Ceramic Cross-Sections: Correct Apparent Thickness from an Oblique Cut

Measurement purpose

Determine whether apparent section width supports a local normal layer thickness after separating physical inclination from image-line orientation.

Specimens and conditions

Layer geometry

Resolved physical interfaces and a locally valid planar-layer or explicit alternative geometry

Section orientation

Known actual plane relative to a specimen datum, including the supported angular range

Equipment and records required

• Section preparation: Appropriate methods preserving the relevant interfaces and orientation record
• Dimensional observation: Calibrated imaging and an independently checked projection calculation

Method sequence

1. Definition

   Select physical interfaces and local thickness direction

   Record: Measurand and section-location map

2. Orientation

   Verify section inclination and measure shortest in-plane spacing

   Record: Uncorrected width and angle evidence

3. Interpretation

   Evaluate correction and sensitivity without masking preparation damage

   Record: Bounded normal thickness or unresolved orientation

Decision and uncertainty

Correct an apparent thickness only with established section geometry and valid interfaces; unknown inclination cannot be supplied from the expected thickness.

Angular uncertainty, nonparallel boundaries, calibration and preparation artifacts affect the result through different mechanisms.

The dimensional-method owner approves the correction; the drawing owner defines the required local or spatial thickness quantity.

Traceable outputs

Identify the two interfaces before measuring their separation

State whether the required quantity is conductor thickness, fired resistor thickness, glass overglaze thickness or the entire printed stack. A bright region in a section is not automatically one homogeneous layer. Overlaps and buried conductors can change the relevant lower boundary. Record the material stage and the location relative to pads, tracks and printed edges.

For a locally planar layer, true thickness is the shortest distance between its parallel physical interfaces. That definition is independent of how the mounted specimen is turned on the microscope stage. A surface step, average mass-derived thickness and local cross-section thickness answer different questions and should not share a value without an established relationship.

Separate an oblique cut from an oblique line drawn on the image

A normal section contains the layer normal and intersects the layer plane at ninety degrees. Tilting the section plane away from that condition increases the shortest interface spacing visible within the section. This is a three-dimensional orientation effect. Turning the finished image on the screen does not remove it.

Within any one section image, drawing a caliper at an angle to the shortest interface separation introduces another enlargement. First measure the shortest in-plane separation between the resolved boundaries. Then consider whether the section itself was oblique. Correcting the image-line angle while ignoring physical section inclination leaves the principal ambiguity unresolved.

Use an explicitly defined angular projection

Let alpha be the angle between the section plane and the local layer plane. Alpha equals ninety degrees for the desired normal section. For parallel planar interfaces, their shortest visible separation w equals true thickness t divided by sin alpha. Equivalently, define theta as departure from a normal cut, so theta equals ninety degrees minus alpha.

The correction then becomes t equals w times cos theta. Define the convention beside every calculation; an angle measured from the substrate plane cannot be substituted as if it were measured from the normal-section condition. At very shallow section angles the apparent width becomes large and increasingly sensitive to orientation error.

Check whether the apparent increase is consistent with inclination

For an illustrative true thickness of ten micrometres, a section thirty degrees away from normal shows approximately 11.55 micrometres. At sixty degrees away from normal it shows twenty micrometres. These are geometric examples, not thickness specifications or measured ChipSimple results.

Conversely, an apparent width of twelve micrometres at a verified thirty-degree departure corresponds to approximately 10.39 micrometres normal thickness. If the departure is only known to lie between twenty-five and thirty-five degrees, the corresponding geometric interval is approximately 9.83 to 10.88 micrometres, before image-scale and boundary uncertainties are added.

Measure orientation from the specimen, not the expected answer

Preserve the mounting and section-location record, including a physical datum that relates the polished plane to the original substrate. A nominal fixture setting is useful process information but may not describe the final plane after uneven grinding. Verify the relationship where an angular correction affects the engineering decision.

Do not infer the angle by assuming that the finished layer must have the expected thickness. That makes the correction circular and can hide a real process change. Where the actual orientation cannot be recovered, report the apparent spacing with its limitation and obtain another measurement rather than assigning a convenient angle.

Test the parallel-layer approximation at the measurement location

Printed edges, overlaps and glaze menisci can produce curved or wedge-shaped interfaces. A single correction derived from the substrate plane does not necessarily describe the local layer normal. Inspect a wider field around the measurement and identify whether the reported quantity is a local value, an average over a defined length or a minimum required by the drawing.

A section through a cylindrical via introduces additional curvature and section-position effects. Off-centre wall intersections cannot be treated as a flat-film example merely by rotating the image. Establish the relevant geometry or use a section through the intended axis before interpreting local wall dimensions.

Keep projection correction separate from preparation quality

Polishing can round exposed edges, smear a conductor or remove a fragile region. Those changes alter the physical boundaries being measured; multiplying by a cosine does not restore them. Retain earlier preparation observations and consider an independently prepared section when the dimensional conclusion depends on a questionable edge.

For an independent comparison, align the quantity and location. A surface profilometer may measure the height above adjacent ceramic, while the section resolves a buried layer between different interfaces. Agreement is useful only after those physical definitions are reconciled. A correction should not be tuned to force agreement between unlike measurements.

Release a traceable corrected quantity with its orientation range

The dimensional record should contain specimen identity, section plane, material interfaces, image calibration, in-plane line direction, measured spacing and orientation evidence. Include the uncorrected result and the angular convention so another reviewer can reproduce the projection. Propagate angle uncertainty along with the ordinary dimensional measurement terms.

ChipSimple can review drawing-specific thickness and section-location requirements with the project team. The accepted outcome is a physically defined local thickness, or a clear statement that orientation prevents that determination. Neither a visually wide layer nor a mathematically corrected number establishes a manufacturing capability without the corresponding specimen and method evidence.