanced cooling circuit design (too few or undersized cooling channels on one side), blocked cooling channels due to scale or contamination buildup, TCU malfunction, or deliberate but poorly calibrated attempts to use differential temperature to compensate for other warpage sources. Maintaining mould temperature balance to within ±5 °C between core and cavity is a basic requirement for warpage-sensitive parts; ±2 °C is required for precision applications.
| Factor | Warpage Mechanism | Solution |
|---|---|---|
| Unequal core/cavity temperature | Differential through-thickness shrinkage → bowing toward hot side | Balance cooling circuits; use separate TCUs for core and cavity |
| Glass fibre orientation | Flow shrinkage < transverse shrinkage; 2:1 to 4:1 ratio | Centre gating; CAE flow simulation; consider fewer fibres or different gate |
| Off-centre gate on symmetric part | Asymmetric flow pattern → asymmetric residual stress → twisting | Relocate gate to centre of part; use multi-gate with balanced runners |
| Abrupt wall thickness change | Thick section shrinks more; stress at transition relaxes as distortion | Taper transitions; core out thick sections; maintain uniform wall thickness |
| Insufficient cooling time | Part ejected above heat distortion temperature; warps freely | Extend cooling phase; verify ejection temperature with thermocouple |
| Excessive injection speed | High shear creates intense orientation and residual stress | Reduce injection speed; optimise gate cross-section to maintain fill speed |
| High melt temperature | Higher thermal contraction; more residual stress from flow | Use lower end of supplier temperature window |
| Low holding pressure | Insufficient packing → high shrinkage differential gate vs. far end | Increase holding pressure; verify screw cushion remains positive |
| Blocked cooling channel | Local hot spot → local high shrinkage → local distortion | Flush and clean cooling circuits; inspect with borescope |
| Unbalanced runner system | Cavity-to-cavity fill pressure difference → different shrinkage per cavity | Balance runner system geometrically; use hot runner with individual valve control |
| Insufficient draft angle | Ejection force deforms part during extraction | Increase draft to minimum 1.5° per side; polish mould in draw direction |
| Unbalanced ejection system | Non-uniform ejection force bends part during extraction | Add ejector pins to balance ejection force; use stripper plate for flat parts |
| Residual crystallisation post-ejection | Continued crystal growth after moulding → dimensional change | Extend cooling time; increase mould temperature to complete crystallisation in-mould |
| Moisture in hygroscopic material | Inadequate drying → steam voids → irregular density → warpage | Dry material to specification before moulding; verify with moisture analyser |
| Improper material selection | High-shrinkage semi-crystalline material with complex geometry | Evaluate amorphous alternative; add GF reinforcement; redesign part geometry |
Mold Design for Shrinkage Control
Mould design is the most fundamental engineering lever for shrinkage control: the cavity dimensions must be deliberately sized larger than the nominal part dimensions by the expected shrinkage percentage, so that as the part contracts after ejection, it reaches exactly the target nominal dimension. Getting this compensation right requires accurate shrinkage data, correct application of compensation formulas, and clear understanding of which dimensions are flow-direction and which are transverse in reinforced materials.
Cavity Dimensions — Shrinkage Compensation Formula
The fundamental formula for mould cavity dimension calculation is:
Dw = Dz × (1 + S/100)
Where: Dw = mould cavity dimension [mm] (the dimension to be machined into the tool), Dz = nominal part dimension [mm] (the dimension required on the finished part), S = material linear shrinkage [%] (from material data sheet or measurement).
Worked Example 1 — Amorphous polymer, ABS housing cover:
Target part dimension: Dz = 50.000 mm. Material: ABS, grade shrinkage from data sheet: 0.4–0.7%, midpoint = 0.55%. Mould cavity dimension: Dw = 50.000 × (1 + 0.55/100) = 50.000 × 1.0055 = 50.275 mm. The cavity should be machined to 50.275 mm. If the actual ABS shrinkage under production conditions is at the upper end (0.7%), the finished part will measure: 50.275 / 1.007 = 49.926 mm — 0.074 mm undersized. If at the lower end (0.4%), the part will measure: 50.275 / 1.004 = 50.075 mm — 0.075 mm oversized. The total variation window from shrinkage uncertainty alone is therefore ±0.075 mm relative to nominal. This must be considered against the part tolerance. If the tolerance is ±0.20 mm, the mould has adequate margin for the uncertainty in shrinkage, and the remaining tolerance budget is available for dimensional scatter from process variation.
Worked Example 2 — Fibre-reinforced semi-crystalline polymer, PA66-GF30 structural bracket:
Target part dimensions: 100.000 mm in the flow direction and 80.000 mm in the transverse direction. Material: PA66-GF30, data sheet shrinkage: 0.5% flow, 1.2% transverse. Flow-direction cavity: Dw_flow = 100.000 × (1 + 0.5/100) = 100.000 × 1.005 = 100.500 mm. Transverse-direction cavity: Dw_trans = 80.000 × (1 + 1.2/100) = 80.000 × 1.012 = 80.960 mm. The differential shrinkage = 0.012 − 0.005 = 0.007. Applied to the 80 mm transverse dimension, this represents 80 × 0.007 = 0.56 mm of anisotropic shrinkage that will produce warpage unless counteracted by part stiffness or cooling asymmetry. For this material-geometry combination, CAE simulation of warpage is strongly recommended before confirming mould cavity dimensions.
In practice, mould cavity dimensions for production tooling in demanding applications are initially set to the calculated Dw value, then adjusted after first-article measurement using the actual shrinkage observed on the production press with production material. This empirical correction is called "mould tuning" or "steelwork correction" and is considered normal practice for tight-tolerance tooling. Mould designers typically specify cavity dimensions to a tighter machining tolerance than required by the part (typically +0/−0.01 mm for critical cavities machined by die-sinking EDM or high-speed milling), so that if the actual shrinkage is slightly higher than predicted, material can be removed from the cavity to increase its dimension, while if shrinkage is lower, cavity insert material can be added by welding or replaced.
Gate Placement for Warpage Reduction
Gate location is the single most consequential design decision for warpage in fibre-reinforced or large flat parts. The gate determines the flow pattern, which determines the fibre orientation distribution, which determines the anisotropic shrinkage pattern, which determines the warpage. For symmetric flat parts (rectangular or circular covers and panels), a centre gate produces a radially symmetric flow pattern, which distributes fibre orientation more uniformly in all directions and reduces the anisotropic shrinkage differential compared to an edge gate. While a perfectly centre-gated flat disc will still exhibit some radial-transverse shrinkage anisotropy, the distribution is axially symmetric and produces a uniform radial bowing rather than saddle-type or twisting warpage.
For long rectangular parts where a single central gate cannot adequately fill the cavity without excessive pressure drop, a fan gate or film gate spanning the short width of the part produces a uniform flow front advancing along the length — similar to centre-gating but in one dimension. This gate type produces highly aligned fibres in the length direction (high anisotropy) but the alignment is uniform and predictable, making warpage compensation through asymmetric cooling more straightforward. Multiple pin gates or valve gates evenly distributed along a long part can also reduce flow length and hence flow-induced anisotropy, at the cost of weld lines between flow fronts.
Draft Angles and Ejection System Design
Insufficient draft angle on mould walls parallel to the draw direction — the direction of mould opening — creates significant ejection forces as the part is extracted from the mould. These forces can easily exceed the part's strength at its ejection temperature (which is significantly lower than room-temperature strength), causing localised bending or permanent deformation that manifests as warpage on the ejected part. Minimum draft angles of 1–2° per side are required for most thermoplastics; textured surfaces may require 3–5° per side depending on texture depth. Engineering-grade amorphous polymers like PC and ABS, which have relatively high stiffness at ejection temperature, are more forgiving of marginal draft than semi-crystalline materials like PP and PE, which have very low stiffness near their crystalline melting point and can deform dramatically under even moderate ejection force.
Ejection system layout must ensure that the ejection force is distributed uniformly across the part base, proportional to the local projected area. Concentrating ejection force at a few points — typical of designs with only four corner ejector pins on a large panel — inevitably produces local bending that appears as warpage. For large flat parts, continuous stripper plates or edge-ring ejection systems are preferred over point ejection, as they distribute force across the entire part periphery simultaneously. For parts with ribs, ejector pins should be placed at the end of ribs where the material is thickest and most resistant to bending, and the pin diameter should be maximised within the rib space to spread the force over a larger area.
Cooling System Design for Warpage Prevention
The cooling system — the network of water channels machined through the mould plates — is the most powerful engineering tool for warpage control. Its design determines the temperature distribution across the mould cavity surface, which controls the through-thickness thermal gradient in the part, which directly drives warpage through differential shrinkage. A well-designed cooling system is as important to part quality as the cavity geometry itself, yet it is frequently treated as an afterthought in mould design, with channels placed wherever space is available rather than where they are needed to achieve thermal balance.
Thermal balance principle: For warpage control in flat or gently curved parts, the temperature difference between the core side and the cavity side of the mould must be maintained at <5 °C across the entire mould surface during steady-state production. For precision parts with tight warpage specifications, the requirement is <3 °C. Achieving this level of thermal balance requires careful positioning of cooling channels at equal distances from the mould surface on both core and cavity halves, using equal channel diameters and equal coolant flow rates, and verifying balance with infrared thermal imaging of the mould surface during production.
Cooling time formula: The minimum cooling time required before safe ejection is given by the following engineering formula:
tc = (s² / (π² × α)) × ln((4/π) × (Tm − Tf) / (Te − Tf))
Where: tc = cooling time [s], s = part wall thickness [mm], α = thermal diffusivity of the polymer [mm²/s], Tm = melt temperature at injection [°C], Tf = mould temperature (average of core and cavity) [°C], Te = ejection temperature [°C] (typically 10–20 °C below HDT/Vicat).
As a worked example with PP: wall thickness s = 3 mm, thermal diffusivity α = 0.08 mm²/s (typical for PP), Tm = 230 °C, Tf = 30 °C, Te = 90 °C (PP HDT approximately 100 °C, so Te = 90 °C is conservative). Calculation: tc = (3² / (π² × 0.08)) × ln((4/π) × (230 − 30) / (90 − 30)) = (9 / 0.789) × ln((1.273) × (200/60)) = 11.41 × ln(4.24) = 11.41 × 1.444 = 16.5 s. Rounding up for a practical cooling time: approximately 18 seconds. This agrees well with typical production experience for 3 mm PP at 230 °C melt with a 30 °C mould.
Conformal cooling is a major advance in cooling system design enabled by additive manufacturing (3D printing of mould inserts in tool steel or H13 using selective laser melting). Conventional straight-drilled cooling channels must follow straight-line paths and cannot closely follow the contour of complex curved cavity surfaces, leaving some regions far from any cooling channel and therefore poorly cooled. Conformal cooling channels — printed with complex 3D geometry that closely follows the mould surface contour at a constant stand-off distance — achieve surface temperature uniformity of <2 °C compared to typically >15 °C with equivalent conventional drilling. The practical result of this improved thermal uniformity is a reduction in warpage of 40–60% in complex parts, as well as cycle time reductions of 20–40% from more effective cooling. The additional cost of 3D-printed inserts (typically 2–4× the cost of conventional machining for the insert alone) is justified for high-volume precision applications where warpage failures are currently causing significant scrap or rework cost.
TCU requirements: For warpage-critical moulding, separate Temperature Control Units (TCUs) should be used for the core and cavity circuits rather than manifolding both through a single TCU. This allows the engineer to deliberately set different temperatures on core and cavity sides — which may be necessary to compensate for the inherent heat imbalance caused by the runner system delivering hot melt to the cavity side. Independent TCUs also allow temperature adjustments during production to correct for warpage without disrupting the entire cooling system. Each TCU circuit should include a flow meter to verify adequate coolant velocity (minimum 0.5 m/s for turbulent flow) and a return temperature sensor to monitor the circuit's heat removal rate.
| Material | Recommended Mould Temperature [°C] | Notes |
|---|---|---|
| PP (homo/copolymer) | 30–50 | Higher for better crystallinity and surface gloss; lower for faster cycle and less shrinkage |
| ABS | 40–70 | Higher for better surface finish; lower end for faster cycles |
| PC (polycarbonate) | 80–100 | High mould temperature essential to avoid surface stress and residual stress birefringence |
| PA6 | 60–80 | Higher promotes crystallinity; note moisture effect on final dimensions |
| PA66 | 70–90 | Higher mould temperature recommended for structural grades |
| POM | 50–90 | Higher temperature gives higher crystallinity and lower post-mould shrinkage variation |
| PBT | 50–80 | Higher temperature for better crystallinity; important for surface gloss |
| PET | 70–100 | Hot mould required for adequate crystallisation in structural applications |
| PC/ABS | 60–80 | Balance between PC requirements and ABS surface quality |
| PMMA | 50–70 | Higher temperature reduces residual stress and birefringence in optical parts |
Holding Pressure and Gate Sealing
The packing phase — specifically the holding pressure applied during it — is the most powerful and most immediately accessible lever the process engineer has for controlling part shrinkage in production. Every modern injection moulding machine allows the holding pressure and hold time to be set independently of injection pressure and injection speed, and adjustments take effect immediately from the next shot. Understanding the physics of holding pressure and gate sealing is therefore not merely academic — it is the day-to-day working knowledge of effective process engineering.
During the packing phase, the injection unit applies continued pressure through the screw to the melt remaining in the barrel and runner system. This pressure is transmitted through the melt to the material filling the cavity, effectively "pumping" additional material into the cavity to compensate for the volume reduction as the polymer cools and densifies. The higher the holding pressure, the more material is pushed in, the higher the in-cavity density, and the lower the final shrinkage of the part. In practice, holding pressure is typically set in the range of 60–80% of peak injection pressure. Values below 50% of injection pressure tend to produce excessive shrinkage; values above 90% risk flash formation on the parting line or core sticking due to excessive cavity pressure.
Gate freeze/seal point: The effectiveness of holding pressure is limited to the period before the gate solidifies. The gate — which is intentionally the thinnest point of the runner system — cools and solidifies before the thick sections of the part. Once the gate is frozen solid, no further material can be pushed into or pulled from the cavity regardless of how much pressure is applied at the screw. The part then undergoes "free shrinkage" — cooling and contracting with no further compensation from the machine. The gate seal point is therefore the critical time limit for effective packing, and all subsequent hold time beyond this point is pure cycle time waste that adds no quality benefit.
Gate seal verification — weight method: The most reliable practical method for determining gate seal time is the "weight method" or "gate seal test." A series of parts are moulded with progressively increasing hold times (for example, 2 s, 4 s, 6 s, 8 s, 10 s) at a fixed holding pressure, with all other parameters constant. Each part is weighed to ±0.01 g precision on an analytical balance. As hold time increases from very short values, part weight increases because more material is packed into the cavity. At some hold time, the part weight stabilises — further increases in hold time produce no additional weight increase — because the gate has frozen and no more material can enter. The hold time at which weight stabilises is the gate seal time. The recommended production hold time is gate seal time plus 1–2 seconds safety margin. This simple test should be repeated whenever gate geometry, material grade, or process temperatures are changed.
Screw cushion: A critical prerequisite for effective holding pressure is maintaining a positive screw cushion — the remaining volume of melt in front of the screw tip at the end of the holding phase. If the screw travels to its forward mechanical stop before the holding phase ends, the hydraulic system can no longer transmit pressure to the cavity because there is no melt "spring" to compress. A minimum cushion of 5–10 mm (typically 5–10% of shot size) must remain throughout the holding phase. Cushion is controlled by setting the shot size (screw back position) appropriately. Inconsistent cushion from shot to shot — indicated by variability in screw end position — suggests a problem with the non-return (check) valve, which should be inspected and potentially replaced.
Multi-stage holding pressure profiles: Rather than applying a constant holding pressure throughout the packing phase, modern injection moulding machines allow multi-stage pressure profiles — for example, a profile that starts at 90 MPa for the first 2 seconds, steps down to 70 MPa for the next 2 seconds, then drops to 50 MPa for the final 2 seconds before gate seal. This stepped profile has two advantages over constant pressure: first, it reduces the risk of flash formation by starting at the high pressure needed to prevent immediate shrinkage near the gate, then reducing pressure as the gate cools and seals; second, it reduces residual stress in the part near the gate by limiting the peak pressure applied to that region, which would otherwise produce compressive stress that drives warpage upon ejection. For warpage-critical parts, a step-down profile should be tried as a process improvement before resorting to mould modifications.
| Material | Typical Holding Pressure [MPa] | Typical Hold Time [s] | Notes |
|---|---|---|---|
| ABS | 50–80 | 5–15 | Moderate; gate seal at 5–10 s for typical gate sizes |
| PC | 60–100 | 8–20 | Higher pressure needed; high melt viscosity |
| PP (unfilled) | 40–70 | 5–15 | Lower viscosity; gate seals relatively quickly |
| PP-GF30 | 60–90 | 5–15 | Higher viscosity due to fibre; similar gate seal time |
| PA6/PA66 (dry) | 50–90 | 5–20 | Low viscosity when dry; gate seals moderately fast |
| PA66-GF30 | 70–110 | 8–20 | High pressure for dense GF material |
| POM | 60–100 | 5–15 | Fast crystallisation; gate seals quickly for thin gates |
| PC/ABS | 50–90 | 8–18 | Similar to PC; requires consistent temperature control |
Shrinkage Calculations: Formulas and Worked Examples
This section provides three fully worked engineering examples that illustrate the complete process of shrinkage calculation: from material selection and data sheet interpretation, through cavity dimension calculation, to tolerance analysis and engineering judgement about achievability. These examples are designed to be directly applicable to real mould design problems.
Example 1 — Amorphous Polymer: ABS Housing Cover
A new ABS housing cover for a consumer electronics product requires a nominally 80.000 mm long body. The part tolerance per the engineering drawing is ±0.20 mm (i.e., the part must measure between 79.800 mm and 80.200 mm in dry-as-moulded condition). Material: commercial ABS, data sheet linear shrinkage 0.4–0.7%.
Step 1 — Select design shrinkage value: For a standard ABS with a relatively well-established shrinkage range, the midpoint of the data sheet range (0.55%) is used for the initial cavity calculation. This places the expected part dimension at the centre of the range, leaving approximately equal margin for actual shrinkage being higher or lower than predicted. Cavity dimension: Dw = 80.000 × (1 + 0.55/100) = 80.000 × 1.0055 = 80.440 mm.
Step 2 — Tolerance analysis: If actual ABS shrinkage is at the lower data sheet limit (0.4%), actual part dimension = 80.440 / 1.004 = 80.12 mm (+0.12 mm from nominal). If actual shrinkage is at the upper limit (0.7%), actual part dimension = 80.440 / 1.007 = 79.88 mm (−0.12 mm from nominal). The total shrinkage-related dimensional spread across the data sheet range is therefore ±0.12 mm relative to nominal. Since the part tolerance is ±0.20 mm, the shrinkage uncertainty alone consumes 60% of the total tolerance budget, leaving only ±0.08 mm for process variation (shot-to-shot repeatability, temperature variation, etc.).
Step 3 — Mould machining accuracy requirement: The mould cavity must be machined to an accuracy consistent with the tolerance budget. If the process variation (shot-to-shot) consumes an additional ±0.05 mm, total allocated = ±0.12 + ±0.05 = ±0.17 mm out of ±0.20 mm available — leaving only ±0.03 mm margin. This is tight, and requires careful process control. Machining accuracy of the cavity to ±0.01 mm (achievable with CNC die-sinking EDM or high-speed milling with in-process measurement) ensures that machining error does not consume additional tolerance budget. The conclusion is that this part is achievable within tolerance with a good production process, but should not be considered to have generous tolerance margin.
Example 2 — Semi-Crystalline Fibre-Reinforced Polymer: PA66-GF30 Structural Bracket
A structural bracket for an automotive engine compartment application has nominal dimensions of 100.000 mm in the primary flow direction and 80.000 mm in the transverse direction. Material: PA66-GF30. Data sheet shrinkage: 0.5% in flow direction, 1.2% in transverse direction (dry-as-moulded). Part tolerance: ±0.30 mm on both dimensions.
Step 1 — Cavity dimension calculation (flow direction): Dw_flow = 100.000 × (1 + 0.5/100) = 100.000 × 1.005 = 100.500 mm.
Step 2 — Cavity dimension calculation (transverse direction): Dw_trans = 80.000 × (1 + 1.2/100) = 80.000 × 1.012 = 80.960 mm.
Step 3 — Differential shrinkage and warpage risk assessment: The differential shrinkage ΔS = 0.012 − 0.005 = 0.007 (0.7 percentage points). Applied to the 80 mm transverse dimension: expected differential deformation = 80 × 0.007 = 0.56 mm. Applying the warpage formula with L = 80 mm and t = 3 mm (assumed wall thickness): δ = 0.007 × 80 / (2 × 3) = 0.56 / 6 = 0.093 mm. This is within the ±0.30 mm tolerance. However, this is a simple beam-bending approximation, and the actual warpage will depend strongly on the part geometry (ribs, bosses, mounting flanges all affect stiffness) and the actual fibre orientation achieved with the selected gate location. CAE simulation is strongly recommended to verify warpage before committing to the mould design. If simulation predicts warpage exceeding 0.2 mm, gate relocation or changes to cooling circuit design should be evaluated in simulation before cutting steel.
Step 4 — Moisture conditioning note: PA66-GF30 absorbs atmospheric moisture after moulding, with equilibrium moisture content of approximately 1.8% at 50% RH. This moisture absorption causes dimensional growth of approximately 0.3–0.5% on the PA66 matrix, which partially offsets moulding shrinkage. Dimensions should be measured in a defined moisture condition (either dry-as-moulded or conditioned to equilibrium) and the mould cavity calculations should correspond to the same condition. For automotive engine compartment components exposed to high humidity, the final service dimensions (conditioned) are typically the engineering requirement, and the mould may need to be calculated using a net effective shrinkage that accounts for moisture expansion.
Example 3 — Post-Mould Shrinkage and Moisture Effects: PA6 Gear Housing
A PA6 gear housing is moulded to a nominal dimension of 100.000 mm. The designer needs to understand the expected dimensional range across different moisture conditions during the service life of the part.
Moulding shrinkage: PA6 unfilled data sheet shrinkage = 0.8–1.5%, midpoint = 1.15%. Mould cavity: 100.000 × 1.0115 = 101.150 mm. Expected dry-as-moulded dimension: 100.000 mm (by definition, if mould is correctly sized).
Moisture absorption: PA6 absorbs moisture at a rate that depends on relative humidity and temperature. At 50% RH (standard ambient), PA6 reaches equilibrium moisture content of approximately 2.5–3.0% by weight. Each 1% moisture content absorbed causes linear dimensional growth of approximately 0.3% in the part. At 2.5% moisture, dimensional growth = 2.5 × 0.3% = 0.75%. Applied to the 100 mm dimension: conditioned dimension ≈ 100.000 + 100 × 0.0075 = 100.750 mm.
At saturation (immersed in water), PA6 can absorb up to 9–10% moisture. Dimensional growth = 9 × 0.3% = 2.7%, giving a saturated dimension of approximately 102.7 mm — 2.7 mm larger than the dry-as-moulded dimension. This is an enormous dimensional change for a precision component.
Engineering recommendation: For any PA6 (or PA66) component with dimensional requirements, the engineering drawings must explicitly specify the dimensional verification condition: "dry-as-moulded (DAM)" or "conditioned to 50% RH at 23 °C." If the component is used in a humid environment or in contact with water, the design must account for moisture-induced dimensional growth as a separate design variable. In some applications, moisture-conditioned dimensions — not dry-as-moulded — are the relevant design state, and the mould cavity calculations should use a reduced effective shrinkage that anticipates the post-mould moisture expansion. This is a critical detail that is frequently overlooked in mould procurement specifications, leading to expensive mid-project corrections when first-article inspection is conducted in a different moisture condition than the production baseline.
Measurement Methods and Quality Control
Accurate measurement of shrinkage and warpage is essential for validating mould cavity dimensions, qualifying a new moulding process, and maintaining dimensional conformance in production. The choice of measurement method depends on the required accuracy, the part size and geometry, the volume of parts to be measured, and the budget available for measurement equipment.
Timing of measurement — ASTM D955 protocol: Before discussing measurement methods, the most critical practical requirement must be stated: injection-moulded parts must not be measured immediately after ejection. Post-mould shrinkage continues for 24–48 hours, and hygroscopic materials (PA, PA66) begin absorbing atmospheric moisture immediately after ejection. ASTM D955 specifies that specimens for shrinkage measurement must be conditioned at 23 °C ± 2 °C and 50% relative humidity ± 5% for a minimum of 40 hours before measurement. ISO 294-4 specifies similar requirements. Any measurement taken before this conditioning period is not comparable to nominal specifications based on these standards and may significantly overestimate actual shrinkage.
CMM (Coordinate Measuring Machine): CMM measurement is the gold standard for dimensional verification of moulded parts with discrete feature dimensions (lengths, diameters, hole positions, parallelism, flatness). Modern contact-probe CMMs achieve measurement repeatability of ±0.001 mm (1 micron), which is well below the typical tolerance requirements for injection moulding. CMM is the preferred method for formal first-article inspection and for drawing-callout verification when specific GD&T tolerances must be certified. Limitations of CMM are its relatively slow throughput (a comprehensive measurement of a complex automotive part may take 30–60 minutes with a manual probe) and its inability to capture full-surface warpage maps — it measures only at programmed probe points, not across the entire surface.
Optical measurement and structured light scanning: Full-field optical measurement systems — such as the GOM ATOS (Automated Topometric Sensor) or Zeiss COMET — project structured light patterns onto the part surface and use stereo cameras to compute a three-dimensional point cloud of the entire visible surface in a single scan (typically 10–60 seconds). This point cloud is then compared against the nominal CAD model, and the result is displayed as a colour deviation map showing where the part is larger or smaller than nominal, and where warpage deflects the surface from its intended position. Full-field scanning is invaluable for warpage diagnosis because it reveals the complete three-dimensional distortion pattern of the part, not just isolated point measurements. Resolution is typically ±0.01–0.05 mm depending on the system and scan volume, which is adequate for most injection moulding applications. The limitation is cost (systems range from €30,000 to €200,000) and the need for skilled operators to set up and interpret scans.
Laser trackers: For large parts — automotive body panels, large structural components — where the scanning volume of structured light systems is insufficient, laser tracker systems (such as Leica Absolute Tracker or FARO Laser Tracker) can measure three-dimensional coordinates across a working volume of several metres with ±0.025 mm accuracy. These systems are used in automotive toolrooms and aerospace manufacturing but are uncommon in typical injection moulding facilities due to their high cost and the need for specialised operator training.
In-line SPC (Statistical Process Control) measurement: For high-volume production of critical-dimension parts, manual CMM measurement of every part is impractical. Automated gauge systems — either dedicated fixtures with air gauging or LVDT transducers, or robot-mounted CMM or laser sensors — can measure one or several key dimensions on every moulded cycle, feeding dimensional data to an SPC system. The SPC system monitors the process capability indices Cp and Cpk in real time and triggers alerts when these fall below the required levels. Standard automotive requirements are Cp ≥ 1.33 and Cpk ≥ 1.33 for standard characteristics, and Cpk ≥ 1.67 for safety-critical characteristics. This automated monitoring allows the process engineer to detect and correct dimensional drift — caused by tool wear, temperature changes, or material variability — before it produces significant quantities of out-of-specification parts.
| Measurement Method | Accuracy | Speed | Cost | Best Use Case |
|---|---|---|---|---|
| CMM (contact probe) | ±0.001–0.005 mm | Slow (30–60 min/part) | Medium–High | Formal first-article; GD&T certification; audit measurements |
| Structured light scanning (GOM ATOS) | ±0.01–0.05 mm | Fast (1–5 min/part) | High | Full-surface warpage mapping; deviation from CAD; development phase |
| Laser tracker | ±0.025 mm | Medium | Very High | Large parts (>1 m); automotive panels; toolroom verification |
| Air gauging (inline) | ±0.001–0.003 mm | Very Fast (<5 s) | Medium | Single critical bore or OD dimension; 100% production inspection |
| Tactile gauge / comparator | ±0.005–0.01 mm | Fast (1–2 min) | Low | Incoming inspection; SPC sampling on simple features |
| Vision system (2D) | ±0.01–0.05 mm | Very Fast (<2 s) | Medium | Inline inspection of flat features, hole positions, part presence |
Troubleshooting Table
The following table provides a systematic reference for diagnosing and correcting the most common shrinkage and warpage problems encountered in injection moulding production. For each problem symptom, the most likely root causes are listed in order of probability, followed by recommended corrective actions. In practice, a systematic approach — eliminate the most likely cause first, then verify the effect — is more effective than attempting multiple simultaneous changes.
| Problem | Most Likely Cause | Corrective Action |
|---|---|---|
| Excessive shrinkage — part undersized | Insufficient holding pressure; early gate seal | Increase holding pressure by 10–15 MPa; verify gate seal time and extend hold time if gate is not sealed |
| Insufficient shrinkage — part oversized | Excessive holding pressure; overpacking cavity | Reduce holding pressure; check for flash; verify screw cushion is not excessive |
| Warpage toward cavity side | Cavity side cooling insufficient; cavity wall hotter than core | Improve cavity-side cooling channels; increase coolant flow rate on cavity circuit; verify mould surface temperatures with IR camera |
| Warpage away from gate | High residual stress from excessive injection speed or melt temperature | Reduce injection speed by 20%; reduce melt temperature by 10–15 °C; evaluate gate enlargement |
| Non-uniform shrinkage along part length | Non-uniform wall thickness; non-uniform packing pressure gradient | Redesign wall thickness transitions to be more gradual; relocate gate closer to thick section; increase holding pressure |
| Sink marks on surface | Insufficient packing in thick section; gate sealing before thick section is packed | Increase holding pressure; enlarge gate to delay seal; add secondary gate adjacent to thick section; core out thick boss |
| Post-mould dimensional drift | Continuation of crystallisation after ejection (semi-crystalline polymer) | Extend mould cooling time; increase mould temperature to allow more in-mould crystallisation; consider annealing at 60–80 °C for 1 hour |
| Cavity-to-cavity dimension variation (multi-cavity mould) | Unbalanced runner system; different fill pressures in different cavities | Balance runner system geometrically (equal runner diameters and lengths to all cavities); measure fill balance with short shots; consider naturally balanced runner layout |
| Anisotropic warpage in GF-reinforced part | Fibre orientation anisotropy; flow vs. transverse shrinkage differential | Change gate location to alter flow pattern; run CAE simulation with fibre orientation prediction; evaluate alternative gate type (fan vs. pin) |
| Voids in thick section — no sink marks | Skin freezes before packing can reach centre; no surface sink because skin is rigid | Add secondary gate or valve gate adjacent to thick section; dramatically increase holding pressure; consider design change to hollow out thick section |
| Ejection-induced warpage — part distorts at ejection | Unbalanced ejector force; excessive ejection speed; insufficient draft angle; part too hot at ejection | Redistribute ejector pins to balance force; reduce ejection speed; increase draft angle; extend cooling time before ejection |
| Cycle-to-cycle dimensional scatter | Inconsistent screw cushion; worn or sticking non-return valve | Check cushion variation over 30 consecutive shots; inspect non-return valve for wear; clean check ring and seat; replace valve if necessary |
| Warpage continuing 24–48 h after moulding | Residual stress relaxation; post-crystallisation in semi-crystalline polymer | Anneal parts at 60–80 °C for 1 hour to accelerate stress relaxation; review cooling time adequacy; increase mould temperature to reduce residual stress |
| Dimensions increasing over time in PA6/PA66 part | Moisture absorption causing dimensional growth | Specify measurement condition on drawing (dry-as-moulded vs. conditioned); adjust mould compensation to account for equilibrium moisture expansion; consider moisture barrier coating for dimensional-critical parts |
| Short-term dimensional instability — varies between early and late production | Insufficient cooling time; mould not at thermal equilibrium; varying process conditions during warm-up | Extend cooling phase by 20%; run 50-shot warm-up cycle before production sampling; install TCUs with active temperature control to reach equilibrium faster |
| Wall thickness-dependent shrinkage variation within same part | Expected physical behaviour from PvT physics; unavoidable in parts with thickness transitions | Account for wall-thickness-dependent shrinkage in mould design (apply different shrinkage values to thick and thin features); reduce thickness transitions; consider this in initial part design stage |
CAE Simulation for Shrinkage and Warpage Prediction
Computer-Aided Engineering (CAE) simulation of the injection moulding process has advanced to the point where it can predict shrinkage distributions and warpage of complex parts with useful engineering accuracy before the mould is manufactured. While simulation cannot replace physical trials entirely, it provides an invaluable tool for optimising gate location, cooling circuit design, and process parameters in the design phase — when changes are inexpensive — rather than after the mould is built, when changes are expensive and time-consuming.
Autodesk Moldflow is the most widely used simulation platform globally, offering a comprehensive suite of analyses including filling, packing, cooling, and warpage (FPCW analysis). Moldflow's warpage solver computes the through-part residual stress and differential shrinkage distributions from the filling and cooling analyses, then uses finite element analysis (FEA) to predict the resulting deformation of the ejected part. Moldflow's material database contains PvT data, viscosity-temperature data, and — critically for warpage prediction — fibre orientation model parameters for hundreds of commercial polymer grades. For GF-reinforced materials, Moldflow includes fibre orientation simulation using the Folgar-Tucker model (or the newer reduced strain closure model) which predicts the distribution of fibre alignment angles throughout the part. This fibre orientation prediction feeds directly into the anisotropic shrinkage calculation, making Moldflow particularly valuable for GF-filled parts where fibre-induced anisotropy dominates warpage.
Moldex3D is a competing simulation platform with strong adoption in Asia-Pacific markets, developed by CoreTech System in Taiwan. Moldex3D is widely regarded as having particularly strong fibre orientation prediction capabilities and is often preferred for long-fibre thermoplastic (LFT) simulation where the fibre length distribution must be tracked through the filling process. Its cooling analysis is comparable to Moldflow, and its warpage solver uses similar FEA principles. The choice between Moldflow and Moldex3D in practice often depends on the existing software infrastructure of the mould designer or OEM, as both platforms have extensive material databases and validation against physical measurements.
Sigmasoft (from SIGMA Engineering, Germany) takes a different approach that focuses particularly on the three-dimensional thermal analysis of the complete mould system — including the mould plates, cooling channels, hot runner manifold, and part — over multiple moulding cycles until thermal equilibrium is achieved. This approach is more computationally intensive than the conventional "single cycle" analysis used in Moldflow and Moldex3D, but it provides more accurate temperature distributions that account for the effects of previous cycles on mould temperature distribution. Sigmasoft is particularly valued for precision tooling applications — automotive optical components, medical device components, precision gears — where even small thermal imbalances have significant effects on part dimensions and where conventional simulations fail to capture the multi-cycle thermal history of the mould.
What simulation predicts: A full FPCW analysis produces the following outputs relevant to shrinkage and warpage control: volumetric shrinkage distribution (shows where shrinkage is highest and lowest across the part); fibre orientation tensor distribution (shows flow vs. transverse orientation in GF-reinforced materials); residual stress map (shows the distribution of frozen-in stress through the part cross-section); cooling time distribution (identifies hot spots where cooling is insufficient); and warpage magnitude and direction (the predicted deflection of the ejected part in all three spatial directions, typically displayed as a colour map with the original part shape as reference).
Accuracy expectations: Engineers new to simulation often expect absolute accuracy — "the simulation says warpage will be 0.5 mm, so the actual part will warp 0.5 mm." This expectation is unrealistic. Absolute accuracy of warpage simulation is typically ±20–30% of the predicted magnitude, depending on material data quality, gate and cooling accuracy in the model, and model assumptions. This means a predicted 0.5 mm warpage could manifest as anything from 0.35 mm to 0.65 mm in practice. However, the true value of simulation is in comparative analysis: when two gate locations are compared in simulation and one produces 0.5 mm predicted warpage while the other produces 0.2 mm, the simulation is reliably indicating that the second option is significantly better — even if neither absolute number is precise. Comparative accuracy (sensitivity to design changes) is typically ±5–10%, making simulation an excellent tool for design optimisation even when absolute accuracy is limited.
When to invest in simulation: CAE simulation represents a significant cost (commercial simulation software licences cost €15,000–50,000/year; simulation service bureaux charge €2,000–10,000 per full analysis) and requires skilled operators who understand both polymer physics and FEA principles. The investment is clearly justified in the following situations: (1) complex geometry with ribs, bosses, and non-uniform wall thickness in GF-reinforced materials, where fibre orientation anisotropy will inevitably produce significant warpage; (2) tight tolerance requirements of <0.1% of nominal dimension, where trial-and-error mould modifications would be prohibitively expensive; (3) large moulds where a single mould modification costs €10,000–100,000+, making the cost of simulation trivially small by comparison; and (4) multi-cavity moulds where runner imbalance and cavity-to-cavity shrinkage differences must be optimised before steel is cut.
Sensitivity analysis workflow: For maximum value from simulation investment, a structured sensitivity analysis workflow is recommended. First, establish a baseline simulation with the initial gate location and cooling circuit design. Then systematically vary one parameter at a time — gate location, cooling channel diameter, coolant temperature, holding pressure — and document the effect of each variation on predicted warpage. This builds a process map showing which parameters have the strongest influence on warpage for the specific part geometry and material, allowing the engineer to focus experimental process development effort on the parameters that matter most. This approach is far more efficient than random process experimentation in the toolroom.
Tederic Injection Molding Machines for Dimensional Control
Achieving consistent dimensional accuracy in injection-moulded parts requires not only correct mould design and process engineering, but also an injection moulding machine that delivers repeatable, precise execution of the programmed process parameters. Shot-to-shot variation in injection pressure, holding pressure, injection speed, and screw position directly translates into variation in part dimensions and shrinkage. TEDESolutions (tedesolutions.pl) is the authorised distributor of Tederic injection moulding machines (tedericglobal.com) in Poland and Central Europe, offering a comprehensive range of machine technologies specifically suited to demanding dimensional applications.
Shot volume repeatability: The most fundamental machine specification for dimensional consistency is shot volume repeatability — the variation in the volume of polymer delivered to the mould from shot to shot. Tederic servo-hydraulic injection moulding machines achieve shot volume repeatability of ±0.1%, which means that on a 100 cm³ shot, the actual delivered volume varies by no more than ±0.1 cm³ between shots. This level of repeatability translates directly into consistent packing of the cavity and consistent part weight, which is the foundation of dimensional consistency. Variation in shot volume is the dominant source of dimensional scatter in production, and reducing it to ±0.1% ensures that holding pressure — rather than shot volume variation — is the controlling factor for dimensional accuracy.
Closed-loop pressure and velocity control: Tederic machines implement fully closed-loop control of both injection velocity and holding pressure. The machine continuously monitors actual screw velocity (via encoder feedback) and actual hydraulic or servo drive pressure (via transducer), and applies real-time corrections to the drive system to maintain the programmed profile within ±1 MPa of setpoint for pressure and within ±1% of setpoint for velocity. This closed-loop control is critical for shrinkage control because it ensures that the holding pressure applied to the mould is exactly as programmed — a machine that delivers inconsistent holding pressure will produce inconsistent shrinkage even with a well-designed mould and correct process settings.
Multi-stage holding pressure profiles: All Tederic machines in the NE1 and G-series ranges support up to 8 independently programmable holding pressure stages, each with its own pressure setpoint and duration. This capability supports the step-down holding pressure profiles described in the holding pressure section of this guide, allowing the process engineer to implement sophisticated packing strategies that minimise residual stress while maintaining effective shrinkage compensation throughout the gate-seal time window. The ability to program complex pressure profiles — and to verify their execution through data logging — is a significant practical advantage for warpage-sensitive applications.
Tederic NE1-series all-electric machines: The NE1-series all-electric injection moulding machines represent Tederic's highest-precision platform, designed specifically for applications requiring the absolute minimum process variation. By replacing hydraulic cylinders and pumps with direct-drive servo motors for each machine axis (injection, clamp, ejector, screw rotation), the NE1 eliminates the main sources of hydraulic variation: oil temperature change, pump pressure variation, and leakage in hydraulic circuits. The result is a shot volume repeatability of ±0.05% — twice as precise as servo-hydraulic — and injection velocity repeatability that enables consistent filling of thin-walled or complex-geometry parts. For optical, medical device, and precision electronics applications where dimensional tolerances of ±0.03–0.05 mm are required on moulded features, the NE1-series provides the machine performance needed to realise these tolerances in production.
Tederic G-series large-tonnage servo-hydraulic machines: For large automotive and structural components requiring clamping forces of 400–3,000 tonnes, the Tederic G-series servo-hydraulic machines deliver precision performance comparable to all-electric at the higher tonnage levels where all-electric technology becomes impractical. The G-series uses a precision servo-controlled variable-displacement hydraulic pump that delivers exactly the flow and pressure required by the process at each instant, with no excess flow being throttled off or dumped — which is both the source of conventional hydraulic machines' pressure variability and of their high energy consumption. The servo pump achieves pressure control accuracy of ±0.5 MPa at up to 180 MPa injection pressure, enabling consistent packing for large-section automotive parts where holding pressure consistency is critical for sink-free Class-A surfaces.
Energy efficiency: Tederic servo systems — whether servo-hydraulic in the G-series or all-electric in the NE1-series — consume 30–50% less electrical energy than conventional fixed-displacement hydraulic machines of equivalent clamping force. For injection moulding operations where machines run 24 hours per day, this energy saving directly reduces the cost per part. The combination of improved dimensional consistency (reducing scrap and rework cost) and lower energy consumption makes the total cost of ownership of Tederic machines highly competitive against conventional hydraulic alternatives, even when the higher initial purchase price is taken into account.
TEDESolutions service and process support: Beyond machine supply, TEDESolutions provides on-site process engineering support across Poland for customers experiencing shrinkage and warpage problems in production. TEDESolutions engineers can perform systematic process analysis — shot weight studies for gate seal verification, mould temperature mapping with infrared cameras, process capability analysis — and implement corrective measures in collaboration with the customer's production team. This service capability is particularly valuable for customers who have encountered unexpected dimensional problems after mould qualification and need rapid resolution to restore production conformance. Contact TEDESolutions at tedesolutions.pl for more information about machine capabilities and process support services available in Poland and Central Europe.
Summary
Shrinkage and warpage are the most consequential dimensional phenomena in injection moulding. Their correct prediction, compensation, and control are fundamental engineering competencies that determine whether a moulded product meets its dimensional specification or generates scrap and rework cost. The following key takeaways summarise the essential engineering knowledge from this guide:
- Shrinkage is physically unavoidable but predictable. Thermoplastic polymers contract as they cool from melt to solid, and this contraction must be pre-compensated through oversized mould cavity dimensions. Shrinkage values range from 0.2–0.5% for amorphous polymers (ABS, PC, PMMA) to 1.8–4.0% for unfilled semi-crystalline polymers (PP, POM, HDPE), with GF-reinforced grades falling in the 0.3–1.5% range depending on fibre content and direction.
- Holding pressure is the most powerful shrinkage control lever. Setting holding pressure at 60–80% of peak injection pressure, with hold time extending to gate seal (verified by the weight method), compensates a significant fraction of in-mould shrinkage by packing additional material into the cavity. Step-down multi-stage pressure profiles reduce residual stress while maintaining effective packing.
- Warpage is caused by differential shrinkage, not average shrinkage. Non-uniform cooling, fibre orientation anisotropy in GF-reinforced materials, asymmetric gate location, and abrupt wall thickness transitions all create internal stress gradients that manifest as geometric distortion after ejection. Addressing warpage requires a systems approach spanning part design, gate location, cooling circuit design, and process parameters — no single adjustment reliably solves a warpage problem.
- GF-reinforced materials require special attention to anisotropy. The ratio of transverse to flow-direction shrinkage in GF-reinforced grades is typically 2:1 to 4:1. This anisotropy is the dominant cause of warpage in filled materials and makes CAE simulation with fibre orientation prediction effectively mandatory for tight-tolerance GF-reinforced parts.
- Mould cavity dimensions must be calculated with material-specific shrinkage values. Use the formula Dw = Dz × (1 + S/100). For GF-reinforced materials, apply different shrinkage values to flow-direction and transverse-direction features. Always use material data sheet values rather than generic reference values for production tooling calculations.
- Cooling system design is the primary warpage control tool. Maintaining temperature balance of <5 °C between core and cavity sides, using separate TCUs for independent circuit control, and adopting conformal cooling for complex geometries can reduce warpage by 40–60%. Cooling time should be calculated using the engineering formula and verified empirically.
- Measurement must follow ASTM D955 conditioning. Parts must be measured after 40 hours at 23 °C and 50% RH to account for post-mould shrinkage stabilisation. For hygroscopic materials (PA, PA66), the moisture condition of measurement must be specified on engineering drawings, as dimensions can change by 0.5–2.0% between dry-as-moulded and conditioned states.
- CAE simulation reduces risk but cannot replace trials. Modern simulation tools (Moldflow, Moldex3D, Sigmasoft) provide ±20–30% absolute accuracy for warpage prediction, but their comparative accuracy (sensitivity to design changes) is ±5–10%, making them invaluable for gate location optimisation, cooling circuit design, and process parameter selection before mould fabrication.
- Machine precision directly affects dimensional consistency. Shot-to-shot variation in injection pressure and holding pressure is a direct contributor to dimensional scatter in production. Machines with closed-loop pressure control, high shot repeatability (±0.1% or better), and multi-stage holding pressure capability — such as Tederic servo-hydraulic and all-electric platforms — provide the process stability needed to achieve Cpk ≥ 1.33 on tight-tolerance features.
- Systematic troubleshooting beats trial-and-error adjustments. Use the root cause logic described in the troubleshooting table: identify the symptom pattern, determine the most likely cause, make one change at a time, verify the effect before proceeding. Changing multiple process parameters simultaneously makes it impossible to isolate which change produced which effect and will ultimately slow down problem resolution.
The convergence of three technologies is continuously raising the achievable standard of dimensional accuracy in injection moulding: CAE simulation that can predict shrinkage and warpage with increasing fidelity before steel is cut; conformal cooling systems manufactured by additive manufacturing that achieve <2 °C surface temperature uniformity across complex geometries; and servo-precision injection moulding machines that deliver holding pressure and shot volume consistency at ±0.05–0.1% repeatability. Together, these technologies are making dimensional deviations of <0.05% of nominal dimension routinely achievable in production for a growing range of part geometries and materials — a standard that was confined to laboratory conditions two decades ago. For manufacturers who invest in the engineering knowledge, tooling quality, and machine technology to apply these capabilities systematically, dimensional quality ceases to be a production problem and becomes a competitive advantage.
If your production is currently experiencing shrinkage or warpage problems, or if you are planning a new tooling project where dimensional accuracy is critical, TEDESolutions is available to support your engineering team with process analysis, machine selection guidance, and on-site troubleshooting across Poland and Central Europe. Contact us at tedesolutions.pl to discuss your specific dimensional control challenges and how Tederic machine technology can help you achieve them.
Потрібна підтримка у виборі термопластавтомата?
Зв'яжіться з нашими експертами TEDESolutions і знайдіть ідеальне рішення для вашого виробництва
Пов'язані статті
Відкрийте для себе більше цінного контенту
Довідкова таблиця швидкостей усадки пластмас (дані на 2025 рік)
Повний довідник на 2025 рік щодо швидкостей усадки при литті під тиском. Повні таблиці даних для 25+ полімерів включаючи PP, PE, ABS, PC, PA6, PBT. Формули компенсації для проектування прес-форм та фактори переробки для точних розмірів деталей.
Як розрахувати силу замкнення в обробці пластмаса: формули та приклади 2025
Оволодійте розрахунком сили замкнення в литті пластмас під тиском за допомогою перевірених формул та покрокових прикладів. Дізнайтеся, як правильно розмірити машини та уникнути дорогих дефектів облою.
Впровадження Цифрових Близнюків у Виробництві Пластмас - Аналіз ROI та Кейс-стаді
Як технологія цифрових близнюків забезпечує 150-400% ROI в операціях лиття пластмас? Кейс-стаді McKinsey та Deloitte розкривають стратегії впровадження, економію витрат та бізнес-переваги для трансформації Індустрії 4.0.