Mold Steel Selection: P20 vs 718 vs H13 vs S136 vs NAK80

Last updated: 2026-09-01 · Reading time: about 11 min

Five steels cover most injection mold builds, and each one exists for a different reason: P20 is the economical all-rounder, 718 is P20 upgraded for large blocks and better polish, H13 is the wear fighter for filled resins, S136 is the corrosion-proof mirror steel, and NAK80 is the pre-hardened finishing specialist. The right choice is not “the best steel” — it is the steel that matches your volume, your resin, and your surface requirement without paying for capabilities your part will never use.

Key facts

  • Steel choice is set by three inputs: production volume, resin aggressiveness, and surface finish requirement.
  • P20 and 718 are pre-hardened mold steels — machined as delivered, no hardening heat treatment needed.
  • H13 and S136 are hardened in service (around 48–52 HRC) for wear resistance and corrosion resistance respectively.
  • NAK80 delivers mirror polish and fine texture straight from a pre-hardened block at about 40 HRC.
  • Upgrading steel costs money once; running the wrong steel costs money every production day.

What does the steel actually decide in a mold?

Steel is rarely the largest row on a mold quotation, but it sets four things you live with for the whole program: how long the tool survives before wear forces repairs, how well visible surfaces can be polished or textured, whether corrosive resin byproducts will pit the cavity, and how cleanly fine features machine in the first place. Pick the steel against the requirement that matters most to your part, and let the quotation reflect that pick explicitly — a quote that does not name the core and cavity steel grades is not finished yet.

Pre-hardened vs hardened-in-service — Pre-hardened steels (P20, 718, NAK80) arrive from the mill already at working hardness and are machined directly. Hardened-in-service steels (H13, S136) are machined soft, then heat treated to working hardness — an extra step that adds lead time and a small dimensional risk, in exchange for much higher surface hardness.

The five steels side by side

Steel Hardness in service Best suited for Polish and finish Watch out for
P20
1.2311		 28–32 HRC, pre-hardened General-purpose tooling, low to mid volumes, non-corrosive resins, prototype and bridge tools. Polishes to a good commercial finish; not a mirror-grade steel. Softest of the five — abrasive or corrosive resins shorten its life quickly.
718 / 718H
1.2738		 28–36 HRC, pre-hardened Larger mold blocks where hardness must stay uniform through thick sections; mid to higher volumes. Polishes cleaner than P20; takes texture well. Still not corrosion-resistant; for corrosive resins pair with coatings or choose S136.
H13
1.2344		 46–52 HRC, hardened Glass-filled and mineral-filled engineering resins, high-wear cores and inserts, long programs. Fair polish; chosen for toughness and abrasion resistance rather than appearance. Heat treatment adds lead time and cost; brittle at thin edges if design ignores it.
S136
1.2083		 48–52 HRC, hardened PVC and flame-retardant grades that release corrosive gas; optical and transparent parts; parts with strict hygiene requirements. Mirror polish to very high gloss; holds polish in production. Hardening cycle and material cost — do not specify it for parts that will never see its benefits.
NAK80 38–42 HRC, pre-hardened Appearance parts needing mirror or fine texture without a hardening step; precise, stable features. Outstanding polish and EDM surface quality for a pre-hardened steel. Material cost is notably above P20/718; overkill for non-visible internal parts.

How does production volume change the answer?

Volume is the first filter because it converts “steel cost” into “cost per shot”. A 20,000-shot program and a 2,000,000-shot program made from the same geometry are different tooling problems:

  • Under roughly 100,000 shots: pre-hardened P20 is usually the rational pick — you are not paying for wear capacity the program will never consume.
  • Mid volumes and larger molds: 718 earns its extra cost when the block is thick (uniform hardness matters) or the finish needs to be better than P20 delivers.
  • Multi-million-shot programs: hardened steels move from “upgrade” to “baseline”. A worn cavity at shot 800,000 costs downtime, rework and possibly a new insert — far more than the hardening cycle would have.
  • The crossover question: if your program sits near the boundary — say 300,000 to 500,000 shots — state the volume honestly in the RFQ and let the quote show both options with their cost-per-shot math. A quote that hides that comparison is selling you one answer instead of the decision.

This is the same arithmetic as cavitation: the sticker price is paid once, the steel’s behavior repeats every cycle. Any quote for a long program should state the steel grade and justify it against the shot count, not just list it.

Which resin forces which steel?

The resin is the second filter, and it can overrule volume. Two mechanisms do the damage, and they call for different steels:

  1. Abrasion. Glass fiber, mineral fill and some reinforcements grind at gates, runners and cores shot after shot. The answer is hardness and toughness — H13 on the wear zones, even if the rest of the mold stays P20 or 718.
  2. Corrosion. PVC decomposition and some flame-retardant systems release acidic gas at melt temperature; moisture-sensitive processing can add condensation damage between runs. The answer is corrosion-resistant steel — S136 on the affected surfaces — because pitting ruins polish and dimensions in ways no repair fully reverses.
  3. Neither. Plain ABS, PP, PE and similar commodity resins are gentle on tooling. Paying for H13 or S136 here buys nothing the part will ever use.

Mixed specifications are normal and legitimate: a common build is a P20 or 718 base with H13 inserts at abrasive gate areas, or S136 cores only on corrosive programs. A line-item quote should let you see exactly where each grade sits.

What does the surface requirement demand?

The third filter is finish. Polished and textured surfaces put different loads on steel purity and hardness:

  • Hidden surfaces / light benching. Any of the five works; default to the least steel that satisfies the other two filters.
  • Commercial polish and standard texture. 718 covers most of this territory reliably.
  • Mirror gloss, optical clarity, fine texture. This is S136 or NAK80 territory. S136 when corrosion risk exists or maximum polish is needed after hardening; NAK80 when you want mirror results without a hardening step and with excellent EDM surfaces.

One coupling worth remembering: a mirror finish on a soft steel will look right at T1 and drift during production. If the finish requirement is severe, the steel has to be able to hold it, not just show it once.

Where do the working steels sit — and what happens to the mold base?

A mold is not one block of steel; it is a stack of plates and inserts, and the grade discussion only applies to the parts that touch the plastic. That distinction matters twice — once for cost, once for honesty in the quotation:

  • Core and cavity plates, inserts, lifters and sliders carry the working surfaces. This is where volume, resin and finish apply, and where H13, S136 or NAK80 earn their place.
  • The mold base itself — spacer plates, clamp plates, guide pins — carries structure, not plastic. Standard base steels at pre-hardened hardness are correct there, whatever the working inserts are made of. Specifying mirror steel for plates that never see the melt is budget burned in a place nobody can see.
  • Sliders and lifters deserve their own row. They wear from movement as well as from the resin; a glass-filled program can need hardened sliders even in a mold whose main cavity stays pre-hardened.

So when you read a quotation, check that the steel rows name components, not just grades. “S136” on the bottom line is ambiguous; “S136 on cavity inserts, 718 on cores, standard base” is a scope you can verify at acceptance — and compare against another factory’s line by line.

How do you check steel claims in a quotation?

Steel grades are one of the easiest rows to quietly downgrade after the deposit, and one of the hardest problems to catch at T1 — because the mold still works on day one. Protect yourself with paper, the same way you protect the rest of the scope in the RFQ checklist:

  1. Name the grade per component in the RFQ — base, core, cavity, inserts — and require the quotation to answer at the same granularity.
  2. Require the mill certificate (steel supplier document with grade and heat number) as part of the deliverables, not a verbal “we use good steel”.
  3. Ask for heat treatment records when hardened steels are specified: target hardness range and the measured result.
  4. Keep the records with the mold file. When the tool ships to your plant, grade documentation travels with it — it is your evidence if a wear question appears a year later.

Does better steel always mean a better mold?

No — and this is the trap in steel selection. An S136 cavity on a plain PP part with no gloss requirement is not quality; it is budget spent in the wrong place, with the same part quality as 718 would have delivered. Real quality is matching: volume, resin and finish define the requirement, and the cheapest steel that satisfies all three is the engineered choice. Conversely, shaving steel grade to cut the sticker price on a two-million-shot glass-filled program is the most expensive saving available in mold sourcing — the tool pays it back in downtime, inserts and early rebuild.

Buyers often ask me which steel is best, and the honest answer is always the same: best for what. Give me your shot count, your resin family and your appearance standard, and the steel practically selects itself. The problems never come from choosing a modest steel for a modest job — they come from silent substitutions after the deposit, which is why I put grade and heat number on paper before any steel is ordered.

— Arthur, Tooling Engineer at aCatMould

Frequently asked questions

What is the most common mold steel for general tooling?

P20 and its variants, because most programs live in the low-to-mid volume range with non-aggressive resins. It is the default answer, and the default worth challenging only when volume, resin or finish gives a reason to.

Can one mold mix several steels?

Yes, and it is often the right call. Base in 718, wear-zone inserts in H13, corrosive-zone cores in S136 — each grade where its property is actually used, documented per component in the quote.

Does hardening the steel change the dimensions?

Heat treatment introduces small, predictable movement, which experienced toolrooms compensate during machining. It is a managed risk, not a reason to avoid hardened steels where they are needed — but it belongs on the schedule and in the DFM conversation.

Which steel for transparent or optical parts?

S136 or NAK80. The deciding tiebreakers: corrosion risk from the resin favors S136; wanting mirror results without a hardening cycle favors NAK80.

How much does the steel choice move the mold price?

Less than most buyers expect on the quotation — steel is one row among many, and machining hours usually outweigh material cost — but enormously on lifetime cost if the choice is wrong for the volume or resin. Ask for the grade to be justified per component instead of asking for a cheaper number.

Should I specify the steel myself in the RFQ?

Specify the requirement — shot count, resin family, finish grade — and ask the factory to propose and justify the grade. Then verify the proposal against the logic in this article. A factory that cannot justify its steel choice has not finished quoting.

Key takeaways

  1. Three inputs decide the steel: production volume, resin aggressiveness, surface finish requirement — in that order of filtering.
  2. P20 is the economical default, 718 for large blocks and better polish, H13 for abrasive filled resins, S136 for corrosive resins and mirror gloss, NAK80 for finishing without hardening.
  3. Mixed steels in one mold are legitimate engineering: put each grade where its property is consumed, and name it per component.
  4. “Better steel” is not a goal — matching steel to requirement is; over-specifying wastes budget exactly like under-specifying does, just quietly.
  5. Verify grades with mill certificates and heat treatment records attached to the mold file, not with verbal assurances.

If you want a steel recommendation you can argue with — grade per component, justified against your volume and resin, with mill certificates in the file — send us your part data. You will get a written DFM and a line-item quotation back, steel rows included, from a named engineer.

Sources
LKM (Lung Kee) — mould bases and mold steel for the toolmaking industry
MatWeb — material property database for resin and steel datasheets
HASCO — the pioneer of mouldmaking: standard mold units and components
DME — global supplier of mold bases, components and hot runner systems