1 partial di erential equations(p.d.e.) · 2019-08-15 · 1 partial di erential equations(p.d.e.)...
TRANSCRIPT
1 Partial Differential Equations(P.D.E.)
Formation of partial differential equations - Lagrange’s Linearequation Solution of standard types of first order partialdifferential equations - Linear partial differential equations ofsecond and higher order with constant coefficients.
1.1 Introduction
In a differential equation if there are two or more independent variablesand the derivatives are partial derivatives then it is called a partialdifferential equation .
Examples:
1.
(∂2z
∂x2
)2
+
(∂2z
∂y2
)3
= xy (z− dependent variable; x and y−
independent variables)
2. x∂z
∂x+ y
∂z
∂y+ t
∂z
∂t= xyt (z− dep. variable; x, y and t− independent
variables)
The order of a partial differential equation if the order of the highestpartial derivative occurring in the equation.
The degree of a partial differential equation is the greatest exponent ofthe highest order.
In the above, e.g.1 is a second order & third degree equation, e.g.2 is afirst order equation & first degree equation.
Note : Throughout this chapter we use the following notations; z will betaken as a dependent variable which depends on two independent variablesx and y so that z = f(x, y).We write
∂z
∂x= p,
∂z
∂y= q,
∂2z
∂x2= r,
∂2z
∂x∂y=
∂2z
∂y∂x= s and
∂2z
∂y2= t.
Thus p2 + q2 = x2 + y2 is a partial differential equation of order 1 andr + xt = s2 − p is a partial differential equation of order 2.
1
2 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
1.2 Formation of Partial Differential Equation by Elimination ofArbitrary Constants
Let f(x, y, z, a, b) = 0. (1)be an equation which contains two arbitrary constants ‘a’ and ‘b’.
Partially differentiating (1) with respect to ‘x’ and ‘y’ we get two moreequations.
Eliminating a and b from these three equations, we get φ(x, y, z, p, q) = 0which is a partial differential equation of order 1.
In this case the number of arbitrary constants to be eliminated is equalto the number of independent variables and we obtain a first order partialdifferential equation. If the number of arbitrary constants to be eliminatedis more than the number of independent variables, we get partial differentialequations of second or higher order.
1.2.1 Examples of Formation of P.D.E. by Elimination of Arbitrary Constants
Example 1.1. Form the partial differential equation by eliminating thearbitrary constants from z = ax+ by + a2 + b2.
Solution: Given z = ax+ by + a2 + b2 (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= a i.e., p = a (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= b i.e., q = b (3)
From (2) and (3)
a = p and b = q
Substituting in (1), we get z = px+ qy + p2 + q2.
Example 1.2. Eliminating the arbitrary constants a and b from z = a2x+b2y + ab.
Solution: Given z = a2x+ b2y + ab. (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= a2 i.e., p = a2 (2)
Differentiating (1) partially w.r.t ‘y’
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 3
∂z
∂y= b2 i.e., q = b2 (3)
From (2) and (3)
a2 = p and b2 = q
Substituting in (1), we get z = px+ qy +√pq.
Example 1.3. Form the PDE by eliminating the arbitrary constants a andb from z = (x+ a)(y + b).
Solution: Given z = (x+ a)(y + b) (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= y + b i.e., p = y + b (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= x+ a i.e., q = x+ a (3)
From (2) and (3)
x+ a = q and y + b = p
Substituting in (1), we get z = pq.
Example 1.4. Form the PDE by eliminating a, b from u = (x2+ a2)(y2+b2).
Solution: Given u = (x2 + a2)(y2 + b2) (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= 2x(y2 + b2) i.e., p = 2x(y2 + b2) (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= 2y(x2 + y2) i.e., q = 2y(x2 + y2) (3)
From (2) and (3)
x2 + a2 =q
2yand y2 + b2 =
p
2x
Substituting in (1), we get 4xyu = pq.
Example 1.5. Find the differential equation of all sphere whose centreslie on the z − axis. [UQ]
4 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Solution: Any point on the z− axis is of the form (0, 0, c). The equationof the sphere with centre (0, 0, c) and radius ‘a’ is given by
(x− 0)2 + (y − 0)2 + (z − c)2 = a2 (1)
Differentiating (1) partially w.r.t ‘x’
2x+ 2(z − c)∂z
∂x= 0 i.e., x+ p(z − c) = 0 (2)
Differentiating (1) partially w.r.t ‘y’
2y + 2(z − c)∂z
∂y= 0 i.e., y + q(z − c) = 0 (3)
From (2) and (3)
z − c =−xp
and z − c =−yq
⇒ −xp
=−yq
∴ qx = py.
Example 1.6. Form the PDE from x2 + y2 + (z − c)2 = a2 [UQ]
Solution: Similar to above Example.
Example 1.7. Find the differential equation of all spheres of fixed radiushaving their centres in the xy − plane. [UQ]
Solution: Any point in the xy−plane is of the form (a, b, 0). The equationof the sphere with center (a, b, 0) and radius r is given by (x− a)2 + (y −b)2 + (z − 0)2 = r2. i.e., (x− a)2 + (y − b)2 + z2 = r2. (1)Here a, b are the two arbitrary constants and r is a given constant.Differentiating (1) partially w.r.t ‘x’
2(x− a) + 2z∂z
∂x= 0 i.e., x− a+ pz = 0 (2)
Differentiating (1) partially w.r.t ‘y’
2(y − b) + 2z∂z
∂y= 0
∂z
∂y= 2y(x2 + y2) i.e., y − b+ qz = 0 (3)
From (2) and (3)
x− a = −pz and y − b = −qz
Substituting in (1), we get z2(p2 + q2 + 1) = r2.
Example 1.8. Form the partial differential equation by eliminating thearbitrary constants a and b from (x− a)2 + (y − b)2 + z2 = 1.
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 5
Solution: Similar to above Example.
Example 1.9. Form the partial differential equation by eliminating thearbitrary constants a and b from log(az − 1) = x+ ay + b.
Solution: Given log(az − 1) = x+ ay + b. (1)Differentiating (1) partially w.r.t ‘x’
1
az − 1.a∂z
∂x= 1 i.e.,
ap
az − 1= 1 (2)
Differentiating (1) partially w.r.t ‘y’
1
az − 1.a∂z
∂y= a i.e.,
q
az − 1= 1 (3)
From (2) ap = az − 1 ⇒ a =1
z − p
From (3) q = az − 1 ⇒ z
z − p− 1
(∵ a =
1
z − p
)(z − p)q = z − z + p
zq = p(1 + q)
Example 1.10. Form the partial differential equation by eliminating a andb from (x− a)2 + (y − b)2 = z2cot2α , where is a constant. [UQ]
Solution: Given (x− a)2 + (y − b)2 = z2cot2α (1)Differentiating (1) partially w.r.t ‘x’
2(x− a) = 2z∂z
∂xcot2α i.e., x− a = pzcot2α (2)
Differentiating (1) partially w.r.t ‘y’
2(y − b) = 2z∂z
∂ycot2α i.e., y − b = qzcot2α (3)
Sub. (2) and (3) in (1), we get
p2z2cot4α + q2z2cot4α = z2cot2α
cot2α(p2 + q2) = 1
p2 + q2 = tan2α
Example 1.11. Form the partial differential equation by eliminating the
arbitrary constants from 2z =x2
a2+y2
b2
Solution: Given 2z =x2
a2+y2
b2(1)
Differentiating (1) partially w.r.t ‘x’
6 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
2∂z
∂x=
2x
a2i.e., p =
x
a2(2)
Differentiating (1) partially w.r.t ‘y’
2∂z
∂y=
2y
b2i.e., q =
x
b2(3)
From (2) and (3), we get a2 =x
pand b2 =
y
qSubstituting in (1), we get
2z = p2 + q2
Example 1.12. Form the partial differential equation by eliminating the
arbitrary constants a and b from z = axey +1
2a2e2y + b.
Solution: Given z = axey +1
2a2e2y + b (1)
Differentiating (1) partially w.r.t ‘x’
∂z
∂x= aey i.e., p = aey (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= axey +
1
2a2e2y.2 i.e., q = axey + a2e2y (3)
Sub. (2) in (3), we get
q = px+ p2
Example 1.13. Form the partial differential equation by eliminating the
arbitrary constants a and b from z = xy + y√x2 + a2 + b
Solution: Given z = xy + y√x2 + a2 + b (1)
Differentiating (1) partially w.r.t ‘x’
∂z
∂x= y + y
2x
2√x2 + a2
i.e., p = y +xy√x2 + a2
(2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= x+
√x2 + a2 i.e., q = x+
√x2 + a2 (3)
From (3), we get√x2 + a2 = q − x
Substituting in (2), we get
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 7
p = y +xy
q − x
p =qy
q − x
i.e., pq − px = qy
i.e., px+ qy = pq
Example 1.14. Form the partial differential equation of all spheres whoseradii are the same.
Solution: The equation of all sphere with equal radius can be taken as(x− a)2 + (y − b)2 + (z − c)2 = r2 (1)where a, b, c are arbitrary constants and r is a given constant. Since thenumber of arbitrary constants is more than the number of independentvariables, we will get the p.d.e. of order greater than 1.Differentiating (1) partially w.r.t ‘x’
2(x− a) + 2(z − c)∂z
∂x= 0 i.e., i.e., (x− a) + (z − c)p = 0 (2)
Differentiating (1) partially w.r.t ‘y’
2(y − b) + 2(z − c)∂z
∂y= 0 i.e., (y − b) + (z − c)q = 0 (3)
Differentiating (2) partially w.r.t ‘x’
1 + (z − c)∂2z
∂x2+∂z
∂x.∂z
∂x= 0 i.e., 1 + (z − c)r + p2 = 0 (4)
Differentiating (2) partially w.r.t ‘y’
1 + (z − c)∂2z
∂y2+∂z
∂y.∂z
∂y= 0 i.e., 1 + (z − c)t+ q2 = 0 (5)
From (4) and (5), we get z − c =−p2 − 1
rand z − c =
−q2 − 1
t
i.e., z − c =−p2 − 1
r=
−q2 − 1
tt(p2 + 1) = r(q2 + 1)
Example 1.15. Obtain the partial differential equation by eliminating
a, b, c fromx2
a2+y2
b2+z2
c2= 1
Solution: Since the number of arbitrary constants is more than thenumber of independent variables, we will get the p.d.e of order greaterthan ’1’.
Givenx2
a2+y2
b2+z2
c2= 1 (1)
Differentiating (1) partially w.r.t ‘x’
8 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
2x
a2+
2z
c2∂z
∂x= 0 i.e.,
x
a2+z
c2p = 0 (2)
Differentiating (1) partially w.r.t ‘y’
2y
b2+
2z
c2∂z
∂y= 0 i.e.,
y
b2+z
c2q = 0 (3)
Differentiating (2) partially w.r.t. ‘x’1
a2+
1
c2
(z∂2z
∂x2+∂z
∂x.∂z
∂x
)= 0 i.e.,
1
a2+
1
c2(zr2 + p2
)= 0 (4)
Differentiating (3) partially w.r.t ‘y’1
b2+
1
c2
(z∂2z
∂y2+∂z
∂y.∂z
∂y
)= 0 i.e.,
1
b2+
1
c2(zt2 + q2
)= 0 (5)
Differentiating (2) partially w.r.t ‘y’
0 +1
c2
(z∂2z
∂y∂x+∂z
∂x
∂z
∂y
)= 0 i.e.,
1
c2(zs+ pq) = 0 (5)
i.e., zs+ pq = 0
Note : Wemay also get different partial differential equations. The answeris not unique.
1.3 Formation of partial differential equation by elimination ofarbitrary Functions
Note: The elimination of one arbitrary function from a given relationgives a partial differential equation of first order while elimination of twoarbitrary function from a given relation gives a second or higher orderpartial differential equation.
1.3.1 Examples of Formation of P.D.E. by Elimination of Arbitrary Functions
Example 1.16. Eliminate f from z = f(x2 − y2).
Solution: Given z = f(x2 − y2) (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= f ′(x2 − y2).2x i.e., p = 2xf ′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= f ′(x2 − y2)(−2y) i.e., q = −2yf ′ (3)
From (2) and (3), we get f ′ =p
2xand f ′ =
−q2y
p
2x=
−q2y
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 9
i.e., py + qx = 0
Example 1.17. Eliminate f from z = xy + f(x2 + y2).
Solution: Given z = xy + f(x2 + y2). (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= y + f ′(x2 + y2).2x i.e., p = y + 2xf ′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= x+ f ′(x2 + y2).2y i.e., q = x+ 2yf ′ (3)
From (2) and (3),we get f ′ =p− y
2xand f ′ =
q − x
2y
p− y
2x=q − x
2y
i.e., py − qx = y2 − x2
Example 1.18. Eliminate f from z = x+ y + f(xy).
Solution: Given z = x+ y + f(xy) (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= 1 + f ′(xy).y i.e., p = 1 + yf ′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= 1 + f ′(xy).x i.e., q = 1 + xf ′ (3)
From (2) and (3), we get f ′ =p− 1
yand f ′ =
q − 1
x
p− 1
y=q − 1
x
i.e., px− qy = x− y
Example 1.19. Obtain PDE by eliminating the arbitrary function f fromxyz = φ(x+ y + z). [UQ]
Solution: Given xyz = φ(x+ y + z) (1)Differentiating (1) partially w.r.t ‘x’
y
(x∂z
∂x+ z
)= φ′(x+ y + z)
(1 +
∂z
∂x
)i.e., y(xp+ z) = (1 + p)φ′ (2)
Differentiating (1) partially w.r.t ‘y’
x
(y∂z
∂y+ z
)= φ′(x+ y + z)
(1 +
∂z
∂y
)i.e., x(yq + z) = (1 + q)φ′ (3)
From (2) and (3), we get
10 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
φ′ =y(xp+ z)
1 + pand φ′ =
x(yq + z)
1 + q
y(xp+ z)
1 + p=x(yq + z)
1 + qi.e., y(xp+ z)(1 + q) = x(yq + z)(1 + p).
Example 1.20. Eliminate the arbitrary function f from z = f(yx
).
Solution: Given z = f(yx
)(1)
Differentiating (1) partially w.r.t ‘x’
∂z
∂x= f ′
(yx
).
(−yx2
)i.e., p =
−yx2f ′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= f ′
(yx
).1
xi.e., q =
1
xf ′ (3)
From (2) and (3),we get f ′ =−px2
yand f ′ = qx
−px2
y= qx
i.e., px+ qy = 0
Example 1.21. Form a partial differential equation by eliminating f from
z = xf
(x
y
). [UQ]
Solution: Given z = xf
(x
y
)(1)
Differentiating (1) partially w.r.t ‘x’
∂z
∂x= xf ′
(x
y
).1
y+ f
(x
y
)i.e., p =
x
yf ′ + f (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= xf ′
(x
y
).
(−xy2
)i.e., q =
−x2
y2f ′ (3)
From (1), f =z
xFrom (3), f ′ =
−qy2
x2
Substituting these in (2), we get p =x
y
(−qy2
x2
)+z
x⇒ p = −qy
x+z
x
i.e., px+ qy = z
Example 1.22. Form PDE by eliminating arbitrary function f from
z = f(xyz
). [UQ]
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 11
Solution: Given z = f(xyz
)(1)
Differentiating (1) partially w.r.t ‘x’
∂z
∂x= f ′
(xyz
)(zy − xy ∂z∂x
z2
)i.e., p =
(zy − xyp
z2
)f ′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= f ′
(xyz
)(zx− xy ∂z∂y
z2
)i.e., q =
(zx− xyq
z2
)f ′ (3)
From (2) and (3), we get
f ′ =pz2
yz − xypand f ′ =
qz2
xz − xyp
pz2
yz − xyp=
qz2
xz − xyppxz − pqxy = qyz − pqxy
i.e., px = qy.
Example 1.23. Form a partial differential equation by eliminating f from
z = x2 + 2f
(1
y+ logx
).
Solution: Given z = x2 + 2f
(1
y+ logx
)(1)
Differentiating (1) partially w.r.t ‘x’
∂z
∂x= 2x+ 2f ′
(1
y+ log x
).1
xi.e., p = 2x+
2
xf ′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= 2f ′
(1
y+ log x
)(−1
y2
)i.e., q =
−2
y2f ′ (3)
From (2) and (3), we get
f ′ =x
2(p− 2x) and f ′
−qy2
2
x
2(p− 2x) =
−qy2
2i.e., px− 2x2 = −qy2.i.e., px+ qy2 = 2x2.
Example 1.24. Form PDE by eliminating arbitrary function f and g fromz = f(x+ ay) + g(x− ay).
Solution: Given z = f(x+ ay) + g(x− ay) (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= f ′ (x+ ay) + g′ (x− ay) i.e., p = f ′ + g′ (2)
12 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= af ′(x+ ay)− ag′(x− ay) i.e., q = af ′ − ag′ (3)
Differentiating (2) partially w.r.t ‘x’
∂2z
∂x2= f ′′ + g′′ i.e., r = f ′′ + g′′ (4)
Differentiating (3) partially w.r.t ‘y’
∂2z
∂y2= a2f ′′ + a2g′′ i.e., t = a2 (f ′′ + g′′) (5)
Using (4), we get t = a2r.
Example 1.25. Form partial differential equation by eliminating arbitraryfunction f and g from z = f(2x+ y) + g(3x− y). [UQ]
Solution: Given z = f(2x+ y) + g(3x− y) (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= 2f ′ (2x+ y) + 3g′ (3x− y) i.e., p = 2f ′ + 3g′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= f ′(2x+ y)− g′(3x− y) i.e., q = f ′ − g′ (3)
Differentiating (2) partially w.r.t ‘x’
∂2z
∂x2= 4f ′′(2x+ y) + 9g′′(3x− y) i.e., r = 4f ′′ + 9g′′ (4)
Differentiating (3) partially w.r.t ‘y’
∂2z
∂y2= f ′′(2x+ y) + g′′(3x− y) i.e., t = f ′′ + g′′ (5)
Differentiating (2) partially w.r.t ‘y’
∂2z
∂x∂y= 2f ′′(2x+ y)− 3g′′(3x− y) i.e., s = 2f ′′ − 3g′′ (6)
Solving (4) and (5), we get f ′′ =1
5(qt− r) and g′′ =
1
5(r − 4t)
Sub. the values of f ′′ and g′′ in (6), we get
s =2
5(9t− r)− 3
5(r − 4t)
5s = 30t− 5r
s = 6t− r
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 13
Example 1.26. Form partial differential equation by eliminating arbitraryfunction f and g from z = f(x+ 2y) + g(2x− y).
Solution: Given z = f(x+ 2y) + g(2x− y) (1)Differentiating (1) partially w.r.t ‘x’
p = f ′ + 2g′ (2)
Differentiating (1) partially w.r.t ‘y’
q = 2f ′ − g′ (3)
Differentiating (2) partially w.r.t ‘x’ : r = f ′′ + 4g′′ (4)Differentiating (3) partially w.r.t ‘y’ : t = 4f ′′ + g′′ (5)Differentiating (2) partially w.r.t ‘y’ : s = 2f ′′ − 2g′′ (6)
Solving (4) and (5), we get f ′′ =4t− r
15, g′′ =
4r − t
15Sub. the values of f ′′ and g′′ in (6), we get
s =2
15(4t− r)− 2
15(4r − t)
15s = 10t− 10r
3s+ 2r = 2t
Example 1.27. Eliminate the arbitrary function φ and fromz = xφ(y) + y(x) and form the partial differential equation.
Solution: Given z = xφ(y) + y(x) (1)Differentiating (1) partially w.r.t ‘x’
p = φ(y) + yϕ′(x) (2)
Differentiating (1) partially w.r.t ‘y’
q = xφ′(y) + ϕ(x) (3)
Differentiating (2) partially w.r.t ‘y’
s = φ′(y) + ϕ′(x) (4)
Now, px+ qy = xφ(y) + xyϕ′(x) + xyφ′(y) + yϕ(x)
= xφ(y) + yϕ(x) + xy(φ′(y) + ϕ′(x))
px+ qy = z + xys [using (1) and (4)]
Example 1.28. Form partial differential equation by eliminating arbitraryfunction f and g from z = xf(3x+ 2y) + g(3x+ 2y).
14 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Solution: Given z = xf(3x+ 2y) + g(3x+ 2y) (1)Differentiating (1) partially w.r.t ‘x’
p = 3xf ′ + f + 3g′ (2)
Differentiating (1) partially w.r.t ‘y’
q = 2xf ′ + 2g′ (3)
Differentiating (2) partially w.r.t ‘x’
r = 3(x.3f ′′ + f ′) + 3f ′ + 9g′′r = 9xf ′′ + 9g′′ + 6f ′ (4)
Differentiating (3) partially w.r.t ‘y’ : t = 4xf ′′ + 4g′′ (5)Differentiating (2) partially w.r.t ‘y’ : s = 6xf ′′ + 2f ′ + 6g′′ (6)(4)− 3× (6) ⇒ r − 3s = 9xf ′′ + 9g′′ + 6f ′ − 18xf ′′ − 6f ′ − 18g′′
= −9xf ′′ − 9g′′
r − 3s = −9(xf ′′ + g′′) = −9
(t
4
)[using(5)]
4r − 12s+ 9t = 0
Example 1.29. Form partial differential equation by eliminating arbitrary
function f and g from z = xf(yx
)+ yφ(x). [UQ]
Solution: Given z = xf(yx
)+ yφ(x) (1)
Differentiating (1) partially w.r.t ‘x’
∂z
∂x= xf ′
(yx
)(−yx2
)+ f
(yx
)+ yφ′(x)
∂z
∂x=
−yxf ′(yx
)+ f
(yx
)+ yφ′(x)
i.e., p =−yxf ′ + f + yφ′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= xf ′
(yx
).1
x+ φ(x) ⇒ ∂z
∂y= f ′
(yx
)+ φ(x)
i.e., q = f ′ + φ (3)
Now,∂2z
∂x∂y= f ′′
(yx
).
(−yx2
)+ φ′(x) i.e., s =
−yx2f ′′ + φ′ (4)
∂2z
∂y2= f ′′
(yx
).1
xi.e., t =
1
xf ′′ (5)
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 15
∴ px+ qy = −yf ′ + xf + xyφ′ + yf ′ + yφ [using (2) and (3)]
px+ qy = xf + yφ+ xyφ′
From (1), px+ qy = z + xyφ′ (6)
Using (5) in (4), s =−yx2
(xt) + φ′ ⇒ φ′ = s+yt
xSub. the value of φ′ in (6), we get
px+ qy = z + xy
(s+
yt
x
)px+ qy = z + xys+ y2t
Example 1.30. Form a p.d.e by eliminating arbitrary function f and gfrom z = f(x+ y)g(x− y).
Solution: Given z = f(x+ y)g(x− y) i.e., z = fg (1)Differentiating (1) partially w.r.t ‘x’
∂z
∂x= fg′ + gf ′ i.e., p = fg′ + gf ′ (2)
Differentiating (1) partially w.r.t ‘y’
∂z
∂y= f (−g′) + gf ′ i.e., q = gf ′ − fg′ (3)
∂2z
∂x2= fg′′ + g′f ′ + gf ′′ + f ′g′ i.e., r = fg′′ + gf ′′ + 2f ′g′ (4)
∂2z
∂y2= −[f(−g′′) + g′f ′] + gf ′′ + f ′(−g′)
t = fg′′ + gf ′′ − 2f ′g′ (5)
From (2) and (3), we get
p2 − q2 = f 2g′2 + g2f ′2 + 2fgf ′g′ − (g2f ′2 + f 2g′2 − 2fgf ′g′)
= 4fgf ′g′
p2 − q2 = 4zf ′g′ [using(1)] (6)
From (4) and (5), we get
r − t = 4f ′g′ (7)
From (6) and (7), we get
p2 − q2 = z(r − t)
16 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
1.3.2 Formation of partial differential equation by elimination of arbitrary functionφ from φ(u, v) = 0, where u and v are function of x, y and z.
Let (u, v) = 0 (1)Differentiating (1) partially w.r.t x and y, we get
∂φ
∂u
∂u
∂x+∂φ
∂v
∂v
∂x= 0 (2)
∂φ
∂u
∂u
∂y+∂φ
∂v
∂v
∂y= 0 (3)
To eliminate φ, it is enough to eliminate∂φ
∂uand
∂φ
∂v
From (2) and (3), eliminating∂φ
∂uand
∂φ
∂vFrom (2) and (3), we get ∣∣∣∣∣∣∣
∂u
∂x
∂v
∂x∂u
∂y
∂v
∂y
∣∣∣∣∣∣∣ = 0
Example 1.31. Form the partial differential equation by eliminating thearbitrary function φ from the relation φ(x2 + y2 + z2, xyz) = 0
Solution: Given φ(x2 + y2 + z2, xyz) = 0 (1)Let u = x2 + y2 + z2, v = xyz
Equation (1) becomes φ(u, v) = 0 (2)Eliminating φ from (2), we get∣∣∣∣∣∣∣
∂u
∂x
∂v
∂x∂u
∂y
∂v
∂y
∣∣∣∣∣∣∣ = 0 (3)
Here,∂u
∂x= 2x+ 2zp,
∂u
∂y= 2y + 2zq
∂v
∂x= xyp+ yz,
∂v
∂y= xyq + xz
From (3),
∣∣∣∣ 2x+ 2zp xyp+ yz2y + 2zq xyq + xz
∣∣∣∣ = 0
i.e., (2x+ 2zp)(xyq + xz)− (xyp+ yz)(2y + 2zq) = 0
i.e., pxz2 − y2 − qy(z2 − y2) = z(y2 − x2)
Example 1.32. Form the partial differential equation by eliminating thearbitrary function from f(x+ y + z, xy + z2) = 0. [UQ]
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 17
Solution: Given f(x+ y + z, xy + z2) = 0 (1)Let u = x+ y + z, v = xy + z2
Equation (1) becomes φ(u, v) = 0 (2) Eliminating φ from (2), we get∣∣∣∣∣∣∣∂u
∂x
∂v
∂x∂u
∂y
∂v
∂y
∣∣∣∣∣∣∣ = 0 (3)
Here,∂u
∂x= 1 + p,
∂u
∂y= 1 + q
∂v
∂x= y + 2zp,
∂v
∂y= x+ 2zq
From(3), ∣∣∣∣ 1 + p y + 2zp1 + q y + 2zq
∣∣∣∣ = 0
i.e., (1 + p)(x+ 2zq)− (1 + q)(y + 2zp) = 0
i.e., (x− 2z)p− (y − 2z)q = y − x
Example 1.33. Form the partial differential equation by eliminating the
arbitrary function g from the relation g(yx, x2 + y2 + z2
)= 0
Solution: Given g(yx, x2 + y2 + z2
)= 0 (1)
Let u =y
x, v = x2 + y2 + z2
Equation (1) becomes g(u, v) = 0 (2)Eliminating φ from (2), we get∣∣∣∣∣∣∣
∂u
∂x
∂v
∂x∂u
∂y
∂v
∂y
∣∣∣∣∣∣∣ = 0 (3)
∂u
∂x=
−yx2,∂u
∂y=
1
x∂v
∂x= 2x+ 2zp
∂v
∂y= 2y + 2zq
From (3),
∣∣∣∣∣∣∣−yx2
2x+ 2zp
1
x2y + 2zq
∣∣∣∣∣∣∣ = 0
i.e.,−yx2
(2y + 2zq)− 1
x(2x+ 2zp) = 0
i.e., xzp+ yzq + x2 + y2 = 0
18 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
1.3.3 Exercises & Solutions
Find the partial differential equations by eliminating the arbitraryconstants a, b&c as the case may be:
1. z = ax+ by [Ans : z = px+ qy]
2. z = a sin by [Ans: r + t = 0]
3. z = axy + b [Ans: px− qy = 0]
4. z = ax+ a2y2 + b [Ans: q = 2yp2]
5. z = ax4 + by4 [Ans: px+ qy = 4z]
6. ax+ by + cz = 1 [Ans: r = 0 (or) s = 0 (or) t = 0]
7. z = ax+ (1− a)y + b [Ans: p+ q = 1]
8. z = (x− a)2 + (y − b)2 + 1 [Ans: z =(p2
)2+(q2
)2+ 1]
Find the partial differential equations by eliminating the arbitraryconstants from the following:
1. z = f(x2 − y2) [Ans: py + qx = 0]
2. z = f(x+ iy) + g(x− iy) [Ans: r + t = 0]
3. xy + yz + zx = f(z
x+ y)
[Ans: p(x+ y)(x+ 2z)− q(x+ y)(y + 2z) = z(x− y)]
4. z = xf(x+ t) + g(x+ t) [Ans: r = 2s− t]
5. z = x2f(y) + y2g(x) [Ans: xyr = 2(px+ qy − 2z)]
6. z = f(x− y) [Ans: p+ q = 0]
7. z = y2 + 2f(+logy) [Ans: px2 + qy = 2y2]
8. z = x+ yf(x2 − y2) [Ans: z = xq + yp]
9. f(xy + z2, x+ y + z) = 0 [Ans: (2z − x)p− (2z − y)q = x− y]
10. φ = (z2 − xy, xz ) = 0 [Ans: px2 − q(xy − 2z2) = zx]
11. z = f(ax+ by) + g(αx+ βy) [Ans: bβr − (aβ + bα)s+ aαt = 0]
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 19
1.4 Lagrange’s Linear Equations
The equation of the form Pp+Qq = R is known as Lagrange’s equation,where P,Q and R are functions of x, y and z.
To solve this equation it is enough to solve the subsidiary equations[(or) Lagrange’s auxiliary equations]
dx
P=dy
Q=dz
Z(1)
If u = a and v = b are two solutions of (1) then the solution of the givenLagrange’s equation is φ(u, v) = 0. Generally, the subsidiary equationscan be solved in two ways1. Method of grouping 2. Method of multipliers
1.4.1 Method of grouping
In the subsidiary equations
dx
P=dy
Q=dz
Z
if the variables can be separated in any pair of equations, then we get asolution of the form u = a and v = b.
20 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
1.4.2 Method of Multipliers
1. Single Multiplier (l,m, n):Choose any three multipliers l,m, n may be constants or functions ofx, y, z we have
dx
P=dy
Q=dz
R=ldx+mdy + ndz
lP +mQ+ nRIf it is possible to choose l,m, n such that lP +mQ+ nR = 0 then
ldx+mdy + ndz = 0.
If ldx + mdy + ndz is a perfect differential of some function, sayu(x, y, z), then du = 0∴ u = a is a solution.
* Note : If lP +mQ + nR 6= 0 for any (l,m, n), then go for DoubleMultipliers as follows
2. Double Multipliers (l1,m1, n1), (l2,m2, n2):Choose Double Multipliers (l1,m1, n1), (l2,m2, n2) may be constants orfunctions of x, y, z we have
dx
P=dy
Q=dz
R=l1dx+m1dy + n1dz
l1P +m1Q+ n1R=l2dx+m2dy + n2dz
l2P +m2Q+ n2RIntegrating both sides of the equation,
l1dx+m1dy + n1dz
l1P +m1Q+ n1R=l2dx+m2dy + n2dz
l2P +m2Q+ n2Rwe get u = 0∴ u = a is a solution.
1.4.3 Working Rule to solve the equation Pp + Qq = R
(i) Form the subsidiary equationsdx
P=dy
Q=dz
R
(ii) Solve the subsidiary equations by the method of grouping or themethod of multipliers or both to get two independent solutions u = a
and v = b.
(iii) The general solution is φ(u, v) = 0
1.5 Examples of Lagrange’s Linear Equations
1.5.1 Examples of Lagrange’s Linear Equations by Method of grouping
Example 1.34. Solve px+ qy = z.
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 21
Solution: Given px+ qy = z
This equation is of the form Pp+Qq = R, then P = x,Q = y and R = z.The subsidiary equations are
dx
P=dy
Q=dz
R
Takingdx
x=dy
yIntegrating, we get logx = logy + loga
i.e.,x
y= a
Similarly fromdy
y=dz
z, we get
y
z= b
Hence the general solution is φ(x
y,y
z) = 0
Example 1.35. Find the general solution of px+ qy = 0. [UQ]
Solution: Given px+ qy = 0This equation is of the form Pp+Qq = R, then P = x,Q = y and R = 0.The subsidiary equations are
dx
x=dy
y=dz
0
Takingdx
x=dy
yIntegrating, we get logx = logy + loga
i.e.,x
y= a
Takingdy
y=dz
0, we get
dz = 0
Integrating, we get z = b
Hence the general solution is φ(x
y, z) = 0.
Example 1.36. Solvey2z
xp+ xzq = y2. [UQ]
22 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Solution: Giveny2z
xp+ xzq = y2
i.e., y2zp+ x2zq = xy2
The subsidiary equations are
dx
y2z=
dy
x2z=
dz
xy2
Takingdx
y2z=
dy
x2z, we get
dx
y2=dy
x2
i.e., x2dx = y2dy
Integrating, we getx3
3=y3
3+ c1
i.e., x3 − y3 = a
Takingdx
y2z=
dz
xy2, we get
dx
z=dz
xi.e., xdx = zdz
Integrating, we getx2
2=z2
2+ c2
i.e., x2 − z2 = b
∴ The general solution is φ(x3 − y3, x2 − z2) = 0.
Example 1.37. Solve yzp+ xzq = xy. [UQ]
Solution: Given yzp+ xzq = xyThe subsidiary equations are
dx
yz=dy
xz=dz
xy
Takingdx
yz=dy
xz, we get
dx
y=dy
x
i.e., xdx = ydy
Integrating, we getx2
2=y2
2+ c1
i.e., x2 − y2 = a
Takingdx
yz=dz
xy, we get
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 23
xdx = zdz
Integrating, we getx2
2=z2
2+ c2
i.e., x2 − z2 = b
∴ The general solution is φ(x2 − y2, x2 − z2) = 0.
Example 1.38. Solve x2p+ y2q = z2.
Solution: Given x2p+ y2q = z2
The subsidiary equations are
dx
x2=dy
y2=dz
z2
Takingdx
x2=dy
y2
Integrating, we get −1
x= −1
y+ c1
i.e.,1
x− 1
y= a
Takingdx
x2=dz
z2Integrating, we get
−1
x= −1
z+ c2
i.e.,1
x− 1
z= b
∴ The general solution is φ(1
x− 1
y,1
x− 1
z) = 0.
Example 1.39. Solve pcotx+ qcoty = cotz.
Solution: Given pcotx+ qcoty = cotzThe subsidiary equations are
dx
cotx=
dy
cot y=
dz
cot z
Takingdx
cotx=
dy
cot y, we have
tanxdx = tan ydyIntegrating, logsecx = logsecy + loga
i.e.,secx
sec y= a
Similarly takingsec y
sec z= b
∴ The general solution is φ
(secx
sec y,sec y
sec z
)= 0.
24 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
1.5.2 Examples of Lagrange’s Linear Equations by Method of multipliers
Example 1.40. Solve (y − z)p+ (z − x)q = x− y.
Solution: Given (y − z)p+ (z − x)q = x− y
The subsidiary equations aredx
y − z=
dy
z − x=
dz
x− y(1)
Each fraction of (1) =dx+ dy + dz
y − z + z − x+ x− y=dx+ dy + dz
0
⇒ dx+ dy + dz = 0Integrating, x+ y + z = aUsing x, y, z as multipliers,
each fraction of (1) =xdx+ ydy + zdz
x (y−z)+y (z−x)+z (x−y)=xdx+ ydy + zdz
0
⇒ xdx+ ydy + zdz = 0
Integrating,x2
2+y2
2+z2
2= c
i.e., x2 + y2 + z2 = b
∴ The general solution is f(x+ y + z, x2 + y2 + z2
)= 0.
Example 1.41. Solve (mz − ny)∂z
∂x+ (nx− lz)
∂z
∂y= ly −mx. [UQ]
Solution: Given (mz − ny) p+ (nx− lz) q = ly −mx
The subsidiary equations aredx
mz − ny=
dy
nx− lz=
dz
ly −mx(1)
Using x, y, z as multipliers, each fraction of (1)=
xdx+ ydy + zdz
x (mz − ny) + y (nx− lz) + z (ly −mx)=xdx+ ydy + zdz
0
⇒ xdx+ ydy + zdz = 0
Integrating, we getx2
2+y2
2+z2
2= c1
i.e., x2 + y2 + z2 = aUsing l,m, n as multipliers, each fraction of (1) =
ldx+mdy + ndz
l (mz − ny) +m (nx− lz) + n (ly −mx)=ldx+mdy + ndz
0
ldx+mdy + ndz = 0
Integrating, lx+my + nz = b∴ The general solution is φ(x2 + y2 + z2, lx+my + nz) = 0.
Example 1.42. Solve x(y2 − z2
)p+ y
(z2 − x2
)q = z
(x2 − y2
). [UQ]
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 25
Solution: Given x(y2 − z2
)p+ y
(z2 − x2
)q = z
(x2 − y2
)The subsidiary equations are
dx
x (y2 − z2)=
dy
y (z2 − x2)=
dz
z (x2 − y2)
Using x, y, z as multipliers, each fraction of (1)=
xdx+ ydy + zdz
x2 (y2 − z2) + y2 (z2 − x2) + z2 (x2 − y2)=xdx+ ydy + zdz
0
⇒ xdx+ ydy + zdz = 0
Integrating, we getx2
2+y2
2+z2
2= c1
i.e., x2 + y2 + z2 = a
Using1
x,1
y,1
zas multipliers, each fraction of (1) =
lxdx+
1ydy +
1zdz
(y2 − z2) + (z2 − x2) + (x2 − y2)=
lxdx+
1ydy +
1zdz
0l
xdx+
1
ydy +
1
zdz = 0
Integrating, log x+ log y + log z = log b
i.e., xyz = b
∴ The general solution is φ(x2 + y2 + z2, xyz) = 0.
Example 1.43. Solve (y + z)p+ (x+ z)q = x+ y. [UQ]
Solution: Given (y + z)p+ (x+ z)q = x+ y
The subsidiary equations aredx
y + z=
dy
x+ z=
dz
x+ y(1)
Each fraction of (1) =dx+ dy + dz
2 (x+ y + z)=
dx− dy
− (x− y)=
dy − dz
− (y − z)
d(x− y)
(x− y)=d(y − z)
(y − z)Integrating, we get log(x− y) = log(y − z) + log a
i.e.,x− y
y − z= a
Takingdx+ dy + dz
2 (x+ y + z)=
dx− dy
− (x− y),we have
1
2
d(x+ y + z)
(x+ y + z)= −d(x− y)
(x− y)
26 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Integrating, we get1
2log(x+ y + z) = − log(x− y) + log b√
(x+ y + z)(x− y) = b
The general solution is φ
(x− y
y − z,√
(x+ y + z)(x− y)
)= 0.
Example 1.44. Solve (x2 − yz)p+ (y2 − zx)q = z2 − xy. [UQ]
Solution: The subsidiary equations are
dx
(x2 − yz)=
dy
(y2 − zx)=
dz
z2 − xy(1)
Each fraction of (1) =
dx+ dy + dz
(x2 − yz) + (y2 − zx) + (z2 − xy)=
dx+ dy + dz
x2 + y2 + z2 − xy − yz − zx
Using x, y, z as multipliers, each fraction of (1) =
xdx+ ydy + zdz
x3+y3+z3 − 3xyz=
dx+ dy + dz
x2+y2+z2 − xy − yz − zxxdx+ ydy + zdz
(x+ y + z) (x2 + y2 + z2 − xy − yz − zx)=
dx+ dy + dz
x2+y2+z2 − xy − yz − zxxdx+ ydy + zdz
x+ y + z=dx+ dy + dz
1
xdx+ ydy + zdz = (x+ y + z)d(x+ y + z)
Integrating,x2
2+y2
2+z2
2=
(x+ y + z)2
2+ c
xy + yz + zx = a
Each fraction of (1) =dx− dy
(x2 − yz)− (y2 − zx)=
dy − dz
(y2 − zx)− (z2 − xy)
i.e.,dx− dy
(x2 − y2) + z (x− y)=
dy − dz
(y2 − z2) + x (y − z)
d (x− y)
(x− y) (x+ y + z)=
d (y − z)
(y − z) (x+ y + z)
d (x− y)
(x− y)=d (y − z)
(y − z)Integrating, log(x− y) = log(y − z) + log b
x− y
y − z= b
∴ The general solution is φ
(xy + yz + zx,
x− y
y − z
)= 0.
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 27
Example 1.45. Solve x2(y − z)p+ y2(z − x)q = z2(x− y).
Solution: The subsidiary equations are
dx
x2 (y − z)=
dy
y2 (z − x)=
dz
z2 (x− y)(1)
Using1
x2,1
y2,1
z2as multipliers
Each fraction of (1) =
1x2dx+
1y2dy +
1z2dz
(y − z) + (z − x) + (x− y)=
1x2dx+
1y2dy +
1z2dz
0
i.e.,1
x2dx+
1
y2dy +
1
z2dz = 0
Integrating,1
x+
1
y+
1
z= a
Using1
x,1
y,1
zas multipliers
Each fraction of (1) =
1xdx+
1ydy +
1zdz
x (y − z) + y (z − x) + z (x− y)=
1xdx+
1ydy +
1zdz
0
i.e.,1
xdx+
1
ydy +
1
zdz = 0
Integrating, log x+ log y + log z = log b
xyz = b
∴ The general solution is φ
(1
x+
1
y+
1
z, xyz
)= 0.
Example 1.46. Solve y2p− xyq = x(z − 2y). [UQ]
Solution: The subsidiary equations are
p− xyq = x(z − 2y) (1)
Takingdx
y2=
dy
−xy, we have
dx
y=
dy
−xxdx = −ydy
Integrating,x2
2= −y
2
2+ c
i.e., x2 + y2 = a
Takingdy
−xy=
dz
x (z − 2y)we have
28 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
dy
−y=
dz
(z − 2y)
i.e., (z − 2y)dy = −ydzi.e., ydz + zdy − 2ydy = 0
i.e., d(yz)− 2ydy = 0
Integrating, yz − y2 = b∴ The general solution is φ
(x2 + y2, yz − y2
)= 0.
Example 1.47. Solve (x2 − y2 − z2)p+ 2xyq − 2xz = 0. [UQ]
Solution: The subsidiary equations aredx
x2 − y2 − z2=
dy
2xy=
dz
2xz(1)
Takingdy
2xy=
dz
2xz, we have
dy
y=dz
z
log y = log z + log a
i.e.,y
z= a
Using x, y, z as multipliersEach fraction of (1) =
xdx+ ydy + zdz
x (x2 − y2 − z2) + 2xy2 + 2xz2=xdx+ ydy + zdz
x3 + xy2 + xz2
=xdx+ ydy + zdz
x (x2 + y2 + z2)
Takingdy
2xy=xdx+ ydy + zdz
x (x2 + y2 + z2)
dy
y=
2 (xdx+ ydy + zdz)
(x2 + y2 + z2)
dy
y=d(x2 + y2 + z2
)(x2 + y2 + z2)
log y = log(x2 + y2 + z2) + logb
i.e.,y
x2 + y2 + z2= b
∴ The general solution is φ
(y
z,
y
x2 + y2 + z2
)= 0.
Example 1.48. Solve (x− 2z)p+ (2z − y)q = y − x. [UQ]
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 29
Solution: The subsidiary equations aredx
x− 2z=
dy
2z − y=
dz
y − x(1)
Each fraction of (1) =dx+ dy + dz
x− 2z + 2z− y + y − x=dx+ dy + dz
0
i.e., dx+ dy + dz = 0
i.e., x+ y + z = a
Each fraction of (1) =ydx+ xdy
xy − 2yz + 2xz− xy=ydx+ xdy
2z(x− y)
Takingydx+ xdy
2z (x− y)=
dz
y − x, we have
ydx+ xdy = −2zdz
i.e., d(xy) = −2zdz
Integrating, xy = −z2 + b
xy + z2 = b
The general solution is φ(x+ y + z, xy + z2
)= 0.
Example 1.49. Solve (z2 − 2yz − y2)p+ (xy + zx)q = xy − zx. [UQ]
Solution: The subsidiary equations are
dx
z2 − 2yz − y2=
dy
xy + zx=
dz
xy − zx(1)
Using x, y, z as multipliers each fraction of
(1) =xdx+ ydy + zdz
x (z2 − 2yz − y2) + y (xy + zx) + z (xy − zx)=xdx+ ydy + zdz
0we have
xdx+ ydy + zdz = 0
Integrating, we getx2
2+y2
2+z2
2= c1
x2 + y2 + z2 = a
Takingdy
xy + zx=
dz
xy − zx, we have
dy
y + z=
dz
y − z
i.e., (y − z)dy = (y + z)dz
ydy − zdy − ydz − zdz = 0
ydy − (ydz + zdy)− zdz = 0
ydy − d(yz)− zdz = 0
30 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Integrating,y2
2− yz − z2
2= c2
i.e., y2 − z2 − 2yz = b
The general solution is φ(x2 + y2 + z2, y2 − z2 − 2yz
)= 0.
Example 1.50. Solve pz − qz = z2 + (x+ y)2.
Solution: The subsidiary equations aredx
z=dy
−z=
dz
z2 + (x+ y)2(1)
Takingdx
z=dy
−z, we have dx = −dy
Integrating, x+ y = a
Takingdx
z=
dz
z2 + (x+ y)2, we have
dx
z=
dz
z2 + a2(∵ x+ y = a)
i.e., dx =zdz
z2 + a2
Integrating, x =1
2log(z2 + a2) + c
2x = log(z2 + a2) + 2c
i.e., 2x− log[z2 + (x+ y)2
]= b
The general solution is φ{x+ y, 2x− log
[z2 + (x+ y)2
]}= 0
Example 1.51. Solve p− q = log(x+ y).
Solution: The subsidiary equations aredx
1=dy
−1=
dzlog(x+y)
(1)
Takingdx
1=dy
−1, we have dx = −dy
Integrating, x+ y = a
Takingdx
1=
dz
log(x+ y), we have
dx =dz
log a(∵ x+ y = a)
i.e., log adx = dz
Integrating, (loga)x = z + b
i.e., [log(x+ y)]x = z + b
x log(x+ y)− z = bThe general solution is φ(x+ y, xlog(x+ y)− z) = 0.
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 31
1.5.3 Exercises & Solutions
Solve the following partial differential equations:
1. z(x− y) = x2p− y2q
[Ans : φ
(1
x+
1
y,
z
x+ y
)= 0
]2. (x2 + y2 + yz)p+ (x2 + y2 − xz)q = z(x+ y)[
Ans : φ(x− y − z,x2 + y2
z2) = 0
]3. (3z − 4y)
∂z
∂x+ (4x− 2z)
∂z
∂y= 2y − 3x[Ans : φ(x2 + y2 + z2, 2x+ 3y + 4z) = 0
]4. (y + z)p+ (z + x)q = x+ y[
Ans : φ
(y − z
x− y, (x− y)
√x+ y + z
)= 0
]5. (y2 + z2)p− xyq + xz = 0
[Ans : φ
(yz, x2 + y2 + z2
)= 0]
6. (y − z)p+ (x− y)q = (z − x)[Ans : (x2 + 2yz, x+ y + z) = 0
]7.y − z
yzp+
z − x
xyq =
x− y
xy[Ans : φ(x+ y + z, xyz) = 0]
8. p cotx+ q cot y = cot z
[Ans : φ
(cos z
cos y,cos y
cosx
)= 0
]9. (a− x)p+ (b− y)q = c− z
[Ans : φ
(a− x
b− y,b− y
c− z
)= 0
]1.6 Solution of Partial Differential Equations
A solution or integral of a partial differential equation is a relationbetween the independent and the dependent variables which satisfies thegiven differential equation.Note: There are two distinct types of solution for partial differentialequations, one type of solution containing arbitrary constants and theother type of solution containing arbitrary functions. Both these types fsolution may be given as solutions of the same partial differentialequation.Any solution of a partial differential equation in which the number of
arbitrary constants is equal to the number of independent variables iscalled the complete integral .
32 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Any solution obtained from the complete integral by giving particularvalues to the arbitrary constants is called a particular integral .Any solution obtained from the complete integral by eliminating
arbitrary constants is called a singular integral .Any solution which contains the maximum number of arbitrary
functions is called as a general integral .Let φ(x, y, z, a, b) = 0 be the complete integral of f(x, y, z, p, q) = 0.
Then the solution obtained by eliminating a and b from the equations
φ(x, y, z, a, b) = 0 (1)
∂φ
∂a= 0 (2)
∂φ
∂b= 0 (3)
is called the singular integral of the partial differential equation.In (1), we put b = g(a) we get
φ(x, y, z, a, g(a)) = 0 (4)
Differentiating (4) w.r.t. ‘a’, we get
∂φ
∂a+∂φ
∂gg′(a) = 0 (5)
The obtained by eliminating ‘a’ from (4) and (5) is called the generalintegral of φ(x, y, z, a, b) = 0
1.6.1 Solution of Partial Differential Equations by Direct Integration
A partial differential equation can be solved by successive integration inall cases where the dependent variable occurs only in the partial derivatives.We illustrate this method in the following Examples.
Example 1.52. Solve∂z
∂x= cos x.
Solution: If z is a function of x only then there is no difference between∂z
∂xand
dz
dx. Therefore to find z we can directly integrate
dz
dx= cos x and
hence the solution is z = sin x + A, where A is arbitrary constant. Buthere z is a function of two variables x and y.∴ Instead of the arbitrary constant A we can take an arbitrary function
of y say f(y).∴ The solution is z = sin x+ f(y).
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 33
Example 1.53. Solve∂2z
∂x∂y= sinx.
Solution: Given∂2z
∂x∂y= sinx. Integrating w.r.t ‘x‘
∂z
∂y= −cosx+ f(y),
where f(y) is an arbitrary function of y. Integrating again w.r.t ‘y’ Z =
−ycosx + F (y) + g(x) where F (y) =
∫f(y)dy is an arbitrary function of
y and g(x) is an arbitrary function of x.
Example 1.54. Solve∂2z
∂x∂y= 0.
Solution : Given∂2z
∂x∂y= 0 Integrating w.r.t ‘x‘
∂z
∂y= f(y), where f(y)
is an arbitrary function of y. Integrating again w.r.t ‘y‘ z = F (y) + g(x),where F (y) is an arbitrary function of y and g(x) is an arbitrary functionof x.
Example 1.55. Solve∂2u
∂y∂x= 4xsin(3xy).
Solution: Given∂2u
∂y∂x= 4xsin(3xy) Integrating w.r.t ‘y‘
∂u
∂x=
−4x cos(3xy)
3x+ f(x) =
−4
3cos(3xy) + f(x) Integrating again w.r.t
‘x‘ u =−4
3
sin(3xy)
3y+ F (x) + g(y) ie, u =
−4
9ysin(3xy) + F (x) + g(y),
where F (x) =
∫f(x)dx is an arbitrary function of x and g(y) is an
arbitrary function of y.
Example 1.56. Solve∂z
∂x= 2x+ 3y;
∂z
∂y= 3x+ cosy
Solution:∂z
∂x= 2x+ 3y Integrating w.r.t ‘x‘ z = x2 + 3xy + f(y) (1)
where f(y) is an arbitrary function of ‘y‘ Differentiating (1) partially
w.r.t ‘y‘∂z
∂y= 3x+ f ′(y) By hypothesis,
∂z
∂y= 3x+ cosy
∴ f(y) = siny + c
∴ from (1), z = x2 + 3xy + siny + c , where c is an arbitrary constant.
Example 1.57. Solve∂2z
∂x∂y= sinxsiny for which
∂z
∂y= −2siny, when
x = 0 and z = 0 when y is an odd multiple ofπ
2.
34 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Solution:∂2z
∂x∂y= sinxsiny
Integrating w.r.t ‘x‘∂z
∂y= −cosxsiny + f(y) (1)
Given∂z
∂y= −2siny when x = 0
Hence, −2siny = −siny + f(y)∴ f(y) = −sinyFrom (1),
∂z
∂y= −cosxsiny − siny
Integrating w.r.t ‘y‘z = cosxcosy + cosy + g(x) (2)
Given z = 0 when y = (2n+ 1)π
2
[yis an odd multiple of
π
2
]Hence g(x) = 0From(2), z = (cosx+ 1)cosy.
Example 1.58. Solve∂2u
∂x∂t= e−t cosx, given that u = 0 when t = 0 and
∂u
∂t= 0 when x = 0. Show also that as t→ ∞ , u→ sinx.
Solution: Given∂2u
∂x∂t= e−t cosx
Integrating w.r.t ‘x‘∂u
∂t= e−t sinx+ f(t)
When x = 0,∂u
∂t= 0
∴ f(t) = 0 Hence∂u
∂t= e−t sinx
Integrating w.r.t ‘t‘u = −e−t sinx+ g(x)When t = 0, u = 00 = −sinx+ g(x) i.e., g(x) = sinx∴ u = −e−t sinx+ sinxu = (1− e−t) sin x.Apply t→ ∞, we have e−t → 0⇒ u = sinx.
1.6.2 Exercises & Solutions
Solve the following partial differential equation using direct integration
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 35
1.∂z
∂x= −x
y
[Ans: z = −x
2
2y+ φ(y)
]2.∂2z
∂x2= sin y
[Ans :z =
x2
2sin y + xf(y) + φ(y)
]3.
∂2z
∂x∂y= x2 + y2
[Ans :z =
x3y
3+y3x
3+ φ(y) + f(x)
]4.∂z
∂x= 4x− 2y;
∂z
∂y= −2x+ 6y
[Ans :z = 2x2 − 2xy + 3y2 + c
]5.∂2z
∂y2− z = 0 when y = 0, z = ex and
∂z
∂y= e−x[
Ans :z = ey coshx+ e−y sinhx]
1.7 Solution of standard types of first order partial differentialequations
A partial differential equation which involves only the first order partial
derivatives p (i.e., =∂z
∂x) and q (i.e., =
∂z
∂x) is called a first order
partial differential equation.The general form of first order partial differential equation is
f(x, y, z, p, q) = 0. We shall see some standard form of such equationsand solve them by special methods.
1.7.1 Type I: f(p, q) = 0 i.e., the equation contain p and q only.
To find Complete IntegralGiven f(p, q) = 0 (1)Let z = ax+ by + c be a trial solution of the equation (1). Then
p =∂z
∂x= a and q =
∂z
∂y= b
From (1), we get
f(a, b) = 0
Hence the complete integral of (1) is
z = ax+ by + c
Solving for b from f(a, b) = 0, we get b = β(a)∴ The complete integral of (1) is
z = ax+ β(a)y + c (2)
36 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
since [number of arbitrary constants(a,c)=number of independentvariables(x,y)=2]
To find Singular IntegralTo obtain the singular integral we have to eliminate a and c from the
equations z = ax+ β(a)y + c,∂z
∂a= 0 and
∂z
∂c= 0
i.e., z = ax+ β(a)y + c
x+ β′(a)y = 0
1 = 0
The last equation being absurd and hence there is no singularintegral .
To find General IntegralPut c = χ(a) in complete integral equation (2), we get
z = ax+ β(a)y + χ(a) (3)
Differentiate (3) w.r.t. ‘a’, we get
z = x+ β′(a)y + χ
′(a) (4)
Eliminating ‘a’ from (3) and (4), we get general integral of the givenp.d.e.
Note: For the equation of the type of f(p, q) = 0 there is no singularintegral.
Example 1.59. Solve pq = 1.
Solution: Given pq = 1 (1)To find Complete Integral :Let z = ax+ by + c be a trial solution of the equation (1). Then
p =∂z
∂x= a and q =
∂z
∂y= b
From (1), we get
ab = 1
i.e., =1
a
Hence the complete integral of (1) is
z = ax+1
ay + c (2)
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 37
since [number of arbitrary constants(a,c)=number of independentvariables(x,y)=2]To find Singular Integral :To obtain the singular integral we have to eliminate ‘a’ and ‘c’ from the
equations z = ax+ β(a)y + c,∂z
∂a= 0 and
∂z
∂c= 0
i.e., z = ax+1
ay + c
x− 1
a2y = 0
1 = 0
The last equation being absurd and hence there is no singular integral.To find General Integral :Put c = χ(a) in complete integral equation (2), we get
z = ax+1
ay + χ(a) (3)
Differentiate (3) w.r.t. ‘a’, we get
z = x− 1
a2y + χ
′(a) (4)
Eliminating ‘a’ from (3) and (4), we get general integral of the givenp.d.e.
Example 1.60. Solve p2 + q2 = 1.
Solution: Given p2 + q2 = 1 (1)Let z = ax+ by + c be a trial solution of (1)
Then p =∂z
∂x= a and q =
∂z
∂y= b
From (1), we get
a2 + b2 = 1
∴ b =√
1− a2
∴ The complete integral is z = ax+√
1− a2y + cThe singular and general integrals are obtained as first example of this
type.
Example 1.61. Solve p+ q = pq.
Solution: Given p+ q = pq (1)Let z = ax+ by + c be a trial solution of (1)
38 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Then p =∂z
∂x= a and q =
∂z
∂y= b
From (1), we get
a+ b = ab
i.e., b =a
a− 1
∴ The complete integral is z = ax+a
a− 1y + c
The singular and general integrals are obtained as in previous procedure.
Example 1.62. Solve pq + p+ q = 0.
Solution: Solution: Similar to above Example
Example 1.63. Find the complete integral of p = eq.
Solution: Given logp = q (1)Let z = ax+ by + c be a trial solution of (1)
Then p =∂z
∂x= a and q =
∂z
∂y= b
From (1), we get
log a = b
∴ The complete integral is z = ax+ (loga)y + c
The singular and general integrals are obtained as first example of thistype.
Example 1.64. Solve√p+
√q = 1.
Solution: Given√p+
√q = 1 (1)
Let z = ax+ by + c be a trial solution of (1)
Then p =∂z
∂x= a and q =
∂z
∂y= b
From (1), we get
√a+
√b = 1
i.e., b =(1−
√a)2
∴ The complete integral is z = ax+(1−
√a)2y + c
The singular and general integrals are obtained as first example of thistype.
Example 1.65. Find the complete integral of p2 + q2 = npq.
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 39
Solution: Given p2 + q2 = npq (1)Let z = ax+ by + c be a trial solution of (1)
Then p =∂z
∂x= a and q =
∂z
∂y= b
From (1), we get
a2 + b2 = nab
i.e., b2 − nab + a2 = 0
∴ b =na±
√n2a2 − 4a2
2
b =a
2
[n±
√n2 − 4
]∴ The complete integral is z = ax+
a
2
[n±
√n2 − 4
]y + c
1.7.2 Type II: z = px+ qy + f(p, q) (Clairaut’s Form)
Suppose a partial differential equation of the form
z = px+ qy + f(p, q) (1)
which is said to be of Clairaut’s form.To find complete integral : The complete integral of (1) is bysubstituting p = a and q = b in (1), i.e.,
z = ax+ by + f(a, b) (2)
where a and b are arbitrary constants.To find singular integral : Partially differentiating (2) w.r.t ‘a’ and ‘b’and then equating to zero, we get
x+∂f
∂a= 0 (3)
y +∂f
∂b= 0 (4)
By eliminating a and b from (2), (3) and (4), we get the singular integral.To find General Integral : Put b = ψ(a) in complete integral equation(2), we get
z = ax+ β(a)y + f(a, β(a)) (5)
Differentiate (3) w.r.t. ‘a’, we get
z = x+ β′(a)y + f
′(a, β(a)) (6)
Eliminating ‘a’ from (5) and (6), we get general integral of the givenp.d.e.
40 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Example 1.66. Find the complete and singular integrals of z = px+ qy+pq.
Solution: Given z = px+ qy + pq (1)The given partial differential equation is in Clairaut’s form.∴ The complete integral is z = ax+ by + ab (2)
where a and b are arbitrary constants.To find singular integral : Partially differentiating (2) w.r.t ‘a’ and ‘b’then equating to zero, we get
x+ b = 0
y + a = 0
∴ a = −y and b = −x
Substituting the values of ‘a’ and ‘b’ in (2), we get
z = −yx− xy + (−y)(−x) = −yx− xy + yx
∴ z = −xywhich is the singular integral.
To find General Integral : Put b = β(a) in complete integral equation(2), we get
z = ax+ β(a)y + aβ(a) (3)
Differentiate (3) w.r.t. ‘a’, we get
z = x+ β′(a)y + aβ
′(a) + β(a) (4)
Eliminating ‘a’ from (3) and (4), we get general integral of the givenp.d.e.
Example 1.67. Solve z = px+ qy + 2√pq.
Solution: Given z = px+ qy + 2√pq (1)
The given partial differential equation is in Clairaut’s form∴ The complete integral is z = ax+ by + 2
√ab (2)
where a and b are arbitrary constants.To find singular integral: Partially differentiating (2) w.r.t a and b thenequating to zero, we get
x+
√b
a= 0 (3)
y +
√a
b= 0 (4)
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 41
∴ x = −√b
aand y = −
√a
bFrom the above two equations we get
xy = 1
which is the singular integral.The general integral is obtained as first example of this type.
Example 1.68. Solve z = px+ qy + p2q2.
Solution: The given equation is in Clairaut’s form.∴ The complete integral is z = ax+ by + a2b2 (1)To find singular integral: Partially differentiating (1) w.r.t a and b thenequating to zero, we get
x = −2ab2 (2)
y = −2a2b (3)
xy = 4a3b3 (4)
From (2),x
b= −2ab
From (3),y
a= −2ab
From (1), z = ab[xb+y
a+ ab
]= ab(−2ab− 2ab+ ab)
z = −3a2b2 (5)
Now, z3 = −27a6b6
i.e., z3 = −27(a3b3)2
Using (4), z3 = −27(xy4
)2∴ z3 =
−27
16x2y2
which is the required singular integral.The general integral is obtained as first example of this type.
Example 1.69. Solve z = px+ qy + p2 − q2.
Solution: The given partial differential equation is in Clairaut’s form∴ The complete integral is z = ax+ by + a2 − b2 (1)
To find singular integral: Partially differentiating (2) w.r.t a and b thenequating to zero, we get
42 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
x+ 2a = 0
y − 2b = 0
∴ a =−x2
and b =y
2Substituting the values of a and b in (1), we get
z =−x2
2+y2
2+x2
4− y2
4
ie., 4z = y2 − x2 which is the singular integral.The general integral is obtained as first example of this type.
Example 1.70. Solve z = px+ qy + p2 + pq + q2.
Solution: The given equation is in Clairaut’s form.∴ The complete integral is z = ax+ by + a2 + ab+ b2 (1)To find singular integral: Partially differentiating (1) w.r.t a and b thenequating to zero, we get
x+ 2a+ b = 0 (2)
y + a+ 2b = 0 (3)
Solving (2) and (3), we get
a =1
3(y − 2x), b =
1
3(x− 2y)
Substituting the values of a and b in (1), we get
3z = xy − x2 − y2
which is the singular integral.The general integral is obtained as first example of this type.
Example 1.71. Solve z = px+ qy +√
1 + p2 + q2.
Solution: The given equation is in Clairaut’s form.
∴ The complete integral is z = ax+ by +√
1 + a2 + b2 (1)To find singular integral: Partially differentiating (1) w.r.t a and b then
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 43
equating to zero, we get
x+a√
1 + a2 + b2= 0 ⇒ x =
−a√1 + a2 + b2
(2)
y +b√
1 + a2 + b2= 0 ⇒ y =
−b√1 + a2 + b2
(3)
From (2) and (3), we get
x2 + y2 =a2 + b2
1 + a2 + b2
1− (x2 + y2) = 1− a2 + b2
1 + a2 + b2
1− x2 − y2 =1
1 + a2 + b2√1 + a2 + b2 =
1√1− x2 − y2
(4)
From (2) and (3), we get
a =−x√
1− x2 − y2(5)
and
b =−y√
1− x2 − y2(6)
Substituting (4), (5) and (6) in (1), we get
z =−x2√
1− x2 − y2− y2√
1− x2 − y2+
1√1− x2 − y2
z =√
1− x2 − y2
z2 = 1− x2 − y2
∴ x2 + y2 + z2 = 1
which is the singular integral.The general integral is obtained as first example of this type.
Example 1.72. Find the complete integral ofpqz = p2(xq + p2) + q2(yp+ q2).
Solution: Given pqz = p2(xq + p2) + q2(yp+ q2)
= p2xq + p4 + q2yp+ q4
Dividing by pq we have
44 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
z = px+ qy +p4 + q4
pq
This is Clairaut’s form.The complete integral is
z = ax+ by +a4 + b4
ab,
where a and b are arbitrary constants.
1.7.3 Type III: f(z, p, q) = 0 ie., equation not containing x and y explicity
Given f(z, p, q) = 0 (1)Let z = f(x+ ay) be the solution of (1)Put x+ ay = u Then
z = f(u) and∂u
∂x= 1,
∂u
∂y= a
p =∂z
∂x=dz
du
∂u
∂x=dz
du
q =∂z
∂y=dz
du
∂u
∂y= a
dz
du
Substituting the values of p and q in (1), we get
f
(z,dz
du, adz
du
)= 0 (2)
This is an ordinary differential equation of first order
Solving fordz
du, we obtain
dz
du= g(z, a)
i.e.,dz
g(z, a)= du
Integrating
∫dz
g(z, a)= u+ c
φ(z, a) = u+ c
ie., (z, a) = x+ ay + c
which is the complete integral of (1).The singular integral can be obtained by the usual methods.
Example 1.73. Find the complete integral of z = pq.
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 45
Solution: Given z = pq (1)Let z = f(x+ ay) be the solution of (1)Put u = x+ ay so that z = f(u)
p =dz
duand q = a
dz
du
Substituting in (1), we get
z = a
(dz
du
)2
dz
du=
√z
a⇒ dz√
z=
1√adu
Integrating, 2√z =
1√au+ c
2√az = u+
√ac
2√az = u+ b,whereb =
√ac
4az = (u+ b)2
4az = (x+ ay + b)2
which is the complete integral.
Example 1.74. Solve z = p2 + q2.
Solution: Given z = p2 + q2 (1)Let z = f(x+ ay) be the solution of (1)Put u = x+ ay so that z = f(u), then
p =dz
duand q = a
dz
du
Substituting in (1), we get
z =
(dz
du
)2
+ a2(dz
du
)2
⇒ z =(1 + a2)
(dz
du
)2
dz
du=
√z
1 + a2⇒√1 + a2
dz√z
=du
Integrating,√
1 + a22√z = u+ b
(1 + a2)4z = (u+ b)2
i.e., 4(1 + a2)z = (x+ ay + b)2
which is the complete integral.
Example 1.75. Solve p(1 + q) = qz.
46 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Solution: Given p(1 + q) = qz (1)Let z = f(x+ ay) be the solution of (1)Put u = x+ ay so that z = f(u), then
p =dz
duand q = a
dz
du
Substituting in (1), we get
dz
du
(1 + a
dz
du
)= a
dz
duz ⇒1 + a
dz
du=az
dz
du=az − 1
a⇒ a
az − 1dz =du
Integrating, log(az − 1) = u+ b
i.e., log(az − 1) = x+ ay + b
which is the complete integral.
Example 1.76. Solve 9(p2z + q2
)= 4.
Solution: Given 9(p2z + q2
)= 4 (1)
Let z = f(x+ ay) be the solution of (1)Put u = x+ ay so that z = f(u), then
p =dz
duand q = a
dz
du
Substituting in (1), we get
9
[(dz
du
)2
z + a2(dz
du
)2]= 4(
dz
du
)2
=4
9(z + a2)⇒ dz
du=
2
3
1√z + a2
3
2
√z + a2dz = du
Integrating,we get3
2
(z + a2
)3/23/2
= u+ b(z + a2
)3/2= x+ ay + b
which is the complete integral.
Example 1.77. Solve p(1 + q2
)= q(z − a).
Solution: Given p(1 + q2
)= q(z − a) (1)
Let z = f(x+ ay) be the solution of (1)Put u = x+ ay so that z = f(u), then
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 47
p =dz
duand q = a
dz
duSubstituting in (1), we get
dz
du
[1 + b2
(dz
du
)2]= b
dz
du(z − a) ⇒
[1 + b2
(dz
du
)2]
=b(z − a)(dz
du
)2
=bz − ba− 1
b2⇒ dz
du=
√bz − ba− 1
bb√
bz − ba− 1dz = du
Integrating, 2√bz − ba− 1 = u+ c
4 (bz − ba− 1)2 = (u+ c)2
i.e., 4 (bz − ba− 1)2 = (x+ by + c)2
which is the complete integral.
Example 1.78. Solve z2(p2 + q2 + 1
)= 1.
Solution: Given z2(p2 + q2 + 1
)= 1 (1)
Let z = f(x+ ay) be the solution of (1)Put u = x+ ay so that z = f(u), then
p =dz
duand q = a
dz
duSubstituting in (1), we get
z2
[(dz
du
)2
+ a2(dz
du
)2
+ 1
]= 1(
dz
du
)2
+ a2(dz
du
)2
+ 1 =1
z2(1 + a2
)(dzdu
)2
=1
z2− 1
(1 + a2
)(dzdu
)2
=1− z2
z2
dz
du=
√1− z2
z√1 + a2
z√1− z2
dz =1√
1 + a2du
Integrating,
∫z√
1− z2dz =
1√1 + a2
∫du (2)
Put 1− z2 = t⇒ −2zdz = dt⇒ zdz = −1
2dt
48 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
(2) ⇒ −1
2
∫dt√t=
1√1 + a2
∫du
−1
22√t =
1√1 + a2
u+ b
−√t =
u√1 + a2
+ b
−√t =
x+ ay√1 + a2
+ b
−√
1− z2 =x+ ay√1 + a2
+ b
which is the complete integral.
Example 1.79. Solve z2(p2 + q2 + 1
)= c2.
Solution: Similar to above Example.
Example 1.80. Solve q2 = z2p2(1− p2
).
Solution: Given q2 = z2p2(1− p2
)(1)
Let z = f(x+ ay) be the solution of (1)Put u = x+ ay so that z = f(u), then
p =dz
duand q = a
dz
du
Substituting in (1), we get
a2(dz
du
)2
= z2(dz
du
)2[1−
(dz
du
)2]
a2
z2= 1−
(dz
du
)2
⇒(dz
du
)2
=z2 − a2
z2
dz
du=
√z2 − a2
z⇒ z√
z2 − a2dz =du
Integrating,
∫z√
z2 − a2dz =
∫du
Put z2 − a2 = t⇒ 2zdz = dt⇒ zdz =1
2dt
1
2
∫dt√t=
∫du
1
22√t = u+ b ⇒
√t =u+ b√
z2 − a2 = x+ ay + b
which is the complete integral.
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 49
1.7.4 Type IV: f(x, p) = g(y, q) (Separable equations)
A first order partial differential equation is called separable if it can bewritten as f(x, p) = g(y, q).Let f(x, p) = g(y, q) = a , where a is an arbitrary constant.
∴ f(x, p) = a, g(y, q) = a Solving for p and q from these two equations weget
p = f1(x, a) and q = g1(y, a)
Since dz =∂z
∂xdx+
∂z
∂ydy
= pdx+ qdy
dz = f1(x, a)dx+ g1(x, a)dyIntegrating, we get
z =
∫f1(x, a)dx+
∫g1(y, a)dy + b
which gives the complete integral.
Example 1.81. Solve p+ q = x+ y
Solution: Given that p+ q = x+ y (1)i.e., p− x = y − q
Let p− x = y − q = a , where a is an arbitrary constant.Now, p− x = a⇒ p = a+ x
i.e., f1(x, a) = a+ xy − q = a⇒ q = y − z
i.e., f2(y, a) = y − a∴ The complete integral is
z =
∫f1(x, a)dx+
∫f2(y, a)dy + b
z =
∫(a+ x) dx+
∫(y − a)dy + b
z =(a+ x)2
2+
(y − a)2
2+ b
2z = (a+ x)2 + (y − a)2 + 2b
Example 1.82. Solve p− x2 = q + y2
Solution: Let p− x2 = q + y2 = a (1)Now, p− x2 = a⇒ p = a+ x2
i.e., f1(x, a) = a+ x2
50 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
q + y2 = a⇒ q = a− y2
i.e., f2(y, a) = a− y2
∴ The complete integral is
z =
∫f1(x, a)dx+
∫f2(y, a)dy + b
z =
∫ (a+ x2
)dx+
∫ (a− y2
)dy + b
z =
(ax+
x3
3
)+
(ay − y3
3
)+ b
Example 1.83. Solve p2 − q2 = x− y
Solution: Given p2 − x = q2 − yLet p2 − x = q2 − y = aNow, p2 − x = a⇒ p =
√a+ x
i.e., f1(x, a) =√a+ x
q2 − y = a⇒ q =√a+ y
i.e., f2(y, a) =√a+ y
∴ The complete integral is
z =
∫f1(x, a)dx+
∫f2(y, a)dy + b
z =
∫ √a+ xdx+
∫ √a+ ydy + b
z =2
3(a+ x)3/2 +
2
3(a+ y)3/2 + b
Example 1.84. Solve p2 + q2 = x+ y
Solution: Similar to above Example.
Example 1.85. Solve√p+
√q = 2x
Solution: Let√p− 2x = −√
q = a√p− 2x = a⇒ p = a+ 2x2
i.e., f1(x, a) = a+ 2x2
−√q = a⇒ q = a2
i.e., f2(y, a) = a2
∴ The complete integral is
z =
∫f1(x, a)dx+
∫f2(y, a)dy + b
z =
∫(a+ 2x)2dx+
∫a2dy + b
z =(a+ 2x)3
3× 2+ a2y + b
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 51
Example 1.86. Solve p2y(1 + x2
)= qx2
Solution: Givenp2(1 + x2
)x2
=q
y
Letp2(1 + x2
)x2
=q
y= a
Nowp2(1 + x2
)x2
= a⇒ p =
√ax√
1 + x2
i.e., f1(x, a) =
√ax√
1 + x2q
y= a⇒ q = ay
i.e., f2(y, a) = ay∴ The complete integral is
z =
∫f1(x, a)dx+
∫f2(y, a)dy + b
z =
∫ √ax√
1 + x2dx+
∫aydy + b
Put 1 + x2 = t⇒ 2xdx = dt
z =
√a
2
∫dt√t+ a
∫ydy + b
=
√a
22√t + a
y2
2+ bz =
√a√
1 + x2 + ay2
2+ b
Example 1.87. Solvex
p+y
q= 1
Solution: Givenx
p= 1− y
q
Letx
p= 1− y
q= a
Nowx
p= a ⇒ p =
x
a∴ f1(x, a) =
x
a
1− y
2= a ⇒ q =
y
a+ 1∴ f2(y, a) =
y
a+ 1∴ The complete integral is
z =
∫f1(x, a)dx+
∫f2(y, a)dy + b
z =
∫x
adx+
∫y
a+ 1dy + b
z =x2
2a+
y2
2(a+ 1)+ b
52 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Example 1.88. Solve p+ q = sin x+ sin y
Solution: Given p− sinx = q − sin yLet p− sinx = q − sin y = a
Now p− sinx = a ⇒ p = a+ sinx ∴ f1(x, a) = a+ sinx
q − sin y = a ⇒ q = a+ sin y ∴ f2(y, a) = a+ sin y
∴ The complete integral is
z =
∫f1(x, a)dx+
∫f2(y, a)dy + b
z =
∫(a+ sinx) dx+
∫(a + siny) dy + b
z = ax− cosx+ ay − cos y + b
1.7.5 Exercises & Solutions
Solve the following partial differential equations
Type I : f(p, q)=0
a) p2 − q2 = 4[Ans: z = ax+
√a2 − 4y + c
]b) p+ sinq = 0 [Ans : z = by − xsinb+ c]
c) p = q [Ans : z = a(x+ y) + c]
Type II : Clairaut’s form:
a) z = px+ qy +p
q− p
[Ans: CI : z = ax+ by +
a
b− a,SI :
y
1− x
]b) (1− x)p+ (2− y)q = 3− z [Ans: CI : z = ax+ by − a− 2b+ 3]
c) z = px+ qy + logpq[Ans: CI : z = ax+ by + logab,SI : z = −2− logxy]
d) z = px+ qy + 3p1/3q1/3[Ans: CI : ax+ by + 3a1/3b1/3,SI : xyz = 1
]Type III : f(z,p, q)=0
a) p2 + pq = z2[Ans: logz =
1√1 + a
(x+ ay) + b
]b) z2 = p2 + q2 + 1
[Ans: cosh−1 z =
1√1 + a2
(x+ ay) + b
]c) p3 + q3 = 8z
[Ans: (1 + a3)z2 =
64
27(x+ ay + b)3
]
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 53
d) p+ q = z [Ans: x+ ay = (1 + alogz + b)]
e) p3 = qz[Ans: 4z = a(x+ ay − b)2
]Type IV : f(x,p )=g(y,q )
a) q2 − p = y − x
[Ans: z =
(a+ x)2
2+
2
3(a+ y)3/2 + b
]b) yp = 2yx+ logq
[Ans: z = x2 + ax+
eay
a
]c) p2 + q2 = x+ y
[Ans: z =
2
3
{(x+ a)3/2 + (y − a)3/2
}+ b
]d) q = xyp2
[Ans: z = 2
√ax+
ay2
2+ b
]e) x2p2 = yq2 [Ans: z = alogx+ 2a
√y + b]
54 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
1.7.6 Equations Reducible to Standard Forms
1.7.7 Type A : f (xmp, ynq) = 0
An equation of the form f (xmp, ynq) = 0 where m and n are constantscan be transformed into an equation of the first type [i.e., f(p, q) = 0]Put X = x1−m and Y = y1−n where m 6= 1 and n 6= 1.
Then p =∂z
∂x=
∂z
∂X
dX
∂x= P (1−m)x−m where P =
∂z
∂X∴ xmp = (1−m)P
q =∂z
∂y=∂z
∂Y
dY
∂y= Q (1− n) y−n where Q =
∂z
∂Y
∴ ynq = (1− n)Q
Hence the given equation reduces to f [(1−m)P, (1− n)Q] = 0which is of the form F(P, Q) = 0 (Type I)
1.7.8 Type B : f (xmp, ynq, z) = 0
An equation of the form f (xmp, ynq, z) = 0 can also be transformed to thestandard type F(P, Q, z) = 0 (Type II) by the substitution X = x1−m
and Y = y1−n where m 6= 1 and n 6= 1.Note: In the above two types if m = 1 put X = log x and if n = 1 , putY = log y:
∴ p =∂z
∂x=
∂z
∂X
dX
∂x= P
1
xwhere P =
∂z
∂Xi.e., xp = P
q =∂z
∂y=∂z
∂Y
dY
∂x= Q
1
ywhere Q =
∂z
∂Y
i.e., yq = Q∴ The equations reduce to either F (P,Q) = 0 or F (P,Q, z) = 0.
1.7.9 Type C : f(zkp, zkq
)= 0 where k is a constant
An equation of the form f(zkp, zkq
)= 0 can be transformed into an
equation of the first type [i.e., f(p, q) = 0 ].Put Z = zk+1 if k 6= −1
Then∂Z
∂x= (k + 1) zkp
P = (k + 1) zkp where P =∂Z
∂x
i.e., zkp =1
k + 1P
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 55
Q = (k + 1) zkq where Q =∂Z
∂y
i.e., zkq =1
k + 1Q
Hence the given equation reduces to f
(1
k + 1P,
1
k + 1Q
)= 0 which is of
the form F(P, Q) = 0Note: If k = −1, then
Put Z = log z
∂Z
∂x=
1
zp
P = z−1p
Similarly,
P = z−1p
Hence the given equation again reduces to the form F (P,Q) = 0.
1.7.10 Type D: f(xmzkp, ynzkq
)= 0
This can be transformed to an equation of the form F (P,Q) = 0 by thesubstitution.
X =
{x1−m if m 6= 1log x if m = 1
Y =
{y1−n if n 6= 1log y if n = 1
Z =
{zk+1 if k 6= −1log z if k = −1
Example 1.89. Solve xp+ yq = 1
Solution: Given xp+ yq = 1 (1)The given differential equation is the form f (xmp, ynq) = 0 with m = 1 ,n = 1.Put X = log x and Y = log y
p =∂z
∂x=
∂z
∂X
∂X
∂x= P
1
x, where P =
∂z
∂X∴ xp = P
Similarly, yq = Q where Q =∂z
∂Y∴ (1) becomes P +Q = 1 (2)
This is of the form F (P,Q) = 0 [Type I]Let z = aX + bY + c be a solution of (2)
56 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
∂z
∂X= a⇒ P = a
∂z
∂Y= b⇒ Q = b
Substituting the values of P and Q in (2), we get
a+ b = 1
i.e., b = 1− a
∴ The complete integral of P +Q = 1 is
z = aX + (1− a)Y + b
∴ The required complete integral is
z = a log x+ (1− a) log y + b
Example 1.90. Solve p2x2 + q2y2 = z2
Solution: The given differential equation can be written as(xp)2 + (yq)2 = z2 (1)This is of the form f (xmp, ynq, z) = 0 with m = 1 , n = 1.Put X = log x and Y = log y
p =∂z
∂x=
∂z
∂X
∂X
∂x= P
1
x,whereP =
∂z
∂X
∴ xp = P
Similarly, yq = Q where Q =∂z
∂Y∴ (1) becomes P 2 +Q2 = z2 (2)
This is of the form F (P,Q, z) = 0 [Type II]Let z = φ(X + aY ) be a solution of (2)Let u = X + aY so that z = φ(u)
Then P =dz
duand Q = a
dz
du
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 57
∴ (2) becomes
(dz
du
)2
+ a2(dz
du
)2
= z2(dz
du
)2 (1 + a2
)= z2
dz
du=
z√(1 + a2)
dz
z=
du√(1 + a2)
Integrating, log z =u√
(1 + a2)+ b
∴ log z =X + aY√(1 + a2
) + b
log z =log x+ a log y√
(1 + a2)+ b
which is the complete integral.
Example 1.91. Solve p2x2 + q2y2 = z
Solution: This problem is to similar to above problem.
Example 1.92. Solve z2 = xypq
Solution: The given equation can be written as z = (xp)(yq) (1)This is of the form f (xmp, ynq, z) = 0 with m = 1 , n = 1.Put X = log x and Y = log y
p =∂z
∂x=
∂z
∂X
∂X
∂x= P
1
x, where P =
∂z
∂X
∴ xp = P
Similarly, yq = Q where Q =∂z
∂Y∴ (1) becomes z2 = PQ (2)
This is of the form F (P,Q, z) = 0 [Type II]Let z = φ(X + aY ) be a solution of (2)Let u = X + aY so that z = φ(u)
Then P =dz
duand Q = a
dz
du
58 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
∴ (2) becomes z2 = a2(dz
du
)2
dz
du=
z√a
dz
z=
√adu
Integrating, log z =√au+ b
i.e., log z =√a (X + aY ) + b
i.e., log z =√a (log x+ a log y) + b
which is the complete integral.
Example 1.93. Solve z2(p2x2 + q2
)= 1
Solution: The given equation can be written as (px)2 + q2 =1
z2(1)
This is of the form f (xmp, q, z) = 0 with m = 1.Put X = log x
p =∂z
∂x=
∂z
∂X
∂X
∂x= P
1
x, where P =
∂z
∂X
∴ xp = P
∴ (1) becomes P 2 + q2 =1
z2(2)
This is of the form F (P, q, z) = 0Let z = φ(X + aY ) be a solution of (2)Let u = X + aY so that z = φ(u)
Then P =dz
duand Q = a
dz
du
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 59
∴ (2) becomes
(dz
du
)2
+ a2(dz
du
)2
=1
z2(dz
du
)2 (1 + a2
)=
1
z2
dz
du=
1
z√1 + a2
zdz =1√
1 + a2du
Integrating,z2
2=
1√1 + a2
u+ b
z2
2=
1√1 + a2
(X + ay) + b
z2
2=
1√1 + a2
(log x+ ay) + b√1 + a2z2 = 2 (log x+ ay) + c, where c = 2b
which is the complete integral.
Example 1.94. Solve p2x4 + y2zq = 2z2
Solution: The given equation can be written as (x2p)2 + (y2q)z = 2z2(1)This is of the form f (xmp, ynq, z) = 0 with m = 2, n = 2. [Type II]Put X = x1−m and Y = y1−n
i.e., X =1
xand Y =
1
y
p =∂z
∂x=
∂z
∂X
∂X
∂x=
∂z
∂X
(− 1
x2
)= −P
x2,whereP =
∂z
∂X
∴ x2p = −PSimilarly y2q = −Q, where Q =
∂z
∂Y∴ (1) becomes P 2 −Qz = 2z2 (2)
This is of the form F (P,Q, z) = 0 [Type II]Let z = φ(X + aY ) be a solution of (2)Let u = X + aY so that z = φ(u)
Then P =dz
duand Q = a
dz
du
60 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
∴ (2) becomes
(dz
du
)2
− a
(dz
du
)z = 2z2(
dz
du
)2
− az
(dz
du
)− 2z2 = 0
dz
du=az ±
√a2z2 + 8z2
2
dz
du= z
[a±
√a2 + 8
2
]dz
z=
[a±
√a2 + 8
2
]du
Integrating, log z =
[a±
√a2 + 8
2
]u+ b
log z =
[a±
√a2 + 8
2
](X + aY ) + b
log z =
[a±
√a2 + 8
2
](1
x+a
y
)+ b
which is the required complete integral.
Example 1.95. Solve x4p2 − yzq = z2
Solution: The given equation can be written as (x2p)2 − (yq)z = z2 (1)This is of the form f (xmp, ynq, z) = 0 with m = 2, n = 1.Put X = x1−m and Y = log y
i.e., X =1
xand Y = log y
p =∂z
∂x=
∂z
∂X
∂X
∂x=
∂z
∂X
(− 1
x2
)= P
(− 1
x2
),whereP =
∂z
∂X
∴ x2p = −P
q =∂z
∂y=∂z
∂Y
∂Y
∂y=∂z
∂Y
(1
y
)= Q
(1
y
),whereP =
∂z
∂Yi.e., yq = Q
∴ (1) becomes P 2 −Qz = z2 (2)This is of the form F (P,Q, z) = 0 [Type II]Let z = φ(X + aY ) be a solution of (2)Let u = X + aY so that z = φ(u)
Then P =dz
duand Q = a
dz
du
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 61
∴ (2) becomes
(dz
du
)2
− a
(dz
du
)z = z2(
dz
du
)2
− az
(dz
du
)− z2 = 0
dz
du=az ±
√a2z2 + 4z2
2
dz
du= z
(a±
√a2 + 4
2
)dz
z=
(a±
√a2 + 4
2
)du
Integrating, log z =
[a±
√a2 + 4
2
]u+ b
log z =
[a±
√a2 + 4
2
](X + aY ) + b
log z =
[a±
√a2 + 4
2
](1
x+ a log y
)+ b
which is the required complete integral.
Example 1.96. Solve z2(p2 + q2) = x2 + y2
Solution: The given equation can be written as (zp)2 + (zq)2 = x2 + y2
(1)Put Z = zk+1, where k = 1 (Type C)i.e., Z = z2
∂Z
∂x= 2z
∂z
∂x
i.e., P = 2zp, where P =∂Z
∂x⇒ zp =
P
2
Similarly, zq =Q
2, where Q =
∂Z
∂y
62 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
∴ (1)becomesP 2
4+Q2
4= x2 + y2
P 2 +Q2 = 4(x2 + y2
)P 2 − 4x2 = 4y2 −Q2
Take P 2 − 4x2 = 4y2 −Q2 = 4a2 (say) [TypeIV ]
Now P 2 − 4x2 = 4a2
⇒ P = 2√x2 + a2
∴ f1(x, a) = 2√x2 + a2
Now, 4y2 −Q2 = 4a2
⇒ Q = 2√y2 − a2
∴ f2(y, a) = 2√y2 − a2
∴ Z =
∫f1 (x, a) dx+
∫f2 (y, a) dy + b
Z = 2
∫ √x2 + a2dx+ 2
∫ √y2 − a2dx+ b
Z = 2
[x
2
√x2+a2+
a2
2sinh−1
(xa
)]+2
[y
2
√y2−a2−a
2
2cosh−1
(ya
)]+b
z2 = x√x2 + a2 + a2 sinh−1
(xa
)+ y√y2 − a2 − a2 cosh−1
(ya
)+ b
which is the required complete integral.
Example 1.97. Solve p2 + q2 = z2(x2 + y2)
Solution: The given equation can be written as
(pz
)2+(qz
)2= x2 + y2(
z−1p)2
+(z−1q
)2= x2 + y2 (1)
Put Z = log z [∵ k = −1] (TypeC)
∂Z
∂x=
1
z
∂z
∂x
i.e., P = z−1p, where P =∂Z
∂x
Similarly, Q = z−1q, where Q =∂Z
∂y
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 63
∴ (1)becomes P 2 +Q2 = x2 + y2
P 2 − x2 = y2 −Q2
Take P 2 − x2 = y2 −Q2 = a2 (say) [TypeIV ]
P 2 − x2 = a2
⇒ P =√x2 + a2
∴ f1(x, a) =√x2 + a2
Now, y2 −Q2 = a2
⇒ Q =√y2 − a2
∴ f2(y, a) =√y2 − a2
∴ Z =
∫f1 (x, a) dx+
∫f2 (y, a) dy + b
Z =
∫ √x2 + a2dx+
∫ √y2 − a2dy + b
Z =
[x
2
√x2+a2+
a2
2sinh−1
(xa
)]+
[y
2
√y2−a2−a
2
2cosh−1
(ya
)]+b
log z =1
2
[x√x2 + a2 + a2 sinh−1
(xa
)+ y√y2 − a2 − a2 cosh−1
(ya
)]+ b
which is the required complete integral.
1.7.11 Exercises & Solutions
Solve the following partial differential equations:
1.x2
p+y2
q= z
[Ans:
3
2z2 =
x3
a+
y3
1− a+ b
]
2. 2x4p2−yzq−3z2=0
[Ans: log z=
a±√a2+24
2
(1
x+a log y
)+y
]
3. xp+ yq = 1
[Ans: z =
1
1 + a(log x+ a log y) + b
]
4.p
x2+
q
y2= z
[Ans: log z =
1
3 (1 + a)
(x3 + ay3
)+ b
]
5. q2y2 = z(z − px)
[Ans: log z =
−1±√1 + 4a2
2a2(log x+ alogy) + b
]
64 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
1.8 Homogeneous Linear Partial Differential Equation
A linear partial differential equation in which all the partial derivativesare of the same order is called homogeneous linear partial differentialequation. Otherwise it is called non homogeneous linear partial equation.A homogeneous linear partial differential equation of nth order with
constant coefficients is of the form
a0∂nz
∂xn+ a1
∂nz
∂xn−1∂y+ · · ·+ an
∂nz
∂yn= F (x, y) (1)
where a0, a1, . . . an are constants.This equation also can be written in the form
(a0Dn + a1D
n−1D′ + · · ·+ anD′n)z = F (x, y)
where D =∂
∂x,D′ =
∂
∂y.
i.e., f(D,D′)z = F (x, y) (2)The solution of f(D,D′)z = 0 is called the complementary function of (2).The particular solution of (2) is called the particular integral function of(2) given symbolically by
P.I. =F (x, y)
f(D,D′)
Hence the complete solution z = complementary function + particularintegral.i.e., Complete solution z = C.F. + P.I.
1.8.1 To find the complementary function (C.F)
Put D = m and D′ = 1 in f(D,D′) = 0, then we get an equation, whichis called the auxillary equation of (2).∴ the auxillary equation of (2) is f(m, 1) = 0i.e., a0m
n + a1mn−1 + · · ·+ an = 0
Let the roots of this equation be m1,m2, . . . ,mn.Case (i): If the roots are real (or imaginary) and distinct. ThenC.F. = f1(y +m1x) + f2(y +m2x) + · · ·+ fn(y +mnx)Case (ii):
(a) If any two roots are equal (ie, m1 = m2 = m) and others are distinct,Then C.F. = f1(y+mx)+xf2(y+m2x)+f3(y+m3x)+· · ·+fn(y+mnx).
(b) If any three roots are equal (i.e., m1 = m2 = m3 = m )and others aredistinct. ThenC.F. =f1(y +mx) + xf2(y +mx) + x2f3(y +mx) + f4(y +m4x)+
· · ·+ fn(y +mnx).
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 65
1.8.2 To find Particular Integral (P.I)
Type 1 : If F (x, y) = eax+by, then
P.I. =1
f(D,D′)eax+by
=1
f(a, b)eax+by provided f(a, b) 6= 0.
If f(a, b) = 0, then
P.I. = x.1
f ′(D,D′)eax+by
where f ′(D,D′) is the partial derivative of f(D,D′) w.r.t D.
Type 2 : If F (x, y) = sin(ax+ by), then
P.I. =1
f(D,D′)sin(ax+ by)
Replace D2 by −a2, D′2 by −b2 and DD′ by −ab in f(D,D′) providedthe denominator is not equal to zero.If the denominator is zero, then
P.I. = x1
f ′(D,D′)sin(ax+ by)
where f ′(D,D′) is the partial derivative of f(D,D′) w.r.t D. Note :Similar formula holds for F (x, y) = cos(ax+ by).
Type 3 : If F (x, y) = xmyn, then
P.I. =1
f(D,D′)xmyn
= [f(D,D′)]−1xmyn
Expand [f(D,D′)]−1 in ascending powers of D,D’ and then operateon xmyn.
Note : In xmyn if m > n, then try to write f (D,D′) as a function
ofD′
Dand if m < n, then try to write f (D,D′) as a function of
D
D′ .
Note :1
Ddenotes integration w.r.t ‘x’,
1
D′ denotes integration w.r.t
‘y’.
Type 4 : If F (x, y) = eax+byφ(x, y), then
66 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
P.I. =1
f(D,D′)eax+byφ(x, y)
= eax+by 1
f(D + a,D′ + b)φ(x, y)
Type 5 : If F (x, y) is any function, resolve f(D,D′) into linear factorssay (D −m1D
′), (D −m2D′), . . . , (D −mnD
′), then
P.I. =1
(D −m1D′)(D −m2D′) . . . (D −mnD′)F (x, y)
Note : (1)1
D −mD′F (x, y) =
∫F (x, c−mx)dx where y = c−mx
(2)1
D +mD′F (x, y) =
∫F (x, c+mx)dx where y = c+mx
1.8.3 Examples of Homogeneous Linear P.D.E.
Example 1.98. Solve(D2 − 5DD′ + 6D′2) z = 0
Solution: The auxillary equation is
m2 − 5m+ 6 = 0 [Replace D by m and D′ by 1]
(m− 2)(m− 3) = 0
i.e., m = 2, 3
C.F. = f1(y + 2x) + f2(y + 3x)
∴ The solution is z = f1(y + 2x) + f2(y + 3x)
Example 1.99. Solve∂3z
∂x3− 4
∂3z
∂x2∂y+ 4
∂3z
∂x∂y2= 0
Solution: Given (D3 − 4D2D′ + 4DD′2)z = 0The auxillary equation is
m3 − 4m2 + 4m = 0
i.e., m(m− 2)2 = 0
i.e., m = 0, 2, 2
C.F. = f1(y + 0x) + f2(y + 2x) + xf3(y + 2x)
∴ The solution is z = f1(y + 0x) + f2(y + 2x) + xf3(y + 2x)
Example 1.100. Solve the equation∂2z
∂x2− 3
∂2z
∂x∂y+ 2
∂2z
∂y2= 0
Solution: Given (D2 − 3DD′ + 2D′2)z = 0The auxillary equation is
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 67
m2 + 3m+ 2 = 0
(m− 1)(m− 2) = 0
i.e., m = 1, 2
C.F. = f1(y + x) + f2(y + 2x)
∴ The solution is z = f1(y + x) + f2(y + 2x)
Example 1.101. Solve the equation(2D2 + 5DD′ + 2D′2) z = 0
Solution: The auxillary equation is
2m2 + 5m+ 2 = 0
(2m+ 1)(m+ 2) = 0
i.e., m =−1
2,−2
C.F. = f1
(y − 1
2x
)+ f2(y − 2x)
∴ The solution is f1
(y − 1
2x
)+ f2(y − 2x)
Example 1.102. Solve(D2 − 2DD′ +D′2) z = ex+2y
Solution: The auxillary equation is
m2 − 2m+ 1 = 0
(m− 1)2 = 0
i.e., m = 1, 1
∴ C.F. is f1 (y − x) + xf2(y − x)
P.I. =1
D2 − 2DD′ +D′2ex+2y
=1
1− 4 + 4ex+2y
= ex+2y
∴ The complete solution z = C.F. + P.I.
i.e., z = f1(y + x) + xf2(y + x) + ex+2y.
Example 1.103. Solve(D2 −D′2) z = ex+2y
Solution: The auxillary equation is
m2 − 1 = 0
(m− 1)(m+ 1) = 0
i.e., m = 1,−1
68 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
∴ C.F. is f1 (y + x) + f2(y − x)
P.I. =1
D2 −D′2ex+2y
=1
1− 4ex+2y
= −1
3ex+2y
∴ The complete solution z = C.F. + P.I.
i.e., z = f1(y + x) + f2(y − x)− 1
3ex+2y.
Example 1.104. Solve(D2 − 7DD′2 + 12D′2) z = ex−y
Solution: The auxillary equation is
m2 − 7m+ 12 = 0
(m− 3)(m− 4) = 0
i.e., m = 3, 4
∴ C.F. is f1(y + 3x) + f2(y + 4x)
P.I. =1
D2 − 7DD′ + 12D′2ex−y
=1
1 + 7 + 12.ex−y
=1
20ex−y
∴ The complete solution z = C.F. + P.I.
i.e., z = f1(y + 3x) + f2(y + 4x) +1
20ex−y.
Example 1.105. Solve(6D2 + 6DD′ +D′2) z = (ex + e−2y
)2Solution: The auxillary equation is
6m2 + 6m+ 1 = 0
i.e., m =−6±
√36− 24
12
m =−3±
√3
6
∴ C.F. is f1(y + (−3 +
√3
6)x) + f2(y − (
3 +√3
6)x)
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 69
P.I. =1
6D2 + 6DD′ +D′2 (ex + e−2y)2
=1
6D2 + 6DD′ +D′2 (e2x + e−4y + 2ex−2y)
=e2x+0y
6D2 + 6DD′ +D′2 +e0x−4y
6D2 + 6DD′ +D′2 +2ex−2y
6D2 + 6DD′ +D′2
=1
24e2x +
1
16e−4y − ex−2y
∴ The complete solution z = C.F. + P.I.
Example 1.106. Solve (D +D′)2z = ex−y
Solution: The auxillary equation is
m2 + 2m+ 1 = 0
i.e., m = −1,−1
∴ C.F. is f1(y − x) + xf2(y − x)
P.I. =1
D2 + 2DD′ −D′2ex−y
=1
0ex−y
∴ P.I. = x1
2D + 2D′ex−y
= x1
0ex−y
∴ P.I. = x21
2ex−y
∴ The complete solution z = C.F. + P.I.
i.e., z = f1(y − x) + xf2(y − x) +x2
2ex−y
Example 1.107. Solve(D4 −D′4)2 z = ex+y
Solution: The auxillary equation is
m4 − 1 = 0
i.e., (m2 − 1)(m2 + 1) = 0
i.e., m = 1,−1, i,−i
∴ C.F. is f1(y + x) + f2(y − x) + f3(y + ix) + f4(y − ix)
70 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
P.I. =1
D4 −D′4ex+y
=1
0.ex+y
∴ P.I. = x1
4D3ex+y
=1
4xex+y
∴ The complete solution z = C.F. + P.I.
i.e., z=f1(y + x) + f2(y − 4x) + f3(y + ix) + f4(y − ix) +1
4xex+y
Example 1.108. Solve(D3 − 3DD′2 + 2D′3) z = ex+y
Solution: The auxillary equation is
m3 − 3m+ 2 = 0
i.e., m = 1, 1,−2
∴ C.F. is f1(y + x) + xf2(y + x) + f3(y − 2x)
P.I. =1
D3 − 3DD′2 + 2D′3ex+y
=1
0.ex+y
∴ P.I. = x1
3D2 − 3D′2ex+y = x
1
0ex+y
P.I. = x21
6Dex+y = x2
1
6ex+y
∴ The complete solution z = C.F. + P.I.
i.e., z=f1(y + x) + xf2(y + x) + f3(y − 2x) + x21
6ex+y
Example 1.109. Solve the equation(D2 + 3DD′ + 2D′2) z = ex cosh y
Solution: The auxillary equation is
m2 + 3m+ 2 = 0
i.e., m = −1,−2
∴ C.F. is f1(y − x) + f2(y − 2x)
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 71
P.I. =1
D2 + 3DD′ + 2D′2ex cosh y
=1
D2 + 3DD′ + 2D′2ex
(ey + e−y
2
)=
1
2
[1
D2 + 3DD′ + 2D′2ex+y +
1
D2 + 3DD′ + 2D′2ex−y
]=
1
2
[1
6ex+y +
1
0ex−y
]P.I. =
1
2
[ex+y
6+ x
1
2D + 3D′ex−y
]=
1
2
[ex+y
6+ x
1
−1ex−y
]∴ P.I. =
1
2
[ex+y
6− xex−y
]∴ The complete solution z = C.F. + P.I.
Example 1.110. Solve(D2 + 9D′2) z = cos(2x+ 3y)
Solution: The auxillary equation is
m2 + 9 = 0
m2 = −9
i.e., m = ±3i
∴ C.F. is f1(y + 3ix) + f2(y − 3ix)
P.I. =1
D2 + 9D′2 cos(2x+ 3y)
=1
−4− 81cos(2x+ 3y)
[Replacing D2 by − 4, D′2 by − 9
]= − 1
85cos(2x+ 3y)
∴ The complete solution z = C.F. + P.I.
i.e., z = f1(y + 3ix) + f2(y − 3ix)− 1
85cos(2x+ 3y)
Example 1.111. Solve(D2 − 2DD′ +D′2) z = sin(2x+ 3y)
Solution: The auxillary equation is
m2 − 2m+ 1 = 0
m = 1, 1
72 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
∴ C.F. is f1(y + x) + xf2(y + x)
P.I. =1
D2 − 2DD′ +D′2 sin(2x+ 3y)
=1
−4+12−9sin(2x+3y)
[Replacing D2 → −4, D′2 → −9, DD′ → −6
]= −sin(2x+ 3y)
∴ The complete solution z = C.F. + P.I.i.e., z = f1(y + x) + xf2(y + x)− sin(2x+ 3y)
Example 1.112. Solve∂2z
∂x2+∂2z
∂y2= cos 2x cos 3y
Solution: Given(D2 +D′2) z = cos 2x cos 3y
The auxillary equation is
m2 + 1 = 0
m2 = −1
m = i,−i
∴ C.F. is f1(y + ix) + f2(y − ix)
P.I. =1
D2 +D′2 cos 2x cos 3y
=1
D2 +D′21
2(cos(2x+ 3y) + cos(2x− 3y))
=1
2
(1
D2 +D′2 cos(2x+ 3y) +1
D2 +D′2 cos(2x− 3y)
)=
1
2
(1
−4− 9cos(2x+ 3y) +
1
−4− 9cos(2x− 3y)
)= − 1
26(cos(2x+ 3y) + cos(2x− 3y))
= − 1
26
(2 cos
(4x
2
)+ cos
(6y
2
))= − 1
13(cos 2xcos3y)
∴ The complete solution z = C.F. + P.I.Aliter :
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 73
P.I. =1
f(D2 +D′2)sin ax sin by (or) cos ax cos by
=1
f(−a2,−b2)sin ax sin by (or) cos ax cos by provided f
(−a2,−b2
)6= 0
∴ P.I. =1
D2 +D′2 cos 2x cos 3y
=1
−4− 9cos 2x cos 3y
∴ P.I. = − 1
13cos 2x cos 3y
Example 1.113. Solve(D2 +DD′ − 6D′2) z = cos(2x+ y)
Solution: The auxillary equation is
m2 +m− 6 = 0
i.e., (m− 2)(m+ 3) = 0
i.e., m = 2,−3
∴ C.F. is f1(y + 2x) + f2(y − 3x)
P.I. =1
D2 +DD′ − 6D′2 cos(2x+ y)
=1
−4−2+6cos(2x+y)
[Replacing D2 → −4, D′2 → −1, DD′ → −2
]=
1
0cos(2x+ y)
∴ P.I. = x1
2D +D′ cos(2x+ y)
P.I. = x2D −D′
4D2 −D′2 cos(2x+ y)
= x2D −D′
−16 + 1cos(2x+ y)
[Replacing D2 by − 4, D′2 by − 1
]= − x
15[2D cos(2x+ y)−D′ cos(2x+ y)]
= − x
15[−4 sin(2x+ y) + sin(2x+ y)]
=x
5sin(2x+ y)
∴ The complete solution z = C.F. + P.I.
Example 1.114. Write the P.I. of(D2 +DD′) z = sin(x+ y)
74 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Solution:
P.I. =1
D2 +DD′ sin(x+ y)
=1
−1− 1sin(x+ y)
[Replacing D2 by − 1, DD′ by − 1
]∴ P.I. = −1
2sin(x+ y)
Example 1.115. Solve the equation∂2z
∂x2− ∂2z
∂x∂y= sin x cos 2y
Solution: Given(D2 −DD′) z = sin x cos 2y
The auxillary equation is
m2 −m = 0
i.e., m = 0, 1
∴ C.F. = f1(y + 0x) + f2(y + x)
= f1(y) + f2(y + x)
P.I. =1
D2 −DD′ sinxcos2y
=1
D2 −DD′ .1
2(sin(x+ 2y) + sin(x− 2y))
=1
2
[1
D2 −DD′ sin(x+ 2y) +1
D2 −DD′ sin(x− 2y)
]=
1
2
[1
−1 + 2sin(x+ 2y) +
1
−1− 2sin(x− 2y)
]=
1
2
[sin(x+ 2y)− 1
3sin(x− 2y)
]∴ The complete solution z = C.F. + P.I.
Example 1.116. Solve(D2 + 3DD′ − 4D′2) z = cos(2x+ y)
Solution: The auxillary equation is
m2 + 3m− 4 = 0
i.e., (m− 1)(m+ 4) = 0
i.e., m = 1,−4
∴ C.F. = f1(y + x) + f2(y − 4x)
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 75
P.I. =1
D2 + 3DD′ − 4D′2 sin y
=1
D2 + 3DD′ − 4D′2 sin(0x+ y)
=1
0 + 0 + 4sin(0x+ y)
[Replacing D2 by 0, D′2 by − 1, DD′ by 0
]P.I. =
1
4sin y
∴ The complete solution z = C.F. + P.I.
Example 1.117. Solve(D3 +D2D′ −DD′2 −D′3) z = 3 sin(x+ y)
Solution: The auxillary equation is
m3 +m2 −m− 1 = 0
i.e., m = 1,−1,−1
∴ C.F. = f1(y + x) + f2(y − x) + xf3(y − x)
P.I. =1
D3 +D2D′ −DD′2 −D′33 sin(x+ y)
=1
−D −D′ +D +D′3 sin(x+ y)[Replacing D2 by − 1, D′2 by − 1, DD′ by − 1
]=
1
03 sin(x+ y)
∴ P.I. = x1
3D2 + 2DD′ −D′23 sin(x+ y)
= x1
−3− 2 + 13 sin(x+ y)[
Replacing D2 by − 1, D′2 by − 1, DD′ by − 1]
= −3
4x sin(x+ y)
∴ The complete solution z = C.F. + P.I.
Example 1.118. Solve(D3 +D2D′ − 2DD′2) z = sin(x− 2y)
Solution: The auxillary equation is
m3 +m2 − 2m = 0
m(m2 +m− 2) = 0
i.e., m = 0, 1,−2
76 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
∴ C.F. = f1(y) + f2(y + x) + f3(y − 2x)
P.I. =1
D3 +D2D′ − 2DD′2sin(x− 2y)
=1
−D −D′ + 8Dsin(x− 2y)[
Replacing D2 by − 1, D′2 by − 4]
=1
7D −D′ sin(x− 2y)
=7D +D′
−49 + 4sin(x− 2y)[
Replacing D2 by − 1, D′2 by − 4]
= − 1
45[7D sin(x− 2y) +D′ sin(x− 2y)]
= − 1
45[7 cos(x− 2y)− 2 cos(x− 2y)]
P.I. = −1
9cos(x− 2y)
∴ The complete solution z = C.F. + P.I.
i.e., z = f1(y) + f2(y + x) + f3(y − 2x)− 1
9cos(x− 2y).
Example 1.119. Solve(D2 − 3DD′ + 2D′2) z = sin x cos y
Solution: The auxillary equation is
m2 − 3m+ 2 = 0
(m− 1)(m− 2) = 0
i.e., m = 1, 2
∴ C.F. = f1(y + x) + f2(y + 2x)
P.I. =1
D2 − 3DD′ + 2D′2 sinx cos y
=1
D2 − 3DD′ + 2D′21
2[sin(x+ y) + sin(x− y)]
=1
2
[1
D2 − 3DD′ + 2D′2 . sin(x+ y) +1
D2 − 3DD′ + 2D′2 sin(x− y)
]=
1
2
[1
−1 + 3− 2. sin(x+ y) +
1
−1− 3− 2sin(x− y)
]
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 77
=1
2
[1
0. sin(x+ y)− 1
6sin(x− y)
]P.I. =
1
2
[x
1
2D − 3D′ . sin(x+ y)− 1
6sin(x− y)
]=
1
2
[x
2D + 3D′
4D2 − 9D′2 . sin(x+ y)− 1
6sin(x− y)
]=
1
2
[x2D + 3D′
−4 + 9. sin(x+ y)− 1
6sin(x− y)
]=
1
2
[x
5[2D sin(x+ y) + 3D′ sin(x+ y)]− 1
6sin(x− y)
]=
1
2
[x
5[2 cos(x+ y) + 3 cos(x+ y)]− 1
6sin(x− y)
]P.I. =
1
2
[x cos(x+ y)− 1
6sin(x− y)
]
∴ The complete solution z = C.F. + P.I.
i.e., f1(y + x) + f2(y + 2x) +1
2
[x cos(x+ y)− 1
6sin(x− y)
].
Example 1.120. Solve(D2 + 3DD′ + 2D′2) z = x2y2
Solution: The auxillary equation is
m2 + 3m+ 2 = 0
(m+ 1)(m+ 2) = 0
i.e., m = −1,−2
∴ C.F. = f1(y − x) + f2(y − 2x)
P.I. =1
D2 + 3DD′ + 2D′2x2y2
=1
D2[1 + 3D′
D + 2D′2
D2
]x2y2=
1
D2
[1 +
(3D′
D+
2D′2
D2
)]−1
x2y2
78 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
=1
D2
[1−
(3D′
D+
2D′2
D2
)+
(3D′
D+
2D′2
D2
)2
− . . .
]x2y2
=1
D2
[1− 3D′
D− 2D′2
D2+
9D′2
D2
]x2y2
=1
D2
[1− 3D′
D+
7D′2
D2
]x2y2
=1
D2
[x2y2 − 3
D(2x2y) +
7
D2(2x2)
]=
1
D2(x2y2)− 1
D3(6x2y) +
1
D4(14x2)
=1
12x4y2 − 1
10x5y +
7
180x6
=1
2
[x
5[2 cos(x+ y) + 3 cos(x+ y)]− 1
6sin(x− y)
]P.I. =
1
2
[x cos(x+ y)− 1
6sin(x− y)
](∵
1
Dnxm =
xm+n
(m+ 1)(m+ 1) . . . (m+ n)
)∴ The complete solution z = C.F. + P.I.
Example 1.121. Solve(D2 − 2DD′ + 2D′2) z = x2y2
Solution: The auxillary equation is
m2 − 2m+ 2 = 0
m =2±
√4− 8
2i.e., m = 1± i
∴ C.F. = f1(y + (1 + i)x) + f2(y + (1− i)x)
P.I. =1
D2 − 2DD′ + 2D′2x3y
=1
D2[1− 2D′
D + 2D′2
D2
]x3y=
1
D2
[1−
(2D′
D− 2D′2
D2
)]−1
x3y
=1
D2
[1 +
2D′
D
]x3y =
1
D2
[x3y +
2
D(x3)
]=
1
D2(x3y) +
2
D3(x3) =
x5y
20+x6
60
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 79
∴ The complete solution z = C.F. + P.I.
Example 1.122. Solve(D2 + 6DD′ + 5D′2) z = xy4
Solution: The auxillary equation is
m2 + 6m+ 5 = 0
(m+ 1)(m+ 5) = 0
i.e., m = −1,−5
∴ C.F. = f1(y − x) + f2(y − 5x)
P.I. =1
D2 + 6DD′ + 5D′2xy4 =
1
5D′2[
D2
5D′2 +65DD′ + 1
]xy4=
1
5D′2
[1 +
(6D
5D′ +D2
5D′2
)]−1
xy4
=1
5D′2
[1−
(6D
5D′ +D2
5D′2
)+
(6D
5D′ +D2
5D′2
)2
− . . .
]xy4
=1
5D′2
[1− 6D
5D′
]xy4 =
1
5D′2
[xy4 − 6y4
5D′
]=
1
5D′2 (xy4)− 6
25D′3 (y4)
=xy6
150− y7
875
∴ The complete solution z = C.F. + P.I.
Example 1.123. Find the general integral of the equation∂2z
∂x2+ 3
∂2z
∂x∂y+ 2
∂2z
∂y2= x+ y
Solution: Given(D2 + 3DD′ + 2D′2) z = x+ y The auxillary equation is
m2 + 3m+ 2 = 0
(m+ 1)(m+ 2) = 0
i.e., m = −1,−2
∴ C.F. = f1(y − x) + f2(y − 2x)
80 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
P.I. =1
D2 + 3DD′ + 2D′2 (x+ y)
=1
D2(1 + 3D′
D + 2D′2
D2
)(x+ y)
=1
D2
[1 +
(3D′
D+
2D′2
D2
)]−1
(x+ y)
=1
D2
[1−
(3D′
D+
2D′2
D2
)+
(3D′
D+
2D′2
D2
)2
− . . .
](x+ y)
=1
D2
[1− 3D′
D
](x+ y) =
1
D2
[x+ y − 3
D
]=
1
D2(x) +
1
D2(y)− 1
D3(3)
=x3
6+x2y
2− x3
2
∴ The complete solution z = C.F. + P.I.
Example 1.124. Solve(D2 + 4DD′ − 5D′2) z = x+ y2
Solution: The auxillary equation is
m2 + 4m− 5 + 2 = 0
(m+ 5)(m− 1) = 0
i.e., m = 1,−5
∴ C.F. = f1(y + x) + f2(y − 5x)
P.I. =1
D2 + 4DD′ − 5D′2 (x+ y2) =1
−5D′2(−D2
5D′ − 4D5D′ + 1
)(x+ y2)
= − 1
5D′2
[1−
(4D
5D′ +D2
5D′
)]−1
(x+ y2)
= − 1
5D′2
[1 +
(4D
5D′ +D2
5D′
)+
(4D
5D′ +D2
5D′
)2
− . . .
](x+ y2)
= − 1
5D′2
[1 +
4D
5D′
](x+ y2) = − 1
5D′2
[x+ y2 +
4
5D′
]= −1
5
1
D′2 (x) +1
D′2 (y2) +
4
5
1
D′3 (1) = −1
5
(xy2
2+y4
12+
2y3
15
)∴ The complete solution z = C.F. + P.I.
i.e., z = f1(y + x) + f2(y − 5x)− 1
5
(1
2xy2 +
1
12y4 +
2
15y3).
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 81
Example 1.125. Solve(D2 − 4DD′ + 4D′2) z = ex−2y cos(2x− y)
Solution: The auxillary equation is
m2 − 4m+ 4 = 0
i.e., m = 2, 2
∴ C.F. = f1(y + 2x) + xf2(y + 2x)
P.I. =1
D2 − 4DD′ + 4D′2ex−2y cos(2x− y)
= ex−2y 1
(D + 1)2 − 4(D + 1)(D′ − 2) + 4(D′ − 2)2cos(2x− y)
= ex−2y 1
D2 + 10D − 4DD′ − 20D′ + 4D′2 + 25cos(2x− y)
= ex−2y 1
−4 + 10D − 8− 20D′ − 4 + 25cos(2x− y)
= ex−2y 1
10D − 20D′ + 9cos(2x− y)
= ex−2y ((10D − 20D′)− 9) cos(2x− y)
(10D − 20D′)2 − 81
= ex−2y (10D − 20D′ − 9) cos(2x− y)
100D2 − 400DD′ + 400D′2 − 81
= ex−2y (10D − 20D′ − 9)
−400− 800− 400− 81cos(2x− y)
= −ex−2y
1681(−20 sin(2x− y)− 20 sin(2x− y)− 9 cos(2x− y))
P.I. =1
1681ex−2y [40 sin(2x− y) + 9 cos(2x− y)]
∴ The complete solution z = C.F. + P.I.
Example 1.126. Solve(D2 −DD′ − 2D′2) z = (y − 1)ex
Solution: The auxillary equation is
m2 −m− 2 = 0
(m− 2)(m+ 1) = 0
i.e., m = 2,−1
∴ C.F. = f1(y + 2x) + f2(y − x)
82 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
P.I. =(y − 1)ex
D2 −DD′ − 2D′2
= ex1
(D + 1)2 − (D + 1)D′ − 2D′2 (y − 1)
= ex1
D2 + 2D + 1−DD′ −D′ − 2D′2 (y − 1)
= ex[1 +
(D2 + 2D −DD′ −D′ − 2D′2)]−1
(y − 1)
= ex[1−
(D2 + 2D −DD′ −D′ − 2D′2)+ . . .
](y − 1)
= ex [1 +DD′ +D′] (y − 1)
= ex [y − 1 +D(1) + 1]
= exy
∴ The complete solution z = C.F. + P.I.i.e., z = f1(y + 2x) + f2(y − x) + yex.
Example 1.127. Solve(D2 +DD′ − 6D′2) z = y cosx
Solution: The auxillary equation is
m2 +m− 6 = 0
(m+ 3)(m− 2) = 0
i.e., m = 2,−3
∴ C.F. = f1(y + 2x) + f2(y − 3x)
P.I. =1
D2 +DD′ − 6D′2y cosx
=1
(D − 2D′)(D + 3D′)y cosx
=1
D − 2D′
∫(c+ 3x) cos xdx where y = c+ 3x
=1
D − 2D′ [(c + 3x) sin x+ 3 cos x]
=1
D − 2D′ [(y − 3x+ 3x) sin x+ 3 cos x] ∵ c = y − 3x
=1
D − 2D′ (y sinx+ 3 cos x)
=
∫[(c1 − 2x) sin x+ 3 cos x]dx where y = c1 − 2x
= [(c1 − 2x)(− cosx)− (−2)(− sinx)] + 3 sinx
= −(c1 − 2x) cos x− 2 sin x+ 3 sin x
= −y cosx+ sinx
P.I. = sin x− y cosx
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 83
∴ The complete solution z = C.F. + P.I.
1.8.4 Exercises & Solutions
Solve the following partial differential equations:
1.∂2z
∂x2+
∂2z
∂x∂y− 2
∂2z
∂y2= 0 [Ans: z = f1(y + x) + f2(y − 2x)]
2. 2∂2z
∂x2+ 5
∂2z
∂x∂y+ 2
∂2z
∂y2= 0 [Ans: z = f1(y − 2x) + f2(2y − x)]
3.(D3 − 2D2D′) 2e2x + 3x2y[
Ans: z = f1(y) + xf2(y) + f3(y + 2x)1
4e2x +
1
60
(3x5y + x6
)]4. r + s− 6t = cos(2x+ y)[
Ans: z = f1(y − 3x) + f2(y + 2x) +1
5x (sin 2x+ y)
]5. r + 2s+ t = 2(y − x) + sin(x− y)[
Ans: z = f1(y − x) + xf2(y − x) + x2y − x3 +1
2x2 (sinx− y)
]6.(D2 + 2DD +D′2) z = 2 cos y − x sin y
[Ans: z = f1(y − x) + xf2(y − x) + x sin y]
7.(D2 + 3DD′ + 2D′2) z = x+ y[
Ans: z = f1 (y − x) + f2 (y − 2x)− x3
3+x2y
2
]8.(2D2 − 2DD′ +D′2) z = 2e3y + ex+y + y2[Ans: z = f1
(y +
1 + i
2x
)+ f2
(y +
1− i
2x
)+
2
9e3y + ex+y +
x2y2
2
]9.(D2 −DD′) z = cos x cos 2y[
Ans: z = f1(y) + f2(y + x) +1
2cos (x+ 2y)− 1
6cos(x− 2y)
]10.
(D2 − 2DD′) z = e2x + x3y[
Ans: z = f1 (y) + f2 (y + 2x)1
4e2x +
x5y
20+x6
60
]
84 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
1.9 Non-Homogeneous Linear P.D.E.
A linear partial differential equation in which all the partial derivativesare not of the same order is called non-homogeneous linear partialdifferential equation.Consider the equation f(D,D′)z = F (x, y) (1)If f(D,D′) is not homogeneous, then (1) is the non-homogeneous linear
equation. As in the of homogenous linear equation , the completesolution is
z = C.F. + P.I.
[Finding P.I. of Non-Homo. P.D.E. is same as in Homo. P.D.E.]
1.9.1 To find the complementary function (C.F.)
Case(1): If (D −m1 − C1) z = 0, then C.F. = eC1xf1 (y +m1x)
Case(2): If (D′ −m1D − C1) z = 0, then C.F. = eC1yf1 (x+m1y)
Case(3): If (D +m1D′ + C1) z = 0 ⇒ [D − (−m1)D
′ − (−C1)] z = 0,
then C.F. = e−C1x (f1(y −m1x))
Case(4): If (D′ +m1D + c1) z = 0 ⇒ [D′ − (−m1)D − (−c1)] z = 0, then
C.F. = e−C1yf1 (x−m1y)
Case(5): If (D −m1D′ − c1) (D −m2D
′ − c2) z = 0, then
C.F. = eC1xf1 (y +m1x) + eC2xf2 (y +m2x)
Case(6): If (D −m1D′ − c1)
2z = 0, then
C.F. = eC1xf1 (y +m1x) + xeC1xf2 (y +m1x)
Case(7): Put D = h and D′ = k in f (D,D′) = 0, then we get an
Aux. eqn.
f(h, k) = 0 (2)
Solve (2) and find the value of h interm of k or k interm of h.
Let the n values of h be denoted by f1(k), f2(k), f3(k), . . . , fn(k).
Then
C.F. =∑
c1ef1(k)x+ky +
∑c2ef2(k)x+ky
+∑
c3ef3(k)x+ky + ...+
∑cnefn(k)x+ky
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 85
1.9.2 Examples of Non-Homogeneous Linear P.D.E.
Example 1.128. Solve (D − 2D′ − 1) z = 0
Solution: C.F. = exf1(y + 2x) [∵ C1 = 1,m1 = 2 as case(1)]
Example 1.129. Solve (D′ − 2D − 3) z = 0
Solution: C.F. = e3yf1(x+ 2y) [∵ C1 = 3,m1 = 2 as case(2)]
Example 1.130. Solve (D + 3D′ + 1) z = 0
Solution: Given (D + 3D′ + 1) z = 0 ⇒ [D − (−3)D′ − (−1)]C.F. = e−xf1(y − 3x) [∵ C1 = 1,m1 = 3 as case(3)]
Example 1.131. Solve (D′ + 5D + 6) z = 0
Solution: Given (D′ + 5D + 6) z = 0 ⇒ [D′ − (−5)D − (−6)]C.F. = e−6yf1(x− 5y) [∵ C1 = 6,m1 = 5 as case(4)]
Example 1.132. Solve (D − 4D′ − 1) (D′ + 5D − 3) z = 0
Solution: Given (D − 4D′ − 1) [D′ − (−5)D − 3] z = 0C.F. = exf1(y + 4x) + e3yf2(x− 5y)
Example 1.133. Solve (D − 2D′ + 2)3z = 0
Solution: Given (D − 2D′ + 2)3z = 0 ⇒ [D − 2D′ − (−2)]
3z = 0
C.F. = e−2xf1(y + 2x) + xe−2xf2(y + 2x) + x2e−2xf3(y + 2x)
Example 1.134. Solve(D2 −DD′ +D′ − 1
)z = 0
Solution:Consider
(D2 −DD′ +D′ − 1
)=(D2 − 1
)−DD′ +D′
= (D − 1)(D + 1)−D′(D − 1)
= (D − 1)(D −D′ + 1)
= [D − 0D′ − 1][D −D′ − (−1)]
C.F. = exf1(y) + e−xf2(y + x)
(OR)
Put D = h and D′ = k in
f (D,D′) = 0 ⇒(D2 +DD′ +D′ − 1
)z = 0
then we get an Aux. eqn.f(h, k) = 0 ⇒
(h2 + hk + k − 1
)= 0 (1)
Solve (1) and find the value of h interm of k or k interm of h.h2 + hk + k − 1 = 0
86 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
Now, h =k ±
√k2 − 4(k − 1)
2=k ±
√(k − 2)2
2=k ± (k − 2)
2i.e., h = k − 1, 1
C.F. =∑
c1e(k−1)x+ky +
∑c2e1x+ky
=∑
c1e−xek(x+y) +
∑c2ex+ky
= e−xf1(x+ y) + exf2(y)
Example 1.135. Solve(D2 −D′2 − 3D + 3D′) z = 0
Solution: Put D = h and D′ = k in
f (D,D′) = 0 ⇒(D2 −D′2 − 3D + 3D′) z = 0
then we get an Aux. eqn.f(h, k) = 0 ⇒
(h2 − k2 − 3h+ 3k
)= 0 (1)
Solve (1) and find the value of h interm of k(h2 − k2 − 3(h− k)
)= 0
(h− k)(h+ k)− 3(h− k) = 0
(h− k)(h+ k − 3) = 0
h = k, 3− k
C.F. =∑
c1ekx+ky +
∑c2e(3−k)x+ky =
∑c1ek(x+y) +
∑c2e3x+k(y−x)
= f1(x+ y) + e3xf2(y − x)
Example 1.136. Solve(2D2 −DD′ −D′2 + 6D + 3D′) z = 0
Solution: Put D = h and D′ = k
A.E.(2h2 − hk − k2 + 6h+ 3k
)= 0
i.e.,[2h2 + h(6− k) + (3k − k2)
]= 0
h =(k − 6)±
√(6− k)2 − 8(3k − k2)
4
=(k−6)±
√k2−12k+36−24k+8k2
4=
(k − 6)±√9k2 − 36k + 36
4
=(k − 6)± (3k − 6)
4=
(k−6)+(3k−6)
4,(k−6)−(3k−6)
4
= k − 3,−k2
MA6351 Transforms and Partial Differential Equations by K A Niranjan Kumar 87
C.F. =∑
c1e(k−3)x+ky +
∑c2e
−k
2x+ ky
=∑
c1e−3xek(x+y) +
∑c2e
k
(y−x
2
)
= e−3xf1(x+ y) + f2
(y − x
2
)1.10 Assignment III[Partial Differential Equations]
1. (i) Form a partial differential equation by eliminating a and b from theexpression (x− a)2 + (y − b)2 + z2 = c2.(ii) Find the partial differential equation of all planes which are at aconstant distance a1 from the origin.
2. (i) Form the partial differential equation by eliminating arbitraryfunction from z = xf(2x+ y) + g(2x+ y).(ii) Form the PDE by eliminating the arbitrary function φ fromφ(x2 + y2 + z2, ax+ by + cz
)= 0.
3. Solve the following:(i) p(1 + q)z = qz
(ii) x2p+ y2q = 0 (iii) xp+ yq = 0
4. Solve the following:(i) z = px+ qy + log pq (ii) z = px+ qy + pq
(iii) z = px+ qy +√
1 + p2 + q2 (iv) z = px+ qy + p2q2
5. Solve the following:(i) p(1 + q) = qz (ii) x4p2 − yzq = z2
(iii) x4p2 + y2qz = 2z2 (iv) z2(p2x2 + q2
)= 1
6. Solve the following:(i) p2 + q2 = x2 + y2 (ii) z2
(p2 + q2
)= x2 + y2
(iii) 4z2q2 = y + 2zp− x (iv) p2 + q2 = z2(x2 + y2
)7. Solve the following:
(i) x2(y − z)p = y2(z − x)q = z2(x− y)(ii)
(x2 − yz
)p+
(y2 − zx
)q = z2 − xy
(iii) (y − xz)p+ (yz − x)q = (x+ y)(x− y)
88 Unit I - PARTIAL DIFFERENTIAL EQUATIONS (P.D.E.)
(iv)(x2 − y2 − z2
)p+ 2xyq = 2zx
(v) x(z2 − y2
)p+ y
(x2 − z2
)q = z
(y2 − x2
)(vi) x
(y2 + z
)p+ y
(x2 + z
)q = z
(x2 − y2
)(vii) (3z − 4y)p+ (4x− 2z)q = (2y − 3x)(viii) (x− 2z)p+ (2z − y)q = y − x(ix) x(y − z)p+ y(z − x)q = z(x− y)
8. Solve the following:(i)(D3 +D2D′ −DD′2 −D′3) z = e2x+y + cos(x+ y)
(ii)∂3z
∂x3− 2
∂3z
∂x2∂y= ex+2y + 4 sin (x+ y)
(iii)(D3 − 7DD′2 − 6D′3) z = cos (x+ 2y) + 4
(iv)(D2 + 2DD′ +D′2) z = sinh(x+ y) + ex+2y
(v)(D2 + 2DD′ +D′2) z = x2y + ex−y
(vi)(D2 −DD′ − 2D′2) z = 2x+ 3y + e3x+4y
(vii)(D2 −D′2) z = ex−y sin(x+ 2y)
(viii)(D2 − 5DD′ + 6D′2) z = y sinx
(ix) r + s− 6t = y cosx
9. Solve the following:(i)(D2 −D′2 − 3D + 3D′) z = xy + 7
(ii)(D2 + 2DD′ − 2D − 2D′2) z = sin(x+ 2y)
(iii)(2D2 −DD′ −D′2 + 6D + 3D′) z = xey
(iv)(D2 + 2DD′ +D′2 − 2D − 2D′) z = e3x+y + 4
(v)(D2 +D′2 + 2DD′ + 2D + 2D′ + 1
)z = e2x+y