Same CAD Volume, Different Revision: A Moved-Hole Test

Two STEP revisions can have identical volume, mass, surface area and overall dimensions while putting a mounting hole in a different place. A property spreadsheet can therefore miss a change that matters at assembly.

Disclosure: Prepared on behalf of CADProps with AI assistance. This is an original synthetic geometry example with checked arithmetic, not a customer story or a claimed test of any CAD application.

Build a small, controlled comparison

Use millimetres and a fixed coordinate frame. Create a rectangular solid spanning x = 0 to 100, y = 0 to 60 and z = 0 to 8. Its uncut volume is 48,000 mm³.

Revision A has one cylindrical through-hole of radius 5 mm, with its axis parallel to z and its centre at (25, 20) in the xy plane. Revision B moves that hole centre to (40, 20). Keep the plate, hole radius, thickness, units and export coordinates unchanged. Do not add fillets or chamfers.

The hole moves 15 mm. The two hole regions do not overlap because their centres are more than two radii apart. Both remain fully inside the plate.

Why four familiar checks still pass

Each revision removes the same cylinder:

Hole volume = π × 5² × 8 = 200π ≈ 628.319 mm³.

So both revisions have:

• Solid volume: 48,000 − 200π ≈ 47,371.681 mm³.

• Bounding extents: 100 × 60 × 8 mm in the specified coordinate frame.

• Surface area: 2(100 × 60 + 100 × 8 + 60 × 8) − 2π × 5² + 2π × 5 × 8 = 14,560 + 30π ≈ 14,654.248 mm².

• Calculated mass: identical if both use the same uniform density. For an assumed density of 0.00785 g/mm³, each is about 371.868 g.

The area expression removes the two circular openings and adds the bore wall. Moving the hole does not change those areas. The density is an example input, not a material specification.

These values are useful screening checks, but their equality is not proof that revisions match. This example does not claim that every geometric property, such as centre of mass, remains equal.

Measure what changed, not only the net total

Let A and B denote the material occupied by the two solids in the same coordinate frame. A Boolean cut A minus B identifies material present only in A; B minus A identifies material present only in B. Their intersection identifies common material.

For this example, the expected ideal results are:

• A-only material: 200π ≈ 628.319 mm³, at the new hole position, where B has a void.

• B-only material: 200π ≈ 628.319 mm³, at the old hole position, which B fills.

• Common material: 48,000 − 400π ≈ 46,743.363 mm³.

The net volume change is zero, but the total non-overlapping difference is 400π ≈ 1,256.637 mm³. Reporting added and removed regions separately exposes the move. These regions describe geometry differences, not a proposed machining sequence.

A practical revision-check procedure

1. Identify the inputs. Record the two filenames, revision labels, export dates and, where practical, file hashes. A newly exported file is not automatically a new engineering revision.

2. Confirm the scope and units. Compare the intended bodies, with no missing assembly components. Use valid closed solids for solid-volume comparisons.

3. Establish alignment. Preserve the agreed datum frame or document the transformation used. A display-centering option is not geometric registration. Arbitrary best-fit alignment may hide a meaningful placement change.

4. Run both cuts. Inspect A-only and B-only regions separately, then the common region. Check the moved feature visually and dimension it from the required datums.

5. Record numerical settings. Note the software/version, model tolerances, Boolean tolerance and any warnings. A failed operation or unavailable value is not a zero difference.

6. Apply the engineering requirements. Compare the changed feature against the drawing and acceptance criteria. A computational tolerance does not certify manufacturing conformity.

A compact record to keep

For this synthetic test, a useful result is: “A → B; mm; unchanged source coordinates; hole centre moved from (25, 20) to (40, 20); volume change 0; A-only 628.319 mm³; B-only 628.319 mm³; common 46,743.363 mm³.” Add actual software settings and observed results when you run it. The figures here are analytical expectations, not observed software output.

Open CASCADE’s official modeling-algorithms documentation explains the Cut and Common operations used in this reasoning.

For a browser-based workflow, the CADProps comparison page documents automatic solid differences for supported closed STEP pairs and separate visual comparison for other supported 3D models. It uses source coordinates rather than automatic alignment, and files are uploaded for server processing. Check the current limits and your organisation’s data policy before using confidential files.

A reliable review keeps the property totals, the difference geometry and the changed feature’s dimensions together. None replaces the other.