Abstract
Dynamic simulation of revolved solids plays an important role in many fields. Aiming at the lacks of solutions in some key aspects, this study establishes governing equation of motion based on theory of variational inequality; designs a compatibility iteration algorithm for solving contact forces; deduces parametric equations of arbitrary cylinder and cone in three-dimensional space; provides corresponding analytical methods to identify contact points between bodies and to calculate volume integral over bodies; proposes a rotation matrix modification approach to conserve volume and shape of rigid body in the case of large rotation. The accuracy, availability, competence, robustness, and application prospects of the presented methodology are demonstrated by several interesting and challenging problems.
Similar content being viewed by others
References
Ginsberg J (2008) Engineering dynamics, vol 10. Cambridge University Press, New York
Wilson EL, Farhoom I, Bathe KJ (1973) Nonlinear dynamic analysis of complex structure. Earthq Eng Struct Dyn 1:241–252
Hilber HM, Hughes TJR, Taylor RL (1977) Improved numerical dissipation for time integration algorithms in structural mechanics. Earthq Eng Struct Dyn 5(3):283–292
Belytschko T (1980) Explicit time integration of structure-mechanical systems. Adv Struct Dyn 1:97–122
Tamma KK, D’Costa JF (1992) New explicit variable time-integration self-starting methodology for computational structural dynamics. Int J Numer Methods Eng 33(6):1165–1180
Chung J, Hulbert GM (1993) A time integration algorithm for structural dynamics with improved numerical dissipation: the generalized-α method. J Appl Mech 60(2):371–375
Hulbert GM, Chung J (1996) Explicit time integration algorithms for structural dynamics with optimal numerical dissipation. Comput Methods Appl Mech Eng 137(2):175–188
Wu YS, Smolinski P (2000) A multi-time step integration algorithm for structural dynamics based on the modified trapezoidal rule. Comput Methods Appl Mech Eng 187(3–4):641–660
Gravouil A, Combescure A (2001) Multi-time-step explicit-implicit method for non-linear structural dynamics. Int J Numer Methods Eng 50(1):199–225
Noels L, Stainier L, Ponthot JP (2004) Combined implicit/explicit time-integration algorithms for the numerical simulation of sheet metal forming. J Comput Appl Math 168(1):331–339
Zhong WX (2004) On precise integration method. J Comput Appl Math 163(1):59–78
Courant R (1943) Variational methods for the solution of problems of equilibrium and vibrations. Bull Am Math Soc 49:1–23
Feng K (1965) Difference schemes based on variational principle. J Appl Comput Mech 2–4:238–262 (In Chinese)
Lucy LB (1977) A numerical approach to the testing of the fission thesis. Astron J 82:1013–1024
Banerjee PK, Butterfield R (1981) Boundary element methods in engineering science. McGraw-Hill, London
Shi GH (1991) Manifold method of material analysis. In: Transactions of the 9th army conference on applied mathematics and computing, Minneapolis, MN, pp 57–76
Babuska I, Melenk JM (1997) The partition of unity method. Int J Numer Methods Eng 40:727–758
Strouboulis T, Babuka I, Copps K (2000) The design and analysis of the generalized finite element method. Comput Methods Appl Mech Eng 181:43–69
Bardenhagen SG, Brackbill JU, Sulsky D (2000) The material-point method for granular materials. Comput Methods Appl Mech Eng 187(3–4):529–541
Calvo B, Martinez MA, Doblare M (2005) On solving large strain hyperelastic problems with the natural element method. Int J Numer Methods Eng 62:159–185
Belytschko T, Gracie R, Ventura G (2009) A review of extended/generalized finite element methods for material modeling. Model Simul Mater Sci Eng 17(4):043001
Zienkiewicz OC, Taylor RL, Zhu JZ (2013) The finite element method: Its basis and fundamentals, 7th edn. Elsevier, Oxford
Fan H, Zheng H, He SM (2016) S–R decomposition based numerical manifold method. Comput Methods Appl Mech Eng 304:452–478
Cundall PA, Strack ODL (1979) A discrete numerical model for granular assemblies. Geotechnique 29(1):47–65
Shi GH (1988) Discontinuous deformation analysis: a new numerical model for the statics and dynamics of block system. PhD thesis. Univ Calif Berkeley, Berkeley, California
Shi GH (1996) Simplex integration for manifold method, FEM, DDA and analytical analysis. In: Proceedings of the 1th internat. Forum on discontinuous deformation analysis, DDA and simulations of discontinuous media. TSI Press, Albuquerque, New Mexico, USA, pp 205–262
Cheng YM, Zhang YH (2000) Rigid body rotation and block internal discretization in DDA analysis. Int J Numer Anal Methods Geomech 24(6):567–578
Shi GH (2001) Three dimensional discontinuous deformation analyses. In: Proceedings of the 38th US rock mechanics symposium, Washington DC, pp 1–21
Zhang YH, Cheng YM (2002) Coupling of FEM and DDA methods. Int J Geomech 2(4):503–517
Doolin DM, Sitar N (2004) Time integration in discontinuous deformation analysis. J Eng Mech 130(3):249–258
Doolin DM (2005) Unified displacement boundary constraint formulation for discontinuous deformation analysis (DDA). Int J Numer Anal Methods Geomech 29(12):1199–1207
MacLaughlin MM, Doolin DM (2006) Review of validation of the discontinuous deformation analysis (DDA) method. Int J Numer Anal Methods Geomech 30(4):271–305
Zhao ZY, Gu J (2009) Stress recovery procedure for discontinuous deformation analysis. Adv Eng Software 40(1):52–57
Miki S, Sasaki T, Koyama T, Nishiyama S, Ohnishi Y (2010) Development of coupled discontinuous deformation analysis and numerical manifold method (NMM–DDA). Int J Comput Methods 7(1):1–20
Koyama T, Nishiyama S, Yang M, Ohnishi Y (2011) Modeling the interaction between fluid flow and particle movement with discontinuous deformation analysis (DDA) method. Int J Numer Anal Methods Geomech 35(1):1–20
Beyabanaki SAR, Bagtzoglou AC (2012) Three-dimensional discontinuous deformation analysis (3-D DDA) method for particulate media applications. Geomech Geoeng 7(4):239–253
Bao HR, Zhao ZY (2012) The vertex-to-vertex contact analysis in the two-dimensional discontinuous deformation analysis. Adv Eng Softw 45(1):1–10
Wu W, Zhu HH, Zhuang XY, Ma GW, Cai YC (2014) A multi-shell cover algorithm for contact detection in the three dimensional discontinuous deformation analysis. Theor Appl Fract Mech 72:136–149
Fan H, He SM (2015) An angle-based method dealing with vertex–vertex contact in the two-dimensional discontinuous deformation analysis (DDA). Rock Mech Rock Eng 48(5):2031–2043
Shi GH (2015) Contact theory. Sci China Technol Sci 8:1–47
Tang CA, Tang SB, Gong B, Bai HM (2015) Discontinuous deformation and displacement analysis: from continuous to discontinuous. Sci China Technol Sci 58(9):1567–1574
Morgan WE, Aral MM (2015) An implicitly coupled hydro-geomechanical model for hydraulic fracture simulation with the discontinuous deformation analysis. Int J Rock Mech Min Sci 73:82–94
Yagoda-Biran G, Hatzor YH (2016) Benchmarking the numerical discontinuous deformation analysis method. Comput Geotech 71:30–46
Fan H, Zheng H, Zhao JD (2017) Discontinuous deformation analysis based on strain-rotation decomposition. Int J Rock Mech Min Sci 92:19–29
Fu GY, Ma GW, Qu XL (2017) Boundary element based discontinuous deformation analysis. Int J Numer Anal Methods Geomech 41(7):994–1015
Fan H, Zheng H, Zhao JD (2018) Three-dimensional discontinuous deformation analysis based on strain-rotation decomposition. Comput Geotech 95:191–210
Fan H, Zheng H, Wang JF (2018) A generalized contact potential and its application in discontinuous deformation analysis. Comput Geotech 99:104–114
Cai YE, Liang GP, Shi GH, Cook NGW (1996) Studying an impact problem by using LDDA method. In: Proceedings of the first international forum on discontinuous deformation analysis (DDA) and simulations of discontinuous Media, Berkeley, CA. TSI Press, Albuquerque, pp 288–295
Lin CT, Amadei B, Jung J, Dwyer J (1996) Extensions of discontinuous deformation analysis for jointed rock masses. Int J Rock Mech Min Sci 33:671–694
Jiang W, Zheng H (2011) Discontinuous deformation analysis based on variational inequality theory. Int J Comput Methods 8:193–208
Zheng H, Zhang P, Du XL (2016) Dual form of discontinuous deformation analysis. Comput Methods Appl Mech Eng 305:196–216
Fan H, Zhao JD, Zheng H (2018) Variational inequality-based discontinuous deformation analysis for discrete particle-block system. Int J Numer Methods Eng 115:1–37
Wu JH, Ohnishin Y, Nishiyama S (2005) A development of the discontinuous deformation analysis for rock fall analysis. Int J Numer Anal Methods Geomech 29:971–988
Fu XL (2008) A two-stage prediction-correction method for solving monotone variational inequalities. J Comput Appl Math 214(2):345–355
Newmark NN (1959) A method of computation for structural dynamics. J Eng Mech Div Proc ASCE 85(EM3):67–94
He BS, Liao LZ (2002) Improvements of some projection methods for monotone nonlinear variational inequalities. J Optim Theory Appl 112:111–128
Funding
This work is supported by General Research Fund Nos. CityU 11272916, CityU 11213517 from the Research Grant Council of the Hong Kong SAR and National Science Foundation of China (Grant No. 51779213).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix 1: Several Parameter Equations
1.1 Parameter Equation of Line Segment
In Fig. 53, any point C(x, y, z) on line AB can be written as
where t is the parameter associated with point C. If 0 ≤ t ≤ 1, point C is on line segment AB; otherwise, point C is outside of the line segment.
1.2 Parameter Equation of Plane
A plane with the normal vector n is determined by points A, B, and C, as shown in Fig. 54. An arbitrary point D(x, y, z) on this plane can be expressed by
where λ and μ are the parameters. Vector τ(m, n, p) is the direction vector of line BC.
1.3 Parameter Equation of Circle
In three-dimensional space, as depicted in Fig. 55, point C(xC, yC, zC) is the center of circle, n(nx, ny, nz) = n(cosα, cosβ, cosγ) is the unit normal vector of the circular plane. α, β, and γ are the angles between n and x-, y-, and z-axis, respectively. By taking point C as the origin of coordinates, we can make a local Cartesian coordinate system o*–u–v–w, in which the positive direction of w-axis is the same as that of vector n. The construction of the coordinate system o*–u–v–w is as follows: We move the unit normal vector n to the origin o of the global Cartesian coordinate system (see Fig. 55), where vectors nx and ny are the two components of vector n regarding to x- and y-axis, respectively. While vector nxy is the projection of vector n onto o–x–y plane. By rotating nxy by 90° around z-axis, we have another vector ξ(ξx, ξy, ξz), as described in Fig. 55. Obviously, through rotating the global Cartesian coordinate system o–x–y–z by γ angle around vector ξ, and then by moving it to point C, we can obtain the local Cartesian coordinate system o*–u–v–w. This is to say that the following coordinate transformation between the two systems holds:
where
where (ξx, ξy, ξz) is the direction cosine of vector ξ. Note that ξz = 0, Eq. (51) can be simplified as
On the other hand, in the local Cartesian coordinate system the circle can be expressed as
where R is the radius of the circle. Additionally, the angle parameter θ is measured according to the rule of right-hand. Substitution of Eqs. (53) into (50) yields the parameter equation of circle as follows:
1.4 Parameter Equation of Cylinder Surface
Thanks to the parameter equation of circle, given a bottom circle with the unit normal vector n(cosα, cosβ, cosγ) and the radius R, in three-dimensional space a cylinder with the height H (see Fig. 56) can be expressed by
where the coefficients Ai, Bi, and Ci (i = x, y, z) are defined by Eq. (52). In addition, θ and s are the two parameters. Point C(xC, yC, zC) and R are the center and radius of the bottom circle, while point C* is the center of the upper circle. In addition, the s-axis is local one-dimensional axis pointing the center C to the center C*. If s = 0, Eq. (53) gives us the bottom circle; while if s = H, the upper circle can be obtained.
1.5 Parameter Equation of Cone Surface
Based on the parameter equation of circle, for arbitrary cone with the unit normal vector n(cosα, cosβ, cosγ) of the bottom circular plane, as sketched in Fig. 57, it can be described as
where θ and s are the two parameters. Point C(xC, yC, zC) and R are the center and radius of the bottom circle, H is the height of the cone. And the s-axis is local one-dimensional axis along the height direction from center C to cone tip V. If s = 0, Eq. (56) gives us the bottom circle; while if s = H, we can get the cone tip.
Appendix 2: Volume Integral of Arbitrary Sphere, Cylinder, and Cone
For an arbitrary sphere, cylinder, and cone in three-dimensional space, during the generation of the governing equation of motion in this study, the involved volume integral terms can be boiled down to \(\iiint_{{}} {\left[ {{\varvec{T}}(x,y,z)} \right]^{\text{T}} dxdydz}\) and \(\iiint_{{}} {\left[ {{\varvec{T}}(x,y,z)} \right]^{\text{T}} {\varvec{T}}(x,y,z)dxdydz}\). The former is related to the body force sub-matrix, while the latter is associated with the mass sub-matrix. The corresponding integrand is the linear function or the quadratic function of x, y, and z, respectively. Indeed, we only need to calculate the volume integrals of ten monomials: 1, x, y, z, x2, y2, z2, yz, zx, and xy with respect to the global Cartesian coordinate system o–x–y–z. For this end, we first calculate the volume integrals of ten monomials: 1, u, v, w, u2, v2, w2, uv, vw, and wu in regard to the local Cartesian coordinate system o*–u–v–w. Then, by means of the coordinate transformation to achieve the original purpose.
For a given sphere in three-dimensional space, by translating the origin of global Cartesian coordinate systemt to the center of the sphere, we get a local Cartesian coordinate system o*–u–v–w, as shown in Fig. 58. By using the local spherical coordinate system, we have
where R is the radius of the sphere.
For a given cylinder in three-dimensional space, we can construct a local Cartesian coordinate system o*–u–v–w, as displayed in Fig. 59, where origin o* of local coordinate system is coincide with center C of the bottom circle of the cylinder. And the positive direction of w-axis is the same as the unit inner normal vector n(cosα, cosβ, cosγ) of the bottom circle. Acoording to the substitution rule for definite integral, we have
where J is the Jacobian matrix, reading
Substitution of the coordinate tranformation given by Eq. (50) into Eq. (59) yields \(\left| {\varvec{J}} \right| \equiv 1\).
On the other hand, with respect to the local Cartesian coordinate system o*–u–v–w, the cylinder surface can be expressed as
where R and H are the radius of the bottom circle and the height of the cylinder, respectively. If the local cylindrical coordinate system is adopted, which means
where
Then, we have
For a specified cone in three-dimensional space, we can also make a local Cartesian coordinate system o*–u–v–w, as shown in Fig. 60, in which the origin o* of local coordinate system is the center C of the bottom circle of the cone. In addition, the positive direction of w-axis is identical to the unit inner normal vector n(cosα, cosβ, cosγ), see Fig. 59. Obviously, there exsits the coordinate transformation described in Eq. (50) between the global and lobal Cartesian coordinate systems.
In the local Cartesian coordinate system o*–u–v–w, the cone surface can be given by
where R and H are the radius of the bottom circle and the height of the cone, respectively. By using the formula for definite integration by substitution and the local cylindrical coordinate system, we have
Ultimately, by substituting Eq. (50) into Eq. (58) and considering Eqs. (63) and (65), one can obtain the ten basic volume integrals over an arbitrary sphere, cylinder, or cone in regard to the global Cartesian coordinate system o–x–y-z, reading
where the word “Body” stands for “Sphere”, “Cylinder”, “Cone”, or an assembly (to be explained in detail in “Appendix 3”). xC, yC, and zC are the global coordinates of the origin of the local Cartesian coordinate system o*-u-v-w. The coefficients Ai, Bi, and Ci (i = x, y, z) are determined by Eq. (52) (For “Sphere”, γ ≡ 0). With the help of Eq. (66), we can readily generate the body force sub-matrix and the mass sub-matix associated with an arbitrary sphere, cylinder, or cone.
Appendix 3: Volume Integral of Some Representative Assemblies
In this study, an assembly is a combination of spheres, cyliners, or cones. Next, we will illustrate the calculation of the volume integral of several typical assemblies.
For an assembly (abbreviated as As-1) from one sphere and one cylinder, as shown in Fig. 61, in which RS and RC (≤ RS) are the radiuses of the sphere and the cylinder, respectively. We can set up a local Cartesian coordinate system o*–u–v–w with the center of the sphere as the origin of system (see Fig. 61). The volume integral of the assembly can be calculated by
where H is the height of the cylinder, and h is the distance between the center of the sphere and the center of the bottom circle of the cylinder.
Consider an assembly (abbreviated as As-2) from one sphere and one cone, as exhibited in Fig. 62, where RS and RC (≤ RS) are the radiuses of the sphere and the cone, respectively. Point A is the cone tip. We can make a local Cartesian coordinate system o*–u–v–w taking the center of the sphere as the origin of system (see Fig. 62). The volume integral of the assembly can be obtained from
where H is the height of the cylinder, and h is the distance between the center of sphere and the center of the bottom circle of the cone.
As for the assembly (referred to as As-3) from one sphere and one cone-tip, as presented in Fig. 63. Here, RS and RC are the radiuses of the sphere and the cone, respectively. While point A is the cone tip. Choosing the center of the bottom circle of the cone as the origin, we can construct a local Cartesian coordinate system o*–u–v–w (see Fig. 63). The volume integral of the overlap region (abbreviated as Ov-1) is given by
where H is the height of the cylinder, and h is the distance between the center of sphere and the cone tip. Further, the volume integral of the assembly can be expressed as
Now, let us see the assembly (called As-4) in Fig. 64, which is the combination of one sphere and two cylinders. The radiuses of the three bodies are equal to each other. For the two cylinder, there is a one of the bottom circles that just coincides with a certain center section of the sphere (see Fig. 64). We can make two tangent planes of the sphere, which are respectively perpendicular to the two cylinders (see the right-hand side of Fig. 64). Then, we can construct two shorter cylinders, i.e. the shorter cyliner1 (referred to as Sc-1) and the shorter cyliner2 (referred to as Sc-2). Thus, the volume integral of the assembly can be approximated by
For the assembly from two completely holing-through cylinders, whose axes intersect with each other at point A with an angle φ, as demonstrated in Fig. 65. Here, R1 and R2 are their radiuses, respetively. And R1 ≤ R2. By choosing the center of bottom circle of the cylinder with the radius R1 as the origin, we have a local Cartesian coordinate system o*–u–v–w (see Fig. 65). H is the distance between points A and o*. The volume integral of the overlap region (referred to as Ov-2) is given by
Then, according to the addition rule we can easily calculate the volume integral of the assembly.
If the cylinder 1 is just pasted on the suface of the cylinder 2, as presented in Fig. 66. We can make a shorter cylinder 1 with height H = |AB| − h = |AB| − R1ctgφ − R2sinφ (see the right-hand side of Fig. 66). Here, point A and angle φ are the intersection point and the intersection angle between their axes, respectively. Obviously, the volume integral of the assembly (referred to as As-6) is approximated to the sum of that of the shorter cylinder 1 and the cylinder 2.
Figure 67 shows an assembly (named as As-7) from one sphere and one cylinder. Here, R1 and R2 (R1 ≥ R2/2) are the radiuses of the cylinder and the sphere, respectively. For this assembly, the center O of the sphere is always on the surface of the cylinder in current study. According to the geometrical relationship, we can determine the central angle φ (see the right-hand side of Fig. 67). After this, the volume integral of the assembly can be computed approximately by
Consider the assembly (referred to as As-8) from one sphere and one cone, as indicated in Fig. 68. For this asssebmly, we always set the center O of the sphere to be on the surface of the cone. Passing through point O a smaller circular section can be obtained. Similar to the assembly As-7, after getting the central angle φ (see the right-hand side of Fig. 68), the volume integral of the assembly can be determined approximately by
Now, let su see the two simple assemblies that is respectively from one cylinder + one cone, or one clinder + one cone, as shown in Fig. 69. For these assemblies, the overlap cones can determined with easy. Based on this, we can compute the volume integrals of the assemblies by applying the addition rule of definite integral.
After getting S, Su, Sv, Sw, Suu, Svv, Sww, Suv, Svw, and Swu, one can easily compute S, Sx, Sy, Sz, Sxx, Syy, Szz, Sxy, Syz, and Szx acoording to Eq. (66).
Appendix 4: Some Sub-Functions
When two bodies get closer and closer, the contact points depend on the closest points between them. To find the relevant closest points, the algebraic or geometric algorithms will be given in this section.
4.1 Closest Point Between Point and Sphere Surface
The problem of the closest points between two sphere surfaces can be reduced to this topic.
As shown in Fig. 70, point A(xA, yA, zA) is outside of sphere O(xO, yO, zO). Point B(xB, yB, zB) is the undetermined closest point. The algebraic algorithm to find point B as follows:
where tB is the parameter associated with point B. R is the radius of the sphere.
4.2 Closest Point Between Point and Cylinder Surface
The problem of the closest points between sphere surface and cylinder surface can be boiled down to this topic.
In Fig. 71, points C and C* are the centers of the bottom and upper circles of a cylinder. Point B B(xB, yB, zB) is the undetermined closest point between point A(xA, yA, zA) and the cylinder surface.
The algebraic algorithm
- i.
Construct a function of parameters θ and s
$$f(\theta ,s) = (x_{\text{B}} - x_{\text{A}} )^{2} + (y_{\text{B}} - y_{\text{A}} )^{2} + (z_{\text{B}} - z_{\text{A}} )^{2}$$(76)where the coordinate components xB, yB, and zB all can be expressed by two parameters θ and s (see Eq. 55).
- ii.
Solve the minimum condition
$$\left\{ \begin{aligned} \frac{\partial f(\theta ,s)}{\partial \theta } = 0 \hfill \\ \frac{\partial f(\theta ,s)}{\partial s} = 0 \hfill \\ \end{aligned} \right.$$(77)Note that Eq. (77) is a system of nonlinear equations.
- iii.
Determine point B by applying parameters θ and s to the parameter equation of cylinder surface (Eq. 55).
The geometric algorithm
- i.
Find the foot of perpendicular D of point A on line CC*.
- ii.
Calculate parameter tB associated with point B on line AD.
$$t_{\text{B}} = R/\left| {\text{DA}} \right|$$(78)where R is the radius of the bottom circle of the cylinder.
- iii.
Determine point B by applying parameter tB to the parameter equation of point (Eq. 48).
4.3 Closest Point Between Point and Cone Surface
The problem of the closest points between sphere surface and cone surface can be transformed to this item.
Points C, V, radius R, and height H determine a cone surface, as presented in Fig. 72. Points A, C, and V are three known points. Point B(xB, yB, zB) is the undetermined closest point on the cone surface between point A and the cone surface.
The algebraic algorithm
- i.
Construct a function of parameters θ and s
$$f(\theta ,s) = (x_{\text{B}} - x_{\text{A}} )^{2} + (y_{\text{B}} - y_{\text{A}} )^{2} + (z_{\text{B}} - z_{\text{A}} )^{2}$$(79)where the coordinate components xB, yB, and zB all can be expressed by two parameters θ and s (see Eq. 56).
- ii.
Solve the minimum condition
$$\left\{ \begin{aligned} \frac{\partial f(\theta ,s)}{\partial \theta } = 0 \hfill \\ \frac{\partial f(\theta ,s)}{\partial s} = 0 \hfill \\ \end{aligned} \right.$$(80)where Eq. (80) is also a system of nonlinear equations.
- iii.
Determine point B by applying parameters θ and s to the parameter equation of cone surface (Eq. 56).
The geometric algorithm
It is easily be known that points A, B, C, D, E, and V are coplanar and ∆ABV is a right triangle.
- i.
Calculate angle ∠AVC through two intersection lines VA and VC.
- ii.
Estimate two distances |VE| and |AB|
$$\begin{aligned} \left| {\text{VE}} \right| = \frac{{\left| {\text{VA}} \right|\sin (\pi /2 - \angle {\text{AVC}} + \arctan (R/H))}}{\sin (\pi /2 - \arctan (R/H))} \hfill \\ \left| {\text{AB}} \right| = \left| {\text{VA}} \right|\sin (\angle {\text{AVC}} - \arctan (R/H)) \hfill \\ \end{aligned}$$(81) - iii.
Find point E by using parameter tE = |VE|/|VC| and Eq. (48).
- iv.
Determine point B applying parameter tB = |AB|/|AE| to Eq. (48).
4.4 Closest Point Between Point and Circle
This topic includes the closest point between sphere surface and circle.
In Fig. 73 line CD is perpendicular to the plane of circle C(xC, yC, zC), point A is not on line CD, which is parallel to the unit normal vector n. Point B(xB, yB, zB) is unknown closest point between the circle and point A(xA, yA, zA). The algebraic algorithm is as follows:
- i.
Construct a function of parameter θ
$$f(\theta ) = (x_{\text{B}} - x_{\text{A}} )^{2} + (y_{\text{B}} - y_{\text{A}} )^{2} + (z_{\text{B}} - z_{\text{A}} )^{2}$$(82)where the coordinate components xB, yB, and zB all can be expressed by parameter θ (see Eq. 54).
- ii.
Solve the minimum condition
$$\frac{\partial f(\theta )}{\partial \theta } = 0$$(83)Here Eq. (83) is a quartic equation of tan(θ/2), which can be solved analytically.
- iii.
Determine point B using the parameter equation of circle (Eq. 54) with parameter θ.
4.5 Closest Points Between Circle and Line
Line AE is outside of circle C(xC, yC, zC), as shown in Fig. 74. Points B(xB, yB, zB) and D(xD, yD, zD), which belong to line AE and circle C, respectively, are the two closest points that are undetermined. This means that the distance between them is the shortest. To find points B and D, The algebraic algorithm:
- i.
Construct a function of parameters θ and t
$$f(\theta ,t) = (x_{\text{B}} - x_{\text{D}} )^{2} + (y_{\text{B}} - y_{\text{D}} )^{2} + (z_{\text{B}} - z_{\text{D}} )^{2}$$(84)where the coordinate components xB, yB, and zB all can be expressed by parameter t (see Eq. 48). While the coordinate components xD, yD, and zD can be described by using parameter θ (see Eq. 54).
- ii.
Solve the minimum condition
$$\left\{ \begin{aligned} \frac{\partial f(\theta ,t)}{\partial \theta } = 0 \hfill \\ \frac{\partial f(\theta ,t)}{\partial t} = 0 \hfill \\ \end{aligned} \right.$$(85)Equation (85) is a system of nonlinear equations.
- iii.
Determine point B using parameter t and Eq. (48), determine point D by applying parameter t to Eq. (54).
It is a time-consuming process to solve the system of nonlinear equations. In order to improve compute efficiency, we propose the following geometric algorithm (see Fig. 75):
- i.
Find the foot of perpendicular B1 of point C on line AE.
- ii.
Calculate the closest point D1 between point B1 and circle C (refer to Fig. 73).
- iii.
Find the foot of perpendicular B2 of point D1 on line AE.
- iv.
Repeat step i–iii until |Bn−1Bn| ≤ 10−6.
In this study, the geometric algorithm is referred to as “projection iterative method”. Our numerical experiments show that a satisfactory solution can be obtained within ten iterative steps and that the geometric algorithm is faster than the above algebraic algorithm in most cases.
4.6 Closest Points Between Circle and Plane
In Fig. 76, points A(xA, yA, zA) and B(xB, yB, zB) are the undetermined closest points between circle C(xC, yC, zC) and place DEFG.
The algebraic algorithm
- i.
Construct a function of parameter θ, λ, and μ
$$f(\theta ,\lambda ,\mu ) = (x_{\text{B}} - x_{\text{A}} )^{2} + (y_{\text{B}} - y_{\text{A}} )^{2} + (z_{\text{B}} - z_{\text{A}} )^{2}$$(86)where the coordinate components xA, yA, and zA can be given by parameter θ (see Eq. 54). The coordinate components xB, yB, and zB are related to parameters λ and μ (see Eq. 49).
- ii.
Solve the minimum condition
$$\left\{ \begin{aligned} \frac{\partial f(\theta ,\lambda ,\mu )}{\partial \theta } = 0 \hfill \\ \frac{\partial f(\theta ,\lambda ,\mu )}{\partial \lambda } = 0 \hfill \\ \frac{\partial f(\theta ,\lambda ,\mu )}{\partial \mu } = 0 \hfill \\ \end{aligned} \right.$$(87)Here, Eq. (87) is a system of nonlinear equations.
- iii.
Determine point A by using parameter θ (Eq. 54) and point B though parameters λ and μ (see Eq. 49).
The geometric algorithm I (“Projection iterative method”, see Fig. 77)
- i.
Find the foot of perpendicular B1 of point C on plane DEFG.
- ii.
Calculate the closest point A1 between point B1 and circle C (refer to Fig. 73).
- iii.
Find the foot of perpendicular B2 of point A1 on plane DEFG.
- iv.
Repeat step i–iii until |Bn−1Bn| ≤ 10−6.
It has been shown that a satisfactory solution can be obtained within twenty iterative steps and that the geometric algorithm I is always faster than the corresponding algebraic algorithm.
The geometric algorithm II (see Fig. 78)
- i.
Construt a point C1 through point C along normal vector nP of plane DEFG.
- ii.
Construt a point C2 through point C along normal vector nC of circle C.
- iii.
Calculate two intersection points A* and A between plane CC1C2 and circle C. Here, point A is just the one of the closest points.
- iv.
Find the foot of perpendicular B of point A on plane DEFG.
Points A and B are the closest points between the circle and the plane.
4.7 Closest Points Between Two Circles
Points A(xA, yA, zA) and B(xB, yB, zB) are the undetermined closest points between circles C1 and C2, as depicted in Fig. 79.
The algebraic algorithm
- i.
Construct a function of parameters θ1 and θ2
$$f(\theta_{1} ,\theta_{2} ) = (x_{\text{B}} - x_{\text{A}} )^{2} + (y_{\text{B}} - y_{\text{A}} )^{2} + (z_{\text{B}} - z_{\text{A}} )^{2}$$(88)where the coordinate components xA, yA, and zA can be given by parameter θ1, and the coordinate components xB, yB, and zB are related to parameter θ2 (see Eqs. 3 or 4).
- ii.
Solve the minimum condition
$$\left\{ \begin{aligned} \frac{{\partial f(\theta_{1} ,\theta_{2} )}}{{\partial \theta_{1} }} = 0 \hfill \\ \frac{{\partial f(\theta_{1} ,\theta_{2} )}}{{\partial \theta_{2} }} = 0 \hfill \\ \end{aligned} \right.$$(89)Here, Eq. (90) is a system of nonlinear equations.
- iii.
Determine points A and B by using parameters θ1 and θ2, respectively (see Eqs. 3 or 4).
The geometric algorithm (“Projection iterative method”, see Fig. 80)
- i.
Find the foot of perpendicular D of point C2 on the plane of circle C1.
- ii.
Calculate the closest point B1 between point D and circle C1 (refer to Fig. 73).
- iii.
Calculate the closest point A1 between point B1 and circle C2 (refer to Fig. 73).
- iv.
Calculate the closest point B2 between point A1 and circle C1(refer to Fig. 73).
- v.
Repeat step iii–iv until |Bn−1Bn| ≤ 10−6.
It has been shown that a satisfactory solution can be obtained within about ten iterative steps and that the projection iterative method is always faster than the above algebraic algorithm.
4.8 Closest Points Between Circle and Cylinder Surface
In Fig. 81, points A(xA, yA, zA) and B(xB, yB, zB) are the two undetermined closest points between circle C1 and a cylinder surface.
The algebraic algorithm
- i.
Construct a function of parameters θ1, θ2, and s
$$f(\theta_{1} ,\theta_{2} ,s) = (x_{\text{B}} - x_{\text{A}} )^{2} + (y_{\text{B}} - y_{\text{A}} )^{2} + (z_{\text{B}} - z_{\text{A}} )^{2}$$(90)where the coordinate components xA, yA, and zA can be given by parameter θ1 (see Eq. 54). The coordinate components xB, yB, and zB are associated with parameters θ2 and s (see Eq. 55).
- ii.
Solve the minimum condition
$$\left\{ \begin{aligned} \frac{{\partial f(\theta_{1} ,\theta_{2} ,s)}}{{\partial \theta_{1} }} = 0 \hfill \\ \frac{{\partial f(\theta_{1} ,\theta_{2} ,s)}}{{\partial \theta_{2} }} = 0 \hfill \\ \frac{{\partial f(\theta_{1} ,\theta_{2} ,s)}}{\partial s} = 0 \hfill \\ \end{aligned} \right.$$(91)Equation (91) is a system of nonlinear equations.
- iii.
Determine points A and B by using parameter θ1 (see Eq. 54), and θ2 and s, respectively (see Eq. 55).
The geometric algorithm (see Fig. 82)
- i.
Find the closest point A between circle C1 and axis line C2C* of the cylinder surface (refer to Figs. 74 or 75).
- ii.
Calculate the closest point B between point A and the cylinder surface (refer to Fig. 71).
- iii.
This geometric algorithm is very efficient.
4.9 Closest Points Between Circle and Cone Surface
Points A(xA, yA, zA) and B(xB, yB, zB) are the two undetermined closest points between circle C1 and a cone surface, as shown in Fig. 83.
The algebraic algorithm
- i.
Construct a function of parameters θ1, θ2, and s
$$f(\theta_{1} ,\theta_{2} ,s) = (x_{\text{B}} - x_{\text{A}} )^{2} + (y_{\text{B}} - y_{\text{A}} )^{2} + (z_{\text{B}} - z_{\text{A}} )^{2}$$(92)where the coordinate components xA, yA, and zA can be given by parameter θ1 (see Eq. 54). The coordinate components xB, yB, and zB are associated with parameters θ2 and s (see Eq. 56).
- ii.
Solve the minimum condition
$$\left\{ \begin{aligned} \frac{{\partial f(\theta_{1} ,\theta_{2} ,s)}}{{\partial \theta_{1} }} = 0 \hfill \\ \frac{{\partial f(\theta_{1} ,\theta_{2} ,s)}}{{\partial \theta_{2} }} = 0 \hfill \\ \frac{{\partial f(\theta_{1} ,\theta_{2} ,s)}}{\partial s} = 0 \hfill \\ \end{aligned} \right.$$(93)Equation (93) is a system of nonlinear equations.
- iii.
Determine points A and B by using parameter θ1 (see Eq. 54), and parameters θ2 and s (see Eq. 55), respectively.
The geometric algorithm (“Projection iterative method”, see Fig. 84)
- i.
Find the closest point A1 between circle C1 and axis C2V of the cone surface.
- ii.
Calculate the closest point B1 between point A1 and the cone surface.
- iii.
Find the closest point A2 between point B1 and circle C1.
- iv.
Repeat step ii–iii until |Bn−1Bn| ≤ 10−6.
Numerical experiments show that a satisfactory solution can be obtained within about ten iterative steps. The projection iterative method is faster than the above algebraic algorithm.
Rights and permissions
About this article
Cite this article
Fan, H., Wang, J. Dynamic Modeling of Sphere, Cylinder, Cone, and their Assembly. Arch Computat Methods Eng 27, 725–772 (2020). https://doi.org/10.1007/s11831-019-09328-w
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11831-019-09328-w