UNIT 7

Structure
7.1 Introduction
Objectives
7.2 What is a Conicoid?
..
.

7.3 Change of Axes

Translation of Axes
Projection
Rotation of Axes
7.4 Reduction to Standard h r m
7.5 Summary

7.1 INTRODUCTION .

You have seen in Block 1 that the general equation pf secondrjdegret in two
variables x and y represents a conic. In analogy with- tl$s,w~can ask: what will a
general second degree equation in three variables r~present?In Black 2 you have
studied some particular forms of second degree equations in three variables, namely,
those representing spheres, cohes and cylinders'. In this unit we study the most
general form of a second degree equation in three variables. The'surfaces,generated
by these equations are called quadrics or copicoids. This name js apt because, as
you will see in Unit 8,iiky &n be formed-by revolving a conic about a line called ':
an axis. . -
Alexis Clairaut (1713-17651, a French mathematician, was one of the pioneers to
study quadric 'surfaces. H e specified'that a surface, in general, can b$ represented
by an equation in three variables. He &esented his ideas in his book 'Recherche
Sur Les Courbes a Double Courbure' in which he gave the equations of several
conicoids like the sphere, cylinder, hyperboloid and ellipsoid.
We start this unit with a small section in which we define a conicoid. In the next
section we discuss rigid body motions in a three-dimensional system. We shall
consider two types of transformations: translation of axes and rotation of axes. You
can see that a conicoid remains unchanged in shape and size under these
transformations. Lastly, we shall discuss how to reduce the equation of conicoid into
a more simple form.

' Objectives
After studying this unit you should be able to
define a general conicoid;
obtain tht new coordinates when a given cooidinate system is translated or
. rotated; ,-

use the fact that translation and rotation of axes are rigid body motions;
check whether a given coniqoid has centre or not;
prove anct apply the fact that if a conicoid has a centre, then it can be reduced
to standard form.
5 >-
 7.2 AT IS A COMGOID?
In this section we shall define surfaces in a three-dimensional coordinate system
which are analogous to conic sections in a two-dimensional system.
Let us start with a definition.
Definition: A general secoad degree equation ih three variables is an equation of the L
form
axz+by2 +~z~+2fyz+2~zx+2hxy+2ux+2vy+2wz+d = 0, .... (1)
where a, b, c, d, f, g, h, u, v, w are real numbers and at least one of a, b, c, f, g,
h is non-zero.
Note that if we put either z = k; a constant, x = k or y = k, in (I), then the
equation reduces to a general second degree equation in two variables, and
therefore, represents a conic.
Now we shall see what a general second degree equation in three variables
represents. Let us first consider'some particular cases of (1).
Case1 :Supposeweput a = b = c = 1 a n d g = h =f = Oin(l).Thenweget the
equation
+
x2 y2 + zZ 2ux + +
2vy + 2wz i- d=0 .... (2)
Does this equation seem familiar to you? In Unit 6 you saw that if 1'

+ +
u2 v2 wi- d > 0, then (2) represents-a sphere with centre (- f, -g, -h) and
radius V u2+v'+w2-d.

Case 2 : Suppose we put u = v = w = d = C, in (I), then we get

+ + + +
ax2 by2 cz2 2fyz 2gzx 2hxy = 0. +
What does this quation represent? You know from Unit 6 that this equation
represents a cone.
Case 3 : If we put a = b = 1, h = 0 and z = k in (I), then it reduces to
x2 +y2 + 2ux + 2vy + d = 0, z=k . (3) ...
This represents a right circular cylinder (see Unit 6, Sec.6.4).
Similarly you can see that if we put x = k or y = k and a = b = 1, h = 0, then
again (3) represents a cylinder.
We will discuss the surfaces represented by (1) in detail in the next uriit.
The particular cases 1 , 2 and 3 suggest that the points whose coordinates satisfy (1)
lie on a surface in the three-dimensional system. Such a surface is called a conicoid
or a quadric. Algebraically, we define a conicoid as follows:
Definition : A conicoid (or quatll-ic surfacrj 1 1 1 the X\r'%-ca~~rcll~~atc
system 1s thc
set S of points (x, y, z) E R that satisfy a general second degree equation in threc
3

variables.
1 So, for example, if

+
F(x.y .z) = ax'+ by2+cz%2fyz 2gzx+2hxy+2ux+2vy+?wz+d=O
is thc second dcgree equation, then
S = .((x,y,z)E @ I F (x,y,z)=O)
Note that S can be empty. For example, if F (x,y,z) = ~ % ~ * + z ~ + 1 = 0 ,
then
S = {(x,y,z) I F (x,y,z) = 0) = 4, the empty set.
In such cases we call the conicoid an imaginary conicoid.
Since the above expression is very lengthy, for convenience we often denote this
conicoid by F (x,y,z,) ='o.
Note : In future, whenever we use the expression F (x, y, z) = 0, we will mean the
cquqtion (1).
In this unit you will see that we can always reduce F(x, y, z) to a'rnuch smaller
quadratic polynomial. To do this we need to transform the axes suitably. Let us see
what this means.
7.3 CHANGE OF AXES
In.Block 1, you saw that a general second degree equation can be traosfomed into
the standard form using a suitable change of axes. You also saw that these standard
forms are very useful for studying the geometrical properties of the conic concerned.
Here we shall show that in th'e case of the three-dimensional system also we can
transform an equation F (x, y, z) = 0 into the corresponding standard form by
means of an appropriate change of coordinate axes. As in the case of the
two-dimensional system, the transformations that we apply are of two types : change
of origin and change of direction of axes. Let us consider these one by one in the
following sub-sections.

7.3.1 Translation of Axes

Here we shall discuss how the coordinates of a point in the three-dimensional system
get affected by shifting the origin from one point to another point, without changing
the direction of the axes. The procedure is thg same as in the two-dimensional case.
Let OX, OY, O Z be the coordinate axes of a three-dimensional system. What
happens when we shift the origin from 0 to another point Q' (see Fig.l)?

Fis.1 : TnnaPtbo of axes through 0'.

Let the coordinates of 0' in the XYZ-system be (,;. b, c). Let O'X', O'Y' and
O'Z' be the new axes which ;ire parallel to the OX. O Y and OZ axes. Suppose P
is a point as given in Fig.1. Let the coordinates of P inathe XYZsystem be
(x, y, z) and in the X'YtZ'-systembe (x', y', 2'). W e first find a relationship
between z and z'. For this we draw a line through P, which is perpendicular t o the
XOY plane. Let it cut the planes XOY and XtO'Y'in L and L', respectively.
Then we have
PL = z and PL' = 2 ' .
From Fig.1 we have
PL = PL' + L'L
Now L'L is also the length of the perpendicular from the point 0' to the XOY
I
plane. Therefore L'L = c
Thus we have z = z! + c.
Similarly we get x = x' +a and y = y' + b.
Hence if we shift the origin from O(0, 0, 0 ) to another point 0' (a, b , c) without
changing the direction of the axes, then the new coordinates of any point
*P(x, y,.z) with respect to the origin 0' will be
X ' = x - a, y' = y - b and'i' = z - c .......(4)
So, for example what will the new coordinates (x', y', 2') of a point P (x, y, z) be
-
when we shift the origin to (2, -1, I)? They will be x' = x 2, y' = y 1 and +
z' = z - 1.
 Conlcoids When we transform the axes in such a way, we say we have shifted the origin to
(2, -1, 1) or translated the axes through (2, -1, 1).
Now, what will the effect of such a transformation be on any equation in x, y, z?
. ,. :,. If, in an equation in the XYZ-system we replace x, y, z, by x f + a , y t + b , z' +c, then
we get the new equation*inthe XrY'Z'-system. For example, when we shift the
origin to (2, -1, I), then the equation of the plane 3x+2y-z = 5 will be transformed
into 3x'+2yr - z' = 2.
Note that the respective coefficients of x, y, z and of x', y', z' remain llnchanged
under a shift in origin. Thus the direction ratios of the normal to a plane dn not '
change when we shift the origin (Recall the definition of direction ratios fre3aA 1 Jnit
4.) Can you guess why this happens? This is obviously because we are nor snifting
the direction of the coordinate planes, we are only shifting the origin
Now let us consider the effect.of translation of axes on a general second degrek
equation.
Theorem 1 : Let the coordinates of a given surface S in a given coordinate system
XYZ satisfy a second degree equation in three variables. Let us shift the origin from
0 to another point 0' giving rise to a new coordinate system X'Y'Z'. Then S is
still represented by a general second degree equation in three variables in the new

In (5) the group of terms

- coordinate system.

having degree 2, namely ax2 +

Proof : Let the given surface S satisfy the equation
+ +
byZ c s t 2fyz+ 2gzx 2hxy F(x,y,z) = ax 2+by 2+cz 2+2fyz+2gzx+2hxy+2ux+2vy+2wz+d = 0.
is called the second degree part For convenience we write the equation in the form
of the equation and the group F(x,y,z) = Cax2+Z2fyz+2Xux+d = 0 .... ( 5 )
o'fthe terns 2ux + 2vy + 2wzis,
called,the h e a r part of the Let (p, q, r) denote the coordinates of Or in the XYZ-system. Consider now the
equation. new system of coordinate axes O'Xr, O'Y', O'Z' parallel to the given system with
origin 0'. You know. that the relation between the coordinates in the original and
new systems are given by

.
X = xl+p
Y = yr+q
z = zl+r 1
J
Substituting these elipressions for x, y, z in (5), we pet
C a ( ~ + ~ ) ~ + Z 2 f ( y + ~ ) ( z + r ) + Z 2 u ( x + p )=
+ d0.
.... (6)

Now we expand the above expression and simplify by collecting like terms. We get
ax'2+by'2+cz'2+2fy'z'+2gz'x'+2hx'y'+2u'x'+2v'y'+2w'z'+d = 0, .... (7)
where
u' = (ap hq+ + gr) + u
+ +
u' = (hp bq fr) v +
w' = (gp fq+ + + er) w .... (8)
and d'=ap2~bq2+c?+2fqr+2grp+2hpq+2~p+2vq+2wr+d
Let G (x', y', z') denote the expression in the left hand side of (7). Then We see that any
point (x', y', z') belongingto §.satisfies the equation G (x', y', z') = 0,which is again a
general second degree equation.
Hence we have proved the result.
If you compare (5) and (7), then you will see that the second degree part of the
equation remain unchanged whereas the linear part changes.' Hence, wesacan
conclude that

( under tlfe transformation of shifting the origin of a coordinate system, the second I
[ degree part of a general second degree equation does not change. '
J
Why don't you try s o m e ' e a e ~ ~ a enow?
s .

What will the new equation of a right circular cone, with vertex 0, axis
..El) a)
-- OZ and semi-vertical angle a be, when we shift the origin t o (-1,1; O)?
.8I b) What does the new equation represent? Sketch the surface. .
332) Obtarn the transformed equation of the following equations when the origin Genemi Thwv of Conicoids
is shifted to (1, -3, 2).
+ +
a) x2 y2 z2 - 4 x + 6y - 27 5 = 0. +
b) x2 - 2y2 - 32 = 0.

Let us now consider the transformation in whic!l.the direction of the axes is changed.
For this we need to understand the concept of a projection. So let us first see what
a projection is.
L
7.3.2 Projection
'In this section we shall talk about an important concept in geometry. This concept
has even been used by artists through centuries for giving depth to their works of
art. Let us define it.
Definition : Let A be a point in the XYZ coordinate system.
The projection of A on a line segment PQ is the foot of the perpendicular drawn
from the point A to the line.
From Fig.2 you can see that the prdjection of A on PQ will be the point 0 where
the plane through A and perpendicular to PQ meets the line PQ.

Fig. 2 :'Pmj&tion of the point A on the line PQ i s 0

Dtfinition : The projection of a line segment A B on a line PQ is the segment A' l3'
of the line PQ. where A' and B' are respectively the projections of the points A
and B on the line PQ. (see Fig. 3)

Fig. 3:The Projection of the line segment AB on the line PQ is the line segment A'B'
 Conicoids : Remark .: From Fig.4 you can see that the length of A rB' = JAB(cos0, where.0
is the angle between AB and A'B' We also call this number the projection of AB
on PQ.
So what will the projection of BA be? It will be (BAI cos (n 0) that is - AB +
cose. .Thisshows that the projection can be positive or negative depending upon the
direction of the line segment.
Whenever you come across the term projection, from the context it will be clear
A whether we are referring to a line segment or a number.
I
* I
I 1 We shall now state a simple result which we shall need white discussing rotation of
axes.
A'. . B' Theorem 2: Suppose Al, A2,.... .., A,.,are n points in space. Then the algebraic sum
~ i g 4:
. of the projections of AIA2,AzA3,.. ....., An-,An on a line is equal to the projection
of A, A, on that line.
We will not prove this result in general,here; but shall give you a proof in a
particular case only. The proof, in any situation, is'on the same lines,
Proof: Consider the situationin Fig.5 concerning 4 points Al, AZ,AS,A4 and,bheir
projections At1,AI2, At3,Af4on a given line L.

Af3 A', Aj *'4

Fig. 5: Projections of the line segments A,A*,A2A3, A3A4 on a $vcn line L.

Then .--- .. --
AllAP2f At2Af34 A13A14= $ ' 1 ~ ' 2 ( , \ ' \ , ,\' ,\',) ()\'I)\', '
A'1A'2f ArZAf4) AllAf4
This shows that the sum of the projections of the line segments A1A2,A2A3,A3A4,
is equal to the projection of A1A4.SO, we have proved the result for the situation
in' Fig. 5.
Now consider another useful result involving projections.
The~r~!m 3 : Let P (xl,yl, 2 , ) and Q (x2, yz, z2)be two points in the XYZ coordinate
system. Then the projection of the line segment PQ on a line with direction cosines
I, m, n is given by
- - + (22 - z1)n.
(x2 XI)] -I- ( ~ 2 ~ l ) m
Prmf : In Unit 4, Sec. 4.3.2,-you have seen that the direction ratios of the line
-
joining P and Q are (x2 - xl), (y2 y,), (22 21). -
Lct ~PQIdenote the d&tagce&etwen P and_Q,-lye.,
--
I PQ ( =. .t/.'(x2 -. x , ) ~%-lyZ +
-' yl;).Z (z2 - z ~ ) ~ .
Then the direction cosines of the line Segment BQ are
%-*I
7
Y2-Yl
-7
-
% 21
-
IPQI IPQI IPQI
10
. Let 0 be the angle between the lines PQ and L. Then the projection of the line
segment PQ on the line L is (PQ(cos 0. But, from Unit 4, Sec. 4.3.3, we laave
 x2- xi1 z in
cos B = - + Y2- Ylm + 22-

lPQl IPQl . IPQI

Therefore, the required projecllon = (x2 xl)l - + (y2 - y!)m f (zi - zl)n
For example, wha$will the projection of the line segment joining 0 (0, 0, 0,) to the
point P (5, 2, 4) on the line hai.ng 2, -3, 6 as,its direction ratios be ? We know that
the direction cosines of the line with direction ratios 2, -3. 6 is.

3
Thus, the projection of OF = 5 X 27 + 2 x (-$ + 4 x 57 = 4.
Now here is an exercise for you.

E3) Let P(6, 3, 2), Q(5, 1, 4), R(3, -4, 7) and S(0, 2, 5) be four points in space..
Find the projection of the line segment PQ on RS.

Now we are in a position to discuss how the coordinates in space are affected by
changing the direction of tlre axes without changing the origin.
7.3.3 Rotation of Axes
Let us now consider the transformation of coordinates when the coordinate system
is rotated about the origin through an angle 8. Eet.the original system be OXYZ.
Suppose we rotate the coordinate axes through an angle 8 in the anti-clockwise
direction. Let OX', OY', OZ' denote the new coordinate axes (see Fig. 6). Suppose
the direction coslnes of OX', OY' and OZ' be l1,ml, nl;
12,1112, n2; I,,m3, n3, respectively.

are obtnined by rotating the axes OX,OY and OZ through an angle 0

Fig.6 : The axes OX', OY' ~ t l d0%'

Let, P be any point in space haviirg coordinates (x, y, z) and (x', y', 2') with respect
to the old and new coordinate systems. Let PN be the perpendicular from P on
OY'. Then
ON = y'
The line segnleut ON i s also the projectioi~of OP on the line 0 ~ ' w i t l ldirection
cosines 1,. rn,, n,. Therefore, by Theorem 2, we have
O N = (x-0)12 t (y-O)m2 + (2--U)n2.
Hence wc get
y' = x12 + ym2 + znz .... (9)
Similarly we can show that
x' = xl, + yml + zn, .... (10)
and
Z' = x13 + ym3 + zn3 .... (1 1)
Therefore given ( 4 , y, z) and the direction cosines of the new co~rdinateaxes, we
can get the new coordinates (x', y', z') using equations (9), (10) and (11).
Now how can we find x, y, z in terms of'x', y'; z'? For this we draw PM
perpendicular to OY (see Fig.6). Then
O M = Y. +.
O M is also equal to the projection of OP on OY. Now let us see what the directibn
cosiries of OY with respect to the new coordinate axes OX', OY', 02' are. We
kn0.w that the direction cosines of OX:, OY', OZ' with respect to OY are m , , m,
and m3. D o you agree that the direction cosines of OY with respect-to OX', OY', '
O Z ' are also m l , m2, m3?(We leave this as an exercise for you to verify.) Therefore.
by Theorem 2, we get
+
4' = OM = (x' - O)ml + ( y t - O)m2 (z' - O)rn3,
i.e., y = ntlxf + m2y' + mnz' ...(12)
Similarly, we get
x = 1 , ~+ ' lzyl + f3z' ...(13)
and
+
z = nlxl nlyr n3z1 + ...(14)
Hence (12), (13) and (14) give the coordinates of x, y, z in terms of x ' , y' and z'.
You may find that these equations are not easy to remember. FO; easy reference,
we arrange the equations in a table, as shown in Table 1.
Table 1: Transformation of ctwrdinates

Note that for finding x, y, z we make use of the elements in the respective columns,
and to find x', y' and zf we make use of the elements in the respective rows.
Let us consider an example.
Example 1 : bind the new equation of the conicoid
3x2 + 5y2 + 3z2 + 2yz + 2zx + 2xy = 1
when the coordinate system is transformed into a new system with the same origin'
and with the coordinate axes having direction ratios -1, 0, 1; 1, -1, 1 ; 1, 2, 1 with
respect to the old system.
Solution : The given surface is
.+ + +
3x2 5y2 3 9 2y2 + 222: 2xy = 1 + ....(15)
Let OX' Yf Z' be the new coordinate system. Then the direction cosines of OX',
OY' and OZ' with respect to the original axes are
1
- 0, -.1 1-. - -.1 -.1 -.1 2 -
1; respectively.
d- .d&&&&dq
We form the rransformation table:

From the table we. have

 . .
Gerwral Theory of Conicoids

To find the new equation, we substitute the expressions of x, y, z in (15). Then we

--.
get

Simplifying each term of the above eirpression and collecting the coefficients of like
terms, we get
+ +
2xf2 3yf2 62'' = 1.
This is the new equation of the conicoid.
You can do the following exercises on the saine lines.

E4) ~ i i the
d new equation of the I'ollowing conicoids when the coordinates system
is changed into a new system with the same origin and direction ratios 1, 2,
3; 1, A2, .I; 4, 1, -2; with respect to the old system.
a) x2 - 5y2 + z2 = 1.
b) x2 + y' + z' -2xy - 2yz +2zx + x -y -I- z. = 0

E5) For the conircoid in Example 1 and E4, calculate the sum of coefficients of the "
square terms in the original equation and in the new equation. Can you infer
a anything from the outcome?
E6) What will the: new equation of thd plane x+y+z = 0 be when the coordinate
system 'XYZ is transformed into another coordinate system X'Y'Z' by the
following equations?

E7) Does a cone remain a cone under ratation of axes? Give reasons for your
answer.

Now let us consider the efiect of cotation of axes on F(x,y,z) = 0.

Theorem 4 : Let S be a conieoid satisfying a second degree equation in a coordinate
System XYZ is transformed into another coordinate system X'Y'Z' by the
direction of axes, without changing the origin, S will still be represented. by a second
degree equation.

Proof : Gppose S is represented by the second degree equation

+
Zax2 + 22fyz + Z 2ux d = 0 in the XYZ-systems.

Let 11, ml, n,;4, m2, n2;13, m3, n3 be the direction cosines of the new coordinate axes
OX' OY', 02' respectively. Then you know that the coordinates (x, y, z),
(x', y', z') of a point in the old and new system respectively, satisfy the following
relationship.
 Let ";-substitute these expressions in the given equation. We consider the second
degree parts and linear parts of the equatibn separately.
The second degree part is Sax2 + '22fyz. When we substitute the expressions for
x,y,z in this part, we get
-
{Ca(E;xl-t!,yl + l,z.')' + 2 f j m , x'+ m, y'+ rn, z') (n,xl+ n,yl+ n3z1)).
The coefficient of x" in thk above expression is
+ + +
(G bm: + cn: 2fm2n2 2gnlll 2hllml), +
SimiIai.1~the coefficient of yt2 intthe above expression is
(alg + bmi + cni + 2fin2n2+ 2gn212+ 2h12m.2),
and that of zt2 is
+
(& + bmi + cn: + 2fm3n3 2gh313 2h13m3). +
Similarly we collect the coefficients of y'z', z'x' and x'y' Then we get an expression
of the form
a'xt2 + bryr2+ c'zt2 + 2f'y'zr +2grz'x' + 2h'x'yr .
.. .(16)
You can also see that the linear part becomes
u'x + v'y + w'z ....(17)
Where
+
u' = ul; vml + wnl
+ +
V' = uZ2 vm2 wn2, and
+
w'= u13 vm3 + wn3
From (16) and (17) we see that the transformed equation is a general second degree
equation. '

Now looking at expression (17), can you say anything about the change in the a

constant term? It remains unchanged under rotation-of axes. Another interesting

fact that you may have o b s e ~ e din the proof above is given in the following exercise.

+
Ef3) Suppose that the second degree part zax2' 2Sfyz of a general second degree
equation translorms into 2 a'xt2;.+ 2Xf1y'z' under rotation of axes. Show that
a+b+c = at+b'+c'.

Well, let us see what we can gather from Theorems 1 and 4. They say that if a
conicoid S is represented by a-second degree equation F(x,y,z) = 0 in one coordinate
system, then it is still represented by a second degree equation in ally other
a coordinate system obtained by a translation or rotation of axes.
Thus
'7
a conicoid remains a conicoid under translation or rotation of axes.
I
In fact, every geometrical figure remains unchanged in shape and size under
translation or rotation of axes. Therefore these transformations are called rigid body
motions.
In this section we have discussed two important transfonfiations of a3hree-
dimensional cobrdinate system. We also said that the importance of these
transformations lies in the fact that wk can use them to reduce any general second
degree equation in variables into a simpler form. h t us see how this happens.

7.4 REDUCTION TO STANDARD FO

In thjs section we shall show that by suitably applying the transformations that we
have discussed in the previous section; we can write the general equation of a
conicoid in a simplerform.
Let us consider a conicoid given by the equation
F(p(l y, z) = Sax 2 + 2Zfyz + S2ux +
d =0
Let us asLume that there exists a new Cartesian coordinate system, obtained by
translating the origin, in w'lich the 1inear;pait of-F(x,y,z) = 0 vanishes. You will
see that this is possible only for aparticular type of conicoids.
 Let the coordinates of O rbe ( q , yo, q)in the new system. Then we know that in
the transformed equation, the second degree part is unchanged and the linear part
1 becomes
U'X' + v'y' +wrzr
where u', v' and w' are as in (8). We have assumed that the linear part vanishes.
Therefore u' = v' = w' = 0. This means that we should have
h
.ax,+ h y 0 + g q + u = O
.hx,+byO+fi,+v=O
igx,,+ fyo+ czo + w = 0
1
In other words: ( 4 , yo, q) is a solution of the system of equations
ax+hy+gz+ u = O
hx+by+fi+v=O
-gx+£y+cz+w=o
I
If the system of equations (18) has a solution for (xo, yo, zo) dR3, then the point,
....(18)

(x,,,yo, q,), is called a centre of the given conicoid and we say that the conicoid has
a centre at ( q , yo, q ) . You will understand why this is called a centre later in Unit 8.
q Now let us assume that the given conicoid S has a centre. We transfer the origin
to the centre (k, yo, q,).Then the transformed equation becomes
Sw2+.28fyz + 2u' x + 2v'y + 2w'z + d' = 0.
(Recall that the second degree part does not change by shifting the origin.)
Since (xo, yo, Q) is a solution to (18), we see that u', = v' = w' = '0. Therefore,
the above equation reduces to
(Zax2 + 2Zfyz) + d' = 0
This equation does not have any linear part.
We have just proved the following result.
Theorem 5 : Suppose that S is a conicoid which is represented by a general second
degree equationF(x,y,z) = 0 in a coordinate system XYZ.Suppose that S has a
ceritre O r (i.e. the system of equitioni'(18) has a solution (x,,, yo, q)). Then, by
shifting the,origin to the ceritre O r ,the equation assumes the form
+
axZ + by2 + cz2 +2fyz 2gzi f 2hxy + d' = 0
in the new coordina!e system X'Y'Z'. See Block 2. MTE-04for

Let us consider an example.

'

Example 2 : Consider the conicoid given by the equation

2x2 + 3y2 f 4z2 - 4x - +
i2y - 242 49 = 0.
Does it have a centre? If so, find it.
Solution : The given equation is
+
F(x, y, z) = 2x2 + 3 ~ ' 4z2 - 4x - 12y - 242 + 49 = 0
We shall first check whether the given equation has a centre or not, that is, if the
sy'stem of equations (18) has a solution or not. Here a = 2, b = 3, c = 4, u = -2,
v = -6, w = -12. Then we have
2x-2=0
3y-6=0
42 - 12 = 0
This set of equations has a solution, namely, (1,2,3). Hence, (1,2,3) is a centre of S.
Note : Let us go back to (18) for a moment. There we saw that if the system of
then (%,yo,%) is a centre of the conic. In Unit
equations has a solution (~,,,~~,q),
5 of the course 'Elementary Algebra', you have seen that the system of equations
has a solution if
*
a h b
A = h b f ' #O
g f c

In fact if A ic 0, then there exists a unique solution.

W h y don't you trysome exercise now?
- --

k9) Check whether the following kfiicoids have a centre or not

a) 3xT + 7 9 + 32' + 10yz -2zx + l h y +4x -12y +42 + 1 = 0
b) x2+y2 +z2 -2yz +2zx - 2 x y - % + 2 y - 2 2 - 3 = 0.
c) 5x2 + 69 2x = 0.-
-
E10) Find a centre of the cohicoid
14x2 + 14y2+ 8z2 4yz - 42%- 8xy + 18x - 18y + 5 = 0. What will its
new equation be if .the origip is shifted to this centre?

We shall now state a theorem without proving it. To prove this we need s o q e
advanced techniques in calculus which are beyond the scope of this course.
Theorem 6 : Suppose S is a given conicoid whose equation is given by 4: (x,y,z) =
ax2 + by2 + cz2 + 2fy2 + 2gzx + +
2hxy 2ux + 2vy + 2wz + d = 0.
with respect to a given system of coordinates XYZ. Then there exists a new
I
Cartesian coordinate system X'Y'Z' obtained by rotating the axes of the given I
system XYZ,without shifting the origin, such that the representation of S in the c
new system takes the form
G(x',y',zt) = a'x' 2+b'y'2 + c'zf2+ 2u1x' + 2v'yr + 2w'z' + d' = 0.
That is, the ne- equati~ndoes not contain the product terms, yz, zx and xy.
Combining Theorems 5 and 6 we have the following result. I

Corollary 1 : Let S be a conicoid given by the equation F(x,y,z) = 0, which has a

centre 0' in a coordinate iystem XYZ. There exists a new coordinate system . I
obtained by shifting the origin from 0 to 0' and then 'rotating the system about
0', in which the equation takes the simpler form I
afX12 + bryt2+ C1zf2 + d' = 0 ....( 19)
If d' # 0 then we can divide throughout by d', and we get the form
+
ax2 by2+ cz2 = 1,.
a' b' c'
whereay- -,b=- - a n d c = - -.
d' d' d'
(19) is called the standard equation of a conicoid .
Recall that in the case of the two-dimensional system also we have seen that we can
reduce any second degree equation into a simple form. ,

Let us now consider some examples.

Example 3 : Show that the conicoid gjven by x2 + 2yz - 4x + 6y 22 = 0 +
I
has a centre. Reduce it to standard form by shifting the origin to the centre, and thev
rotating the axes to get a new system in which the direction ratios of the new axes .
are given by

with respect to the original coordinate systern.

I
We first check ilzether the conicoid has a centre or Aot. Using (18) we see that
(2,0,0) is a centre of the given conicoid.

Shifting the origin from (0,0,0) to (2,0,0), we get the new equation as - 1
x2 + 2yz - 4 = D I :
Now we apply a rotation of axes to the new equation. We note that the direction P
cosines of the new axes are I
I

f i o m Sec.7.3.3 we have
 General T b w y of Cdcddr

I
Substituting these equations in the given equation of the conicoid, we see that

which is in the standard form:

Now here .is an exercise for you.

E l l ) Find the standard equation of the following cdnicoids.

a) x2+y2+z2-2x-2y-22-1 = 0, by shifting the origin to the centre.
b) 3 ~ ~ + 5 ~ ~ + 3 ~ ~ - 2 y ~ + 2 ~ ~ - - 2 ~ y + 2 ~=+ 0,
1 2by~ shifting
+ 1 0 ~ +the
20
origin to the centre and then rotating the system so that the direction ratios
of the new axes are
-1,0,1; l , l , l ; 1,-2,l.

We will stop our discussion on general theory ot conicoids for now, though we shall
refer to them off and on in the following units. 1n the next unit we will discuss the
surfaces ,formed by (19).
Let US now do a quick review of what we have covered in this unit.

7;5 SUMMARY
In this unit we have covered the following points:
1) . A general second egree equation in three variables
P
a~~+by~+c~+~2fyz+2gzx+2hxy+2~~+2vy+2wz+d=d
. represents a conicold. \
2) Translation of axes: the tral~sformafionof a coordinate system in which the
origin is shifted to another point without changing the-direction of the axes.
The equations of transformations are given by
x=xt+a
y=yf+b
z=zf+c
3) Rotation of axes: the transformation of a coordinate system in which the
direction of axes is changed without shifti& the origin. The equations of
transformation are given by the following table: .

where li, mi, ni, i = 1, 2, 3 are the direction cosines of the axes.
4) A conicoid remains a conicoid under a translation or rotation of axes.
5) There is a Cartesian coordinate system in which the equation of a conicoid with
a centre takes the standard form
atxt2+ b'yj2 + cfzt2+ d t = 0.
And, now you may like to check whether you have achieved the objectives of this
unit (see Sec.7.1). If you would like to see our solutions to the exercises in this
unit., we have given them in the following section.
7.6 SOL
ES)a) In Unit 6 of Block 2 you have seen that the equation of a right circular cone
with vertex 0,axis OZ and semi-vertical angle a is ,
+ ?j = z2tanza.
x2
When we shift the origin t o (-1,1,0), then the coordinates in the new system
are given by
X' = ~ + 1 ,r Y' = y-1, Z' = Z,
i.e. x = x' - 1 y = yl+l, z = z' I
Substituting far x,y,z in the given equation of the cone, we get
(x' ( y l + 1)2 = z'3an2a. I

b) This equation represents a right circular cone with vertex at the point (-1,1,0)
axis along the line parallel t o the z-axis through the vertex and semi-verti-91
angle a (see E l in Sec.6.2 of Unit 6, Black 2).
E2) The equations of transformation are given by
X = x'f 1, y = y'-3, Z = z1+2
a) the given equation is
x2+?i-2-4~+6y-2~+5 = 0
Substituting for x,y,z, in the above equation, we get
( ~ ' + 1 ) ~(y'-3)2+(~'+2)2-4(~'+
+ 1)+6(y1-3)-2(z1+2) +5 =0
Simplifying, we get
xf2+ y'2+z12-2x'+2z1--7 = 0
which represents-a sphere
b) The transformed equation is
7 0.
~ ' ~ - 2 y ' ~ + 2 ~ ' - 4 ~ ' - 3 2 ' -=
E3) The direction ratios of RS are -3, 6, -2.
3 =i,
Therefore, the direction cosines of RS are -5, 6 -7.2
Hence the projection of PQ on RS is

E4) a) The given equation is

X2-59+ z 2 = 1
The direction ratios of the axes are given by .
1, 2, 3; 1, -2, 1; 4, 1, -2.
Therefore the direction cosines of the axes are given by

The transformation Table is

I
I
'd
Then we have
1 1
x =- .I + ,
. y' +-
3 I
,
4- fiz' I
I
 Substituting this in the given equation, we get Cmrrl Theory at Conicdde

simptifying we get

8
b) The new equation is x12 + yt2,+ zt2+ y1z'-- 8 . H'Z' = 1.
2/ZT JTl Jf4
E5) From E4(a) we get
'a+b+c = -5+2 -318
and a1+b'+c' = i:
- -6 + - = -3-
This sllows that urldet rotation the sum of coefficients of the square terms remains
unaltered in value, i.e., aib+c = a'+b'+ct.
I
similarly you can observe the same for E4(b) and Example 1 also.
!
E6) From the given equations of transformations we see that the coordinate system
is changed into another coordinate system with the same origin and the
direction cosines of the new axes, with respect to the old system, are given by

To get the new equation, we substitute the values of x,y,z in the equation
x+y+z = 0. Then we get

Therefore under the transformation the plane x+y+z = 0 becomes the plane .
= 0.
2'

E7) Yes. This is because with vertex at the origin is represented by a homogeneous
second degree equation.
ax2+by2+cz2+2hxy+2fyz+2gzx= 0.
Now, from the proof of Theorem 2, we see that under rotation of axes, the
above equation becomes
a'xtZ+b'y'2+~'~'2+2h'~'y'+2f'y'~'+2g'~'x' '= 0,
which is again a homogeneous second degree equation.
Therefore, it represents a cone.

m) From the proof of theorem 2, we have.

a' = alt + bm12 + cn12+ 2fm1n1 + 2gn 1 11 x + 2hllmj
'
b' = a12' + bmzZ+ cn22+ 2 f w 2 +'2gnz lz + 2h12mz
c' = aZ: + bm32 +. cn32+ 2fm,n3 + 2gn3 l3 + 2hlf13
then a' + b' + c' = a(llz + lZ2+ 1:) + b(mT + 'm22 i rnc)+ c(ni2
+ n22+ n3') +2f(m6'nl + m2 % + m3 n3) +2g(nlli +
412 +' n&t+ 2h,CJlml + 12m2+
We know that

3 3 . 3
Z 8=1,= 2 m:,= S .nt
ill i= 1 i= 1
'(see Unit 4, Block 2).

Also, since the axes are mutually perpendicular, by condition for

pehndicularity givec in Unit 4, ~lock.2;we get .'
 Therefore, we get a' + b' + c' = a +b+c
E9) a) Weahave to see whether the following system of linear equations is I
consistent or not. I
ax+hy+gz+u = 0 I
hx+by+fz+v = 0 II

gx+fy+cz+w = 0 I

-.
H e r e a ~ = 3 , b ~ 7 , ~ = 3 , f = 5 , g = - 1 , h = 5 , u = 2 ,-
w = -2. So we have
v 6,
= I

,
3x+5y-z+2 = Q
5x+7y+Sz-6 = 0
-x+.5y+3z-2 = 0
. On solving this system of equations, we find that it has a unique solution 1
given by x = 3, -
1 y = 1 and z = -4 Therefore the given cooicoid has a
3' I

unique centre at ( 1 1 4 )
- 3 9 -5,5
b) This conicoih has infiniiely many centres.
c) This conicoid has no centre since the system of equations (18) is
inconsisten in this case.
E10) The conicoid has a centre at (- 12, 12, 0). Now we shift the origin from (0,0,0)
to (- - 1 -,1 0). The equations of transformation are
2' 2
x = x' -A-. 1 y y' + 1T , = 2'
2" , P

Substituting for x,y,z in the given equation, we get the new equation as
1 4 ~ ~ + 1 4 y ~ + 8 ~ ~ - 4 y z - 4 z x=
- 84.~ ~ I
I
E l l ) a) We know that the given equation represents a sphere with centre at
(1,1,1) (see Unit 5). Then by shifting the origin to the centre we get the I
I
standard eqyation as !
~ ' ~ + y ' ~ + z=' 24.
b) The given equation is 1
3x2+ 5y2+3z2- 2Yz+2z~-2~Y+2x+ 12y+ 1oz+ 20 = 0. I
We first check whether the conicoid represented by this equation has a I
centre. i
I
Herea=3,b=5,c=3,h= -1,g=1 , f = -l,u=l,v=U, ,
w = 5 and d = 20. The system of equations for transformations are given 1
by
3x--y+z+l = 0
. -x+5y-z+6 = 0 , .. , a
j
I
x-y+3z+5 = 0
Sblving these equations, we get x = - 1, y --and
1
a n t r e is (-- -2 -I$). 5 3
5 -
z = -I3 Hence a
a.
I
6' 3'
Now we shift the origin to the centre. Then we get the new equation as
3x2 + 5y2+ 322- 2yz 2zx - 2xy + dl = (I$+ i
Now we apply the rotation of Ges. The equations of tiansformation are I
X = -x'+y'+zt I
y 5 yl-kL r'
z = x'+y'+zl I
Substituting for x,y,z in the given equition of the conicoid, we get
3(-x'+Y'+z')~ + 5(y1-2z1)'+ ~ ( x ' + Y ' + z ' )-
~ 2(y1-22') (x1+y'+z') + I
~ ( X ' + ~ ' + ' Z(-xl+y'+z')
') -2 (-xl+y'+z') (yl-2z') d' = 0 +
+ +
i.e., 4xt2 9yf2 36zt2 d' = 0 + i
This is the,standard form of the given' conicoid. I
I
I
20 *
I
11
-
 UNIT 8 CENTRAL CONICOIDS ..

Structure
8.1 Introduction
i . Objectives
. 3.2 A Conicoid's Centre .
. 8.3- Classification of Central Conicoids
&.J Ellipsoid -
8.5 Hyperboloid of One sheet
8.6 Hyperboloid of Two Sheets
8.7 Intersection With a Line or a Plane
Line Intersections
planar Intersections
8.8 Summary
8.9 Solutions/Answers

8.1 INTRODUCTION
In the previous unit you were introduced to certain surfaces in a three-dimensional
system, which are called conicoids or quadric surfaces. There we discussed some
general theory of conicoids and showed that a conicoid remains a conicoid under
translation and rotation of axes. You also saw that some conicoids possess a centre
and some don't. Based on this the conicoids are classified into two types-central
and non-central conicoids. In this unit we shall concentrate only on central conicoids.
We first observe that a central conicoid can be reduced to a simpler fm by an
appropriate change of axes. Then we use thesp Simplex f o r m to discuss the
$
geometrical properties of different types of cent~alconicoids. You will pee that there
are four types of central conicoids - cone, ima@ary ellipsoid,conicoid,' yperboloid
t
of one sheet and hyperboloid of two sheets. You"Iia~ealreadystudi d cones in
detail in Unit 6. The remaining three reaI conicoids are the three-dimensiond versions
of the central conics that you studied in Block 1, namely, ellipses and hyperbolas,
Ancient mathematicians like Euclid, Archimedes and Appolonius were familiar with
the above mentioned geometrical objects, though they did not study them
analytically. The analytical study of these objects started niuch later, with the
application of the three-dimensional coordinate system to geometry. The first
mathematician to suggest the extension of the two-dimensional coordinate system
to three dimensions. was the Swiss mathematician John Bernoulli (1667-1748). But
the actual application of space coordinates to geometry was done by another Swiss
rriathematician Jacob Hermann (1678-1733). He applied them to obtain the equations
of several types of quadric surfaces.
Even though we are only concerned with central conicoids in this unit, in the first
section we shall consider some necessary general theory of conicoids which we have
.not covered in the previous unit. We shall define a centre of a conicoid and obtain
a characterisation for central conicoids. Then we shall discuss the four different
types of central conicoids in separate sections. W=will end this unit with a discussion
on sections obtained by intersecting a central dbnicoid by a line or a plane. In this
connection we also discuss tangents to a central conicoid.
In the next unit we will tackle no?-central conicoids. That will be easier for you to
grasp if you ensure that you have achieved the objectives given below.

Objectives
After studying this unit you should be able to : ,
I
check whether a conicoid is central or not if you are given its equation;
9
obtain standard fdrms of an ellipsoid, hyperboloid of onessheet and hyperboloid
t - of two sheets;