Worked examples — Conservative vector fields — potential functions
Before we start, one reminder of the three tools we lean on, and why each:
- The gradient — the arrow field built from a hill (see Gradient and directional derivative). We reverse it to find the hill.
- The cross-partial / curl test — a fast yes/no filter so we don't hunt for a hill that doesn't exist (see Curl of a vector field, Clairaut's theorem (equality of mixed partials)).
- The Fundamental Theorem of Line Integrals (FTLI) — turns work integrals into subtraction (see Line integrals of vector fields).
The scenario matrix
Every conservative-field problem falls into one of these case classes. Each row is a "cell" the exam can hit; the last column says which example below covers it.
| Cell | Case class | What makes it tricky | Example |
|---|---|---|---|
| C1 | Plain 2D, both signs positive | baseline recipe | Ex 1 |
| C2 | 2D with negative / mixed-sign components | sign-tracking in | Ex 2 |
| C3 | Not conservative (curl ) | recipe must fail loudly | Ex 3 |
| C4 | Degenerate / punctured domain (hole) | test passes but field is NOT conservative | Ex 4 |
| C5 | FTLI to compute work over a path | endpoints only, ignore the curve | Ex 5 |
| C6 | 3D field, integrate three times | leftover depends on 2 vars | Ex 6 |
| C7 | Zero field / constant field (limiting/degenerate input) | boundary of the theory | Ex 7 |
| C8 | Real-world word problem (physics energy) | translate force potential | Ex 8 |
| C9 | Exam twist: closed loop + "which path?" trap | loop vs. hole | Ex 9 |
We now hit each cell.
Ex 1 — Cell C1: the clean baseline
Ex 2 — Cell C2: mixed signs everywhere
Ex 3 — Cell C3: the recipe must fail
Ex 4 — Cell C4: the punctured-domain trap (degenerate domain)
on the plane minus the origin. Does the curl test guarantee a potential?
Forecast: Parent note warned us. Guess: test passes but field is NOT conservative.

- Compute . With , quotient rule gives
- Compute . With , Why these steps? We must see with our own hands that everywhere the field is defined — so the naive test says "conservative."
- Take the loop integral around the unit circle , . Here , so and . Why this step? A conservative field has zero loop integral (FTLI, ). A nonzero loop is a hard disproof.
Answer: Not conservative on the punctured plane — the hole at the origin breaks the "simply connected" requirement (look at the red loop enclosing the missing point in the figure). See Green's theorem for why the enclosed hole matters.
Verify: symbolically, yet .
Ex 5 — Cell C5: FTLI to compute work along a path
Using from Ex 1, compute from to along any curve.
Forecast: Since Ex 1 proved it conservative, the wiggly path is irrelevant — only two numbers matter. Guess whether the work is positive or negative first.
- Recall the hill. (drop ; it cancels in subtraction). Why this step? FTLI says work , so we need at the two endpoints only.
- Evaluate at .
- Evaluate at .
- Subtract. Why this step? is the total "downhill/uphill" change; the path in between is invisible.
Answer: .
Verify: the in cancels (), so the answer is path- and constant-independent ✓.
Ex 6 — Cell C6: full 3D reconstruction
. Find in 3D.
Forecast: Three components mean three matching conditions. Guess: conservative, and has an piece plus an piece.
- Curl test (all three pairs).
- , ✓
- , ✓
- , ✓ Why this step? In 3D, needs all three cross-partial equalities (see Curl of a vector field).
- Integrate in . . Why this step? Now the invisible leftover depends on both remaining variables and .
- Match . , so (no ).
- Match . , so . Why these steps? Each further derivative peels away one variable's dependence from the leftover until only a constant remains.
Answer:
Verify: ✓.
Ex 7 — Cell C7: the degenerate limiting inputs
(a) The zero field . (b) The constant field . Are these conservative? Find .
Forecast: These are the "boundary" of the theory. Guess: both trivially conservative — a flat floor and a tilted plane.
- Zero field, test. ✓. Integrate: , and . Why this step? A constant hill () has zero slope everywhere — its gradient is the zero arrow field. This is the degenerate floor.
- Constant field, test. ✓. Integrate : ; match . Why this step? A constant force is the gradient of a flat tilted plane — uniform gravity is exactly this.
Answer: (a) . (b)
Verify: (a) ✓. (b) ✓.
Ex 8 — Cell C8: real-world energy problem
A particle feels the force (a 2D spring pulling toward the origin, , units N/m·m). It moves from to . How much work does do? Take .
Forecast: The particle starts and ends at the same distance … wait, , . It moves to a point closer/farther? Guess the sign of the work first (does the spring help or resist?).
- Test / recognise the potential. : ✓ conservative. Integrate: ; . So . Why this step? Spring force is minus the gradient of energy; here is the negative of the usual potential energy . This links to Conservation of energy in physics.
- Apply FTLI. Work . Why this step? Only the endpoints' distances from origin matter — the path is irrelevant for this conservative spring.
- Plug . Work joules. Why this step? Negative work: the particle ended farther from the origin (), so the inward spring resisted the motion. Sign matches intuition.
Answer: J.
Verify: units: ·length = (N/m)·m = N·m = J ✓. Sign negative because (moved against the pull) ✓.
Ex 9 — Cell C9: the exam-style closed-loop twist
Let (conservative, from the parent note). A student computes around the boundary of the triangle . They spend 20 minutes parametrising three edges. What's the shortcut answer, and where's the trap?
Forecast: Closed loop + conservative. Guess the answer instantly before any integral.
- Check conservativeness. ✓, domain all of (simply connected, no hole). Why this step? The C4 trap only bites when a hole is enclosed; here there is none, so the shortcut is legal.
- Use loop . For a conservative field on a hole-free domain, with , giving . Why this step? FTLI with equal endpoints — no parametrisation needed.
Answer:
The same-looking question with the punctured field of Ex 4 around a loop enclosing the origin gives , not . The distinguishing question is always: "Is the field conservative on a domain containing the whole loop's interior?" If a hole is enclosed, the shortcut is void — fall back to Green's theorem or direct computation.
Verify: direct sum of the three edge integrals of around the triangle totals (checked numerically).
Recall
Recall Which cell am I in?
Given a new field, what is your very first move? ::: Run the cross-partial / curl test — it filters out non-conservative fields before you waste time hunting for . The test passes but the domain has a hole around which the loop winds — conclusion? ::: You may NOT declare it conservative; check a loop integral (Cell C4). Closed loop, conservative field, no enclosed hole — value? ::: Exactly (Cell C9). Constant nonzero force field — is it conservative? ::: Yes; its potential is a tilted plane (Cell C7).
"Test, then trust the hole." Curl-zero is necessary; a hole-free domain makes it sufficient.