This report presents the design procedures and average run lengths of two multivariate cumulative sum (CUSUM) quality control procedures. The first CUSUM procedure reduces each multivariate observation to a scaler and then forms a CUSUM of the scalers. The second CUSUM procedure forms a CUSUM vector directly from the observations. These two procedures are compared to each other and to the multivariate Shewhart chart. Other multivariate quality control procedures are mentioned. Robustness, the...

Topics: DTIC Archive, Crosier, Ronald B, CHEMICAL RESEARCH DEVELOPMENT AND ENGINEERING CENTER ABERDEEN...

New three-level designs for fitting second-order response surfaces are obtained by three methods. New designs for 9, 10, and 13 factors are obtained by combining two-level factorial designs and incomplete block designs. New designs for 6, 8 and 10 factors are obtained by design rotation; these designs are rotations of central composite designs. Two sequences of design (for 7, 11 and 15 factors) are obtained by using a Hadamard matrix to reduce v+1 linearly dependent variables to a design for v...

Topics: DTIC Archive, Crosier, Ronald B, CHEMICAL RESEARCH DEVELOPMENT AND ENGINEERING CENTER ABERDEEN...

First-order saturated designs can have the two desirable properties of being an orthogonal design and having two levels only if n=O mod 4, where n is the number of design points. For other cases, researchers have developed D-optimal two level designs. Recently, two-level saturated designs that are efficient for submodels containing only a few of the factors were developed. These p-efficient designs (p is the number of parameters in a submodel) have two properties not possessed by the D-optimal...

Topics: DTIC Archive, Crosier, Ronald B., EDGEWOOD RESEARCH DEVELOPMENT AND ENGINEERING CENTER ABERDEEN...

Designs for fitting first-order response surfaces over spherical regions are often constructed from the vertices of regular geometric figures. Designs constructed from cubes (the two-level factorial designs) are naturally oriented so that the designs treat the factors equally, but the simplex designs are usually presented in an orientation that gives the factors unequal, asymmetric ranges. The unequal, asymmetric ranges of the coded design factors create a problem in applying the designs. The...

Topics: DTIC Archive, Crosier, Ronald B, EDGEWOOD RESEARCH DEVELOPMENT AND ENGINEERING CENTER ABERDEEN...

The central composite design and designs based on the icosahedron and the dodecahedron are given for four component mixture systems. A uniform shell design is applied to the development of an obscurant smoke. Oils, cements, gasolines, and perfumes, are mixtures of components. The development of better products of these types involves experimentation with different formulations, or blends, of the components. Keywords: Surface methodology, Shell design, Geometry, Review, Central composite.

Topics: DTIC Archive, Crosier, Ronald B, CHEMICAL RESEARCH DEVELOPMENT AND ENGINEERING CENTER ABERDEEN...

The shell designs consist of two series of related designs (the uniform shell designs and the simplicial shell designs) that are used for fitting second-order response surfaces over spherical regions. The shell designs for two, four, six, and eight factors require fewer experimental runs than the central composite designs or the Box-Behnken designs, but they have not been used much because of their unequal treatment of the design factors. Rotations of the shell designs that provide an equal...

Topics: DTIC Archive, Crosier, Ronald B, EDGEWOOD RESEARCH DEVELOPMENT AND ENGINEERING CENTER ABERDEEN...

The present chemical warfare (CW) agent toxicity estimates are not suitable for use with the general population (GP) because they are framed for male soldiers. A method was created to convert the median effective dose and probit (or Bliss) slope to estimates applicable to the GP. It was assumed that individual susceptibilities have a log-normal distribution. Two mathematical models were developed to describe a healthy or sensitive subpopulation (SP). In the tail model, the SP consists of all...

Topics: DTIC Archive, Crosier, Ronald B, EDGEWOOD CHEMICAL BIOLOGICAL CENTER ABERDEEN PROVING GROUND MD,...

Civil defense planning requires estimates of the toxicity of chemical warfare agents to the general public, but the current toxicity estimates are for male soldiers. In ECBC-TR-224 and ECBC-TR-337, individual susceptibilities for both the general population and the military subpopulation were modeled by a lognormal distribution. The assumption of a lognormal distribution of individual susceptibilities for both the general population and a subpopulation cannot be correct. This report presents an...

Topics: DTIC Archive, Crosier, Ronald B, EDGEWOOD CHEMICAL BIOLOGICAL CENTER ABERDEEN PROVING GROUND MD...

Response surface designs are often compared by optimal design theory criteria, such as the D- and G-efficiencies. For designs over spherical regions such criteria are calculated after scaling the designs to have the same diameter. But in practice designs are applied by scaling the coded factor ranges (not the diameter) to the ranges of the experimental factors. The result of this discrepancy between theory and practice is that the mathematical efficiencies may apply to regions of varying sizes....

Topics: DTIC Archive, Crosier, Ronald B, EDGEWOOD RESEARCH DEVELOPMENT AND ENGINEERING CENTER ABERDEEN...

All rotations of a response surface design can be obtained by making planar rotations in the set of planes defined by every pair of the coordinate axes. Standard multivariate optimization routines may then be used to optimize a criterion for the desirability of a rotation as a function of the set of planar angles. Criteria for the symmetry of the design (such as the same set of factor levels for all factors) and for the size of the design region (changed by rescaling after the rotation) are...

Topics: DTIC Archive, Crosier, Ronald B, EDGEWOOD RESEARCH DEVELOPMENT AND ENGINEERING CENTER ABERDEEN...

The present chemical warfare agent toxicity estimates are not suitable for use With the general population (GP) because they are framed for male soldiers. A method was created to convert the median effective dose and Bliss slope to estimates applicable to the GP. It was assumed that individual susceptibilities have a log-normal distribution. Two mathematical models were developed to describe a healthy or sensitive subpopulation (SP). In the tail model, the SP consists of all individuals having...

Topics: DTIC Archive, Crosier, Ronald B, EDGEWOOD CHEMICAL BIOLOGICAL CENTER ABERDEEN PROVING GROUND MD,...

Civil defense planning requires estimates of the toxicity of chemical warfare agents to the general public, but the current toxicity estimates are for male soldiers. Individual susceptibilities for both the general population and the military subpopulation are modeled by a lognormal distribution. Historical military demographics are used to estimate the size of the subpopulation from which military personnel are drawn. Limits on the median effective dose and the probit slope of a subpopulation...

Topics: DTIC Archive, Crosier, Ronald B, EDGEWOOD CHEMICAL BIOLOGICAL CENTER ABERDEEN PROVING GROUND MD...