Clear Filters
Clear Filters

ODE23s, Differential equation, problem solving

1 view (last 30 days)
Hello everybody. I have a serious problem solving the diff. equation, expressed below. Any help;
function [kap, alf0, Pstar, Astar]= nofilter()
imu=sqrt(-1);
rho=1000;RR=0.3;
Jb=0.1; %diastato
CC=0.1;%diastato
chr=0.1;
cht=0.1;
U=0.5;
span=0.34/2;
alam1=90;
alam2=90;
chm=0.5*(chr+cht);
span3d=2*span;
area=2*(chm*span);
AR=(span3d^2)/area;
AR=10;
B=14.75; %adiastato
BB=B*rho*U*0.34*chr*chr*RR; %diastato
pic=0.25;
II=Jb/pi*rho*area*chr*chr*chr; %adiastato
C=CC/1/2*pi*rho*area*chr*U^2; %adiastato
a=span3d*RR^2/chr*area;
b=RR^2/chr^2;
e=RR/chr;
d=RR*span3d/area;
alf0=linspace(1, 60, 2);
kap=linspace(0.025,0.2,2);
for i=1:length(alf0)
for j=1:length(kap)
om(j)=pi*kap(j)*2*U/chr; % reduced velocity k, Huxham, k=pi*khuxham
T(j)=2*pi/om(j);
alf0r(i)=alf0(i)*pi/180;
theo(j)=besselh(1,2,kap(j))./(besselh(1,2,kap(j))+imu*besselh(0,2,kap(j))); %theo function, lift deficiency factor
Re=U*chr*10^6; %reynolds
CD0=0.075/(log10(Re)+2)^2; %skin friction
fprintf('Calculating for %d\n', i, j);
[t,y]=ode23s(@(t,y) nofilter(t,y, i, j), [0 , 100],[0, 0]);
[Astar(i,j), Pstar(i,j)] = processsignal(t,y);
Pstar(i,j) = 0.5*BB.*(Pstar(i,j))./0.5*rho* 2*RR.*(Astar(i,j)).*cos(((Astar(i,j)))).*U^3;
end
end
function dYdt=nofilter(t,Y, i, j)
thi(i,j)=alf0r(i).*cos(om(j)*t);
k1=8*II+2*a*cos(thi(i,j)).*(cos(Y(1))-Y(1)*sin(Y(1)));
dYdt=[Y(2);(1/k1)*((Y(2)).*(-2*B+theo(j).*(AR/AR+2).*b.*cos(thi(i,j)).*(-cos(Y(1)+Y(1).*sin(Y(1)))))+...
(Y(2).^2)*2*a*cos(thi(i,j))*(2*sin(Y(1))*Y(1)+Y(1)*cos(Y(1)))-Y(1)*C+...
theo(j)*(AR/AR+2)*e*(thi(i,j).*cos(thi(i,j))+(3/4-pic)*2*thi(i,j).*cos(thi(i,j)))+...
d*(thi(i,j)+(1/2-pic)*2*thi(i,j).*cos(thi(i,j)))-e*(1/pi)*sin(thi(i,j))*(CD0+1.6*(thi-atan(2*e*Y(2).*(cos(Y(1))-Y(1).*sin(Y(1))))).^2))];
end
end
function [Aout,Pstar] = processsignal(t,y)
% process signal
Y = y(:,1);
Ydot = y(:,2);
k = 2:length(Y)-1;
Kmax = find( Y(k)>Y(k-1) & Y(k)>Y(k+1) );
Kmax = Kmax+1; %correct index
Kmin = find( Y(k)<Y(k-1) & Y(k)<Y(k+1) );
Kmin = Kmin+1; %correct index
ti_max = 0.5*(t(Kmax(1:end-1))+t(Kmax(2:end)));
ti_min = 0.5*(t(Kmin(1:end-1))+t(Kmin(2:end)));
N = 5; % number of final cycles for averagingaw
Aout = 0.5*(mean(Y(Kmax(end-N,end))) - mean(Y(Kmin(end-N,end))));
subplot(2,1,1), plot(t,Y,t(Kmax),Y(Kmax),'ob',t(Kmin),Y(Kmin),'ob')
subplot(2,1,2), plot(t(Kmax),Y(Kmax),'o',t(Kmin),abs(Y(Kmin)),'o')
Jint = find(t>=t(Kmax(end-N)) & t<t(Kmax(end)));
Pstar = (1/4)*trapz(t(Jint),Ydot(Jint).^2); % integral,1/4
end

Answers (0)

Categories

Find more on Programming in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!