Finite element modeling of crash test behavior for windshield laminated glass
Highlights
► Windshield FE models were set up for the simulations of glass and PVB. ► Investigate the influence of failure stress of glass on the windshield behavior. ► The behavior of the G-P-T (5 mm mesh) model agrees well with test results. ► The cracked area and acceleration level were determined by glass failure stress. ► A 50 MPa failure stress of glass best predicts the crash behavior of windshield.
Introduction
In vehicle-to-pedestrian accidents, head injuries are one of the most common injury types and can lead to lifelong disability or death [1], [2], [3]. The automotive windshield, with which pedestrians come into frequent contact, has been identified as one of the main contact sources for pedestrian head injuries. Otte [1] reported that the windshield was the most frequent vehicle sources of head injury in 543 accident cases, and Yoshiyuki [4] reported that the windshield's glass is the leading source of head injury for adult pedestrians according to the IHRA Pedestrian Safety Working Group's summary report.
Safety glass is widely used in automotive structures in order to reduce the injury severity of pedestrians in vehicle–pedestrian collisions. Nowadays, polyvinyl butyral (PVB) laminated windshields are typically used in automobiles. The laminated glass is obtained by pressing two pieces of glass plate and one piece of PVB film together at a high pressure and temperature. The function of PVB is to keep the layers of glass bonded even when broken, and its high strength prevents the glass from breaking into large sharp pieces, thereby greatly reducing the possibility of injury caused by shards of flying glass. In limited time dynamics, the elastic behavior for small deformations of the composite is determined by the glass. For large deformations, the PVB-interlayer plays a dominant role because the brittle glass cannot withstand large strains. When the glass layers fail, the PVB interlayer still has its load-carrying capacity, which can be observed experimentally. The energy-absorbing properties of the laminated glass are obtained through a combination of the energy dissipation as a consequence of the fracture behavior of the glass and the visco-elastic deformation of the PVB interlayer. Based on the experimental observation of glass under indentation loading, the cracks can be classified into four major types: cone cracks, radial cracks, median vent cracks and lateral cracks [5], [6].
In order to produce safer glass, many experimental studies related to the mechanical properties of automobile glass have been carried out. Generally speaking, experiments for studying the mechanical response of such composite materials in terms of different strain rates can be categorized into two types: quasi-static and dynamic investigations [7]. Timmel et al. [8] conducted a four-point bending test, in which the experimental setup consisted of a laminated glass plate (total thickness: 6.72 mm, 0.72 mm PVB) bearing/supported by two cylinders to evaluate three different material models in explicit FE solver (LS-DYNA) as well as fit the best material model with experiment data to describe the behavior of the PVB laminated glass. Muralidhar et al. [9] and Rahulkumar et al. [10] used a compressive shear strength experiment to study the debonding phenomenon occurring in the glass polymer interface. In the case of dynamic investigations, Charpy, drop-weight tests and the Split Hopkinson Pressure Bar (SHPB) or Kolsky's apparatus have been widely used [11], [12]. The ball drop experiments were conducted on laminated glass to investigate its temporal response after the impact of a hard sphere [13]. In order to determine the failure criterion for laminated glass in the case of impact, a wide range of experiments with glass planes were carried out by Pyttel et al. [14]. Meanwhile, Xu et al. [15] carried out a set of dynamic compression impact experiments on PVB specimens using the SHPB method at strain rates ranging from 700/s to 4500/s. Finally, headform tests were carried out to determine the contact stiffness characterization of windshields for accident reconstruction as part of the MADYMO program [16]. However, detailed studies have seldom focused on the simulation of windshield laminated glass; indeed, it remains a challenge to accurately simulate the impactor acceleration and fracture pattern of windshield laminated glass in impact tests.
In the current study, windshield models are set up based on the LS-DYNA code; the mechanical behavior of these models is simulated and compared with tests from the literature. The objective of the current study is to investigate the mechanical behavior of windshield laminated glass in a numerical simulation.
Section snippets
Development of windshield FE models
In this study, five windshield finite-element models are set up according to the LS-DYNA code. Based on the different material laws for glass and PVB, the five models are divided into three groups: single layer, double layer, and triple layer. All of windshield FE models are developed with shell elements by projecting the geometry of real-world vehicle windshield. The overall quality of mesh is controlled in the process of modeling as shown in Table 1.
The single layer glass model consisting of
Windshield FE model validation against EEVC tests
Table 3 lists the peak values of the headform's linear acceleration. The peak values include two values: first peak value and second peak value. When compared with the EEVC test results, the peak values of G-P-T (5 mm mesh) in the simulation show the best results and concur with the impact test.
Fig. 5 shows a comparison of the headform's linear acceleration history curves for G-P-T (5 mm mesh) with the headform's impact test results. The magnitude and pulse duration of the headform's linear
Discussion
In this study, a series of numerical simulations were carried out in order to determine a method for simulating the mechanical behavior of laminated glass. The headform's impact simulations were conducted at 11.1 m/s and 6 m/s using different impact positions on the windshield. Five windshield FE models were set up based on the LS-DYNA code.
According to Xu's study [22], the cracking process of glass can be divided into three phases: radial crack phase, circular crack phase, and plasticity
Conclusions
Laminated glass models were set up using different combinations of glass and PVB with various connection hypotheses and two mesh sizes based on the LS-DYNA code. The results indicated that the G-P-T (5 mm mesh) model is the most accurate for representing a windshield model. This kind of laminated glass model consists of two layers of a shell element with a tied element connection—namely, the shell element layer represents glass and the membrane element layer represents PVB.
The cracked area and
Acknowledgments
Thanks for the financial support of the China Scholarship Council (CSC) and the MAIF Foundation for their support.
References (22)
- et al.
Damage development in carbon/epoxy laminates under quasi-static and dynamic loading
Compos Sci Technol
(1999) - et al.
A finite element model for impact simulation with laminated glass
Int J Impact Eng
(2007) - et al.
Mechanical behaviour in tension of cracked glass bridged by an elastomeric ligament
Acta Mater
(2000) - et al.
Review of experimental techniques for high rate deformation and shock studies
Int J Impact Eng
(2004) - et al.
Failure criterion for laminated glass under impact loading and its application in finite element simulation
Int J Impact Eng
(2011) - et al.
Experimental study on mechanical behavior of PVB laminated glass under quasi-static and dynamic loadings
Compos Part B Eng
(2011) - Otte D. Severity and mechanism of head impacts in car-to-pedestrian accidents. In: Proceedings of the International...
- Fildes B, Gabler HC, Otte D, Linder A, Sparke L. Pedestrian impact priorities using real-world crash data and harm. In:...
- Neal-Sturgess CE, Carter E, Hardy R, Cuerden R, Guerra L, Yang JK. APROSYS European in-depth pedestrian database. In:...
- Yoshiyuki M. Summary of IHRA pedestrian safety WG activities (2005) – proposed test methods to evaluate pedestrian...
Review indentation fracture: principles and applications
J Mater Sci
Cited by (0)
- 1
Tel.: +86(0)73182655374; fax: +46(0)317723656.