Interference Fit in Silicone Overmolding: Matching LSR Shrinkage Against Substrate Tolerance

When liquid silicone is overmolded onto a rigid substrate — a plastic or metal core — and mechanical retention depends on the silicone gripping that substrate, the quality of that grip is governed by a specific interaction that's easy to get wrong: the interference fit between the shrinking silicone and the dimensionally stable substrate. Because LSR shrinks significantly and somewhat variably as it cures (as discussed more broadly in the context of silicone shrinkage versus rigid plastics), and because the substrate has its own dimensional tolerance, the actual grip force on any given part is the product of two variables stacking together — and designing for the nominal case alone often produces parts at the tolerance extremes that either grip too loosely or generate excessive stress.

Why Interference Fit in Overmolding Is a Stacked-Tolerance Problem

In a purely mechanical interference fit between two rigid parts, both components have defined tolerances and the interference is the difference between them. Silicone overmolding adds a distinct complication: the silicone's final dimension isn't set by machining tolerance but by shrinkage behavior during cure, which introduces its own variability on top of the substrate's tolerance.

• The silicone shrinks onto the substrate as it cures, and it's precisely this shrinkage that creates the grip — the silicone is effectively trying to contract to a dimension smaller than the substrate it surrounds, and the resulting interference is what holds the two together.
• The actual grip force depends on both how much the silicone shrinks and where within its tolerance band the substrate dimension actually falls — meaning the real interference is a combination of two independent variables, each with its own range, not a single fixed value.
• Designing for only the nominal (average) case ignores what happens at the tolerance extremes: a substrate at the small end of its tolerance combined with silicone at the high end of its shrinkage range produces very different grip than a substrate at the large end combined with low silicone shrinkage.

What Goes Wrong at Each Tolerance Extreme

The key insight from this table: the design has to work across the full combined range of both variables, not just at the nominal midpoint — because real production parts will land across the whole span, and both extremes carry their own distinct failure modes.

Why Both Extremes Are Genuine Failure Modes, Not Just "Slightly Off"

• The loose extreme isn't merely a cosmetic looseness — for parts where the silicone's mechanical grip is the retention mechanism (grips, seals, handles, protective overmolds), insufficient interference can allow the silicone to rotate, slide, or eventually separate from the substrate under use, which is an outright functional failure.
• The tight extreme isn't just "a snug fit" — excessive interference means the silicone is under sustained tensile stress in service, which can accelerate stress-related degradation over time, increase the risk of tearing during assembly or use, and interact poorly with the compression set and long-term dimensional behavior discussed in the context of silicone's shape-recovery properties.

How the Calculation Should Actually Be Approached

Rather than designing to nominal dimensions and hoping production variation stays close to the average, a more robust approach accounts for the combined tolerance stack directly:

• Establish the substrate's actual dimensional tolerance range, not just its nominal dimension, from the substrate supplier's real capability rather than an idealized drawing value.
• Establish the silicone's realistic shrinkage range for the specific compound and part geometry — validated through trial data rather than a single generic shrinkage percentage, since shrinkage varies with formulation and section geometry.
• Evaluate grip across the full combined range — specifically checking the loose extreme (small substrate + low shrinkage) and the tight extreme (large substrate + high shrinkage) — to confirm that even the worst-case combinations still produce acceptable grip without excessive stress.
• Adjust the design target so the acceptable grip window spans the full combined tolerance range, which sometimes means tightening substrate tolerance, adjusting the nominal interference, or incorporating mechanical retention features (undercuts, grooves, texture) that supplement pure interference grip and reduce sensitivity to the tolerance stack.

The Role of Supplemental Mechanical Retention

Because pure interference fit is inherently sensitive to the combined tolerance stack, many well-designed overmolded parts don't rely on interference alone:

• Mechanical retention features — grooves, undercuts, through-holes, or surface texture on the substrate that the silicone flows into and locks around — provide retention that's far less sensitive to the exact interference value, since the grip comes partly from mechanical interlock rather than solely from contraction pressure.
• This is often the more robust design choice for parts where retention is functionally critical, precisely because it reduces dependence on holding both the substrate tolerance and the silicone shrinkage within a narrow combined range.

Questions to Ask a Mold or Overmolding Supplier

• Has the interference fit been evaluated across the full combined tolerance stack of both substrate dimension and silicone shrinkage range, or designed only to nominal dimensions?
• What actual shrinkage data — for this specific compound and geometry — was used, and how was it validated?
• Does the design rely on interference fit alone, or does it incorporate mechanical retention features to reduce sensitivity to the tolerance stack?
• What is the substrate supplier's real dimensional tolerance capability, and has that actual range (not just the drawing nominal) been factored into the grip calculation?

The Practical Takeaway

Interference fit in silicone overmolding is a stacked-tolerance problem where silicone shrinkage variability and substrate tolerance combine to determine the real grip on each part. Designing to the nominal case alone reliably produces parts at the extremes that either grip too loosely or carry excessive stress — both genuine failure modes. Evaluating the design across the full combined range, and supplementing pure interference with mechanical retention features where retention is critical, is what separates an overmold that works consistently in production from one that works on the sample bench but fails intermittently at volume.