fib_symposium_2017, 030, v2 (final): ’Required thickness of flexurally rigid baseplate for . . .

Required Thickness of Flexurally Rigid Baseplate for Anchor Fastenings

Longfei Li 1
1
Dr. Li Anchor Profi GmbH, Gustav-Stoll-Weg 7, D-72250 Freudenstadt;
PH(+49) 7441 407 3833; FAX(+49) 7441 407 7139; email: [email protected]

Abstract
The current design methods according to prEN1992-4 for multiple anchor fastenings
with baseplates are based on the assumption that the baseplates are sufficiently rigid
such that anchor tension forces are calculated using plane distribution method, as by
analogy with beam bending where plane sections remain plane. But there is yet no
provision on the level of stiffness required to achieve plane distribution of anchor
tension forces. This can result in possible vulnerability in the practice of anchor
design, because the baseplate may not be actually rigid enough to ensure plane
distribution and there is no provision to check. The assumed rigid baseplate without
checking its required stiffness may lead to the application of smaller anchor bolts and
relatively thin (elastic) baseplates. This is unsafe since the real anchor tension forces
on elastic baseplate can be much higher than those calculated with “plane distribution
method”.
In this paper, a method for identifying the required baseplate thickness to achieve the
conditions of rigid baseplate assumption is presented and illustrated with sample
calculations. A design method with elastic baseplate is proposed and discussed.
Key words: Anchorage design, rigid, elastic baseplate

1. Introduction
The design method for multiple anchor fastenings according to prEN1992-4 /1/ is
based on the assumption that the anchor tension forces on baseplates are distributed
linearly. In the practice, e.g. using most free anchorage design software products, the
anchor tension forces are calculated by the plane distribution method, see Fig.1-1 c).

Fig. 1-1 Example of multiple anchor fastening with baseplate /4/

1
2 fib_symposium_2017, 030, v2 (final): ’Required thickness of flexurally rigid baseplate for . . .

The forces from external loads are distributed onto the baseplate using the beam
section method, see also Fig.1-1c). With the forces on the baseplate determined,
stresses in the base plate are calculated through plate analysis, for example, using
finite element analysis (FEA), assuming the baseplate is elastic. The baseplate
thickness is then determined by the condition that under design actions the maximum
stress in the baseplate does not exceed its design yield stress.
Tests conducted by university Stuttgart /2/ on models using thicknesses as obtained
from the commonly used method described in the previous paragraph have shown
much higher anchor tension forces than those assumed by the plane distribution
method. This means that the current design method in the practice may underestimate
the anchor tension forces and may be unsafe in many application cases.
In order to understand and solve this problem, there is a working party “required
stiffness of baseplate” in the fib task group T2.9. In this paper, the research results of
fib T2.9.8 conducted mainly by Dr. Li Anchor Profi GmbH on stiffness conditions for
flexurally rigid baseplate and design method for elastic baseplate are presented.

2. Elastic baseplate model

2.1 Basic assumptions
For the elastic analysis of baseplate, e.g. shown
schematically in Fig 2.1-1, the Kirchhoff plate
theory /3/ is used with the baseplate being
elastically bedded on concrete base. Bolts and
concrete areas are represented by elastic springs.
In the analytical model, it is assumed that the
following parameters are known and can be
taken for the calculation.
- Baseplate thickness tfix [mm] and E-modulus, e.g. 210,000 N/mm² for steel baseplate.
- Concrete bedding factor Cc [N/mm³] is derived from /5, 6/ with 15fc,cube, e.g. 375
N/mm³ for C20/25.
- Anchor stiffness or anchor spring constant CA [kN/mm] at SLS may be determined
by pullout tests with single anchors in uncracked concrete (fig. 2.1-2) /7/.

Fig. 2.1-2 Determination of anchor stiffness CA by pullout tests /7/

- For headed studs with shaft cross section As, bearing area AH and effective anchorage
depth hef , the anchor spring constant may be expressed by

2
 fib_symposium_2017, 030, v2 (final): ’Required thickness of flexurally rigid baseplate for . . . 3

1
CA =
hef 1

E  As CC  AH
- Profile is welded onto the baseplate so that the action forces from the profile may
be calculated by beam section method rather conservatively, see Fig.1-1 c).
2.2 Numerical solution
A finite element program has been developed as part of a commercial anchorage design
software /8/ to investigate the behavior of the elastic baseplate model. The process is
iterative that from an initial spring setup, bolts in compression and concrete areas in
tension are progressively eliminated, much like a beam on elastic foundation. Both
rectangular and triangular elements are available in the FEA model so that baseplates of
any shape can be analyzed. The influence of shear effect can also be investigated as an
additional option. The commercial anchorage design software/8/ can also perform rigid
baseplate analysis using the plane distribution method.
2.3 Verification of the elastic baseplate model by tests
In order to validate the proposed elastic baseplate model, models of 5 sets of tests
from /2, 9/ were analyzed using the computer program /8/. In all tests, the thicknesses
of baseplates are so chosen that their max. stress does not exceed the yield stress.
Solutions were performed for both rigid and elastic baseplate method. The bonded
anchor spring constants CA=EAs/l (l=tfix+hef) /10/ were used for the latter method.
Figures 2.3-1 to 2.3-3 show the dimensions of the tested multiple-anchor fastenings
/2/ and the comparisons of calculated and tested results. For observing the current
situation in the anchorage design practice using plane distribution method, the anchor
tension forces calculated by this method are also plotted in the figures. It is evident
that the elastic baseplate model provides very accurate anchor tension forces. The
plane distribution method underestimates them dramatically in these 3 tests.
Figs. 2.3-4 and 2.3-5 show the test models of /9/ (tfix=20mm) and the comparisons of
calculated and tested results. An anchor spring constants of CA= 39.8 kN/mm /9/ was
used for the elastic baseplate analysis. It is again evident that the elastic baseplate
analysis provides very accurate anchor tension forces.

Fig. 2.3-1 Comparison of calculated and tested max. anchor tension force, No 1

3
4 fib_symposium_2017, 030, v2 (final): ’Required thickness of flexurally rigid baseplate for . . .

Fig. 2.3-2 Comparison of calculated and tested max. anchor tension force, No 2

Fig. 2.3-3 Comparison of calculated and tested max. anchor tension force, No 3

Fig. 2.3-4 Comparison of calculated and tested max. anchor tension force, No 4

4
 fib_symposium_2017, 030, v2 (final): ’Required thickness of flexurally rigid baseplate for . . . 5

Fig. 2.3-5 Comparison of calculated and tested max. anchor tension force, No 5
The calculated anchor tension force by the plane distribution method agrees well with
the tested one shown in figure 2.3-4. But in figure 2.3-5, it underestimates the maximum
anchor tension forces by about 12%.

3. Stiffness condition for flexurally rigid baseplate

The results in section 2.3 show that the elastic baseplate model can accurately predict
anchor tension force distributions on baseplate. With this in mind, the following stiffness
condition for plane distribution method is proposed:
a) The baseplate is treated as elastic under design actions σEd ≤ σRd = fyk /γM and
b) Stiffness condition is met if the max. anchor tension forces, Nr, max , using plane
distribution method and Ne, max , using the elastic baseplate method are equal:
N r ,max  N e,max
This stiffness condition will be illustrated by two examples. The first example in fig. 3-1

Fig. 3-1 Determination of required thickness of rigid baseplate, Example 1

5
6 fib_symposium_2017, 030, v2 (final): ’Required thickness of flexurally rigid baseplate for . . .

shows a fastening with baseplate and one bonded anchor with a design load of 5 kN.
The max. stress for a 20mm thick plate(S235) fulfills the stress condition σEd (186) ≤
σRd (fyk /γM = 235/1.5 = 214) N/mm². The anchor tension force, Ne, max, from elastic
baseplate method is 15 kN, compared with Nr,max of 9 kN using plane distribution
method. To fulfil the stiffness condition with Ne, max = 9 kN, the plate thickness has to be
increased to 41 mm.
The second example in Fig. 3-2 is a fastening with 2 bonded anchors and a design load
of 20 kN. The maximum stress for a 23 mm thick baseplate fulfills the stress condition
σEd (179) ≤ σRd (214) N/mm². The max. anchor tension force is, Ne, max = 16.8 kN due to
prying force, whereas Nr, max = 10 kN. The thickness has to be increased to 35 mm to
fulfil the stiffness condition of Ne, max = 10 kN.

Fig. 3-2 Determination of required thickness of rigid baseplate, Example 2

4. Required additional proof of multiple-anchor resistance with elastic baseplate

The investigation in section 2 shows that the elastic baseplate model can accurately
calculate the anchor tension force distribution. A stiffness condition for plane
distribution method was then proposed. As shown in Figs. 3-1 and 3-2, this stiffness
condition may lead to unrealistic plate thicknesses /13,14/. That means, there are
application cases where the plane distribution method cannot be used. In these cases,
elastic baseplate model must be used. But the current design method /1/ lacks any
provision for the elastic baseplate model /11/. For example, the resistance of the
tested 9-anchor group /12/ (fig. 4-1) at concrete cone failure cannot be calculated by
/1/ because the reduction factor for non-uniformly distributed anchor tension loads
cannot be considered by load eccentricity on anchors.
Therefore additional proof to /1/ is necessary for anchor groups at concrete cone
failure if the elastic baseplate is to be used where anchor tension forces are not
distributed in a plane. For this additional proof, the use of Ψec,N =1,0, NEd, max ≤
NRd,c/n, as shown in fig. 4-2 may be proposed conservatively.

6
 fib_symposium_2017, 030, v2 (final): ’Required thickness of flexurally rigid baseplate for . . . 7

fc,150 = 43.4 N/mm², hef =65 mm, tfix=15mm, CA= 76.1 kN/mm

Fig. 4-1 Calculated anchor tension forces of test /12/ by /8/

Fig. 4-2 Schematic representation of the additional proof for anchor group with
non-plane distribution of anchor tension forces on elastic baseplate

Table 4-1 compares the design resistances NRd of anchor groups calculated according
to the above proposal and those with the tested failure loads NRu /12/. The calculated
minimum safety factor γMc = NRu/NRd =1.6 is higher than the minimum required of 1.5.

Table 4-1 Verification of the proposed additional proof for elastic baseplate
Baseplate stiffness Concrete cone resistance [kN] Safety factor
tfix [mm] NRd, calculated NRu, tested/12/ NRu/NRd Req. NRu/NRd
10 33.1 93 2.8
15 43.9 94 2.1 ≥ 1.5
20 56.3 92 1.6
Rigid 90.7 - -

7
8 fib_symposium_2017, 030, v2 (final): ’Required thickness of flexurally rigid baseplate for . . .

5. Conclusions
According to prEN1992-4 /1/, the anchor tension forces on baseplate are assumed to
have a plane distribution when designing multiple anchor fastenings with baseplate.
In order to fulfil this condition, the baseplate has to be sufficiently stiff. Many base
plates used in practice may not be stiff (thick) enough to meet this condition.
An elastic baseplate model has been developed, taking into account the plate relative
deflection and concrete prying forces. The model can very accurately calculate the
anchor tension force distribution with output in good agreement with test results.
A stiffness condition for the plane distribution method has been defined. Baseplate
thicknesses were calculated for a number of cases, fulfilling this condition. Results
indicate that the required thickness for the plane distribution method to be valid is
unrealistic in many cases /13,14/.
An additional proof to /1/ for concrete cone failure is necessary in order to use the
elastic baseplate model. The procedure for this additional proof has been proposed
and verified with test results. With this additional proof, elastic baseplates could be
generally used in the design of multiple anchor fastenings.

6. References
/1/ Draft FprEN 1992-4 Eurocode 2 — Design of concrete structures — Part 4 Design
of fastenings for use in concrete, 2016-03-17
/2/ Fichtner, S.: Untersuchungen zum Tragverhalten von Gruppenbefestigungen unter
Berücksichtigung der Ankerplattendicke und einer Mörtelschicht, Dissertation
Universität Stuttgart 2011
/3/ Cook, R. D. Concepts and applications of FEA, 2E, John Wiley & Sons Inc, 1981
/4/ Li, L.: Befestigungsplanung mit dem herstellerunabhängigen Programm Anchor
Profi, Seminar-Unterlagen der Ingenieurakademie West e.V. 2016
/5/ Lieberum, K.-H.: Das Tragverhalten von Beton bei extremer Teilflächenbelastung,
Dissertation TH Darmstadt, 1987
/6/ Hunziker, Armin: In Stahlbeton eingespannte Stahlprofile und Stahlbolzen,
Bautechnik 5/1984
/7/ Li, L. Relevant Anchor Stiffness for Design of Multiple Anchor Fastenings,
presentation on fib T2.9 meeting in Philadelphia, October 2016
/8/ Dr. Li Anchor Profi GmbH: Anchorage Design Software Anchor Profi 3.0.0, 2016
/9/ Mallée, R., etc.: Befestigung von Ankerplatten mit Dübeln, Beton- und
Stahlbetonbau 94, 1999
/10/ Li, L.: Commentary on the New Proposal for “Tension Loads on Anchors”,
presentation on fib T2.9 meeting in Milano, May 2016
/11/ DRAFT MINUTES of fib TG 2.9 meeting “Fastenings to structural concrete and
masonry” (formerly SAG 4) TU Dresden (Dresden, Germany) 26-27 October 2015
/12/ FMPA Stuttgart: Bericht über Belastungsversuche an einbetonierten
Kopfbolzengruppen, Bericht II.4-14 151, 16. Mai 1983
/13/Schneider, H.: Zum Einfluss der Ankerplattensteifigkeit auf die Ermittlung der
Dübelkräfte bei Mehrfachbefestigungen, August 1999
/14/ Li, L.: Erforderliche Dicke für die biegesteife Ankerplatte, Stellungnahme zu
CEN/TC 250 N 1454, Abschnitt 6.2.1 Tension loads, an DIN, NA 005-07-01-01 AK,
March 2016