Thermoforming looks like a shaping process and is actually a thermal one. The press does not bend material into a shape; heat makes the material soft enough to accept one, and everything that follows is a race against cooling. Once that is understood, the variables that repeat and the variables that drift stop being mysterious.
Most production problems in a small insole workshop are thermal problems wearing mechanical clothes. A pair that comes out thinner on one side, a shell that springs back slightly after trimming, a batch that feels different from last month’s — these are usually differences in how the material was heated and how it cooled, not differences in the tooling.
We run thermoforming daily in our own plant, and we supply assessment systems to clinics that form in-house, so we see both the factory version and the small-room version of the same process. The variables are identical in both. What differs is how much of the process is documented, and documentation is what separates a repeatable product from a lucky one.
The shape is a side effect of the heat
Thermoformable insole materials share a property: above a certain temperature they become soft enough to deform without tearing, and below it they hold whatever shape they were given. The window between those two states is narrower than most operators assume, and its position depends on the material, its thickness and its history.
That is why two operators using the same press and the same blank can produce different pairs. One heats until the material “looks ready”; the other heats for a set time at a set temperature. The first is judging a surface appearance that reflects the outside of the blank, while the shape is being decided by the core.
It is also why a machine change rarely solves a consistency problem on its own. A better controller narrows the variation in one variable. It does not write the routine, and it does not stop an operator from removing the blank early because the queue is long.
Temperature: measure the blank, not the platen
Every press has a setpoint, and the setpoint describes the platen, not the material. The gap between them is where variation lives: contact quality, blank thickness, how evenly the blank sits, and how long it has been in the room before heating.
The practical rule is to read the setpoint as a starting condition and to watch the blank as the thing being controlled. Whatever verification method is used — a surface temperature check, a test form, a bend test — it should be applied to the blank and recorded, because that is the measurement that predicts the result.
Uniformity across the blank matters as much as the absolute value. A blank that is hotter at one end forms unevenly, and the unevenness shows up as thinning on the deep side of the arch. Operators often respond by increasing the setpoint, which makes the hot end hotter and the problem worse.
Ambient conditions deserve a note. A room that is cold in the morning and warm by mid-afternoon changes how quickly a blank heats and how quickly it cools. A routine written in July will behave differently in January unless the routine includes a check rather than only a duration.
There is one more subtlety worth recording. The temperature a blank reaches is not the temperature it was set to, and the difference is not constant across the life of a heating element. A routine that records only settings will therefore record the intention rather than the event. Recording the verification result alongside the setting is what turns a setting into evidence, and evidence is what allows a workshop to explain a batch six months later.
Dwell: the variable that hides in plain sight
Dwell is the time the blank spends under heat before forming. Too short and the core stays stiff while the surface softens, which produces a shell that forms at the surface and springs back later. Too long and the material can sag, thin at the edges or lose the property that made it worth buying.
Both faults look similar at the bench: a pair that does not match the model closely enough. The difference appears later. Under-dwelled pairs tend to relax after trimming, so a shape that was acceptable at inspection is subtly different a week later. Over-dwelled pairs tend to be thin exactly where the shape is deepest, which is where load concentrates.
Dwell also interacts with material thickness in a way that surprises operators moving between product tiers. A thick blank needs more time to reach core temperature; a thin blank reaches it quickly and then has nothing left to give. A single dwell setting across a mixed order is a recipe for one product being right and the other being wrong.
The manageable version is a small table: material family, thickness band, dwell, verification result. It does not need to be sophisticated. It needs to exist, because the moment it exists an operator can follow it and a supervisor can audit it.
Cooling under load is the step nobody owns
After forming, the shape has to be held while the material returns below its softening range. If it is released early, it relaxes; if it is held unevenly, it distorts. This is the step most often left to habit, and it is the step that most often explains a pair that looks right and behaves wrong.
Cooling under load means the model or the foot stays in contact until the shell is stable. That takes time which is invisible on a quotation and visible in a queue, which is exactly why it gets shortened when the room is busy. A written minimum, checked occasionally with a timer, is the only defence.
Uniform cooling matters too. A shell cooled on a cold bench on one side and in still air on the other can develop a slight twist that nobody notices until the fitting appointment. Cooling fixtures exist for this reason; a simple cradle that supports the shell evenly does most of the work.
There is a training point here that is easy to miss. Operators understand heating because it is visible. Cooling is invisible, so it feels optional. The routine has to make it explicit, with the same status as the heating step, or it will be compressed whenever production is under pressure.
Blank history: moisture, storage and batch
Material is not inert before it reaches the press. Sheets absorb moisture from the air, and moisture turns to vapour when heated, which shows up as surface bubbles, inconsistent forming or a shell that feels different from the previous batch of the same grade.
Storage discipline is therefore part of the process, not an administrative detail. Sealed storage, stock rotation and a note of when a carton was opened all reduce variation. In humid climates this matters enough to change the routine, not merely the housekeeping.
Batch changes deserve a first-off check even when the grade is nominally identical. Suppliers change slightly, and a new batch can behave differently under the same settings. Running one pair and comparing it against a retained reference is cheap; discovering the difference across a full order is not.
Keeping a reference pair from each approved batch is the simplest version of this discipline. It costs one pair per batch and it gives the workshop a physical standard to compare against when something feels off.
Where drift comes from over months
Even a documented routine drifts, because the equipment around it ages. Heating elements lose efficiency, thermocouples read slightly differently, seals harden and jigs wear at the contact points. None of these fail suddenly; all of them change the result gradually.
The usual symptom is a slow increase in the number of pairs that need rework, with no obvious cause. Operators compensate by nudging settings, which restores output and hides the trend. Six months later the settings are far from where they started and nobody can explain why.
A periodic verification closes that loop: run a reference specification, compare against the retained pair, record the result. Monthly is enough for most small workshops, and the record itself is what makes the drift visible before it becomes a complaint.
Operator drift is the other half. A routine followed by one person becomes a set of personal habits, and when that person leaves, the habits leave with them. Written routines with a verification step survive turnover; verbal traditions do not.
First-off checks and honest scrap
Every production run should begin with a first-off pair that is inspected against the specification before the rest of the batch proceeds. This is standard practice in manufacturing and surprisingly rare in small insole workshops, where the first pair often becomes the first sale.
The check is short: shell geometry against the model, thickness at the deepest point, trim line, heel cup symmetry, surface condition. If the first pair is outside tolerance, the batch does not start, and the settings are reviewed rather than adjusted by feel mid-run.
Scrap discipline is the harder cultural change. A pair that is out of tolerance is not a discount item and not a training pair; it is evidence that the process moved. Recording it, with the settings used, converts a loss into information. Workshops that hide scrap pay for it twice.
The same logic applies to re-makes from fitting appointments. A re-make is a signal about the specification or the process, and the useful response is to find which one moved. Treating it as an inconvenience to be absorbed is how a workshop ends up with a re-make rate it cannot explain.
When to stop relying on heat forming
Heat forming is the right process for a wide range of products, and it stays right as long as the specification can tolerate its variation. The question is not whether it is good or bad but whether the product you are selling needs more control than it provides.
The signal to move on is repeatability across batches rather than quality within one. If two pairs made a month apart from the same specification are visibly different, and the routine is documented and verified, then the process is the limit. At that point matched tooling or machining becomes worth its cost.
Volume changes the calculation too. Tooling is expensive per model and cheap per pair at volume; at low volume it is the reverse. A workshop that expects to produce a few dozen pairs of a given model should stay with heat forming, and a workshop producing the same model continuously should not.
There is a middle path many workshops use: heat form the shell, then finish by machining the trim line and the top surface to a controlled geometry. It keeps the low tooling cost of forming and adds dimensional control where it matters most, which is usually the interface with the shoe.
Documenting a routine that survives
A useful routine fits on one page and contains four things: the settings, the verification method, the acceptance criteria and the record. Settings without verification are a suggestion. Verification without acceptance criteria is an opinion.
Photographs help more than prose in a workshop. A photograph of a correctly formed shell next to a model, a photograph of an acceptable trim line, a photograph of the cooling cradle in use — these communicate faster than paragraphs and they survive translation.
Finally, the routine should name an owner. A document without an owner is a document that gets followed until the room is busy. Naming a person, and giving them the authority to stop a batch, is what makes the document operational rather than decorative.
Frequently asked questions
Do you manufacture the ovens and presses used for thermoforming?
No. Our scope is foot assessment systems and finished custom orthotic insoles. We run thermoforming in our own plant and can describe the process in detail, but we do not build or sell forming equipment.
Why do two pairs from the same specification look different?
Usually because heat and cooling were not controlled the same way. Dwell time, blank temperature, how long the shell stayed under load while cooling, and the moisture history of the blank all change the result. A written routine with a verification step is what closes that gap.
How can I tell whether a blank has been heated enough?
By checking the blank rather than the platen: a surface temperature reading, a test form, or a bend test on a scrap corner. The setpoint describes the machine; the blank is what actually decides the shape.
Can we form insoles without a cooling fixture?
Yes, but the shell has to be supported evenly while it cools and it has to stay under load long enough to be stable. A simple cradle does most of what a fixture does, and it costs far less than the rework caused by cooling a shell on a flat bench.
What is the minimum order if we send specifications to your factory?
Two hundred pairs for insoles, mixable across models and sizes, with samples dispatched within three days of specification approval. That lets a clinic compare factory output against its own forming before committing to production volumes.
How often should the process be verified?
Monthly is enough for most small workshops: run a reference specification, compare the output against a retained pair, and record the result. The record is what makes gradual drift visible before it becomes a customer complaint.
What support comes with your assessment systems?
Each system ships with a dedicated WhatsApp group, direct access to our engineers, remote assistance where the site allows it, and over-the-air updates for units connected to the internet. Machine warranty runs twelve months on non-human damage with free replacement parts.
What we would do next
Take one specification you produce regularly, run it three times across a week, and compare the pairs against each other rather than against the model. That exercise will tell you more about your process than any equipment upgrade, and it costs a few blanks. If the comparison shows the process is the limit and you would rather not invest in tooling, our insole categories page shows what we produce to specification, and the FAQ page covers minimum order and sample timing.
This article is intended for professional and educational reference. It is not medical advice. Foot assessment equipment produces measurements and images; clinical interpretation and any decision about care remain the responsibility of a qualified professional. Certification scope is limited to the documentation we hold: European Union certification and ISO quality system certification.