Environmental Stress Cracking: Why Plastic Parts Crack Without Being Dropped

A pair of glasses sits on a desk. Nobody drops it, nobody sits on it. A few months later, a fine crack appears next to a screw boss on the temple. A smartwatch band clip, a phone case, an earbud stem, a laptop hinge cover -- the story is often the same: a plastic part fails at a load far below anything it saw in a drop test.

The usual cause is environmental stress cracking (ESC). It is one of the most common reasons plastic parts fail in the field, and one of the easiest to miss in development, because it needs three things at once -- and a standard lab test usually supplies only one or two of them.

Three Ingredients: Plastic, Strain and a Chemical

ESC happens when:

A susceptible plastic -- most often an amorphous polymer such as polycarbonate (PC), PC/ABS blends, ABS, acrylic (PMMA) or polystyrene,
is held under a sustained tensile strain -- from a press fit, a screw boss, a snap fit, a bent clip, moulded-in stress or a thermal mismatch,
while in contact with an active chemical -- skin oil (sebum), sunscreen, hand cream, insect repellent, cosmetics, cleaning agents, alcohol wipes, lubricants or some adhesives.

Remove any one of the three and the part usually survives. Put all three together and a part that would carry the same strain in air for years can craze and crack in days or weeks.

The chemical does not dissolve the part. It diffuses into the most strained regions of the surface, lets the polymer chains slide apart more easily and turns a slow, harmless creep into tiny crazes -- microscopic, fibril-bridged cracks. Under continued strain the crazes grow and join until a visible, usually brittle-looking crack runs through the wall.

Why Drop Tests Miss It

A drop test applies a large strain for a few milliseconds, in clean lab air. ESC needs a modest strain held for a long time in the presence of a chemical. The two failure modes live in completely different corners of the load map:

A part can pass every drop and still crack in the field from ESC.
The crack origin is typically at a sustained stress raiser -- a boss, a sharp internal corner, a weld line, a gate -- not at the impact point.
Fracture surfaces look brittle, even for tough plastics like PC that are famous for ductile behaviour.

That is why ESC is normally treated as a separate reliability topic, with its own tests and its own design limits.

Critical Strain: The Number That Matters

For a given plastic and a given chemical, there is a strain level below which cracking does not start within the test time. This is called the critical strain. In air, many engineering plastics tolerate sustained strains of a few percent. In contact with an aggressive chemical, the critical strain of a susceptible plastic can drop well below 1 % -- sometimes to a few tenths of a percent.

Two practical points follow:

Strain, not stress, is the useful measure. Plastics relax under a constant strain, so the stress at a press fit falls over time while the strain stays. ESC test methods therefore usually fix the strain.
The chemical decides the limit. The same grade can be robust against water and alcohol and sensitive to sunscreen or oleic acid. A material datasheet that says "good chemical resistance" without naming the chemical and the strain says very little.

How ESC Is Tested

Public test methods give a common language. Among them:

ISO 22088 (Plastics -- Determination of resistance to environmental stress cracking), a multi-part series. Part 3 describes the bent strip method: a test bar is clamped over a curved fixture so that its outer surface is held at a known strain, the chemical is applied, and the bar is inspected for crazes and cracks after a set time.
ASTM D543 (Standard Practices for Evaluating the Resistance of Plastics to Chemical Reagents), which includes a practice for specimens held under strain while exposed to a reagent.

A typical bent-strip study runs several strain levels, often with specimens kept warm and humid to mimic skin contact, and checks them after a few days. A sensitive way to quantify the damage is to pull the exposed bars in a tensile test afterwards and compare their elongation at break with unexposed bars: a large loss of elongation shows damage long before a crack is visible.

The strain on the bar's surface comes from simple beam geometry: for a bar of thickness t bent over a radius R, the outer-fibre strain is about t / (2R). That makes the test cheap and repeatable -- and it gives a strain number that can be compared directly with simulation results.

These standards are good references; the chemicals, strain levels, temperatures and pass criteria a particular product must meet are set by each product team.

Where Simulation Comes In

ESC is a long-term problem, but the strain that drives it is a static quantity that structural simulation computes well:

Assembly strain. Press fits, screw torque, snap fits that stay deflected, clips holding a lens or a display, heat stakes -- each leaves a sustained strain in the plastic.
Thermal strain. A metal insert in a plastic housing, a glass lens in a plastic frame or a battery pressing on a cover strains the plastic every time the temperature changes. A product left in a hot car combines high temperature, high strain and, often, sunscreen.
Worn strain. Glasses temples spread when worn, watch bands bend around the wrist, earbud stems flex in the ear -- the product is strained every day the user wears it.

A simulation of these states produces a map of sustained strain across every plastic part. Compared with the critical strain of the chosen material against the chemicals it will meet, the map shows immediately which features are at risk -- usually a short list of bosses, corners and clips -- long before any part is moulded.

Simple rules of thumb help to read the map:

Look at the maximum principal strain on the surfaces a user or a chemical can reach, not deep inside the wall.
Use the strain that stays after assembly or under steady wear, not the peak of a short event.
Pay extra attention to weld lines and gates, where the moulded material is weaker and may carry moulded-in stress on top of the assembly strain.

Design Levers

When a feature shows up with too much sustained strain, there are several ways out:

Reduce the interference. Smaller press-fit interference, a looser tolerance stack or a compliant feature lowers the sustained strain directly.
Round the corners. Generous radii at bosses, ribs and clip roots spread the strain.
Change the fastening. A screw boss with thicker walls, a different thread-forming screw or a metal insert can move the strain away from exposed surfaces.
Let snap fits relax. A snap fit should deflect during assembly and return to near zero strain afterwards, not stay preloaded.
Choose the material against the chemical. Semi-crystalline plastics (for example polyamides, PBT, PP) and certain blends are much less sensitive to many skin-contact chemicals than unfilled amorphous grades. Recycled-content grades can behave differently from their virgin counterparts and are worth testing separately.
Control the process. Higher mould temperature, adequate packing and an annealing step reduce moulded-in stress.
Keep the chemical away. Coatings, sealed seams and careful selection of adhesives, greases and cleaning agents used in assembly all help.

Closing Thoughts

Environmental stress cracking is a slow, quiet failure: no impact, no overload, just a modest strain, a common chemical and time. It is also predictable. Once the critical strain of a material against the relevant chemicals is known, a simulation of the assembled, heated and worn product shows where the sustained strain exceeds it -- early enough to change a radius, an interference or a material instead of a tool.

AutoCAE.AI builds AI-assisted simulation workflows that help hardware teams check sustained-strain hot spots in plastic parts alongside drop, impact, thermal and stiffness load cases. Visit www.autocae.ai to learn more.