Why Does a Silicone Dog Paw Cleaner Cup Lose Its Shape After Repeated Use?

Failure analysis / shape retention · reviewed October 2026

The first batch sells. Six weeks later the reviews mention a cup that "won't stand up straight any more", a mouth that has gone wavy, or a wall that folds inward when the dog steps in. The material is still silicone. Nothing has torn. The part has simply stopped being the shape it was moulded into.

Shape loss in a flexible silicone cup is usually described as a material problem, and the first response is to ask for a harder compound. That is the most expensive way to answer the question, because a harder compound changes the feel of the product and may not fix a structure that was never supported in the first place.

Before changing anything, it helps to separate three different failures that all look like

Definition. Compression set is the deformation that remains after a compressed rubber part is released, expressed as a percentage of the compression applied. It is not the same thing as the temporary squeeze a flexible cup is designed to survive.

Name the deformation before you explain it

The wall folds inward during handling

This is a stability problem. It appears when the unsupported wall is tall relative to its thickness, or when the compound is soft enough that a normal grip exceeds the wall's buckling resistance. It is often worst on the side where the hand squeezes while the other side is free.

The rim goes wavy and the opening stops being round

A wavy rim is a different failure. It usually appears before the product reaches a customer, and it points at release, at rim geometry or at how the part was supported after demoulding. A rim that is oval from day one will not become round with use.

The base rocks and the cup will not stand

A rocking base is a flatness problem, and flatness on a moulded rubber part is sensitive to the base junction, to shrinkage at a thick section, and to how the part cooled. It is rarely caused by the user.

Why the distinction matters

Each failure has a different cheapest fix. A collapsing wall responds to geometry or hardness. A wavy rim responds to release and mould closure. A rocking base responds to the base design and the cooling path. Treating all three as "soft material" wastes a trial cycle and often makes two of them worse.

Write down which of the three you are looking at, with photographs, before contacting anyone. A supplier can work with a named failure. A supplier cannot work with "the cups are bad".

Three clocks are running at the same time

Silicone recovers quickly, which makes the first ten seconds misleading. The deformation customers complain about has usually accumulated on a slower clock.

Seconds: elastic recovery

Release a squeeze and the cup springs back. If it returns within a second or two, the material is behaving elastically and the problem is not yet visible. This is the behaviour most people test in a showroom, and it is the least informative.

Hours to weeks: creep under sustained load

Leave the cup compressed, or store it with something resting against the wall, and the material continues to move. Creep is the reason a cup that spent a month in a warehouse carton with another cup pressed into it can arrive slightly oval. It is also the reason packaging design belongs in the same conversation as the compound.

Months: permanent set

Compression set is the residual deformation after a compressed sample is released. ISO 815-1 describes the test at a reference strain of 25 percent for most hardnesses, with higher-hardness compounds tested at lower strain, and it separates the physical part of the effect from the chemical part that grows with temperature and time. The standard's abstract is a clear explanation of why a warm, loaded part changes permanently while a cool, unloaded one recovers.

The compression set figure is expressed as a percentage: zero means full recovery, one hundred means no recovery at all. For a paw cup, a low single-digit figure at room temperature is unremarkable; the interesting number is what happens after a few weeks at summer warehouse temperature with the cup nested inside another cup.

Is the compound too soft, or the structure unsupported?

Hardness is one variable among four

Wall thickness, cup diameter, unsupported height and rim geometry all contribute to stiffness. A 40 Shore A cup with a 2.5 mm wall can be stiffer in the hand than a 50 Shore A cup with a 1.8 mm wall, because the second moment of area grows with the cube of thickness while hardness moves linearly at best. Changing hardness without looking at the wall is a guess dressed as an engineering decision.

Durometer readings also have limits. ASTM D2240 is an empirical indentation test, and the standard states plainly that no simple relationship exists between indentation hardness and other fundamental properties. That sentence is worth quoting the next time a hardness change is proposed as a cure for a geometry problem.

The unsupported wall ratio

Measure the height of wall that has no nearby support, and divide it by the wall thickness. A ratio around 40 behaves very differently from a ratio around 70. This single number explains most collapse complaints, and it can be calculated from the drawing before a single part is moulded.

Thickness transitions

A base that is 4 mm thick meeting a wall that is 1.8 mm creates a hard step. Rubber shrinks differently across that step, and the base junction becomes the place where the part either holds its shape or starts to curl. A tapered transition costs nothing in the mould and removes a stress concentration.

The same logic applies at the rim: a thick rolled rim on a thin wall pulls the wall outward as it cures, and the opening grows slightly oval. Thinning the rim or moving the roll inward often fixes an ovality problem that hardness cannot touch.

Could the mould be the cause rather than the design?

Release stretch leaves a memory

To demould a deep cup, the part has to pass over a core. If the rim is gripped, stretched or peeled rather than lifted, the material near the rim is strained beyond its elastic limit for a moment, and part of that strain stays. The result is an oval mouth that looks like a design failure and is actually a release failure.

You can test for this without a laboratory: measure the opening immediately after demoulding and again after twenty-four hours. A part that changes measurably in that window was strained during release or is still cooling under load.

Wall variation from an off-centre core

If the core sits slightly off-centre, one wall is thicker than the other. The thick side cools and shrinks differently, and the cup develops a lean or a flat spot. Sectioning one part and measuring the wall at eight points is unpleasant work that answers the question in an hour.

Batch-to-batch cure variation

Cure time and temperature interact with compound age and with how the mould is loaded. A batch that is slightly undercured will feel softer, recover more slowly and take a set sooner. That is why a complaint about "this batch is different" deserves a batch record rather than a compound change.

A diagnostic sequence for one afternoon

You need four cups from the same batch, one from an older batch if possible, a caliper, a flat surface and a clock.

1. Photograph the failure with a ruler in frame. Note where it starts: rim, wall, base.
2. Measure the opening across two axes and the height at four points. Write the numbers down.
3. Place the cup on a flat surface and try to rock it. A sheet of paper under one edge is enough to see the gap.
4. Squeeze to a fixed gap for sixty seconds, release, and time the return to the original opening dimension.
5. Repeat with a cup from a different batch and compare. One part is an anecdote; two batches is a pattern.
6. Store one cup nested inside another for a week at room temperature, then measure again. This reproduces the warehouse, not the bathroom.

By the end of that sequence you will know whether you are dealing with a design limit, a release problem, a batch problem or a packaging problem. Those four have four different fixes, and only one of them involves changing the compound.

What should you change first, in order of cost?

Start with the changes that do not touch the mould: packing method, storage instruction, and the strain the product sees in the carton. Then look at geometry that can be changed inside the existing cavity, such as a local rib or a transition radius, if the tool allows it. Only then consider a new tool or a different compound.

A hardness change is the most visible change to the customer and the least predictable in its effect, because it alters feel and cleaning contact at the same time as it alters stiffness. It should be the last experiment, not the first.

When the part has to be seen

Some causes cannot be separated from photographs. A rim that is oval in a picture could be a release problem, a mould closure problem or a storage problem, and the three look identical on a phone screen. At FromRubber we ask for the measurements from the sequence above, the batch or order number, and one photograph of the affected area with a ruler in frame. With those three items we can usually narrow the cause to one process stage before anyone ships a sample.

FromRubber is the export brand of Dongguan Bohao Electronic Technology Co., Ltd., a Dongguan-based silicone moulder running its own tooling shop and certified to ISO 9001:2015, ISO 14001:2015 and IATF 16949:2016.

We do not name a root cause from photographs alone, and we do not recommend changing hardness before the wall distribution has been measured.

If you are already in that conversation with a customer and need a second opinion, send the details to nani@fromrubber.com or karl@fromrubber.com. WeChat and WhatsApp: +86 18676210913.

Questions that come up when cups lose their shape

Diagnose the structure before you change the material

Shape loss is a message about where the load went. Read the message before replying to it: measure the opening, time the recovery, check the base on a flat surface, and compare two batches. In most cases the answer is visible in the geometry or in the way the part was released, not in the hardness number on the datasheet.

A cup that keeps its shape is not a cup made of harder rubber. It is a cup whose wall, rim and base were designed to carry the loads that handling and storage actually apply.