Material selection is usually treated as a box to tick after the part is designed, when in fact it is the decision that fixes the tool. The resin sets the shrinkage the cavity has to be cut for, the mold temperature the cooling system has to reach, the steel that will survive a million shots, the drying that decides whether the parts are strong or merely the right shape, and the tolerance the tool can hold for as long as it exists. Choose it late, or choose it by price per kilo, and everything downstream is a correction.
The four questions that actually decide the resin
Most of the shortlist is settled by four questions, and none of them is “which plastic is best”, because there is no such thing. The questions are: what does the part have to survive, what does it have to hold, what will the mold have to do to make it, and what can you actually buy.
The environment covers temperature at both ends, chemicals including cleaning agents, UV, and time. A part that sees 70 °C in a car interior in summer and 20 °C below zero in winter is a different requirement from one that lives indoors, and the two ends rule out different resins.
The load is about stiffness, creep and impact, and the three point at different materials. A part that must not deflect under a sustained load at 60 °C needs either glass or a higher family; a part that must survive a drop needs impact strength, which glass reduces.
The process is the part most often skipped. The resin determines the melt temperature, the mold temperature, the shrinkage, whether it needs drying, whether it corrodes or abrades the steel, and how it behaves in a thin wall. All of that is fixed before the mold is designed, and all of it costs money to change afterwards.
Availability is unglamorous and decisive. A grade that exists, in the color, in the volume, with a local distributor, beats a better grade that arrives in twelve weeks with a minimum order larger than the annual requirement.
The order matters. Environment and load narrow the choice to a family; the process consequences decide whether the tool can be built for it at a sensible price; availability decides whether the answer is real. Teams that start at availability, or at the resin their last product used, usually end up at the first question later, with the tool already cut.
The five families, and what each is for
It is worth knowing the families rather than a list of forty grades, because within a family the trade-offs are small and between families they are large. Almost every selection problem is really a question of which family, followed by which grade inside it.
| Family | Examples | What it is good at | Where it runs out | What it does to the tool |
|---|---|---|---|---|
| Commodity | PP, HDPE, PS, rigid PVC | Cheap per part, forgiving to mold, chemically inert (PP and PE especially) | Heat, stiffness, and UV unless stabilized | Soft on steel. PVC is the exception and attacks it |
| Engineering | ABS, PC, PA6 and PA66, POM, PBT, PMMA | Stiffness and toughness at moderate temperature | Above 100 to 120 °C continuous, and most of them absorb water | Higher melt temperatures, tighter windows, and several are corrosive or deposit on vents |
| High temperature | PPS, PSU, PES, PEI, PEEK | Continuous service above 150 °C, with dimensional stability | Price, and the mold temperatures they demand | Mold temperatures of 130 to 200 °C: a chiller will not do it, and the tool needs oil and often cartridge heaters |
| Elastomer | TPE, TPU, TPV | Grip, sealing, and overmolding onto a rigid substrate | Creep under load, and heat | Different shrinkage from the substrate, and the bond is mechanical or chemical rather than automatic |
| Filled and reinforced | PP-GF30, PA66-GF30, PBT-GF30, mineral-filled grades | Stiffness and creep resistance at temperature, without switching family | Anisotropy, and the wear they put on the tool | Abrasive: gates, hot runner tips and vent lands wear out faster than on any unfilled grade |
Two rows in that table drive most of the tooling consequences on this page. The high temperature family is not expensive because of the polymer alone; it is expensive because a mold at 150 °C is a different machine. With water at atmospheric pressure you cannot exceed about 90 °C at the tool, so anything above that needs a pressurized water unit or oil, and the cooling design has to be laid out for it from the start rather than adapted afterwards. That is the subject of mold cooling design, and it is the reason a resin swap late in a project is rarely a resin swap alone.
The filled and reinforced family is where most of the surprises are, because glass is sold as a way to get stiffness without changing material, and it does that well — while quietly changing shrinkage, warpage, wear and appearance at the same time.
Shrinkage: the number that shapes the tool
Every cavity is cut oversize by the shrinkage factor, which means the resin choice is the first number on the tool drawing. What matters for the finished part is not the average shrinkage but the width of the band: a resin that shrinks 2 % plus or minus 0.1 can be held to a tight tolerance, while one that shrinks anywhere between 1.0 and 2.5 % cannot be held to anything better than the band allows, however carefully the mold is made.
| Resin | Shrinkage, flow direction | What that means for tolerance | Note |
|---|---|---|---|
| PP | 1.0–2.5 % | Wide band, and the mold temperature moves it | Crystallinity depends on how fast the part cools, so a cold tool and a hot tool give different parts |
| HDPE | 1.5–3.0 % | The widest of the common resins | Cut the cavity oversize by a large factor and accept a broad band |
| ABS | 0.4–0.7 % | Tight, and predictable | Amorphous, which is exactly why it is used for visible housings |
| PS | 0.4–0.7 % | Tight | Amorphous and stiff, but brittle |
| PC | 0.5–0.8 % | Tight | Amorphous; the band stays narrow across a wide process window |
| PA6 | 0.6–1.5 % | Medium, and it moves after molding | The dry dimension is not the service dimension: the part grows as it takes up moisture |
| PA66 | 0.8–1.5 % | Medium | Same as PA6, slightly less |
| POM | 1.8–2.5 % | Medium to wide, but consistent | High and repeatable; the cavity is cut much oversize and the result is predictable |
| PBT | 0.4–1.0 % | Tight | 0.2–0.5 % with 30 % glass, at the cost described below |
| PMMA | 0.2–0.8 % | Tight | Optical grades, so the surface matters as much as the dimension |
| Rigid PVC | 0.2–0.6 % | Tight | Corrosive, so the steel matters more than the shrinkage |
The practical rule is that a resin with a wide shrinkage band cannot deliver a tight tolerance, and no amount of tool precision changes that. This is the same conclusion as in injection molding tolerances, seen from the other end: there, the question is what to put on the drawing; here, it is which resin makes the drawing achievable in the first place.
Two entries deserve a note. PA6 and PA66 are dimensionally stable in the dry state and grow afterwards as they absorb moisture, which means the dimension measured at the press is not the dimension in service. A part that must fit a mating part a month later has to be specified wet, or the design has to tolerate the growth. POM shrinks a great deal but does it consistently, which makes it easier to work with than its number suggests: a predictable 2 % is easier to cut for than an unpredictable 1 %.
Why adding glass does not make a part more accurate
Glass fiber reduces shrinkage, and the reduction is real. What it also does is make shrinkage direction-dependent, because the fibers align with the flow and restrain contraction along their length far more than across it. The result is a part that shrinks less overall and less evenly, which on a large part can be worse than the unfilled grade it replaced.
| Grade | Shrinkage along flow | Shrinkage across flow | What it does to a 200 mm part |
|---|---|---|---|
| Unfilled PP | 1.5–2.0 % | 1.5–2.0 % | Shrinks a lot, but equally in both directions, so it stays close to its intended shape |
| PP-GF30 | 0.4–0.8 % | 1.0–1.5 % | About 1.5 mm difference between the two directions: a box that is measurably narrower along the flow |
| PA66-GF30 | 0.3–0.5 % | 0.8–1.2 % | About 1.2 mm: a flat panel that dishes, because one direction pulls more than the other |
| PBT-GF30 | 0.2–0.5 % | 0.6–0.9 % | About 0.9 mm, and the reason glass-filled PBT connectors are designed with ribs rather than flat walls |
The numbers in the last column are the argument. On a 200 mm part, a glass-filled grade can differ by more than a millimeter between the direction the melt flowed and the direction across it, while the unfilled grade that shrinks twice as much does so equally in both directions and therefore stays closer to its designed shape. A flat panel in PA66-GF30 dishes; the same panel in unfilled PA66 stays flat and is simply bigger than the cavity.
That is why glass-filled parts are designed with ribs and crowned surfaces rather than flat walls, and why the gate position becomes a dimensional decision rather than a cosmetic one. It is also why the DFM review has to know the grade before it can say anything useful about warpage.
Moisture: the mistake that cannot be undone
This is the single most expensive routine error in injection molding, and it is expensive because it is invisible. Moisture in the melt does not produce a weak part that can be re-run; it produces a part whose polymer chains have already been cut, and no subsequent operation repairs that.
| Resin | Water absorbed in 24 h | At saturation | Drying before molding | What happens if it is skipped |
|---|---|---|---|---|
| PA6 | 1.3–1.9 % | 8–10 % | 80 °C for 4 h, or 100 °C for 2 h, desiccant dryer | Hydrolysis: the polymer chain is cut, impact strength falls permanently, and re-drying does not restore it |
| PA66 | 1.0–1.5 % | about 8.5 % | The same | The same, and it is the most common expensive mistake in the whole list |
| PC | about 0.15 % | about 0.35 % | 120 °C for 4 h | Splay and silver streaks, and a measurable drop in impact strength |
| PBT | about 0.1 % | about 0.4 % | 120 °C for 4 h | Hydrolysis in the barrel, and parts that crack instead of bending |
| PET | about 0.1 % | about 0.4 % | 120 to 140 °C, dehumidified | The same, and faster, because PET hydrolyses more readily than PBT |
| ABS | 0.2–0.45 % | 1.0–1.5 % | 80 °C for 2 to 4 h | Silver streaks, and a weld line that is weaker than the design assumed |
| PMMA | about 0.3 % | about 2.0 % | 80 °C for 3 h | Bubbles in an optical part, which is scrap rather than a repair |
| POM | about 0.2 % | about 0.8 % | 80 °C for 2 to 3 h, and not hotter | Formaldehyde release and a smell; overheating in the dryer is worse than the moisture it was meant to remove |
| PP, HDPE | under 0.01 % | under 0.02 % | Normally not required | Nothing, which is a large part of why the cheapest resins are also the most forgiving |
The mechanism is hydrolysis. Water at melt temperature reacts with the polymer chain and breaks it, reducing molecular weight. The part comes out with the right shape, the right color and a normal surface, and fails a drop test it should have passed. Because the damage is chemical rather than visible, it is usually attributed to the design or to the molder’s process, and the actual cause — a dryer that was set to 80 °C when the resin needed 120 °C, or a hopper left open overnight — is never found.
The row worth reading twice is the last one. PP and HDPE absorb essentially nothing, need no drying, and tolerate a molder with modest equipment. That is a real advantage, and it is not about mechanical properties at all: it is about how much of the process can go wrong without consequence. A specification that chooses PA66 for a part that PP could have made has bought stiffness and accepted a permanent dependency on drying discipline.
Heat: two different numbers that get confused
Every datasheet gives a heat deflection temperature, and most selection decisions use it as if it were the service temperature. It is not, and the gap between the two columns in the table below is where a lot of field failures come from.
| Resin | HDT at 1.82 MPa | Continuous service, no load | Mold temperature needed | What that does to the tool |
|---|---|---|---|---|
| PP | 50–60 °C | about 100 °C | 20–60 °C | A chiller and standard circuits; nothing special is required |
| ABS | 80–100 °C | 70–85 °C | 40–80 °C | Standard |
| PC | 125–135 °C | 115–130 °C | 80–120 °C | A hot water unit, and a cooling layout designed for it rather than inherited |
| PA66 | 70–90 °C | 100–120 °C | 60–90 °C | Standard, but a low mold temperature costs crystallinity and therefore strength |
| POM | 90–110 °C | 90–100 °C | 80–110 °C | Hot water |
| PBT, unfilled | 55–65 °C | 120–140 °C | 60–90 °C | Standard |
| PBT-GF30 | 200–210 °C | about 140 °C | 60–90 °C | The glass closes the gap between the two temperature columns |
| PET | 70–80 °C | 100–120 °C | 100–140 °C | A hot mold is not optional: below about 100 °C the part stays amorphous and soft |
| PPS | about 135 °C unfilled | 200–220 °C | 130–150 °C | An oil unit, and a different class of tool design problem |
| PEEK | 150 °C unfilled, 315 °C with glass | about 250 °C | 160–200 °C | Oil at the top of its range, plus cartridge heaters in the tool |
HDT is measured under a load — 1.82 MPa in the figures below — and it is the temperature at which a standard test bar deflects a given amount under that load. Continuous service temperature is what the material tolerates with no load at all. So PBT shows an HDT of 55 to 65 °C and a continuous service temperature of 120 to 140 °C, and both numbers are correct; they answer different questions.
The consequence is practical. A part that must hold a shape under load at 100 °C cannot be made from unfilled PBT, even though the same grade is fine in the same application at 120 °C with no load on it. Glass changes the first number dramatically — PBT-GF30 goes from about 60 °C to over 200 °C — which is why the filled grades exist, and why the second column barely moves.
The right-hand column is the tooling consequence, and it is easy to miss. A resin that needs 100 °C of mold temperature cannot be run on a chiller, and a tool built with standard circuits and quick couplings will not reach it. The cooling system has to be specified for it, and the press needs a hot water or oil unit that the molder may not own.
Chemical resistance: the two that catch people
Chemical resistance is usually checked against the product the part will contain, and almost never against what the part will be cleaned with. The second is the more common failure, because cleaning agents are aggressive by design and are applied to a part that is under stress from assembly.
| Medium | Attacks | What survives it | How the failure shows up |
|---|---|---|---|
| Alkaline cleaners and ammonia | PC, and to a lesser extent ABS | PP, PE, POM | Stress cracking weeks after assembly, starting at a molded-in stress point |
| Acids, and acidic food | PA, and POM at higher concentrations | PP | Loss of strength that gets diagnosed as a design error, because nothing visible has changed |
| Alcohols and fuels | PA swells, PS crazes | POM, PBT, PP | A part that fits when dry and jams a week later |
| Chlorinated solvents | Most thermoplastics, POM quickest | PP at low stress, briefly | Cleaning the part destroys it, and the cleaner gets blamed |
| Oils and greases, hot | PP and PE swell at temperature | PA, PBT, POM | Slow swelling that appears as a loose fit after a season in service |
| UV and weathering | PP, PE and ABS without a stabilizer | ASA, PMMA, PC with a coating | Chalking and color shift within one outdoor season |
The two rows that cause the most trouble are the first and the third. Polycarbonate under alkaline cleaners is a classic: the part passes every mechanical test, is assembled with a screw boss under load, is wiped down with an ammonia-based cleaner in the factory, and cracks a fortnight later at the boss. Nothing was wrong with the material or the design; the combination was. PA in alcohols and fuels is the other, and it presents as a fit problem rather than a material problem, because the part swells gradually and stops fitting rather than breaking.
Where a part is both stressed and cleaned, the safe combination is a semi-crystalline resin with good chemical resistance — PP, POM or PBT — and a design that keeps molded-in stress low. That last part matters as much as the resin: stress cracking needs both a chemical and a stress, and removing either one removes the failure.
What the resin does to the tool
This is the section that connects the resin to the invoice, and it is the one most often left out of a material discussion. The same part geometry, in a different resin, can need a different steel, a different heat treatment, a different venting routine and a different maintenance interval.
| Resin | Corrosive? | Abrasive? | Steel that holds up | What it does to venting |
|---|---|---|---|---|
| PP, PE, PS, ABS | No | No | P20 is enough | Standard venting and standard cleaning |
| PC, PMMA | No | No | P20 or 718H | Volatiles leave a pale deposit on the vent land |
| PA, POM | Mildly, from decomposition products | No | 718H, or S136 if the tool runs hot | POM releases formaldehyde when overheated, and the vents block faster than on any other common resin |
| PBT, PET | Mildly | No | 718H | Standard, but drying failures show up as splay rather than as a burning part |
| Rigid PVC | Yes, aggressively | No | S136 or a similar stainless, hardened | The decomposition products are acidic and attack both the steel and the vent lands |
| Flame-retardant grades | Often yes | Sometimes | S136 hardened, or a coating | Several flame-retardant systems release acids that corrode the lands and the water circuits |
| Glass-filled | No | Severely | H13, or S136 hardened past 50 HRC, with carbide at the gate | Abrasion widens the land until it flashes instead of venting |
| PPS, PEEK | Mildly at melt temperature | Yes, when filled | H13 or harder | Mold temperatures above 130 °C change the whole venting picture |
Three rows are worth acting on immediately. Rigid PVC releases hydrogen chloride as it decomposes, which attacks steel, so a PVC tool is built in a stainless grade such as S136 and hardened, and the tool is cleaned far more often than one running PP. Putting PVC through a P20 tool will work for a while and then show up as pitting on the cavity surface. Glass-filled grades wear the places where flow is fastest: gates, hot runner tips and vent lands, in that order of severity. That is a mold steel selection consequence, not a processing one, and it is cheaper to decide at the resin stage than to re-cut a gate later.
The third is venting. Every resin that leaves a deposit — POM, PVC, flame-retardant systems, and to a lesser degree PC and PA — closes its own vent lands faster than PP does, and needs a cleaning interval written into the tool documentation. What that interval is, and what happens when it is missed, is set out in mold venting.
Cost is not the price per kilo
The comparison that decides most projects is price per kilogram of resin, and it is the wrong comparison, because parts are sold by the piece. Four other numbers sit between the resin price and the cost of the part, and any of them can reverse the ranking.
| What you are comparing | The number that actually matters | Why the price per kilo misleads |
|---|---|---|
| Resin price | Price per kilogram | Parts are sold by the piece, not by the kilogram, and the piece weighs what the density says it weighs |
| Density | Grams per cubic centimeter | POM at 1.41 and PP at 0.905 differ by more than half in mass for an identical part |
| Cycle time | Seconds per shot | A resin that needs a hotter mold and a longer cool adds to every part for the whole life of the tool |
| Scrap rate | Percent | A hygroscopic resin with a marginal dryer produces scrap continuously, and the resin price never shows it |
| Tool cost | One-off | A corrosive or abrasive resin changes the steel, and that changes the tool price before the first part is made |
| Tool life | Shots before refurbishment | An abrasive grade can halve the life of the same tool, which is a cost per part rather than a cost per tool |
Density is the one that surprises people. Two identical parts, one in POM at 1.41 g/cm³ and one in PP at 0.905, differ by more than half in mass, so a resin that looks cheaper per kilogram can be more expensive per part. Cycle time is the second: a grade that needs a mold at 100 °C and a longer cool adds a few seconds to every shot for the life of the tool, which on a multi-cavity tool is the largest number in the calculation. The full structure of what a tool and a part cost, including the parts of it that are not resin at all, is in injection mold cost.
The last two rows are the ones with the longest shadow. A corrosive resin changes the steel, which changes the tool price before a single part exists; an abrasive resin shortens the tool’s life, which is a cost per part rather than a cost per tool. Neither appears in a resin quote, and both are decided by the resin choice.
Six substitutions that usually work
Most selection work is not choosing from a blank sheet; it is adjusting an existing choice because of price, supply, or a regulation. These are the substitutions that are usually safe, with what they cost you stated honestly.
| From | To | What you gain | What you give up |
|---|---|---|---|
| ABS | PP | Chemical resistance, lower cost, and no drying at all | Stiffness, gloss, and dimensional stability |
| ABS | PC/ABS | Heat resistance and impact strength | Cost, and drying becomes mandatory |
| PA66 | PA6 | Easier processing and a wider window | A little heat resistance and stiffness |
| PA66 | PBT | Dimensional stability, and no growth after molding | Impact strength at low temperature |
| POM | PBT | Better dimensional stability and no formaldehyde risk | Creep resistance and lubricity |
| PC | PMMA | Optical clarity and UV stability without a coating | Impact strength, dramatically |
Two of them are worth expanding. ABS to PP is the most common cost reduction in consumer goods, and it works because PP brings chemical resistance and no drying requirement along with the lower cost. What it does not bring is stiffness or gloss, so a housing that drops from ABS to PP usually needs a thicker wall or a rib pattern, which is a design change rather than a material change. PA66 to PBT is the standard answer to a moisture problem: PBT absorbs a tenth of what PA66 does, so the part stops growing after molding, and the price is impact strength at low temperature, which matters for anything that might be dropped in winter.
What is not on the list is anything that crosses from amorphous to semi-crystalline without a design review. That change moves shrinkage by a factor of two or three, and the tool has already been cut.
A worked example: a power tool housing
The requirements, as a customer would state them: a drop of 1.5 m onto concrete without cracking, no visible deformation at 70 °C, resistance to hand oils and workshop grease, a visible surface that takes a color match, wall thickness of 2.5 mm, and 50 000 parts a year.
ABS passes the drop and the appearance, and fails the temperature: an HDT of 80 to 100 °C is measured under load, and a housing in a hot workshop will creep at the boss. It also has poor resistance to some greases. PC/ABS raises the heat resistance into the right range and keeps most of the impact strength, at the cost of mandatory drying and a mold temperature around 80 to 100 °C. PA66-GF30 is the stiffest and the most heat resistant, and it is wrong for this part: the visible surface shows glass, the shrinkage is anisotropic so the housing warps, and the material is abrasive, which shortens tool life for a part that is mostly cosmetic. PP with talc is cheap, chemically excellent, and too flexible for a housing with a battery in it.
The answer is PC/ABS, and the tooling consequences follow from it rather than being chosen: the cooling system must reach 80 to 100 °C, which means a hot water unit and circuits designed for it; the resin must be dried at 120 °C for four hours, which means the molder needs a dryer that reaches it; the shrinkage is 0.5 to 0.7 %, so the cavity is cut with a small allowance and the tolerance on the boss positions can be tight; and the gate position is both a flow decision and a cosmetic one, because a weld line on a visible surface is a reject. All four of those are in what the trial has to prove before the tool is signed off.
Six selection mistakes
These are the ones that recur often enough to be worth naming.
- Using HDT as the service temperature. It is measured under load, so it is the conservative number for a loaded part and a pessimistic one for anything else. Using it the other way round produces parts that soften in service.
- Assuming glass makes a part more accurate. It reduces shrinkage and adds anisotropy, and on a large flat part the second effect wins.
- Specifying a hygroscopic resin without checking the molder’s dryer. A dryer rated to 80 °C cannot dry PC at 120 °C, and the parts will look correct and fail later.
- Choosing a corrosive or abrasive grade and keeping the same steel. PVC on P20, or glass-filled nylon with an unhardened gate, is a tool-life decision taken by accident.
- Comparing resins by price per kilogram. Density, cycle time and scrap rate are all in the part cost and none of them is in the resin quote.
- Choosing the resin after the tool is designed. Every parameter on this page — shrinkage, mold temperature, steel, venting — is cheaper to accommodate on a drawing than in hardened steel.
There is a seventh that is not a mistake so much as an omission: not writing the grade down. A tool card that says “ABS” instead of the grade, the color house and the drying condition leaves the next repair, or the next molder, guessing at the one variable that decides whether the part is right. For a tool that will run for years and possibly move between plants, that is the cheapest documentation in the project.
Ten checks before the resin is frozen
This is written to be used at the point the resin is about to be fixed, which should be before the mold design is released rather than after.
| # | Check | Why it is on the list |
|---|---|---|
| 1 | Service temperature, asked twice: with load, and without | They are different numbers, and using the wrong one produces a part that softens in service |
| 2 | Chemical environment, including what the part is cleaned with | Cleaning agents cause more stress cracking than the product ever does |
| 3 | Whether the part lives outdoors, and for how long | Unstabilized PP and ABS fail within a season, and the fix is a grade rather than a design |
| 4 | The tightest dimension, and whether the shrinkage band can hold it | If the band is wider than the tolerance, no toolmaker can close the gap |
| 5 | Whether the resin needs drying, and whether the molder can reach the temperature | A dryer that tops out at 80 °C cannot dry PC at 120 °C, whatever the setting says |
| 6 | The mold temperature the resin needs, and whether the tool was designed for it | Above about 80 °C the cooling design changes, and retrofitting it is expensive |
| 7 | Whether the resin is abrasive or corrosive, and whether the steel matches | This is a tool-life decision made at the resin stage and paid for at the tool stage |
| 8 | The wall thickness the design allows, and the flow length it gives | Flow length to thickness ratio decides whether the grade fills the part at all |
| 9 | Any flammability, food-contact or medical requirement | These reduce the choice to an approved list, and nothing else on the list matters if the grade is not on it |
| 10 | Whether the grade is available, in the color, at the volume, in the region | A perfect choice that cannot be bought in the required color is not a choice |
Two items carry most of the weight. The first is asking about temperature twice, because it is the question most often answered with a single number and the two numbers differ by a factor of two on some resins. The tenth is the one that keeps a technically correct choice from failing in practice: a grade that cannot be bought, in the color, at the volume, from a local distributor, is not a decision, it is a hope.
The rest of the chain follows from the resin and is covered elsewhere on this site: the process itself in plastic injection molding, the runner consequences in hot runner versus cold runner and cavity count, the tooling consequences in mold steel selection, and the maintenance that a corrosive or abrasive grade makes unavoidable in mold maintenance. Where the volume is low enough that the tool is built in aluminum, the resin choice narrows further, for reasons set out in low volume injection molding.
When the choice needs to be made against drawings, samples and a target cost rather than on a table like the ones above, that work sits with plastic material selection and engineering support — comparing candidate grades on the actual part, checking the shrinkage against the tolerance that matters, and confirming that the molder can dry and run what has been specified before the steel is cut.
Frequently asked questions
What is the most common plastic used for injection molding?
By volume, polypropylene, followed by polyethylene and then ABS. PP dominates because it is cheap, chemically resistant, needs no drying, and tolerates a wide process window, which makes it the default for containers, caps, housings and anything where the requirement is “adequate and inexpensive”. ABS takes most of the visible consumer housings because it is amorphous, so it shrinks predictably and takes a good surface. The most common choice is not the best material; it is the one that fails least often for the least cost.
How do I choose between ABS and PC/ABS?
By temperature under load, and by whether you can dry the resin. ABS has an HDT of 80 to 100 °C and needs drying for two to four hours at 80 °C; PC/ABS raises the heat resistance by roughly 20 to 30 °C, keeps most of the impact strength, and needs 120 °C for four hours. If the part never exceeds 70 °C under load and the molder has a basic dryer, ABS is enough. If it sits in a hot enclosure or carries a load at temperature, PC/ABS is the answer and the drying is not optional.
Does glass-filled plastic hold tighter tolerances?
It holds a lower average shrinkage, which is not the same thing. Glass reduces shrinkage along the flow far more than across it, so a 30 % glass grade can differ by a millimeter or more between the two directions on a 200 mm part. That produces warpage, and warpage is harder to hold than a uniform shrink. Glass helps when the requirement is stiffness or creep resistance at temperature; it does not reliably help when the requirement is dimensional accuracy, and on flat parts it often makes it worse.
Why do my nylon parts change size after molding?
Because polyamide absorbs moisture from the air and swells as it does. PA6 takes up 1.3 to 1.9 % in the first 24 hours and can reach 8 to 10 % at saturation, and the dimensional change follows the moisture content roughly linearly. The part measured at the press is dry; the part in service is not. For anything with a tight fit, specify the dimension in the conditioned state, condition the parts deliberately before measuring, or move to PBT, which absorbs about a tenth as much.
What happens if I mold PA66 without drying it?
The water reacts with the polymer at melt temperature and cuts the chain, a process called hydrolysis. The parts come out with the correct shape and surface, and with reduced molecular weight, which means reduced impact strength and reduced elongation. Unlike a moisture defect such as splay, this cannot be seen and cannot be reversed by re-drying; the polymer is already damaged. It is the main reason a molder’s dryer capacity should be checked before a hygroscopic grade is specified, not after the first failed drop test.
Can I use the same mold for a different plastic?
Sometimes, and the check is not the part but the tool. Shrinkage has to be close enough that the cavity can still produce a part in tolerance, which rules out most moves between amorphous and semi-crystalline families. The mold temperature the new resin needs has to be reachable with the existing cooling, which rules out moving to PC or PBT on a tool built for PP. And the steel has to survive the new resin, which rules out PVC or glass-filled grades on a P20 tool without accepting a short life. A change within a family — ABS to PC/ABS, or PA66 to PA6 — is usually feasible; anything else needs the tool checked against all three.
Is a more expensive resin always the better choice?
No, and the reason is that cost per part is not cost per kilogram. Density decides how much resin a part uses, cycle time decides how long it occupies the press, scrap rate decides how much is thrown away, and tool life decides how often the mold is rebuilt. A resin at twice the price per kilogram with half the density and no drying requirement can produce a cheaper part. The comparison that works is the cost of a thousand finished parts including material, press time and scrap, and it usually takes an hour to build and saves a project.
When should the resin be chosen in a mold project?
Before the mold design is released, and ideally at the same time as the part design review. The resin fixes the shrinkage the cavity is cut for, the mold temperature the cooling system must reach, the steel that will survive the run, the drying the molder must be equipped for, and the venting maintenance interval. All of those are drawing decisions; once the steel is cut they become machining decisions, and after heat treatment they become expensive ones. The practical rule is that a resin change after tool kick-off should be treated as a design change, with the schedule and cost that implies.
