Shot blasting profile: anchor pattern and adhesion

Two 3 mm CRCA brackets come off the same laser, go through the same powder batch at 70 µm and go into the same 480-hour salt spray. One passes. The other shows red rust along the cut edge at 200 hours. The difference is often not the powder or the pretreatment but the blast: one was profiled at 35 µm Rz, the other was degreased and coated with the original rolled surface intact.

Shot blasting profile is the anchor a coating grips. Too little and the film sits on a smooth surface with nothing to key into; too much and the peaks stand proud of a thin film and rust first. The target is a range, not a maximum, and the range depends on the coating system that follows.

This article covers the Rz range for powder and wet systems, the difference between angular grit and round shot, how to measure profile on the shop floor, the salt and dust that blasting can drive into the surface, and the point at which the blast itself damages a thin panel.

What profile does a coating need?

Powder at 60–80 µm needs 25–40 µm Rz. A high-build wet system at 160 µm and above needs 40–75 µm. Below 20 µm the anchor is inadequate and adhesion falls to a level where a cross-hatch or pull-off test exposes it; above 100 µm the peaks are taller than a single powder coat and the film over each peak is too thin to protect the steel.

Rz is the average peak-to-valley height over the assessment length, and it is the number the ISO 8503-1 comparators represent. It is not the same as Ra, the arithmetic average roughness that a machining drawing usually quotes. A grit-blasted surface can read Ra 6 µm and Rz 45 µm, because the profile is spiky rather than sinusoidal. Quoting Ra for a blast specification is the first place a coating enquiry goes wrong.

The relationship between profile and film is geometric. A 70 µm powder film over a 35 µm profile leaves 35 µm of film above the peaks and fills the valleys. The same film over a 90 µm profile leaves the peaks effectively bare, even though a DFT gauge placed on a peak reads 70 µm and a gauge placed in a valley reads 160 µm. The gauge is not lying; it is measuring a surface that should never have been blasted that deep for that film.

Shot blasting profile: the numbers to specify

The Ganesh column reflects a batch chamber rather than an inline machine. The 16 ft × 16 ft chamber takes frames, panels and brackets on hangers, and the operator sets pressure and standoff per part family. That is flexible, and it is also why the profile has to be written on the route card rather than assumed from a machine setting.

Shot versus grit: same number, different anchor

Round shot peens. It strikes the surface, deforms it and leaves a dimpled profile with rounded valleys, and because it cold-works the surface it also induces compressive stress. That is useful on parts that will carry fatigue load, and less useful when the job is adhesion, because a rounded dimple gives the coating less mechanical key than a sharp cut.

Angular grit cuts. Each particle is a small chisel, and the profile it leaves has undercuts and sharp edges that a coating can flow into and grip. At the same pressure and exposure time, grit produces a deeper profile than shot; to reach the same Rz with shot you blast longer, which puts more energy into the panel.

For a powder line on CRCA, angular grit at 25–40 µm Rz is the default. For a part where fatigue or stress corrosion matters, shot peening to a controlled intensity is a separate specification and should not be conflated with blast cleaning for coating.

Media selection also carries a contamination risk. A chamber that has run carbon steel with steel grit should not then blast stainless or aluminium. Embedded iron particles from the media rust on the surface within weeks, and on stainless they also initiate pitting. Non-ferrous media, a dedicated chamber or a thorough clean-down between materials is the control.

Pressure, standoff and the distortion limit

Blasting energy is pressure times time, and thin sheet punishes both. A 1.5 mm panel blasted at 7 bar with a 45° nozzle will oil-can across an unsupported span, and the distortion appears after coating, when it cannot be pressed out. The practical controls are:

• reduce pressure to 4–5 bar on sheet below 2 mm;
• increase standoff to 300 mm and work in passes rather than holding in one place;
• support the panel on a flat bed or a frame rather than hanging it free;
• blast both faces if the part is a visible panel, so the stress is balanced.

The last point is often skipped. A panel blasted on one face only is left with a compressive layer on that face and nothing on the other, and it bows towards the blasted side. On a 1.5 mm canopy door that is 3–5 mm of bow over a metre, which is enough to show at a gasket line.

Measuring profile without a laboratory

The fast method is the ISO 8503-1 comparator. Two comparator plates, one for grit and one for shot, are each divided into fine, medium and coarse segments, and the operator compares the blasted part by eye and by touch under good light. It takes seconds and it depends on the operator, which is why it is a screening check rather than a record.

The recordable method is replica tape. A foam-backed tape is pressed onto the surface, the foam compresses into the valleys, and a micrometer reads the total thickness; subtracting the tape backing gives the peak-to-valley profile. It is a spot check, it costs a tape per reading, and it produces a number that can go on a batch record alongside the DFT readings.

Whichever method is used, the reading has to be taken on the production surface, not on a test coupon blasted separately. A coupon in the corner of the chamber sees a different angle and a different exposure from the part in the middle, and the difference is often 10–15 µm.

The contamination that comes with the profile

Blasting does not only change the topography; it can drive contamination into it. Three sources matter:

Recycled abrasive. Fines, dust and salts accumulate in a recycled abrasive stream. The fines reduce cutting efficiency and the salts sit in the valleys under the coating. A separator that removes fines below 20 µm and a regular abrasive top-up keep the stream in condition.

Compressed air. Air from a compressor carries oil and water unless it is filtered and dried. A coalescing filter and a desiccant dryer at the chamber inlet are basic; a visible oil sheen on a white cloth held at the nozzle is the field test.

Salt on the incoming part. A part that has been stored near a pickling line or transported in a humid container can arrive with chloride on the surface, and blasting drives it into the profile rather than removing it. A Bresle extraction before blasting, or a conductivity check on the rinse after pretreatment, is the evidence.

The acceptance number for residual chloride before coating is commonly 20 mg/m². Above that, blistering and underfilm corrosion appear at 200–400 hours of salt spray, and no film build fixes it because the salt is under the coating.

What Ganesh actually runs

Blasting happens in a 16 ft × 16 ft chamber, which takes fabricated frames and panels on hangers rather than coil or cut sheet on a conveyor. Abrasive, pressure and standoff are set per part family and written into the route, and the parts move from the chamber to the seven-tank pretreatment line so the blasted surface does not sit long enough to flash rust.

That sequence is the point. A profile is only useful if the next stage reaches it before the surface oxidises, and the seven-tank line — alkaline clean, rinse, activation, conversion coating, rinse, demineralised rinse and dry-off — is what converts a freshly blasted surface into a coated one. The shot blasting capability page describes the chamber, and seven-tank pretreatment covers what happens next.

Where we are limited is thickness. A 0.8–1.0 mm panel cannot take a full-pressure blast without distorting, and on a visible part we would rather say so than deliver a part that is flat on the jig and bowed on the assembly line. Above 2 mm the chamber handles the full range, including the 40–75 µm profile a C4 wet system needs.

What to put on the drawing or the route card

1. Blast grade, Sa 2½ or Sa 3, with ISO 8501-1 named.
2. Profile range in Rz, matched to the coating system, not a single maximum.
3. Abrasive type and size, grit or shot, and the restriction on sharing media between materials.
4. Pressure limit for thin sheet, with the thickness at which it applies.
5. Areas to be masked, such as machined faces, bores and threads that must not be profiled.
6. Residual chloride limit, and the test method used to check it.
7. Hold time to pretreatment, stated as hours rather than "as soon as possible".
8. Adhesion acceptance, pull-off value to ISO 4624 and the failure mode required.

The profile article that pairs with this one is powder coating laser-cut edges, which covers what happens at the arris where the film thins, and phosphating and conversion coatings covers the layer that sits on the profile. The surface treatment hub indexes both. If the coating system is still open, send the duty and the drawing to request a quote and the blast and film numbers can be set together.

When blasting is the wrong preparation

A part with a machined finish. Blasting a bearing bore or a sealing face destroys the tolerance. Mask it, or blast before machining and accept that the profile is removed where the cut goes.

Thin visible panels below 1 mm. The distortion risk outweighs the adhesion gain. Use a chemical pretreatment and a coating that bonds to a cleaned rolled surface, or change the design so the panel is supported.

Stainless with shared media. If the chamber cannot be dedicated or cleaned, do not blast stainless in it. Embedded iron from carbon-steel media is a corrosion site that pickling will not always remove.

A part that will be welded after coating. Blasting before welding is wasted at the weld zone, because the weld burns the coating back and the profile with it. Weld, clean the weld, then blast the assembly.

A part already carrying salt. Blasting drives chloride into the profile. Wash and dry first, or blast and then run a salt-removal rinse before pretreatment, and verify with a conductivity reading rather than a visual check.