Thermoforming: The 1 Bar Ceiling, the 5:1 Draw That Leaves a Fifth of the Sheet in the Wall, and the Corner That Always Runs at Half the Average

Engineering guide to vacuum and pressure thermoforming: areal draw ratio and the 1 + 4H/D law, average versus corner wall thickness, the 101 kPa vacuum ceiling and membrane stress, vent hole sizing and evacuation velocity, plug assist geometry and syntactic foam, chill marks, cooling time, draft and shrink allowances, tooling life, and India EPR and FSSAI compliance for food-contact trays.

·

Thermoforming: The 1 Bar Ceiling, the 5:1 Draw That Leaves a Fifth of the Sheet in the Wall, and the Corner That Always Runs at Half the Average

A 180 × 120 × 40 mm polypropylene food tray, 0.8 mm starting sheet, six cavities, five-second cycle. The drawing asks for a 0.45 mm wall. The first article comes back with an average wall of 0.44 mm — dead on — and a bottom corner at 0.31 mm. Two metres up the line there is a gloss mark on the rim where the sheet touched a cold insert, and a 1.8 mm web has formed between cavities three and four.

Every one of those four numbers was decided before the sheet went into the oven. The average is conservation of mass and nothing else. The corner is the average multiplied by a geometry factor nobody measures. The chill mark is a heat-transfer coefficient times a contact time. The web is two cavities competing for material that has to travel a longer path than the one between the cavities.

Thermoforming has a reputation as the low-precision end of plastics. The tool is single-sided, the pressure is one atmosphere, cycle times are measured in seconds. But the process is unusual in one respect that matters commercially: a single dimensionless ratio — formed surface area over original sheet area — predicts the mean wall thickness of the part to within a few percent, before any simulation and before any tool is cut. Once you have that ratio, the part's mass, its corner risk and the sheet you must buy all follow.

This guide walks the process in the order a quote meets it: the oven that sets the temperature field, the draw ratio that sets the average, the corner that sets the gauge, the pressure that sets the detail, the plug that redistributes without adding, the cooling that sets the cycle, and the tooling and compliance arithmetic that decide whether the job belongs in India at all.

Two processes wearing one name

The trade splits thermoforming at roughly 1.5 mm of sheet thickness, and the split is not cosmetic. Below it, sheet arrives on a roll, is indexed through an in-line oven, formed, trimmed and stacked inside one machine, and a part costs a few rupees. Above it, sheet arrives as cut blanks, is clamped in a frame by hand or robot, formed on a stationary tool, and then travels to a separate CNC for trimming.

· Thin gauge (packaging) · Heavy gauge (industrial)

Sheet thickness · 0.15 – 1.5 mm · 2.5 – 12 mm+

Feed · Continuous roll, indexed by chain rails · Individual cut blanks

Heating · In-line oven, sheet moving · Stationary oven, sheet clamped

Cycle · 2 – 10 s · 1 – 20 min

Mould temperature · 20 – 40 °C, actively cooled · 60 – 90 °C, often heated

Trimming · In-line punch or steel rule die · Separate 5-axis CNC or robot

Tooling cost · Low per cavity, many cavities · High per tool, few cavities

Typical tolerance · ±0.2 – 0.5 mm · ±0.5 – 2.0 mm

Wall thickness spread · ±10% · ±15 – 25%

The two are different businesses. A roll-fed line with 4 to 40 cavities runs 20 to 60 cycles a minute and is economic only when one product runs for weeks; changeover means new mould, new trim tool, new stacker and a new oven profile. A single-station cut-sheet machine is the least productive option in the building and the only practical one for parts over about 1.5 m or runs under a few hundred, where setup dominates.

The physical sequence is identical in both: clamp, heat, form, cool, demould, trim. Around 70–80% of the defects a shop actually fights come from the heating and cooling halves rather than the forming half, which is the opposite of what the machine's control panel suggests.

The oven is the machine

A thermoplastic sheet has to be brought into a window where its modulus has fallen by two to three orders of magnitude — from ~2 GPa to single-digit megapascals — but where it still has enough melt strength to carry its own weight across the clamp frame. That window is narrow, maybe 40 °C wide, and it is different for every resin.

Radiant ceramic heaters run at 540–650 °C on production machines and transfer heat by Stefan-Boltzmann:

With an emissivity of 0.9, a 600 °C heater facing a 25 °C sheet delivers 29 kW/m² per face; facing a sheet already at 150 °C it still delivers 28 kW/m², because the fourth power is dominated by the heater. Raising the heater to 650 °C gives 37 kW/m², to 700 °C gives 45 kW/m². That insensitivity to sheet temperature is why radiant heating is roughly constant-flux, and why the heater is controlled by distance and dwell rather than by a thermocouple on the plastic.

Constant flux has a consequence that explains the whole thin/heavy gauge divide. Heat enters both faces, so each half of the sheet must conduct the full incident flux across a distance of s/2. The quasi-steady temperature difference between surface and core is:

Polymer thermal conductivity is 0.15–0.25 W/m·K. At 29 kW/m² and k = 0.2 W/m·K:

Sheet · Energy to +140 K · Quasi-steady gradient at 29 kW/m²

0.8 mm · 153 kJ/m² · 58 K

1.2 mm · 229 kJ/m² · 87 K

3.0 mm · 662 kJ/m² · 218 K

6.0 mm · 1,323 kJ/m² · 435 K

At 1.2 mm the gradient fits inside the process window: the surface runs hot, the core catches up, and the whole sheet arrives together in a few seconds. At 6 mm the required gradient is 435 K, which is larger than the entire window from glass transition to degradation. The sheet cannot be heated at that flux without cooking the surface, so the heater is turned down, the process becomes conduction-limited, and the time constant scales as s^2/\alpha. That is the entire reason a 1 mm tray forms in seconds and a 6 mm cover takes minutes: not the forming, not the cooling alone, but the fact that heat has to diffuse through a thickness that is six times larger and a squared distance.

Two practical rules compress this. For polyolefins and PVDF in a forced-air oven, allow six minutes per millimetre of thickness; for PVC and PETG, three minutes per millimetre. Neither rule is a substitute for a surface pyrometer, but both are within a factor of 1.5 of what a correct oven profile produces, which is enough to size an oven.

The forming window itself, from the standard resin tables:

Resin · Forming window · Max areal draw ratio · Shrinkage

PS · 150 – 190 °C · 8:1 · 0.4 – 0.6%

HIPS · 160 – 205 °C · 8:1 · 0.45 – 0.65%

ABS · 150 – 205 °C · 10:1 · 0.4 – 0.8%

Plasticised PVC · 110 – 150 °C · 10:1 · 1.0 – 2.5%

LDPE · 125 – 175 °C · 6:1 · 2.5 – 3.2%

HDPE · 140 – 190 °C · 8:1 · 2.2 – 3.2%

PP · 140 – 165 °C · 6:1 · 2.0 – 2.8%

PC · 175 – 230 °C · 8:1 · 0.5 – 0.7%

Amorphous PET · 125 – 165 °C · 6:1 · 0.2 – 0.5%

PETG · 140 – 175 °C · 6:1 · 0.5 – 0.7%

The published numbers differ by 10–20 K between suppliers, and deliberately so: two-sided radiant heat of heavy-gauge sheet runs hotter than thin-gauge roll stock, and a 350 °F number from one extruder's data sheet is not a specification for another's. What matters is that the window is a band, not a point, and that the top of the band is where sag and surface defects live while the bottom is where residual stress and springback live.

The draw ratio is conservation of mass

The sheet has a fixed amount of material. Forming does not add any. Every square millimetre of formed surface has to come from somewhere on the flat sheet, so the average thickness of the formed part is fixed by the ratio of areas before anything else happens.

The reciprocal is the average retained fraction. For a cylindrical cup of diameter D and depth H, the formed area is the side plus the base, and the sheet consumed is the footprint:

That is the single most useful line in thermoforming. It says draw depth is measured in diameters, not millimetres, and that a cup twice as deep needs an area ratio five times the footprint, not two.

H/D · R_A · Average wall as % of sheet · 1.2 mm sheet becomes

0.2 · 1.80 · 55.6% · 0.67 mm

0.3 · 2.20 · 45.5% · 0.55 mm

0.5 · 3.00 · 33.3% · 0.40 mm

0.75 · 4.00 · 25.0% · 0.30 mm

1.0 · 5.00 · 20.0% · 0.24 mm

1.5 · 7.00 · 14.3% · 0.17 mm

2.0 · 9.00 · 11.1% · 0.13 mm

A square tray is computed the same way, from its own areas. A 200 × 300 × 40 mm tray has 1,000 cm² of formed surface over a 600 cm² footprint — R_A = 1.67, average wall 60% of sheet. Push the depth to 80 mm and R_A climbs to 2.33 and the average falls to 43%. A 100 mm cube, R_A = 5.

The material's maximum draw ratio is a real constraint, and it correlates with hot strength rather than with elongation at room temperature. Amorphous resins with broad rubbery plateaus — ABS, HIPS, plasticised PVC — tolerate 8:1 to 10:1. Semi-crystalline resins — PP, HDPE, PET — give 6:1 to 8:1 because the crystal phase holds the sheet together until it melts, then lets go all at once. Anything above about 3:1 is an engineering problem, not a drawing call: it needs plug assist, pressure forming, or a different part.

The corner is not the average

The average is where the arguments start. It is also the least useful number on the part, because no point on the part has to be the average. Material distribution is set by the order in which the sheet touches the tool, and the corners are last.

A bottom corner of a deep tray is the only place where the sheet is stretched in two directions simultaneously: it has to travel down the wall and around the corner radius at the same time, at the point furthest from the clamp, in material that has already been chilled by contact with the sidewall on the way down. Production data on single-cavity female forming puts corner thickness at 25–40% of original on aggressive draws and 30–50% of the average on ordinary ones. A flat bottom may hold 60–80%; the flange holds 90–100% because it never stretches at all.

That produces the gauge inversion every thermoformer runs and every buyer is surprised by:

where f_{corner} is the corner's fraction of the average. For a tray at R_A = 2.3, a 0.30 mm corner minimum and a corner factor of 0.45:

You cannot buy 1.0 mm sheet and hit a 0.30 mm corner, however good the tool is, because the corner does not negotiate. It gets what the distribution gives it. The consequence is that the sheet gauge is set by the thinnest point the specification cares about, and the cost of the part is set by the gauge. This is why the most valuable conversation at quoting stage is not about the mould — it is about whether the corner really needs 0.30 mm, or whether 0.25 mm with a 6 mm corner radius would carry the load.

Opening radii is free and it is the change most often available and least often made. Minimum inside corner radius should be one sheet thickness, preferred two to three; a 5° draft instead of 1° cuts the stretch on a 100 mm wall by about 7%; and rounding a square outline reduces corner thinning by 15–30% for no tooling cost at all.

The 1 bar ceiling

Vacuum forming has one hard limit and it is not a machine setting. The driving pressure is the difference between atmosphere and whatever the pump can leave in the cavity, and the pump can leave essentially nothing. So the forming pressure is 101.3 kPa — 14.7 psi, one bar, and not a pascal more.

Spread over a large area that is not trivial. One square metre of sheet at 101 kPa sees 101 kN, about 10.3 tonne-force. But spread over a thick sheet it becomes small, because the stress that actually deforms the material is a membrane stress that scales with radius over thickness:

Condition · p · t = 0.8 mm · t = 2 mm · t = 6 mm

Vacuum · 0.101 MPa · 1.27 MPa · 0.51 MPa · 0.17 MPa

Low pressure · 0.30 MPa · 3.75 MPa · 1.50 MPa · 0.50 MPa

Pressure forming · 0.70 MPa · 8.75 MPa · 3.50 MPa · 1.17 MPa

High pressure · 1.00 MPa · 12.5 MPa · 5.00 MPa · 1.67 MPa

(at R = 20 mm)

Read the table across rather than down. Vacuum alone puts 1.3 MPa into a 0.8 mm sheet at a 20 mm radius and 0.17 MPa into a 6 mm sheet at the same radius — a factor of seven, purely from thickness. That is why heavy-gauge work cannot reproduce fine texture under vacuum, why it needs radii measured in tens of millimetres, and why the same tool that produces crisp detail in 1 mm PETG produces mush in 4 mm acrylic.

Pressure forming closes the gap by putting compressed air on the back of the sheet. Industrial machines run 3–7 bar, and deep-detail systems go to 10 bar, which is four to ten times the vacuum ceiling. It buys sharper radii, moulded-in texture and lettering, and roughly double the dimensional precision, at the cost of a pressure box, a sealed upper platen and a tool that must withstand being a pressure vessel.

The second problem with vacuum is that it destroys itself at the worst moment. The instant the sheet seals against the mould face, the remaining cavity volume is isolated, and the pressure differential across the sheet is set by how fast that trapped air can escape through the vent holes. Not by the pump. A 0.5 m² mould vented on a 40 mm grid with 0.8 mm holes has 312 holes and 157 mm² of total flow area. Evacuating 5 litres through that in 0.3 s requires an average velocity of:

106 m/s through a 0.8 mm hole is a choked, compressible flow with a large pressure drop. The hole is not a pipe, it is a restriction, and the real cavity pressure at that instant is well above the pump's ultimate vacuum. That is the physical origin of two defects that are usually blamed on temperature: incomplete draw into corners, and chill marks where a partially-formed sheet dwells against a cold mould.

Vent sizing is therefore a design decision with a table:

Material / gauge · Vent diameter · Spacing

Thin gauge < 1.5 mm, HIPS, PET · 0.5 – 0.8 mm · 25 – 50 mm in detail zones

Medium 1.5 – 3 mm, ABS, PETG · 0.8 – 1.0 mm · 40 – 60 mm

Heavy 3 – 6 mm, ABS, acrylic · 1.0 – 1.2 mm · 50 – 75 mm

Very heavy > 6 mm, PC · 1.2 – 1.5 mm · 50 – 75 mm

Polycarbonate, any gauge · ≤ 0.65 mm (dimpling) · back-drill to 3 mm

Holes go where the sheet touches last — the deepest point of every pocket, the base of every wall, the centre of every large flat, every sharp internal corner. Putting them on cosmetic faces is how a witness mark ends up in the customer's first article. Every hole should be back-drilled to a larger diameter from the reverse side so the restriction is short. Slots have less tendency to dimple than round holes of the same area and are worth using where the trim line will remove them.

Plug assist redistributes; it does not add

Plug assist is the standard answer for anything deeper than about 0.7:1, and more than 90% of thin-gauge parts are formed with it. A mechanical plug — usually a syntactic foam, acetal or lacquered felt — descends into the softened sheet and pre-stretches it before pressure is applied.

The geometry rules are simple enough to hold in your head:

What the plug does is change the sequence of stretching. It pushes material down toward the base and the corners before the sheet has a chance to bridge across the cavity mouth, so the material that would have become a thick flange becomes a thick corner instead. What it cannot do is change the average. That is fixed by the areal ratio and no tool geometry moves it by a gram.

Plug material matters more than plug shape, for a reason that is thermal rather than mechanical. The plug is in contact with sheet at 150–170 °C, and every degree it removes locally raises the sheet's viscosity exactly where you are asking it to stretch. Syntactic foam — hollow glass microspheres in an epoxy matrix, roughly 50% air by volume — has a thermal conductivity low enough that a working plug settles at an equilibrium temperature near 100 °C against a 150 °C sheet, while solid aluminium would run much closer to sheet temperature and chill it on contact. That is the whole selection criterion: low conductivity, low friction, machinable, durable.

The 2010 Polymer Engineering & Science study on plug-assisted polystyrene forming is worth knowing because it inverts the usual priorities: plug displacement, sheet temperature, plug temperature and plug shape dominated the wall-thickness distribution, while heat transfer was a secondary effect. The dominant mechanism was temperature-dependent friction between plug and sheet. Friction sets how much the sheet slips over the plug, and slip sets how much material reaches the base. That is why a polished steel plug and a felt plug of identical geometry produce different parts, and why swapping plug material without re-optimising geometry produces rejects.

Chill marks and webbing

A chill mark is a local gloss or texture defect where the sheet touched the mould while still hot enough to take a impression, then froze before it could conform. It is a heat-transfer problem with a short time constant:

For a 1 mm sheet at 160 °C against an 80 °C mould, \rho = 1050 kg/m³ and c_p = 1300 J/kg·K, a contact coefficient of 500 W/m²·K and 0.5 s of contact gives 14.7 K of local cooling. Half a second costs fifteen degrees. At 1.0 s it is 29 K, which is most of the usable window, and the material there has effectively stopped stretching.

That is the arithmetic behind two machine specifications that otherwise look like marketing. Servo-driven platens move at 700–1,000 mm/s; pneumatic drives reach 250–300 mm/s. The time between first contact and full vacuum is the exposure, and a machine that halves it halves the chill. It is also why colder moulds require faster tables, why hot moulds (60–90 °C on heavy-gauge machines) tolerate slower ones, and why central chilled-air cooling on a heavy-gauge machine shortens the cycle by 10% for 20 K of air temperature and 25% for 40 K.

Webbing is different: it is excess material bridging between two adjacent deep features, forming a sheet-like web that has to be cut out by hand. It is a geometry problem before it is a setting. Two tall cavities close together give the sheet a short path between them that is easier to take than the long path down both walls. Remedies, in order of effectiveness: increase the gap between features, open the inside corners, reduce the height difference between adjacent draws, orient the part so the draw is along the machine direction of the sheet, and as a last resort relocate the web into a zone that trims away. Raising sheet temperature makes webbing worse, not better, because it lowers the resistance of the short path.

Cooling sets the cycle, and the cycle sets the price

Cooling is pure conduction and it obeys a square law. For a sheet cooled from both faces to a demould temperature, the classical solution gives:

with \alpha the thermal diffusivity. For most unfilled thermoplastics \alpha is 0.05–0.25 mm²/s; 0.1 mm²/s is a fair working value. Taking a 165 °C sheet, a 40 °C mould and an 85 °C demould temperature, the logarithmic term evaluates to 0.812:

Sheet · Calculated cooling · As run, with air assist

0.5 mm · 0.2 s · 2 – 3 s

1.0 mm · 0.8 s · 4 – 6 s

2.0 mm · 3.3 s · 10 – 15 s

3.0 mm · 7.4 s · 18 – 26 s

4.0 mm · 13.2 s · 30 – 45 s

6.0 mm · 29.6 s · 60 – 120 s

The gap between the calculated and the as-run column is the honest part of this table. The Fourier solution assumes perfect thermal contact, instantaneous pressure and a demould criterion based on the average temperature. Real parts have contact resistance at the mould face, a sheet that is still forming while it cools, and a demould criterion based on the hottest point rather than the average. A factor of two to four between theory and practice is normal, and it grows with thickness because thicker parts are also stiffer and more likely to be demoulded conservatively.

This is where the money is. A 600 × 400 × 120 mm ABS cover at 3 mm runs roughly 12 parts an hour on a single-cavity machine; the same footprint at 1.2 mm on a six-cavity roll-fed line runs over 4,000. The material cost difference is only 3:1. The machine-hour cost difference is 100:1, and it is entirely cooling and handling.

Tooling: what to make it from, and how long it lasts

The tool only ever sees one atmosphere in vacuum forming, so it does not need to be strong. It needs to be thermally right, dimensionally stable at 60–90 °C, and vented.

Volume · Tool material · Tool life · Notes

1 – 50 · Wood, MDF, tooling board · 1 – 20 · Fast, cheap, moves with humidity

50 – 1,000 · Filled epoxy, cast resin · 100 – 1,000 · Good surface, poor conductivity

1,000 – 100,000 · Cast aluminium · 10,000+ · Best compromise

100,000+ · Machined aluminium, temperature-controlled · 500,000 · Cooling channels essential

Aluminium is the default for a reason that is thermal, not structural: it pulls heat out of the sheet, which shortens the cycle and reduces residual stress. A low-conductivity epoxy tool forms the same shape and produces a part that cools slowly and warps more.

The design rules that decide whether the tool works:

Heavy-gauge tolerances are not thin-gauge tolerances and should not be quoted as if they were. Across a 300 mm face, expect ±0.3 mm on thin gauge and 1.0–2.0 mm on heavy; wall thickness ±10% versus ±15–25%; trim position ±0.2 mm with hard tooling versus ±0.5 mm with CNC and ±1.5 mm by hand.

Trimming is a separate cost centre and often the surprise. Thin gauge trims in-line with a steel rule die or a punch — effectively free per part once the die exists. Heavy gauge goes to a 5-axis router or a robot, and that operation can cost more than the forming.

What the part costs

Two worked examples with 2026 Indian trade prices. Sheet: PP at ₹135/kg, HIPS ₹130–175/kg, PETG ₹390–450/kg, ABS ₹175/kg — all indicative ranges from sheet suppliers, and all negotiable at volume.

A. 180 × 120 × 40 mm PP food tray, 0.8 mm, 6 cavities, 5 s cycle. Net part 15.7 g. At a 200 × 140 mm pitch the skeleton is 23%, so 20.4 g of sheet per part, ₹2.76 of material. Machine at ₹1,500/h over 4,320 parts/h is ₹0.35. A ₹1,80,000 tool amortised over 200,000 parts is ₹0.90.

B. 600 × 400 × 120 mm ABS cover, 3 mm, 1 cavity, 5 min cycle. Net 756 g before thinning; allow 25% for trim and draw, so 945 g at ₹175/kg is ₹165. Machine at ₹900/h over 12 parts/h is ₹75. A ₹1,20,000 tool over 5,000 parts is ₹24.

The two examples are the same process and completely different businesses. In A the material is 69% of cost and the tool is 22%. In B the material is 62%, the machine 28%, and the tool 9%. Neither is dominated by labour, which is why thermoforming survived the shift of low-volume plastics work offshore better than machining did — and why the break-even against injection moulding sits in the hundreds-to-tens-of-thousands range rather than at a fixed number.

The Indian regulatory layer

Thermoforming is the process behind most of the rigid food packaging sold in India, and two regulatory regimes now sit on top of it.

Plastic Waste Management Rules, as amended in March 2026. Rigid plastic packaging is Category I, and it carries a mandatory recycled-content obligation: 30% in 2025-26, 40% in 2026-27, 50% in 2027-28 and 60% from 2028-29. Flexible packaging is 10/10/20/20. The targets do not apply where another regulator — FSSAI, CDSCO, the Central Insecticides Board, or a mandatory Indian Standard — prohibits recycled plastic in that application, and a food-contact shortfall from 2025-26 can be carried forward for up to three years provided at least a third is made good each year.

For a thermoformer this is not paperwork; it is a materials decision. Recycled PP and HIPS for industrial trays is available and increasingly specified. Recycled material for direct food contact is a different question, and the exception is narrower than the market assumes.

FSSAI packaging regulation. Any material in direct contact with food must be food grade, must conform to the relevant Indian Standard, and must pass an overall migration limit of 60 mg/kg or 10 mg/dm² per IS 9845 with no visible colour migration. Specific migration limits apply to metals — barium 1.0 mg/kg, cobalt 0.05, copper 5.0, iron 48, lithium 0.6, manganese 0.6, zinc 25. Recycled PET is permitted for food packaging only as and when standards and guidelines are notified, and packaging must be marked accordingly.

The practical consequence for a job shop is that "food grade" is a chain, not a property: the resin grade, the regrind fraction, the colour masterbatch, the release agent and the printing ink all have to be inside the declaration. A tray formed from virgin sheet on a tool that was last used with an industrial masterbatch is not automatically food grade.

Scale. The Indian thermoform packaging market was estimated at US1.43 billion in 2025, growing at 7.7% to about US2.59 billion by 2033, with PET the largest material segment. Globally the thermoforming packaging market is put near US55 billion and the thermoforming machine market near US1.6 billion, with Asia-Pacific the largest region at roughly US$647 million. Machine building is domestic too — the extrusion and thermoforming lines from Rajkot ship to 78 countries, which means a shop buying a roll-fed line in Coimbatore or Ahmedabad can buy it with Indian service.

Design rules worth putting on the drawing

Rule · Value · Why

Draw ratio · ≤ 2:1 areal for vacuum only · Beyond it, plug assist or pressure forming

Depth to width · ≤ 0.5 unassisted female · Deeper needs pre-stretch

Inside corner radius · ≥ 1 × sheet thickness, 2–3 × preferred · Corner is where thinning concentrates

Draft, female · 3° minimum · Shrink grips a male tool, releases a female

Draft, male · 5° minimum · Part shrinks onto the tool

Wall tolerance · ±10% thin gauge, ±15–25% heavy · Do not call out a nominal wall without a corner minimum

Flatness · 0.5 mm / 300 mm thin, 1.5–3.0 mm heavy · Residual stress

Rib depth · ≤ 3 × width, generous root radius · Ribs starve the adjacent wall

Texture · +2–3° draft · Texture is draft in reverse

Undercut · ≤ 0.5 mm, else split tool · Single-sided tool has no ejection mechanism

And on the RFQ, four numbers that decide the price and are almost never supplied: the minimum wall at the worst corner, the resin grade and its food-contact status, the acceptable trim method, and the annual volume with its expected life. A thermoformer quoting from a 3D model alone will quote the average, and the average is the one number the part does not have to meet.

What the arithmetic actually says

Thermoforming's reputation for low precision is earned at the corner and nowhere else. The average wall of a formed part is fixed by conservation of mass and predictable to a few percent from a single area ratio — 1 + 4H/D for a cup, the same idea for anything else. The forming pressure is capped at 101 kPa and nothing about the machine changes that, which is why thick sheet needs big radii and deep detail needs compressed air. The plug redistributes material but cannot add it. The corner gets 25–50% of the average because it is last, coldest, and stretched in two directions. And the cycle is a square law in thickness, which is why the same process produces a ₹4 tray and a ₹264 cover.

None of that is discovered at first article. All of it is available at quoting, from a footprint, a depth and a sheet gauge — which is the difference between a price that holds and a price that gets renegotiated after the tool is cut.

More FabFlow blog posts