Chapter Three: Linear Programming: Computer Solution and Sensitivity Analysis

PROBLEM SUMMARY
1. QM for Windows 2. QM for Windows and Excel 3. Excel 4. Graphical solution; sensitivity analysis 5. Model formulation 6. Graphical solution; sensitivity analysis (3–5) 7. Sensitivity analysis (3–5) 8. Model formulation 9. Graphical solution; sensitivity analysis (3–8) 10. Sensitivity analysis (3–8) 11. Model formulation 12. Graphical solution; sensitivity analysis (3–11) 13. Computer solution; sensitivity analysis (3–11) 14. Model formulation 15. Graphical solution; sensitivity analysis (3–14) 16. Computer solution; sensitivity analysis (3–14) 17. Model formulation 18. Graphical solution; sensitivity analysis (3–17) 19. Computer solution; sensitivity analysis (3–17) 20. Model formulation 21. Graphical solution; sensitivity analysis (3–20)
90

35. Model formulation 36. Graphical solution; sensitivity analysis (3–35) 37. Computer solution; sensitivity analysis (3–35) 38. Model formulation; computer solution 39. Sensitivity analysis (3–38) 40. Model formulation; computer solution 41. Sensitivity analysis (3–40) 42. Model formulation 43. Computer solution; sensitivity analysis (3–42) 44. Model formulation 45. Computer solution; sensitivity analysis (3–44) 46. Model formulation 47. Computer solution; sensitivity analysis (3–46) 48. Model formulation 49. Computer solution; sensitivity analysis (3–48) 50. Computer solution

PROBLEM SOLUTIONS
1.
x2
110 100

22. Computer solution; sensitivity analysis (3–20) 23. Model formulation 24. Graphical solution; sensitivity analysis (3–23) 25. Computer solution; sensitivity analysis (3–23) 26. Model formulation 27. Graphical solution; sensitivity analysis (3–26) 28. Computer solution; sensitivity analysis (3–26) 29. Model formulation 30. Graphical solution; sensitivity analysis (3–29) 31. Computer solution; sensitivity analysis (3–29) 32. Model formulation 33. Model formulation; computer solution 34. Computer solution; sensitivity analysis
0 10 20 30 40 50 60 70 80 90 100 110 80 70 60 50

A 40 B
30 20

optimal point: x1 = 15.29 x2 = 38.24 Z = 4,205.88

C

Z 10
x1

17

2.

QM for Windows establishes a “template” for the linear programming model based on the user’s specification of the type of objective function, the number of constraints and number of variables, then the model parameters are input and the problem is solved. In Excel the model “template” must be developed by the user. Changing cells: B10:B12 Constraints: B10:B12 0 G6 F6 G7 F7 Profit: B10 * C4 B11 * D4

6.

x2
300 250 200

*A: x1 = 0 x2 = 160 Z = 2,560 B: x1 = 128.5 x2 = 57.2 Z = 2457.2 C : x1 = 167 x2 = 0 Z = 2,004 B Z
0 50 100

A
150 100

3.

50

Point A is optimal
250 300 350

150 C 200

x1

B12 * E4 (a) A: 3(0) + 2(160) + s1 = 500 s1 = 180 4(0) + 5(160) + s2 = 800 s2 = 0 B: 3(128.5) + 2(57.2) + s1 = 500 s1 = 0 4(128.5) + 2(57.2) + s2 = 800 s2 = 0 C: 2(167) + 2(0) + s1 = 500 s1 = 0 4(167) + 5(0) + s2 = 800 s2 = 132

4.

The slope of the constraint line is –70/60. The optimal solution is at point A where x1 = 0 and x2 = 70. To change the solution to B, c1 must increase such that the slope of the objective function is at least as great as the slope of the constraint line, –c1/50 = –70/60 c1 = 58.33 Alternatively, c1 must decrease such that the slope of the objective function is at least as great as the slope of the constraint line, –30/c2 = –70/60 c2 = 25.71 Thus, if c1 increases to greater than 58.33 or c2 decreases to less than 25.71, B will become optimal.

(b) Z = 12x1 + 16x2 and, x2 = Z/16 – 12 x1/16 The slope of the objective function, –12/16, would have to become steeper (i.e., greater) than the slope of the constraint line 4x1 + 5x2 = 800, for the solution to change. The profit, c1, for a basketball that would change the solution point is, 4/5 = –c1/16 5c1 = 64 c1 = 12.8 Since $13 > 12.8 the solution point would change to B where x1 = 128.5, x2 = 57.2. The new Z value is $2,585.70. For a football, –4/5 = –12/c2 4c2 = 60 c2 = 15 Thus, if the profit for a football decreased to $15 or less, point B will also be optimal (i.e., multiple optimal solutions). The solution at B is x1 = 128.5, x2 = 57.2 and Z = $2,400. 18

5. (a) x1 = no. of basketballs x2 = no. of footballs maximize Z = 12x1 + 16x2 subject to 3x1 + 2x2 ≤ 500 4x1 + 5x2 ≤ 800 x1, x2 ≥ 0 (b) maximize Z = 12x1 + 16x2 + 0s1 + 0s2 subject to 3x1 + 2x2 + s1 = 500 4x1 + 5x2 + s2 = 800 x1, x2, s1, s2 ≥ 0

(c) If the constraint line for rubber changes to 3x1 + 2x2 = 1,000, it moves outward, eliminating points B and C. However, since A is the optimal point, it will not change and the optimal solution remains the same, x1 = 0, x2 = 160 and Z = 2,560. There will be an increase in slack, s1, to 680 lbs. If the constraint line for leather changes to 4x1 + 5x2 = 1,300, point A will move to a new location, x1 = 0, x2 = 250, Z = $4,000. 7. (a) For c1 the upper limit is computed as –4/5 = –c1/16 5c1 = 64 c1 = 12.8 and the lower limit is unlimited. For c2 the lower limit is, –4/5 = –12/c2 4c2 = 60 c2 = 15 and the upper limit is unlimited. Summarizing, ∞ ≤ c1 ≤ 12.8 15 ≤ c2 ≤ ∞ For q1 the upper limit is ∞ since no matter how much q1 increases the optimal solution point A will not change. The lower limit for q1 is at the point where the constraint line 3x1 + 2x2 = q1 intersects with Objective Coefficient Ranges Variables x1 x2 Lower Limit No limit 15.000 Current Values 12.000 16.000 Upper Limit 12.800 No limit

point A where x1 = 0, x2 = 160, 3x1 + 2x2 = q1 3(0) + 2(160) = q1 q1 = 320 For q2 the upper limit is at the point where the rubber constraint line (3x1 + 2x2 = 500) intersects with the leather constraint line (4x1 + 5x2 = 800) along the x2 axis, i.e., x1 = 0, x2 = 250, 4x1 + 5x2 = q2 4(0) + 5(250) = q2 q2 = 1,250 The lower limit is 0 since that is the lowest point on the x2 axis the constraint line can decrease to. Summarizing, 320 ≤ q1 ≤ ∞ 0 ≤ q2 ≤ 1,250 (b) Z = 2560.000 Variable x1 x2 Constraint c1 c2 Value 0.00 160.000 Slack/Surplus 180.00 0.00 Reduced Cost 0.800 0.000 Shadow Price 0.00 3.20

Allowable Increase 0.800 No limit

Allowable Decrease No limit 1.000

Right Hand Side Ranges Constraints c1 c2 Lower Limit 320.000 0.000 Current Values 500.000 800.000 Upper Limit No limit 1250.000 Allowable Increase No limit 450.000 Allowable Decrease 180.000 800.000

19

(c) The shadow price for rubber is $0. Since there is slack rubber left over at the optimal point, extra rubber would have no marginal value. The shadow price for leather is $3.20. For each additional ft.2 of leather that the company can obtain profit would increase by $3.20, up to the upper limit of the sensitivity range for leather (i.e., 1,250 ft.2). 8.(a) x1 = no. of units of A x2 = no. of units of B maximize Z = 9x1 + 7x2 subject to 12x1 + 4x2 ≤ 60 4x1 + 8x2 ≤ 40 x1,x2 ≥ 0 (b) maximize Z = 9x1 + 7x2 + 0s1 + 0s2 subject to 12x1 + 4x2 + s1 = 60 4x1 + 8x2 + s2 = 40 x1, x2, s1, s2 ≥ 0 9.
x2 30 25 20 15 10 5

(b) The constraint line 12x1 + 4x2 = 60 would move inward resulting in a new location forpoint B at x1 = 2, x2 = 4, which would still be optimal. (c) In order for the optimal solution point to change from B to A the slope of the objective function must be at least as flat as the slope of the constraint line, 4x1 + 8x2 = 40, which is –1/2. Thus, the profit for product B would have to be, –9/c2 = –1/2 c2 = 18 If the profit for product B is increased to $15 the optimal solution point will not change, although Z would change from $57 to $81. If the profit for product B is increased to $20 the solution point will change from B to A, x1 = 0, x2 = 5, Z = $100. 10.(a) For c1 the upper limit is computed as, –c1/7 = –3 c1 = 21 and the lower limit is, –c1/7 = –1/2 c1 = 3.50 For c2 the upper limit is, –9/c2 = –1/2 c2 = 18 and the lower limit is, –9/c2 = –3 c2 = 3 Summarizing, 3.50 ≤ c1 ≤ 21 –3 ≤ c2 ≤ 18 (b) Z = 57.000 Variable x1 x2 Constraint c1 c2 Value 4.000 3.000 Slack/Surplus 0.000 0.000 Reduced Cost 0.000 0.000 Shadow Price 0.550 0.600

A: x1 = 0 x2 = 5 Z = 35 *B: x1 = 4 x2 = 3 Z = 57 C: x1 = 5 x2 = 0 Z = 45 A B C
0 5 10 15 20 25 30 35 40

Point B is optimal
x1

(a) A: 12(0) + 4(5) + s1 = 60 s1 = 40 4(0) + 8(5) + s2 = 40 s2 = 0 B: 12(4) + 4(3) = 60 s1 = 0 4(4) + 8(3) + s2 = 40 s2 = 0

C: 12(5) + 4(0) + s1 = 60 s1 = 0 4(5) + 8(0) + s2 = 40 s2 = 20

20

Objective Coefficient Ranges Variables x1 x2 Lower Limit 3.500 3.000 Current Values 9.000 7.000 Upper Limit 21.000 18.000 Allowable Increase 12.000 11.000 Allowable Decrease 5.500 4.000

Right Hand Side Ranges Constraints c1 c2 Lower Limit 20.000 20.000 Current Values 60.000 40.000 Upper Limit 120.000 120.000 (a) Allowable Increase 60.000 80.000 Allowable Decrease 40.000 20.000

(c) The shadow price for line 1 time is $0.55 per hour, while the shadow price for line 2 time is $0.60 per hour. The company would prefer to obtain more line 2 time since it would result in the greatest increase in profit. 11.(a) x1 = no. of yards of denim x2 = no. of yards of corduroy maximize Z = $2.25x1 + 3.10x2 subject to 5.0x1 + 7.5x2 ≤ 6,500 3.0x1 + 3.2x2 ≤ 3,000 x2 ≤ 510 x1, x2 ≥ 0 (b)maximize Z = $2.25x1 + 3.10x2 + 0s1 + 0s2 + 0s3 subject to 5.0x1 + 7.5x2 + s1 = 6,500 3.0x1 + 3.2x2 + s2 = 3,000 x2 + s3 = 510 x1, x2, s1, s2, s3 ≥ 0 12.

5.0(456) + 7.5(510) + s1 = 6,500 s1 = 6,500 – 6,105 s1 = 395 lbs. 3.0(456) + 3.2(510) + s2 = 3,000 s2 = 0 hrs. 510 + s3 = 510 s3 = 0 therefore demand for corduroy is met.

(b) In order for the optimal solution point to change from B to C the slope of the objective function must be at least as great as the slope of the constraint line, 3.0x1 + 3.2x2 = 3,000, which is –3/3.2. Thus, the profit for denim would have to be, –c1/3.0 = –3/3.2 c1 = 2.91 If the profit for denim is increased from $2.25 to $3.00 the optimal solution would change to point C where x1 = 1,000, x2 = 0, Z = 3,000. Profit for corduroy has no upper limit that would change the optimal solution point. (c) The constraint line for cotton would move inward as shown in the following graph where point C is optimal.

x2 1600 1400 1200 1000 800 600 A 400 200

A: x1 = 0 x2 = 510 Z = $1,581 *B: x1 = 456 x2 = 510 Z = $2,607 C: x1 = 1,000 x2 = 0 Z = $2,250

x2 1600 1400 1200 1000 800

C, optimal x1 = 1,000 x2 = 0 Z = $2,250 B C

B

600 A 400 200

C
0 200 400 600 800 1000 1200 1400 1600

x1 0 200 400 600

D
800 1000 1200 1400 1600

x1

21

13. Z = 2607.000 Variable x1 x2 Constraint c1 c2 c3 Value 456.000 510.000 Slack/Surplus 395.000 0.000 0.000 Reduced Cost 0.000 0.000 Shadow Price 0.000 0.750 0.700

Objective Coefficient Ranges Variables x1 x2 Lower Limit 0.000 2.400 Current Values 2.250 3.100 Upper Limit 2.906 No limit Allowable Increase 0.656 No limit Allowable Decrease 2.250 0.700

Right Hand Side Ranges Constraints c1 c2 c3 Lower Limit 6015.000 1632.000 0.000 Current Values 6500.000 3000.000 510.000 Upper Limit No limit 3237.000 692.308 Allowable Increase No limit 237.000 182.308 15.
x2 8 7 6 5 4 3 2 1

Allowable Decrease 395.000 1368.000 510.000

(a) The company should select additional processing time, with a shadow price of $0.75 per hour. Cotton has a shadow price of $0 because there is already extra (slack) cotton available and not being used so any more would have no marginal value. (b) 6,105 ≤ q1 ≤ ∞ 1,632 ≤ q2 ≤ 3,237 0 ≤ q3 ≤ 692.308 The demand for corduroy can decrease to zero or increase to 692.308 yds. without changing the current solution mix of denim and corduroy. If the demand increases beyond 692.308 yds., then denim would no longer be produced and only corduroy would be produced. x1 = no. of days to operate mill 1 x2 = no. of days to operate mill 2 minimize Z = 6,000x1 + 7,000x2 subject to 6x1 + 2x2 ≥ 12 2x1 + 2x2 ≥ 8 4x1 + 10x2 ≥ 5 x1, x2 ≥ 0 22 0 ≤ c1 ≤ 2.906 2.4 ≤ c2 ≤ ∞

A: x1 = 0 x2 = 6 Z = 42,000 A B: x1 = 1 x2 = 3 Z = 27,000

*C: x1 = 4 x2 = 0 Z = 24,000

B

14.

C
0 1 2 3 4 5 6 7 8

x1

(a) 6(4) + 2(0) – s1 = 12 s1 = 12 2(4) + 2(0) – s2 = 8 s2 = 0 4(4) + 10(0) – s3 = 5 s3 = 11 (b) The slope of the objective function, –6000/7,000 must become flatter (i.e., less) than the slope of the constraint line, 2x1 + 2x2 = 8, for the solution to change. The cost of operating Mill 1, c1, that would change the solution point is, –c1/7,000 = –1 c1 = 7,000 Since $7,500 > $7,000, the solution point will change to B where x1 = 1, x2 = 3, Z = $28,500. (c) If the constraint line for high-grade aluminum changes to 6x1 + 2x2 = 10, it moves inward but Objective Coefficient Ranges Variables x1 x2 Lower Limit 0.000 6000.000 Current Values 6000.000 7000.000 Upper Limit 7000.000 No limit

does not change the optimal variable mix. B remains optimal but moves to a new location, x1 = 0.5, x2 = 3.5, Z = $27,500. 16. Z = 24000 Variable x1 x2 Constraint c1 c2 c2 Value 4.000 0.000 Slack/Surplus 12.000 0.000 11.000 Shadow Price 0.000 –3000.000 0.000

Allowable Increase 1000.000 No limit

Allowable Decrease 6000.000 1000.000

Right Hand Side Ranges Constraints c1 c2 c3 Lower Limit No limit 4.000 No limit Current Values 12.000 8.000 5.000 Upper Limit 24.000 No limit 16.000 Allowable Increase 12.000 No limit 11.000 Allowable Decrease No limit 4.000 No limit

(a) There is surplus high-grade and low-grade aluminum so the shadow price is $0 for both. The shadow price for medium-grade aluminum is $3,000 indicating that for every ton that this constraint could be reduced, cost will decrease by $3,000. (b) 0 ≤ c1 ≤ 7,000 6,000 ≤ c2 ≤ ∞ ∞ ≤ q1 ≤ 24 4 ≤ q2 ≤ ∞ ∞ ≤ q3 ≤ 16

17.

x1 = no. of acres of corn x2 = no. of acres of tobacco maximize Z = 300x1 + 520x2 subject to x1 + x2 ≤ 410 105x1 + 210x2 ≤ 52,500 x2 ≤ 100 x1, x2 ≥ 0

(c) There will be no change.

23

18.

x2
600 500 400 300 200 100

A: x1 = 0 x2 = 100 Z = 52,000 B: x1 = 300 x2 = 100 Z = 142,000

*C: x1 = 320 x2 = 90 Z = 142,800 D: x1 = 410 x2 = 0 Z = 123,000

The profit for corn must be greater than $520 for the Bradleys to plant only corn. (c) If the constraint line changes from x1 + x2 = 410 to x1 + x2 = 510, it will moveoutward to a location which changes the solution to the point where 105x1 + 210x2 = 52,500 intersects with the axis. This new point is x1 = 500, x2 = 0, Z = $150,000. (d) If the constraint line changes from x1 + x2 = 410 to x1 + x2 = 360, it moves inward to a location which changes the solution point to the intersection of x1 + x2 = 360 and 105x1 + 210x2 = 52,500. At this point x1 = 260, x2 = 100 and Z = $130,000. 19. Z = 142800.000 Variable x1 x2 Constraint c1 c2 c3 Value 320.000 90.000 Slack/Surplus 0.000 0.000 10.000 Shadow Price 80.000 2.095 0.000

A B
0 100 200 300

C

Point C is optimal
600 700 800

400D 500

x1

(a)

x1 = 320, x2 = 90 320 + 90 + s1 = 410 s1 = 0 acres uncultivated 90 + s3 = 100 s3 = 10 acres of tobacco allotment unused

(b) At point D only corn is planted. In order for point D to be optimal the slope of the objective function will have to be at least as great (i.e., steep) as the slope of the constraint line, x1 + x2 = 410, which is –1. Thus, the profit for corn is computed as, –c/520 = –1 c1 = 520

Objective Coefficient Ranges Variables x1 x2 Lower Limit 260.000 300.000 Current Values 300.000 520.000 Upper Limit 520.000 600.000 Allowable Increase 220.000 80.000 Allowable Decrease 40.000 220.000

Right Hand Side Ranges Constraints c1 c2 c3 Lower Limit 400.000 43050.000 90.000 Current Values 410.000 52500.000 100.000 Upper Limit 500.000 53550.000 No limit Allowable Increase 90.000 1050.000 No limit Allowable Decrease 10.000 9450.000 10.000

24

(a) No, the shadow price for land is $80 per acre indicating that profit will increase by no more than $80 for each additional acre obtained. The maximum price the Bradley’s should pay is $80 and the most they should obtain is at the upper limit of the sensitivity range for land. This limit is 500 acres, or 90 additional acres. Beyond 90 acres the shadow price would change. (b) The shadow price for the budget is $2.095. Thus, for every $1 dollar borrowed they could expect a profit increase of $2.095. If they borrowed $1,000 it would not change the amount of corn and tobacco they plant since the sensitivity range has a maximum allowable increase of $1,050. 20. x1 = no. of sausage biscuits x2 = no. of ham biscuits maximize Z = .60x1 + .50x2 subject to .10x1 ≤ 30 .15 x2 ≤ 30 .04x1 + .04x2 ≤ 16 0..01x1 + .024x2 ≤ 6 x1, x2 ≥ 0

(a)x1 = 300, x2 = 100, Z = $230 .10(300) + s1 = 30 s1 = 0 left over sausage .15(100) + s2 = 30 s2 = 15 lbs. left over ham .01(300) + .024(100) + s4 = 6 s4 = 0.6 hr. (b) The slope of the objective function, –6/5, must become flatter (i.e., less) than the slope of the constraint line, .04x1 + .04x2 = 16, for the solution to change. The profit for ham, c2, that would change the solution point is, –0.6/c2 = –1 c2 = .60 Thus, an increase in profit for ham of 0.60 will create a second optimal solution point at C where x1 = 257, x2 = 143 and Z = $225.70. (Point D would also continue to be optimal, i.e., multiple optimal solutions.) (c) A change in the constraint line from, .04x1 + .04x2 = 16 to .04x1 + .04x2 = 18would move the line outward, eliminating both points C and D. The new solution point occurs at the intersection of 0.01x1 + .024x2 = 6 and .10x = 30. This point is x1 = 300, x2 = 125, and Z = $242.50. 22.

21.
x2 600 500 400 300 200 100

Z = 230.000
A: x1 = 0 x2 = 200 Z = 100 B: x1 = 120 x2 = 200 Z = 172 C: x1 = 257 x2 = 143 Z = 225.70 *D: x1 = 300 x2 = 100 Z = 230 E: x1 = 300 x2 = 0 Z = 180 D E Point D is optimal
x1

Variable x1 x2 Constraint c1 c2 c3 c4

Value 300.000 100.000 Slack/Surplus 0.000 15.000 0.000 0.600 Shadow Price 1.000 0.000 12.500 0.000

A

B C

0

100

200

300

400

500

600

700

800

Objective Coefficient Ranges Variables x1 x2 Lower Limit 0.500 0.000 Current Values 0.600 0.500 Upper Limit No limit 0.600 Allowable Increase No limit 0.100 Allowable Decrease 0.100 0.500

25

Right Hand Side Ranges Constraints c1 c2 c3 c4 Lower Limit 25.714 15.000 12.000 5.400 Current Values 30.000 30.000 16.000 6.000 Upper Limit 40.000 No limit 17.000 No limit Allowable Increase 10.000 No limit 1.000 No limit Allowable Decrease 4.286 15.000 4.000 0.600

(a) The shadow price for sausage is $1. For every additional pound of sausage that can be obtained profit will increase by $1. The shadow price for flour is $12.50. For each additional pound of flour that can be obtained, profit will increase by this amount. There are extra ham and labor hours available, so their shadow prices are zero, indicating additional amounts of those resources would add nothing to profit. (b) The constraint for flour, indicated by the high shadow price. (c) .50 ≤ c1 ≤ ∞ 25.714 ≤ q1 ≤ 40 The sensitivity range for profit indicates that the optimal mix of sausage and ham biscuits will remain optimal as long as profit does not fall below $0.50. The sensitivity range for sausage indicates the optimal solution mix will be maintained as long as the available sausage is between 25.714 and 40 lbs. 23. x1 = no. of telephone interviewers x2 = no. of personal interviewers minimize Z = 50x1 + 70x2 subject to 80x1 + 40x2 ≥ 3,000 80x1 ≥ 1,000 40x2 ≥ 800 x1, x2 ≥ 0 24.
x2 80 70 60 50 40 30 20 10

(a) The optimal point is at B where x1 = 27.5 and x2 = 20. The slope of the objective function –50/70, must become greater (i.e., steeper) than the slope of the constraint line, 80x1 + 40x2 = 3,000, for the solution point to change from B to A. The cost of a telephone interviewer that would change the solution point is, –c1/70 = –2 c1 = 140 This is the upper limit of the sensitivity range for c1. The lower limit is 0 since as the slope of the objective function becomes flatter, the solution point will not change from B until the objective function is parallel with the constraint line. Thus, 0 ≤ c1 ≤ 140 Since the constraint line is vertical, it can increase as far as point B and decrease all the way to the x2 axis before the solution mix will change. At point B, 80(27.5) = q1 q1 = 2,200 At the axis, 80(0) = q1 q1 = 0 Summarizing, 0 ≤ q1 ≤ 2,200 (b) At the optimal point, B, x1 = 27.5 and x2 = 20.

A: x1 = 12.5 x2 = 50 Z = 4,125 *B: x1 = 27.5 x2 = 20 Z = 2,775

80(27.5) – s2 s2 40(20) – s3 s3

= 1,000 = 1,200 extra telephone interviews = 800 =0

A

(c) A change in the constraint line from 40x2 = 800 to 40x2 = 1,200, moves the lineup, but it does not change the optimal mix. The new solution values are x1 = 22.5, x2 = 30, Z = $3,225.

B Point B is optimal
x1

0

10

20

30

40

50

60

70

26

25. Z = 2775.000 Variable x1 x2 Constraint c1 c2 c3 Value 27.500 20.000 Slack/Surplus 0.000 1200.000 0.000 Shadow Price –0.625 0.000 –1.125

Objective Coefficient Ranges Variables x1 x2 Lower Limit 0.000 25.000 Current Values 50.000 70.000 Upper Limit 140.000 No limit Allowable Increase 90.000 No limit Allowable Decrease 50.000 45.000

Right Hand Side Ranges Constraints c1 c2 c3 Lower Limit 1800.000 No limit 0.000 Current Values 3000.000 1000.000 800.000 Upper Limit No limit 2200.000 2000.000 Allowable Increase No limit 1200.000 1200.000 Allowable Decrease 1200.000 No limit 800.000

(a) Reduce the personal interview requirement; it will reduce cost by $0.625 per interview, while a telephone interview will not reduce cost; i.e., it has a shadow price equal to $0. (b) 25 ≤ c2 ≤ ∞ 1,800 ≤ q1 ≥ ∞ x1 = no. of gallons of rye x2 = no. of gallons of bourbon maximize Z = 3x1 + 4x2 subject to x1 + x2 ≥ 400 x1 ≥ .4(x1 + x2) x2 ≤ 250 x1 = 2x2 x1 + x2 ≤ 500 x1, x2 ≥ 0

27.

x2 600 500 400 300

A: x1 = 266.7 x2 = 133.3 Z = 1,333.20 *B: x1 = 333.3 x2 = 166.7 Z = 1,666

26.

200

B
100

A Z

Feasible solution line Point B is optimal

0

100

200

300

400

500

600

700

x1

27

(a) Optimal solution at B: x1 = 333.3 and x2 = 166.7 (333.3) + (166.7) – s1 = 400 s1 = 100 extra gallons of blended whiskey produced .6(333.33) – .4(166.7) – s2 = 0 s2 = 133.3 extra gallons of rye in the blend (166.7) + s3 = 250 s3 = 83.3 fewer gallons of bourbon than the maximum (333.3) + (166.7) + s4 = 500 s4 = 100 gallons of blend production capacity left over (b) Because the “solution space” is not really an area, but a line instead, the objective function coefficients can change to any positive value and the solution point will remain the same, i.e., point B. Observing the graph of this model, no matter how flatter or steeper the objective function becomes, point B will remain optimal.

30.
x2 160 140 120 100 80

A: x1 = 0 x2 = 66,666.7 Z = $86,667 *B: x1 = 70,833.33 x2 = 24,166.67 Z = $116,416.67 A C: x1 = 95,000 x2 = 0 Z = $114,000 B

60 40 20

C
0 20 40 60 80 100 120 140 160 x1

28. Z = 1666.667 Variable x1 x2 Constraint c1 c2 c3 c5 Value 333.333 166.667 Slack/Surplus 100.000 133.333 83.333 0.000 Shadow Price 0.000 0.000 0.000 3.333

(a) The optimal solution point is at B where x1 = $70,833.33, and x2 = $24, 166.67. The slope of the objective function, –1.2/1.3, must become flatter than the slope of the constraint line, .18x1 + .30x2 = 20,000, for the solution point to change to A (i.e., only cattle). The return on cattle that will change the solution point is –1.2/c2 = –.18/30 c2 = 2 Thus, the return must be 100% before Alexis will invest only in cattle. (b) Yes, there is no slack money left over at the optimal solution. (c) Since her investment is $95,000, she could expect to earn $21,416.67.

29.

x1 = $ amount invested in land x2 = $ amount invested in cattle maximize Z = 1.20x1 + 1.30x2 subject to x1 + x2 ≤ 95,000 .18x1 + .30x2 ≤ 20,000 x1, x2 ≥ 0

28

Objective Coefficient Ranges Variables x1 x2 Lower Limit –2.000 –6.000 Current Values 3.000 4.000 Upper Limit No limit No limit Allowable Increase No limit No limit Allowable Decrease 5.000 10.000

Right Hand Side Ranges Constraints c1 c2 c3 c4 c5 Lower Limit No limit No limit 166.667 –250.000 400.000 Current Values 400.000 0.000 250.000 0.000 500.000 Upper Limit 500.000 133.333 No limit 500.000 750.000 Allowable Increase 100.000 133.333 No limit 500.000 250.000 Allowable Decrease No limit No limit 83.333 250.000 100.000

(a)–2.0 ≤ c1 ≤ ∞ –6.0 ≤ c2 ≤ ∞ Because there is only one effective solution point the objective function can take on any negative (downward) slope and the solution point will not change. Only “negative” coefficients that result in a positive slope will move the solution to point A, however, this would be unrealistic. (b)The shadow price for production capacity is $3.33. Thus, for each gallon increase in capacity profit will increase by $3.33. (c)This new specification changes the constraint, x1 – 2x2 = 0, to x1 – 3x2 = 0. This change to a constraint coefficient cannot be evaluated with normal sensitivity analysis. Instead the model Objective Coefficient Ranges Variables x1 x2 Lower Limit –1.333 –9.000 Current Values 3.000 4.000 Upper Limit No limit No limit

must be solved again on the computer, which results in the following solution output.

Z = 1625.000 Variable x1 x2 Constraint c1 c2 c3 c5 Value 375.000 125.000 Slack/Surplus 100.000 175.000 125.000 0.000 Shadow Price 0.000 0.000 0.000 3.250

Allowable Increase No limit No limit

Allowable Decrease 4.333 13.000

Right Hand Side Ranges Constraints c1 c2 c3 c4 c5 Lower Limit No limit No limit 125.000 –500.000 400.000 Current Values 400.000 0.000 250.000 0.000 500.000 Upper Limit 500.000 175.000 No limit 500.000 1000.000 29 Allowable Increase 100.000 175.000 No limit 500.000 500.000 Allowable Decrease No limit No limit 125.000 500.000 100.000

31. Z = 116416.667 Variable x1 x2 Constraint c1 c2 Value 70833.333 24166.667 Slack/Surplus 0.000 0.000 Shadow Price 1.050 0.833

Objective Coefficient Ranges Variables x1 x2 Lower Limit 0.780 1.200 Current Values 1.200 1.300 Upper Limit 1.300 2.000 Allowable Increase 0.100 0.700 Allowable Decrease 0.420 0.100

Right Hand Side Ranges Constraints c1 c2 Lower Limit 66666.667 17100.000 Current Values 95000.000 20000.000 Upper Limit 111111.111 28500.000 Allowable Increase 16111.111 8500.000 Allowable Decrease 28333.333 2900.000

(a) The shadow price for invested money is $1.05.Thus, for every dollar of her own money Alexis invested she could expect a return of $0.05 or 5%. The upper limit of the sensitivity range is $111,111.11, thus, Alexis could invest $16,111.11 of her own money before the shadow price would change. (b) This would change the constraint, .18x1 + .30x2 = 20,000 to .30x1 + .30x2 = 20,000. In order to assess the effect of this change the problem must

be solved again using the computer, as follows. Z = 86666.667 Variable x1 x2 Constraint c1 c2 Value 0.000 66666.667 Slack/Surplus 28333.33 0.000 Shadow Price 0.000 4.333

Objective Coefficient Ranges Variables x1 x2 Lower Limit No limit 1.200 Current Values 1.200 1.300 Upper Limit 1.300 No limit Allowable Increase 0.100 No limit Allowable Decrease No limit 0.100

Right Hand Side Ranges Constraints c1 c2 Lower Limit 66666.667 0.000 Current Values 95000.000 20000.000 Upper Limit No limit 28500.000 30 Allowable Increase No limit 8500.000 Allowable Decrease 28333.333 20000.000

32.

maximize Z = 140x1 + 205x2 + 190x3 + 0s1 + 0s2 + 0s3 + 0s4 subject to 10x1 + 15x2 + 8x3 +s1 = 610 x1 – 3x2 + s2 = 0 .6x1 – .4x2 – .4x3 – s3 = 0 x2 – x3 – s4 = 0 x1, x2, s1, s2, s3, s4 ≥ 0

Z = 9765.596 Variable x1 x2 x3 Constraint c1 c2 c3 c4 Value 22.385 16.789 16.789 Slack/Surplus 0.000 27.982 0.000 0.000 Reduced Cost 0.000 0.000 0.000 Shadow Price 16.009 0.000 –33.486 –48.532

Objective Coefficient Ranges Variables x1 x2 x3 Lower Limit –237.857 132.000 117.000 Current Values 140.000 205.000 190.000 Upper Limit 171.739 325.227 No limit Allowable Increase 31.739 120.227 No limit Allowable Decrease 377.857 73.000 73.000

Right Hand Side Ranges Constraints c1 c2 c3 c4 Lower Limit 0.000 –27.982 –21.217 –20.890 Current Values 610.000 0.000 0.000 0.000 Upper Limit No limit No limit 11.509 28.154 Allowable Increase No limit No limit 11.509 28.154 Allowable Decrease 610.000 27.982 21.217 20.890

33.

(a) and (b) minimize Z = $400x1 + 180x2 + 90x3 subject to x1 ≥ 200 x2 ≥ 300 x3 ≥ 100 4x3 – x1 – x2 ≤ 0 x1 + x2 + x3 = 1,000 x1, x2, x3 ≥ 0

(c) Z = 206000.000 Variable x1 x2 x3 Constraint c1 c2 c3 c4 31 Value 200.000 600.000 200.000 Slack/Surplus 0.000 500.000 100.000 0.000 Shadow Price –220.000 0.000 0.000 18.000

Objective Coefficient Ranges Variables x1 x2 x3 Lower Limit 180.000 90.000 No limit Current Values 400.000 180.000 90.000 Upper Limit No limit 400.000 180.000 Allowable Increase No limit 220.000 90.000 Allowable Decrease 220.000 90.000 No limit

Right Hand Side Ranges Constraints c1 c2 c3 c4 c5 Lower Limit 0.000 No limit No limit –500.000 500.000 Current Values 200.000 100.000 100.000 0.000 1000.000 Upper Limit 700.000 600.000 200.000 2500.000 No limit Allowable Increase 500.000 500.000 100.000 2500.000 No limit Allowable Decrease 200.000 No limit No limit 500.000 500.000

34.(a) x1 x2 x3 x4 Z

36.7142 58.6371 0 63.5675 9,177.85
x2 (1000s)

36.
18 16 14 12 Optimal 10 8 6 4 2 0 2 4 6 8 10 12 14 x1 (1000s) 16 18 20 22

B: x1 x2 Z

8,000 10,000 $11,500

(b) 34.6871 q1 61.5335 43.0808 q2 71.2315 q3 65.4686 55 q4 529.0816 c1 747.9999 350.3345 c2 3,488.554 c3 1,363.636 c4 1,761.476 20.132 c5 64.4643 (c) process 1 time is the most valuable with a dual value of $7.9275 (d) Product 3(x3) is not produced; it would require a profit of $65.4686 to be produced. 35. Maximize Z subject to: $0.17x1 $0.50x1 0.75x2

B A C

(a) c2 (b) s1

.50 $140

(c) There would be no feasible solution.

0.25x2 $4,000 (printing budget) x1 x2 18,000 (total copies, rack space) x1 8,000 (entertainment guide) x2 8,000 (real estate guide) x1, x2 0

32

37.

x1 x2 Z

8,000 10,000 $11,500

the original solution with a 140,000 ft2 store, thus, given these conditions, Mega-Mart should not purchase the land. 40.(a) Maximize Z subject to: x1 x3 x2 x1 x3 x2 $0.97x1 0.83x2 0.69x3

(a) The dual value of rack space is 0.75 so an increase in rack space to accommodate an additional 500 copies would result in increased advertising revenue of $375. An increase in rack space to 20,000 copies would be outside the sensitivity range for this constraint and require the problem to be solved again. The new solution is x1 8,000, x2 10,560 and Z $11,920 (b) 7,000 is within the sensitivity range for the entertainment guide (6,250 q3 10,000). The dual value is $0.25 thus for every unit the distribution requirement can be reduced, revenue will be increased by $0.25, or $250. Thus, Z $11,750 38. (a) Maximize Z 4.25x1 5.10x2 4.50x2 5.20x4 4.10x5 4.90x6 3.80x7 subject to: x1 x2 x3 140,000 xi 15,000, i x4 x5 x6 x7 x8

324 cartons

x1 + x 2 + x3 ≤ 324 cartons x3 ≥ x1 + x 2 x3 ≥3 x1 x 2 ≤ 120 (b) x1 = 54,x2 = 108,x3 = 162,Z 5 $253.80 41.(a) The shadow price for shelf space is $0.78 per carton, however, this is only valid up to 360 cartons, the upper limit of the sensitivity range for shelf space. (b)The shadow price for available local dairy cartons is $0 so it would not increase profit to increase the available amount of local dairy milk. (c) The discount would change the objective function to, maximize Z 0.86x1 0.83x2 0.69x3

1, 2, …,7

xi ≥ 15, 000, i = 1, 2, … , 7 xi ≤ .20, i = 1, 2, … , 7 Σxi x8 = .10 x 4 + x6 + x 7 xi ≥ 0 (b) x1 x2 x3 x4 x5 x6 x7 x8 Z 15,000 26,863 20,588.24 26,862.75 15,000 15,000 15,000 5,686 $625,083

and the constraint for relative demand would change to x3 ≥ 1.5 x1 the resulting optimal solution is, x1 = 108, x2 = 54, x3 = 162, Z $249.48

Since the profit declines the discount should not be implemented. 42. x1 = road racing bikes x2 = cross country bikes x3 = mountain bikes maximize Z subject to 600x1 400x2 900x3 120 300x3 $12,000

39.(a) A 20,000 ft2 increase in store size to 160,000 ft2 would increase annual profit to $718,316. This is a $93,233 increase in profit. Given the price of the land ($190,000) relative to the increase in profit, it would appear that the cost of the land would be offset in about 2 years, therefore the decision should be to purchase the land. (b) The decrease in profit in all departments would result in a new solution with Z $574,653. This is a reduction of $50,430 annually in profit from 33 43.

1,200x1 1,700x2 x1 x2 x3 20 8x1 12x2 16x3 x3 2(x1 x2) x1, x2, x3 0 x1 = 3 x3 = 6 Z 3,600

(a) More hours to assemble; the dual value for budget and space is zero, while the dual value for assembly is $30/hour. (b) The additional net sales would be $900. Since the cost of the labor is $300, the additional profit would be $600. (c) It would have no effect on the original solution. $700 profit for a cross country bike is within the sensitivity range for the objective function coefficient for x2. 44. Maximize Z subject to: $0.35x1 0.42x2 0.37x3

47.

x1 x2 x3 Z

20 33.3334 26.6667 $703,333.40

(a) The sensitivity range for x2 is 7,500 c2 8,774.999. Since $7,600 is within this range the values for x1, x2, and x3 would not change, but the profit would decline to $683,333.30 (i.e., less the difference in profit, ($600)(x2 33.3334) (b) One ton of grapes; the dual value is $23,333.35 (c) Grapes: (0.5)($23,333.35) $11,666.68 Casks: (4)($3,833.329) $15,333.32 Production: $0 Select the casks. (d) $6,799; slightly less than the lower band of the sensitivity range for cj. 48. Minimize Z $37x11 46x21 46x22 46x23 50x33 42x41 42x42 subject to: .7x11 .6x21 .5x31 .7x12 .6x22 .5x32 .7x13 .6x23 .5x33 x11 x12 x21 x22 x31 x32 x41 x42 37x12 37x13 50x31 50x32 42x43 .3x41 .3x42 .3x43 x13 x23 x33 x43 400 tons 250 tons 290 tons 350 tons 530 tons 610 tons 490 tons

0.45x1 0.41x2 0.50x3 x1 x2 x3 2,000 x1 200 x2 200 x3 200 x1 x2 x3 x1, x2, x3 0 45. x1 x2 x3 Z 1,000 800 200 $760

960

(a) Increase vending capacity by 100 sandwiches. There is already excess assembly time available (82 minutes) and the dual value is zero whereas the dual value of vending machine capacity is $0.38. $38 in additional profit. (b) x1 x2 Z 1,000 1,000 $770

The original profit is $760 and the new solution is $770. It would seem that a $10 difference would not be worth the possible loss of customer goodwill due to the loss of variety in the number of sandwiches available. (c) Profit would increase to $810 but the solution values would not change. If profit is increased to $0.45 the solution values change to x1 1,600, x2 200, x3 200. 46.(a) Maximize Z subject to: 7,500x1 8,200x2 10,500x3

49. x13 350 tons x21 158.333 tons x22 296.667 tons x23 75 tons x31 610 tons x42 240 tons Z $77,910 Mine 1 Mine 2 Mine 3 Mine 4 350 tons 530 tons 610 tons 240 tons

Multiple optimal solutions exist (a) Mine 4 has 240 tons of “slack” capacity. (b) The dual values for the 4 constraints representing the capacity at the 4 mines show that mine 1 has the highest dual value of $61, so its capacity is the best one to increase. (c) The sensitivity range for mine 1 is 242.8571 c1 414.2857, thus capacity could be increased 34

.21x1 .24x2 .18x3 x1 x2 x3 80 12x1 14.5x2 16x3 x3 (x1 x2)/2 x1, x2, x3 0

17 2,500

by 64.2857 tons before the optimal solution point would change. (d) The effect of simultaneous changes in objective function coefficients and constraint quality values cannot be analyzed using the sensitivity ranges provided by the computer output. It is necessary to make both changes in the model and solve it again. Doing so results in a new solution with Z $73,080, which is $4,830 less than the original solution, so Exeter should make these changes. 50. minimize Z = 8.2x1 + 7.0x2 +6.5x3 + 9.0x4 + 0s1 + 0s2 + 0s3 + 0s4 subject to 6x1 + 2x2 + 5x3 + 7x4 – s1 = 820 .7x1 – .3x2 – .3x3 – .3x4 – s2 = 0 –.2x1 + x2 + x3 – .2x4 + s3 = 0 x3 – x1 – x4 –- s4 = 0 ***** Input Data ***** Max. Z = 8.2x1 + 7.0x2+ 6.5x3 + 9.0x4 Subject to c1 c2 c3 c4 6x1 + 2x2 + 5x3 + 7x4 ≥ 820 .7x1 – .3x2 – .3x3 – .3x4 ≥ 0 –.2x1 + 1x2 + 1x3 – .2x4 ≤ 0 –1x1 + 1x3 – 1x4 ≥ 0
0 50

.16x1 + .20x2 + s3 = 40 32.8x1 + 20x2 + s4 = 6,000 x1, x2, s1, s2, s3, s4 ≥ 0 (c)
x2
400 350 300 250 200 clay molding

A: x1 = 0 x2 = 181.03 Z = $43,447.20 *B: x1 = 56.70 x2 = 154.64 Z = $47,886.60 C : x1 = 100 x2 = 120 Z = $47,800

D: x1 = 136.36 x2 = 76.36 Z = $44,234.80 E: x1 = 182.93 x2 = 0 Z = $34,756.70

A
150 100 50

B C
baking

D
100 150 E 200

glazing 250 300 350 400

x1

(d)x1 = 56.70, x2 = 154.64 .30(56.7) + .25(154.64) + s1 = 60 s1 = 4.33 hr. of molding time .27(56.7) + .58(154.64) + s2 = 105 s2 = 0 hr. of baking time .16(56.7) + .20(154.64) + s3 = 40 s3 = 0 hr. of glazing time 32.8(56.7) + 20(154.64) + s4 = 6,000 s4 = 1,047.42 lbs. of clay (e)The optimal solution is at point B. For point C to become optimal the profit for a large tile, x1, would have to become steeper, than the constraint line for glazing, .16x1 + .20x2 = 40: –c1/240 = .16/.20 c1 = 192 This is the upper limit for c1. The lower limit is at point A which requires an objective function slope flatter than the constraint line for baking, –c1/240 = .27/.58 c1 = 111.72 Thus, 111.72 ≤ c1 ≤ 192

***** Program Output ***** Infeasible Solution because Artificial variables remain in the final tableau.

CASE SOLUTION: MOSAIC TILE COMPANY
(a)maximize Z = $190x1 + 240x2 subject to .30x1 + .25x2 ≤ 60 hr.– molding .27x1 + .58x2 ≤ 105 hr. – baking .16x1 + .20x2 ≤ 40 hr.– glazing 32.8x1 + 20x2 ≤ 6,000 lb. – clay x1, x2, ≥ 0 (b)maximize Z = $190x1 + 240x2 + 0s1 + 0s2 + 0s3 + 0s4 subject to .30x1 + .25x2 + s1 = 60 .27x1 + .58x2 + s2 = 105 35

The same logic is used to compute the sensitivity range for c2. The lower limit is computed as, –190/c2 = –.16/.20 c2 = 237.5

The upper limit is, –190/c2 = .27/.58 c2 = 408.15 The sensitivity ranges for the constraint quantity values are determined by observing the graph and seeing where the new location of the constraint lines must be to change the solution point. For the molding constraint, the lower limit of the range for q1 is where the constraint line intersects with point B, .30(56.7) + .25(154.64) = q1 q1 = 55.67 The upper limit is ∞ since it can be seen that this constraint can increase indefinitely without changing the solution point. Thus, 55.67 ≤ q1 ≤ ∞ For the baking constraint the lower limit of the range for q2 is where point C becomes optimal, and the upper limit is where the baking constraint intersects with the x2 axis (x2 = 200). .27(100) + .58(120) = q2 q2 = 96.6 At x2 axis: .27(0) + .58(200) = q2 q2 = 116 Thus, 96.6 ≤ q2 ≤ 116 For the glazing constraint the lower limit of the range for q3 is at point A, and the upper limit is where the glazing constraint line, .16x1 + .20x2 = 40, intersects with the baking and molding constraints (i.e., x1 = 80.28 and x2 = 143.68). At A: .16(0) + .20(181.03) = q3 q3 = 36.21 At C: (g)

For the clay constraint the upper limit is ∞ since the constraint can increase indefinitely. The lower limit is at the point where the constraint line intersects with point B: At B: 32.8(56.7) + 20(154.64) = q4 q4 = 4,952.56 Thus, 4,952.56 ≤ q4 ≤ ∞ (f) The slope of the objective must be flatter than the slope of the constraint that intersects with the x2 axis at point A, which is the baking constraint, –190/c2 = .27/.58 c2 = $408.14

Problem Title: Case Problem: Mosaic Tile Company ***** Input Data ***** Max. Z = 190x1 + 240x2 Subject to c1 c2 c3 c4 .30x1 + .25x2 ≤ 60 .27x1 + .58x2 ≤ 105 .16x1 + .20x2 ≤ 40 32.8x1 + 20x2 ≤ 6000

***** Program Output ***** Final Optimal Solution At Simplex Tableau : 2 Z = 47886.598 Variable x1 x2 Constraint c1 c2 c3 Value 56.701 154.639 Slack/Surplus 4.330 0.000 0.000 1047.423 Shadow Price 0.000 10.309 1170.103 0.000

At intersection of constraints: .16(80.28) + .20(143.68) = q3 q3 = 41.58 Thus, 36.21 ≥ q3 ≥ 41.58

c4

36

Objective Coefficient Ranges Variables x1 x2 Lower Limit 111.724 237.500 Current Values 190.000 240.000 Upper Limit 192.000 408.148 Allowable Increase 2.000 168.148 Allowable Decrease 78.276 2.500

Right Hand Side Ranges Constraints c1 c2 c3 c4 Lower Limit 55.670 96.600 36.207 4952.577 Current Values 60.000 105.000 40.000 6000.000 Upper Limit No limit 116.000 41.577 No limit Allowable Increase No limit 11.000 1.577 No limit Allowable Decrease 4.330 8.400 3.793 1047.423

(h)Since there is already slack molding hours left over, reducing the time required to mold a batch of tiles will only create more slack molding time. Thus, the solution will not change. (i) Additional clay will have no effect on the solution since there is already slack clay left. Thus, Mosaic should not agree to the offer of extra clay. (j) Although an additional hour of glazing has the highest shadow price of $1,170.103, the upper limit of the sensitivity range for glazing hours is 41.577. Thus, with an increase of only 1.577 hours the solution will change and a new shadow price will exist. In order to assess the full impact of a 20 hour increase in glazing hours the problem should be solved again using the computer with this change. This new solution results in a profit of $49,732.39 an increase in profit of only $1,845.79. The reason for this small increase can be observed in the graphical solution; as the glazing constraint increases it quickly becomes a “non-binding” constraint with a new solution point. (k)A reduction of 3 hours is within the sensitivity range for kiln hours. However, the shadow price for kiln hours is $1,170.103 per hour. Thus, a loss of 3 kiln hours will reduce profit by (3)(1,170.103) = $3,510.31.

CASE SOLUTION: “THE POSSIBILITY” RESTAURANT –– CONTINUED
The solution is, x1 x2 Z 40 20 $800

(A)The question regarding a possible advertising expenditure of $350 per day requires that the sensitivity range for q1 be computed. q1: s3: 20 + 7 0 s4: 14 – 1.1 0 7 20 1.1 14 7 2.86 1.1 12.72 s1: 40 – 2 0 s2: 20 – 0 2 40 20 7 20 20 Summarizing, –20 and, 2.86 12.72 2.86 12.72 20

Since q1 = 60 + , = q1 – 60. Therefore, –26 ≤ q1 – 60 ≤ 12.72 57.14 ≤ q1 ≤ 72.72 Thus, an increase of 10 meals does not affect the shadow price for mean demand, which is $800. An increase of 10 meals will result in increased profit of ($8)($10) = $80, which exceeds the advertising expenditure of $30. The ad should be purchased.

37

(B)The reduction in kitchen staff from 20 to 15 hours requires the computation of the sensitivity range for q2. q2: s3: 20 – 10 0 s4: 14 + 4 0 – 20 20 4 14 70 1 4 –3.5 s1: 40 – 40 0 s2: 20 + 4 0 – 40 40 4 20 70 1 4 –5 Summarizing, –5 and, 3.5 1 3.5 1 10

CASE SOLUTION: JULIA’s FOOD BOOTH
(A)x1 = pizza slices, x2 = hot dogs, x3 = barbeque sandwiches The model is for the first home game, maximize Z subject to: $0.75x1 1.05x2 1.35x3

$0.75x1 .0.45x2 0.90x3 1,500 24x1 16x2 25x3 55,296 in2 of oven space. x1 x2 + x3 x2 ≥ 2.0 x3 x1, x2, x3 0

Since q2 = 20 + , = q2 – 20. Therefore, –3.5 ≤ q2 – 20 ≤ 1 16/5 ≤ q2 ≤ 21 A reduction of 5 hours to 15 hours would exceed the lower limit of the sensitivity range. This would result in a change in the solution mix and the shadow price, so the impact could not be totally ascertained from the optimal simplex tableau. solving the model again with q2 = 15 results in the following new solution. s1 s3 x1 x2 Z 5.45 81.82 49.09 5.45 $676.36

*Note that the oven space required for a pizza slice was determined by dividing the total space required by a pizza, 14 x 14 = 196 in2, by 8, or approximately 24 in2 per slice. The total space available is the dimension of a shelf, 36 in. x 48 in. = 1,728 in2, multiplied by 16 shelves, 27,648 in2, which is multiplied by 2, the times before kickoff and halftime the oven will be filled = 55,296 in2. Solution: x1 = 1,250 pizza slices x2 = 1,250 hot dogs x3 = 0 barbecue sandwiches Z = $2,250 Julia should receive a profit of $2,250 for the first game. Her lease is $1,000 per game so that leaves her with $1,250. Her cost of leasing a warming oven is $100 per game, thus she will make a little more than what she needs to, i.e., $1,000, for it to be worth her while to lease the booth. A “tricky” aspect of the model formulation is the $1,500 used to purchase the ingredients. Since the objective function reflects net profit, the $1,500 is recouped and can be used for the next home game to purchase food ingredients; thus, it’s not necessary for Julia to use any of her $1,150 profit to buy ingredients for the next game. (B) Yes, she would increase her profit; the dual value is $1.50 for each additional dollar. The upper limit of the sensitivity range for budget is $1,658.88, so she should only borrow approximately $158. Her additional profit would be $238.32 or a total profit of $2,488.32. 38

Notice that simply using the shadow price of $16 for staff time (hr) would have indicated a loss in profit of only (5hr)(16) = $80, or Z = $720. The actual reduction in profit to $676.36 is greater. The final question concerns an increase in the coefficient for c1 from $12 to $14. This requires the computation of the range for c1. (C)The final question concerns an increase in the coefficient for c1 from $12 to $14. This requires the computation of the range for c1. c1, basic: –8 –2 –8 –2 –8 –2 –4 0 8 –4 4 – 16 + 4 – 16 + 4 – 16 + 4 0 16 4

Since c1 = 12 + , = c1 – 12. Therefore, –4 c1 – 12 4 –8 c1 16 Since c1 = $14 is within this range the price increase could be implemented without affecting Pierre’s meal plans.

(C) Yes, she should hire her friend. It appears impossible for her to prepare all of the food items given in the solution in such a short period of time. The additional profit she would get if she borrowed more money as indicated in part B would offset this additional expenditure. (D) The biggest uncertainty is the weather. If the weather is very hot or cold, fans might eat less. Also, if it is rainy weather for a game or games, the crowd might not be as large, even though the games are all sellouts. The model results show that Julia will reach her goal of $1,000 per game - if everything goes right. She has little slack in her profit margin, thus it seems unlikely that she will achieve $1,000 for each game.

39

Sign up to vote on this title
UsefulNot useful

Master Your Semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master Your Semester with a Special Offer from Scribd & The New York Times

Cancel anytime.