In this project, we will calculate the Morse index of FBMS catenoid inside a Spheroid. Consider a catenoid inside and spheroid. It is genus zero and has two boundary components. This surface describe by a conformal map 𝛴=X(M) for M cylinder [-T,T]×S1 with coordinates (t,𝜃). The critical catenoid is isometric to the immersion X(t,𝜃)=c(coshtcos𝜃,coshtsin𝜃,t),where 𝜃[0,2𝜋) and t[-T,T].
Figure 1:X(t,𝜃) with c=1 and t[-1,1]
Now we can calculate tangent vector fields of critical catenoid. Xt(t,𝜃)=c(sinhtcos𝜃,sinhtsin𝜃,1)X𝜃(t,𝜃)=c(-coshtsin𝜃,coshtcos𝜃,0)
Figure 2:Xt(t,𝜃) with c=1 and t[-1,1]
Figure 3:X𝜃(t,𝜃) with c=1 and t[-1,1]
Figure 4:X(t,𝜃),Xt(t,𝜃),X𝜃(t,𝜃)
Observe that Xt×X𝜃=|e1e2e3csinhtcos𝜃csinhtsin𝜃c-ccoshtsin𝜃ccoshtcos𝜃0|=(0-c2coshtcos𝜃)e1-(0+c2coshtsin𝜃)e2+(c2coshtsinhtcos2𝜃+c2coshtsinhtsin2𝜃)e3=(-c2coshtcos𝜃,-c2coshtsin𝜃,c2coshtsinht(cos2𝜃+sin2𝜃))=(-c2coshtcos𝜃,-c2coshtsin𝜃,c2coshtsinht)That implies|Xt×X𝜃|=(-c2coshtcos𝜃)2+(-c2coshtsin𝜃)2+(c2coshtsinht)2=c4cosh2tcos2𝜃+c4cosh2tsin2𝜃+c4cosh2tsinh2t=c4cosh2tcos2𝜃+c4cosh2tsin2𝜃+c4cosh2tsinh2t=c2cosh2t(1+sinh2t)=c2cosh4t=c2cosh2t Next, we can calculate normal derivative n1 as n1=Xt×X𝜃|Xt×X𝜃|=1cosht(-cos𝜃,sin𝜃,sinht) (3) by (1) and (2) We know general equation in spherical coordinate for the Spheroid is E=(acos𝜃sinv,asin𝜃sinv,bcosv),where v[0,𝜋] and a and b are constants.
Figure 5:E with a=1 and b=2
Now we try to deduce Spheroid equation in cylindrical coordinates. Reason behind that is our catenoid equation is in cylindrical coordinates. Spheroid equation isx2a2+y2a2+z2b2=1 In cylindrical coordinate system, we can write x=rcos𝜃y=rsin𝜃z=tFrom (4), we can write r2cos2𝜃a2+r2sin2𝜃a2+t2b2=1r2a2+t2b2=1t2b2=1-r2a2t=±b1-r2a2Therefore we can write Spheroid equation E is cylindrical coordinatesE=(rcos𝜃,rsin𝜃,±b1-r2a2)
Figure 6:E
So in yz-plane we have the projection of E as Ep, (set 𝜃=𝜋/2)Ep=(r,±b1-r2a2)
Figure 7:Ep with a=1, b=2
Or I can write this way. From (5) I haver2a2+t2b2=1r=±a1-t2b2r=a1-t2b2(6) since r>0. Therefore we can write Spheroid equation E is cylindrical coordinatesE=(a1-t2b2cos𝜃,a1-t2b2sin𝜃,t)
Figure 8:E
So in yz-plane we have the projection of E as Ep, (set 𝜃=𝜋/2 and 𝜃=-𝜋/2)Ep=(±a1-t2b2,t)Ept=(atb211-t2b2,1)
Figure 9:Ep
Also, projection of Catenoid to yz-plane is Xp=c(cosht,t)
Figure 10:Xp at c=1
Here we can see that, to intersect both catenoid and spheroid, |c|<|a|.
Figure 11:Ep and Xp at c=0.8.
We can find tangent vector fields of Xp as Xpt=c(sinht,1)
Figure 12:Vector plot of Xpt at c=0.8
Now I need to find normal vector for Ep. To do that, first I'm trying to find the curvature. Let us take the curve equation of the Ep is 𝛾Ep. Therefore 𝛾Ep=(±a1-t2b2,t)Also, we can write 𝛾'Ep=(atb211-t2b2,1)=(atb2bb2-t2,1)=(atb1b2-t2,1)|𝛾'Ep|=a2t2b2(b2-t2)+1=a2t2+b4-b2t2b2(b2-t2)=t2(a2-b2)+b4b2(b2-t2)Then arc length function is s(t)=0tt2(a2-b2)+b4b2(b2-t2)dt=1b0tt2(a2-b2)+b4b2-t2dtLet t=bsin𝜃dt=bcos𝜃d𝜃and t0,𝜃0 and tt,𝜃arcsin(tb). Sos(t)=1b0arcsin(tb)b2sin2𝜃(a2-b2)+b42b21-b2sin2𝜃bcos𝜃d𝜃=0arcsin(tb)sin2𝜃(a2-b2)+b2d𝜃=b0arcsin(tb)1-(1-a2b2)sin2𝜃d𝜃=bE(arcsin(tb)|(1-a2b2)), where E(arcsin(tb)|(1-a2b2)) is elliptic integral of the second kind. Consider, EpEpt=(±a1-t2b2,t)(atb211-t2b2,1)=-a2tb2+tThis proves position vector is not normal to the curve. Next I'm trying to find unit tangent vector to Ep. We know that, from (8), Ept=(atb211-t2b2,1)=(atbb2-t2,1)Therefore T=Ept|Ept|=(atbb2-t2,1)a2t2b2(b2-t2)+1=(atbb2-t2,1)a2t2-b2t2+b4b2(b2-t2)=(atbb2-t2,1)(a2-b2)t2+b4b2(b2-t2)=(atbb2-t2,1)(a2-b2)t2+b4bb2-t2=(at(a2-b2)t2+b4,bb2-t2(a2-b2)t2+b4)dTdt=(a(a2-b2)t2+b4-at2t(a2-b2)2(a2-b2)t2+b4(a2-b2)t2+b4,b(-2t2b2-t2)(a2-b2)t2+b4-bb2-t2(2(a2-b2)t2(a2-b2)t2+b4)(a2-b2)t2+b4)=(a[(a2-b2)t2+b4-t2(a2-b2)]((a2-b2)t2+b4)3/2,-bt((a2-b2)t2+b4)-b(b2-t2)(a2-b2)t((a2-b2)t2+b4)3/2b2-t2)=(ab4((a2-b2)t2+b4)3/2,-bt((a2-b2)t2+b4+(b2-t2)(a2-b2))((a2-b2)t2+b4)3/2b2-t2)=(ab4((a2-b2)t2+b4)3/2,-bt((a2-b2)t2+b4+(b2-t2)(a2-b2))((a2-b2)t2+b4)3/2b2-t2)=(ab4((a2-b2)t2+b4)3/2,-bt((a2-b2)t2+b4+(a2-b2)b2-(a2-b2)t2)((a2-b2)t2+b4)3/2b2-t2)=(ab4((a2-b2)t2+b4)3/2,-bt(b4+a2b2-b4)((a2-b2)t2+b4)3/2b2-t2)=(ab4((a2-b2)t2+b4)3/2,-a2b3t((a2-b2)t2+b4)3/2b2-t2)|dTdt|=a2b8((a2-b2)t2+b4)3+a4b6t2((a2-b2)t2+b4)3(b2-t2)=a2b8(b2-t2)((a2-b2)t2+b4)3(b2-t2)+a4b6t2((a2-b2)t2+b4)3(b2-t2)=a2b8(b2-t2)+a4b6t2((a2-b2)t2+b4)3(b2-t2)=a2b10-a2b8t2+a4b6t2((a2-b2)t2+b4)3(b2-t2)=a2b10-t2a2b6(b2-a2)((a2-b2)t2+b4)3(b2-t2)=ab3b4+t2(-b2+a2)((a2-b2)t2+b4)32(b2-t2)=ab3((a2-b2)t2+b4)b2-t2 Then unit normal vector can be written as N=dTdt|dTdt|. So by (9), (10) and definition of N, N=(ab4((a2-b2)t2+b4)31/2,-a2b3t((a2-b2)t2+b4)31/2b2-t2)ab3((a2-b2)t2+b4)b2-t2=(bb2-t2((a2-b2)t2+b4)1/2,-at((a2-b2)t2+b4)1/2)=(bb2-t2(a2-b2)t2+b4,-at(a2-b2)t2+b4)