TM 5-826-6/AFR 93-5
CHAPTER 4
SAMPLING TECHNIQUES
4-1. General. Inspection of an entire feature may require considerable effort, especially if the feature is very large. This
may be particularly true for flexible pavements containing much distress. Because of the time and effort involved, frequent
surveys of the entire feature may be beyond available manpower, funds, and time.
4-2. Inspection. All sample units should be surveyed if a pavement feature contains only five or fewer sample units. A
statistical sampling plan was developed to obtain an adequate estimate of the PCI determined by inspecting only a portion
of the sample units within normal size features (i.e., 20 + 5 slabs for rigid pavements, and 5,000 + 1,000 sq ft for flexible).
Use of the statistical sampling plan described herein will considerably reduce the time required to inspect a feature without
significant loss of accuracy. However, this sampling plan is optional, and inspection of the entire feature may be desirable
in some cases.
a. Determination of pavement feature. The first step in the condition survey is the designation of pavement features.
Each facility such as a runway, taxiway, etc., is divided into segments or features that are definable in terms of the same
design, the same construction history, the same traffic area, and generally the same overall condition. General features
can be determined from pavement design and construction records and can be further subdivided as deemed necessary
based on a preliminary survey. It is important that all pavement in a given feature be such that it can be considered
uniform. For example, the center part of some runways in the traffic lanes should be separate features from the portion
outside the traffic lanes.
b. Selection of sample units to be inspected. Assuming a normal distribution of data, the number of sample units to
be surveyed to provide a 95-percent confidence level can be determined from
2
N
(eq 4-1) n =
2
e
2
4 (N - 1) +
where
n = number of sample units to be inspected
N = total number of sample units in feature
= one standard deviation in PCI between sample units within the feature
e = allowable error in determining the true PCI
Data collected during the development of the PCI procedure gave average values of 10 and 15 for flexible and rigid
pavements, respectively. These values should be used for determining the number of sample units to be inspected unless
field data and experience indicate the higher or lower values would be more appropriate for military airfield pavements.
The number of samples n obtained from the equation is the minimum number to be inspected in a given pavement feature
in order to have 95 percent confidence that the PCI is within 5 points. The above equation is presented graphically in
figure 4-1 to simplify its use in application.
c. The minimum number of sample units that must be surveyed to obtain an adequate estimate of the PCI of a
feature is selected from figure 4-1. Once the number of sample units η has been determined from figure 4-1 the spacing
interval of the units is computed from
(eq 4-2) i = N
η
where
i = spacing interval of units to be sampled
N = total number of sample units in the feature
η = number of sample units to be inspected
4-1
Previous Page
