root/OptimizingMATLABCode/Truss/paramSweepSerial.m @ 10
| 10 | anderm8 | function [nVals,aVals,peakVals,computeTime] = paramSweepSerial(nNum,aNum)
|
|
% [nVals,aVals,peakVals,computeTime] = paramSweepSerial(nNum,aNum)
|
|||
%
|
|||
% This is a parameter sweep study of the effect of the number of elements
|
|||
% and element cross sectional area on the displacement at the tip of a
|
|||
% cantilevered truss.
|
|||
%
|
|||
% Inputs:
|
|||
%
|
|||
% nNum: Number of nodes to try along bottom cord (scalar)
|
|||
%
|
|||
% aNum: Number of cross sectional areas to try (scalar)
|
|||
%
|
|||
% Outputs:
|
|||
%
|
|||
% nVals: vector with the number of nodes tested
|
|||
%
|
|||
% aVals: vector of cross sectional areas tested
|
|||
%
|
|||
% peakVals: maximum deflection at the truss tip for each combination of
|
|||
% nVals and aVals
|
|||
%
|
|||
% computeTime: parfor-loop computation time
|
|||
%
|
|||
%
|
|||
% Copyright 2015 The MathWorks, Inc.
|
|||
%
|
|||
% Two inputs
|
|||
narginchk(2,2);
|
|||
%% Initialize Problem
|
|||
% Vectors of number of elements and cross section areas to sweep
|
|||
nVals = (1:nNum)+10; % number of segments, start with 11
|
|||
aVals = linspace(100, 200, aNum); % cross sectional area
|
|||
% Grid of all combinations
|
|||
[nGrid, aGrid] = meshgrid(nVals, aVals);
|
|||
% Peak value results matrix
|
|||
peakVals = nan(nNum,aNum);
|
|||
%% Parameter Sweep
|
|||
t0 = tic;
|
|||
for ii = 1:numel(aGrid)
|
|||
% Solve ODE
|
|||
Y = trussCantilever(nGrid(ii),aGrid(ii));
|
|||
% Determine peak deflection in Y direction at the tip
|
|||
peakVals(ii) = max(Y(:,end));
|
|||
end
|
|||
computeTime = toc(t0);
|
|||
end
|