Citation: Özgür Boyacıoğlu Kalkan. On normal curves and their characterizations in Lorentzian n-space[J]. AIMS Mathematics, 2020, 5(4): 3510-3524. doi: 10.3934/math.2020228
[1] | Roa Makki . Some characterizations of non-null rectifying curves in dual Lorentzian 3-space D31. AIMS Mathematics, 2021, 6(3): 2114-2131. doi: 10.3934/math.2021129 |
[2] | Yanlin Li, Kemal Eren, Soley Ersoy . On simultaneous characterizations of partner-ruled surfaces in Minkowski 3-space. AIMS Mathematics, 2023, 8(9): 22256-22273. doi: 10.3934/math.20231135 |
[3] | Muslum Aykut Akgun . Frenet curves in 3-dimensional δ-Lorentzian trans Sasakian manifolds. AIMS Mathematics, 2022, 7(1): 199-211. doi: 10.3934/math.2022012 |
[4] | Yasin Ünlütürk, Talat Körpınar, Muradiye Çimdiker . On k-type pseudo null slant helices due to the Bishop frame in Minkowski 3-space E13. AIMS Mathematics, 2020, 5(1): 286-299. doi: 10.3934/math.2020019 |
[5] | Samah Gaber, Abeer Al Elaiw . Evolution of null Cartan and pseudo null curves via the Bishop frame in Minkowski space R2,1. AIMS Mathematics, 2025, 10(2): 3691-3709. doi: 10.3934/math.2025171 |
[6] | Haiming Liu, Jiajing Miao . Gauss-Bonnet theorem in Lorentzian Sasakian space forms. AIMS Mathematics, 2021, 6(8): 8772-8791. doi: 10.3934/math.2021509 |
[7] | Kemal Eren, Hidayet Huda Kosal . Evolution of space curves and the special ruled surfaces with modified orthogonal frame. AIMS Mathematics, 2020, 5(3): 2027-2039. doi: 10.3934/math.2020134 |
[8] | Emel Karaca . Non-null slant ruled surfaces and tangent bundle of pseudo-sphere. AIMS Mathematics, 2024, 9(8): 22842-22858. doi: 10.3934/math.20241111 |
[9] | Ayman Elsharkawy, Ahmer Ali, Muhammad Hanif, Fatimah Alghamdi . Exploring quaternionic Bertrand curves: involutes and evolutes in E4. AIMS Mathematics, 2025, 10(3): 4598-4619. doi: 10.3934/math.2025213 |
[10] | Beyhan YILMAZ . Some curve pairs according to types of Bishop frame. AIMS Mathematics, 2021, 6(5): 4463-4473. doi: 10.3934/math.2021264 |
In the Euclidean space E3, three classes of curves exists, which are called rectifying, normal and osculating curves all satisfying Cesaro’s fixed point condition [1]. The rectifying curve in E3 is defined as a curve whose position vector always lies in its rectifying plane which is spanned by the tangent vector T and the binormal vector B [2]. The relationship between the rectifying curves and the notion of centrodes in mechanics was introduced in [3]. Many authors in their papers have investigated rectifying curves in Euclidean and Lorentz-Minkowski space [4,5,6,7,8,9].
Similarly, a normal curve in Minkowski 3-space E31 is defined in [10] as a space curve whose position vector always lies in its normal plane which is spanned by the normal vector N and the binormal vector B of the curve. According to this definition, the position vector of a normal curve satisfies α(s)=λ(s)N(s)+μ(s)B(s) for some differentiable functions λ(s) and μ(s) in arclength function s. Spacelike, timelike and null normal curves in Minkowski space are studied in [10] and [11]. Spacelike and timelike normal curves in Minkowski space-time are investigated in [12]. The relations between rectifying and normal curves in Minkowski 3-space are obtained in [13]. The characterizations of normal curves in Galilean space are obtained in [14] and [15]. Moreover in [16], the definition and concept of a normal curve is extended to the general case En.
In this paper, by using similar methods as in [8], we introduce the normal curves in the Lorentzian n -space En1. We characterize null and non-null normal curves in terms of their curvature functions and obtain necessary and sufficient conditions for any curve to be a normal curve.
Let En1 denote the Lorentzian n -space. For vectors X=(x1,x2,...,xn) and Y=(y1,y2,...,yn) in En1
⟨X,Y⟩=−x1y1+n∑i=2xiyi |
is called Lorentzian inner product. Since ⟨,⟩ is an indefinite metric, recall that a vector v∈En1 can have one of three causal characters; it can be spacelike ⟨v,v⟩⟩0 or v=0; timelike if ⟨v,v⟩⟨0 and null (lightlike) if ⟨v,v⟩=0 and v≠0. The pseudo-norm (length) of a vector v is given by ‖v‖=√|⟨v,v⟩| [17].
We define the curve of α=α(s) to be an arclength parametrized non-null curve in En1. Let {T(s),N(s),B1(s),...,Bn−2(s)} be the moving frame along α, where the vectors T(s),N(s), B1(s),...,Bn−2(s) are mutually orthogonal vectors satisfying
⟨T,T⟩=ε1=±1,⟨N,N⟩=ε2=±1and⟨Bi,Bi⟩=εi+2=±1,i={1,2,...,n−2}. | (2.1) |
Then the Frenet equations of the curve α are as follows [18]:
T′(s)=ε2k1(s)N(s),N′(s)=−ε1k1(s)T(s)+ε3k2(s)B1(s),B′1(s)=−ε2k2(s)N(s)+ε4k3(s)B2(s),B′i(s)=−εi+1ki+1(s)Bi−1(s)+εi+3ki+2(s)Bi+1(s),B′n−2(s)=−εn−1kn−1(s)Bn−3(s). | (2.2) |
If the curve is not arclength parametrized, then the right-hand sides of (2.2) must be multiplied by the speed ν of α. We recall the functions ki(s) are called the i -th curvatures of for i={1,2,...,n−1}. All the curvatures satisfy ki(s)⟩0 for all s∈I,1⩽i⩽n−2. If kn−1(s)=0 for all s∈I, then Bn−2(s) is a constant vector and the curve lies in a (n−1) -dimensional affine subspace orthogonal to Bn−2 which is isometric to the Lorentzian (n−1) -space En−11. Thus the curve lies in a hyperplane if and only if in every point the position vector of a curve lies in the orthogonal complement of Bn−2. Analogously, if in every point the position vector of an arclength parametrized curve lies in the orthogonal complement of the tangent vector T, then the curve α lies on some hyperquadrics. Indeed, we see that the derivative of ⟨α,α⟩ is zero; hence ⟨α,α⟩ is a constant and thus lies on some hyperquadrics. Here the converse is also true. From this reasoning, we study curves for which in every point the position vector of the curve lies in the orthogonal complement of the tangent vector T [8].
Definition 2.1. A curve α:I⊂R→En1 is a normal curve if the orthogonal complement of T(s) contains a fixed point for all s∈I. Hereafter, since the orthogonal complement of T(s) is defined by T⊥(s)={v∈En1|⟨v,T(s)⟩=0}, the position vector of a normal curve holds (2.3) with λ,μ1,μ2,…μn−2 differentiable real functions
α(s)=λ(s)N(s)+μ1(s)B1(s)+...+μn−2(s)Bn−2(s). | (2.3) |
Let us note that the hyperquadrics Sn−11 and Hn−10 are defined by
Sn−11={ν∈En1|⟨v,v⟩=1},Hn−10={ν∈En1|⟨v,v⟩=−1} |
respectively. In the rest of this paper, we assume that all the curvatures of the curve are not identically zero.
In this section we give some characterizations of null normal curves in En1, n⟩4.
Let α:I⊂R→En1 be a null curve parametrized by the pseudo-arclength such that {α′(s),α″(s),...,αn(s)} is a basis of Tα(s)En1 for all s. Then there exists only one Frenet frame satisfying the equations
T′=N,N′=k1T−B1,B′1=−k1N+k2B2,B′2=−k2T+k3B3,B′i=−kiBi−1+ki+1Bi+1,i∈{3,...,n−3}B′n−2=−kn−2Bn−3 | (3.1) |
where
⟨T,T⟩=⟨B1,B1⟩=0,⟨T,B1⟩=1,⟨N,N⟩=⟨B2,B2⟩=...=⟨Bn−2,Bn−2⟩=1.. |
Let α(s) be a null normal curve in En1, parametrized by pseudo-arclength s. Then its position vector satisfies the equation
α(s)=λ(s)T(s)+μ1(s)N(s)+μ2(s)B2(s)+...+μn−2(s)Bn−2(s) | (3.2) |
for some differentiable functions λ(s),μ1(s),μ2(s),...,μn−2(s). Differentiating (3.2) with respect to s and by using (3.1), we obtain the system of equations
λ′+μ1k1−μ2k2=1, | (3.3) |
λ+μ1′=0, | (3.4) |
μ1=0, | (3.5) |
μ2′−μ3k3=0, | (3.6) |
μi′−μi−1ki−μi+1ki+1=0, i∈{2,3...,n−3}, | (3.7) |
μ′n−2+μn−3kn−2=0. | (3.8) |
From the Eqs. (3.4) and (3.5) we get μ1(s)=0 and λ(s)=0. Considering the differentiable functions λ(s),μ1(s),μ2(s),...,μn−2(s), we get the following theorem.
Theorem 3.1. Let α:I⊂R→En1 be a null normal curve in En1, parametrized by pseudo-arclength s. Then the following statements are hold:
ⅰ. The components of the position vector of α are
λ(s)=0, μ1(s)=0,μi(s)=i−2∑i=0μi,k(s)∂k∂sk(−1k2(s)) i∈{2,3,...,n−2} |
where the functions μi,k can be inductively defined by
μ1,0=0,μ2,0=1 |
and for i∈{3,4,...,n−2}
μi,0(s)=ki−1(s)μi−2,0(s)+μ′i−1,0(s)ki(s),μi,k(s)=ki−1(s)μi−2,k(s)+μi−1,k−1(s)+μ′i−1,k(s)ki(s),μi,i−3(s)=μi−1,i−4(s)+μi−1,i−3(s)ki(s),μi,i−2(s)=μi−1,i−3(s)ki(s). | (3.9) |
ⅱ. If k2,k3,…,kn−2 are nonzero constants, then α lies in pseudosphere Sn−11(r),r∈R+0.
Proof. ⅰ. Let α(s) be a null normal curve in En1, parametrized by pseudo-arclength s. Then its position vector is given by (3.2). Then the equation system (3.3)–(3.8) gives
λ(s)=0, μ1(s)=0, |
μ2(s)=−1k2(s), μ3(s)=1k3(s)(−1k2(s))′. | (3.10) |
Considering the functions μ1,0,μ2,0,μ3,0 and μ3,1, we have
μ1(s)=μ1,0(s)(−1k2(s)),μ1,0(s)=0,μ2(s)=μ2,0(s)(−1k2(s)),μ2,0(s)=1,μ3(s)=1k3(s)(−1k2(s))′,μ3,0(s)=0,μ3,1(s)=1k3(s). |
By induction from (3.7), we obtain
μi(s)=i−2∑k=0μi,k(s)∂k∂sk(−1k2(s)),2⩽i⩽n−2. | (3.11) |
Here the functions μi,k are defined by (3.9). This proves the statement (ⅰ).
ⅱ. If k2,k3,…,kn−2 are nonzero constants, then the components μ2,μ3…,μn−2 of the position vector of α(s) are constant numbers. Then the position vector of α(s) is
α(s)=μ2B2(s)+μ3B3(s)+...+μn−2Bn−2(s). |
From the last equation, we get
⟨α(s),α(s)⟩=μ22+μ23+...+μ2n−2=r2,r∈R+0, |
which means that α(s) lies in Sn−11(r) with center at the origin and the radius r. This proves the statement (ⅱ).
Theorem 3.2. Let α:I⊂R→En1 be a null curve in En1 with nonzero curvatures. Then α(s) is congruent to a normal curve if and only if
(n−4∑k=0μn−2,k(s)∂k∂sk(−1k2(s)))′+kn−2(s)(n−5∑k=0μn−3,k(s)∂k∂sk(−1k2(s)))=0. | (3.12) |
Proof. If α(s) is a null normal curve in En1, writing (3.11) for i=n−3 and i=n−2 in (3.8), we obtain (3.12).
Conversely, assume that (3.12) holds. Then we define the vector m(s)∈En1 given by
m(s)=α(s)−μ2(s)B2(s)−...−μn−2(s)Bn−2(s) | (3.13) |
with μ2,μ3,...,μn−2 as in (3.10) and (3.11). If we differentiate (3.13) with respect to s and using (3.1)
m′(s)=((n−4∑k=0μn−2,k(s)∂k∂sk(−1k2(s)))′+kn−2(s)(n−5∑k=0μn−3,k(s)∂k∂sk(−1k2(s))))Bn−2(s) | (3.14) |
gives m′(s)=0. Then m(s) is a constant vector and so α(s) is congruent to a null normal curve.
In this section, we first characterize the non-null normal curves in terms of their curvatures.
Let α:I⊂R→En1 be an arclength parametrized non-null normal curve in En1. The position vector of the curve satisfies (2.3) for smooth functions λ,μ1,μ2,...μn−2. Differentiating (2.3) with respect to s and using (2.2), we have
T(s)=−ε1k1(s)λ(s)T(s)+(λ′(s)−ε2k2(s)μ1(s))N(s)+(ε3k2(s)λ(s)+μ′1(s)−ε3k3(s)μ2(s))B1(s)+n−3∑i=2(εi+2ki+1(s)μi−1(s)+μ′i(s)−εi+2ki+2(s)μi+1(s))Bi(s)+(μ′n−2(s)+εnkn−1(s)μn−3(s))Bn−2(s) |
It follows that
1+ε1k1(s)λ(s)=0, | (4.1) |
λ′(s)−ε2k2(s)μ1(s)=0, | (4.2) |
ε3k2(s)λ(s)+μ′1(s)−ε3k3(s)μ2(s)=0, | (4.3) |
εi+2ki+1(s)μi−1(s)+μ′i(s)−εi+2ki+2(s)μi+1(s)=0, i∈{2,3,...,n−3} | (4.4) |
μ′n−2(s)+εnkn−1(s)μn−3(s)=0. | (4.5) |
This system consists of n equations and (n−1) curvature functions, the function λ and (n−2) functions μi. Thus the coefficient functions μi can be expressed in terms of the curvature functions, derivatives of the curvature functions and the function λ. From (4.1), we have
λ(s)=−ε1k1(s). | (4.6) |
Using the coefficient (4.6) in (4.2), we get
μ1(s)=−ε1ε21k2(s)(1k1(s))′. | (4.7) |
Similarly using the coefficient (4.7) in (4.3), we obtain
μ2(s)=−ε1ε3k3(s)[ε3k2(s)k1(s)+ε2(1k2(s)(1k1(s))′)′]. |
When the other coefficient functions are calculated, long and complex expressions with curvature functions appear. Considering the functions μ1,0,μ2,0,μ2,1 and μ2,2, we have
μ1(s)=μ1,1(s)(1k1(s))′,μ1,1(s)=−ε1ε21k2(s), | (4.8) |
μ2(s)=μ2,0(s)(1k1(s))+μ2,1(s)(1k1(s))′+μ2,2(s)(1k1(s))′′ | (4.9) |
where
μ2,0(s)=−ε1k2(s)k3(s),μ2,1(s)=−ε1ε2ε3(1k3(s))(1k2(s))′,μ2,2(s)=−ε1ε2ε3(1k3(s))(1k2(s)). | (4.10) |
Similarly introducing the functions μ3,0,μ3,1,μ3,2 and μ33 we have
μ3(s)=μ3,0(s)(1k1(s))+μ3,1(s)(1k1(s))′++μ3,2(s)(1k1(s))′′+μ3,3(s)(1k1(s))′′′ | (4.11) |
where
μ3,0(s)=−ε1ε4k4(s)(k2(s)k3(s))′,μ3,1(s)=−ε1ε4k4(s)[ε2ε4(k3(s)k2(s))+(k2(s)k3(s))+ε2ε3(1k3(s)(1k2(s))′)′],μ3,2(s)=−ε1ε2ε3ε4k4(s)[1k3(s)(1k2(s))′+(1k3(s)(1k2(s)))′],μ3,3(s)=−ε1ε2ε3ε4k4(s)(1k3(s))(1k2(s)). | (4.12) |
By induction from (4.4), we obtain
μi(s)=i∑k=0μi,k(s)∂k∂sk(1k1(s)),1⩽i⩽n−2. | (4.13) |
Here the functions μi,k can be inductively defined by
{μ1,0(s)=0,μ1,1(s)=−ε1ε21k2(s),μ2,0(s)=−ε1k2(s)k3(s),μ2,1(s)=−ε1ε2ε3(1k3(s))(1k2(s))′,μ2,2(s)=−ε1ε2ε3(1k3(s))(1k2(s)),μi,0(s)=εi+1ki+1(s)(εi+1ki(s)μi−2,0(s)+μ′i−1,0(s)),μi,k(s)=εi+1ki+1(s)(εi+1ki(s)μi−2,k(s)+μ′i−1,k(s)+μi−1,k−1(s)),μi,i−1(s)=εi+1ki+1(s)(μi−1,i−2(s)+μ′i−1,i−1(s)),μi,i(s)=εi+1ki+1(s)μi−1,i−1(s). | (4.14) |
where k∈{1,2,...,i−3} and i∈{3,4,...,n−2}. Substituting Eqs. (4.6) and (4.13) into (2.3), we get the position vector of the normal curve as:
α(s)=−ε1k1(s)N(s)+n−2∑i=1(i∑k=0μi,k(s)∂k∂sk(1k1(s)))Bi(s). | (4.15) |
Then based on the Eqs system (4.1)–(4.5), we state the following theorem:
Theorem 4.1. Let α:I⊂R→En1 be an arclength parametrized curve in En1 with nonzero curvatures. Then α(s) is congruent to a normal curve if and only if
(n−2∑k=0μn−2,k(s)∂k∂sk(1k1(s)))′+εnkn−1(s)(n−3∑k=0μn−3,k(s)∂k∂sk(1k1(s)))=0 | (4.16) |
with μi,k inductively defined by the system (4.14).
Proof. If α(s) is a normal curve, writing (4.13) for i=n−3 and i=n−2 in (4.5), we obtain (4.16).
Conversely, assume that (4.16) holds. Then we define the vector m(s)∈En1 given by
m(s)=α(s)−λ(s)N(s)−μ1(s)B1(s)−...−μn−2(s)Bn−2(s) | (4.17) |
with λ(s)=−ε1k1(s) and μ1(s),μ2(s),...,μn−2(s) as in (4.8), (4.9) and (4.13). If we differentiate (4.17) with respect to s and by using (2.2)
m′(s)=((n−2∑k=0μn−2,k(s)∂k∂sk(1k1(s)))′+εnkn−1(s)(n−3∑k=0μn−3,k(s)∂k∂sk(1k1(s))))Bn−2(s) | (4.18) |
gives m′(s)=0. Then m(s) is a constant vector and so α(s) is congruent to a normal curve.
Now, assume that all the curvature functions k1,k2,...,kn−1 of normal curve are nonzero constants. Then, we give the following result:
Theorem 4.2. For odd n, there exists no normal curve with nonzero constant curvatures and for even n, every curve with nonzero constant curvatures is a normal curve in En1.
Proof. Assume that there exists a normal curve with its nonzero constant curvatures k1,k2,...,kn−1. From (4.1), (4.2), (4.3) and (4.4), it follows that
λ=−ε1k1,μ1=0,μ2=−ε1k2k1.k3,μ3=0. |
For i∈{4,5,...,n−2} Eq. (4.4) gives
μi+1(s)=εi+2ki+2(s)(εi+2ki+1(s)μi−1(s)+μ′i(s)). |
By induction we obtain that
μ2m−1=0, | (4.19) |
μ2m=−ε1m∏i=1k2im+1∏i=1k2i−1. | (4.20) |
For odd n, with the help of (4.19) and (4.20), Eq. (4.16) takes the following form
ε1εnk2k4...kn−3kn−1k1k3...kn−4kn−2=0. |
However, since we assume all curvatures to be nonzero, this leads to a contradiction. For even n, according to Theorem 4.1., since the curvature functions obviously satisfy the relation (4.16) then α(s) is congruent to normal curve. Thus the proof is completed.
Example 4.1. A curve β:I⊂R→En1 which has all its curvatures constant is parametrized by
β(s)=(a1cosh(b1s),a1sinh(b1s),a2cos(b2s),a2sin(b2s),...,atcos(bts),atsin(bts)) | (4.21) |
for even n=2t and by
β(s)=(a1cosh(b1s),a1sinh(b1s),a2cos(b2s),a2sin(b2s),...,atcos(bts),atsin(bts),cs) | (4.22) |
for odd n=2t+1. Here c,ai,bi∈R and all bi are different numbers for i={1,2,...,t}.
From the parametrization (4.21), for even n, we obtain the derivative of ⟨β,β⟩ is zero. Then ⟨β,β⟩=constant and thus β lies on some hyperquadrics in En1. This means that in every point the position vector of the curve lies in the orthogonal complement of the tangent vector T. So the curve β is a normal curve. Also we can easily show that the curve β with all its curvatures constant is a normal curve since ⟨β(s),T(s)⟩=0. From the parametrization (4.22), for odd n, the curve β with all its curvatures are constant is not a normal curve since ⟨β(s),T(s)⟩≠0.
Theorem 4.3. Let α=α(s) be an arclength parametrized curve, lying fully in the n -dimensional Lorentzian space with nonzero curvatures. Then α is a normal curve if and only if α lies in some hyperquadrics in En1.
Proof. First assume that α(s) is congruent to a normal curve. It follows, by straightforward calculations using Theorem 4.1., we obtain
2ε2(1k1(s))(1k1(s))′+2ε3(1k2(s)(1k1(s))′)(1k2(s)(1k1(s))′)′+2ε4{1k3(s)[ε3k2(s)k1(s)+ε2(1k2(s)(1k1(s))′)′]}×{1k3(s)[ε3k2(s)k1(s)+ε2(1k2(s)(1k1(s))′)′]}′+...+2εi+2(i∑k=0μi,k∂k∂sk(1k1(s)))(i∑k=0μi,k∂k∂sk(1k1(s)))′+...+2εn(n−2∑k=0μn−2,k∂k∂sk(1k1(s)))(n−2∑k=0μn−2,k∂k∂sk(1k1(s)))′=0. |
On the other hand, the previous equation is differential of the equation
ε2(1k1(s))2+ε3(1k2(s)(1k1(s))′)2+ε4{1k3(s)[ε3k2(s)k1(s)+ε2(1k2(s)(1k1(s))′)′]}2+...+εi+2(i∑k=0μi,k∂k∂sk(1k1(s)))2+...+εn(n−2∑k=0μn−2,k∂k∂sk(1k1(s)))2=r,r∈R. | (4.23) |
Then using (4.6)–(4.13) in (4.17), we get ⟨α(s)−m,α(s)−m⟩=r. Consequently α(s) lies in some hyperquadrics in En1.
Conversely, if α(s) lies in some hyperquadrics in En1, then ⟨α(s)−m,α(s)−m⟩=r, r∈R where m(s)∈En1 is a constant vector. By taking the derivative of the previous equation with respect to s, we obtain ⟨α(s)−m,T(s)⟩=0, which means that α(s) is a normal curve.
Theorem 4.4. Let α:I⊂R→En1 be an arclength parametrized normal curve in En1 with nonzero curvatures. Then the following statements are hold:
ⅰ. The normal component and the first binormal component of the position vector of the curve are given by
⟨α(s),N(s)⟩=−ε1ε2k1(s),⟨α(s),B1(s)⟩=−ε1ε2ε3k2(s)(1k1(s))′. | (4.24) |
ⅱ. The first binormal component and the second binormal component of the position vector of the curve are given by
⟨α(s),B1(s)⟩=−ε1ε2ε3k2(s)(1k1(s))′,⟨α(s),B2(s)⟩=−ε1ε3ε4k3(s)[ε3k2(s)k1(s)+ε2(1k2(s)(1k1(s))′)′]. | (4.25) |
ⅲ. The second binormal component and the third binormal component of the position vector of the curve are given by
⟨α(s),B2(s)⟩=−ε1ε3ε4k3(s)[ε3k2(s)k1(s)+ε2(1k2(s)(1k1(s))′)′],⟨α(s),B3(s)⟩=−ε1ε5k4(s)[ε2(k3(s)k2(s)(1k1(s))′)+ε4{1k3(s)[k2(s)k1(s)+ε2ε3(1k2(s)(1k1(s))′)′]}′], | (4.26) |
ⅳ. The jth binormal component and the (j+1)th binormal component of the position vector of the curve are given by
⟨α(s),Bj(s)⟩=εj+2j∑k=0μj,k(s)∂k∂sk(1k1(s)),⟨α(s),Bj+1(s)⟩=εj+3j+1∑k=0μj+1,k(s)∂k∂sk(1k1(s)) | (4.27) |
where 3⩽j⩽n−3 and μj,k is introduced by (4.14).
ⅴ. The distance function ρ(s)=‖α(s)‖ satisfies ρ2(s)=|a| for some a∈R
ⅵ. The distance function ρ(s)=‖α(s)‖ is constant and the binormal component αB(s) of the position vector of the curve has nonconstant length.
Conversely, if α(s) is an arclength parametrized curve with nonzero curvatures and one of the above statements holds in En1, then α(s) is congruent to a normal curve.
Proof. To prove (ⅰ), (ⅱ), (ⅲ) and (ⅳ), assume that α(s) is an arclength parametrized normal curve in En1. Taking the inner product of the two sides (4.15) with N(s),B1(s),B2(s),...,Bj(s),Bj+1(s) where 3⩽j⩽n−3 respectively, we obtain the statements (ⅰ), (ⅱ), (ⅲ), (ⅳ).
Conversely, assume that (ⅰ) is given. Differentiating ⟨α(s),N(s)⟩=−ε1ε2k1(s) with respect to s and by using (2.2), we get ⟨α(s),T(s)⟩=0, which means α(s) is congruent to a normal curve. Similarly since the statements (ii), (iii), (iv) holds, then α(s) is a normal curve.
To prove (v), assume that α(s) is an arclength parametrized normal curve. Then multiplying (4.3), (4.4), and (4.5) with εi+2μi(s) where i∈{1,2,...,n−2} respectively,
ε3μ1(s)(μ1′−ε3k3(s)μ2(s))=−ε2(1k1(s))(1k1(s))′,εi+2μi(s)(εi+2ki+1(s)μi−1(s)+μi′−εi+2ki+2(s)μi+1(s))=0,εnμn−2(s)(εnkn−1(s)μn−3(s)+μ′n−2(s))=0 |
and adding these equations we get n−2∑i=1εi+2μ2i=−ε2(1k1(s))2+a for a∈R. From (4.15), we have
ρ2(s)=|⟨α(s),α(s)⟩|=|ε2λ2+n−2∑i=1εi+2μ2i|=|a|. |
Conversely, differentiating ρ2(s)=|⟨α(s),α(s)⟩|=|a| with respect to s, we get ⟨α(s),T(s)⟩=0. Thus, α(s) is congruent to a normal curve.
ⅵ. Decompose the position vector of a curve α(s) in its normal and binormal component, i.e.,
α(s)=ε2⟨α(s),N(s)⟩N(s)+αB(s). |
From αB(s)=n−2∑i=1μi(s)Bi(s), we have ‖αB(s)‖=√|−ε2(1k1(s))2+a|. Thus, the binormal component has nonconstant length. The distance function ρ is proved in (ⅴ).
Conversely since ρ2(s)=‖α(s)‖2 is constant, ⟨α(s),T(s)⟩=0. Hence, α(s) is congruent to a normal curve.
Lemma 4.1. Let α:I⊂R→En1 be an arclength parametrized curve with non-null vector fields N,B1,B2,…,Bn−2, lying fully in En1, then α(s) is congruent to a normal curve if and only if there exists a differentiable function f(s) such that
f(s)kn−1(s)=(n−2∑k=0μn−2,k(s)kn−1(s)∂k∂sk(1k1(s))),f′(s)=−εnkn−1(s)(n−3∑k=0μn−3,k(s)∂k∂sk(1k1(s))) | (4.28) |
Applying similar methods as in [19,20,21] together with Lemma 4.1., we obtain the following theorems for normal curves in En1.
Theorem 4.5. Let α(s) be an arclength parametrized curve in En1 with nonzero curvatures and timelike principal binormal Bn−2. Then α(s) is congruent to a normal curve if and only if there exist constants a0,b0∈R such that
f′(s)kn−1(s)={a0−∫(f″(s)kn−1(s)−f′(s)k2n−1(s)k′n−1(s)−f(s)kn−1(s))sinhθ(s)ds}sinhθ(s)−{b0−∫(f″(s)kn−1(s)−f′(s)k2n−1(s)k′n−1(s)−f(s)kn−1(s))coshθ(s)ds}coshθ(s) | (4.29) |
where
Proof. Let α(s) is congruent to a normal curve with εn=−1. According to Lemma 4.1., there exists a differentiable function f(s) such that the relation (4.28) holds. Let us determine the differentiable functions θ(s),a(s) and b(s) by
θ(s)=∫s0kn−1(s)ds, | (4.30) |
a(s)=−f′(s)kn−1(s)sinhθ(s)+f(s)coshθ(s)+∫[f″(s)kn−1(s)−f′(s)k2n−1(s)k′n−1(s)−f(s)kn−1(s)]sinhθ(s)ds, | (4.31) |
b(s)=−f′(s)kn−1(s)coshθ(s)+f(s)sinhθ(s)+∫[f″(s)kn−1(s)−f′(s)k2n−1(s)k′n−1(s)−f(s)kn−1(s)]coshθ(s)ds. | (4.32) |
By using (4.28), we find a′(s)=0 and b′(s)=0. Thus a(s)=a0 and b(s)=b0∈R. Multiplying (4.31) and (4.32) respectively with sinhθ(s) and −coshθ(s), adding the obtained equations, we get (4.29).
Conversely let a0,b0∈R are constants such that the relation (4.29) holds. Let us define the differentiable function f(s) by
f(s)=1kn−1(s)[n−2∑k=0μn−2,k(s)kn−1(s)∂k∂sk(1k1(s))]. |
By the derivative of (4.29) with respect to s, we obtain
f(s)={a0−∫(f″(s)kn−1(s)−f′(s)k2n−1(s)k′n−1(s)−f(s)kn−1(s))sinhθ(s)ds}coshθ(s)−{b0−∫(f″(s)kn−1(s)−f′(s)k2n−1(s)k′n−1(s)−f(s)kn−1(s))coshθ(s)ds}sinhθ(s). |
As a result of this and (4.29), we obtain f′(s)=(n−3∑k=0μn−3,k(s)∂k∂sk(1k1(s)))kn−1(s). Thus Lemma 4.1. implies that α(s) is congruent to a normal curve.
For the curves with spacelike principal binormal Bn−2, we obtain the following theorem, which can be proved in a similar way as Theorem 4.5.
Theorem 4.6. Let α(s) be an arclength parametrized curve in En1 with nonzero curvatures and spacelike principal binormal Bn−2. Then α(s) is congruent to a normal curve if and only if there exist constants a0,b0∈R such that
f′(s)kn−1(s)={∫(f″(s)kn−1(s)−f′(s)k2n−1(s)k′n−1(s)−f(s)kn−1(s))cosθ(s)ds−a0}cosθ(s)−{∫(f″(s)kn−1(s)−f′(s)k2n−1(s)k′n−1(s)−f(s)kn−1(s))sinθ(s)ds−b0}sinθ(s). | (4.33) |
This study gives normal curves and examines some characterizations of normal curves in Lorentzian n -space En1. We determine necessary and sufficient condition for a null and non-null curve to be congruent to a normal curve in Lorentzian n -space En1. We characterize normal curves in terms of their curvature functions. The results of this study may also be developed to other different spaces.
The author declares no conflicts of interest in this paper.
[1] | K. Honda, J. Inoguchi, Cesaro's method for Cartan framed null curves, Preprint, 2003. |
[2] |
B. Y. Chen, When does the position vector of a space curve always lie in its rectifying plane?, Am. Math. Mon., 110 (2003), 147-152. doi: 10.1080/00029890.2003.11919949
![]() |
[3] | B. Y. Chen, F. Dillen, Rectifying curves as centrodes and extremal curves, Bull. Inst. Math. Acad., 2 (2005), 77-90. |
[4] | K. İlarslan, E. Nešovic, M. Petrović-Torgašev, Some characterizations of rectifying curves in the Minkowski 3-space, Novi. Sad. J. Math., 33 (2003), 23-32. |
[5] | S. Izumiya, N. Takeuchi, New special curves and developable surfaces, Turk. J. Math., 28 (2004), 153-163. |
[6] | K. İlarslan, E. Nešovic, Some characterizations of rectifying curves in the Euclidean space E4, Turk. J. Math., 32 (2008), 21-30. |
[7] | T. Ali, M. Önder, Some characterizations of spacelike rectifying curves in the Minkowski space-time, Glob. J. Sci. Front. Res. Math. Decision Sci., 12 (2012), 57-64. |
[8] |
S. Cambie, W. Goemans, I. Van Den Bussche, Rectifying curves in then n -dimensional Euclidean space, Turk. J. Math., 40 (2016), 210-223. doi: 10.3906/mat-1502-77
![]() |
[9] | Ö. B. Kalkan, H. Öztürk, On rectifying curves in Lorentzian n-space Evn, C. R. Acad. Bulg. Sci., 72 (2019), 158-169. |
[10] | K. İlarslan, Spacelike normal curves in Minkowski 3-space E13, Turk. J. Math., 29 (2005), 53-63. |
[11] | K. İlarslan, E. Nešovic, Timelike and null normal curves in Minkowski space E13, Indian J. Pure Appl. Math., 35 (2004), 881-888. |
[12] |
K. İlarslan, E. Nešovic, Spacelike and timelike normal curves in Minkowski space-time, Publ. I. Math. Nouvelle série, 85 (2009), 111-118. doi: 10.2298/PIM0999111I
![]() |
[13] | M. Grbovic, E. Nešovic, Some relations between rectifying and normal curves in Minkowski 3-space, Math. Commun., 17 (2012), 655-664. |
[14] | H. Öztekin, Normal and rectifying curves in Galilean space G3, Proc. IAM, 5 (2016), 98-109. |
[15] | H. Öztekin, A. O. Öğrenmiş, Normal and rectifying curves in pseudo-Galilean space G13 and their characterizations, J. Math. Comput. Sci., 2 (2012), 91-100. |
[16] | Ö. Bektaş, Normal curves in the n-dimensional Euclidean space, Adv. Differ. Equ., 456 (2018), 1-12. |
[17] | B. O'Neill, Semi-Riemannian Geometry with Applications to Relativity, Academic Press, London, 1983. |
[18] | E. İyigün, K. Arslan, On Harmonic Curvatures of Curves in Lorentzian n-Space, Commun. Fac. Sci. Univ. Ank. Series A1, 54 (2005), 29-34. |
[19] | Ç. Camcı, K. İlarslan, E. Šućurović, On pseudohyperbolical curves in Minkowski space-time, Turk. J. Math., 27 (2003), 315-328. |
[20] |
S. Breuer, D. Gottlieb, Explicit characterization of spherical curves, P. Am. Math. Soc., 27 (1971), 126-127. doi: 10.1090/S0002-9939-1971-0270275-2
![]() |
[21] | Y. C. Wong, On an explicit characterization of spherical curves, P. Am. Math. Soc., 34 (1972), 239-242. |
1. | Xin Zhao, Donghe Pei, Pedal Curves of the Mixed-Type Curves in the Lorentz-Minkowski Plane, 2021, 9, 2227-7390, 2852, 10.3390/math9222852 | |
2. | Safaa Mosa, Mohamed El-Fakharany, Mervat Elzawy, Normal Curves in 4-Dimensional Galilean Space G4, 2021, 9, 2296-424X, 10.3389/fphy.2021.660241 |