Composite Structures 58 (2002) 497–503

www.elsevier.com/locate/compstruct

Three-parameter vs. two-parameter Weibull distribution

for pultruded composite material properties
Maha Alqam a, Richard M. Bennett a,*
, Abdul-Hamid Zureick b

a
Department of Civil and Environmental Engineering, The University of Tennessee, Knoxville, TN 37996-2010, USA
b
School of Civil Engineering, Georgia Institute of Technology, Atlanta, GA 30332-0355, USA

Abstract
The three-parameter and two-parameter Weibull distributions are compared using 26 mechanical property data sets of fiber-
reinforced polymeric (FRP) composite materials manufactured by the pultrusion process. Both strength and stiffness properties were
examined. The probability distributions were compared on the basis of goodness of fit, nominal design values, and allowable load to
achieve uniform reliability. It is recommended that the two-parameter Weibull distribution be used to characterize FRP composite
material properties. The primary basis for this recommendation is small differences in nominal design values and small differences in
allowable loads between the two-parameter and three-parameter Weibull distributions. Other supporting reasons for the recom-
mendation are similar observed significance levels in distribution fitting, computational efficiency, and the fact that the location
parameter of the three-parameter Weibull distribution is near the first order statistic.
Ó 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Composites; Probability; Statistical analysis; Weibull distribution; Reliabililty; Maximum likelihood; Method of moments

1. Introduction [10] to examine the compressive strength property data

of flat pultruded panels, the three-parameter Weibull
The determination of the statistics and probability distribution was used by Abdallah et al. [1] to assess the
distributions of the random variables describing mate- reliability of glass-fiber reinforced polymeric pultruded
rial properties plays an important role in the develop- rods and by Zureick and Steffen [18] to compute the 95%
ment of probabilistic based design specifications. The lower confidence limit on the 5th percentile of the com-
choice of the probability distribution chosen to repre- pressive property data of E-glass polyeter and vinylester
sent the material property data will have a large effect on pultruded single angles.
the calculated reliability. Assuming different distribu- This paper examines the two-parameter vs. the three-
tions for the material properties can result in computed parameter Weibull distribution for characterizing FRP
probabilities of failure that vary by over an order of composite material properties. Only the Weibull distri-
magnitude. This is the result of the lower tail behavior of butions are considered in this paper, although other
different cumulative distribution functions, which has probability distributions may also fit the data. The
become known as the tail-sensitivity problem in struc- purpose of the paper is to determine the adequacy of the
tural reliability [3]. two-parameter Weibull distribution or if there is justi-
The Weibull distribution [14] is often used to describe fication for using the three-parameter Weibull distribu-
the strength of fiber-reinforced polymer (FRP) compos- tion for FRP material properties.
ites [8,12]. Typically, the two-parameter Weibull distri- The three-parameter Weibull cumulative distribution
bution is used, although the three-parameter Weibull function is given by:
distribution is more robust and may provide a better "  b #
characterization of the data. While the two-parameter xd
FX ðxÞ ¼ 1  exp  xPd ð1Þ
Weibull distribution was previously used by Mottram h

*
Corresponding author. Tel.: +1-865-974-7540; fax: +1-865-974- in which h is the scale parameter, b is the shape param-
2669. eter, and d is the location parameter. If d ¼ 0, the distri-
E-mail address: [email protected] (R.M. Bennett). bution becomes the two-parameter Weibull distribution.
0263-8223/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 2 6 3 - 8 2 2 3 ( 0 2 ) 0 0 1 5 8 - 7
498 M. Alqam et al. / Composite Structures 58 (2002) 497–503

Zanakis [15] documents seventeen different methods E-glass/polyester pultruded box-shaped components in
for obtaining the parameters of the three-parameter another case [16]. Coupons from the former and latter
Weibull distribution. Two common parameter estima- study are labeled VG and PG, respectively. Coupon
tion methods used in engineering are the modified mo- data from a study [10] of E-glass/polyester plates is also
ment method and the maximum likelihood method, with included in this study, and labeled as PG-M. Both
those being the methods considered in this paper. The stiffness and strength parameters are examined, with a
two-parameter and three-parameter Weibull distribu- total of 26 data sets and over seven hundred data points
tions are compared on the basis of the goodness-of-fit of being considered. A summary of the data sets is given in
the distribution to FRP material property data and Table 1, which gives the sample size and coefficient of
lower tail behavior. The lower tail behavior is examined variation of each data set.
through the analysis of nominal design values and
probabilistic based allowable loads.
Data used for the comparisons are taken from two
2. Parameter estimation
different studies that examined the short-term axial
compressive strength of E-glass/vinylester pultruded
Two methods will be used for parameter estimation,
I- and box-shaped components in one case [17], and
the modified moment method and the maximum likeli-
the short-term eccentric axial compressive strength of
hood method. Moment methods are based on equating

Table 1
Description of data sets and location parameter for three-parameter Weibull distribution of d^=x1 and d^=x
Material property Sample size (2) Coefficient of Location parameter, d^
(1) variation (3) Modified moment method Maximum likelihood method
d^=x1 (4) d^=x (5) d^=x1 (6) d^=x (7)
Longitudinal tensile strength
VG 1–6 30 0.074 0.984 0.883 0.996 0.894
VG 7–12 30 0.130 0.949 0.760 0.925 0.741
VG 13–18 24 0.102 0.957 0.811 0.983 0.833
VG 19–24 24 0.077 0.913 0.786 0.979 0.843
PG 30 0.069 0.800 0.680 0.878 0.746
Longitudinal compressive strength
VG 1–6 30 0.073 0.985 0.886 0.994 0.894
VG 7–12 30 0.093 0.756 0.605 0.932 0.747
VG 13–18 24 0.114 0.969 0.819 0.995 0.841
VG 19–24 24 0.106 0.948 0.793 0.940 0.786
PG 30 0.122 0.935 0.750 0.938 0.752
PG-M 51 0.092 0.560 0.427 0.819 0.624

Shear strength
VG 19–24 24 0.072 0a 0a 0a 0a
PG 18 0.098 0.950 0.816 0.988 0.848
Transverse compressive strength
PG-M 52 0.060 0.974 0.871 0.982 0.877
Longitudinal tensile modulus
VG 1–6 30 0.068 0.949 0.836 0.938 0.826
VG 7–12 30 0.096 0a 0a 0a 0a
VG 13–18 24 0.116 0.917 0.742 0.962 0.778
VG 19–24 24 0.046 0.974 0.902 0.981 0.908
PG 30 0.063 0.947 0.839 0.947 0.838
Longitudinal compressive modulus
VG 1–6 30 0.054 0.940 0.843 0.973 0.873
VG 7–12 30 0.087 0.852 0.703 0a 0a
VG 13–18 24 0.132 0.940 0.753 0.984 0.788
VG 19–24 23 0.040 0.749 0.682 0a 0a
PG 30 0.040 0.941 0.864 0.962 0.884
Shear modulus
VG 19–24 24 0.113 0.841 0.666 0a 0a
PG 18 0.104 0.895 0.742 0.970 0.804
a
Indicates location parameter was set equal to zero.
 M. Alqam et al. / Composite Structures 58 (2002) 497–503 499

sample moments to the corresponding distribution mo- mum likelihood method. This method requires solving
ments. The mean value or the first moment, l, of the the following simultaneous Eqs. [4].
Weibull distribution is: ! !1
  X n
^ X
n
^
1 ðxi  d^Þ lnðxi  d^Þ
b
ðxi  d^Þ b

l ¼ d þ hC 1 þ ð2Þ i¼1 i¼1

b !
1 X
n
where C is the gamma function. The variance or second   lnðxi  d^Þ=n ¼ 0 ð8Þ
moment about the mean, r2 , is: b^ i¼1
    
2 1
2 2
r ¼h C 1þ 2
C 1þ ð3Þ Pn !1
b b ^b b^
i¼1 ðxi  dÞ
h^ ¼ ð9Þ
By equating the mean l to the sample mean x and the n
variance r2 to the sample variance s2 , Eqs. (2) and (3)
b^ Xn
b^1
Xn
1
can be used to estimate the parameters of the Weibull
^
ðxi  d^Þ  ðb^  1Þ ðxi  d^Þ ¼ 0 ð10Þ
distribution. An estimate of the location parameter, d, is ^b
h i¼1 i¼1
also required, with a common estimate for d being x1 , Eq. (10) can be difficult to solve. This is apparent from
the first order statistic (the first data point when order- the graph shown in Fig. 1, in which the left hand side
ing from smallest to largest). A better estimate of the of Eq. (10) is plotted for different values of d^ for
location parameter is [15]: the longitudinal tensile strength of data set VG19-24.
x1 xn  x22 The values of h^ and b^ are held constant and equal to the
d^ ¼ ð4Þ maximum likelihood estimates (h^ ¼ 58:7 MPa and
x1 þ xn  2x2
b^ ¼ 2:18) while d^ is varied. The change in curvature near
in which d^ is the estimate of the location parameter, d, the root may cause many standard root-finding tech-
and xi is the ith order statistic of the sample of size n. niques, such as Newton–Raphson, to fail. Even if the
Improvements have been made to the basic moment method does not fail, the flat slope for most of the graph
method, resulting in the modified moment method [4]. coupled with the steep slope near the root often results
Parameters of the Weibull distribution are estimated in slow convergence. The best technique is generally to
from the following equations: start with a value just slightly less than the minimum
   
2 C 1 þ 2
 C 2
1 þ 1 data point, reduce this value until there is a change in
s ^
b ^
b
¼ h ð5Þ sign, and then use some standard root finding technique
ðx  x1 Þ2
 i2
^
1  n1=b C 1 þ b^ 1 over the interval in which the sign changes.
In five out of the 26 data sets examined, the maximum
s
h^ ¼ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
    ð6Þ likelihood estimator of the location parameter was less
C 1 þ b2^  C2 1 þ b1^ than zero. Often in these cases, the estimators were
unreasonable. For example, the maximum likelihood
  estimate of the location parameter for the longitudinal
1
d^ ¼ x  h^C 1 þ ð7Þ tensile modulus for specimens VG7-12 was approxi-
b^
mately negative eight times the mean value. The shape
in which a hat indicates the estimate of the parameter. parameter was 139.5, while the typical range of the
Eqs. (5)–(7) will be used in this paper as one estimate of
the parameters of the three-parameter Weibull distri-
bution.
In two cases out of the 26 data sets examined, the
modified moment method resulted in a location pa-
rameter less than zero. In these two cases, the location
parameter was set to zero as a negative location pa-
rameter would imply the possibility of negative values
for the material properties. For cases in which there was
a positive estimate of the location parameter (d^), the
average values of d^=x1 and d^=x are 0.90 and 0.77, re-
spectively, where x1 and x are the minimum (first order
statistic) and sample mean of the data set. Values of the
location parameter in terms of the first order statistic
and the mean value are given in Table 1.
The second method of estimating the parameters of
the Weibull distribution that will be used is the maxi- Fig. 1. Value of Eq. (10) vs. location parameter, d.
500 M. Alqam et al. / Composite Structures 58 (2002) 497–503

shape parameter is 6–25. In these five cases, the location The modified moment method and the maximum
parameter was taken as zero, or the two-parameter likelihood method were used to determine parameters of
Weibull distribution was used. the three-parameter Weibull distribution for each of the
The maximum likelihood estimator of the location 26 data sets being considered in this paper. The maxi-
parameter is generally just slightly less than the mini- mum likelihood method was also used to determine
mum data point. For the 19 cases with a positive loca- parameters for the two-parameter Weibull distribution
tion parameter, the average values of d^=x1 and d^=x are by setting d ¼ 0 and only using Eqs. (8) and (9). Al-
0.96 and 0.82, respectively. The location parameter from though the parameters for the two-parameter Weibull
the maximum likelihood method is given in Table 1. distribution could be determined using moment meth-
ods, the maximum likelihood method is the generally
used method for the two-parameter Weibull distribution
3. Goodness of fit [12]. An OSL was obtained for each of the distributions
for each data set. The results are shown in Table 2.
Many methods, such as Chi-square, the Kolmogo-
rov–Smirnov, exist for determining the goodness of fit of
a probability distribution to a set of data. The Ander-
son–Darling test was chosen for this study as it is more Table 2
OSL using the Anderson–Darling test for different Weibull distribu-
sensitive to the tail behavior [9], and has been recom- tions
mended for statistical analysis of composites [12]. The
Material OSL
sensitivity to the tail behavior is particularly useful in
property (1) Two-parameter Three-parameter Weibull
structural engineering applications, where the tail is
Weibull (2)
important in computing the structural reliability. The Modified Maximum
Anderson–Darling statistic is obtained as: moment likelihood
method (3) method (4)
Z 1 e 2
½ F n ðxÞ  F0 ðxÞ Longitudinal tensile strength
A2n ¼ n dF0 ðxÞ ð11Þ VG 1–6 0.009 0.823 0.817
1 F0 ðxÞ½1  F0 ðxÞ
VG 7–12 0.302 0.045 0.072
in which Fen ðxÞ is a step function that jumps at the order VG 13–18 0.138 0.224 0.146
statistics of x, and F0 ðxÞ is the hypothesized continuous VG 19–24 0.008 0.208 0.420
PG 0.699 0.871 0.860
cumulative distribution function. The Anderson–Dar-
ling statistic is a measure of the square of the error be- Longitudinal compressive strength
tween the data and the hypothesized distribution VG 1–6 0.001 0.536 0.320
VG 7–12 0.212 0.475 0.727
weighted so that the tails of the data are more important VG 13–18 0.088 0.183 0.117
than the central portion. For computation purposes, the VG 19–24 0.469 0.272 0.303
Anderson–Darling statistic can be obtained as: PG 0.214 0.131 0.096
n   PG-M 0.122 0.255 0.157
X 1  2i   
A2n ¼ ln½F0 ðxðiÞ Þ þ ln 1  F0 ðxðnþ1iÞ Þ n Shear strength
i¼1
n VG 19–24 0.517 0.610a 0.517a
ð12Þ PG 0.073 0.544 0.486
Transverse compressive strength
in which xðiÞ is the ith order statistic of the data set. For PG-M 0.020 0.546 0.424
the Weibull distribution, an observed significance level
Longitudinal tensile modulus
(OSL) is obtained as follows [12]:
VG 1–6 0.179 0.505 0.522
1 VG 7–12 0.022 0.030a 0.022a
OSL ¼ ð13Þ VG 13–18 0.010 0.023 0.012
1 þ exp½0:10 þ 1:24lnAD þ 4:48AD  VG 19–24 0.177 0.622 0.508
PG 0.180 0.504 0.414
in which
  Longitudinal compressive modulus
0:2 2 VG 1–6 0.153 0.518 0.483
AD ¼ 1 þ pffiffiffi An ð14Þ
n VG 7–12 0.052 0.012 0.052a
VG 13–18 0.010 0.042 0.036
The OSL is the probability of obtaining a value of the VG 19–24 0.299 0.386 0.299a
test statistic at least as large as that obtained from the PG 0.504 0.753 0.678
data if the hypothesis that the data are actually from Shear modulus
the distribution being tested is true. Typically, a 5% VG 19–24 0.065 0.312 0.065a
significance level is used, so that the null hypothesis is PG 0.116 0.598 0.701
a
only rejected if the OSL is less than 0.05. Indicates a location parameter of 0.0.
 M. Alqam et al. / Composite Structures 58 (2002) 497–503 501

The two-parameter Weibull distribution cannot be Table 3

rejected at the 5% significance in 19 out the 26 cases. The Ratio of 5-percentile value to mean value
three-parameter Weibull distribution is rejected in five Material Ratio of 5-percentile value to mean value
cases for parameters determined with the modified mo- property (1) Two-parameter Three-parameter Weibull
ment method and in three cases for parameters deter- Weibull (2)
Modified Maximum
mined with the maximum likelihood method. In most moment likelihood
cases in which the three-parameter Weibull was rejected, method (3) method (4)
the two-parameter Weibull was also rejected for the Longitudinal tensile strength
same data set. VG 1–6 0.825 0.904 0.909
The average OSL for the two-parameter Weibull VG 7–12 0.754 0.817 0.814
VG 13–18 0.800 0.857 0.863
distribution was 0.178, with the average OSL for the
VG 19–24 0.817 0.876 0.888
three-parameter Weibull distribution being 0.386 and PG 0.863 0.879 0.886
0.356 for the modified moment method and the maxi-
Longitudinal compressive strength
mum likelihood method, respectively. Although strictly
VG 1–6 0.821 0.906 0.908
speaking the OSL cannot be used for ranking distribu- VG 7–12 0.805 0.839 0.853
tions, higher values of the OSL do indicate a higher VG 13–18 0.772 0.852 0.857
significance level. Therefore, it appears that the three- VG 19–24 0.785 0.847 0.847
parameter Weibull distribution is slightly preferable to PG 0.765 0.822 0.825
PG–M 0.811 0.834 0.841
the two-parameter Weibull distribution, though the two-
parameter Weibull distribution cannot be rejected in Shear strength
most cases. VG 19–24 0.876 0.867 0.876
PG 0.802 0.862 0.874
Abdallah et al. [1] compared the two-parameter and
three-parameter Weibull distribution for compressive, Transverse compressive strength
tensile, and flexural strength and stiffness of glass fiber PG-M 0.857 0.910 0.913
reinforced pultruded rods. Distributional parameters Longitudinal tensile modulus
were obtained by using linear regression of the data VG 1–6 0.857 0.894 0.895
when plotted on Weibull probability paper. The two- VG 7–12 0.857 0.824 0.857
VG 13–18 0.785 0.825 0.838
parameter and three-parameter Weibull distribution
VG 19–24 0.899 0.931 0.934
were compared on the basis of the correlation coefficient PG 0.871 0.900 0.903
between the best fit line and the data. The correlation
Longitudinal compressive modulus
coefficient for the two-parameter Weibull distribution
VG 1–6 0.884 0.912 0.918
averaged 0.902 for the 22 data sets considered, while the VG 7–12 0.856 0.854 0.856
correlation coefficient averaged 0.924 for the three- VG 13–18 0.748 0.814 0.826
parameter Weibull distribution. This is consistent with VG 19–24 0.926 0.927 0.926
the present results in that the three-parameter Weibull PG 0.914 0.932 0.936
distribution in general provides a slightly better fit of the Shear modulus
data than the two-parameter Weibull distribution, al- VG 19–24 0.760 0.815 0.760
though the two-parameter Weibull distribution provides PG 0.784 0.837 0.855
a reasonable fit.

although would need to be included in actual code de-

velopment.
4. Nominal design value The three-parameter Weibull distributions resulted in
5-percentile values that averaged 5% greater than the
Nominal design values are used in design practice. 5-percentile value from the two-parameter Weibull dis-
Typically in structural engineering, the nominal design tribution. The 5-percentile value from the maximum
value for material properties is the 5-percentile value [5]. likelihood method parameters was just slightly greater
The 5-percentile value for each of the three Weibull than that obtained from the modified moment method
distributions being considered is shown in Table 3. parameters, about half a percent. Using the two-
Statistical uncertainty, or the uncertainty from using a parameter Weibull distribution would result in lower
finite sample size to estimate the parameters, was not estimates of the 5-percentile value relative to the three-
accounted for in any of the design values. Several parameter Weibull distribution. This is consistent with
methods are available for accounting for statistical un- the results of Tenn [13] who found that the two-
certainty, with no agreed upon method. Rather than parameter Weibull distribution resulted in lower allow-
potentially biasing the results by accounting for statis- able design values than the three-parameter Weibull,
tical uncertainty, it was not included in the calculations, although he found only about a 1% decrease.
502 M. Alqam et al. / Composite Structures 58 (2002) 497–503

5. Effect of distribution on allowable load Table 4

Ratio of allowable load from three-parameter Weibull to two-pa-
rameter Weibull for a reliability index of 3.00
A performance function is considered as follows.
Material property (1) Ratio of allowable load to allowable load
g ¼RLD ð15Þ for two parameter Weibull distribution

where R is the resistance, L is a live load, D is a dead Modified moment Maximum likelihood
method (2) method (3)
load, and g < 0 is failure. The load is considered to be
half dead load and half live load. The dead load is as- Longitudinal tensile strength
VG 1–6 1.07 1.07
sumed to follow a normal distribution with a coefficient VG 7–12 1.18 1.18
of variation of 0.10, and the live load is assumed to VG 13–18 1.09 1.09
follow an Extreme Type I distribution with a coefficient VG 19–24 1.08 1.08
of variation of 0.25. These are approximately the sta- PG 1.01 1.01
tistics and load ratio applicable to both building and Longitudinal compressive strength
bridge loads [6,11]. VG 1–6 1.08 1.07
Each data set was used for the resistance, R, and a VG 7–12 1.07 1.08
dead and live load, D and L, were determined using first- VG 13–18 1.15 1.15
VG 19–24 1.12 1.12
order reliability methods [2] such that the reliability PG 1.16 1.16
index would be 3.00. Both the three-parameter Weibull PG-M 1.05 1.06
distribution determined using the modified moment
Shear strength
method and the maximum likelihood method were used. VG 19–24 0.99 1.00
The ratios of the allowable load from the three-pa- PG 1.09 1.09
rameter Weibull distribution to the allowable load from
Transverse compressive strength
the two-parameter Weibull distribution are shown in PG-M 1.03 1.03
Table 4.
The three-parameter Weibull distributions resulted Longitudinal tensile modulus
VG 1–6 1.02 1.02
in allowable loads that averaged 7% greater than al- VG 7–12 0.95 1.00
lowable loads from the two-parameter Weibull distri- VG 13–18 1.10 1.11
bution, with the modified moment method giving on VG 19–24 1.00 1.00
the average just slightly higher allowable loads. Given PG 1.01 1.01
the fact that different chosen probability distributions Longitudinal compressive modulus
can result in allowable loads that vary by as much as VG 1–6 1.00 1.00
50% [7], the 7% variation in allowable load is not very VG 7–12 0.99 1.00
significant. VG 13–18 1.20 1.21
VG 19–24 1.00 1.00
The allowable load was also examined for a live load PG 1.00 1.00
to dead load ratio of four instead of one. This was based
on composite structures being lighter, resulting in higher Shear modulus
VG 19–24 1.16 1.00
live to dead load ratios. For a live load to dead load PG 1.12 1.13
ratio of 4, the allowable load from the three-parameter
Weibull distributions averaged 3% greater than the al-
lowable load using the two-parameter Weibull distri- used. This is because of the relative ease in estimating
bution. The decrease in the ratio of allowable loads was the distribution parameters, the less likelihood in ob-
expected since the resistance statistics become less im- taining a negative location parameter, and the similarity
portant as the uncertainty in the load increases with in the results to those obtained using the maximum
increasing live to dead load ratios. The 3% difference is likelihood method.
virtually insignificant. The two-parameter Weibull distribution could not be
rejected at the 5% level for 19 out of the 26 data sets.
Although the two-parameter Weibull resulted in lower
6. Evaluation of results nominal design values and lower allowable loads than
either of the three-parameter Weibull distributions, the
Not much difference was observed between the differences were well less than ten percent. This differ-
modified moment method and the maximum likelihood ence is insignificant in most structural engineering ap-
method for the three-parameter Weibull distribution. plications.
Both resulted in similar OSL, nominal design values, Even though the three-parameter Weibull distribu-
and allowable loads. If a three-parameter Weibull dis- tion is more tobust than the two-parameter Weibull
tribution is to be used to characterize the data, it is distribution, there are some problems with the three-
recommended that the modified moment method be parameter Weibull distribution. The basic conceptual
 M. Alqam et al. / Composite Structures 58 (2002) 497–503 503

problem is that there is a nonzero lower bound, which References

implies that for loads less than the lower bound there is
no chance of failure. Average values of the lower bound, [1] Abdallah MH, Abdin EM, Selmy AI, Khashaba UA. Reliability
analysis of GFRP pultruded composite rods. Int J Qual Reliab
or location parameter, were within 10% of the first order
Manage 1996;13(2):88–98.
statistic, and were about 80% of the mean value. If ad- [2] Ang AH-S, Tang W. In: Probability Concepts in Engineering
ditional data were to be collected such that, for example, Planning and Design, vol. II. New York, NY: John Wiley & Sons;
the sample size was doubled, the scale parameter would 1984. p. 11.
undoubtedly decrease. [3] Ditlevsen O. Uncertainty Modeling with Applications to Multi-
dimensional Civil Engineering Systems. New York, NY: Mc-
Graw-Hill; 1981.
[4] Dodson B. Weibull Analysis. Milwaukee, Wisconsin: ASQ Qual-
7. Conclusions ity Press; 1994.
[5] Ellingwood BR. Load and Resistance Factor Design (LRFD) for
Structures using Fiber-Reinforced Polymer (FRP) Composites.
The three-parameter and two-parameter Weibull
NIST GCR 00-793, National Institute of Standards and Tech-
distributions were compared using 26 sets of data on nology, 2000.
fiber-reinforced polymeric pultruded composite material [6] Galambos TV, Ellingwood B, MacGregor JG, Cornell CA.
mechanical properties. Both strength and stiffness Probability based load criteria: assessment of current design
properties were examined. It is recommended that the practice. J Struct Div, ASCE 1982;108(ST5):959–77.
two-parameter Weibull distribution be used to charac- [7] Gromala DS, Sharp DJ, Pollock DG, Goodman JR. Load and
resistance factor design for wood: the new US wood design
terize FRP composite material properties. The primary specification. In: International Timber Engineering Conference,
basis for this recommendation is small differences in Tokyo, Japan. 1990. p. 311–8.
nominal design values and small differences in allowable [8] King RL. Statistical methods for determining design allowable
loads between the two-parameter and three-parameter properties for advanced composite materials. In: 15th Reinforced
Weibull distributions. Other supporting reasons for the Plastics Congress, British Plastics Federation, Nottingham, En-
gland. 1986. p. 79–85.
recommendation are similar OSL in distribution fitting, [9] Lawless JF. Statistical models and methods for lifetime data. New
computational efficiency, and the fact that the location York, NY: John Wiley and Sons; 1982.
parameter of the three-parameter Weibull distribution is [10] Mottram JT. Compression strength of pultruded sheet material.
near the first order statistic. The last reason implies there J Mater Civil Eng 1994;6(2):185–200.
is a load near the lowest data point that can be applied [11] Nowak AS. Calibration of LRFD bridge code. J Struct Eng,
ASCE 1995;121(8):1245–51.
to the structural member for which there is no chance of [12] Rust SW, Todt FR, Harris B, Neal D, Vangel M. Statistical
failure. This seems counterintuitive, as it seems reason- methods for calculating material allowables for MIL-HDBK-17.
able that there would be some chance of failure at any In: Test Methods for Design Allowables for Fibrous Composites.
load level, albeit the probability could be quite small. ASTM STP 1003; 1989. p. 136–49.
[13] Tenn LF. Statistical analysis of fibrous composite strength data.
In: Test Methods and Design Allowables for Fibrous Composites.
ASTM STP 734; 1981. p. 229–44.
Acknowledgements [14] Weibull W. A statistical distribution function with wide applica-
bility. Appl Mech 1951;18:293–7.
This work was partially supported by the Federal [15] Zanakis SH. A simulation study of some simple estimators for the
three-parameter Weibull distribution. J Stat Comput Simul
Highway Administration (FHWA) under Contract 1979;9:101–16.
No. DTFH61-00C-00022. Mr. Eric Munley serves as [16] Zureick A, Kang J, Butz T. Eccentrically loaded pultruded
the Contracting Officers Technical Representative for members in compression. Research Report SEMM 01-3, Georgia
FHWA. Additional funds were provided by the Georgia Institute of Technology, Atlanta, 2001.
Institute of Technology and the University of Tennessee. [17] Zureick A, Scott D. Short-term behavior and design of fiber-
reinforced polymeric slender members under axial compression.
The support from all of these organizations is gratefully J Compos Construct 1997;1(4):140–9.
acknowledged. The results presented herein represent [18] Zureick A, Steffen R. Behavior and design of concentrically
the views of the authors and not those of the sponsors. loaded pultruded angle struts. J Struct Eng 2000;126(3):406–16.