6th Int. Conf. Design and Analysis of Protective Structures – DAPS 2017, 29.11.-1.12.

, Melbourne, Australia

Scaling Rules for the Analysis of Blast Loaded Concrete Structures – a

Abstract

A consistent set of scaling rules for concrete structures under blast loading is reviewed. The rules,
analyzed with the Buckingham Π-Theorem, are in accordance with the cube root scaling rule [1].
Assumptions made by their adaption in experiments are discussed by focusing on possible constraints.
Furthermore, the rules are revalidated by experiments available in literature. Whereas blast waves scale
well, many experiments denote a mismatch in the structural response. A material or structural dependent
size effect is often stated to cause the mismatch, though in some cases, the performed experiment suffers
from insufficient scaling. To quantify the effect of insufficient scaling, the experiments performed by
Wang et al.[2], where blast loaded reinforced concrete slabs were tested at different scales, are modeled
numerically. A finite element model, based on the RHT model and explicit consideration of
reinforcement by beam elements, demonstrates that insufficient scaling can explain a major portion of
the mismatch in the given case. It is concluded that, under certain limitations, scaled experiments are a
powerful method to analyze blast loaded concrete structures when the related scaling rules are followed
correctly.

Keywords: Blast loading, cube root scaling, Buckingham Π-Theorem, size effect

1. Introduction

Scaled models have already been applied for years in engineering practice to analyze complex
scenarios [3]. Examples include wind tunnel, shaking table and blast loading experiments. In certain
scenarios, using scaled models makes the experimental investigation possible in the first place. To get a
sufficiently scaled model, i.e. to achieve similarity between scaled model and full scale prototype, it is
necessary that critical physical phenomena act in the same way at both scales. This can be assured by
keeping dimensionless quotients, which express the relation between related physical phenomena,
constant. For the special case that the scenario under investigation is fully described by equations, the
quotients can be obtained by bringing the equations in a dimensionless form. However, for many
applications a closed form expression is unknown. In this case, the Buckingham Π-Theorem, an
established way to perform dimensional analysis, can help determine the necessary quotients. Under the
assumption that equation F with solution f, expresses the behavior of the scenario, Bridgman 1922 [4]
states: “If the equation F(q1,q2,…,qn) = 0 is to be a complete equation, the solution has the form
f(Π1, Π2,…, Πn-k) = 0, where the Π’s are the n - k products of the parameters q1,q2,... and are
dimensionless in the fundamental units.” Generally, k equals the number of fundamental dimensions
(units) – in classical mechanics three (e.g. the SI base units for length, mass and time: meter, kilogram
and second). Other dimensions, for example forces, are then derived as quotients of the base units:
[force] = [mass length / second2]. At this point, it should be noted that not the scenario itself, but a
model thereof is investigated when performing and analyzing scaled experiments. Thus, obtained results
are only as good as underlying modeling assumptions. Unconsidered parameters q or violations of
similarity conditions (Π terms) underline the limitations of the chosen modeling approach.
In the paper at hand, a consistent set of Π terms, useful to scale blast loaded concrete structures, is
reviewed. Since loading and structural response shall be scaled by the same rules, they must both follow
the same consistent set of Π terms. The widely used cube root scaling rule, also denoted as Hopkinson
1
 Schaufelberger et al.: Scaling Rules for the Analysis of Blast Loaded Concrete Structures – a Critical Literature Review

or Hopkinson-Cranz scaling, is used as starting point for blast scaling in section 2. Π terms for scaling
RC (reinforced concrete) structures, consistent with the terms for blast scaling are then outlined in
section 3. A quantification of similarity violations (i.e. scaling rule violations) for a selected case using
a FE (Finite Element) model in section 4, and final conclusions close the paper.

2. Blast Scaling

Since blast wave phenomena scale well and the handling of small charges is much easier than that
of large ones, empirical formulas, expressing the scale independent blast behavior, have already been
developed and utilized in the early 20th century (Hopkinson 1915 and Cranz 1926).

2.1 Definition of cube root scaling rule

The cube root scaling rule is illustrated in Fig. 1 for a full scale prototype and a model scaled by
the scaling factor λ. It states that for charges with different charge mass M (Mprototype and Mmodel,
characterized by radius rprototype = r and rmodel = λr), similar blast waves are observed at constant scaled
distance Z (i.e. for distance R holds Rprototype = R and Rmodel = λR).
. 2.1
√
Similar means that blast waves have the same peak pressure p, scaled duration ̅ and scaled impulse .̅
̅ . 2.2
√

̅ . 2.3
√
Although scaled distance Z is not dimensionless, the statement of the cube root scaling rule is in
accordance with the Buckingham Π-Theorem. For a more detailed description than the subsequent one,
see [1].

Fig. 1: Cube root scaling of blast waves in the style of Baker et al. [5]. Top: Full scale blast wave caused by detonation of
a charge with radius r, observed at distance R and characterized by peak pressure p, duration t and impulse I. Bottom: model
scaled by factor λ.

2.2 Blast scaling with dimensional analysis

To perform a dimensional analysis, parameters determining the physical behavior of the scenario
under investigation have to be defined: A blast wave, caused by an air burst of a spherical charge with
radius r, is characterized by peak pressure p, impulse I and duration time t. The observed response is
determined by distance from source R, explosive mass M (commonly expressed in TNT equivalents),
density of air ρ0 and sound velocity in air c0. Hence, the scenario is determined by eight parameters;
consequently, five dimensionless Π terms have to be defined:

2
 6th Int. Conf. Design and Analysis of Protective Structures – DAPS 2017, 29.11.-1.12., Melbourne, Australia

. 2.4

.
Considering that Π terms are invariant, i.e. they must have the same value in model and prototype
scenario, 2.4 can be reformulated in terms of scaling factors λi. Scaling factors thereby express the
relation between the respective model and prototype quantity, e.g. λR = Rmodel / Rprototype
: 1 → 2.5a
Since the same materials are used on both scales, material properties remain invariant λρ,λc = 1. As
a consequence, pressure is also invariant (2.5b). From the third Π term it results that time t scales same
as radius r (2.5c)
: 1 → → 1 2.5b

: 1 → → 2.5c
The important scaling factor for charge mass is then obtained by applying the found relations to the
fourth term:
: 1 → → 2.5c

Finally, the scaling factor for the impulse is determined by the fifth Π term:
: 1 → → 2.5d

At this point, it should be noted that, although seven parameters (and therefore seven scaling
factors) were introduced, only λr can be varied independently. Therefore λ = λr defines the only
independent scaling factor. The crucial relations between the scaling factors (unequal to one) scale –
under the above stated model assumptions – air bursts and resulting blast waves. Applying the scaling
relation (e.g. Mmodel = λM Mprototype) along with the respective expression for the scaling factors 2.5 a, b,
and c in 2.1, 2.2 and 2.3 proves the cube root scaling rule.
. 2.6

̅ . 2.7

̅ . 2.8

Due to their scale independence demonstrated in 2.6 to 2.8, Z, ̅ and ̅ are useful quantities to
characterize blast waves.

2.3 Validity of cube root scaling rule

In the 1960s field tests with different charge weights were performed to analyze the cube root
scaling rule. In Fig. 2 experimental results as given in [6][7][8] for surface bursts with hemispherical
charges from 5 short tons (4536 kg) to 500 short tons (453 600 kg) are illustrated. Plotting peak pressure
and scaled impulse over scaled distance for different charge weights demonstrates the existence of a
3
 Schaufelberger et al.: Scaling Rules for the Analysis of Blast Loaded Concrete Structures – a Critical Literature Review

functional relation (proposed by Kingery in 1966 [6]). The scale independence of the function relation
validates the applicability of the cube root scaling rule for surface bursts. Similar relations were also
found for air burst [9] and, within limitations, for buried explosions [10].

Fig. 2: Surface burst with hemispherical charge, functional relations for peak pressure and scaled impulse as given by
Kingery [6] along with measurement in single tests [6][7][8].

3. Scaling Concrete Structural Response

Whereas scaling blast waves is well established and a commonly used method to experimentally
investigate detonations, the scalability of concrete response is more questionable. In many static
experiments, a pronounced mismatch between scaled models and full scale prototype, denoted as size
effect, is observed [11]. However, the application of scaled model to analyze complex RC structures
under blast loading can also be found [12][13].

3.1 Analyzing scalability of concrete structural response using dimensional analysis

In the following, the already demonstrated terms for blast scaling are expanded and their
applicability for RC structures is analyzed. Again, it shall be assumed that the same material is used for
model and prototype. Material properties are compressive strength fc, fracture energy G, failure strain εf.
Structural properties are length quantities (e.g. length, width, thickness) and reinforcement cross
section A. Furthermore, due to rate dependent material behavior, the strain rate is also considered. For
the sake of simplicity, the introduction of quantities having the same dimensions is omitted as much as
possible (e.g. concrete density, yield stress of steel, etc.) It is assumed that, if not mentioned differently,
quantities with the same dimensions also exhibit the same scaling factor.
. → 1 →

. → 1 → 1

. → 1 →
3.1
. → 1

. → 1 →

1
. → 1 →
The scaling of scenario’s geometry is determined by and . Whereas expresses that all
lengths quantities are simply scaled by the scaling factor (which is generally known as geometric scaling
and in good accordance with ), softens the scaling rule for reinforcement. It states that keeping

4
 6th Int. Conf. Design and Analysis of Protective Structures – DAPS 2017, 29.11.-1.12., Melbourne, Australia

the ratio between rebar cross section and the product of two length quantities constant, suffices. Since
the two length quantities can, for example, be the width and thickness of a concrete slab, is in
general fulfilled by keeping the reinforcement ratio constant, i.e. the ratio of rebar cross section and slab
cross section. This is of high practical importance, since in many cases scaling reinforcement diameters
is impossible. The violation of the Π terms scaling material properties ( and ) has to be made
accountable for observed mismatches in the response of RC structures in many static experiments (see
for example [11]), often denoted as size effect [14][15]. As compressive strength and failure strain
remain invariant, the fracture energy should be scaled by the scaling factor (see ). Therefore, by using
the same material, the fracture energy in the scaled model is too large. Since the impact of fracture
energy in highly dynamic scenarios, like blast loaded concrete structures, is less clear, the influence of
violating will be one aspect of discussion in the following. Furthermore, since in the scaled model
occurring strain rates are too high (see ), rate dependent failure stresses are also
overestimated in the scaled model. Both similarity violations result in a strengthening of scaled models.

3.2 Discussion of selected experiments

In literature, numerous examples dealing with the scaling of RC structures can be found. Three well
documented examples are demonstrated in the following; given conclusions on scalability are discussed.

3.2.1 Beam under close-in detonation

Zhang et al. [16] investigated the failure behavior of RC beams under close-in detonations on three
scales (1/1, 4/5, 3/5) and two different scaled distances (0.44 m/kg1/3 and 0.50 m/kg1/3). All beams were
reinforced with four bars in horizontal directions (two in upper and two in lower layer) and stirrups with
a spacing of 60 mm. In all beams, steel bars with a diameter of 6 mm were used for horizontal
reinforcement, meaning that the reinforcement ratio was not kept constant. The authors observed that,
although the failure modes are similar, the amount of damage is much higher at larger scales (Fig. 3 left)
The observed mismatches (scaling factor 3/5: 35 % for 0.44 m/kg1/3 and 40 % for 0.5 m/kg1/3, see also
Fig. 3 right) are unacceptably large. Explaining observed mismatch by size effect suffers from the fact
that similarity violations – like improper reinforcement scaling – have not been quantified.

Fig. 3: Beam under close-in detonation, Zhang et al. [16]. Left: more severe damage is observed for larger scale beams,
(scaled distance of 0.50 m/kg1/3) scabbed zones coded in gray. Right: Central deflection, normalized by beam height plotted
over scaling factor, for scaled distances of 0.44 m/kg1/3 and 0.50 m/kg1/3.

3.2.2 Slab under contact detonation

Contact detonation experiments on RC slabs at full and 1/4 scale are outlined by Forsén [17]. Aim
of the presented experimental series was to find a scale independent relation between charge weight and
crater size. Therefore in a first step several 1/4 scaled tests were performed. In a next step, the found
functional relation between crater size and charge weight was validated by two full scale tests. In the
full scale slab ribbed reinforcement bars with a diameter of 10 mm at spacing of 300 mm were used. For
the 1/4 scaled slabs, due to the unavailability of 2.5 mm ribbed bars, thread having the respective scaled

5
 Schaufelberger et al.: Scaling Rules for the Analysis of Blast Loaded Concrete Structures – a Critical Literature Review

diameter but smooth surface was used. Thus, diameter and spacing were both scaled by the scaling factor
so that the term from equation 3.1 was fulfilled. Fig. 4 demonstrates the scale independence of the
functional relation between crater size and charge weight and hence underlines the applicability of
scaling rules for analyzing contact detonations.

Fig. 4: RC slab under contact detonation, Forsén [17]. Full scale prototype and 1/4 scaled model tests demonstrate scale
independence of functional relation between crater diameter and charge weight, both normalized by slab thickness.

3.2.3 Slab under close-in detonation

Wang et al. [2] investigated the scalability of RC slabs under close-in detonations. Slabs of three
different sizes (1250 x 1250 x 50 mm3, 1000 x 1000 x 40 mm3, 750 x 750 mm x 30 mm3), resulting in
scaling factors 1, 4/5 and 3/5 were blast loaded at a scaled distance of 0.59 m/kg1/3 (tests with scaled
distance of 0.52 m/kg1/3 were also performed but are not discussed in the following). Slabs were clamped
at two sides; the explosive was placed above slab’s center. Similar to the experiments on RC beams
demonstrated in 3.2.1, in this experiment reinforcement was also scaled insufficiently. In all cases a
rebar mesh with spacing of 75 mm and 3 mm radius was used (Fig. 5).

Fig. 5: Scaling of reinforcement in experiments on RC slabs under close-in blast loading [2].

Applying the same reinforcement mesh in all cases results in constant reinforcement cross section
per length. For full scale prototype and 3/5 scaled model:
,
3.8 3.2

, 3.3
3.8

6
 6th Int. Conf. Design and Analysis of Protective Structures – DAPS 2017, 29.11.-1.12., Melbourne, Australia

However, the dimensionless reinforcement ratio (reinforcement cross section per concrete cross
section) is not invariant.
,
, 0.8 % 3.4
,

,
, 1.3 % 3.5
,

As a result, in the scaled model, the amount of reinforcement is higher than in the prototype.
Furthermore, the used scaled charge was too small (for 3/5 scaled model: 130 g instead of 138 g). Both
similarity violations result in a “strengthening” of the scaled model. Although cracks are more
pronounced at larger scales (Fig. 6 left), the observed general failure behavior is very similar. Due to the
fixed supports on two sides, the detonation causes a bending of the slab, which results in more
pronounced cracks along the center line, parallel to the support. The deviation of the central deflection
between the scales is more pronounced than differences in the crack pattern. The central deflection is
underestimated by 21 % in the 3/5 scaled model (Fig. 6 right).

Fig. 6: RC slab under close-in detonation, scaled distance of 0.59 m/kg1/3[2]. Left: back face crack pattern on the three
investigated scales (black: full scale prototype; blue: 4/5 scaled model; red: 3/5 scaled model). Right: central deflection
normalized by slab thickness plotted over the scaling factor.

Since a quantification of both similarity violations, which result in a strengthening of the scaled
model, is not available, claiming the existence of a size effect, based on the experimental outcome only,
is inappropriate. Quantification of similarity violations shall be achieved with the help of a FE model.

4. Quantifying the Impact of Similarity Violations Using Finite Element Analysis

The FE analysis of the close-in detonation experiment, presented in 3.2.3, was performed using the
commercial software LS-DYNA. Full scale and 3/5 scaled slab were discretized using ten volume
elements over the thickness. Using cubic shaped elements resulted in an element edge of 5 mm for full
scale and 3 mm for the 3/5 scaled slab. Reinforcement, assumed to be at the slab’s mid-plane, was
modeled with beam elements having 80 % of the length of the volume element edge. Beam and volume
elements were coupled with a constraint (*CONSTRAINED_BEAM_IN_SOLID [18]). The blast load
was applied with the *LOAD_BLAST_ENHANCED function [18], which calculates the pressure acting
on a surface based on charge weight and distance to charge center. As outlined in [19], the blast load
function bases on scale independent functional relations for surface ([6] and demonstrated in Fig. 2) and
air [9] bursts.
For describing concrete, the RHT (Riedel-Hiermaier-Thoma) model [20] with the standard
parameter set for 35 MPa concrete was employed. The RHT model, which in its implementation in LS-
DYNA is also widely used for modeling concrete under highly dynamic loading [21], bases on a
7
 Schaufelberger et al.: Scaling Rules for the Analysis of Blast Loaded Concrete Structures – a Critical Literature Review

nonlinear equation of state and a three surface strength model. For describing the behavior of failed
concrete, which cannot withstand tensile, and shear but compressive forces, a softening approach is
incorporated in the model. It shall be mentioned that fracture energy is not an input parameter of the
failure model [22] but a consequence of stress, chosen element size and actual failure strain. The failure
strain – itself dependent on the stress state – is a function of acting pressure and dimensionless input
parameters only. So, except for the strain rate dependence, the RHT model follows the assumption that
concrete response is scale independent. A simplified Johnson-Cook model [23] was chosen for the steel
reinforcement. Parameters as follows: density ρ = 7896 kg/m3, Young’s modulus E = 208.7 GPa,
Poisson’s ratio: ν = 0.28, yield stress A = 600 MPa, yield parameters B = 275 MPa, n = 0.36, c = 0.022.

Fig. 7: Validation of the FE model for RC slab under close-in blast loading: Comparing simulated back faces with
experimental results given in [2]. 3/5 scaled model (left) and full scale prototype (right). Simulations correspond to case 1 in
Fig. 8.

Simulating the full scale prototype and the 3/5 scaled model slab (as investigated in [2]) served to
validate the numerical model for the scenario at hand. Fig. 7 demonstrates that the same failure modes
and similar crack patterns are observable in experiment and simulation. The numerically predicted
vertical displacement slightly over predicts the experimental finding: full scale 22 mm instead of 19 mm,
scaled 19 mm instead of 15 mm (upscaled values). However, also due to uncertainties in the
experimental setup, the model is still considered good enough to investigate the impact of similarity
violations.

Fig. 8: Better following similarity results in a decrease of scaling mismatch. Top: maximum central deflection for full
scale prototype and 3/5 scaled model in experiment and simulation. Bottom: Respective mismatch: case 1: validation, setting
8
 6th Int. Conf. Design and Analysis of Protective Structures – DAPS 2017, 29.11.-1.12., Melbourne, Australia

as in [2]; case 2: correctly scaled charge weight; case 3: correctly scaled charge weight and reinforcement.

In the next steps the setup of the scaled model was subsequently “optimized” by neutralizing the
effects of insufficient scaling.
 case1: validation (Fig. 7), setup as close as possible to the experiment
 case2: correctly scaled charge weight (138 g instead of 130 g in the scaled model)
 case3: ideal scaling of reinforcement: multiplying spacing and diameter by the scaling factor.
As demonstrated in Fig. 8, the subsequent increase of similarity leads to a reduction of mismatch
from 15 % (case 1) to 7 % (case 2) to almost 0 % (case 3). The downfall of scaling mismatch to quasi
zero for a proper scaled FE model is due to the fact that underlying model assumptions, namely the RHT
and the Johnson-Cook model as well as the blast loading, follow the above outlined scaling rules. At
this point, it should also be mentioned that strain-rate dependent enhancing of failure stress was
considered in all simulations; but for the scenario at hand, the non-invariance of strain rates has a
negligible influence on final deflection.
Knowing that 15 % mismatch observed in case 1 diminishes for a proper scaled model, it can be
concluded that the fast portion of experimentally observed strengthening (21 % mismatch) is caused by
similarity violations. Whether a material or structural dependent size effect, caused by keeping the
fracture energy constant, is responsible for the remaining 6 % mismatch still remains open.

5. Conclusion

Although widely used, the validity of scaled models for analyzing RC structures in blast loading
scenarios is still questioned. Experiments denoting a pronounced size effect for comparatively simple
structures [2] [16], as well as experiments demonstrating their scalability [17] are available in literature.
Within this paper, fundamental scaling rules for RC structures under blast loading were reviewed.
Whereas scaling of blast loading with the cube root scaling rule seems undisputed, the scalability of
concrete response is less clear. It is especially critical that if one assumes concrete behavior is determined
by strength and fracture energy, similarity conditions are violated by using the same material at different
scales. One experiment, a close-in detonation with observed pronounced strengthening [2], was analyzed
in more detail. A FE model, allowing quantification of similarity violations in the setting, demonstrates
that in the specific case, a fast portion of experimentally observed strengthening is caused by insufficient
scaling of reinforcement and charge weight, and is therefore not an intrinsic size effect.
Encouraged by the outcome of the herein presented study, it can be concluded that scaled
experiments are a powerful method to analyze blast loaded concrete structures. Furthermore, it can also
be proposed that the impact of similarity violations, which in many experiments are impossible to avoid,
should be investigated with a numerical model. Thereby examining the used material models, with
respect on the scalability of their fundamental parameters, can give further insight in the quality of the
expected result.

Acknowledgements
This work was supported by Defence Science and Technology Agency (DSTA), Singapore. The financial
support of the DSTA is thankfully acknowledged. Any opinions, findings and conclusions expressed in
this paper are those of the writers and do not necessarily reflect the view of DSTA, Singapore.

References

[1] Baker, W.E., Westine, P.S. and Dodge, F.T., Similarity Methods in Engineering Dynamics; ISBN
0-87671-564-1, Hayden Book Company, 1973.
[2] Wang, W., Zhang, D, Lu, F., Wang, S.C. and Tang, F, Experimental Study on Scaling the
Explosion Resistance of a One-Way Square Reinforced Concrete Slab Under a Close in Blast
Loading, International Journal of Impact Engineering, Vol.49, pp.158-164, 2012.
[3] Coutinho, C.P., Baptista, A.J. and Rodrigues, J.D., Reduced Scale Models Based on Similitude
Theory: A Review up to 2015; Engineering Structures, Vol.119, pp. 81-94, 2016.
[4] Bridgman, P.W., Dimensional Analysis, Yale University Press, 1922.
[5] Baker, W.E., Cox, P.A., Westine, P.S., Kulesz, J.J. and Strehlow, R.A., Explosion Hazards and
Evaluation, ISBN 0-444-41756-7, Elsevier, 1983.
9
 Schaufelberger et al.: Scaling Rules for the Analysis of Blast Loaded Concrete Structures – a Critical Literature Review

[6] Kingery, C.N., Air Blast Parameters versus Distance for Hemispherical TNT Surface Burst,
Ballistic Research Laboratories, Report No.1344, Aberdeen, Maryland, 1966.
[7] Kingery, C.N., Keefer, J.H. and Day, J.D., Surface Air Blast Measurements from a 100-Ton TNT
Detonation, Ballistic Research Laboratories, Report No.1410, Aberdeen, Maryland, 1962.
[8] Reisler, R.E., Keefer, J.H. and Giglio-Tos, L., Basic Air Blast Measurements from a 500-Ton
TNT Detonation Project 1.1 Operation Snowball, Ballistic Research Laboratories Report
No.1818, Aberdeen, Maryland, 1967.
[9] Kingery, C.N. and Bulmash, G., Airblast Parameters from TNT Spherical Air Burst and
Hemispherical Surface Burst, Ballistic Research Laboratories Report No. ARBRL-TR-02555,
Aberdeen, Maryland, 1984.
[10] Chabai, A.J., Scaling Dimensions of Craters Produced by Buried Explosions, Sandia Laboratory,
Report No. SC-RR-65-70, Albuquerque, New Mexico, 1965.
[11] Bažant, Z.P., Ožbolt, J and Eligehausen, R., Fracture Size Effect: Review of Evidence for
Concrete Structures, Journal of Structural Engineering Vol.120, pp.2377-2398, 1994.
[12] Wachtell, S., Comparison of Blast Response – Scaled vs. Full-Size Concrete Structures, Annals
of the New York Academy of Science Vol.152, No.1, pp. 829-850, 1968.
[13] Ichino, H., Ohno, T. and Ishikawa, N., Internal Explosion Tests for Scaled Subsurface
Magazines; International Journal of Protective Structures Vol.7, No.1, pp.100-122, 2016.
[14] Bažant, Z.P., Size Effect in Blunt Fracture: Concrete Rock, Metal, Journal of Engineering
Mechanics Vol.110, No.4 pp.518-535, 1984.
[15] Carpinteri, A., Fractal nature of material microstructure and size effects on apparent
mechanical properties, Mechanics of Materials Vol.18, pp.89-101, 1994.
[16] Zhang, D., Yao, S., Lu, F., Chen, X., Lin, G., Wang, W. and Lin, Y., Experimental Study on the
Scaling of RC-Beams Under Close-In Blast Loading, Engineering Failure Analysis Vol.33,
pp.497-504, 2013.
[17] Forsén, R.: Experiments Used for Comparison of Blast Damage to Full Scale and One Fourth
scale Reinforced Concrete Structures; Minutes of the 24th Explosive Safety Seminar 2, St. Louis,
Missouri, 1990.
[18] Livermore Software Technology Corporation (LSTC), LS-DYNA Keyword User’s Manual
Volume I; (r: 6319), 2015.
[19] Schwer, L., Teng, H. and Souli, M., LS-DYNA Air Blast Techniques: Comparisons with
Experiments for Close-in Charges, 10th European LS-DYNA Conference, Würzburg, Germany,
2015.
[20] Riedel, W.: Beton unter Dynamischen Lasten Meso - und Makromechanische Modelle und ihre
Parameter; Fraunhofer EMI, ISBN 3-8167-6340-5, Fraunhofer IRB Verlag, 2004
[21] Grunwald, C., Schaufelberger, B., Stolz, A., Riedel, W. and Borrvall, T., A General Concrete
Model in Hydrocodes: Verification and Validation of the RHT model in LS-DYNA, International
Journal of Protective Structures, Vol.8, No.1, 2017.
[22] Homquist, T.J., Johnson, G.R. and Cook, W.H., A Computational Constitutive Model for
Concrete Subjected to Large Strains, High Strain Rates, and High Pressures, 14th International
Symposium on Ballistics, pp.591-600, Quebec, Canada, 1993.
[23] Johnson, G.R. and Cook, W.C., A Constitutive Model and Data for Metals Subjected to Large
Strains, High Strain Rates and High Temperature, Proceedings of the 7th International
Symposium on Ballistics, pp.541-547, The Hague, Netherlands, 1983.

10