Designation: E 1155 – 96 (Reapproved 2008)

Standard Test Method for

Determining FF Floor Flatness and FL Floor Levelness
Numbers1
This standard is issued under the fixed designation E 1155; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (e) indicates an editorial change since the last revision or reapproval.
This standard has been approved for use by agencies of the Department of Defense.

1. Scope 3.1.2 flat—even, plane, homoloidal, free of undulation.

1.1 This test method covers a quantitative method of mea- 3.1.2.1 Discussion—For the purposes of this test method,
suring floor surface profiles to obtain estimates of the floor’s flatness will be measured by calculating curvature value, q,
characteristic FF Flatness and FL Levelness Face Floor Profile between all 12-in. reading points separated by 24 in. The
Numbers (F-Numbers) using the inch-pound system of units. curvature value is the difference between successive elevation
differences. The mean and standard deviation of all the
NOTE 1—A complete metric companion to Test Method E 1155 has curvature values for a given test section are then converted
been developed, Test Method E 1155M; therefore, no metric equivalents
according to the equations in this test method to get the
are shown in this test method.
dimensionless FF Flatness Number.
1.2 The text of this test method references notes and 3.1.3 floor profilometer—a Type I device (see 6.1.1) that
footnotes that provide explanatory material. These notes and produces a continuous record of the elevation of a single point
footnotes (excluding those in tables and figures) shall not be moving along a line on the floor’s surface.
considered as requirements of this test method. 3.1.4 horizontal—level, normal to the direction of gravity.
1.3 This standard does not purport to address all of the 3.1.5 inclinometer—a Type II device (see 6.1.2) that mea-
safety concerns, if any, associated with its use. It is the sures the angle between horizontal and the line joining the two
responsibility of the user of this standard to establish appro- points of contact with the floor’s surface.
priate safety and health practices and determine the applica- 3.1.6 level—horizontal, normal to the direction of gravity.
bility of regulatory limitations prior to use. 3.1.6.1 Discussion—For the purposes of this test method,
levelness will be measured by collecting elevation differences
2. Referenced Documents
at points spaced 10 ft apart and that will be described by the FL
2.1 ASTM Standards: 2 Levelness number (dimensionless).
E 1155M Test Method for Determining FF Floor Flatness 3.1.7 longitudinal differential floor profilometer, n—a Type
and FL Floor Levelness Numbers [Metric] II device (see 6.1.2) that produces a continuous record of the
2.2 ACI Standard: elevation difference between two points moving along a line on
ACI 117-90 Standard Specifications for Tolerances for Con- the floor’s surface, which two points remain separated by a
crete Construction and Materials3 fixed distance.
3. Terminology 3.1.8 sample measurement line—a sample measurement
line shall consist of any straight line on the test surface along
3.1 Definitions of Terms Specific to This Standard: which measurements are taken, with the limitations listed in
3.1.1 elevation—height, altitude, vertical location in space. 7.3.
Elevation measurements are always made parallel to the 3.1.9 sign convention—where up is the positive direction;
direction of gravity. down is the negative direction. Consequently, the higher the
reading point, the more positive its hi value, and the lower the
1
This test method is under the jurisdiction of ASTM Committee E06 on
reading point, the more negative its hi value. Similarly, the
Performance of Buildings and is the direct responsibility of Subcommittee E06.21 elevation difference from a low point to a high point (that is, an
on Serviceability. uphill difference) is positive, while the elevation difference
Current edition approved Jan. 15, 2008. Published January 2008. Originally from a high point to a low point (that is, a downhill difference)
approved in 1987. Last previous edition approved in 2001 as E 1155 – 96 (2001).
2
For referenced ASTM standards, visit the ASTM website, www.astm.org, or is negative.
contact ASTM Customer Service at [email protected]. For Annual Book of ASTM 3.1.10 test section—a test section consists of any subdivi-
Standards volume information, refer to the standard’s Document Summary page on sion of the test surface with the limitations listed in 7.2.
the ASTM website.
3
Available from American Concrete Institute (ACI), P.O. Box 9094, Farmington
Hills, MI 48333-9094, http://www.aci-int.org.

Copyright. © ASTM International. 100 Barr Harbour Dr. P.O. box C-700 West Conshohocken, Pennsylvania 19428-2959, United States

1
Copyright by ASTM Int'l (all rights reserved); Thu Oct 22 14:47:39 EDT 2009
Downloaded/printed by
Greg Gidez (Hensel Phelps Construction Co.) pursuant to License Agreement. No further reproductions authorized.
 E 1155 – 96 (2008)
3.1.11 test surface—on any one building level, the entire NOTE 2—When the traffic patterns across a floor are random, (as is
floor area of interest constitutes the test surface, with the generally the case) evaluation of the floor’s FF Flatness and FL Levelness
limitations listed in 7.1. will necessarily involve a random sampling of the surface, since all of the
infinite potential profiles to be seen by the traffic can not possibly be
3.1.12 vertical—parallel to the direction of gravity. measured. In those instances when the traffic across a floor will be
3.2 Symbols: confined to specific paths, however, the requirement for random sampling
3.2.1 Ai—area of Test Section i. is eliminated, since the floor can indeed be inspected exactly as it will be
3.2.2 di—difference in elevation (in inches) between read- seen by all of the traffic. In these special cases, rather than inferring the
ing points Pi and Pi−1 (i $ 1). condition of the traffic paths from a random sample, it is far more useful
3.2.3 Ff—Face FF Flatness Number (dimensionless). to measure each of the traffic paths directly using continuous recording
floor profilometer configured to run exactly in the traffic wheel paths. Such
3.2.4 Ffi—composite FF Flatness Number for Test Section i.
direct simulation measurements eliminate the inherent uncertainties of
3.2.5 Fl—Face FL Levelness Number (dimensionless). statistical sampling and provide profile information immediately appli-
3.2.6 Fli—composite FL Levelness Number for Test Section cable to the correction of the surface in way of the future traffic.
i.
3.2.7 hi—elevation (in inches) of Reading Point Pi (i $ 0). 6. Apparatus
3.2.8 nj—number of reading points in Test Sample j (nj $ 6.1 Point Elevation Measurement Device:
12). 6.1.1 Type I Apparatus—If a Type II apparatus (see 6.1.2) is
3.2.9 Nmin—minimum number of 10-ft elevation difference not used for this test, then an apparatus capable of measuring
readings required per the test section. the elevations of a series of points spaced at regular 12-in.
3.2.10 qi—arithmetic difference (in inches) between eleva- intervals along a straight line on the floor surface shall be used.
tion differences di and di−1 (i $ 2). Examples of satisfactory Type I point elevation measurement
3.2.11 rxj—number of readings of Variable xobtained from devices include, but are not limited to the following:
Sample j. 6.1.1.1 Leveled Straightedge, with gage (for example, tri-
3.2.12 sxj—standard deviation of Variable x in Sample j. square, dial indicator, etc.) to measure vertical distance from
3.2.13 Vxj—variance of Variable x in Sample j. the upper straightedge surface to floor.
3.2.14 zi—difference in elevation (in inches) between Read- 6.1.1.2 Leveled Straightedge, with graduated wedges or
ing Points Pi and Pi−10 (i $ 10). shims to measure vertical distance from lower straightedge
4. Summary of Test Method surface to floor.
6.1.1.3 Optical Level, with vernier or scaled target.
4.1 Straight lines are marked at various locations on the
6.1.1.4 Laser Level, with vernier or scaled target.
floor surface. Point elevations are then measured at regular
6.1.1.5 Taut Level Wire, with gage to measure vertical
12-in. intervals along each line. The elevation differences
distance from wire to floor.
between all adjacent reading points are calculated, and a
straight line approximation to the surface profile along each 6.1.1.6 Floor Profilometer.
measurement line is produced and evaluated for consistency 6.1.2 Type II Apparatus—If a Type I apparatus (see 6.1.1) is
with visual observation of the floor surface. not used for this test, then an apparatus capable of measuring
4.2 The arithmetic differences between all adjacent 12-in. the elevations of a series of points spaced at regular 12-in.
elevation differences and the elevation differences between all intervals along a straight line on the floor surface shall be used.
points separated 10 ft are then calculated. Estimates of each Examples of satisfactory Type II point elevation measurement
test section’s floors FF Flatness and FL Levelness F-Numbers devices include, but are not limited to the following:
are obtained through statistical analyses of these calculated 6.1.2.1 Inclinometer, having 12-in. contact point spacing.
profile values. Finally, the F-Numbers for each test section are 6.1.2.2 Longitudinal Differential Floor Profilometer, having
combined to arrive at a composite set of F-Numbers for each 12-in. sensor wheel spacing.
test surface. 6.2 Ancillary Equipment:
6.2.1 Measurement Tape, graduated in feet.
5. Significance and Use 6.2.2 Chalk Line (or other means for marking straight lines
5.1 This test method provides statistical (and graphical) on the test surface).
information concerning floor surface profiles. 6.2.3 Data Recording Means—This procedure requires the
5.2 Results of this test method are used primarily to: recording of both verbal and numeric information. Examples of
5.2.1 Establish compliance of randomly trafficked floor satisfactory data recording means include, but are not limited
surfaces with specified FF Flatness and FL Levelness toler- to the following:
ances, 6.2.3.1 Manual Data Sheet.
5.2.2 Evaluate the effect of different construction methods 6.2.3.2 Magnetic Tape Recorder, (voice or direct input).
on resulting floor surface flatness and levelness, and 6.2.3.3 Paper Chart Recorder.
5.2.3 Investigate the curling and deflection of floor surfaces. 6.2.3.4 Direct Computer Input.
5.3 Results of this test method shall not be used to enforce
NOTE 3—Since the bias of the results obtained with this test method
contract flatness and levelness tolerances on those floor instal- will vary directly with the accuracy of the particular measurement device
lations primarily intended to support the operation of fixed- employed, all project participants should agree on the exact test apparatus
path vehicle systems (for example, narrow aisle warehouse to be used prior to the application of this test method for contract
floors). specification enforcement.

2
Copyright by ASTM Int'l (all rights reserved); Thu Oct 22 14:47:39 EDT 2009
Downloaded/printed by
Greg Gidez (Hensel Phelps Construction Co.) pursuant to License Agreement. No further reproductions authorized.
 E 1155 – 96 (2008)
7. Organization of Test Area NOTE 4—Since construction joints are a discontinuity in the floor
surface, measuring across them would introduce statistical anomalies into
7.1 Test Surface—On any one building level, the entire floor this test method. Construction joints are therefore excluded from the
area of interest shall constitute the test surface. generation of F-Number statistics. However, since traffic will nevertheless
7.1.1 When this test method is used to establish compliance pass across many of the construction joints, a separate measurement and
of randomly trafficked floor surfaces with specified FF Flatness analysis of the joints may be required in order to provide a quantitative
and FL Levelness tolerances, each portion of the surface which measure of the roughness of the joints themselves. Some joints may never
has a unique specified set of tolerances must be treated as a see traffic, for example, those along a wall. The particular joints required
to be analyzed may be specified in contract specifications, along with a
separate surface.
maximum allowable value for qi.
7.2 Test Section—A test section shall consist of any subdi-
vision of a test surface satisfying the following criteria: 8. Procedure
7.2.1 No test section shall measure less than 8 ft on a side,
nor comprise an area less than 320 ft2. 8.1 Record the name and location of the subject building;
7.2.2 No portion of the test surface shall be associated with the installation date of the subject floor; the subject floor’s
more than one test section. specified Ff and Fl values; the make, model, and serial number
7.2.3 When testing a concrete floor, no test section boundary of the test apparatus to be used; the date of the test; and the
shall cross any construction joint. name of the individual making the test.
7.3 Sample Measurement Line—A sample measurement NOTE 5—When this test is used to evaluate the compliance of a new
line shall consist of any straight line on the test surface concrete floor with contract flatness and levelness specifications, the
satisfying the following criteria: timeliness of the test vis-a-vis the date of the floor’s installation is of
7.3.1 No sample measurement line shall measure less than critical importance. Since most concrete floors will change shape signifi-
cantly within a few days after installation, owing to inevitable shrinkage
11 ft in length.
and deflection, the American Concrete Institute (see ACI 117-90) now
7.3.2 No portion of any sample measurement line shall fall requires that specified concrete floor tolerances be checked within 72 h
within 2 ft of any slab boundary, construction joint, isolation after floor installation in order to ensure that an accurate gage of the
joint, block-out, penetration, or other similar discontinuity. surface’s “as-built” shape is assessed.
7.3.2.1 Exception—Shrinkage crack control joints formed
8.2 Lay out the test surface.
either by partial depth sawcuts or by partial depth inserts shall
be ignored. 8.2.1 Divide the entire test surface into test sections. Assign
7.3.2.2 Exception—If the area to be excluded from mea- a different identification number to each test section, and
surement exceeds 25 % of the test section area, then the 2-ft record the locations of all test section boundaries.
boundary exclusion shall not apply. 8.2.2 Within the restrictions described in 7.3, 7.6, and 8.2.3,
7.3.3 Measurement lines may not be placed parallel to each determine the number and location of all sample measurement
other closer than 4 ft. lines to be used in each test section. Assign a different
7.4 Type I Test Sample (Measured With Type I identification number to each sample measurement line, and
Apparatus)—A Type I test sample shall consist of not less than record the locations of all sample measurement line starting
twelve sequential point elevation measurements made at regu- and stopping points. Mark or otherwise physically delineate
lar 12-in. intervals along a single sample measurement line. each sample measurement line on the test surface.
7.5 Type II Test Sample (Measured With Type II 8.2.3 The sample measurement lines within each test sec-
Apparatus)—A Type II test sample shall consist of not less tion shall be arranged so as to blind the test results (to the
than eleven sequential measurements of the elevation differ- extent possible) to any surface profile anisotropies resulting
ences between adjacent reading points spaced at regular 12-in. from the floor’s method of construction. Accomplish this by
intervals along a single sample measurement line. distributing the sample measurement lines uniformly across the
7.6 Minimum Number of zi Readings Per Test Section—The entire test section and either:
number (or length) of Type I or Type II test samples to be 8.2.3.1 Orienting all lines at 45° to the longest construction
collected within each test section shall be sufficient to yield (in joint abutting the test section, (not corner-to-corner diagonals)
aggregate) not less than Nmin individual measurements of zi, (see Fig. 1), or
where Nmin is calculated as follows: 8.2.3.2 Placing equal numbers of lines of equal aggregate
Nmin 5 2=A ~320 # A # 1600! (1)
length both parallel to and perpendicular to the longest test
section boundary. See Fig. 1.
5 A/30 ~A . 1600!
8.2.3.3 When the short dimension (width) of the slab being
where: measured is less than 25 ft, all measurement lines must be 45°
A = test section area, ft2. diagonals in accordance with 8.2.3.1.
7.7 Construction Joints—Where construction joints are re- 8.3 Collect Type I or Type II test samples, or both, from
quired to be measured, periodic measurements of the 24-in. each test section sufficient (in aggregate) to satisfy the mini-
curvature qi shall be taken, transverse to and centered on the mum zi reading requirement prescribed in 7.6. No upper limit
construction joint. At least one qi measurement shall be taken is placed upon the number of test samples that may be collected
on each straight section of joint, with a maximum interval from a single test section. All data collected on all survey lines
between measurement locations not to exceed 10 ft. These measured in a given test section shall be incorporated into the
measurement locations shall be recorded. calculations of F-Numbers. Data shall only be excluded when

3
Copyright by ASTM Int'l (all rights reserved); Thu Oct 22 14:47:39 EDT 2009
Downloaded/printed by
Greg Gidez (Hensel Phelps Construction Co.) pursuant to License Agreement. No further reproductions authorized.
 E 1155 – 96 (2008)

FIG. 1 Location of Sample Measurement Lines on Test Section

it can be demonstrated that the test apparatus reported inaccu- 9.3 Calculate the difference in elevation between all adja-
rate values or that the test procedure of this test method was not cent reading points:
followed. In the event that data is excluded, the entire survey 9.3.1 If analyzing a Type I test sample, calculate the
line shall be considered unusable; no single measurement of di, elevation differences, di, between all adjacent reading points as
qi, or zi may be excluded. follows:
8.3.1 Subdivide each sample measurement line into 12-in.
di 5 hi 2 hi21 ~in.! (3)
long intervals. The points marking the ends of these 12-in.
intervals are the sample reading points. Designate the starting where:
point of each sample as P0 and then sequentially number each i $ 1.
successive reading point down the sample measurement line as Each Type I Test Sample j will therefore result in (nj − 1)
P1, P2, P3, etc. calculated di values. Whenever Point Pi is higher than Point
8.3.2 For each test sample, measure and record in sequence: Pi−1, the value for di will be positive. Conversely, whenever
8.3.2.1 If a Type I apparatus is used, the elevations (in Point Pi is lower than Point Pi−1, the value for di will be
inches) of all sample reading points, or negative.
8.3.2.2 If a Type II apparatus is used, the differences in 9.3.2 If analyzing a Type II test sample, designate all di
elevation (in inches) between all adjacent sample reading values in accordance with 9.1.2.1.
points.
9.4 For each Test Sample j, calculate the profile curvatures,
9. Calculation qi, between all reading points separated by 24 in. as follows:
9.1 Calculate the elevations of all reading points: qi 5 di 2 di21 ~in.! 5 hi 2 2 3 hi21 1 hi22 ~in.! (4)
9.1.1 If analyzing a Type I test sample, designate the where: i $ 2.
elevation measurements collected at Reading Points P0, P1, P2, Each test sample will result in (nj − 2) calculated qi values.
. . . Pi, etc. as h0, h1, h2, . . . hi, etc. A positive qi value will denote a trough, while a negative qi
9.1.2 If analyzing a Type II test sample: value will denote a crest.
9.1.2.1 Designate the elevation difference measurements
9.5 For each Test Sample j, calculate the elevation differ-
collected between Reading Points P0 and P1, and P2 and P3,
ences, zi, between all reading points separated by 10 ft as
and . . . Pi−1 and Pi, etc. as d1, d2, d3, . . . di, etc.
follows:
9.1.2.2 Let h0 = 0.
9.1.2.3 Calculate the elevations, hi, of all reading points as zi 5 hi 2 hi210 ~in.! (5)
follows: where:
hi 5 hi21 1 di ~in.! (2) i $ 10.
where: Each test sample will result in (nj − 10) calculated z values.
i $ 1. A positive zi value will denote an uphill change in elevation
9.1.2.4 Each Type II test sample will therefore result in nj from Pi−10 to Pi, while a negative zi value will denote a
calculated hi values. downhill change in elevation from Pi−10 to Pi.
9.2 Produce a straight line graph between each of the nj 9.6 For each Test Sample j, calculate the mean, qij, of all (nj
calculated hi values. This is a straight line approximation of the − 2) qi values.
floor surface profile. Evaluate each straight line profile ap- 9.6.1 Add all (nj − 2) qi values in Sample j as follows:
proximation subjectively to confirm that it appears to represent nj21
the actual floor surface profile. This serves as a subjective ( qi 5 q2 1 q3 1 q4 1 . . . 1 qn 21 ~in.!
j
(6)
i52
quality control check to ensure that no gross anomalies are
present in the data before reporting the results of this test 9.6.2 Divide this sum by (nj − 2) to obtain the mean value of
method. the qi values in Sample j as follows:

4
Copyright by ASTM Int'l (all rights reserved); Thu Oct 22 14:47:39 EDT 2009
Downloaded/printed by
Greg Gidez (Hensel Phelps Construction Co.) pursuant to License Agreement. No further reproductions authorized.
 E 1155 – 96 (2008)
nj21 4.57
( qi Ffj 5 ~dimensionless! (16)
i52 ~3·Sqi 1 |q̄ij|!
q̄ij 5 n 2 2 ~in.! (7)
j
where:
9.7 For each Test Sample j, calculate the standard deviation Sqi = standard deviation of the qi values in Sample j (from
Sqj, of all (nj − 2) qi values. 9.7.3), and
9.7.1 Add the squares of all (nj − 1) qivalues as follows: q̄ij = absolute value of the mean of the qi values in Sample
nj21 j (from 9.6.2).
( qi2 5 q22 1 q32 1 q42 1 . . . 1 q2n 21 ~in.2!
i52 j
(8) 9.11 Calculate the composite FF Flatness Number estimate
for each test section by combining all of the FF Flatness
9.7.2 Multiply the sum of all (nj − 2) qi values obtained in Number estimates obtained from the individual test samples
9.5.1 by the mean value of qij, obtained in 9.5.2, subtract this within that test section.
product from the sum of the squares of all (nj − 2) qi values 9.11.1 The following equation is used to combine the FF
obtained in 9.6.1, and divide this difference by (nj − 3) to Flatness or FL Levelness F-Number estimates derived from
obtain the variance Vqj, of the qi values in Sample j as follows: two different test samples into a single composite F-Number
nj21 nj21 estimate:
( qi 2 2 q̄i i ( qi

Œ
j
i52 52
Vqj 5 ~in. 2! (9) rj 1 r k
nj 2 3 Fj 1 k 5 Fj·Fk (17)
rk·Fj2 1 rj·Fk2

9.7.3 Take the square root of the variance, Vqj, of the qi where:
values in Sample j to obtain the standard deviation, Sqj, of the Fj+k = F-Number estimate derived by combining Samples
qi values in Sample j as follows: j and k,
Fj = F-Number estimate derived from Sample j,
Sqj 5 =Vqj ~in.! (10) Fk = F-Number estimate derived from Sample k,
9.8 For each Test Sample j, calculate the mean, z̄ij, of all (nj rj = number of qi or zi readings in Sample j used to
− 10) zi values. derive Fj, and
rk = number of qi or zi readings in Sample k used to
9.8.1 Add all (nj − 10) zi values in Sample j as follows:
derive Fk.
nj21
9.11.2 Using the equation given in 9.11.1, calculate the
( zi 5 z10 1 z11 1 z12 1 . . . 1 zn 21 ~in.! j
(11)
i 5 10 composite FF Flatness Number estimate for each test section
9.8.2 Divide this sum by (nj − 10) to obtain the mean value by combining (iteratively) all of the flatness F-Number esti-
of z̄ij of the zi values in Sample j as follows: mates obtained from the individual test samples within that test
nj21
section. The number of readings to be associated with each
( zi successive FF Flatness Number estimate, Fj+k, will be the sum
i 5 10
z̄ij 5 n 2 10 ~in.! (12) (rj + rk).
j
NOTE 6—Since FF Flatness Numbers may be combined only with other
FF Flatness Numbers, and FL Levelness Numbers may only be combined
9.9 For each Test Sample j, calculate the standard deviation,
with other FL Levelness Numbers, the complete description of any floor
Szj, of all (nj − 10) zi values. surface requires the identification of two separate and distinct values: Ff
9.9.1 Add the squares of all (nj − 10) zi values as follows: Flatness Number and Fl Levelness Number.
nj21
zi2 5 z102 1 z112 1 z122 1 ... 1 z2 nj21 ~in.2!
9.11.3 Sample Problem Illustrating F-Number Combination
(
i 5 10
(13)
Procedure—Three test samples containing 40, 60, and 80
9.9.2 Multiply the sum of all (nj − 10) zivalues obtained in readings respectively are collected from a certain test section.
9.8.1 by the mean value, zi, obtained in 9.8.2, subtract this Analysis in accordance with 9.10 yields the following indi-
product from the sum of the squares of all (nj − 10) zi values vidual sample Ff estimates:
obtained in 9.9.1, and divide this difference by (nj − 11) to Sample One:
obtain the variance, Vzj, of the zi values in Sample j as follows: Ff1 = 20, and
rq1 = number of qi readings in Sample One = 40.
ni21 nj21
( zi 2 2 z̄i i 5(10 zi
i 5 10 j
Sample Two:
Vzj 5 nj 2 11 ~in.2! (14) Ff2 = 30, and
rq2 = number of qi readings in Sample Two = 60.

9.9.3 Take the square root of the variance, Vzj, of the zi Sample Three:
values in Sample j to obtain the standard deviation, Szj, of the Ff3 = 40, and
rq3 = number of qi readings in Sample Three = 80.
zi values in Sample j as follows:
9.11.3.1 To combine these three individual sample results
Szj 5 =Vzj ~in.! (15)
into a single composite Ff estimate for the entire test section,
9.10 Estimate the FF Flatness Number, Ffj, for each Test use the equation given in 9.11.1 by first combining the Ff
Sample j as follows: estimates given for any two of the test samples, and then

5
Copyright by ASTM Int'l (all rights reserved); Thu Oct 22 14:47:39 EDT 2009
Downloaded/printed by
Greg Gidez (Hensel Phelps Construction Co.) pursuant to License Agreement. No further reproductions authorized.
 E 1155 – 96 (2008)
combining this interim Ff result with the Ff estimate from the where:
remaining sample as follows: r = total number of qi or zi readings used to calculate the

Πrq1 1 rq2 F-Number:

FF1 1 2 5 FF1·FF2 (18) F2Number 90 % Confidence Interval
rq2·FF12 1 rq1·FF22

FF1 1 2 5 20·30 Œ 40 1 60
60·202 1 40·302
5 ~100 2 CI90 %!·F/100 to ~100 1 CI90 %!·F/100
9.15 Within each test surface, combine each test section’s
(24)

5 24.5 F-Numbers by area-weighting to obtain overall F-Numbers as

follows:
where:
Ff1+2 = FF Flatness Number estimate derived by combin- ( Ai·Ffi
Overall Ff 5 ( A (25)
ing Samples One and Two. i

9.11.3.2 Calculate the combined sample size, rq1+2, by add- ( Ai·Fli

ing the number of readings contained in each sample as Overall Fl 5 ( A (26)
i
follows:
rq1 1 2 5 rq1 1 rq2 5 40 1 60 5 100 (19) 10. Report
10.1 In tabular form, list the calculation results obtained in
9.11.3.3 Now combine this interim result with the
9.1, 9.3, 9.4, and 9.10-9.14 inclusive for each test section.
F-Number estimate derived from the remaining sample to
Whenever reporting Ff or Fl estimates for test sections, always
obtain the desired composite F-Number estimate as follows:
show the associated 90 % confidence interval in parentheses
FF ~112!13 5 FF112·FF3 Œ rq112 1 rq3
rq3·FF1122 1 rq112·FF32
(20)
immediately following the F-Number estimate.
10.2 An example illustrating correct F-Number reporting
format is as follows:
FF ~112!13 5 24.5·40 Œ 100 1 80
30·24.52 1 100·402
10.2.1 On a particular test section, the estimated Ff value of
24.5 has a 90 % confidence interval of 23.0 to 26.0. Whenever
5 28.8 this F-Number estimate appears in the report, it is followed
where: immediately by its associated 90 % confidence interval en-
Ff(1+2)+3 = FF Flatness Number estimate derived by com- closed in parentheses:
bining Samples One, Two, and Three. Ff 24.5 ~23.0 2 26.0! (27)
9.11.3.4 The combined sample size, rq(1+2)+3, is calculated as
follows: 10.3 When required by contract specifications, the maxi-
rq ~112!13 5 rq112 1 rq3 5 100 1 80 (21) mum qi value on each straight section of construction joint
shall be reported.
5 180
10.4 List the calculated overall F-Number results for the
Therefore, based upon 180 readings of qi, the subject test entire test surface. This is the only F-Number that is reported
section has an estimated Flatness F-Number of Ff 28.8. without a confidence interval.
9.12 Estimate the FL Levelness Number for each Test
NOTE 7—A plot of profile elevation as a function of horizontal distance
Sample j as follows: may also be reported from the point elevation readings obtained in 9.1,
12.5 along any or all sample measurement lines. These plots may be useful in
FLi 5 ~dimensionless! (22) visualizing the FF Flatness and FL Levelness of the survey lines.
~3·Szi 1 | z̄ij|!

where: 11. Factors Influencing Precision and Bias of This Test

Szj = standard deviation of zi in Sample j (from 9.8.3), Method
and 11.1 Since every test section contains an infinity of qi and zi
| z̄ij| = absolute value of the mean of the zi values in values, all of which cannot be measured, estimation of Ff or
Sample j (from 9.8.2). Flfrom a finite sample of qi or zi values involves the math-
9.13 Using the equation in 9.11.1, calculate the composite ematics of statistical inference. The means and standard
FL Levelness Number estimate for each test section by com- deviations of the qi and zi values contained in each test sample
bining (iteratively) all of the FL Levelness Number estimates are calculated and used as estimates for the means and standard
obtained from the individual test samples within that test deviations of the entire infinite qi and zi populations. The
section. The number of readings to be associated with each equations given in 9.9 and 9.11 translate these estimated
successive FL Levelness Number estimate, Fj+k, will be the population statistics into their corresponding F-Numbers.
sum (rj + rk). 11.2 On level floors, as the sample size increases, the mean
9.14 Calculate the 90 % confidence interval, CI90 %, associ- values, qi, and zi, approach zero and can be ignored. Ff and Fl
ated with each FF Flatness and FL Levelness Number estimate are then reduced to functions only of the standard deviations Sq
as follows: and Sz:
CI90 % 5 21.82 ~log10r!3 1 19.4 ~log10r!2 2 71.69 ~log10r! 1 92.62 % 1.523
FF 5 S (28)
(23) q

6
Copyright by ASTM Int'l (all rights reserved); Thu Oct 22 14:47:39 EDT 2009
Downloaded/printed by
Greg Gidez (Hensel Phelps Construction Co.) pursuant to License Agreement. No further reproductions authorized.
 E 1155 – 96 (2008)
4.167 11.7 The results obtained with this test method will also
Fl 5 S (29)
z vary with the precision and bias of the particular elevation
11.3 Since 99.73 % of all values in a normal distribution fall measurement apparatus employed. Instrument inaccuracies
within 63 standard deviations of the population mean, will always result in a lowering of the reported F-Numbers.
99.73 % of all qireadings will have an absolute value less than
12. Precision and Bias
or equal to the following:
12.1 Precision—The repeatability standard deviation for
4.57
qi99.73 % 5 3~Sq! 5 F ~in.! (30) both Ff and Fl is less than 0.25.
F
12.1.1 These values were obtained by conducting repeated
while 99.73 % of the zi readings will have an absolute value measurements of the same test section. For these within-
less than or equal to the following: laboratory tests, the same layout pattern was used, with the
12.5 same operators and apparatus. When operators and equipment
zi99.73 % 5 3~Sz! 5 F ~in.! (31) models (of the same fundamental type of measurement appa-
L

11.3.1 These 3(S) values have been arbitrarily defined as the ratus) are varied, but the layout pattern is retained, the
dimensional limits that characterize a given F-Number. They repeatability standard deviation goes up to about 0.3. The small
provide a simple means for giving the F-Numbers dimensional standard deviation obtained indicates that neither the operator
significance, that are greater than these 3(S) values. It is nor the equipment model (within this particular type apparatus)
incorrect, however, to think of F-Numbers as directly limiting has a very significant influence on repeatability. The results do
the magnitudes of qi or zi. F-Numbers do not prohibit the not vary widely with changes in Ff and Fl. Future versions of
incidence of any specific qi or zi value; they only limit the this test method will be published with precision and bias as
percentage of all qi or zi readings that can have that value. generated by using other measurement apparatus, when suffi-
cient data can be made available.
11.4 Since F-Numbers are derived directly from the qi and
12.1.2 The reproducibility standard deviation for Ff is as
zi statistics, the accuracy of the composite and estimates
follows:
obtained in 9.10 and 9.12 will depend upon the following:
11.4.1 The degree to which the sample qi and zi distributions 10 |Cz|Cz 20 |Cz|Cz 30 |Cz|Cz 40 |Cz|Cz 50 |Cz|Cz 60 |Cz|Cz 70 |Cz|Cz 80
| 0.25 | 0.74 | 1.22 | 1.70 | 2.18 | 2.66 | 3.14 |
match the actual qi and zi distribution, and
11.4.2 The total number of sample qi and zi readings used to 12.1.2.1 These values were also obtained by having various
estimate the means and standard deviations of the entire, testing agencies across the United States conduct repeated
infinite qi and zi populations. measurements of the same test section, in this case, always
11.5 As the size of a sample increases, so does the prob- with different layout patterns. The larger resulting standard
ability that the sample’s statistics will accurately represent deviations obtained indicate that variation in line placement is
those of the entire population. Confidence in the ability of a the dominant factor in obtaining differing values of Ff. In
certain size sample to estimate the statistics of an entire general, precision of results is best at the lower Ff values, and
population is expressed in terms of confidence intervals. The decreases with increasing Ff-Number. It should be noted that
90 % confidence interval calculated in 9.13 is a direct measure the data used to generate this precision and bias statement was
of the degree of statistical uncertainty that will be associated generated through the use of many different instruments, all of
with each F-Number estimate obtained with this test method. which were one of three different models of one Type II
This 90 % confidence interval may be interpreted as follows: instrument. Although several other devices have been used at
11.5.1 Given a sample size of r readings, nine times out of various times to collect data for this test method, no compara-
ten the actual F-Number of the floor will fall between plus or tive data is available from any of these devices. Future versions
minus CI90 % percent of the F-Number estimate obtained with of this test method will be published with precision and bias as
this test method. generated by using other measurement apparatus, when suffi-
cient data can be made available.
11.6 The F-Number estimate obtained in 9.9 or 9.11 is the
12.1.3 The reproducibility standard deviation for Fl is as
midpoint of a range of possible F-Numbers characterizing the
follows:
test section with some known probability. The width of each
such F-Number range will vary directly with the confidence 10 |Cz|Cz 20 |Cz|Cz 30 |Cz|Cz 40 |Cz|Cz 50 |Cz|Cz 60 |Cz|Cz 70 |Cz|Cz 80
level demanded and inversely with the number of sample | 0.70 | 1.75 | 2.00 | 2.75| 2.0 | 2.9 | 4.5 |
readings used to compute the midpoint estimate. The probabil- 12.1.3.1 One way to interpret these results is to state the
ity that a particular range will actually contain the true expected or probable error, that is equal to 0.6746 times the
F-Number of the floor increases as the width of the range standard deviation. For example, when measuring a floor
increases. Likewise, the greater the number of sample readings whose known true Ff value is 15, 50 % of the time it is
used to estimate the range midpoint, the greater the probability expected that the test result be greater than 15.23 or less than
that the true F-Number of the floor will be near that midpoint. 14.77, and 50 % of the time, one would expect the deviation to
be within this range. Another way to interpret the precision
NOTE 8—Since it is impossible to be certain of the exact FF Flatness or
FL Levelness Number of any floor (no matter how many readings are
results is to look at the 95 % reproducibility limits, that are 2.8
taken, some statistical uncertainty will always exist), report the results of times the standard deviation. In this case, 95 % of the time, the
this test method as a range of possible F-Numbers at the prescribed 90 % measured result will fall between 14.05 and 15.95. This test
confidence level. method contains an estimate of the 90 % confidence limits, that

7
Copyright by ASTM Int'l (all rights reserved); Thu Oct 22 14:47:39 EDT 2009
Downloaded/printed by
Greg Gidez (Hensel Phelps Construction Co.) pursuant to License Agreement. No further reproductions authorized.
 E 1155 – 96 (2008)
are required to be provided in any report with the test result 12.2 Bias—The procedure in this test method has no bias
(see section 9.13 and Section 10). As the F-Number goes up because the values of Ff and Fl are defined in terms of this test
(either Ff or Fl) the uncertainty regarding the true value also method.
rises, because much smaller errors can create the same amount
of difference in the test result. For example, when measuring a 13. Keywords
floor whose known true Ff value is 35, 50 % of the time it is
expected that the test result be greater than 36.1 or less than 13.1 Ff; FF Flatness; Fl; FL Levelness; floor; floor profile;
33.9, and 50 % of the time, it is expected that the deviation be floor profilometer; F-Number
within this range, and 95 % of the time, the measured result
will fall between 30.5 and 39.5.

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or [email protected] (e-mail); or through the ASTM website
(www.astm.org).

8
Copyright by ASTM Int'l (all rights reserved); Thu Oct 22 14:47:39 EDT 2009
Downloaded/printed by
Greg Gidez (Hensel Phelps Construction Co.) pursuant to License Agreement. No further reproductions authorized.