2149

AN INNOVATIVE DISTRIBUTED BASE-ISOLATION SYSTEM FOR MASONRY

BUILDINGS: THE REINFORCED CUT-WALL

Mauro SASSU1 And Christian RICCI2

SUMMARY

The article reports the results of experimental tests conducted on an innovative masonry-building
seismic isolator, named "reinforced cut-wall". The system is simple to construct and inexpensive
to implement, yet delivers high static performance. It is made up of a layer of low load-bearing
capacity mortar and an underlying sheet of elastomer waterproofing interposed between the base
of the walls and the foundation head and reinforced by a series of vertical steel rods anchored to
both the wall and foundation by concrete castings. The experimental trials were conducted on pairs
of cellular blocks, 20x20x50 cm, separated by a 5 cm-thick layer of mortar and 3mm elastomer
sheathing; the rods used ranged from 8 to 12 mm in diameter. Such specimens were subjected to a
constant vertical force, simulating the actions of permanent in-service loads, and a cyclic history of
horizontal forces (max. 0.4 times the vertical load), representing an earthquake. Test results,
expressed as a series of hysteresis diagrams, reveal a high degree of energy dissipation,
attributable to the malt-elastomer joint, and efficient elastic recovery due to the reinforcing bars
which also furnish suitable vertical load-bearing capacity. Numerical analyses, conducted by
means of a rheological model calibrated with the help of the test results and applied to the case of a
simple masonry building, have confirmed the high-level performance of the proposed seismic
isolating system.

INTRODUCTION

Background

The idea of placing a vibration filter at the base of masonry buildings to absorb the energy of seismic shocks
dates back to at least the turn of the century. Actually, practices adopted on empirical grounds date back much
earlier, such as the ancient Chinese who used to place a deformable layer between the ground and house
foundations (Buckle 1990) (Zhou 1993), a technique taken up by L. Wright in 1921 in the celebrated Tokyo
Imperial Hotel (Arnold 1989). However, specific scientific proposals for isolating masonry structures from the
ground have aimed to combine mechanical efficiency with ease of application and economy of components
compatible with the "poor" building systems in which they are to be inserted. Some examples (fig. 1) from the
early 1900s include the proposal by Calantarient (Olariu 1994), which calls for interposing a layer of talcum and
sand between the foundation and masonry base and J. Bechtold's proto-structure (Zhou 1998), made up of a stiff
plate sliding over rollers, upon which the building was then laid (analogous to the cylindrical hinges discovered
at the foot of some one-storey Chinese buildings damaged by earthquakes). Such design solutions call for
seismic isolation systems arranged uniformly around a structure's base, and can therefore be classified as
"distributed systems". An alternative strategy is to position a series of support devices in suitable points along the
foundation head, whence the designation "localized" isolation systems. While the former normally offer the
advantage of a more uniform distribution of the consequent passive-type dynamic control effects throughout the
entire building, the second can generally be applied with specially developed support devices or prefabricated

1
Dept of Structural Engineering, University of Pisa (Italy)fax:++39.050.554597 e-mail: sassum@ing.unipi.it
2
Dept of Structural Engineering, University of Pisa (Italy)fax:++39.050.554597 e-mail: sassum@ing.unipi.it
components, which can in many cases be more easily retrofitted to existing buildings or their most significant
sections.

The "reinforced cut-wall" proposal

The seismic isolation technique dealt with here stems from a recent proposal (Sassu 1999) for a system of the
distributed type, called "reinforced cut-wall". It consists of (fig. 2) a layer of mortar of rather modest mechanical
properties overlaying waterproof elastomer sheathing laid between the foundation and base of the masonry walls
to be isolated. Both layers are moreover reinforced by a series of vertical metal rods anchored to the cast-
concrete foundation and building’s wall base.
The design of the mortar joint is such that it will remain integral when subjected to low-level seismic actions, but
will crack in the face of high-intensity earthquakes, thereby dissipating the mechanical energy of the tremors.
The elastomer layer instead has a two-fold function: firstly, to prevent the rise of humidity through capillarity,
and secondly, to aid the mortar in dissipating mechanical energy by slipping in the horizontal plane. Lastly, the
suitably dimensioned metal rods serve to furnish the necessary elastic restraint for the building to return to its
original configuration after cessation of the oscillations, as well as main vertical load-bearing capacity.
The design of the described seismic isolation system moreover enables construction to be carried out through the
assembly of prefabricated blocks containing the reinforcing bars, mortar layer and elastomer sheath. Therefore,
the operations to be performed on-site are limited to casting the concrete for consolidating the system's lower
portion with the foundation, and sealing the sheath along the borders of the prefabricated elements. Such a
procedure enables better control of the quality of the energy-dissipating mortar layer, whose strength must be
proportioned according to the desired elastic limit-state actions. In order to permit major relative displacements
between foundation and masonry panel, it has been also designed an alternative version of the system (large
displacement version), where the weak mortar is extended into the upper holes connected to wall’s base,
containing completely the superior part of the steel bars. At present we referee about experimental tests on the
first type of base isolator (small displacement version).

EXPERIMENTAL TESTS

A first series of experimental trials were carried out on 12 sample block pairs. The specimens (fig. 4) were
constructed from two cellular blocks, 20x20x50 cm., each bearing holes for inserting the reinforcing bars and
fitted together with an intervening 5 cm-layer of mortar and underlying 3 mm-thick elastomer sheath. Two
different types of mortar were tested: lime mortar (type A) and a 1:3 mixture of mortar and concrete (type B).
Traditional enhanced-adherence concrete reinforcing rods, whose diameter was varied from 8, 10 and 12 mm,
were arranged perpendicular to the mortar-elastomer layer and set in groups of two or four per hole and
consolidated with cast concrete. The samples were introduced into the testing apparatus, as per the scheme
presented in figure 3. Each was subjected to constant axial load by the action of two prestressed high strength
steel rods controlled by two cylindrical dynamometers arranged co-axially to the rods. This load was designed to
simulate the presence of the in-service vertical loads experienced by the walls. In addition to this, a shearing load
was produced by a 100 KN single-action oil-pressure jack governed by a hydraulic pump able to apply variable
load histories over time. This load was intended to simulate the cyclic actions produced by a seismic disturbance.
Displacement readings were effected by means of 3 inductive transducers (toll. +/- .001 mm) placed on the
vertical faces of the samples to measure the components of relative motion in the plane of the jacks between base
and summit. Moreover, each dynamometer was fitted with 4 unidirectional strain gauges.
Conventional crushing tests were first performed to assess the strength of each component (block - cast concrete
- mortar – steel bars). Then, samples 1 and 4 were subjected to preliminary measurements under conditions
designated as "empty", i.e., in the absence of the mortar isolating layer, and subjected to a maximum vertical
load of 3.0 N/mm2, obtaining proper stress-strain curves, denoting satisfactory mechanical behavior (fig. 5).
These were carried out to assess both the effect of the prestressing rods on the test itself and the load-bearing
capacity of the vertical reinforcement in response to the combined compressive and bending stress, as well as
their sensitivity to any elastic instability phenomena or geometric non-linearity due to P-δ effects. In order to
account for any deformation effects due to viscosity, each specimen was then subjected to a first series of
experimental trials consisting of 6 cyclic tests, each of which was represented by a brief loading-unloading cycle
with a constant-intensity vertical load of 0.5 N/mm2.
Subsequently, the mortar joint was repaired and a second test series performed, also consisting of 6 identical
loading-unloading cycles with a vertical pressure of 1.0 N/mm2. In both cases the maximum horizontal action

2 2149
was fixed at 0.4 times the vertical one, as exceeding this value could have reduced the resistant section of the
joint due to the load's eccentricity.

Each test yielded data in tabular form, as in table 2, containing the readings from the two extensometers and
three displacement transducers, as well as the load applied by the jack. The relative force-displacement diagrams
were thereby plotted for the direction parallel to the isolation joint, together with the corresponding hysteresis
cycles undergone by each single sample.

RESULTS

Mechanical behavior.

The trials revealed an overall mechanical behavior evolving through three distinct stages, more or less apparent
depending upon the magnitude of the vertical load and the strength of the mortar:

Stage 1 (elastic): the isolating joint manifests an essentially elastic response, with nearly complete continuity
between the foundation and base of the masonry element, so that the integrity of the cut-wall was guaranteed;
Stage 2 (elastomer slip): the elastomer sheath shows early signs of slipping between the two connection edges,
though signs of damage either to the block or the mortar joint are lacking; a corresponding fall in the shear
stiffness of the isolation layer is observed;
Stage 3 (mortar rupture): the isolation joint is affected by growing cracks; the fractures occur along the vertical
direction or somewhat inclined to it, depending upon the axial loads applied. In the latter case (inclined cracks), a
generalized increase in the joint's transverse stiffness is observed . This is due to the formation of compressed
zones within the joint itself which favor a post-elastic "reticular" response on the part of the mortar-vertical
reinforcement assembly.

The role of the vertical reinforcement rods, tested for a maximum a/g ratio of 0.4, results to be decisive for the
proper functioning of the isolating device because, in contrast to previous similar proposals, they insure a nearly
complete return of the slip plane to its initial configuration. For its part, the elastomer sheath imparts an
appreciable degree of transverse slipping to the joint in the absence of hardening. The joint mortar, on the other
hand, provides considerable energy dissipation subsequent to the formation of cracks, while at the same time
benefiting from the effective post-elastic strengthening effect afforded by the compressed zones, thereby
granting considerable load-bearing capacity to the joint even under high-stress conditions.

Rheological model

A rheological model has been developed in order to aid in the interpretation of the data obtained (fig. 6). Briefly,
the model represents the joint as a series of elastic-plastic restraints, whose characteristics vary through the three
stages observed during the tests. On the basis of the experimental results, each of the structural elements was
assigned a proper stiffness K and limit values of the slip coefficient C. In the cases where stiffness was recovered
due to the formation of compressed zones, a supplemental elastic restraint, Kp, active in the presence of assigned
transverse displacements, was added. Overall, the apparent stiffness relative to the start of each test is given by

Ksp1= Kb,eq+1/(1/ Km+1/ Kg)

in which Kb,eq is the portion of the stiffness consequent to the sample's reinforcing rods, Km is that due to the
mortar joint, while Kg is the stiffness contributed by the waterproofing sheath. Moreover, at limit state and when
compressed zones are present, the rheological model calls for an increase in stiffness defined by

Ksp3= Kb.eq + Kp

in which Kp is the experimental stiffness of the mortar compressed zone.

3 2149
The annexed table present as sample the calculations of the stiffness contributed by both the sheathing-mortar
combination (Ke=1/[1/Km + 1/Kg]), as well as the mortar one (Kp) and the steel rods. It can be seen that the
transverse stiffness afforded by this last element is supplemented by a circular ring of mortar that maintains its
connection to the rod throughout the test.
The thickness of this ring, assessed under the constraining assumption of a perfect fit sliding at the two ends of
the cast, ranges between 6.5 and 8.2 mm.

Seismic analysis of a masonry building.

Lastly, by way of example, a simulation was performed of the effects of such an isolator applied to the base of
the simple square-plan masonry building, 12 m on a side, illustrated in annexed figure. The analysis was
performed in conformity with the guidelines in Euro Code n.8, using the simplified response spectrum (report
4.1, part 1-1) with type C soil. The center wall, whose total mass is 126,38 tons, is subjected to the greatest
stress.
Here, varying degrees of plane shear were obtained depending upon the isolator type used. The values, in fact,
varied as a function of the fundamental period of the mass-isolator assembly itself. It was furthermore found that
varying the isolator's reinforcement exclusively affected the specific period exhibited by the building. On the
other hand, the in-service value of plane shear, above which slipping between the wall and foundation begins to
occur, depends essentially on the type of mortar used in the joint. The analyses conducted yielded limit
acceleration values of between 0.069 and 0.076 g for the type 1 mortar and from 0.084 to 0.096 g for type 2.
These therefore represent the ranges to adopt as the design values when the aim is to maintain the limit in-service
state for low-intensity earthquakes, while the hysteresis response should be applied only in the event of seismic
events of considerable intensity.

CONCLUSIONS

The proposed base isolating system for masonry buildings examined in the tests described in the foregoing
provides a good combination of implementation ease and mechanical properties. It moreover offers the added
benefit of waterproofing the building's groundwork, thanks to the insertion of the elastomer sheath. In particular,
from the perspective of mechanics, a good deal of design freedom results from its adoption, as the parameters of
isolator stiffness and limit strength can be adjusted through quite a wide range, depending on the choice of
reinforcement and mortar. In addition, slipping, and the consequent energy dissipation, can be allowed for in the
limited cases of earthquakes of certain preset intensities.
The results obtained also show the way for further development and research: more accurate analytical
assessments are in fact possible through the development of a suitable mechanical model for the joint, while the
possibilities of prefabricating the isolating system by adopting an on-site assembly procedure for its constituent
elements appears quite promising. Finally, from the experimental point of view, it would seem opportune to
broaden the range of horizontal stresses with respect to the vertical ones, extending the analysis to ratios greater
than the rather limited maximum adopted in the present work.

REFERENCES

Arnold, C. & Reitherman, B. 1989. Bearing masonry and earthquakes – the Imperial Hotel of Tokyo, Costruire
in Laterizio n.8. Milano: P.E.G..
Arya, A.S. 1984. Sliding concept for mitigation of earthquake disaster to masonry buildings, Proc. 8th World
Conf. Earthquake Eng., San Francisco.
Buckle, G. & Mayes, R. 1990. Seismic Isolation: History, Application and Performance – A world view,
Earthquake Spectra,Vol.6, n.2.
Kelly, J.M..1986. Aseismic base isolation: review and bibliography. Soil Dynamics and Earthquake
Engineering, Vol.5, n.4.
Li, L. 1984. Base isolation measure for aseismic buildings in China, Proc. 8th World Conf. Earthquake Eng., San
Francisco.
Olariu, L. 1994. Passive control and base isolation: state of the art lecture, Proc. 10th European Conf. Earthquake
Eng., Wien.
Sassu, M. 1999. A non conventional device for energy dissipation on masonry buildings: the reinforced cut-wall,
Proc. Int. Workshop ASSISI-99, Assisi. CICOP: Firenze.
Zhou, F.L. 1993. Most recent developments on seismic isolation of civil buildings and bridges in P.R.China,
Proc. Int. Post-S.M.I.R.T. Conf. Seminar, Capri.
Zhou, F.L., Lu, X., Wang, Q., Feng D. & Yao, Q. 1998. Dynamic analysis on structures base isolated by a ball

4 2149
system with restoring property, Earthquake Engineering and Structural Dynamics, n.27.

fig.1: Examples of distributed base-isolation

fig.2:Schemes of reinforced cut-wall:

(a)-small displacement, (b)- large displacements.

5 2149
 fig.3: Scheme of apparatus trials.

fig.4: Specimen geometry

35000
32500
30000
27500
25000
22500
20000
17500
15000
12500
10000
7500
5000
2500
0
0 0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 0,45 0,5 0,55 0,6 0,65 0,7 0,75 0,8

Vertical displacements (mm)

fig.5: Compression test of base isolator

6 2149
 Trial data: specimen 6
Mortar type: 1
Vertical load: 5 kg/cmq
Width jack: 27,5 cm

red white central Horizontal Dynamometer Dynamometer Average Total

horizontal vertical
reading Exstens. Extens. Exstens. load 02 00 displaceme load Notes
nt
n. 1/1000mm 1/1000mm 1/1000mm daN daN daN mm daN

1 32 -91 -20 0 -2505 -2480 0,0 -4985 1° cycle

2 466 96 236 520 -2441 -2478 0,29 -4919
3 974 385 596 1050 -2415 -2530 0,68 -4945
4 1561 756 1047 1500 -2376 -2593 1,15 -4969
5 1472 749 1044 1050 -2363 -2568 1,11 -4931
6 1004 494 739 570 -2363 -2466 0,77 -4829
7 281 57 214 0 -2376 -2376 0,21 -4752
8 737 294 502 515 -2363 -2441 0,54 -4804 2° cycle
9 1211 571 834 1015 -2363 -2504 0,90 -4867
10 1590 781 1082 1500 -2350 -2568 1,18 -4918
11 1486 762 1064 1020 -2337 -2543 1,13 -4880
12 1026 511 759 550 -2350 -2466 0,79 -4816
13 299 67 229 0 -2363 -2376 0,22 -4739
14 769 318 534 530 -2350 -2428 0,57 -4778 3° cycle
15 1244 596 864 1000 -2337 -2504 0,93 -4841
16 1617 806 1108 1500 -2337 -2568 1,20 -4905
17 1526 796 1097 1050 -2337 -2543 1,17 -4880
18 1029 512 764 550 -2324 -2453 0,79 -4778
19 330 83 247 0 -2350 -2364 0,25 -4714
20 773 319 539 515 -2337 -2414 0,57 -4751 4° cycle
21 1264 613 885 1020 -2337 -2491 0,95 -4828
22 1635 818 1126 1500 -2324 -2568 1,22 -4892
23 1533 801 1102 1050 -2324 -2543 1,17 -4867
24 1023 507 757 540 -2324 -2453 0,79 -4778
25 298 64 226 0 -2350 -2364 0,22 -4714
26 825 355 583 530 -2324 -2428 0,61 -4753 5° cycle
27 1318 655 931 1030 -2324 -2504 0,99 -4828
28 1667 845 1163 1500 -2310 -2580 1,25 -4891
29 1524 803 1113 1030 -2310 -2530 1,17 -4840
30 1048 524 777 570 -2324 -2441 0,81 -4765
31 360 103 273 0 -2337 -2351 0,27 -4688
32 793 343 566 510 -2324 -2402 0,59 -4726 6° cycle
33 1300 650 926 1030 -2310 -2491 0,99 -4801
34 1677 862 1173 1500 -2310 -2568 1,26 -4878
35 1474 783 1072 960 -2310 -2517 1,14 -4828
36 685 306 516 300 -2310 -2389 0,53 -4699
37 303 66 232 0 -2324 -2339 0,23 -4663

1750

1500

1250
Horizontal load (daN)

1000

750

500

250

0
0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 1,10 1,20 1,30

Horizontal displacement (mm)

table 2: Technical schedule on specimen n. 6/mortar type 1

7 2149
 Specimen n° 4: 4φ8 Specimen n° 6: 4φ12
Experimental Experimental stiffness stiffness Incremental Experimental Experimental stiffness stiffness Incremental
n. stiffness stiffness steel bars phase 4 phase 4 n. stiffness stiffness steel bars phase 4 phases 4
stiffness stiffness
cycle phase 1 phase 2 phase 3 cycle phase 1 phase 2 phase 3
Ksp1 Ksp2=Kb.eq Ke Ksp3=Kb.eq+ Kp Ksp1 Ksp2=Kb.eq Ke Ksp3=Kb.eq+ Kp
Kp Kp
1 9298 5698 3600 2784 -2914 1 17931 13590 4341 9574 -4016
2 8306 8015 291 9231 1216 2 15606 13889 1717 17321 3432

3 7606 7569 37 10435 2866 3 15588 13056 2532 17857 4801

4 7746 7059 687 12368 5309 4 16094 13289 2805 17778 4489
5 7123 7083 40 12051 4968 5 13590 13158 432 18077 4919

6 8281 6761 1520 11395 4634 6 15938 13333 2605 16786 3453

fig.6: Rheological model of base isolator.

18
isolator mortar stiffness fundamental Total shear Shear force of a/g Shear
force
type type phase 1 period force single b.i. phase 1 of single
b.i.
35 (daN/cm) (sec.) (daN) phase 1
1 2 43525 0,0697 28037 1168,2 0,086 502
2 2 33801 0,0791 28750 1197,9 0,084 503
3 2 57153 0,0609 27364 1140,2 0,088 502
4 2 92958 0,0477 26367 1098,6 0,091 500
5 2 210879 0,0317 25151 1048,0 0,096 503
35
6 2 170325 0,0353 25422 1059,3 0,088 466

Seism
direction

Fig.7: Applicative example (EC8- static method

8 2149