Elsevier

Journal of Biomechanics

Volume 48, Issue 2, 21 January 2015, Pages 224-232
Journal of Biomechanics

Comparison of explicit finite element and mechanical simulation of the proximal femur during dynamic drop-tower testing

https://doi.org/10.1016/j.jbiomech.2014.11.042Get rights and content

Abstract

Current screening techniques based on areal bone mineral density (aBMD) measurements are unable to identify the majority of people who sustain hip fractures. Biomechanical examination of such events may help determine what predisposes a hip to be susceptible to fracture. Recently, drop-tower simulations of in-vitro sideways falls have allowed the study of the mechanical response of the proximal human femur at realistic impact speeds. This technique has created an opportunity to validate explicit finite element (FE) models against dynamic test data. This study compared the outcomes of 15 human femoral specimens fractured using a drop tower with complementary specimen-specific explicit FE analysis. Correlation coefficient and root mean square error (RMSE) were found to be moderate for whole bone stiffness comparison (R2=0.3476 and 22.85% respectively). No correlation was found between experimentally and computationally predicted peak force, however, energy absorption comparison produced moderate correlation and RMSE (R2=0.4781 and 29.14% respectively). By comparing predicted strain maps to high speed video data we demonstrated the ability of the FE models to detect vulnerable portions of the bones. Based on our observations, we conclude that there exists a need to extend the current apparent level material models for bone to cover higher strain rates than previously tested experimentally.

Introduction

It is estimated that each year over 1.6 million people suffer a hip fracture worldwide (Johnell and Kanis, 2006). Beyond its devastating effects on mobility and quality of life (Boonen et al., 2004, Ekstrom et al., 2009, Hallberg et al., 2004, Magaziner et al., 2003, Nevalainen et al., 2004, Osnes et al., 2004), this type of injury has been associated with increased risk of death (Cooper, 1997, Haleem et al., 2008). These fractures are often associated with a short, low trauma fall; with falls to the side having a high fracture rate (Parkkari et al., 1999). Current screening techniques based on areal bone mineral density (aBMD) measurements (Stone et al., 2003) are unable to identify the majority of people who sustain hip fractures. There is, therefore, a need to study the biomechanics of this problem with the ultimate goal of determining what predisposes a hip to fracture.

In sideways fall in-vitro experimental studies, forces have most often been applied using materials testing machines (Bouxsein et al., 1999, Cheng et al., 1997, de Bakker et al., 2009, Eckstein et al., 2004) with the actuator displacing the greater trochanter (GT) or femoral head in the medial–lateral direction while the shaft is supported. Constant loading rates of 0.1 m/s or lower have been generally used, while the impact speed associated with a sideways fall from standing is 3.0 m/s or higher (Feldman and Robinovitch, 2007, Robinovitch et al., 2004). This rate is beyond the reach of most current materials testing machines. Loading rate is important, however, as it has been shown to affect the mechanical properties of bone (Carter and Hayes, 1977, Courtney et al., 1994a, Linde et al., 1991). Given this rate dependency and the heterogeneity of the mechanical properties of bone, internal stress distributions could be different between dynamic and quasi-static loading, potentially resulting in different fracture patterns being created at different loading rates (Gilchrist et al., 2014). Testing femoral specimens at quasi-static loading rates in a classical sideways fall loading configuration, where the distal end of each specimen is allowed to rotate about a pivoting point (see e.g. Dragomir-Daescu et al., 2011, Grassi et al., 2012), ideally represents a statically determinant structural problem; this means that the reaction forces are independent of the bone׳s mechanical properties. When testing bones dynamically in the same configuration, the nature of the problem changes. Force equilibrium must take inertial forces into account, and these are ultimately tied to the stiffness and mass of the bone and surrounding anatomical structures, as well as the input energy.

Validated subject-specific Finite Element (FE) models based on X-ray computed tomography (CT) scans for studying the mechanics of the proximal femur have been extensively published (Bessho et al., 2007, Dragomir-Daescu et al., 2011, Duchemin et al., 2008, Grassi et al., 2012, Keyak, 2001, Keyak et al., 1998, Koivumaki et al., 2012, Schileo et al., 2008, Trabelsi et al., 2009, Trabelsi et al., 2011). These models have almost invariably followed a quasi-static structural approach and been validated against quasi-static experimental in-vitro models using local strains, overall stiffness of the bone and ultimate force as primary foci of validation. However, the recent introduction of drop-towers to simulate in-vitro the effect of a sideways fall on the mechanical response of the proximal human femur at realistic impact speeds (Fliri et al., 2013, Gilchrist et al., 2013), has opened up the possibility of comparing dynamic FE model predictions to the outcome of dynamic testing, which is the specific aim of the present study.

Section snippets

Experimental testing

The experimental setup and testing are described in detail in previous studies (Gilchrist et al., 2013, Gilchrist et al., 2014) and will only briefly be explained here for clarity. Sixteen fresh frozen proximal femora from 15 females and 1 male were obtained from a tissue donation bank (National Disease Research Interchange, Philadelphia, PA). Before testing, the specimens were imaged using two modalities: first they were scanned at 41 µm isotropic voxel size in an HR-pQCT scanner (XtremeCT,

Experimental testing

Force–time results for all 15 tests showed similar features (Fig. 4), including one or more small and short peaks – likely due to the impact of the bypass masses – followed by a steep rise in force up to the peak value.

Force–time response and force–displacement response comparison

The specimen force–time plots (Fig. 5) show similar relationships between the experimental and FE data, with the force rising at similar rates and times. However, there is a time delay in the FE model force trace compared to their corresponding experimental tests. The

Discussion

The aim of the present study was to compare FE model derived whole bone stiffness and force response to the outcome of experimentally simulated sideways falls occurring at realistic impact speeds. In general, we found statistically significant moderate correlations between the experimentally and computationally derived results for bone stiffness and energy absorption compared at equivalent displacements. No correlation was found between experimental and FEA derived force at equivalent

Conflict of interest

None.

Acknowledgement

Funding from the Natural Sciences and Engineering Research Council of Canada (NSERC RGPIN 288148-09) and the Swiss National Science Foundation (Project no. 205321_144435).

References (48)

  • S. Haleem et al.

    Mortality following hip fracture: trends and geographical variations over the last 40 years

    Inj.—Int. J. Care Inj.

    (2008)
  • B. Helgason et al.

    On the mechanical stability of porous coated press fit titanium implants: a finite element study of a pushout test

    J. Biomech.

    (2008)
  • B. Helgason et al.

    Development of a balanced experimental–computational approach to understanding the mechanics of proximal femur fractures

    Med. Eng. Phys.

    (2014)
  • N. Kelly et al.

    Experimental and numerical characterisation of the elasto-plastic properties of bovine trabecular bone and a trabecular bone analogue

    J. Mech. Behav. Biomed. Mater.

    (2012)
  • J.H. Keyak

    Improved prediction of proximal femoral fracture load using nonlinear finite element models

    Med. Eng. Phys.

    (2001)
  • J.E. Koivumaki et al.

    CT-based finite element models can be used to estimate experimentally measured failure loads in the proximal femur

    Bone

    (2012)
  • F. Linde et al.

    Mechanical properties of trabecular bone. Dependency on strain rate

    J. Biomech.

    (1991)
  • E.F. Morgan et al.

    Trabecular bone modulus–density relationships depend on anatomic site

    J. Biomech.

    (2003)
  • K.K. Nishiyama et al.

    Proximal femur bone strength estimated by a computationally fast finite element analysis in a sideways fall configuration

    J. Biomech.

    (2013)
  • S.N. Robinovitch et al.

    Effect of the "squat protective response" on impact velocity during backward falls

    J. Biomech.

    (2004)
  • E. Schileo et al.

    An accurate estimation of bone density improves the accuracy of subject-specific finite element models

    J. Biomech.

    (2008)
  • E. Schileo et al.

    Subject-specific finite element models can accurately predict strain levels in long bones

    J. Biomech.

    (2007)
  • N. Trabelsi et al.

    Validation of subject-specific automated p-FE analysis of the proximal femur

    J. Biomech.

    (2009)
  • N. Trabelsi et al.

    Patient-specific finite element analysis of the human femur—a double-blinded biomechanical validation

    J. Biomech.

    (2011)
  • Cited by (40)

    • Implementation and validation of finite element model of skull deformation and failure response during uniaxial compression

      2021, Journal of the Mechanical Behavior of Biomedical Materials
      Citation Excerpt :

      Critical strain thresholds were set for compression and tension, and when elements exceeded these thresholds their stiffness was reduced. In simulating the femur, both Bessho et al. (2007) and Ariza et al. (2015) used much larger elements with sizes on the order of millimeters, which represented a mixture of bone and marrow. Each element was assigned a nonlinear material response that included postyield softening and was dependent on the BVF of the element.

    • Bone Remodeling Process: Mechanics, Biology, and Numerical Modeling

      2021, Bone Remodeling Process: Mechanics, Biology, and Numerical Modeling
    View all citing articles on Scopus
    View full text