Soft parts sell products. The grip on a kitchen tool, the strap on a wearable, the bumper on a rugged case, the entire body of a baby product — these are the surfaces a customer touches in a store, and the ones that decide whether the thing feels like a $12 item or a $60 item. The problem is that "soft" is not one material. It is at least three material families with completely different tooling, cost structures, and failure modes, and picking the wrong one is a mistake you usually discover after the tool is cut.
LSR, HCR, and TPE: three different products
Two of these are real silicone. One is not silicone at all, which is the source of most confusion in early quotes.
| Property | LSR (liquid silicone rubber) | HCR (solid / high-consistency silicone) | TPE / TPU |
|---|---|---|---|
| Process | Injection molded, thermoset, automated | Compression or transfer molded, often hand-loaded | Injection molded like a thermoplastic |
| Typical tool cost | $12,000–$60,000 | $3,000–$20,000 | $5,000–$40,000 |
| Cycle time | 20–60 seconds | 2–10 minutes | 15–45 seconds |
| Economical volume | High (tens of thousands and up) | Low to medium | Medium to high |
| Temperature range | About -60°F to 400°F | Similar | Roughly 0°F to 175°F |
The short version: silicone wins on temperature, UV resistance, aging, biocompatibility, and that dry premium touch. TPE wins on cost and on bonding easily to rigid plastics in a two-shot tool. HCR is the low-volume, cheap-tooling option most founders never hear about, because big molders quote LSR by default.
If your volume is a few thousand parts a year, ask specifically for a compression-molded HCR quote before you accept an LSR tooling number. The detailed process economics are in liquid silicone rubber molding and compression molded rubber vs injected TPE.
Choosing durometer for grip
Durometer is the number everyone gets wrong first, usually by going too soft. Softer feels better in a five-second showroom test and worse after a month, because it picks up dirt, deforms, and tears.
- Shore A 20–30: gel-like. Seals, cushions, skin-contact pads. Too soft for anything gripped hard.
- Shore A 40–50: the sweet spot for hand-tool grips and handles. Compresses enough to conform to the palm, firm enough to transmit torque without squirming.
- Shore A 55–70: wearable straps, bumpers, buttons with a defined click, anything that needs shape retention.
- Shore A 70–90: structural elastomer — feet, gaskets under compression, protective corners on rugged products.
A useful rule: if the user applies torque through the soft layer, do not go below about Shore A 45, and keep the layer under about 3 mm over the rigid core — a thick soft overmold feels vague because the rigid part underneath moves relative to the skin. Our shore hardness guide covers how the scales relate and how to specify a tolerance, which matters — durometer is normally held to about ±5 points, and that spread is visible in feel.
Always test durometer on the actual geometry: a 50A grip on a thin oval handle feels nothing like the same compound on a fat cylinder.
Surface texture, tack, and the dust problem
The biggest complaint about soft silicone products is that they collect lint. Silicone has natural surface tack and attracts dust, and on a black product that is visible from across a store. Four countermeasures, and you generally need two:
- Texture the tool. A matte or fine-stipple texture drastically reduces apparent dust and fingerprints compared with a gloss finish. Choose deliberately; see mold texture selection. On silicone, textures also reduce the sticky drag that makes a part hard to pull out of a pocket.
- Ask for a low-tack or self-lubricating compound. Suppliers offer LSR grades with reduced surface friction specifically for this. It is a compound choice, not a post-process.
- Post-cure. A proper post-cure removes volatiles that contribute to tack; plasma treatment is available for demanding cases at added cost.
- Choose the color knowing the risk. Black and dark navy show lint worst. Mid-grays, warm neutrals, and saturated colors hide it. This is a real design constraint, not a preference.
Test dust pickup at the prototype stage — rub the part on a fleece jacket, pocket it for a day, photograph it — because the fix lives in the tool and the compound.
Color, translucency, and what silicone will not do
Silicone colors beautifully in the pastel-to-saturated range and can be made water-clear, which TPE struggles to match. It pigments through the material rather than sitting on the surface, so scratches do not show a different color underneath — a real durability advantage over painted or coated rigid parts, and one reason it beats soft-touch coating on products with a long service life.
Limitations to plan around:
- Color matching is harder than with thermoplastics. Expect a match to a Pantone chip to be approximate, and expect a slight shift between LSR and any rigid part molded in the same color. Approve physical plaques from the actual compound, never a screen.
- Printing on silicone is difficult. Standard pad printing inks do not adhere. You need silicone-specific inks with a primer, laser marking on a two-layer compound, or molded-in features. Debossed and embossed graphics are the reliable route.
- Silicone does not bond to most adhesives. Plan the joint as a mechanical or chemical bond designed in from the start, not something you glue later.
- Fluorescents and true whites are limited next to what is possible when coloring plastic parts, because the pigment chemistry that survives the cure is narrower.
Bonding soft to rigid
Almost every soft product is really an assembly of a soft skin on a rigid skeleton, and the interface is where products fail. Three approaches, in descending order of reliability and cost.
Two-shot / overmolding
The rigid part is molded first, then the soft material is shot over it in a second station or tool — the strongest and cleanest result. With TPE it is chemistry: pick a grade formulated to bond to your substrate, typically PP-bonding or ABS/PC-bonding. LSR does not naturally bond to thermoplastics, so you need a self-adhesive grade, a primer, or a substrate that tolerates the cure temperature (PC, PA, PBT, or metal — not standard ABS). See overmolding and two-shot injection molding.
Mechanical interlock
Design through-holes, undercuts, dovetails, or a lip that the soft material flows through and locks around. This works with any material pair, is forgiving of chemistry, and is what you fall back on when the substrate cannot take a chemical bond. Keep at least 1.5 mm of soft material through the hole so the anchor does not tear.
Separate part, assembled
A stretched-on sleeve or snapped-in gasket. Cheapest tooling, and colors change without a new rigid tool — which is why wearables use separate straps. The risk is a visible gap line.
Tooling and part design rules for soft parts
- Draft is different. Soft parts can be pulled from zero-draft or mildly undercut steel because the material deforms — a design freedom rigid plastics do not give you.
- Watch wall transitions. Silicone tolerates thickness variation far better than thermoplastics, but thin webs under 0.5 mm tear during demolding and in the field.
- Flash is the normal defect. LSR is very low viscosity and finds any gap in the tool, so deflashing is a real line cost. Put the parting line where a fingertip will not feel it.
- Specify post-cure if it matters. It drives off volatiles and is required for food-contact and medical parts, so it belongs on the drawing.
- Biocompatibility for skin contact. Anything worn against skin needs the right grade and, for medical use, testing under ISO 10993.
Projects House designs and engineers soft-touch and silicone products end to end — durometer and compound selection, overmold interfaces, tool design, and supplier qualification. Tell us what has to feel good in the hand through the contact form.