2D-3D Visualization

Fold, rotate, and project without losing the reference frame

Use this lesson to turn a flat pattern into a solid, reverse a sequence of paper folds, and reconcile two-dimensional views. Keep a fixed coordinate convention: x = width, y = depth, z = height for solid objects. For paper, x runs rightward and y upward on the sheet. The coordinates define the practice problems; they are not an extra calculation requirement on the PAT.

1. What a view tells you—and what it leaves unknown

A top orthographic view retains width and depth; a front view retains width and height; a right-end view retains depth and height. Parallel viewing rays collapse the dimension along the viewing direction. Internal edges and hidden-line conventions can provide further structural clues, so a view is more than its outer outline.

A front drawing alone cannot establish depth. For example, consider a stack with four columns of heights 1, 2, 3, 3. If it is explicitly one cube deep, the total is 1 + 2 + 3 + 3 = 9 cubes. If a second identical depth row exists behind it, the same front outline can represent 18 cubes.

One-cube-deep practice stack, viewed from front:

z=3        [C][D]
z=2     [B][C][D]
z=1  [A][B][C][D]
      x=1  2  3  4

Labels identify columns; each bracket is one cube. Every upper cube has support. Count by layers: 4 + 3 + 2 = 9. Count by columns: 1 + 2 + 3 + 3 = 9. Agreement is a useful completeness check. Do not infer a filled volume from a silhouette unless the problem or official drawing conventions support that inference.

For the PAT's painted-face task, the number of cubes is only the first step: classify exposed faces with the table-facing bottom unpainted. A “shadow” is not evidence that every hidden space must contain a cube. Official PAT conventions.

2. Fold this cube net one hinge at a time

         [U]
          |
    [L]—[F]—[R]—[B]
          |
         [D]

Each symbol is one equal square. Hold F still as the front. Fold U upward into the top, D into the bottom, and L/R into the side faces; B closes the back.

Faces sharing an edge in a valid net stay adjacent. The converse is false: edges that were cut apart can meet when the cube closes. For a straight three-square strip, its first and third squares become opposite. A two-edge path that turns a corner is different: L–F–U also has one intervening square, but L and U are adjacent.

Do not extend the shortcut to a strip of four. In L–F–R–B, L/R and F/B are opposite; L/B are adjacent. When a net has no convenient straight strip, track the 90° folds instead of counting squares along an arbitrary path.

Marked faces must satisfy two tests

First check which faces meet; then check the orientation of the marks. Suppose U, L, and R each have a centered dot while F, B, and D are blank. The folded cube has three dotted faces in total. However, all three cannot be shown as the three exterior faces meeting at one visible corner, because L and R are opposite. A candidate corner view showing three dotted faces is impossible.

A centered dot checks face identity but not rotation within the face. For an off-center dot or arrow, track its position relative to the shared edge as the face turns. Adjacency alone cannot validate that candidate.

3. Rotation is not a screen-side rule

For a mark on the page, clockwise rotations give:

0°          90°         180°        270°
 ↑           →           ↓           ←

This is a rotation in the drawing plane, viewed from the same side. In 3D, always state the axis and viewing direction before saying “clockwise.” A feature can move from screen-left to screen-right through ordinary rotation, so switching sides does not prove reflection.

A proper rotation preserves distances, connectivity, and handedness. A chiral object cannot be rotated onto its mirror image. An unmarked cube is achiral; its reflection can be superimposed. For a flat shape, turning the sheet over in 3D must not be confused with rotating it in its plane.

Home exercise: put distinct marks on three adjacent faces of a box. Predict their positions after a turn, then turn the box and check. Explain the mismatch if your prediction fails; no claim about brain growth is needed to make this exercise useful.

4. Reverse paper folds with coordinates

A reflection across x = a sends (x, y) to (2a − x, y). Across y = b, it sends (x, y) to (x, 2b − y). These formulas express equal perpendicular distances from the crease. In a diagram, you can count equal grid steps instead.

Two perpendicular half folds

Start with a 4 × 4 square. Fold left onto right at x = 2, then bottom onto top at y = 2. Punch a small interior hole at (3.5, 2.5) in the upper-right packet, away from boundaries.

1. Undo the last fold, y = 2: add (3.5, 1.5).
2. Undo x = 2: add (0.5, 2.5) and (0.5, 1.5).

Holes shown at cell centers on the original 4 × 4 sheet:

         x=.5  1.5  2.5  3.5
y=3.5      ·    ·    ·    ·
y=2.5      ●    ·    ·    ●
y=1.5      ●    ·    ·    ●
y=0.5      ·    ·    ·    ·

There are four holes, near the side edges and middle of the sheet—not near all four outer corners. A punch “near a corner of the folded packet” does not tell you it is near a corner of the original sheet.

Two parallel half folds

Use a unit square. Fold the right half left at x = 0.5, then fold the right half of that packet left at x = 0.25. Punch at (0.05, 0.40).

Undo x = 0.25 to get x = 0.05 and 0.45. Undo x = 0.5 to add x = 0.95 and 0.55. Sorted, the four positions have x coordinates 0.05, 0.45, 0.55, 0.95, all with y = 0.40. The gaps are 0.40, 0.10, 0.40: collinear does not mean evenly spaced.

Why 2ⁿ is conditional

Complete halving of the whole packet doubles its layers. For n such folds, an interior punch through every layer gives 2ⁿ distinct holes. In an arbitrary partial fold, the layer count can vary across the packet. A punch crossing a crease or edge needs separate geometric analysis rather than automatic doubling.

Try it: three complete half folds give eight layers. An interior punch through all eight gives eight holes; that count still does not determine where the holes go. Undo the actual sequence to locate them.

5. Compare angle openings consistently

Mentally align one ray of each angle to a common direction, then compare the second rays. The opening is unchanged by rotating the whole angle. Acute means between 0° and 90°; obtuse means between 90° and 180°.

Labeled practice: A = 35°, B = 90°, C = 65°, D = 20°. The ascending order is D < A < C < B. All of A, C, and D are acute, so categorization alone cannot order them. This exercise checks the logic of ranking; use the official visual examples to practice estimating closely spaced, unlabeled angles.

If two candidates are close, compare that pair directly. A longer ray or a steeper-looking line on the screen is not itself a larger opening. Do not infer exact degree values from an imprecise text sketch.

6. A repeatable spatial check

For each practice problem, state the reference view, track one transformation at a time, and check the result against every supplied constraint. For cubes, reconcile a layer total with a column total. For folds, verify the coordinates and number of layers. For nets, check opposite faces before the markings.

The PAT's 60 minutes for 90 questions allows 40 seconds per item on average. Work accurately within an overall budget; an instruction to “value accuracy” is not permission to leave most items unanswered. There is no guessing penalty. Put down a provisional choice, flag difficult items, and keep moving. Current exam guide.