CNC machining solutions in chemistry

• 1
  What actually attacks a machined part in a chemical plant
• 2
  Alloy selection for wetted chemistry parts
• 3
  Why 5-axis setups change sealing and flow surfaces
• 4
  Surface finish, passivation and what machining leaves behind
• 5
  When a machined part is the wrong answer
• 6
  Which machining approach fits which chemical part

Most chemical service failures we see are not mechanical. The part was strong enough. It corroded, pitted, or cracked at a weld toe or a sharp internal corner. General corrosion is the easy case to design around: you pick a thicker wall and accept a known metal loss rate. Localised attack is the one that ends a campaign early, because it starts in a crevice, under a gasket, or in a stagnant dead leg where the flow never sweeps the surface.

Three mechanisms do most of the damage. Chloride pitting breaks through the passive film on 304 and 316 stainless once the temperature climbs, which is why 316L is the floor and not the ceiling for hot brine or hypochlorite. Crevice corrosion needs a geometry, not just a chemistry: a gasket face, a thread root, a pressed-in sleeve. Stress corrosion cracking needs a tensile stress and a susceptible alloy, and a machined surface with torn smeared metal supplies both.

Machining influences all three. A ground or polished surface at Ra 0.2–0.8 μm has fewer pit initiation sites than an as-machined surface at Ra 1.6–3.2 μm. A radiused internal corner drains and cleans better than a sharp one. A single-setup 5-axis part has no re-clamping marks, no mismatch on a seal face, and no tolerance stack from four fixtures.

So the first question is not which alloy. It is what the surface and the geometry will do to that alloy after six months in the line.

• 1
  PittingChlorides above roughly 60 °C break down the passive film on 304 and 316.
• 2
  CreviceGasket lands, thread roots and dead legs trap liquid that cannot be renewed.
• 3
  CrackingTensile stress plus chloride plus heat; smeared machined surfaces help it start.

For dilute acids at ambient temperature, 316L covers a lot of ground: pump housings, agitator shafts, sample probe bodies, small valve components. It machines cleanly, welds well, and the molybdenum content gives it a useful margin over 304. Below about 60 °C in neutral chloride service it is usually the economical answer.

Above that, or wherever chlorides and low pH combine, the answer moves to higher alloys. 17-4PH (SUS630) gives you strength and reasonable corrosion resistance in one part, useful for shafts and stems where 316L would need a larger section. Duplex and super duplex grades resist chloride stress cracking far better, but they machine tougher, so expect more tool wear and a slower cycle.

Titanium TA2 and TC4 (Ti-6Al-4V) suit oxidising chloride environments and heat exchanger hardware. Inconel handles hot aggressive service where nothing else survives, at a real cost in tool life and cycle time. Beryllium copper and C101 copper appear in analyser and electrical feedthrough parts where conductivity is the point. PTFE, PEEK and PP get machined for insulators, seals and low-load wear parts.

Two practical notes. First, the alloy must match the whole wetted circuit, not just one part, or you move the corrosion to the weakest component. Second, dissimilar metal contact in a wet line creates a galvanic couple, so check the fitting and the fastener as carefully as the body.

• 1
  316LDefault for dilute acids and ambient chloride service; easy to machine.
• 2
  17-4PHShafts and stems needing strength plus moderate corrosion resistance.
• 3
  Titanium and InconelHot chloride or oxidising service; slower cutting, shorter tool life.
• 4
  PEEK and PTFEInsulators, seals and low-load parts; no galvanic couple with metal.

A gasket face is a flatness problem and a finish problem at the same time. If the face is cut in two setups on a 3-axis machine, any re-clamping error shows up as a step or a taper, and the gasket has to absorb it. On a simultaneous 5-axis center the tool reaches the face at a constant angle in one setup, so flatness and finish stay consistent across the whole land. Our tolerance floor is ±0.005 mm.

Curved and contoured passages are the second case. A spiral or tangential inlet on a cyclone, a curved transfer line, an impeller shroud: these are hard to reach with a 3-axis spindle because the tool has to stay normal to a changing surface. Continuous multi-axis motion keeps the contact angle and the chip load steady, which keeps the finish uniform and avoids the chatter that leaves a rough patch for corrosion to start in.

The third case is part count. A valve body with ports on four faces would need four or five setups on a 3-axis machine. Each setup adds a tolerance stack and a chance of a mismatch. Machining it in one setup removes most of that error and shortens the production path. We run 16 simultaneous 5-axis centers with travels up to 4,000 × 400 × 150 mm, plus 12 four-axis mills, 27 three-axis machines and 16 mill-turn centers.

None of this helps if the drawing does not call out the critical surfaces. Mark the seal faces, the wetted bore and the datum together, and state flatness, finish and concentricity separately. Otherwise the shop has to guess, and the guess will be wrong somewhere. For reference, we hold Ra 0.2–0.8 μm on fine finishes, Ra 0.8–1.6 μm on high finishes, and Ra 1.6–3.2 μm as-machined.

• 1
  One setupGasket lands stay flat across the whole face; no re-clamp step.
• 2
  Constant contact angleEven chip load on contoured passages; fewer chatter marks.
• 3
  Fewer datumsPorts on multiple faces keep their position relative to each other.

Every cutting edge leaves something. A worn tool smears metal instead of shearing it, and that smeared layer has a different microstructure from the bulk. It is also where pits start. Fresh, sharp tooling and a controlled feed rate matter more in chemical service than in most other work, because the surface is the barrier between the alloy and the process fluid.

After machining, stainless parts are usually passivated to remove free iron and restore the chromium oxide layer. This is not optional on wetted surfaces. Bead blasting, tumbling, brushing and polishing are all available; the choice depends on whether you need a smooth pit-resistant surface or a matte one that hides handling marks. Electroless nickel and hardcoat anodising add a barrier layer on aluminium parts, though neither survives strong caustic service.

Laser marking is useful for alloy identification, heat numbers and flow direction arrows, with a minimum character height of 1.5 mm. Keep marks off sealing lands and out of wetted bores. An engraved mark inside a flow passage is a crevice waiting to happen, and a mark across a gasket face is a leak path.

Hardness and residual stress also matter. Heavy roughing followed by a light finish cut leaves a different stress state in the skin than a uniform series of light passes. On parts that see cyclic pressure, that difference shows up as cracking months later. If the drawing calls out a stress-relief step, do it before the finish cut, not after.

• 1
  Sharp toolingAvoids the smeared layer that becomes a pit initiation site.
• 2
  PassivationRemoves free iron and restores the passive film on stainless.
• 3
  Marking placementKeep laser marks off seal faces and out of wetted bores.
• 4
  Stress reliefDo it before the finish cut so the final surface stays clean.

Machining is subtractive and it is best at relatively compact, high-value geometry. A large, thin-walled vessel is not a good candidate: the material cost and the distortion risk both climb faster than the benefit. Fabricated and welded construction is usually cheaper there, and the weld quality can be controlled by procedure and inspection.

Very high-volume simple parts also leave the machining envelope. If you need 200,000 identical washers, stamping or die casting wins on unit cost, and we would say so. Machining earns its place in low to medium volume, in prototypes, in spares for equipment that is already installed, and in geometry that a mould cannot release or a die cannot form.

Some chemistries simply exceed any metal. Hot concentrated caustic attacks aluminium and dissolves the oxide on many stainless grades. Hydrofluoric acid attacks glass and titanium alike. In those lines the wetted part is a polymer, a ceramic or a lined component, and the machined metal is the housing around it. Recognising that early saves a redesign.

There is also a lead-time question. A machined replacement for an obsolete pump part can be made and shipped in days, which matters when a line is down. We quote and return a free DFM analysis within 12 hours, start production within 24 hours, and ship parts in 3–5 days. No minimum order quantity, from one prototype to 10,000+ part runs.

• 1
  Large thin-wall vesselsFabrication and welding usually beat machining on cost and risk.
• 2
  Very high volume simple partsStamping, casting or moulding wins on unit cost.
• 3
  Extreme chemistryUse polymer, ceramic or lined parts and machine only the housing.