FACTA UNIVERSITATIS (NIŠ)

Ser. Math. Inform. Vol. 36, No 5 (2021), 1155-1168

https://doi.org/10.22190/FUMI210804084O
Original Scientific Paper

ON THE BI-P -HARMONIC MAPS AND THE CONFORMAL MAPS

Seddik Ouakkas and Abderrazak Halimi

Laboratory of Geometry, Analysis, Control and Applications, University of Saida,

Dr Moulay Tahar, BP 138 EN-NASR, 20000 Saida, Algeria

Abstract. The objective of this paper is to study the bi-p-harmonicity of a conformal

maps. We establish necessary and sufficient condition for a conformal map to be bi-p-
harmonic and we construct several examples of this type of maps.
Keywords: p-harmonic map, bi-p-harmonic map, conformal map.

1. Introduction
m n
Let φ : (M , g) −→ (N , h) be a smooth map between two Riemannian mani-
folds. Then φ is said to be harmonic if it is a critical point of the energy functional
: Z
1
E (φ) = |dφ|2 dvg
2 M
with respect to compactly supported variations. Equivalently, φ is harmonic if it
satisfies the associated Euler-Lagrange equations given as follows:

τ (φ) = T rg ∇dφ = 0,

τ (φ) is called the tension field of φ. The map φ is said to be biharmonic if it is a

critical point of the bi-energy functional:
Z
1
E2 (φ) = |τ (φ)|2 dvg .
2 M
Received August 04, 2021. accepted November 05, 2021.
Communicated by Uday Chand De
Corresponding Author: Seddik Ouakkas, Laboratory of Geometry, Analysis, Control and Appli-
cations, University of Saida, Dr Moulay Tahar, BP 138 EN-NASR, 20000 Saida, Algeria | E-mail:
seddik.ouakkas@univ-saida.dz
2010 Mathematics Subject Classification. Primary 31B30, 53C43; Secondary, 58E20, 53C18

© 2021 by University of Niš, Serbia | Creative Commons License: CC BY-NC-ND

1156 S. Ouakkas and A. Halimi

The biharmonicity of φ is characterized by the following equation:

2
τ2 (φ) = −T rg ∇φ τ (φ) − T rg RN (τ (φ), dφ)dφ = 0,

where ∇φ is the connection in the pull-back bundle φ−1 (T N ) and, if (ei )1≤i≤m is a
local orthonormal frame field on M , then

2  
T rg ∇φ τ (φ) = ∇φei ∇φei − ∇φ∇M ei τ (φ) .
ei

We will call the operator τ2 (φ), the bi-tension field of the map φ. A generalization
of harmonic and biharmonic maps, p-harmonic and bi-p-harmonic maps are defined
as follows : Let p ≥ 2, the p-energy functional of φ is defined by
Z
1
Ep (φ) = |dφ|p dvg .
p M

φ is said to be p-harmonic if it is a critical point of the p-energy functional (with

respect to any variation of compact support). Equivalently, φ is p-harmonic if it
satisfies the associated Euler-Lagrange equations:
p−2
τp (φ) = |dφ| {τ (φ) + (p − 2) dφ (grad ln |dφ|)} = 0,

τp (φ) is called the p-tension field of φ, one can refer to [1], [12] and [15] for more
details on p-harmonic maps. The bi-p-energy of φ is defined by (see [4]) :
Z
1
E2,p (φ) = |τp (φ) |2 dvg .
2 M

Equivalently, φ is bi-p-harmonic if it satisfies the following equation:

p−2 p−2
τ2,p (φ) = −T rg ∇φ |dφ| ∇φ τp (φ) − |dφ| T rg RN (τp (φ) , dφ) dφ
(1.1) 
p−4

− (p − 2) T rg ∇φ h∇τp (φ) , dφi |dφ| dφ = 0,

where
p−2 p−2 p−2
T rg ∇φ |dφ| ∇φ τp (φ) = ∇φei |dφ| ∇φei τp (φ) − |dφ| ∇φ∇e e τp (φ)
i i

and
 
p−4 p−4
T rg ∇φ h∇τp (φ) , dφi |dφ| dφ = ∇φei |dφ| h∇τp (φ) , dφi dφ (ei )
p−4
− |dφ| h∇τp (φ) , dφi dφ (∇ei ei ) .

τ2,p (φ) is called the bi-p-tension of φ. Following Jiang’s notion (see [9]), we define
stress bi-p-energy tensor associated to the bi-p-energy functionals by varying the
 On the bi-P -harmonic mpas and the conformal maps. 1157

functionals with respect to the metric on the domain (see [11]). For any X, Y ∈
Γ (T M ), we have
1 2 p−2

dφ, ∇φ τp (φ) g (X, Y )


S2,p (φ) (X, Y ) = |τp (φ)| g (X, Y ) + |dφ|
2 n    o
(1.2) − |dφ|
p−2
h dφ (X) , ∇φY τp (φ) + h dφ (Y ) , ∇φX τp (φ)
p−4

dφ, ∇φ τp (φ) h (dφ (X) , dφ (Y )) .


− (p − 2) |dφ|
The stress bi-p-energy tensor of φ satisfies the following relationship
divS2,p (φ) = −h (τ2,p (φ) , dφ) .
The notion of bi-p-harmonic maps was introduced by A.M.Cherif [4] where he gave
the Euler-Lagrange equations associated with the bi-p-energy and he proved a Li-
ouville type theorem for this class of maps. It is important to recall that the
p-biharmonic maps are the critical points of the p-bi-energy functional
Z
1
Ep,2 (φ) = |τ (φ) |p dvg ,
p M
and this type of maps was studied in [3], [5] and [8]. This paper is a continuation
of Cherif’s work [4] on bi-p-harmonic maps where we study the bi-p-harmonicity
of a conformal map φ : (M n , g) −→ (N n , h) (n ≥ 3), we calculate τ2,p (φ) and we
prove that any conformal map is bi-p-harmonic if and only if the gradient of its
dilation satisfies a certain second-order elliptic partial differential equation. From
these results, we construct new examples of bi-p-harmonic maps.

2. The main results

In the first we give the relation between τ2,p (φ) and τp (φ).
Proposition 2.1. Let φ : (M m , g) −→ (N n , h) be a smooth map, then the relation
between τ2,p (φ) and τp (φ) is given by the following equation

2 
p−2
τ2,p (φ) = − |dφ| T rg ∇φ τp (φ) + T rg RN (τp (φ) , dφ) dφ
  
p−4 2
+ (p − 2) |dφ| h∇τp (φ) , dφi dφ grad ln |dφ|
p−4
(2.1) − (p − 2) |dφ| dφ (grad h∇τp (φ) , dφi)
−2
− (p − 2) |dφ| h∇τp (φ) , dφi τp (φ)
(p − 2) p−2 φ
− |dφ| ∇grad ln|dφ|2 τp (φ) .
2 ( )
Proof of Proposition 2.1. Let us choose {ei }1≤i≤m to be an orthonormal frame
on (M, g). By definition, we have
p−2 p−2
τ2,p (φ) = −T rg ∇φ |dφ| ∇φ τp (φ) − |dφ| T rg RN (τp (φ) , dφ) dφ
(2.2) 
p−4

− (p − 2) T rg ∇φ h∇τp (φ) , dφi |dφ| dφ .
1158 S. Ouakkas and A. Halimi

p−2
For the term T rg ∇φ |dφ| ∇φ τp (φ), we obtain

p−2 p−2 p−2

T rg ∇φ |dφ| ∇φ τp (φ) = ∇φei |dφ| ∇φei τp (φ) − |dφ| ∇φ∇e e τp (φ) ,
i i

a simple calculation gives us

 
p−2 p−2 p−2
∇φei |dφ| ∇φei τp (φ) = |dφ| ∇φei ∇φei τp (φ) + ei |dφ| ∇φei τp (φ)
p−2 (p − 2) p−2
= |dφ| ∇φei ∇φei τp (φ) + |dφ| ∇grad(ln|dφ|2 ) τp (φ) ,
2

then
p−2 p−2
2
T rg ∇φ |dφ| ∇φ τp (φ) = |dφ| T rg ∇φ τp (φ)
(2.3) (p − 2) p−2 φ
+ |dφ| ∇grad ln|dφ|2 τp (φ) .
2 ( )
 
p−4
We will develop the term T rg ∇φ h∇τp (φ) , dφi |dφ| dφ , we have

 
p−4
T rg ∇φ h∇τp (φ) , dφi |dφ| dφ
p−4 p−4
= ∇φei |dφ| h∇τp (φ) , dφi dφ (ei ) − |dφ| h∇τp (φ) , dφi dφ (∇ei ei )
 
p−4 p−4
= |dφ| h∇τp (φ) , dφi ∇φei dφ (ei ) + ei |dφ| h∇τp (φ) , dφi dφ (ei )
p−4
− |dφ| h∇τp (φ) , dφi dφ (∇ei ei )
p−4 p−4
= |dφ| h∇τp (φ) , dφi ∇φei dφ (ei ) − |dφ|
h∇τp (φ) , dφi dφ (∇ei ei )
 
p−4 p−4
+ |dφ| ei (h∇τp (φ) , dφi) dφ (ei ) + h∇τp (φ) , dφi ei |dφ| dφ (ei )
p−4 p−4
= |dφ| h∇τp (φ) , dφi τ (φ) + |dφ| dφ (grad h∇τp (φ) , dφi)
p−4 p−4
 
2

+ |dφ| h∇τp (φ) , dφi dφ grad ln |dφ| .
2

Using the fact that

−p+2 (p − 2)  
2

τ (φ) = |dφ| τp (φ) − dφ grad ln |dφ| ,
2

it follows that
 
p−4 −2
T rg ∇φ h∇τp (φ) , dφi |dφ| dφ = |dφ| h∇τp (φ) , dφi τp (φ)
p−4
(2.4) + |dφ| dφ (grad h∇τp (φ) , dφi)
  
p−4 2
− |dφ| h∇τp (φ) , dφi dφ grad ln |dφ| .
 On the bi-P -harmonic mpas and the conformal maps. 1159

By replacing (2.3) and (2.4) in (2.2), we deduce that


2 
p−2
τ2,p (φ) = − |dφ| T rg ∇φ τp (φ) + T rg RN (τp (φ) , dφ) dφ
  
p−4 2
+ (p − 2) |dφ| h∇τp (φ) , dφi dφ grad ln |dφ|
p−4
− (p − 2) |dφ| dφ (grad h∇τp (φ) , dφi)
−2
− (p − 2) |dφ| h∇τp (φ) , dφi τp (φ)
(p − 2) p−2 φ
− |dφ| ∇grad ln|dφ|2 τp (φ) .
2 ( )
Theorem 2.1. Let φ : (M n , g) −→ (N n , h) (n ≥ 3) be a conformal map of dilation
λ , then the bi-p-tension of φ is given by
τ2,p (φ) = (n − p) np−3 λ2p−4 dφ (H (λ, n, p)) ,
where
H (λ, n, p) = (n + p − 2) grad (∆ ln λ)


n2 − 5np + 4n − 2p2 + 8p − 8 
2

− grad |grad ln λ|
2
2
− (p − 1) n − 3np + 4n − 2p2 + 8p − 8 |grad ln λ| grad ln λ
2

− 2 n − p2 + 3p − 2 (∆ ln λ) grad ln λ + 2nRicci (grad ln λ) .

Lemma 2.1. Let φ : (M m , g) −→ (N n , h) be a smooth map. For any vector filed

X and for any smooth function f on M , we have

2
2
T rg ∇φ f dφ (X) = f T rg ∇φ dφ (X) + 2∇φgradf dφ (X) + (∆f ) dφ (X) .

Proof of Theorem 2.1. The fact that the map φ is conformal of dilation λ gives
us p−2
2 p−2
τ (φ) = (2 − n) dφ (grad ln λ) , |dφ| = nλ2 , |dφ| = n 2 λp−2
and  
2
grad ln |dφ| = 2grad ln λ.
Then p−2
τp (φ) = (p − n) n 2 λp−2 dφ (grad ln λ) .
By replacing the expression of τp (φ) in (2.1), we obtain

2
τ2,p (φ) = − (p − n) np−2 λp−2 T rg ∇φ λp−2 dφ (grad ln λ)
− (p − n) np−2 λp−2 T rg RN λp−2 dφ (grad ln λ) , dφ dφ

− (p − 2) (p − n) np−2 λp−2 ∇φgrad ln λ λp−2 dφ (grad ln λ)

(2.5)
2
− (p − 2) (p − n) np−3 λp−4 ∇λp−2 dφ (grad ln λ) , dφ dφ (grad ln λ)



− (p − 2) (p − n) np−3 λp−4 dφ grad ∇λp−2 dφ (grad ln λ) , dφ



+ 2 (p − 2) (p − n) np−3 λp−4 ∇λp−2 dφ (grad ln λ) , dφ dφ (grad ln λ) .


1160 S. Ouakkas and A. Halimi

We will simplify the terms of this last equation.

2
For the term T rg ∇φ λp−2 dφ (grad ln λ), we have

2
2
T rg ∇φ λp−2 dφ (grad ln λ) = λp−2 T rg ∇φ dφ (grad ln λ)
+ 2∇φgradλp−2 dφ (grad ln λ)
+ ∆λp−2 dφ (grad ln λ) .

The fact that φ is conformal gives us (see [13])

2   
2
T rg ∇φ dφ (grad ln λ) = dφ (grad∆ ln λ) + 2dφ grad |grad ln λ|
2
− (n − 2) |grad ln λ| dφ (grad ln λ)
− (∆ ln λ) dφ (grad ln λ) + dφ (Ricci (grad ln λ))

and
2
2∇φgradλp−2 dφ (grad ln λ) = 2 (p − 2) λp−2 |grad ln λ| dφ (grad ln λ)
  
2
+ (p − 2) λp−2 dφ grad |grad ln λ| .

A simple calculation gives

 
2
∆λp−2 = (p − 2) λp−2 ∆ ln λ + (p − 2) |grad ln λ| ,

then

2
T rg ∇φ λp−2 dφ (grad ln λ) = λp−2 dφ (grad∆ ln λ)
  
2
+ pλp−2 dφ grad |grad ln λ|
(2.6) 2
− n − p2 + 2p − 2 λp−2 |grad ln λ| dφ (grad ln λ)

+ (p − 3) λp−2 (∆ ln λ) dφ (grad ln λ)
+ λp−2 dφ (Ricci (grad ln λ)) .

The fact that φ conformal also gives us the following formulas (see [13])

n−2  
2

T rg RN (dφ (grad ln λ) , dφ) dφ = − dφ grad |grad ln λ|
2
(2.7) − (∆ ln λ) dφ (grad ln λ)
+ dφ (Ricci (grad ln λ))

and
2
∇φgrad ln λ λp−2 dφ (grad ln λ) = (p − 1) λp−2 |grad ln λ| dφ (grad ln λ)
(2.8) 1  
2

+ λp−2 dφ grad |grad ln λ| .
2
 On the bi-P -harmonic mpas and the conformal maps. 1161


For the term ∇λp−2 dφ (grad ln λ) , dφ , we have

p−2
∇λ dφ (grad ln λ) , dφ = T rg h ∇λp−2 dφ (grad ln λ) , dφ


= h ∇ei λp−2 dφ (grad ln λ) , dφ (ei )

= λp−2 h (∇ei dφ (grad ln λ) , dφ (ei ))

+ ei λp−2 h (dφ (grad ln λ) , dφ (ei ))


 
2
= λp−2 λ2 ∆ ln λ + nλ2 |grad ln λ|
2
+ (p − 2) λp−2 λ2 |grad ln λ| .
Then
 
2
∇λp−2 dφ (grad ln λ) , dφ = λp ∆ ln λ + (n + p − 2) |grad ln λ| .


(2.9)

Finally, using the following formulas

grad (λp (∆ ln λ)) = λp grad∆ ln λ + pλp (∆ ln λ) grad ln λ
and
   
2 2 2
grad λp |grad ln λ| = λp grad |grad ln λ| + pλp |grad ln λ| grad ln λ,

we obtain
grad ∇λp−2 dφ (grad ln λ) , dφ = λp grad∆ ln λ + pλp (∆ ln λ) grad ln λ


 
2
(2.10) + λp (n + p − 2) grad |grad ln λ|
2
+ p (n + p − 2) λp |grad ln λ| grad ln λ
If we replace (2.6), (2.7), (2.8),(2.9) and (2.10) in (2.5), we conclude that
τ2,p (φ) = (n − p) np−3 λ2p−4 dφ (H (λ, n, p)) ,
where
H (λ, n, p) = (n + p − 2) grad (∆ ln λ)


n2 − 5np + 4n − 2p2 + 8p − 8 
2

− grad |grad ln λ|
2
2
− (p − 1) n − 3np + 4n − 2p2 + 8p − 8 |grad ln λ| grad ln λ
2

− 2 n − p2 + 3p − 2 (∆ ln λ) grad ln λ + 2nRicci (grad ln λ) .

Theorem 2.2. Let φ : (M n , g) −→ (N n , h) (n ≥ 3) be a conformal map of dilation

λ , then φ is bi-p-harmonic if and only if


n2 − 5np + 4n − 2p2 + 8p − 8 
2

(n + p − 2) grad (∆ ln λ) − grad |grad ln λ|
2
2
− (p − 1) n2 − 3np + 4n − 2p2 + 8p − 8 |grad ln λ| grad ln λ

− 2 n − p2 + 3p − 2 (∆ ln λ) grad ln λ + 2nRicci (grad ln λ) = 0.

1162 S. Ouakkas and A. Halimi

If we consider a conformal map φ : (Rn , g) −→ (N n , h) (n ≥ 3) where we suppose

that the dilation λ is radial, then the bi-p-harmonicity of φ is equivalent to an
ordinary differential equation.

Corollary 2.1. Let φ : (Rn , g) −→ (N n , h) (n ≥ 3) be a conformal map of dilation

λ where we suppose that the dilation λ is radial (λ = λ (r) , r = |x|). By setting
0
β = (ln λ) , we get (see [13])

∂ 2

2
 ∂
grad ln λ = β , |grad ln λ| = β 2 , grad |grad ln λ| = 2ββ 0
∂r ∂r
and
 
0 n−1 00 n−1 0 n−1 ∂
∆ ln λ = β + β, grad∆ ln λ = β + β − β .
r r r2 ∂r

Using Theorem 2.2, we deduce that φ is bi-p-harmonic if and only if β satisfies the
following differential equation :
(2.11)
(n + p − 2) (n − 1) 0
(n + p − 2) β 00 − n2 − 5np + 6n − 4p2 + 14p − 12 ββ 0 +


β

r
(n + p − 2) (n − 1) 2 p2 − 3p − n + 2 (n − 1) 2
− β + β
r2 r

+ (p − 1) −n2 + 3np − 4n + 2p2 − 8p + 8 β 3 = 0.

To solve equation (2.11) , we will study two types of solutions. In the first case, we
look at the solutions which are written in the form β = ar , a ∈ R∗ , we obtain the
following result.

Corollary 2.2. Let φ : (Rn , g) −→ (N n , h) (n ≥ 3) be a conformal map of dilation

0
λ where we suppose that (ln λ) = β = ar , a ∈ R∗ . Then φ is bi-p-harmonic if and
only if a is solution of the following algebraic equation :
(2.12)
a2 n2 p − a2 n2 − 3a2 np2 + 7a2 np − 4a2 n − 2a2 p3 + 10a2 p2 − 16a2 p + 8a2 + an2
− 2anp2 + 11anp − 12an + 6ap2 − 20ap + 16a + 2n2 + 2np − 8n − 4p + 8 = 0.

Remark 2.1. Equation (2.12) leads us to two types of solutions

1.  
p
2 (n − 2) n + n (17n − 16)
a=− p
(3n2 − 6n + 4) n (17n − 16) − 13n3 + 42n2 − 28n
and
1p 3
p= n (17n − 16) − n + 2,
4 4
where n ≥ 3.
 On the bi-P -harmonic mpas and the conformal maps. 1163

2.
A (n, p) − 12n − 20p − 2np2 + 11np + n2 + 6p2 + 16
a=
8n + 32p + 6np2 − 2n2 p − 14np + 2n2 − 20p2 + 4p3 − 16
or
A (n, p) + 12n + 20p + 2np2 − 11np − n2 − 6p2 − 16
a=− ,
8n + 32p + 6np2 − 2n2 p − 14np + 2n2 − 20p2 + 4p3 − 16
where
s
4 (n − 1)2 p4 − 4 (n − 1) (n − 4) p3 +
12n3 − 35n2 + 8n + 16 p2


A (n, p) = 3 2 2 2
−2n 4n − 3n − 16n + 16 p + n (3n − 4)

and
1p 3
p 6= n (17n − 16) − n + 2
4 4

Remark 2.1 allows us to study the following examples. The examples to be cited
correspond to the cases where a = −2 and a = −1.

Example 2.1. We consider the inversion φ : Rn \ {0} −→ Rn \ {0} (n ≥ 3) defined by

x 1
φ (x) = |x|2 . φ is a conformal map with dilation λ = r 2 . We deduce that φ is bi-p-harmonic

if and only if
1 1p 5
p=− n+ −20n + 12n2 + 9 + , n≥4
2 4 4
or
3 1p
p=− n+ n (17n − 16) + 2, n ≥ 3.
4 4

Example 2.2. Let φ : Rn \ {0} −→ mathbbR × S n−1 given in polar coordinates by

φ (rθ) = (ln r, θ) , r > 0, θ ∈ S n−1 ⊂ Rn .

φ is a conformal map with dilation λ = r1 . We conclude that φ is bi-p-harmonic if and

only if
n
p= , n≥4
2
or
3 1p
p=− n+ n (17n − 16) + 2, n ≥ 3.
4 4
a
As a second particular case, we will look for the solutions of the form β = 1+r 2 , a ∈
R∗ .

Corollary 2.3. Let φ : (Rn , g) −→ (N n , h) (n ≥ 3) be a conformal map of dilation

0 a ∗
λ where we suppose that (ln λ) = β = 1+r 2 , a ∈ R . Then φ is bi-p-harmonic if

and only if a is solution of the following system:



 n5 p + 2n5 − 3n4 p2 − 6n4 p − 4n4 + n3 p3 + 6n3 p2
3
+14n p − 4n3 + 3n2 p4 − 6n2 p3 − 12n2 p2 + 4n2 p




+8n2 − 2np5 + 4np4 − 2np3 + 16np2 − 24np

(2.13)

 +4p4 − 16p3 + 16p2 = 0
and




3an2 − 2anp2 + anp − 2ap2 + 8ap − 8a + 2n2 + 2np + 4p − 8 = 0

1164 S. Ouakkas and A. Halimi

Remark 2.2. To solve this system, we distingue three cases

1.
2
p = n, a= , n ≥ 3.
n−2
In this case, the conformal map is n-harmonic so bi-n-harmonic.
2.
n 6n − 8
p= , a= 2 , n ≥ 4.
2 n − 8n + 8
Then φ is bi-p-harmonic non-p-harmonic.
3.
1 p 
p= −16n + 4n2 + 8n3 + n4 + 4 + n2 + 2
2n
and
2n2 + 2np + 4p − 8
a=− , n ≥ 3.
3n2 − 2np2 + np − 2p2 + 8p − 8
Then φ is bi-p-harmonic non-p-harmonic.

As the last result of this paper, we calculate the stress bi-p-energy tensor for a
conformal map.

Theorem 2.3. Let φ : (M n , g) −→ (N n , h) be a conformal map of dilation λ, then

we have
S2,p (φ) (X, Y )
p − n p−3 2p−2  2

2
= n λ n (n + p − 4) − 2 (p − 2) |grad ln λ| g (X, Y )
(2.14) 2
+ (p − n) (n − p + 2) np−3 λ2p−2 (∆ ln λ) g (X, Y )
− 2 (p − n) np−2 λ2p−2 (∇d ln λ (X, Y ) − (p − 2) X (ln λ) Y (ln λ)) ,

and the trace of S2,p (φ) is given by

T rg S2,p (φ)
p − n p−2 2p−2 2
(2.15) = n λ (n (n + p − 4) − 2 (p − 2) (p − 4)) |grad ln λ|
2
2
− (p − n) np−2 λ2p−2 (∆ ln λ) .

By using the fact that

 
2
∆λk = kλk ∆ ln λ + k |grad ln λ| ,

we obtain the following corollary :

Corollary 2.4. Let φ : (M n , g) −→ (N n , h) be a conformal map of dilation λ

where n 6= p, then
2
T rg S2,p (φ) = − (p − n) np−2 λ2p−2 T (λ) ,
 On the bi-P -harmonic mpas and the conformal maps. 1165

where
n (n + p − 4) − 2 (p − 2) (p − 4) 2
T (λ) = ∆ ln λ + |grad ln λ|
2 (n − p)
and
n(n+p−4)−2(p−2)(p−4)
T rg S2,p (φ) = 0 if and only if the function λ 2(n−p) is harmonic.

Remark 2.3. Let φ : (Rn , g) −→ (N n , h), (n 6= p) be a conformal map of dilation λ

where we suppose that the dilation λ is radial. By setting β = (ln λ)0 , we deduce that the
trace of S2,p (φ) is zero if if and only if β satisfies the following differential equation :

n−1 n + 2p − 4 2
(2.16) β0 + β+ β = 0.
r 2
The general solution of this equation is given by :
(
2(n−2)
A(n−2)r n−1 −(n+2p−4)r
, n 6= 2, A ∈ R
β= 2
(n+2p−4)r ln r+Ar
, n = 2, A ∈ R.

Remark 2.4. Let φ : (Rn , g) −→ (N n , h), (n 6= p, n 6= 2) be a conformal map of dilation

λ where we suppose that the dilation λ is radial. we will look for the solutions of the form
β = ar , a ∈ R∗ . we deduce that the trace of S2,p (φ) is zero if if and only if

2 (n − 2)
(2.17) a=− , n + 2p − 4 6= 0.
n + 2p − 4

For example, if we consider the conformal map φ : Rn \ {0} −→ R × S n−1 given in polar
coordinates by φ (rθ) = (ln r, θ), we conclude that for this map φ the trace of S2,p (φ) is
zero if if and only if n = 2p.

Proof of Theorem 2.3. Let us choose {ei }1≤i≤n to be an orthonormal frame on

(M, g). By definition, we have

1 2 p−2

dφ, ∇φ τp (φ) g (X, Y )


S2,p (φ) (X, Y ) = |τp (φ)| g (X, Y ) + |dφ|
 2   
(2.18) − |dφ|
p−2
h dφ (X) , ∇φY τp (φ) − |dφ|
p−2
h dφ (Y ) , ∇φX τp (φ)
p−4

dφ, ∇φ τp (φ) h (dφ (X) , dφ (Y )) .


− (p − 2) |dφ|

Using the fact that

p−2
τp (φ) = (p − n) n 2 λp−2 dφ (grad ln λ) ,

we obtain
2 2 2
(2.19) |τp (φ)| = (p − n) np−2 λ2p−2 |grad ln λ| .
1166 S. Ouakkas and A. Halimi


For the term dφ, ∇φ τp (φ) , we have
dφ, ∇φ τp (φ) = h dφ (ei ) , ∇φei τp (φ)



p−2
h dφ (ei ) , ∇φei λp−2 dφ (grad ln λ)


= (p − n) n 2

p−2
= (p − n) n 2 λp−2 h dφ (ei ) , ∇φei dφ (grad ln λ)


p−2
+ (p − n) n 2 ei λp−2 h (dφ (ei ) , dφ (grad ln λ))


p−2
 
2
= (p − n) n 2 λp ∆ ln λ + n |grad ln λ|
p−2 2
+ (p − n) (p − 2) n 2 λp |grad ln λ| .
It follows that
p−2
 
2
dφ, ∇φ τp (φ) = (p − n) n 2 λp ∆ ln λ + (n + p − 2) |grad ln λ| .


(2.20)
   
It remains to simplify h dφ (X) , ∇φY τp (φ) and h dφ (Y ) , ∇φX τp (φ) , we have
  p−2
 
h dφ (X) , ∇φY τp (φ) = (p − n) n 2 h dφ (X) , ∇φY λp−2 dφ (grad ln λ)
p−2
= (p − n) n 2 λp ∇d ln λ (X, Y )
p−2 2
+ (p − n) n 2 λp |grad ln λ| g (X, Y )
p−2
+ (p − n) (p − 2) n 2 λp X (ln λ) Y (ln λ) ,
which gives us
  p−2
h dφ (X) , ∇φY τp (φ) = (p − n) n 2 λp ∇d ln λ (X, Y )
p−2 2
(2.21) + (p − n) n 2 λp |grad ln λ| g (X, Y )
p−2
+ (p − n) (p − 2) n 2 λp X (ln λ) Y (ln λ) .
A similar calculation gives
  p−2
h dφ (Y ) , ∇φX τp (φ) = (p − n) n 2 λp ∇d ln λ (X, Y )
p−2 2
(2.22) + (p − n) n 2 λp |grad ln λ| g (X, Y )
p−2
+ (p − n) (p − 2) n 2 λp X (ln λ) Y (ln λ) .
By substituting (2.19), (2.20), (2.21) and (2.22) in (2.18) and using the fact that
p−2 p−2 p−4 p−4
|dφ| =n 2 λp−2 , |dφ| =n 2 λp−4 ,
we deduce that
S2,p (φ) (X, Y )
p − n p−3 2p−2  2

2
= n λ n (n + p − 4) − 2 (p − 2) |grad ln λ| g (X, Y )
2
+ (p − n) (n − p + 2) np−3 λ2p−2 (∆ ln λ) g (X, Y )
− 2 (p − n) np−2 λ2p−2 (∇d ln λ (X, Y ) − (p − 2) X (ln λ) Y (ln λ)) .
 On the bi-P -harmonic mpas and the conformal maps. 1167

To complete the proof, let’s calculate the trace of S2,p (φ), we have

T rg S2,p (φ) = S2,p (φ) (ei , ei )

p − n p−3 2p−2  2

2
= n λ n (n + p − 4) − 2 (p − 2) |grad ln λ| g (ei , ei )
2
+ (p − n) (n − p + 2) np−3 λ2p−2 (∆ ln λ) g (ei , ei )
− 2 (p − n) np−2 λ2p−2 (∇d ln λ (ei , ei ) − (p − 2) ei (ln λ) ei (ln λ)) ,

then
p − n p−2 2p−2 2
T rg S2,p (φ) = n λ (n (n + p − 4) − 2 (p − 2) (p − 4)) |grad ln λ|
2
2
− (p − n) np−2 λ2p−2 (∆ ln λ) .

Acknowledgment

The authors were supported in part by Directorate General for Scientific Research
and Technological Development.

REFERENCES

1. D. Allen: Relations between the local and global structure of fnite semigroups.
Ph. D. Thesis, University of California, Berkeley, 1968.
2. P. Erdős: On the distribution of the roots of orthogonal polynomials. In: Pro-
ceedings of a Conference on Constructive Theory of Functions (G. Alexits, S. B.
Steckhin, eds.), Akademiai Kiado, Budapest, 1972, pp. 145–150.
3. A. Ostrowski: Solution of Equations and Systems of Equations. Academic Press,
New York, 1966.
4. E. B. Saff and R. S. Varga: On incomplete polynomials II. Pacific J. Math. 92
(1981), 161–172.
5. M. A. Aprodu : Some remarks on p-harmonic maps. Stud. Cercet. Mat., 50
(1998), 297–305.
6. P. Baird and S. Gudmundsson: p-harmonic maps and minimal submanifolds.
Mathematische Annalen, 294 (1992), 611–624.
7. X. Cao and Y. Luo: On p-biharmonic submanifolds in non-positively curved
manifolds. Kodai Math. J., 39 (2016), 567–578.
8. A. M. Cherif: On the p-harmonic and p-biharmonic maps. Journal of Geometry,
109 (41) (2018), 1–11.
9. D. Djebbouri and S. Ouakkas: Some results of the f -biharmonic maps and
applications. Arab. J. Math., 24 (1) (2018), 70–81.
10. S. Dragomir and A. Tommasoli: On p-harmonic maps into spheres. Tsukuba
J. Math., 35 (2) (2011), 161–167.
11. A. Fardoun: On equivariant p-harmonic maps. Ann. Inst. Henri Poincar e, 15
(1) (1998), 25–72.
1168 S. Ouakkas and A. Halimi

12. Y. Han and Y. Luo: Several results concerning nonexistence of proper p-

biharmonic maps and Liouville type theorems. Journal of Elliptic and Parabolic
Equations, 6 (2020), 409–426.
13. G. Y. Jiang: 2-harmonic maps and their first and second variational formulas.
Chin. Ann. Math. Ser., A7 (1986), 389–402.
14. S. Kawai: A p-harmonic maps and convex functions. Geometriae Dedicata, 74
(1999), 261–265.
15. K. Mouffoki and A. M. Cherif: On the p-biharmonic submanifolds and stress
p-bienergy tensors. arXiv : 2104.12562, 2021.
16. N. Nakauchi and S. Takakuwa: A remark on p-harmonic maps. Nonlinear
Analysis, Theory, Methods and Applications, 25 (2) (1995), 169–185.
17. S. Ouakkas and D. Djebbouri: Conformal Maps, Biharmonic Maps, and the
Warped Product. Mathematics, 15 (4) (2016), doi : 10.3390/math 4010015.
18. Y. L. Ou: p-Harmonic morphisms, biharmonic morphisms, and non-harmonic
biharmonic maps. Journal of Geometry and Physics, 56 (2006), 358–374.
19. Z. Wei: Some results on p-harmonic maps and exponentially harmonic maps
between Finsler manifolds. Appl. Math. J. Chinese Univ., 25 (2) (2010), 236–242.