This document summarizes a PhD student's dissertation on crashworthiness simulations of an engineered cementitious composite safety barrier. It describes the methodology, material models used, and simulations conducted including beam-drop weight impact tests, barrier-drop weight impact tests, and barrier-vehicle crash tests. The student adjusted material parameters to match experimental beam-drop test results and used these adjusted models to predict barrier-drop and crash test outcomes. The goal was to investigate barrier responses through vehicle crash simulations using a 2007 Chevrolet Silverado model. Evaluation criteria from highway safety reports were also discussed.

1. Tofail AhmedApril, 2018
Supervisor: Dr. Mi G. Chorzepa
Committee Members: Dr. Stephan A. Durham, Dr. Sung-Hee Sonny Kim
Crashworthiness Simulations for
High Levels of Impact of an
Engineered Cementitious
Composite Safety Barrier

2.  Methodology
 Selection of commercially available
material models
 Beam-drop weight impact test
simulation
 Adjustment of material parameters to
achieve beam-drop weight impact test
results
 Successful verification and validation of
beam-drop weight impact model
 Developing Type S1 barrier model
using adjusted material models
 Prediction of barrier-drop weight
impact test results
 Prediction of vehicle crash test results
 Project Goal:
Investigation of responses of
Type S1 median barrier through
vehicle crash test simulation
Barrier is made of tire chips
and steel fiber modified concrete
Overview

3.  Beam-drop weight impact test
simulation
 Barrier-drop weight impact test
simulation
 Barrier-vehicle crash test
simulation
FEA Models

4.  2007 Chevrolet Silverado
 Developed at Center for Collision and Safety
Analysis (CCSA) laboratory at George Mason
University [1,2]
 Verified and Validated by the developers
Barrier-vehicle crash test simulation
1 2007 Chevrolet Silverado Finite Element Course Validation, Presentation
https://doi.org/10.13021/G8SC8K
2 Components and Full-scale Tests of the 2007 Chevrolet Silverado Suspension System,
Report https://media.ccsa.gmu.edu/cache/NCAC-2009-R-004.pdf

6. Test level 3-11 [1,2]
1 National cooperative Highway Research Program (NCHRP) Report 350
2 Manual for Assessing Safety Hardware (MASH) 2009,2016
3 Procedures for Verification and Validation of Computer Simulations Used for Roadside Safety Applications,
AASHTO, NCHRP, TRB DOI 10.17226/17647
 Length of rigid barrier 23 m
 Impact velocity 62 mph (100 km/h)
 Impact angle 25 degrees
 Vehicle designation 2270P:
four door, two-drive, half-ton pickup truck
weighing 5000 lb (2270 kg)
 2007 Chevrolet Silverado:
four door, 5152 lb (2337 kg)
NCHRP reports [3] a case study where a FIAT
UNO and a Peugeot 206 were used for
experiments and in the corresponding FE
analysis a vehicle model, ‘Geo Metro’,
developed by National Crash Analysis Center
(NCAC) was used. These cars and the FE
model were different but complied with the
requirements for 900 kg small car category.

7. Evaluation Criteria [1,2]
1 National cooperative Highway Research Program (NCHRP) Report 350
2 Manual for Assessing Safety Hardware (MASH) 2009,2016
 Barrier should contain and redirect the
vehicle or not bring the vehicle to a
controlled stop
 Vehicle should not penetrate, underride, or
override the installation
 Control lateral deflection of the barrier is
acceptable
 Barrier should readily activate in predictable
manner by breaking away, fracturing, or
yielding
Structural Properties Vehicle Trajectory
Preferable Criteria
 Vehicle doesn’t intrude into
adjacent traffic lanes
 Exit angle is less than 60 percent of
test impact angle
It is expected that the modified concrete barrier results in larger energy absorption, larger
deformation, and reduced impact force when compared to standard concrete barrier

8. Barrier-Drop weight Impact Test Simulation

9. Barrier-Drop weight Impact Test Simulation: Results
Figure: Impact force Time history Figure: Energy Time history

10. Strain Rate Dependency of Steel
Figure: stress (lbf/in2) – strain curve for mild steel in compression [1] Figure: Results from Dynamic Tensile and Compression Tests on Mild Steel [1]
1 Jones, N. (1990), Structural Impact, Cambridge, Cambridge University

11. Strain Rate Dependency of Steel
𝜎 𝑑
𝜎𝑠
=
𝜀
𝐶
1
𝑝
𝜎 𝑑 = dynamic yield strength
𝜎𝑠 = static yield strength
𝜀 = strain rate
𝐶, 𝑝 are the Cowper-Symonds relationship parameters
C = 40.4 s-1 and p = 5, gives reasonable agreement with
the experimental data for mild steel assembled by
Symonds [1]
1 Jones, N. (1990), Structural Impact, Cambridge, Cambridge University

12. Strain Rate Dependency of Concrete
Figure: Change in Compressive Strength of Concrete with Change in
Strain Rate from Experimental Data [1]
Figure: Dynamic Increase Factor (DIF) versus Strain Rate (in a log scale) [2]
1 Bischoff, P.H., and S.H. Perry, “ Impact Behavior of Plain Concrete Loaded in Uniaxial Compression,” Journal of Engineering Mechanics, June 1995, pp. 685-
693
2 Ross, C.A., and J.W. Tedesco, “Effects of Strain-Rate on Concrete Strength,” Presented at the ACI 1991 Spring Convention, Washington, D.C., March 1992

13. Strain Rate Dependency of Concrete
For roadside safety applications, the strain rate effect increases the peak strength by 15-20%
in case of compression, and 100% in case of tension [1]
1 Users Manual for LS-DYNA Concrete Material Model 159, U.S. Department of Transportation, Federal Highway Administration, FHWA-HRT-05-062, May 2007
2 Bischoff PH, Schluter F-H, editors. Concrete structures under impact and impulsive loading, synthesis report. Bulletin d’information, No 187. Dubrovnik; Comite Euro-
Internationale du Beton; September 1988
Constitutive relationship [2]
Compression:
𝐹𝑟𝑎𝑡𝑒 =
𝑓 𝑐𝑑
𝑓𝑐
= (
𝜀
𝜀0
) 𝛼
𝜀 ≤ 30𝑠−1
𝑓 𝑐𝑑
𝑓𝑐
= 𝛾
3
𝜀 𝜀 > 30𝑠−1
𝜀0 = 30 × 10−6
𝑠−1
𝛼 =
1
5+
3
4
𝑓𝑐
𝑙𝑜𝑔𝛾 = 6𝛼 − .492
Tension:
𝐹𝑟𝑎𝑡𝑒 =
𝑓 𝑡𝑑
𝑓𝑡
= (
𝜀
𝜀0
) 𝛿
𝜀 ≤ 30𝑠−1
𝑓 𝑡𝑑
𝑓𝑡
= 𝜂
3
𝜀 𝜀 > 30𝑠−1
𝜀0 = 3 × 10−6
𝑠−1
𝛿 =
1
10+
1
2
𝑓𝑐
𝑙𝑜𝑔𝜂 = 6𝛿 − .492
𝑓𝑐𝑑,𝑓𝑡𝑑 = dynamic compressive and tensile strength respectively
𝑓𝑐,𝑓𝑡 = static compressive and tensile strength respectively
𝜀 = strain rate

14. Tri-axial Behavior

15. Cap Formulation
Figure: Yield surface with (right) and without (left) cap formulation
Elastic limit of concrete will increase with corresponding increase in confining pressure.
At some point, this strength will start to decrease even if the confinement pressure increases.
This is because of the internal porosity of the material. The function of the cap formulation in
material models is to capture this behavior. In RHT and CSCM this decrease in strength is
modeled through an ellipsoid shape

16. Material Models
 Riedel-Hiermaier-Thoma (RHT) Model [1]
 Continuous Cap Surface Model (CSCM) [2,3]
1 Riedel, W., et al. (2009). "Numerical assessment for impact strength measurements in concrete materials." International Journal of Impact Engineering 36(2): 283-293.
2 Users Manual for LS-DYNA Concrete Material Model 159, U.S. Department of Transportation, Federal Highway Administration, FHWA-HRT-05-062, May 2007
3 Evaluation of LS-DYNA Concrete Material Model 159, U.S. Department of Transportation, Federal Highway Administration, FHWA-HRT-05-063, May 2007

17. Material Models at a glance

18. User Input of Model Parameters
 Relevant experimental studies are performed
 Curve fitting from experimental test results
 Different parameter values are obtained from the curve fitting equation
 These values are used as a starting point for analytical study
 Repeated analytical studies are performed while input parameters are
being adjusted until expected results are achieved
 Applicable for both RHT and CSCM
 Only parameters of interest should be adjusted
 Selection of parameters for adjustment is crucial

19. User Input of Model Parameters: CSCM
 Provides an interpolation method for determination of all the input values
 Adjustment of input values after interpolation, if necessary, is recommended
𝑃 = 𝐴 𝑝 𝑓𝑐
′ 2
+ 𝐵𝑝 𝑓𝑐
′
+ 𝐶 𝑝
𝑃 = 𝐴 𝑝 𝐷 2
+ 𝐵𝑝 𝐷 + 𝐶 𝑝
Automatic
unconfined compression strength
maximum aggregate size
Not reliable, model developers had problem with it
Manual
P is the parameter of interest

20. User Input of Model Parameters: CSCM
Manual
𝑃 = 𝐴 𝑝 𝑓𝑐
′ 2 + 𝐵𝑝 𝑓𝑐
′ + 𝐶 𝑝
𝑃 = 𝐴 𝑝 𝐷 2 + 𝐵𝑝 𝐷 + 𝐶 𝑝

21. Beam-Drop Weight Impact Test Simulation
Compared with experimental study performed by Masud and
Chorzepa (2015)
 500 lb drop weight (226.796 kg)
 Drop height 20 ft
 End velocity 10.943 m/s (24.45 mph)
 Unconfined compressive strength 46.3 MPa (6715 psi)
 Maximum Impact Force 916 kN
 Maximum mid-span displacement 115.74 mm

22. Beam-Drop Weight Impact Test Simulation
RHT
CSCM

23.  Mesh sensitivity
Results: RHT
Internal energy time history

24. Results: RHT
Mid-span displacement time historyImpact force time history

25. Results: CSCM
Internal energy time history
 Mesh sensitivity

26. Results: CSCM
 Mesh sensitivity
Stress displacement curve in a tensile test with
unregulated softening behavior

27. Results: CSCM
Fracture energy values used in parametric study of Texas T4
Bridge Rail Simulation

28. Results: CSCM

29. Results: CSCM
Displacement time history

30. Program of Study
Term Course No. Course Name Credit hours
Spring 2017 CVLE 8350 Nonlinear Finite Element Analysis 3
ENGR 8990 Advanced Bridge Design 3
Summer 2017 ENGR 8990 Advanced Construction Materials 3
Fall 2017 CVLE 6330 Advanced Structural Analysis 3
ENGR 8950 Grad Seminar 1
ENGR 8990 Advanced Steel Design 3
Spring 2018 CVLE 8640 Advanced Strength of Materials 3
ENGR 8990 Structural Dynamics 3
Summer 2018 ENGR 8990 Composite Material Design 3

31. Questions & Discussion