2020, Volume 37, Issue 8
Wrinkling deformation is a common instability mode for flexible membrane structures. The numerical simulation on this problem is challenging. Based on the continuum and tension field theory (TFT), a complementarity co-rotational finite element method (FEM) for the wrinkling analysis of pneumatic membrane structures is proposed. By using the co-rotational approach, the finite deformation is decomposed into a rigid body motion in the global coordinate system and small strain deformation in the local coordinate system of the element. The tangent stiffness matrix of a spatial 3-node triangular membrane element is derived. It includes three parts: material stiffness, rotational stiffness and balanced projection stiffness matrices, and covers the influence of a follower load on the elemental stiffness. In the elemental local coordinate system, a wrinkling model is constructed based on the constitutive relation of bi-modulus material, which can judge the status of one element, i.e., ‘taut’, ‘wrinkled’ or ‘slack’. Furthermore, the oscillation of internal force during the iterative solution is eliminated by establishing an equivalent linear complementarity problem. The stability of the algorithm is improved. Numerical examples show that the proposed method can accurately predict the displacement, stress and wrinkling region of pneumatic membrane structures. Compared with the existing methods such as ‘quasi-dynamic’ and ‘penalty’ ways, the proposed method does not require additional solving techniques to ensure convergence. It is convenient for engineering applications.
A modified generalized differential quadrature (MGDQ) method is utilized to investigate the coupling vibration and buckling characteristics of functionally graded material (FGM) beams with even porosity distribution in thermal environment and under the action of an initial axial mechanical force. Various types of temperature distributions are considered through the thickness direction, and the material properties are temperature-dependent according to modified Voigt mixture power-law model with porosity. Using an n-th order generalized beam theory (GBT), the free vibration and buckling governing equations for this system are derived by Hamiltonian principle as a unity. The control parameters for three different boundary conditions are proposed, and the MGDQ method can be utilized to solve the coupling vibration response with MATLAB computational procedure. Based on the duality between the static and dynamic behaviors of the structure, the buckling responses are obtained by writing loop subprogram, which can greatly simplify decoupling process and improve calculation efficiency. The effects of various beam theories, boundary conditions, different types of temperature rise, initial axial force, thermal-mechanical loads, porosity, material graded index and slenderness ratios on the vibration and buckling behaviors of FGM beams are discussed, and the significant duality for the two different mechanical behaviors of the structure are also revealed by several numerical examples.
The explicit numerical algorithm for the near-field wave motion of fluid-saturated porous media in time domain is investigated based on u -p dynamic formulation. The wave motion equations are decoupled, and dynamic coupling is eliminated by the diagonalization of the mass matrix and pore fluid compression matrix. Based on the decoupled wave motion equations, the central difference method and Newmark constant average acceleration method are adopted for the solution of solid-phase displacement and velocity, respectively. The formulation of pore fluid pressure is derived based on the backward difference method. Then the explicit staggered calculating formulas for the dynamic response of fluid-saturated porous media are derived, and a new full explicit numerical algorithm for the near-field wave motion of fluid-saturated porous media in time domain is developed. The rationality of matrix diagonalization in the algorithm is validated. The numerical results gained by the proposed algorithm accord well with the corresponding analytical results. This indicates the accuracy of the proposed algorithm. Combining the time domain numerical calculation method proposed with the transmission artificial boundary method, it is applied to the near field wave motion problem of fluid-saturated porous media, and the seismic response of saturated soil site is calculated and studied. The numerical results of the seismic response of saturated soil field accord with the elastic wave motion theory. This indicates the applicability of the developed algorithm to the near-field wave motion problem of fluid-saturated porous media. The stability characteristic of the developed algorithm is investigated based on the transfer matrix of the iterative calculating formulas of the algorithm. In the developed algorithm, all the variables of dynamic response are calculated in an iterative pattern. Thusly, this algorithm has the basic characteristic of the full explicit numerical algorithm in time domain. In the developed algorithm, all the components of the dynamic response are solved by recursive and iterative modes, which avoids solving the coupled dynamic equations. This developed algorithm has high computational efficiency and is an effective algorithm for solving near-field wave motion problems in fluid-saturated porous media in time domain.
A new semi-analytical model of torque and rotational angle relationship for multi-bundled conductors is proposed. It is suitable for the calculation of large-span large-rotational angles and twisting scenarios. The model considers the vertical and horizontal displacement of the spacer and the tension variation of the sub-wire during the torsion process. It also considers the nonlinear distribution of the rotational angle along the span length. The calculation results are compared with the finite element method and experimental results which show that the geometric nonlinearity of multi-bundled conductors has a significant effect on the recovery torque. The semi-analytical model has a high calculation accuracy in both small and large rotational angle scenarios and can provide a basis for determining whether the conductors are twisted. The relationship between T-θ curves and the number of spacers is analyzed in detail to investigate the effectiveness of increasing the number of spacers in suppressing the twisting. It is found that only a small increase in the number of spacers can ensure the safety by the fact that the stiffness remains positive and the conductor can be restored by itself even if it is twisted by 180°. For the case of extremely large torque, more spacers are needed to maintain a positive torsional stiffness when it is twisted by 360°, which means that the conductor still has the ability to restore after twisting occurs.
Accidental overload, prolonged environmental erosion and the deformation caused by temperature change, etc., easily lead to strength reduction and cracking of deep flexural members, eventually affecting their serviceability. A total of 15 high-strength lightweight aggregate concrete (HSLWAC) deep flexural members are tested to systematically study their damage evolution laws under shear load. The influencing factors for the shear service ability including shear span-depth ratio (a/h0), bearing plate width (lb) and beam sectional height are considered. The relationship between diagonal crack width and load level is established based on the trend of crack propagation, and the application of diagonal crack width limiting value in current codes to such members is discussed. Test results show that the maximum diagonal crack widths of specimens exhibit an increasing trend with the increase of a/h0 and lb, and a sufficient amount of horizontal and vertical web reinforcement can effectively inhibit the diagonal cracking of such members. Both the normalized diagonal cracking strength and the maximum diagonal crack width are less affected by the size effect. During the entire shear failure process, the diagonal crack width shows a significant correlation with the applied load, and the serviceability performance index on the basis of normal weight concrete specimens is able to predict the serviceability behavior of HSLWAC specimens. However, it is difficult to satisfy the diagonal crack width limiting value for HSLWAC members designed according to the minimum stirrup ratio in Chinese code, so the minimum stirrup ratio should be appropriately increased.
To investigate the effect of the amount and position of the heated compartments on fire behaviour of continuous slabs, this paper conducted fire tests on three continuous slabs. One has fire exerted at the side span only, one has fire exerted at both the side and the mid spans, and the other has fire exerted at all the three spans. The furnace temperatures of each span, the temperature of concrete and steel, the slab deflection, the restraint forces at slab corners and failure modes were obtained. Adopting the plastic damage model on the ABAQUS platform, the numerical analyses were conducted to predict the temperature, the deflection and the mechanical mechanism of tested slabs. The comparison between the test results and numerical results were also conducted, showing that the amount and position of the heated spans have significant effects on the cracking pattern, the deflection and the failure mode of continuous slabs. The short vertical slab cracks appear on the bottom surface of the heated spans, the cracks paralleled to the short-span appear on the top surface of the heated spans and the cracks are adjacent to interior supports. The mid-span vertical deflection of the side span gradually increases with the increase of temperature, and the increase of the deflection of mid span mainly depends on the heated effect of its adjacent spans. In addition, numerical results show that the moment mechanisms of the heated span during the heating stage is different with that during cooling stage. The moments near the internal supports are larger than those at other places. The axial forces in the continuous slab are compressive, and the maximum equivalent plastic tensile strains mainly concentrate near the internal supports (top surface) and the outer sides (bottom surface).
Based on the results of an experimental study on the seismic performance of 4 full-scale concrete columns reinforced with high performance ferrocement laminates (HPFL) under constant axial forces and cyclic horizontal loads, the numerical simulation analysis of the specimens is carried out. The performance characteristics of the seismic capacity, ductility, stiffness and energy dissipation capacity of the reinforced columns are further studied. At the same time, a method for strengthening RC columns with seismic damage under loading and a simplified calculation method for flexural bearing capacity of such structures are also proposed. Based on this, using reinforcement method and finite element analysis methods proposed, the main factors affecting the seismic performance of RC columns with seismic damage under loading are studied, including axial compression ratio, shear span ratio, transverse mesh reinforcement ratio and reinforcement form. The research shows that: the finite element simulation values, theoretical values and experimental values are in a good agreement. After the HPFL reinforcement, the bearing capacity, ductility and energy consumption capacity of the components are significantly improved, and the degradation rate of stiffness is significantly reduced. After the longitudinal reinforcement of the reinforcement layer is anchored into the base, the improvement of the seismic performance is more superior. With the increase of the axial compression ratio, the bearing capacity is greatly increased, and the ductility is insufficient. As the shear-span ratio increases, the bearing capacity and ductility of the specimens decrease. With the increase of load level difference, the later deformation capacity is obviously weaker. Compared with the ring reinforcement, the spiral reinforcement has the same initial stiffness, the former has better ductility and the latter has higher bearing capacity.
Based on ABAQUS platform, a refined numerical finite element model of joints in prestressed steel reinforced concrete beam-concrete filled steel tubular composite column frame was established, and the lateral hysteretic and monotonic load-displacement curves at the column top were calculated. On the basis of the comparison of the calculated monotonic and the measured hysteresis curves, the failure process of the joints under lateral loading at the column top was studied, and the stress for concrete, steel skeleton, steel bars and prestressing tendons were carefully investigated, in which the failure mechanism was discussed. Otherwise, the influences of axial compression ratio, prestressing level, steel tube ratio and stirrup ratio in the panel zone on the lateral load-displacement curve at column top and shear-shear deformation in the panel zone were studied according to the results of parameter analysis. Finally, the formulas for calculating the shear capacity in the panel zone of the joint were proposed. The results showed that the steel-tube, stirrups and prestressed tendons reached the yield and the concrete crushed in the panel zone when the lateral load reached the peak value, which could be deemed as the sign for calculating the shear capacity. The presented formula for calculating the shear capacity could be used for a reference in engineering design.
In order to reasonably determine the fatigue strength of an orthotropic steel bridge deck, the fatigue test achievements at home and abroad are analyzed, and the effective fatigue test data is selected. Combined with fatigue cracking mechanism of typical details, fatigue strength categories (fatigue strength corresponding to 2 million cycles) for orthotropic steel bridge deck are proposed, which adapts to the anti-fatigue design and construction level of China. Research results show that: the fatigue strength of deck-to-rib detail is greatly influenced by the weld type, which belongs to Category 50 for fillet weld and Category 60 for partial penetration weld; for the rib-to-diaphragm detail, the fatigue strength of the welded toe of the diaphragm is Category 70, and that of the longitudinal rib web is Category 55; the fatigue strength of diaphragm cutout detail belongs to Category 70; the fatigue strength of rib-to-rib detail is closely relevant to the splice gap, which indicates that Category 40 is suitable for the splice gap less than 3 mm, Category 70 is suitable for the splice gap between 4 mm and 6 mm, and Category 100 is suitable for the splice gap between 8 mm and 12 mm. The fatigue strength of diaphragm weld toe in rib-to-diaphragm detail, of diaphragm cutout detail and of rib-to-rib detail is consistent with the categories in current JTG D64 specification of China. The fatigue strength of deck-to-rib detail and of rib wall weld toe in rib-to-diaphragm detail is lower than that of the categories in JTG D64 specification. When performing the anti-fatigue design or the fatigue assessment for orthotropic steel bridge decks, reasonable fatigue strength categories should be determined by the comprehensive factors of the detail structure, manufacture quality, actual service state and others.
Earthquake damage to freestanding nonstructural components has occurred in past earthquakes. Ceramic vase is one of the most vulnerable freestanding nonstructural components especially in museums and historical buildings. The vase is commonly placed on the floor without any mechanical attachment between the floor and its bottom. Under the action of friction force, there are basically four types of response mode, i.e., rest, slide, slide-rock, and rock. These kinetic behaviors are determined by friction coefficients and floor motions. Shaking table testing is conducted to understand the seismic behavior of freestanding vase, where four representative motions, consisting of two historic and two artificial ones, are selected and generated. Marble stone panel is fixed on the surface of the shaking table to simulate the floor, which is very popular in modern residential and office buildings. The kinetic friction coefficient is acquired with a slow-pull test, used to define the possible response modes. The rocking and sliding responses of the vase are observed by accelerometers and high-speed camera. The results show that: the response mode of the vase is highly dependent on the input peak acceleration, i.e., the higher the peak acceleration, the larger the rocking angles. In addition, the dynamic response varies with the excitation of different input motions, indicating that the frequency contents of the input motion also affect the rocking intensity and that the dynamic response of AC156 is larger than that of El Centro. It is also found that the ratio of the acceleration at the center of gravity to the local gravitational acceleration is equal to the kinetic friction coefficient in the test. The experimental and closed form theoretical analysis results were generally agreed with each other.
Ultra-high toughness cementitious composites (UHTCC) have excellent performance in toughness and energy absorption. The damage law of UHTCC functionally graded slabs under contact explosion is numerically investigated using LS-DYNA software to design a protective structure with excellent performance. The anti-explosion property of the structure is discussed considering the effects of the target material, explosive content, reinforcement arrangement and energy absorption layer thickness. The results show that the UHTCC functionally graded slabs exhibit excellent blast resistance under blast loading as a consequence of the reduction of craters, scabbing and target damage, and the increase in blast waves absorption. Especially, the target damage can be effectively reduced by arranging the tie ribs and rationally setting the thickness of the energy absorption layer.
It introduces the basic method and process of seismic performance assessment of RC frame structures with energy dissipation and isolated devices based on the FEMA P-58 theory. A typical multi-story RC frame structure is designed, based on which buckling restrained braces and isolation bearings are added to form a BRB-frame structure and an isolated frame structure. The finite element models of the three structures are established in OpenSees software. Appropriate seismic records are selected and scaled to analyze the structural responses of the three structures. The intensity-based assessment method in FEMA P-58 is used. The structural response and seismic loss results of the ordinary frame structure, BRB-frame structure and isolated frame structure under four seismic intensities levels (i.e., frequently occurred earthquake, design level earthquake, maximum considered earthquake, very rare level earthquake) are compared. The results of seismic performance assessment show that the use of either an isolated frame structure or an BRB-frame structure can effectively reduce the repair cost and repair time of the building under earthquakes. Compared with the ordinary frame, the repair cost and repair time of the isolated structure under rare earthquakes can be reduced by 65% and 58%, respectively, and those of the BRB-braced frame structure can be reduced by 47% and 34%, respectively.
Ten push-out experimental specimens were cast and tested to study the interfacial bond-slip behavior of section steel and high-performance fiber concrete (HPFC). The main parameters investigated were strength grade, concrete cover thickness and embedded length. The failure process and load-slip (P-S) curves were obtained. The influence regularities of design parameters on the characteristic bond strength were studied, and the formulae of characteristic bond strength were established. The effective bond stress was introduced and deduced. Through the analysis of the whole process of the effective bond stress slip curve, the development of bond stress was obtained. The results show that : the characteristics bond strength were increased by the increase of the concrete cover thickness and the concrete strength; the calculated value of the characteristics bond strength appeared were similar to that of the test values; the effective bond stress reflected the development and transformation of bond stress at the interface between section steel and high-performance-fiber-concrete; the proportional relationship among the components of the bonding force was derived by process analysis. The research provides an experimental support for analyzing the mechanical behaviors of steel reinforced high-performance-fiber-concrete.
The meso-finite element model of concrete with random three-dimensional polyhedral aggregate is firstly established by developing MATLAB and ABAQUS Python script programs. A highly efficient C++ program is then developed to insert 3D cohesive elements with zero-thickness into the aggregate-mortar interfaces and within the mortar matrix, to realistically simulate their complicated crack initiation and propagation. Parametric studies are carried out to investigate the influence of cohesive fracture properties on the load-carrying capacity and cracking process. The simulation results show that: the macroscopic stress-displacement curve is mainly affected by the tensile strength and fracture energy of mortar and interface viscous crack unit, and the position and shape of crack surface are determined by the relative ratio of tensile strength and fracture energy of mortar and interface bonded crack unit. The mechanical response of concrete reflects the characteristics of crack development, which are not only determined by fracture material parameters, but also affected by meso-structural factors such as aggregate size and shape. The meso-structural features of aggregates also have significant effects on the complicated 3D cracking process.
To develop a probabilistic limit state assessment method for concrete bridge components, a probabilistic transformation principle was introduced to analyze the rating period and rating reference period. The individual risk criterion, social risk criterion, life quality index and cost optimization principle were employed to determine the target reliability index in an operation stage. Both non-stationary and stationary probabilistic models were considered to derive the rating values of load effect and resistance. In addition, the reliability theory was adopted to conduct the calibration of partial factors and an existing bridge was used to complete a case study. The results show that the rating period and rating reference period can be taken as 6 years and 10 years, respectively, considering the actual bridge maintenance conditions. For the ductile components with safety level I, II, and III, the target reliability index of safety assessment is suggested as 3.37, 3.13 and 2.85, respectively. When the stationary-probabilistic-vehicle load effect models of a general operating state and an intensive operating state are employed, the characteristic value of safety assessment can be taken as 0.705 and 0.805 times of design vehicle load effect, respectively. When the non-stationarity of vehicle load process is considered in the safety assessment, the extreme value distribution within a rating reference period can be modeled by a generalized extreme value distribution based on the discretization of continuous stochastic process, and the corresponding 0.95 fractile can be adopted. For the ductile component of safety level I, the partial factors of dead load effect and resistance are proposed as 1.056 and 1.194, respectively, and the partial factors of vehicle load effect corresponding with general operating state and intensive operating state are proposed as 1.081 and 1.054, respectively. The achievements mentioned above can be referenced for the adjustments of the current safety assessment method of existing bridge component specified in the standard.
The seismic failure mode of uniform damage refers to that the structure has a uniform damage and story drift along the height and develops a global energy dissipation mechanism, and it is an ideal and expected failure mode. A uniform damage-based seismic optimization design approach for RC frame structures considering the soil-structure interaction is proposed. By taking the uniformity of maximum inter-story drift ratio (IDR) as the design objective, the seismic optimization design approach was developed. Moreover, the optimization procedure accounted for the IDR distribution and component rotation demands. The component’s sectional reinforcements were selected as the design variables, and the material cost and the reinforcement ratio were considered as the design constraints. Based on the Beam on Nonlinear Winkler Foundation (BNWF) model, the numerical model of the RC frame structure considering soil-structure interaction was established. Two RC frame structures with 5 and 12 stories were employed as the prototype structures. The effects of convergence parameters on the convergence speed and stability were investigated. The transfer of story reinforcements in the optimization was studied. The change of component rotation and IDR distribution after the optimization was analyzed. The analytical results indicate that the developed approach can achieve a uniform IDR distribution and a smaller maximum IDR, thus improving the seismic performance.
A two-segment multi-span curved bridge with a small radius were established and a series shaking table tests were carried out in order to reveal the mechanism of earthquake damage due to adjacent girder pounding of small radius curved bridges. The paper analyzed the effect of the change of adjacent segment period ratio on such responses as the displacement at the end of the curved bridge, the relative displacement of pier and girder and the pounding force under different ground motions. The dynamic response of adjacent girder pounding effects are analyzed under the influence of lateral eccentric effect of single column pier excited by near fault earthquake. There are four main conclusions in this paper. Firstly, the pounding effect of adjacent girders firstly increases and then decreases with the increase of the period ratio under near-fault pulse earthquakes. Secondly, the pounding response increases gradually with the decrease of the period ratio of adjacent girders under the near-fault non-pulse ground motions and far-field earthquake motions. Thirdly, the lateral eccentricity of the single-column pier at the bottom of the girder increases the change of the movement form of the main girder under the earthquake action, resulting in an increase of torsion and lateral displacements. Fourthly, such increased torsion and lateral displacements often result in non-uniform pounding effect appear, which aggravates the local damage of the main girder. The major findings of this paper can provide a scientific basis for better seismic design of curved bridges considering seismic pounding.
Based on the classical shell theory and von Karman geometric nonlinear theory, the displacement-type geometric nonlinear governing equations and simply supported boundary conditions for functionally graded shallow circular spherical shells were derived. The uniform temperature field and the external uniform pressure were considered in the derivation. The two-point boundary value problem posed by this set of governing equations and the boundary conditions was solved with the shooting method. The numerical results of axisymmetric deformation of the shells were obtained. The effects of the geometric parameters of the shell, the transverse gradient properties of the shell’s materials, the volume fraction index and elasticity modulus of the constituent materials, and uniform temperature field on the buckling equilibrium paths, upper/lower critical loads and equilibrium configurations of the shell were investigated. The numerical results show that the upper critical load of the shells decreases significantly with the increase of the volume fraction index and the decrease of the elasticity modulus of the constituent materials. The effects of the volume fraction index on the lower critical load of the shells is complicated. The rise of the uniform temperature brings obvious increase/decrease of the upper/lower critical loads of the shells. The transverse gradient properties of the shell’s materials on the effects of the buckling equilibrium paths and post buckling stable configurations of functionally graded shallow circular spherical shells with simply supported edges are very significant. Two numerical tables and some numerical curves are given for the convenience of designers.
The bending static properties and fatigue properties of rollable laminates made of glass fiber braided composites under large deformation are studied. The relationship between strain and displacement under large deformation is obtained by bending static tests. The load of fatigue tests is determined by finite element simulation of static tests and comparing the results of finite element simulation with static tests. The bending fatigue life and failure mode of different laminates, together with fatigue life curve of the same ply composite laminates under large deformation are studied. The results show that the composite laminates have obvious non-linear behavior in large deformation bending, and the bending fatigue performance of (±45°) laminates is better than that of (0°/90°) laminates. Besides, holding the ratio of minimum strain to maximum strain constant, there is a linear relationship between maximum strain and logarithmic fatigue life.
Based on the linear three-dimensional elasticity theory and "Incremental Deformation Mechanics" theory, the wave equations of acoustoelastic Lamb waves propagating along the non-principal symmetry axes direction of fiber reinforced composite lamina were derived by using Legendre orthogonal polynomial expansion method when the initial stresses were applied horizontally and vertically. The wave equations were solved numerically. In order to verify the accuracy of the method, the dispersion curves of phase velocity for isotropic material obtained by the author’s method were compared with that from "Disperse®". The stress distributions of wave structure without initial stress were computed and compared with the initial condition of stress free boundary by taking a fiber reinforced composite lamina as an example. The effects of horizontal and vertical initial stresses on the dispersion curves of Lamb waves were studied. Because the Lamb wave A0 mode was more sensitive to the acoustoelastic effect, the influence of initial stresses on the displacement distributions of wave structure was investigated.
The dynamic stall in the flight of the aircraft with a large angle of attack will lead to the self-excited torsion or pitch motion of the structure, causing nonlinear stall flutter, which will directly affect the flight safety and structural safety of the aircraft. In current study, the standard Leishman-Beddoes nonlinear unsteady aerodynamic model is modified by Mach number to make it suitable for the calculation of dynamic stall aerodynamic in the case of low speed incompressibility. Then, based on the two-element segment aeroelastic model, Newmark time-domain propulsion method is adopted for the calculation of engineering stall flutter. The stall flutter wind tunnel test is designed and completed based on the calculation results. The experimental results show that the stall flutter calculation results based on L-B model are in a good agreement with the experimental results in most experimental cases. The research results also verified that: the modified L-B model can be used for stall flutter engineering analysis and for the limit cycle oscillation evaluation of low-speed aircraft with high aspect ratio, and the stall flutter speed and limit cycle amplitude are considerably affected by the initial angle of attack at the same time.
Due to the tectonic action, the interface of layered rock may be partially folded. The investigation of the influence of irregular interface on the scattering of cavities is of great significance for the seismic safety evaluation of the surface structure. Based on a substructure method, a complex scattering problem is transformed into a radiation problem and the evaluation of wave response of a layered half-space with regular boundary conditions (free field). The Fourier transform and dual variables are used to obtain the first-order ordinary differential wave equation. Besides, soil layers are merged by using the precise integration algorithm and load boundary conditions are applied to obtain the Green's function, and the dynamic stiffness is obtained. Based on a precise integration algorithm, the coefficient matrix of an improved transfer matrix method is proposed, and the wave motion of the layered half-space is calculated. The improved transfer matrix method has no restrictions on soil layer thickness and the number of layers. The accuracy and effectiveness of the proposed method are validated by comparing them with those of the results of the previous study. The scattering effects of horseshoe holes embedded in a complex layered half-space are investigated. The results show that the effect of local folds on the magnitude of surface displacement is related to factors such as incident wave type, incident wave frequency, and local fold geometry. The peak surface displacement is affected by the combined action of horseshoe-shaped cavity and local folds, and its characteristics are not obviously related to the type of incident wave.
The dynamic response of aviation dummy under vertical impact is studied by combing experimental study and numerical simulation. Firstly, 14 g and 19 g dynamic impact tests were carried out to compare the lumbar response of aviation dummy under different impact loads. Secondly, the numerical analysis model for aviation dummy/seat restraint system was developed and validated. Then, parametric studies were conducted to investigate the effects of seatback angle and seat pitch angle on occupant injury and seat responses. Results show that the ratios of peak lumbar force, peak longitudinal friction and seat pan pressure are greater than the ratio of peak load pulse under both 14 g and 19 g impacts, so the lumbar force, longitudinal friction force and seat pan pressure all have amplification effect on the peak load pulse. The 14 g pulse has a longer duration, and it causes greater Y-axis peak moment of lumbar and Y-axis peak moment of seat pan than 19 g pulse. The peak compression load of the lumbar and the peak pressure of the seat pan both have a quadratic function relationship with the seat back angle. When the seat back angle is about 110°, the risk of lumbar injury of occupant is the largest. The peak compression load of lumbar and the peak pressure of seat pan increase with the increase of the pitch angle of seat, showing a quadratic function relationship, and the growth gradually becomes gentle.