- Foundation
- About 14 min
- 5 practice questions
Slack, binding constraints and checking a plan
Test whether a plan satisfies every constraint, measure the unused amount (slack) of each resource, and identify binding constraints, which are the ones that meet at a corner point.
If something looks wrong, please tell us.
Before solving a linear programming problem it helps to know how to judge a single plan: is it allowed, and how much of each resource does it leave unused? This language of slack and binding constraints is used in every later lesson. It also explains why a corner point is where two constraints are "tight" at once.
Is a plan feasible?
A plan (x, y) is feasible when it satisfies every constraint, including x ≥ 0 and y ≥ 0. Substitute the numbers into each inequality in turn; one failure is enough to make the plan infeasible. Suppose a problem has the constraints 2x + y ≤ 10 and x + 3y ≤ 15 with x, y ≥ 0. The plan (3, 3) gives 2(3) + 3 = 9 ≤ 10 and 3 + 9 = 12 ≤ 15, so it is feasible. The plan (4, 4) gives 2(4) + 4 = 12 > 10, so it is infeasible.
Slack and surplus
For a constraint of the form ax + by ≤ c, the slack is c − (ax + by): the amount of the resource left unused. For a constraint ax + by ≥ c the surplus is (ax + by) − c: the amount by which the requirement is exceeded. A feasible plan has no negative slack or surplus. A constraint with zero slack or surplus is binding (or tight): the plan uses the resource completely, so the point lies exactly on that boundary line.
A corner point of a two-variable feasible region lies on at least two boundary lines, so at least two constraints (counting x ≥ 0 and y ≥ 0) are binding there.
Worked example 1: slack at a plan
For the constraints above, find the slack at the plan (3, 3). In 2x + y ≤ 10 the left side is 9, so the slack is 10 − 9 = 1. In x + 3y ≤ 15 the left side is 12, so the slack is 15 − 12 = 3. No constraint is binding, so the plan is feasible but sits in the interior of the region, not on its boundary. Since the plan leaves resources unused, a better plan may exist.
Worked example 2: finding the corners and their binding constraints
Take the same constraints. The corners come from pairs of binding lines. Solving 2x + y = 10 and x + 3y = 15 (multiply the first by 3 and subtract: 6x + 3y − x − 3y = 30 − 15, so 5x = 15, x = 3, y = 4) gives the corner (3, 4). The other corners are (0, 0), (5, 0) and (0, 5).
| Corner | 2x + y ≤ 10 | x + 3y ≤ 15 | Binding constraints |
|---|---|---|---|
| (0, 0) | slack 10 | slack 15 | x ≥ 0 and y ≥ 0 |
| (5, 0) | slack 0 | slack 10 | 2x + y ≤ 10 and y ≥ 0 |
| (3, 4) | slack 0 | slack 0 | both resource limits |
| (0, 5) | slack 5 | slack 0 | x + 3y ≤ 15 and x ≥ 0 |
Every corner has two binding constraints, as the rule says. Now suppose the objective is to maximise Z = 3x + 2y. The values are Z = 0 at (0, 0), 15 at (5, 0), 17 at (3, 4) and 10 at (0, 5). The maximum is 17 at (3, 4), where both resource limits are binding: both resources are fully used.
Worked example 3: a surplus in a minimisation problem
A diet must supply at least 12 units of vitamin and at least 12 units of mineral, with foods giving 3x + y units of vitamin and x + 3y units of mineral. Take the plan (4, 4). Vitamin: 3(4) + 4 = 16, surplus 16 − 12 = 4. Mineral: 4 + 3(4) = 16, surplus 4. Both requirements are met with spare, so the plan is feasible but not the cheapest possible. The plan (3, 3) gives 12 for each: surplus 0 in both, so both are binding, and (3, 3) is a corner point.
Why this matters
When you read a final answer you should ask which constraints are binding. A binding resource limit is a bottleneck: relaxing it could improve the result. A resource with positive slack at the optimum is not limiting the plan. Later lessons use this idea to explain the value of extra resources.
Key takeaways: feasibility means every constraint holds; slack is c minus the left side for ≤ constraints, surplus is the left side minus c for ≥ constraints; zero means binding; a corner has at least two binding constraints.
Check your understanding
-
1. With the constraint 2x + y ≤ 10, what is the slack at the plan (2, 3)?
- 7
- 0
- 3
- 13
Show answer
Answer: C. 2(2) + 3 = 7, so the slack is 10 − 7 = 3.
-
2. With the constraint x + 3y ≤ 15, what is the slack at the plan (1, 4)?
- 2
- 13
- 0
- 4
Show answer
Answer: A. x + 3y = 1 + 12 = 13, so the slack is 15 − 13 = 2.
-
3. A constraint is binding at a plan when:
- its slack is the largest
- the plan is the origin
- the constraint is a ≥ constraint
- its slack or surplus is zero
Show answer
Answer: D. Binding means the plan lies exactly on the boundary line, so the unused amount is zero.
-
4. How many binding constraints does a corner point of a two-variable feasible region have at least?
- 1
- 2
- 3
- 0
Show answer
Answer: B. A corner is where two boundary lines meet, so at least two constraints (possibly x ≥ 0 or y ≥ 0) are binding.
-
5. In a diet problem the requirement is 3x + y ≥ 12. At (4, 4) the surplus is:
- 4
- 0
- 12
- 16
Show answer
Answer: A. 3(4) + 4 = 16, and 16 − 12 = 4.
Sign in to keep track of the lessons you finish.