17-4 PH Stainless: Machining and Heat Treatment Notes

17-4 PH is the stainless steel we reach for when a customer needs 316-level corrosion resistance on the datasheet and three times 316's strength on the same line — valve internals, sensor housings, firearm and aerospace hardware, pump shafts, surgical instrument blanks. It is also the stainless we get the most quoting questions about, because its whole personality depends on a heat-treatment decision that happens after machining: the "condition" (H900, H1025, H1150…) printed at the end of the material callout. Specify it carelessly and you get parts that miss hardness, bores that shrank out of tolerance after aging, or threads that galled the first time a technician ran a screw into them.

This page is our shop notes on 17-4 PH machining and its heat treatment: what the alloy actually is, the full condition chart with real numbers, why we machine it in Condition A and age afterwards, how much it shrinks during aging (with a planning table you can apply to a drawing), the cutting and tapping practices that keep it predictable, and a decision table against 316, 15-5 PH and 440C. This is a grade-specific deep dive — for the general question of when heat treatment sits in the process chain, start with our heat-treatment sequencing guide, then come back here for the 17-4 specifics.

What 17-4 PH Actually Is

17-4 PH is a precipitation-hardening martensitic stainless steel, standardised as UNS S17400, EN 1.4542, and covered by ASTM A564 (bar), A693 (plate) and a long list of AMS specifications for aerospace stock. The name decodes the alloy: roughly 17% chromium and 4% nickel, with 3–5% copper. The copper is the interesting part. In the solution-treated state that copper sits dissolved in the matrix; age the steel at 480–620 °C and it precipitates as fine copper-rich particles that block dislocation movement — which is how a "soft" solution-treated part gains 300–400 MPa of strength in a simple furnace cycle with no quench and no carburising gas.

• Corrosion resistance: comparable to 304 in most atmospheres, somewhat below 316 in chlorides. It is a genuine stainless, passivatable per ASTM A967 — but not the "marine grade" some buyers assume it is.
• Strength: yield strength from ~725 MPa (H1150) up to ~1170 MPa (H900) — against ~205 MPa for annealed 316L. In the H900 condition it is one of the strongest stainless grades you can buy as bar stock.
• Magnetism: strongly magnetic in every condition. That surprises buyers coming from austenitic grades and occasionally trips quality documentation — it is correct behaviour, not a defect.
• Weldability: weldable, but a welded 17-4 part must be re-solution-treated and re-aged to recover properties in the heat-affected zone. For small machined parts, designing the joint out (or moving to a threaded/brazed joint) is usually the cheaper answer.

One naming trap worth clearing up: "17-4" and "630" are the same alloy (630 is the ASTM cast/wrought designation), and you will also see it as X05CrNiCuNb16-4 or SUS630 on Asian drawings. Same metal, same conditions, different paperwork.

The Condition Chart: H900 Through H1150

The heat-treatment condition is a single number — the aging temperature in °F. Higher aging temperature = lower strength, but better toughness and better stress-corrosion behaviour. These are typical minimum values for bar per ASTM A564; always verify against the certificate for the actual heat:

Condition	Aging treatment	UTS (min)	Yield (min)	Hardness (typ)	Elongation	Best for
Condition A	Solution treated 1040 °C, air cooled	~1100 MPa	~1000 MPa	32–36 HRC	~15%	The machining state — not a service condition
H900	480 °C × 1 h, air cool	1310 MPa (190 ksi)	1170 MPa (170 ksi)	40–48 HRC	10%	Maximum strength, minimum toughness
H925	495 °C × 1 h	1170 MPa	1070 MPa	~43 HRC	10%	Slight toughness gain over H900
H1025	550 °C × 1 h	1070 MPa (155 ksi)	1000 MPa (145 ksi)	35–39 HRC	12%	The workhorse: strength/toughness balance
H1075	580 °C × 1 h	1000 MPa (145 ksi)	860 MPa (125 ksi)	33–37 HRC	13%	General structural hardware
H1100	595 °C × 1 h	965 MPa	795 MPa	~34 HRC	14%	Transitional
H1150	620 °C × 1 h	930 MPa (135 ksi)	725 MPa (105 ksi)	28–32 HRC	16%	Best toughness and chloride-SCC resistance
H1150D	Double aged (H1150 + re-age)	~860 MPa	~725 MPa	~30 HRC	16%	Maximum toughness for critical parts

Two reading notes on this table. First, Condition A is not a delivery condition for a finished part — it is the state the bar arrives in and the state you machine in. A part that ships "as-machined, Condition A" will age itself unpredictably in service (welding nearby, friction heating, even shop-floor warmth over years) and its dimensions will drift. Second, the strength range across conditions is enormous — H900 gives you 60% more yield than H1150 — so the condition number on the drawing is doing more work than the alloy name. When a drawing says "17-4 PH" with no condition, that is a question for the customer, not an assumption for the supplier to make.

The Machining Playbook: Cut in Condition A, Age Afterwards

Almost all 17-4 machining happens in Condition A, at roughly 32–36 HRC, and the whole process chain is arranged around that fact:

• Turning and milling. In Condition A, 17-4 cuts like a medium-alloy steel — carbide tooling at moderate speeds, coated grades (TiAlN) preferred. Typical turning parameters for Condition A bar: cutting speeds around 140–180 m/min with carbide inserts, positive rake geometry, and generous coolant. Expect tool life between 303 and 316: it is harder and more abrasive than the austenitic grades but far more predictable than the hardened martensitics.
• Do not machine H900. At 44+ HRC the alloy grinds tools rapidly; carbide life collapses and even ceramic struggles on interrupted cuts. If a feature was missed before aging, the recovery is grinding or EDM, not a milling cutter. This single fact is why process discipline (see below) matters more for 17-4 than for almost any other stainless.
• Tapping and threading is the pain point. 17-4 work-hardens and galls, and 17-4 tap breakage is the classic scrapped-part story. Our practice: keep thread engagement to 65–75% of theoretical depth (a full 100% thread gains almost no strength and doubles tap torque), use spiral-point or spiral-flute taps with TiAlN coating, and prefer thread milling for anything above M6 or below M3 — a thread mill that snaps is a rework, a broken tap in a blind hole is usually the bin.
• Drilling. Standard HSS-Co works for small holes but carbide or coated tools pay for themselves on any production quantity. Peck cycles with good coolant delivery through the tool; do not let the drill dwell at the bottom of a blind hole, or the work-hardened layer will fight the next tool that touches the surface.
• Work hardening. Less aggressive than 304/316, but real: keep feeds up, avoid spring passes on finishing, and never let a tool rub. A glazed, work-hardened skin is the main cause of "the insert lasted fine on the first part and died on the second".

For context, here is how 17-4 sits against the other common machined stainless grades (machinability index, 303 free-machining = 100):

Grade (condition)	Machinability index	Hardness when machined	Watch out for
303	~100	~85 HRB	Poor corrosion resistance; sulfide inclusions hurt polish and fatigue
304	~60	~85 HRB	Work hardening, stringy chips, galling
316	~55	~90 HRB	Same as 304 plus more abrasive; slower everything
17-4 PH (Condition A)	~48	32–36 HRC	Tapping, galling, aging shrinkage after machining
17-4 PH (H900)	~15	44+ HRC	Essentially a grinding job — avoid entirely
440C (annealed → hardened)	~45 annealed	~95 HRB → 58 HRC	Distortion and scale in hardening; finishing almost all after hardening

The comparison against our 303 vs 304 vs 316 comparison is useful for expectations: 17-4 in Condition A machines a little slower than 316, and the delta shows up mostly in the threading and finishing operations, not in roughing. The big cost difference versus the austenitic grades is not the cutting — it is the heat-treatment step and the dimensional discipline around it.

Aging Shrinkage: Plan the Final Size at Condition A

This is the single most expensive thing buyers and junior programmers get wrong with 17-4. When the part ages, the precipitation of copper phases contracts the material. The linear shrink is small in percentage terms but enormous in tolerance terms, and it scales with the aging temperature:

Aging condition	Typical linear shrink	On a 200 mm length	On a Ø25 mm bore	On a Ø100 mm flange OD
H900	~0.04–0.05%	0.08–0.10 mm	0.010–0.013 mm	0.04–0.05 mm
H1025	~0.05%	~0.10 mm	~0.013 mm	~0.05 mm
H1075	~0.06%	~0.12 mm	~0.015 mm	~0.06 mm
H1150	~0.08–0.10%	0.16–0.20 mm	0.020–0.025 mm	0.08–0.10 mm

Values are typical mill-and-handbook figures for small sections; the actual shrink depends on section size, aging soak and the heat treater's furnace practice — always confirm the compensation with whoever runs the furnace, and keep test coupons with the batch.

How we apply this in practice:

1. Everything cut in Condition A gets a shrink allowance on the features that will be final after aging — typically by offsetting the tool path (or, for critical bores, leaving grind stock).
2. Features with tolerances tighter than ~±0.025 mm are finished after aging — honed bores, ground diameters, lapped seats, wire-EDM slots. Aging first and finishing after costs one extra setup; skipping it costs the batch.
3. Symmetric parts age more predictably than asymmetric ones. A shaft shrinks uniformly; a flanged housing with one heavy boss can warp as well as shrink, which is why thin, asymmetric 17-4 parts get a stress-relief evaluation before aging — the same sequencing logic described in our sequencing guide.
4. Threads should be cut oversize-compensated or cut after aging where gaging matters; a class-fit thread that was perfect in Condition A can close up visibly after H1150.

One more subtlety: aging also relieves a little of the machining residual stress, so flatness can move slightly even where the linear shrink is compensated. For thin 17-4 parts with flatness calls under 0.05 mm, we rough, semi-finish, age, then finish — the standard rest-and-recut logic that any stress-relieved alloy demands.

Passivation and Corrosion Notes

17-4 leaves the machine bright and active, and it will surface-rust in a humid shop or a salt-laden environment if it is not passivated. Standard practice is ASTM A967 — nitric or citric passivation — after all machining and heat treatment are complete. Three practical notes:

• Sequence matters: passivate after aging. Aging in a slightly oxidising atmosphere can leave a heat tint that undermines the passive film; passivating at the end, after any final finishing, gives the clean chromium-rich surface the spec is aiming at.
• H900 and chlorides are a cautious pairing. The peak-strength H900 condition has the lowest fracture toughness and the highest susceptibility to stress-corrosion cracking in chloride environments. For parts that see salt water, de-icing salt, or offshore service, H1025 or higher is the usual engineering answer — the ~10–20% strength sacrifice buys a large improvement in SCC behaviour.
• Do not confuse "stainless" with "316". In pitting resistance 17-4 sits below 316 and roughly at 304. If the part's corrosion requirement is written as "must resist salt spray like the marine spec", 17-4 may not be the right alloy no matter how attractive its strength is — see the decision table below.

17-4 vs 316 vs 15-5 PH vs 440C: The Decision Table

If the part needs…	Choose	Why
High strength + ordinary corrosion exposure (valve stems, fasteners, sensor bodies, actuators)	17-4 PH (H1025/H1075)	~3–5× the yield of 316 at similar atmospheric corrosion performance; machineable in Condition A
Maximum strength in a small part, benign environment	17-4 PH (H900)	190 ksi class strength from bar stock; mind the toughness and SCC caveats
Corrosion in chlorides first, strength second	316 / 316L	Better pitting resistance; no aging step; but ~205 MPa yield means cross-sections grow
Aerospace forging-grade toughness through the section	15-5 PH	Same family chemistry, vacuum remelted; better through-thickness ductility — the aerospace successor for forgings
Maximum hardness / wear (bearings, blades, dies)	440C	58–60 HRC achievable; gives up corrosion resistance and is far harder to machine after hardening
Food, pharma or marine wetted parts with strength needs	17-4 or 316, case by case	17-4 passivates well but pits before 316; the medium and cleaning chemistry decide

A note on 15-5 PH, because RFQs increasingly name it: mechanically it overlaps 17-4 almost exactly, and in bar form either will usually work. The difference is metallurgical cleanliness — 15-5 is produced with refined remelting practice that gives better ductility and toughness through the thickness of heavy sections. If your drawing comes from an aerospace prime, the 15-5 callout is likely frozen; do not substitute without written approval. For commercial parts, quote both if the stock lead times differ — 17-4 is stocked far more widely, including every Chinese machining cluster, and the price gap is real but modest (typically 5–15%).

The Cost Picture: Where 17-4 Money Actually Goes

17-4 bar typically runs 1.5–2.5× the price of 304 per kilogram, and 316 runs between them. On a material-heavy part that premium dominates the piece-price delta. But the structural costs of a 17-4 part are elsewhere:

• Heat treatment: a batch charge for aging (it is a simple one-hour furnace cycle — far cheaper than carburising or full hardening of 440C, but on small lots the per-part share of the batch charge can exceed the material premium).
• Second finishing setup: any part with post-aging finishing adds a setup. On tight-tolerance parts this is unavoidable and should be quoted in, not discovered.
• Thread strategy: thread milling instead of tapping adds machine minutes; snapped taps add scrapped parts. The few extra minutes are almost always the cheaper line item.
• Yield: 17-4 scrap is not recoverable into the same alloy stream the way 304 chips are, which nudges raw-material surcharges on some buyers' contracts.

For most small machined parts — where machining time dominates — the total 17-4 premium over a 316 version lands in the 10–25% band, and it buys three to five times the yield strength. You can model where your own part sits with our free CNC machining cost calculator, then send the drawing for a firm number: we quote 17-4 regularly and will flag the shrink-compensation and finishing decisions up front rather than after the first batch.

How We Run 17-4 Jobs

Our standard process for a 17-4 PH part, which is worth sharing because it is the checklist we wish every RFQ arrived with:

1. Confirm the condition on the drawing (and passivation spec). No condition = engineering query before quoting, not an assumption.
2. Buy Condition A bar from a mill with full traceability — heat number, certificate, hardness on the certificate. 17-4's properties are forged in the mill's solution treatment; there is no recovering a bad heat at the machine shop.
3. Rough and semi-finish in Condition A, with shrink compensation on aging-affected features and grind/hone stock on sub-±0.025 mm features.
4. Age at the specified condition, with test coupons (or the parts themselves) hardness-checked against the condition band — H900 parts come back at 40 HRC or there is a furnace conversation to have.
5. Finish after aging: grind, hone, lap or thread-mill the critical features; verify CMM dimensions against the post-aging state, which is the only state the customer will ever see.
6. Passivate per ASTM A967, then final inspect and pack. Certificates (material, hardness, heat treat, passivation, dimensional) ship with the batch.

The same discipline, expressed as a general principle rather than a grade-specific recipe, is in our sequencing guide; the 17-4-specific numbers above — shrink table, condition chart, machinability index — are what turn that principle into a first-pass-good batch. For buyers working across materials, our aerospace machining overview shows how 17-4, 15-5 and titanium parts share a quality system, and our Ra chart covers what finish to expect from aged 17-4 after grinding versus turning.

Frequently Asked Questions

Can you machine 17-4 PH in the H900 condition?

Practically, no — beyond light grinding, EDM or single light turning passes with premium tooling, H900 at 44+ HRC destroys carbide cutters and makes tapping impossible. The economical route is always: machine in Condition A, age, then finish the few features that need it by grinding or honing. If a drawing arrives as a finished part in H900 with tight tolerances, we plan post-aging finishing into the quote from the start.

Does 17-4 PH rust?

It can. With ~15–17.5% chromium and no molybdenum, 17-4 passivates well but sits below 316 (and slightly below 304) in pitting resistance. Unpassivated chips and parts will show surface rust in a humid shop. After ASTM A967 passivation it performs well in atmospheric and many washdown environments; in continuous chloride exposure, specify H1025+ for SCC resistance and consider 316 instead if corrosion, not strength, is the governing requirement.

How much does 17-4 shrink during heat treatment?

Expect roughly 0.04–0.05% linear contraction at H900 growing to 0.08–0.10% at H1150. On a 200 mm dimension that is 0.08–0.20 mm — several times a typical ±0.05 mm tolerance, which is why the shrink must be compensated or the feature finished after aging. The exact value depends on section size and furnace practice, so we confirm compensation with the heat treater and keep coupons with each batch.

What is the difference between 17-4 and 15-5 PH?

Same family of chemistry and nearly identical strength in the same conditions. 15-5 is made with refined (vacuum) remelting practice, which improves toughness and ductility through the thickness of heavy sections — it is the standard answer for aerospace forgings. In ordinary bar machining for commercial parts, either usually works; for aerospace drawings the callout is frozen and substitution needs approval.

Which condition of 17-4 should I specify?

H900 for maximum strength in benign environments; H1025 for the common strength/toughness balance; H1075 for general hardware; H1150 (or H1150D) for best toughness and chloride-SCC resistance. If you are unsure, tell us the load case and environment — the condition is a one-word change on the drawing but a large change in part behaviour, and it is much cheaper to choose it correctly than to re-heat-treat a batch.

Can 17-4 PH be welded?

Yes, but the weld and heat-affected zone lose the aged properties, and recovery requires a full re-solution-treatment plus re-aging cycle. For small machined parts, a threaded, press-fit or brazed joint usually beats welding on total cost; where welding is unavoidable (large housings, repair work), plan the re-heat-treatment and the resulting dimensional change into the process from the beginning.