ISOM3730 Quality and Process Management

Question Set 2: Statistical Process Control

*************************************************************************************************

Question 1
Webster Chemical Company produces mastics and caulking for the construction industry. The product is blended in large
mixers and then pumped into tubes and capped. Webster is concerned whether the filling process for tubes of caulking is
in statistical control. Several samples of eight tubes are taken and each tube is weighed in ounces.

Tube Number
Sample 1 2 3 4 5 6 7 8 Avg Range
1 7.98 8.34 8.02 7.94 8.44 7.68 7.81 8.11
2 8.23 8.12 7.98 8.41 8.31 8.18 7.99 8.06
3 7.89 7.77 7.91 8.04 8.00 7.89 7.93 8.09
4 8.24 8.18 7.83 8.05 7.90 8.16 7.97 8.07
5 7.87 8.13 7.92 7.99 8.10 7.81 8.14 7.88
Avgs

Assuming that taking 5 samples is sufficient, is the process in statistical control?

Conclusion on process variability? UCLR = 0.84 LCLR = 0.06

Conclusion on process average? UCLx = 8.2 LCLx = 7.86

(Question 2-3)
Four samples of three observations each have been Observation #
taken from the assembly process and you are asked Sample 1 2 3
to construct both the X and R charts: #1 12.3 12.2 12.1
#2 11.9 12.2 12.2
#3 12.0 12.2 11.8
#4 12.1 11.8 11.8
Question 2
The upper control limit of the X chart is:
A. 11.831 B. 12.356 C. 12.050 D. 12.269 E. 11.744

Question 3
The lower control limit of the R chart is:
A. 0 B. 0.3 C. 0.4 D. 0.684 E. 0.771

Question 4
The length of industrial filters is a quality characteristic of interest. Thirty samples, each of size 5, are chosen from the
process. The data yields an average length of 110 mm, with the process standard deviation estimated to be 4mm.
a) Find the warning limits for a three-sigma control chart for the average length.

If the process mean shifts to 112 mm,

b) What is the chance of detecting the shift at the third sample point drawn after the shift?

If the process mean shifts to 115 mm,

c) Calculate the minimum number of samples needed to be drawn such that the chance of detecting this shift is at
least 80%.

1
Question 5
The following are means of samples of 5. The data pertain to the depth to shoulder of fragmentation bomb heads
manufactured during the war by the American Stove Company. The measurements are in inches.

Group number x
1 0.4402
2 0.4390
3 0.4448
4 0.4432
5 0.4428
6 0.4382
7 0.4358
8 0.4440
9 0.4366
10 0.4368

The above samples were taken every 15 minutes. The production rate was 400 per hour. Historical data show that, when
the process is under control, the mean and standard deviation of individual depth-to-shoulder measurements are 0.4490
and 0.0144 respectively. A set of control limits with  2  x is used to monitor the production process. Assume that
depth-to-shoulder measurements are normally distributed. It is known that, if a shift in process mean occurs, the
corresponding Type II error (within 1 test) is 0.524.

a) What is the standard deviation of x ? Hence, what are the control limits? Based on the control limits, determine
from the 10 samples if the process appears to be out of control.
b) What is the magnitude of Type I error with the above control limits?
c) What is the probability that if a shift in process mean occurs, it would be detected at the n th sample after the
shift? How many samples would you have to take to have a chance of approximately 0.85 of catching the shift
within these samples after the shift?

Question 6
The diameter of cotton pins produced by an automatic machine is a characteristic of interest. Based on historical data, the
process average diameter is 15mm with a process standard deviation of 0.8mm. If samples of size 4 are randomly
selected from the process:
a) Find the 3σ control limits for the average diameter.
b) What is the probability of a false alarm?
c) If the process mean shifts to 14.5mm, what is the probability of not detecting this shift on the first sample
plotted after the shift?
d) How many samples, on average, would it take to detect this shift?

Question 7
Before cement is delivered to a civil engineering site, a number of small bricks are made from it. Four are chosen at
random and measured for the compressive strength (measured in N m−2 × 109). This is known to be normally distributed
with standard deviation 6. The mean compressive strength of the four bricks is expected to be 51.
a) Find the 3σ control limits for the average compressive strength.
b) What is the probability of failing to detect the shift if the batch with the mean strength is indeed 47?

2
Question 8
Random samples of size 5 of the length of a connector pin are selected. For each sample, the sample mean and range are
calculated (in millimeters), as shown in the table below.
a) Find the trial 3-sigma control limits for the mean and range chart.
b) Assuming special causes for the out-of-control points, find the revised control limits.

Data for Length of Connector Pin (in mm)

Sample Average Length Range Sample Average Length Range Sample Average Length Range
1 50.3 4 11 52.5 5 21 52.1 2
2 48.4 2 12 47.8 3 22 49.5 3
3 48.5 5 13 49.6 5 23 50.7 4
4 49.1 4 14 50.8 9 24 52.6 5
5 52.6 3 15 48.5 4 25 52.3 6
6 46.2 4 16 49.3 6 26 48.6 5
7 50.8 3 17 53.3 3 27 51.0 4
8 52.2 4 18 52.1 4 28 52.3 3
9 49.5 5 19 50.2 5 29 50.6 2
10 51.7 4 20 51.9 4 30 51.5 4

Question 9
Discuss the interpretation of each of the following control charts.