Wall Thickness in Vacuum Forming: Why Corners Thin and How to Control It

Why a formed part cannot have uniform walls

A vacuum formed part starts as a flat sheet of uniform thickness and ends as a three-dimensional shape. The surface area of the shape is larger than the area of the sheet that formed it. The material has to come from somewhere, and the only somewhere available is thinning.

That is not a manufacturing defect that better tooling eliminates. It is conservation of material. The useful question is not "how do I get uniform walls" but "where do I need thickness, and how much can I afford to lose everywhere else".

Three mechanisms decide where the thinning happens:

• Stretch ratio. Where the sheet is stretched furthest from its original position, it thins most. On a deep draw, that is the corner between the flat and the wall, and the top of the wall.
• Temperature. Hotter material stretches more easily, so a hot zone thins preferentially. A sheet that is 10 °C hotter in the centre will be noticeably thinner at the centre.
• Contact cooling. Once the sheet touches the mold it stops stretching — it is pinned by friction and cooled by conduction. Material that touches early stays thick; material that is still free keeps drawing down. This is why the last point to touch is usually the thinnest, and why a mold with well-placed vacuum holes can actually improve thickness distribution by pinning the sheet early in the right places.

Draw ratio: the number that predicts everything

The draw ratio is the relationship between the depth of the part and its width. It is the single best predictor of how much thinning you will get, and it is available at the design stage before any tooling exists.

Percentages are minimum wall at the worst point — normally the corner between the flat and the wall, or the top of a tall narrow wall. Values assume a reasonable corner radius; a sharp corner will be worse than the table suggests.

Note that the draw ratio uses the minimum width, not the average. A long, narrow trough — say 600 mm long, 80 mm wide, 100 mm deep — has a draw ratio of 1.25 even though it looks shallow, because the material has to come in from the long sides.

The thickness prediction, and how far to trust it

The simplest useful model treats the formed wall as a truncated cone and assumes the material stretches uniformly. For a simple female cavity of width W and depth D:

The model is worth using because it is directionally right and it is available in thirty seconds. It is worth distrusting because:

• It assumes the flange does not draw in. In reality the flange is clamped, so more material comes from the walls than the model predicts — actual walls are thinner than the formula says on a clamped sheet.
• It ignores temperature gradients, which can move the answer by 20 %.
• It says nothing about corners, which are always the thinnest point and usually the failure point.
• It assumes a circular or square section. On a long narrow part the ends behave differently from the sides.

Use it to decide whether a design is feasible and to set expectations. Use a formed sample to set the specification.

Where the material actually goes

On a typical female-molded box, thickness varies systematically:

On a male (positive) mold the pattern is different: the top face is the last to form and is thinnest, while the flange and the lower walls are relatively thick.

The last row is worth attention. In a multi-feature part, the material between two deep draws is drawn toward both, and if there is not enough of it the web tears or bridges. If there is too much, the web ends up as a thick ridge that shows on the outside. Web design is one of the more delicate parts of vacuum forming a complex part, and it is why so many multi-pocket packaging tools have generously radiused webs.

Plug assist: the single most effective control

Plug assist is a mechanical helper — a male plug, usually of syntactic foam, felt, wood or aluminium, that pushes the hot sheet down into the cavity before the vacuum is applied. Its job is to pre-distribute material: it drags material from the flange area into the cavity so that the walls start the vacuum phase with more material than they would otherwise have.

The effect on thickness distribution is large. On a deep draw, plug assist typically improves the minimum wall from 45 % of nominal to 60 – 70 %.

The plug must never be so large that the sheet is pinched between plug and cavity wall — that produces a hard line and often a tear at the point of pinch.

Plug assist versus pressure forming

Pressure forming uses compressed air on the sheet's free side, with the mold vented, to push the sheet onto the mold at up to 3 – 7 bar instead of 1 bar of vacuum. It gives sharper detail, better texture reproduction and tighter radii — but it does not fundamentally change thickness distribution, because it still stretches the sheet from the same starting area. If thickness distribution is the problem, plug assist is the answer; pressure forming answers a different problem.

Pre-stretch and bubble forming

Instead of a plug, some processes pre-stretch the sheet by blowing it into a bubble before it is drawn onto the mold. Blowing a bubble thins the sheet uniformly and, more usefully, converts the sheet into a pre-shaped dome that has already used some of its stretch in a controlled way.

Bubble forming is used mainly in thin-gauge packaging and in some heavy-gauge work where the geometry is complex and a plug would mark the surface. It has two advantages: it is uniform, and it leaves no plug marks. It has one disadvantage: it is hard to control precisely, because the bubble's final shape depends on temperature distribution and on how the sheet's own weight interacts with it.

Design moves that buy thickness back

When the thickness at a corner is unacceptable, there are design changes that help more than process changes. They are worth knowing because they are free — they cost nothing to change on a drawing and they avoid tooling compromises.

In practice, the first two changes solve most thickness problems. Buyers resist them because they change the part — but a part with a 0.4 mm wall where 1.2 mm was specified is already a changed part.

How to measure thickness properly

Thickness on a formed part is measured in three different ways, and they do not agree with each other. Knowing which one the specification refers to prevents most arguments.

Ultrasonic gauges need the material's sound velocity to be set correctly. On a filled or foamed sheet the factory default will be wrong and the reading will be systematically high or low.

The measurement that matters most is the corner. A caliper cannot reach it and an ultrasonic gauge reads the tangent. The only reliable method is to cut the part through the corner and measure the section — which is why first-article approval on a deep part should always include a sectioned sample.

Writing a thickness specification that a molder can meet

A usable thickness specification has four parts:

1. Nominal sheet thickness and material, with the tolerance on the sheet itself. Sheet gauge varies — a "3 mm" sheet may be 2.85 – 3.15 mm, and that variation propagates straight through to the part.
2. Nominal formed wall — the thickness at a defined reference location, usually the flat face.
3. Minimum wall — a floor value, with the locations where it applies. This is the number that governs structural adequacy.
4. Measurement method — ultrasonic or sectioned, at named points, on a named sample size.

An example of a specification that works in practice:

That specification is achievable, measurable and arguable — which is what a specification is for. The common failure is a drawing that says only "wall thickness 2.5 mm", which is neither achievable nor measurable.

Why thickness matters more than anything else structurally

Wall thickness is the dominant structural variable in a formed part, and the reason is that stiffness scales with the cube of thickness while weight scales linearly. That relationship makes the corner — the thinnest point — the place where a part fails, and it makes small thickness gains disproportionately valuable.

Relative to a 1.0 mm wall, for the same material and the same unsupported span. The cube relationship is why a 20 % thickness gain is worth 73 % more stiffness — and why a 20 % thickness loss at a corner is far worse than it looks.

Three practical consequences follow:

• A thin corner is not a cosmetic issue. A part designed at 2 mm with a corner at 0.9 mm has 9 % of the intended stiffness at that corner. It will flex, craze and eventually crack there, and no amount of ribbing elsewhere compensates.
• Adding thickness is the least efficient way to add stiffness. Doubling the sheet thickness doubles the material cost and the heat time for an eightfold stiffness gain — but adding a rib or a crown to a flat face achieves more for far less material. Design the geometry before specifying thicker sheet.
• Ribs on a formed part are shallow and rounded. A vacuum formed rib cannot be as sharp or as deep as a moulded one — typically 1.5 – 3 mm deep with a generous root fillet — and the sheet thins as it wraps the rib. A formed rib 2 mm deep on a 2 mm sheet may be 1.1 mm at its tip. Count it as roughly half the stiffness benefit of the same rib in an injection moulded part.

A worked example: enclosure with a deep draw

A practical case to show how the numbers interact. An equipment enclosure, 400 × 300 mm footprint, 220 mm deep, HIPS, 3 mm sheet, corner radii 15 mm. The client's drawing specified "wall 3 mm minimum".

Step 1 — draw ratio. Minimum width 300 mm, depth 220 mm: draw ratio 0.73. Feasible, but not easy.

Step 2 — prediction. t = 3 × (300²) ÷ (300 + 440)² = 3 × 90,000 ÷ 547,600 = 0.49 mm at the wall. In practice, with a clamped flange, actual measurement on the first trial was 0.72 mm at the wall base, 0.61 mm mid-wall and 0.44 mm at the corner.

Step 3 — the gap. The client needed 3 mm everywhere, and the corner was at 0.44 mm — 15 % of nominal. No process change closes that gap.

Step 4 — options.

The option chosen was radius increase + plug assist + 4 mm sheet, giving a measured minimum of 1.38 mm at the corner and 2.1 mm at the wall. The drawing was rewritten to specify 2.0 mm nominal / 1.3 mm minimum rather than 3 mm everywhere.

The lesson is the general one: on any part with a draw ratio above about 0.6, the wall thickness specification has to be a distribution, not a single number. Writing it as a distribution at the design stage is much cheaper than discovering it at first article.

Frequently asked questions