I will appreciate any suggestion on how I could have a solution to this.

SX=1000*[1 2 3];
SY=2000*[1.5 2 3];
SXY = 1258[1 2 3];
a = [0.3 0.6 0.9];
syms rb
for j=1:1:3
if pwmid(j)<=pwc(j)
SRR(j)=0.5*(SX(j)+SY(j)).*(1-(a(j).^2)/rb^2)+0.5*(SX(j)-SY(j)).*(1+(3*a(j).^4/rb^4)-(4*a(j).^2/rb^2))*cos(2*thbkso(j))...
+SXY(j).*(1+(3*a(j).^4/rb^4)-(4*a(j).^2/rb^2))*sin(2*thbkso(j))+(a(j).^2/rb^2).*pwmid(j);
STT(j)=0.5*(SX(j)+SY(j)).*(1+(a(j).^2)/rb^2)-0.5*(SX(j)-SY(j)).*(1+(3*a(j).^4/rb^4))*cos(2*thbkso(j))...
-SXY(j).*(1+(3*a(j).^4/rb^4))*sin(2*thbkso(j))-(a(j).^2/rb^2).*pwmid(j);
SRT(j)=(0.5*(SX(j)-SY(j)).*sin(2*thbkso(j))+SXY(j).*cos(2*thbkso(j))).*(1-(3*a(j).^4/rb^4)+(2*a(j).^2/rb^2));
SIGMA1A(j)=0.5*(STT(j)+SRR(j))+0.5*((STT(j)-SRR(j)).^2+4*SRT(j).^2).^0.5;
SIGMA3A(j)=0.5*(STT(j)+SRR(j))-0.5*((STT(j)-SRR(j)).^2+4*SRT(j).^2).^0.5;
C0FUN(j)=SIGMA1A(j)-SIGMA3A(j);
rbsoln{j}=double(vpasolve(C0FUN(j)==C0(j),rb));
cell(rbsoln);
rw(j)=rbsoln{j}(1);
rbkt_art(j) = rbsoln{j}(1)-a(j);
else
rw(j)=a(j);
rbkt_art(j)=rbkt_int(j);
end
end

8 Comments

pwmid=[1000 2000 3000]; pwc = [500 1000 3500];
thbkso is a matrix with known values...any values can be used; but my concern is the use of solve or vpasolve not giving me favorable results
I do not understand what your question is? vpasolve() is going to give you a solution if it can find one. If it is not the "right" solution then you need some way of characterizing the "right" solution. For example is there a range of values that the "right" solution will always be in?
C0(j) has not been assigned a meaning.
Sorry
SXY = 125*[1 2 3]; C0 = 100*[1 2 3];
Thanks

Sign in to comment.

Answers (3)

All solutions to those equations are strictly imaginary for the parameters you give.
For example, for j = 1, the solutions are
(3/10)*sqrt(5)*sqrt(roots([+8105,-9500,+4790,-1164,+117]))
and the negatives of those.

2 Comments

Please explain what you mean when you said you were concerned about solve or vpasolve "not giving favorable results" ?
If you want all of the results, then you may have to use solve() instead of vpasolve(), and you might have to double() the result of solve() to get numeric values. I do not have the Symbolic Toolbox so I cannot check exactly what would be returned.

I could have made a mistake along the way, but if I got it right then:

for j = 1 : 3
  A = 32 * SXY(j) * sin(thbkso(j)) * cos(thbkso(j))^3 * SX(j) - 32 * SXY(j) * sin(thbkso(j)) * cos(thbkso(j))^3 * SY(j) - 16 * SXY(j) * sin(thbkso(j)) * cos(thbkso(j)) * SX(j) + 16 * SXY(j) * sin(thbkso(j)) * cos(thbkso(j)) * SY(j) - C0(j)^2 + SX(j)^2 - 2 * SX(j) * SY(j) + 4 * SXY(j)^2 + SY(j)^2;
    B =  - 32 * cos(thbkso(j))^4 * SX(j)^2 * a(j)^2 + 64 * cos(thbkso(j))^4 * SX(j) * SY(j) * a(j)^2 + 128 * cos(thbkso(j))^4 * SXY(j)^2 * a(j)^2 - 32 * cos(thbkso(j))^4 * SY(j)^2 * a(j)^2 - 8 * sin(thbkso(j)) * cos(thbkso(j)) * SX(j) * SXY(j) * a(j)^2 - 8 * sin(thbkso(j)) * cos(thbkso(j)) * SXY(j) * SY(j) * a(j)^2 + 16 * sin(thbkso(j)) * cos(thbkso(j)) * SXY(j) * a(j)^2 * pwmid(j) + 28 * cos(thbkso(j))^2 * SX(j)^2 * a(j)^2 - 64 * cos(thbkso(j))^2 * SX(j) * SY(j) * a(j)^2 + 8 * cos(thbkso(j))^2 * SX(j) * a(j)^2 * pwmid(j) - 128 * cos(thbkso(j))^2 * SXY(j)^2 * a(j)^2 + 36 * cos(thbkso(j))^2 * SY(j)^2 * a(j)^2 - 8 * cos(thbkso(j))^2 * SY(j) * a(j)^2 * pwmid(j) - 2 * SX(j)^2 * a(j)^2 + 8 * SX(j) * SY(j) * a(j)^2 - 4 * SX(j) * a(j)^2 * pwmid(j) + 16 * SXY(j)^2 * a(j)^2 - 6 * SY(j)^2 * a(j)^2 + 4 * SY(j) * a(j)^2 * pwmid(j);
    C = 128 * sin(thbkso(j)) * cos(thbkso(j))^3 * SX(j) * SXY(j) * a(j)^4 - 128 * sin(thbkso(j)) * cos(thbkso(j))^3 * SXY(j) * SY(j) * a(j)^4 + 48 * cos(thbkso(j))^4 * SX(j)^2 * a(j)^4 - 96 * cos(thbkso(j))^4 * SX(j) * SY(j) * a(j)^4 - 192 * cos(thbkso(j))^4 * SXY(j)^2 * a(j)^4 + 48 * cos(thbkso(j))^4 * SY(j)^2 * a(j)^4 - 48 * sin(thbkso(j)) * cos(thbkso(j)) * SX(j) * SXY(j) * a(j)^4 + 80 * sin(thbkso(j)) * cos(thbkso(j)) * SXY(j) * SY(j) * a(j)^4 - 32 * sin(thbkso(j)) * cos(thbkso(j)) * SXY(j) * a(j)^4 * pwmid(j) - 40 * cos(thbkso(j))^2 * SX(j)^2 * a(j)^4 + 96 * cos(thbkso(j))^2 * SX(j) * SY(j) * a(j)^4 - 16 * cos(thbkso(j))^2 * SX(j) * a(j)^4 * pwmid(j) + 192 * cos(thbkso(j))^2 * SXY(j)^2 * a(j)^4 - 56 * cos(thbkso(j))^2 * SY(j)^2 * a(j)^4 + 16 * cos(thbkso(j))^2 * SY(j) * a(j)^4 * pwmid(j) + 7 * SX(j)^2 * a(j)^4 - 18 * SX(j) * SY(j) * a(j)^4 + 4 * SX(j) * a(j)^4 * pwmid(j) - 8 * SXY(j)^2 * a(j)^4 + 15 * SY(j)^2 * a(j)^4 - 12 * SY(j) * a(j)^4 * pwmid(j) + 4 * a(j)^4 * pwmid(j)^2;
    E = 288 * SXY(j) * a(j)^8 * sin(thbkso(j)) * cos(thbkso(j))^3 * SX(j) - 288 * SXY(j) * a(j)^8 * sin(thbkso(j)) * cos(thbkso(j))^3 * SY(j) - 144 * SXY(j) * a(j)^8 * sin(thbkso(j)) * cos(thbkso(j)) * SX(j) + 144 * SXY(j) * a(j)^8 * sin(thbkso(j)) * cos(thbkso(j)) * SY(j) + 9 * SX(j)^2 * a(j)^8 - 18 * SY(j) * a(j)^8 * SX(j) + 36 * SXY(j)^2 * a(j)^8 + 9 * SY(j)^2 * a(j)^8;
    sols_plus = sqrt( roots([A, B, C, 0, E]) );
    sols{j} = [sols_plus; -sols_plus];
  end

I am not certain of these coefficients; I am concerned that the previous solution did not have a 0 in the x^1 position but this does.

Sign in to comment.

Thanks Walter...yes, the solutions are all imaginary with the current inputs

Categories

Find more on Mathematics in Help Center and File Exchange

Asked:

on 12 Aug 2015

Commented:

on 14 Aug 2015

Community Treasure Hunt

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

Start Hunting!