UPSC 2026 Maths Optional Paper 2 Q6c — Step-by-Step Solution
15 marks · Section B
Two-Dimensional and Axisymmetric Flow · Mechanics & Fluid Dynamics · asked 2× in 14 yrs · Read the full method →
Question
Suppose that there is an infinitely long cylinder of radius 'a' placed in a uniform stream having velocity −Vi^. Then using Milne-Thomson’s circle theorem, what will be its complex velocity potential? If in addition, a circulation round the cylinder of k is produced, find out the stagnation points. Do they produce a lifting tendency in the vertical direction? Explain how.
Technique
The circle theorem is a machine: give it the potential of the undisturbed stream, and it returns the potential with the cylinder inserted — so the only real work is writing f(z) for the stream −Vi^ correctly (note the minus: this is the mirror image of the textbook +Ui^ case, and it flips the sign of the lift at the end). Circulation is then superposed as a line vortex at the centre, which leaves ∣z∣=a a streamline because its ψ is a function of r alone. Stagnation points come from dw/dz=0, a quadratic in z whose discriminant changes sign at ∣k∣=4πVa — the classification by circulation strength is where the marks that most scripts leave on the table are. The lift follows from Blasius (or Kutta–Joukowski), and the physical “how” from Bernoulli applied to the surface speed.
Solution
Throughout: two-dimensional, steady, irrotational flow of an inviscid incompressible fluid of density ρ, in the complex plane z=x+iy; the cylinder’s cross-section is the circle ∣z∣=a and the fluid occupies ∣z∣≥a. The complex potential is w(z)=ϕ+iψ, with the fundamental relation
dzdw=u−iv.
Take V>0, so that the stream runs in the negativex-direction.
Step 1 — The undisturbed stream.
The stream velocity is q=−Vi^, i.e. u=−V, v=0. Hence dzdw=−V and, ignoring an irrelevant additive constant,
f(z)=−Vz.(1)
This has no singularity in the finite plane, so the circle theorem applies with no restriction. (Written out: f=−V(x+iy) gives ϕ=−Vx, ψ=−Vy — streamlines y=const traversed in the sense of decreasing x. Correct.)
Step 2 — State Milne-Thomson’s circle theorem before using it.
Circle theorem. Let f(z) be the complex potential of a two-dimensional irrotational flow of incompressible inviscid fluid in which there are no rigid boundaries and all of whose singularities lie in ∣z∣>a. Then on introducing the circular cylinder ∣z∣=a into the flow, the complex potential becomes
w(z)=f(z)+f(zˉa2),
and ∣z∣=a is a streamline of the new flow.
Two one-line justifications, both worth marks:
(i) The boundary condition holds. On ∣z∣=a we have zzˉ=a2, so a2/zˉ=z and therefore
w=f(z)+f(z)=2Ref(z)∈R⟹ψ=Imw=0 on ∣z∣=a,
i.e. the circle is the streamline ψ=0 — exactly the condition for a rigid cylinder.
(ii) The flow far away is unchanged. If f is singular at z0 with ∣z0∣>a, the added term is singular only at the inverse pointa2/zˉ0, whose modulus a2/∣z0∣<a — inside the cylinder, hence outside the fluid. So no new singularity is introduced into ∣z∣>a.
Step 3 — Apply the theorem.
With f(z)=−Vz and V real, f(zˉa2)=(−Vzˉa2)=−Vza2. Therefore
w(z)=−V(z+za2)(∣z∣≥a).(2)
Two checks. As ∣z∣→∞, dzdw→−V, recovering the stream −Vi^. On z=aeiθ, z+za2=aeiθ+ae−iθ=2acosθ, so w=−2Vacosθ is real and ψ=0. ✓
Step 4 — Superpose the circulation.
The circle theorem produces the acyclic (circulation-free) flow. A circulation k round the cylinder is obtained by placing a line vortex at the centre z=0 — which is inside the cylinder, hence legitimate.
Convention (state it, because the sign of the final force depends on it): k is the circulation taken positive in the anticlockwise sense. The corresponding potential is
wv(z)=−2πiklogz.
Indeed, with z=reiθ, wv=−2πiklnr+2πkθ, so
ϕv=2πkθ,ψv=−2πklnr,qθ=r1∂θ∂ϕv=2πrk,
and ∮qθrdθ=k round any circuit enclosing the origin. Because ψv depends on r alone, it is constant on ∣z∣=a, so adding it does not disturb the boundary condition. Hence
w(z)=−V(z+za2)−2πiklogz,∣z∣≥a.(3)
Check on the circulation. For any circuit C round the cylinder, ∮Cdzdwdz=Γ+iQ (Γ = circulation, Q = flux). Here dzdw=−V+z2Va2−2πzik has residue −2πik at z=0, so the integral is 2πi(−2πik)=k: circulation k, zero flux (no source). ✓
Step 5 — Stagnation points.
Stagnation points are the zeros of dzdw lying in the fluid ∣z∣≥a:
dzdw=−V(1−z2a2)−2πzik=0.
Multiplying by −z2 (legitimate, since z=0 in the fluid):
Vz2+2πikz−Va2=0.(4)
This quadratic has
product of roots=V−Va2=−a2,z=2V−2πik±4V2a2−4π2k2=−4πVik±a2−16π2V2k2.(5)
Introduce the dimensionless circulation
λ=4πVak,
so that (5) becomes the compact
z±=a(±1−λ2−iλ),λ=4πVak(6)
Step 6 — Classify by the strength of the circulation. (This is the part answers skip.)
(i) Weak circulation, ∣k∣<4πVa (i.e. ∣λ∣<1). The square root is real and
∣z±∣2=a2[(1−λ2)+λ2]=a2,
so both stagnation points lie on the cylinder, at z=aeiθs with
sinθs=−λ=−4πVak,cosθs=±1−λ2.(7)
They are placed symmetrically about the y-axis and, for k>0, below the horizontal diameter (third and fourth quadrants). For k=0, θs=0,π: the points z=±a, the front and rear generators of the plain symmetric flow.
(ii) Critical circulation, ∣k∣=4πVa (∣λ∣=1). The two roots coalesce:
z=−iasgn(k),
a single double stagnation point at the lowest point of the cylinder when k>0 (highest when k<0). The critical value is
kcrit=4πVa
(iii) Strong circulation, ∣k∣>4πVa (∣λ∣>1). Now a2−k2/16π2V2=iaλ2−1 is imaginary and
z=ia(−λ±λ2−1),
so both roots lie on the y-axis. Their product is −a2, hence ∣z+∣∣z−∣=a2: one is inside the cylinder (it is the inverse point, not part of the fluid and not a physical stagnation point) and one is outside. For k>0 the physical one is
z=−ia(λ+λ2−1),∣z∣=a(λ+λ2−1)>a,
a single stagnation point standing off the cylinder, directly below it; the streamline through it closes on itself, trapping a body of fluid that is carried round with the cylinder.
Figure to draw (small, half a text-column, and it earns its space because it makes the classification and the lift argument visible at a glance): draw the circle ∣z∣=a with centre O and the two axes. Along the top and bottom draw three horizontal streamlines with arrowheads pointing in the −x direction, labelled "−Vi^"; crowd the lines above the cylinder and space them out below it. Draw a curved arrow anticlockwise round the circle labelled “circulation k>0”. Mark the two stagnation points S1,S2 on the lower half of the circle, symmetric about the y-axis, with the angle θs from the positive x-axis to OS2 and the annotation sinθs=−k/(4πVa). Add a thick upward arrow through O labelled Y=ρVk. In a small inset beside it, draw the same circle with a single point below it on the y-axis, labelled ”∣k∣>4πVa: stagnation point leaves the cylinder”. Do not add streamlines inside the circle — there is no flow there.
Step 7 — The force on the cylinder: Blasius’ theorem.
Blasius’ theorem. For steady two-dimensional irrotational flow, the force per unit length on a cylinder is given by
X−iY=2iρ∮C(dzdw)2dz,
where C is any contour enclosing the cylinder and no other singularity.
From (3),
dzdw=−V+z2Va2−2πzik.
Only the coefficient of z−1 in the square is needed (all other terms integrate to zero). Squaring, the z−1 term arises solely from the cross-product of −V with −2πzik:
(dzdw)2=⋯+2(−V)(−2πzik)+⋯=⋯+πiVk⋅z1+⋯,
(the terms z2Va2⋅(−V), z2Va2⋅(−2πzik) and the two squares contribute only z−2,z−3,z−4). Hence Resz=0(dzdw)2=πiVk and
X−iY=2iρ⋅2πi⋅πiVk=2iρ⋅2i2Vk=−iρVk.
Comparing real and imaginary parts:
X=0,Y=ρVk(8)
X=0 is d’Alembert’s paradox: no drag. The entire force is transverse to the stream — i.e. vertical.
Step 8 — “Explain how”: the physical mechanism, from Bernoulli.
On the cylinder z=aeiθ, equation (3) gives
ϕ=−2Vacosθ+2πkθ,q=a1∂θ∂ϕ=2Vsinθ+2πak.(9)
(Setting q=0 in (9) returns sinθ=−4πVak — an independent confirmation of (7), obtained without touching the quadratic (4).)
Bernoulli’s equation for the steady irrotational flow gives p+21ρq2=p∞+21ρV2, so where the fluid is faster, the pressure is lower. Compare the crown and the keel:
For k>0 this is positive: the flow is faster over the top and slower underneath, hence ptop<pbot, hence a net upward force. That is the mechanism.
Integrating it out confirms (8) without Blasius. The pressure force on an element is −pn^adθ with outward normal n^=(cosθ,sinθ), so
Y=−a∫02πpsinθdθ=2ρa∫02πq2sinθdθ
(the constant part of p contributes nothing, since ∫02πsinθdθ=0). With (9),
q2sinθ=4V2sin3θ+πa2Vksin2θ+4π2a2k2sinθ,
and ∫02πsin3θdθ=∫02πsinθdθ=0, ∫02πsin2θdθ=π. Hence
Y=2ρa⋅πa2Vk⋅π=ρVk.✓
The same computation with cosθ in place of sinθ gives X=0, since ∫sin2θcosθ=∫sinθcosθ=∫cosθ=0 over a period. ✓
Step 9 — Answer the question asked.
Yes — provided k=0, and the tendency is exactly vertical. The reasoning, in the order an examiner wants it:
With k=0 the potential (2) is even in y: the flow is symmetric about the x-axis, the stagnation points sit at θ=0,π on the horizontal diameter, the pressure distribution is up–down symmetric, and Y=0. No lift.
Circulation destroys that symmetry, and the migration of the stagnation points off the horizontal diameter to sinθs=−k/(4πVa) is precisely the visible signature of the asymmetry. Their displacement and the lift are two faces of the same broken symmetry — which is the sense in which the stagnation points “produce” the lift.
Quantitatively — the Kutta–Joukowski theorem. In vector form, for circulation Γ=kk^ (anticlockwise) in a stream U∞,
in agreement with (8). The magnitude is ρV∣k∣ per unit length, the direction is perpendicular to the stream — vertical — upward when k>0 (anticlockwise), downward when k<0.
4. Carry the sign of the stream. For the usual textbook stream +Vi^ the same anticlockwise k gives Y=−ρVk, i.e. downward. Reversing the stream to −Vi^, as this paper does, reverses the lift. Quoting the book formula without this reversal is the single commonest way to lose the last marks here.
5. Validity across the classification. Equation (8) holds for everyk: the lift does not care whether the stagnation points are on the cylinder or standing off below it. What changes at ∣k∣=4πVa is only their location. ■
Answer
Circle theorem: Stagnation points: Force per unit length: w(z)=−V(z+za2);with circulation k(anticlockwise):w(z)=−V(z+za2)−2πiklogz.z=−4πVik±a2−16π2V2k2=a(±1−λ2−iλ),λ=4πVak.∣k∣<4πVa:two points on the cylinder ∣z∣=a, at sinθs=−4πVak;∣k∣=4πVa:coincident at z=−iasgnk;∣k∣>4πVa:one point off the cylinder at z=−ia(λ+λ2−1).X=0(d’Alembert),Y=ρVk.Yes: a vertical lift of magnitude ρV∣k∣,upward for k>0,downward for k<0.We post more of this — worked solutions, CSAT trap breakdowns, guide chapters — a few times a week on Telegram. Free, no sign-in.Join→
This solution is part of the Maths Coverage Map — 14 years, mapped. Get the take-away PDF free.
You're subscribed — new solutions, tools, and guides will land in your inbox.