The most common question we get is some version of “what’s your best coating?” It’s the wrong question, and answering it directly usually leads to a tool that comes back six weeks later with the same failure it had before.
There is no best coating. There is a correct match between a specific failure mode, a specific steel, a specific geometry, and a specific process. Get that match right and you can see tool life go up two to ten times without touching anything else in your production process. Get it wrong and you’ve spent money on a surface that was never going to solve the problem you have.
Here’s how to think through it.
Start with how the tool is failing, not with what you want applied
Pull a failed tool and look at it before you call anyone. What you see tells you most of what you need to know.
Aluminum or magnesium welded to the surface, material dragging, galling. This is adhesive wear. Molten metal is chemically bonding to your steel. The fix is a surface the cast metal doesn’t want to stick to, which means low wettability and good release, not just hardness. Hardness alone will not stop soldering.
Fine crazing, a spiderweb pattern, cracks running through the cavity. This is thermal fatigue, better known as heat checking. It comes from the surface expanding and contracting through thousands of thermal cycles while the core stays cooler. What helps here is compressive residual stress at the surface and a treatment that runs below the tempering temperature of your steel so you aren’t softening the substrate while you protect it.
Uniform material loss, rounded edges, washed-out detail, dimensions drifting. Abrasive wear. Glass fill, mineral fill, carbon fiber, and scale are all cutting your tool a little at a time. This is the one case where raw hardness genuinely is the answer, and where a hard coating earns its cost.
Pitting, rust in water lines, degradation in H2S or brine service. Corrosion. You need a barrier with chemical resistance, and if the affected surfaces include internal passages, you’ve just eliminated every line-of-sight process from consideration.
Most real tools show two or three of these at once. That’s normal, and it’s the reason single-process shops so often get it wrong.
The three families, and what each one is actually for
Diffusion treatments (nitriding, nitrocarburizing, and processes like DYNA-BLUE) don’t add a layer on top of your steel. Nitrogen and carbon diffuse into the surface and form hardened compounds within the material itself. Because nothing is sitting on top, nothing can delaminate, chip, or flake off. The hardness is integral to the part.
The practical advantages are significant. These processes run below the tempering temperature of most tool steels, so you get no distortion and no need to re-harden. They’re non-line-of-sight, which means deep ribs, internal cooling channels, tight corners, and blind features all get treated as evenly as the open faces. And they generate compressive residual stress, which is exactly what resists heat checking.
The tradeoff is a ceiling on hardness. Surface hardness in the range of 75 HRC equivalent is excellent, but it isn’t 90.
Hard coatings (TiN, CrN, CrWN, TiAlN, applied by PVD) go the other direction. They are extremely hard, often 90 HRC equivalent or above, and only two to five microns thick. For pure abrasive wear against filled resins or abrasive slurries, nothing in the diffusion family competes.
The tradeoffs are real, though, and they’re the ones that get skipped in the sales conversation. PVD is line-of-sight, so a deep pocket or an internal bore is going to get uneven coverage or none at all. And a coating is only as good as what’s underneath it.
Duplex processes combine both. You diffusion-treat first to build a hardened, load-bearing case, then apply the PVD coating on top of it. This isn’t a marketing bundle. It solves a specific engineering problem, and it’s worth understanding why.
The load support problem that quietly kills coating jobs
Picture a hard coating as a pane of glass. Lay it on a steel table and it will take real weight. Lay the same pane on a mattress and it cracks under a fraction of that load, not because the glass got weaker but because the thing under it moved.
That’s what happens when you put a two-micron PVD coating directly onto a tool steel at 45 to 50 HRC. Under impact or high contact pressure, the substrate deforms elastically underneath a coating that has almost no ability to deform with it. The coating fractures and spalls, and the tool comes back looking like the coating “failed,” when the substrate is what actually failed.
The industry name for this is the eggshell effect, and it explains a large share of the coating disappointments manufacturers describe to us. The coating wasn’t wrong. It just had nothing holding it up.
Diffusion-treating first builds a hardened case that provides that support, transitioning gradually from the very hard surface into the tougher core. That’s the entire logic behind a duplex process like DYNA-MAX. If your application involves impact, high tonnage, or heavy contact stress and a straight coating has already let you down once, load support is almost certainly what you were missing.
Geometry and dimensions will narrow the field for you
Two constraints tend to settle the decision faster than any performance argument.
The first is geometry. If the surfaces that fail are inside something (a bore, a cooling passage, a deep narrow rib, a blind pocket), line-of-sight processes are out. Not “less effective.” Out. Diffusion doesn’t care about line of sight because it’s a chemical process happening at every exposed surface simultaneously.
The second is dimensional tolerance. A finished, ground, polished tool that has to hold tenths cannot go through a process that runs above its tempering temperature and comes back moved. Low-temperature diffusion treatments are dimensionally stable in a way that high-temperature processes are not, which is why they can be applied to finished tooling and to refurbished tooling after weld repair.
A working framework
Ask these in order:
- What’s the dominant failure mode? Adhesive, abrasive, thermal fatigue, or corrosion. Look at the failed tool, don’t guess.
- Where is it failing? Open faces, or somewhere a line-of-sight process can’t reach?
- What’s the substrate, and what’s its condition? Steel grade, hardness, whether it’s been welded or previously treated.
- How much movement can the part tolerate? Finished and in tolerance, or still has stock?
- Is there impact or heavy contact loading? If yes, load support is not optional.
Answer those five honestly and the right process is usually obvious. Answer none of them and you’re picking a coating by color.
Bring us a failed tool
The most useful thing you can send us isn’t a print. It’s the tool that failed, or good photos of it. Failure modes leave signatures, and thirty seconds of looking at the actual wear pattern is worth more than an hour of describing the application.
We run diffusion treatments, hard coatings, duplex processes, and vacuum hardening in house, which means we have no incentive to steer you toward the one process we happen to sell. Sometimes the right answer is the least expensive option on our list. We’d rather tell you that and have your tooling actually last.
Schedule a call and tell us what’s failing. Or run the numbers yourself first with our savings calculator to see what a change in tool life is worth against your current cycle counts.