Where Should Roughness Be Measured on Curved or Finned Cast Surfaces?
Roughness should be measured at drawing-defined locations that represent the surface function and the actual blasting exposure. On curved or finned castings, that usually means separate locations for accessible flats, cylindrical walls, fin sides or roots, and edge-transition zones. The easiest flat should not stand in for geometry that receives different impact energy.
The selected instrument must fit the location and separate roughness from part form. Where a contact stylus cannot seat or travel correctly, buyers may need a qualified optical, replica or correlated witness method rather than an improvised reading.
How Curvature Affects a Stylus Trace
A stylus moving along a cylinder or small radius records both surface texture and geometric form. Filtering may remove some form, but instrument settings and radius influence the result. A fixture that tilts between readings adds another source of variation. Define whether the trace runs axially or circumferentially, the angular location, radius range and method used to remove form.
Curvature should remain within the instrument and procedure capability. If the stylus skid or body contacts the part, or the tip loses normal contact, the result is not valid simply because the instrument displayed a number.
Geometry | Measurement Concern | Planning Response |
|---|---|---|
Large-radius cylinder | Form enters trace | Fix axial/circumferential direction and filtering |
Small-radius boss | Stylus seating and travel are limited | Use qualified optical/replica method or no numerical zone |
Fin side | Vibration and narrow support | Support part, define side and instrument access |
Fin root | Stylus cannot reach or align | Optical/replica study plus visual coverage control |
Hole edge | Tip falls into transition | Set exclusion distance and inspect edge separately |
How to Choose Representative Blast-Exposure Zones
Map direct-impact, glancing-impact and shadowed areas from the approved fixture and nozzle path. Fin tips and exposed rims may receive maximum energy; roots and deep pockets may receive minimum coverage. Measure at least one stable location from each functionally important exposure class during first article. The central broad face can remain a routine trend point only after it correlates with the wider map.
Also separate cast skin, trimmed surfaces and machined zones. Their starting topography and hardness can produce different readings under one blast recipe. If a machined land is intentionally masked, record it as a protected functional surface rather than a failed low-roughness location.
When Optical, Replica or Witness Methods Are Appropriate
Optical instruments can measure without stylus contact and may reach textures on limited curvature, but focus, reflectivity, stitching, filtering and areal-versus-profile parameters must be controlled. A replica can transfer a difficult surface to an accessible measurement, but replica material, curing, removal and reconstruction add uncertainty. A witness coupon is useful only if its response is correlated with the actual part location.
Method | Strength | Qualification Limitation |
|---|---|---|
Contact stylus | Established profile parameters on accessible paths | Access, direction and curvature |
Optical measurement | Non-contact and areal data | Reflectivity, focus, filter and parameter correlation |
Surface replica | Transfers inaccessible feature | Replication fidelity and removal technique |
Witness coupon | Easy routine monitoring | May not reproduce cast skin or shadowed geometry |
Visual master | Covers whole-zone appearance | Does not quantify functional roughness |
How to Document Locations So Suppliers Repeat Them
Use a drawing detail or controlled coordinate map with datum references, angular positions, distances from edges, trace direction and location size. A photograph with a hand-drawn circle can support training but is weak as the only production definition. Identify the instrument program and fixture orientation in the inspection plan.
Measure enough times during capability work to understand operator and placement variation. If repositioning error is close to the acceptance width, improve the fixture or method before using the value to reject production. The sand blasting process should not be adjusted to chase measurement noise.
What to Do When the Critical Surface Cannot Be Measured Directly
Start with the function. If the inaccessible fin root only needs visual uniformity, a master and coverage inspection may be more appropriate than a forced Ra requirement. If it is a coating-preparation surface, qualify the complete cleaning, pretreatment and coating performance at that geometry using sectioning or adhesion evidence where suitable. If it is a seal or bearing interface, it may need redesign, masking or machining rather than an unverifiable blasted finish.
A correlated witness should be placed and oriented to experience the relevant blast condition, then periodically confirmed against destructively measured or optically inspected parts. The post-process plan should state the correlation and reaction when the witness moves outside its limit.
Inspection reports should identify any location that could not be measured as planned. Replacing it with another point requires engineering approval and an updated map; otherwise every batch can migrate toward the easiest surface until the data no longer represents the critical geometry.
Valid roughness data comes from representative geometry, controlled access and a repeatable method. Curved and finned castings usually need a measurement map, not one convenient reading used for the entire component.
