Injection Mold Steel Selection Guide: P20 vs 718H vs H13 vs S136

The steel in an injection mold is the one decision that follows the
tool for its entire life. It sets how long the tool survives, how good the surface of the
part can be, how much the tool costs, and whether the mold can be repaired when something
goes wrong. And it is usually decided in the last ten minutes of a conversation, after
the price has already been quoted. This guide compares the steels that actually get used
in production tooling — P20, 718H, H13, S136 and NAK80 — on hardness, polishability,
wear resistance, machinability and cost, and explains how to pick between them for the
program you are actually running.

Short answer

Choose P20 when you are running unfilled resins, the tool is under
about 500 mm across, and the surface does not need to be polished — it arrives pre-hardened
at around 30 HRC, machines fast and costs the least. Choose 718H when the
tool is larger, runs higher cavity pressure, or will see more wear than P20 tolerates.
Choose H13, hardened to 48 to 52 HRC, for glass-filled and abrasive
resins, for hot runners, and for die casting dies. Choose S136 for
stainless, mirror-polished or medical and food-contact tooling. Choose NAK80
for cosmetic cores that need a good as-machined finish without a full polish
operation.

The rule underneath all of these is simple: steel should be specified against
the resin, the volume and the surface requirement — in that order. Most
expensive steel that has been specified for the wrong reason is simply money thrown away,
and most tool failures that are blamed on the machine are actually failures of steel that
was chosen too softly for the resin in the tool.

Why steel decides more than tool life

Steel is usually discussed in terms of hardness, and hardness is the least interesting
property to optimise on its own. What steel actually determines is the whole
cost-per-part curve for the program, because it sets three things at
once: the price of the tool, how long the tool lasts before it needs repair, and how
good the parts coming out of it can look.

Consider the arithmetic of a high-volume program. A pre-hardened P20 tool might cost
40 percent less to build than a hardened H13 tool. But if the resin is 30 percent glass
filled, P20 may only survive 250,000 shots against 800,000 for H13. At 500,000 parts a
year, the second tool is clearly the cheaper option — the extra capital cost is recovered
well before the first tool needs rebuilding, and the risk of an unplanned tool repair
mid-program disappears. This is why quoting steel as a line item without the program
attached is meaningless, and why two quotations for the same drawing can differ by 40
percent and both be correct.

There is a second, less obvious consequence. Steel determines the achievable surface
finish, because polishability and corrosion resistance are properties of the steel itself.
A textured cosmetic surface on a P20 cavity will begin to dull and lose its texture
noticeably faster than the same texture on S136, and that is a visible change to the
customer, not just a tool-room concern. If the part has a Class A face, the cavity steel
is the first specification to fix and the last one to trade away.

The five steels, and what each is for

Almost all production injection tooling in the world is built from one of five steel
families. Each is a modification of the same basic idea — a hardened tool steel with
enough carbon to take a hard edge and enough alloy content to survive the cycle — and
they differ mainly in how much they cost to build and how they behave against wear,
polish and corrosion.

Two of these need a little more explanation, because they are the ones most often
chosen for the wrong reason.

P20 versus 718H. These are the same family and the same price band,
and the choice between them is usually made on tool size and resin. 718H has a higher
nickel and molybdenum content, which gives it better toughness and better resistance to
the cavitation erosion that appears where the melt front repeatedly breaks and reforms
around a sharp corner. If the tool is large, has deep cores, or runs a resin with any
significant filler, 718H is the safer choice. If the tool is a straightforward housing
in unfilled ABS, P20 is entirely adequate and cheaper to build.

H13 versus S136. These are close relatives and the choice between
them is essentially about corrosion and polish, not about hardness. S136 carries enough
chromium to resist corrosion, which matters for PVC and flame-retardant grades that
degrade and release corrosive gases, and for any tooling that will be washed down or
handled by operators. It also takes a finer polish. H13 cannot be polished to the same
level and will corrode in aggressive service. If the cavity must be a mirror, S136 is the
answer, and the extra cost is paid once at build time rather than repeatedly in rejected
parts.

Full comparison table

The build cost index in that table is relative to a P20 tool of the same size, and it
is the number most likely to be misread. It reflects only the tooling cost, not
the total program cost. A P20 tool at index 1.0 that has to be rebuilt in year two, and
again in year four, is far more expensive than an H13 tool at index 1.35 that runs for
the life of the program.

Matching steel to your resin

Resin choice drives steel choice more strongly than almost anything else, and the
mechanism is straightforward: filler content determines abrasion. A glass fibre in the
resin is a cutting tool, and every cycle the melt flows past a sharp cavity corner it
takes a little material with it.

Two rows in that table are worth dwelling on, because they are the ones that generate
surprises.

Flame retardant without filler still changes the steel. A flame
retardant grade has no abrasive glass in it, but it does generate corrosive gases during
degradation, and a cavity that has been running for a hundred thousand shots in such a
resin will show corrosion pitting on a P20 or H13 cavity. S136 resists this, and in a
PC/ABS appliance enclosure running for seven years, that resistance is what keeps the
tool from needing a cavity rebuild in year three.

Thermosets are the hardest case. BMC and phenolic compounds contain
high mineral loading and are abrasive beyond what most moulded thermoplastics ever see.
They also run at high temperatures and long cure times, which loads the tool thermally
as well as mechanically. If a program involves a thermoset, steel selection should be a
specific conversation with the toolmaker rather than a line on a quotation.

Steel, polish and surface finish

The achievable surface of a moulded plastic part is a property of the cavity steel
before it is a property of the process. Two tools running the same resin on the same
machine, with the same cycle, will produce visibly different surfaces if one cavity is
P20 and the other is polished S136.

The SPI (Society of the Plastics Industry) finish grades are the standard language,
and the practical bands are:

There is a practical consequence here that is often missed: if the design
requires a fine surface, the steel specification has to be written into the drawing
package before the tool is quoted. It is not something to add later, because
changing the cavity steel of an existing tool means new cavity inserts — and possibly a
new tool — and the cost of that change lands at the worst possible moment, in the middle
of a production ramp.

It is also worth being honest about a common expectation gap. A Class A surface on a
textured P20 cavity is not achievable. If a customer asks for “the same finish as the
current part” and the current part came out of a polished S136 cavity, then specifying
P20 on the replacement tool on the grounds that it is cheaper will produce a tool that
cannot make the part. The cheapest correct steel is always cheaper than the cheapest
wrong one plus the rejected parts.

What steel actually costs you

Tool steel cost is a line item on the quotation, but it is not a line item in the
program. The comparison that matters is steel cost per part over the life of the
program, which pulls in three separate effects.

Build cost. The direct premium of moving from P20 to hardened H13 is
typically 30 to 40 percent of the tool price, and S136 adds a further 5 to 15 percent
because of the longer polish cycle. On a small tool this is a modest absolute number.
On a large multi-cavity tool it is significant but still not usually the deciding
factor.

Rebuild avoidance. This is where the real money sits. A cavity
rebuild mid-program costs the tool steel, the machining, the downtime, the remoulding
and the schedule recovery — and it happens at a point where the production schedule is
usually tightest. If hardened steel removes one rebuild from a four-year program, it
pays for itself, and it also removes a risk that is difficult to price.

Scrap cost. This is the one that is least often considered. If the
cavity polish degrades and the visible surface of the part goes out of specification,
the cost lands as rejected parts, not as a tool-room expense. A cosmetic part rejected
at the customer’s incoming inspection is far more expensive than the tool premium that
would have prevented it. This is the argument for S136 on a visible part, and it holds
even at moderate volume where the tool premium looks hardest to justify.

There is a fourth factor that is not a cost but changes the calculation: tool
availability and lead time. Pre-hardened steel ships in a day and machines
immediately. Hardened steel requires heat treatment after rough machining, which adds
one to two weeks to the build schedule, plus a tempering cycle after EDM work. If the
program has a fixed start date, that time has to be in the plan from the beginning, not
discovered at the point where the tool is half built.

Tool life and how to extend it

Tool life is quoted as a number of shots, and the number is meaningful only with the
resin and the operating conditions attached. The ranges below are the realistic bands
for a well-built tool in normal production, and the bottom of each band is what a tool
actually experiences if the process is not controlled.

Tool life is extended by five things, and all five are process decisions rather than
tool-room decisions:

1. Hold the resin specification. Substituting a cheaper grade with a
   different filler load changes the wear rate completely, and it does it invisibly. This
   is the single most common cause of early tool failure we see.
2. Control the melt temperature. Every degree above the recommended
   window accelerates the chemical breakdown of the resin and the resulting corrosive
   by-products.
3. Clean the cavity on a schedule. Degraded resin left in a cavity for
   a month hardens and continues to abrade the steel.
4. Keep the vents open. A blocked vent raises local cavity pressure,
   which forces the steel harder at exactly the points designed to be relieved.
5. Log the shots. A tool that has run 300,000 shots is a different
   engineering question from one that has run 30,000, and most suppliers cannot tell the
   difference without being told.

How to specify steel on a quote

To make sure the steel in the quotation matches the steel you need, five pieces of
information have to be on the table before the quote is issued. Without them, a supplier
is guessing, and the guess is usually P20 because it is the default.

With those five answered, the steel decision becomes mechanical rather than
controversial. It is also the fastest way to make two quotations comparable, because
without a resin, a volume and a surface requirement, two quotations for the same drawing
are measuring different things and the numbers cannot be compared.

Our mold steel selection page lists the
grades we machine and the reasoning behind each recommendation, and our
mold design guide covers the design decisions
that determine how much wear the tool will actually see. For the practical side of
getting a tool built and running, the
mold maintenance page covers the routines that
extend tool life, and the
precision mold making page covers how the
cavities are cut and finished.

Three steel selection cases

Steel selection is general until you apply it to a specific program. These three are
real decisions from our own tooling history, and in each case the obvious answer was
wrong for a reason that only shows up once the volume and the resin are known.

Case 1: The 80,000-part PA66-GF30 housing

A medical device customer wanted a 250 g housing in PA66 with 30 percent glass fibre,
produced at 80,000 parts a year, with no visible cosmetic surface. The first quotation
came back in P20, because the part had no cosmetic requirement and the volume looked
moderate.

That was the wrong call, and the reason was in the resin rather than the part. PA66-GF30
is not an abrasive because of any feature on the housing — it is abrasive because of the
glass. Every cycle the melt passes the sharp corners where the bosses meet the walls and
takes a little material with it. In production the tool lost cavity detail around 140,000
shots and the wear was not distributed evenly, because the flow front repeatedly broke and
reformed at the same corner. The decision was to rebuild with hardened
H13 inserts and to move the gate slightly to shift the wear pattern, and the rebuild
extended the tool past 700,000 shots. The initial saving on P20 had cost more than the
upgrade would have.

Case 2: The glossy cover that had to move to S136

A consumer electronics customer had a 45 g PC/ABS cover with a visible Class A face,
produced at 250,000 a year. The tool was specified in H13 with a fine texture, because the
original tool on the previous product had been H13 and the finish had looked acceptable
at the start of the program.

It did not stay acceptable. H13 in a flame-retardant PC/ABS grade develops surface
corrosion over time, and while the visual effect was slow, the parts that came off the
tool in year two were rejected at incoming inspection for a slightly dull, pitted face.
The complaint was not about the tool, it was about the parts, and it reached the customer
as a supplier quality issue rather than as a tooling issue. Moving the cavity to
polished S136 cost more at build time and removed the entire failure mode. The
general lesson is that on a visible face the steel specification is a quality
requirement, not a cost decision, and it should be written into the drawing package
before the tool is quoted.

Case 3: The insert retrofit that saved the program

An automotive customer ran a 900,000-part-per-year program in PP-T20 in a P20 tool
that had been in service for four years. Tool life was being consumed by the fact that
the customer had quietly switched resin to a higher talc-filled grade for cost reasons,
and the tool was wearing at three times its expected rate.

Rather than rebuild the mold, the cavity inserts were replaced with hardened H13
inserts in the existing mold base. Because the base plates were large and the inserts
were designed to be removable, the swap took two weeks and cost a fraction of a new tool.
The tool was then expected to run past 1.5 million shots. The point is worth stating
plainly: wear is often an insert problem, not a tool problem, and the
cheapest fix is frequently to ask the supplier whether the inserts can be upgraded
without rebuilding the base.

Tool steel terms quick reference