You are on page 1of 7

Materials and Structures, 1994, 27, 33-39

The durability of structural sandwich elements


K. B E R N E R *
lngenieurburo fiir Baustatik, Darmstadt, Germany

J. M. D A V I E S
Department of Civil Engineering and Construction, University of Salford, Salford M5 4 WT, UK

A. H E L E N I U S
University of Technology, Espoo, Finland

L. H E S E L I U S
Partek Corporation, Pargas, Finland
Structural sandwich elements typically have two thin metal faces and a lightweight core. The
core may be polyurethane or polyisocyanurate foamed in situ, or it may be formed from either
rigid plastic foam or mineral wool slabstock. It is particularly in the latter case that durability
problems may arise which have not been properly addressed by the industry. This paper
considers appropriate test regimes for considering the durability of the adhesive bond between
the core and the faces. In the case of panels with a core formed of mineral wool lamellae, it
also considers possible degradation of the core material. The results of these tests demonstrate
that commercially available mineral wools do not have uniform durability. It is necessary to
pay particular attention to this factor when choosing core material for structural sandwich
panels. Although the research described in this paper was conducted with mineral wool core
material in mind, it is believed that the procedures are equally applicable to other materials.
They are being codified in European recommendations for sandwich parcels with additional
recommendations for panels with mineral wool core material, published jointly by the
European Convention for Constructional Steelwork (ECCS) and the International Council for
Building Research Studies and Documentation (C/B).

1. INTRODUCTION
With the continuing development of new methods of
construction and, in particular, the use of new materials,
problems arise in predicting the durability or the expected
life of a structure. This prediction cannot, of course, be
based upon practical experience for that would inhibit
development and result in technology that was already
old before it could be introduced on to the market. It
follows that the 0nly safe way to introduce new products
into practice is to first research the life expectancy using
accelerated ageing or degradation. The difficulty is that
there are no agreed methods available for this and,
indeed, different materials degrade in different ways and,
therefore, require different approaches. The result is that
products may be introduced on to the market without
any knowledge of their expected life.
A good example of this dilemma is to be found when
developing sandwich panels with new types of core
material. CIB Commission W56, 'Lightweight Constructions', has prepared recommendations for the design of
these elements, which typically consist of two thin metal
* All the authorsare membersof CIB CommissionW56,'Lightweight
Constructions'.
0025-5432/94{3RILEM

faces separated by a lightweight core, and is paying


particular attention to the test methods and criteria which
are necessary to define acceptable levels of durability and
structural life for this type of product. This paper
describes some of the background research which has led
to the current proposals.
2. STRUCTURAL SANDWICH ELEMENTS
During the last two decades, lightweight sandwich
elements have found increasing use in the construction
industry. Typically, these elements have two thin metal
faces and a rigid plastic foam core, and are used as the
wall and roof cladding of buildings. Today, they represent
a particularly sophisticated prefabricated building product
of high quality.
There are a number of core materials available such
as polyurethane, polyisocyanurate, expanded or extruded
polystyrene and phenolic foam. These have two potential
disadvantages, namely poor fire resistance and, with the
exception of expanded polystyrene, the use of environmentally damaging gases such as CFCs to blow the foam
during the manufacturing process. Although great efforts
are being made to reduce or eliminate the use of harmful

34

Berner, Davies, Helenius a n d Heselius

gases, the poor performance in fire persists and it is


necessary to look for alternative lightweight core
materials with good fire resistance for applications
where fire resistance is required.
This consideration has led to the development of
sandwich elements using mineral wool lamellae as the
core material, and a number of manufacturers in Europe
now produce cladding panels of this type with good
mechanical properties and proven fire resistance.
2.1 D u r a b i l i t y

Sandwich elements are extremely efficient in structural


terms by virtue of the composite action between the faces
and the core. This action makes demands on the shear
rigidity of the core and the shear bond between the core
and the face material. If either of these is degraded, the
strength of the panel will be reduced.
From the point of view of durability, the most
important property is the bond between the core and the
face material. When the face material is metal, it is
necessary to consider the possibility of internal oxidation
(corrosion) within the element. This can result from small
amounts of condensation on the colder face within the
sandwich as a consequence of variations of temperature
and humidity. It is unlikely to present a problem in
continuously foamed panels with polyurethane or polyisocyanurate cores because these contain no voids and
because of the impervious nature of the foam. However,
it is a significant factor with more pervious cores such
as mineral wool, or where slabstock is used with profiled
faces so that significant voids arise within the sandwich,
and in all cores with poor-quality material.
In order to eliminate internal corrosion of the metal
faces, it is necessary to incorporate corrosion protection
between the metal faces and the core. There are then
three critical surfaces, namely the connection between the
core and the protective layer, the corrosion protection
itself, and the connection between the protective layer
and the metal.
If an adhesive is used to bond the core to the faces, as
is the case when the core consists of mineral wool lamellae
or foamed plastic slabstock, an additional layer of
material is introduced. This adhesive layer introduces
two more critical surfaces making a total of five in all.

face
t
adhesive -~ t
face
t

small
/
load ~ e s t
specimen
~'/
J
200rnmlong x 20mm w i d e ~ / "
" ;t"

Fig. 1 Wedge test on adhesive layer.

These must all be considered with regard to their strength


and long-term durability. However, by careful choice of
materials it is possible to ensure an excellent and durable
bond between the faces and the core.
The second critical element in the consideration of
durability is the core material itself. This has to be able
to withstand climatic attack, which may involve both
temperature and humidity, without being degraded
beyond the limits implied in the design.
Panels with rigid plastic foam core material, which
generally has a high percentage of closed cells, are
relatively impervious to attack by humidity so that the
critical factor is temperature. Conversely, in the case of
panels with mineral wool core material, or other similar
core materials comprising fibrous material and an
adhesive binder, the critical factor is likely to be the
internal humidity.
3. TEST M E T H O D S
In order to quantify the durability of panels with mineral
wool core material, a study has been made which has
included consideration of a number of different test
methods. These included both methods for testing the
properties of the bond between the core and the face,
and methods for accelerating the ageing of the constituent elements. The experience gained with mineral wool
core material may well be applicable to other similar
materials.
3.1 T e s t s o f t h e a d h e s i v e b o n d

The main test method used for investigating the adhesive


bond between the faces and the core was the wedge test
according to ASTM D3762. In this test, a thin wedge is
driven into the adhesive layer (Fig. 1) and is maintained
there under a small load (approximately 3 N) for the
duration of the test. During the test, the growth of the
crack in the adhesive is observed.
The wedge test may be performed under ambient or
accelerated ageing conditions. In this project, accelerated
ageing was achieved by immersing the specimen under
load for 24 h in warm water at a temperature of 70~
Under these conditions, a suitable acceptance criterion
has been found to be that the initial crack, which generally

ss steel wedge
60ramlong /~t~---x 30mm wide / J ] ,
^
~1..""_
" ~- 2t

M a t e r i a l s and S t r u c t u r e s

35

Fig. 2 Degradation of tensile strength of mineral wool after different methods of ageing.
has a length of about 10-20 mm, should not grow by
more than 100%. In addition, it was required that the
crack should remain within the adhesive itself and should
not grow by separation of the adhesive from the face.
The disadvantage of the wedge test is that is does not
give any quantitative values of the strength properties.
However, it does appear to provide an excellent guide
to the correct choice of materials.
Numerical values for the strength of the adhesive bond
were obtained separately using the tension shear test
method according to ASTM D1002-72. This test was also
carried out both before and after ageing as described in
the next section.

3.2 Ageing of the elements


It is immediately apparent that the two factors which
primarily influence the durability of the internal elements
of a sandwich panel (core material, adhesive layer,
protective coating, etc.) are temperature and humidity.
These are likely to vary in some more or less random
fashion during the life of the panel.
A number of test methods have been used by other
investigators who have attempted to introduce some
cyclic variation of temperature and humidity into the test
regime. It was found that for mineral wool core material
and its associated adhesives, ageing cycles had little effect
and that near-maximum degradation could be obtained
by simply ageing the test sample for 24 h under conditions
of 100% humidity and a temperature of 70~
The degradation in the tensile strength of a typical
mineral wool composite after some different methods of
ageing is shown in Fig. 2. These tests were performed on

samples of size 200 mm x 150 mm x 100 mm with the


faces intact. Further comparisons are given later.
3.3 Investigation of the moisture content of panels in use
In a subsidiary investigation, the actual climatic conditions inside mineral wool panels installed on buildings
in Southern Finland during the years 1986-90 were
measured. The majority of the panels were about 5 years
old at the time of measurement and measurements were
made in south-facing walls. The moisture content was
measured by weighing the samples before and after drying
and the results obtained are summarized in Table 1.
Clearly, the elements have remained essentially dry, as the
figures in Table 1 are typical of dry mineral wool.
4. T E S T RESULTS
The test results reported here are part of an extensive
investigation carried out in Finland at the Technical
University of Tampere and in the research laboratory of
the Partek Corporation.
4.1 Results of wedge tests
Fig. 3 shows some typical results of wedge tests on
both acceptable and unacceptable material combinations. For the left-hand group of results, using pretreated steel, copper and stainless steel, the results are
acceptable because the increase of crack length during
ageing is within the limits defined in section 3 and because
the crack is cohesive (i.e. it remains within the adhesive
layer),

36

Berner, Davies, Helenius and Heselius

Table 1 Field measurements of moisure content


Measurement
No.
I
2
3
4
5
6

Locationa

j
j
j
m
m
m

Mean values

Moisture content (wt%)


Inner surface

Mid-point

Outer surface

0.23
0.16
0.09
0.30
0.17
0.09

0.20
0.16
0.11
0.20
0.16
0.10

0.21
0.17
0.10
0.24
0.19
0.11

0.17

0.15

0.17

a Locationsj are in the joints between elements; locations m are in the middle of the element.

Fig. 3 Results of wedge tests.


Unacceptable results were obtained with untreated or
inadequately pre-treated aluminium, stainless steel and
copper as shown by the right-hand group of results.
Relatively large increases in crack length were accompanied by failure of the bond between the adhesive layer and
the metal.

4.2 Strength of the adhesive bond


The effect of ageing on the shear strength of the adhesive
bond is shown in Fig. 4. Here, the adhesive used was a
two-component polyurethane adhesive and the specimens
were approximately 0.8 m x 0.9 m in size. Ageing was
achieved by applying 100 temperature cycles between
temperatures of - 30~ and + 80~ A constant moisture
content of 0.4 wt% was achieved by carrying out the tests
in a sufficiently moist atmosphere. It can be seen that
ageing caused no significant loss of shear bond strength
in the adhesive used.

4.3 Degradation of strength of different types of


mineral wool
It is apparent from the foregoing that ageing is a
significant factor which must be taken into account in
design. In order to do this, it is convenient to define the
degree of degradation arising from the ageing process in
terms of a 'degradation factor', defined as the ratio of
the tensile strength measured between the faces of the
panel after ageing to the tensile strength before ageing.
The degradation factor of available panels has been found
to be very variable covering the whole range between
good and poor durability, as shown in Fig. 5.
Some actual results for panels aged by the recommended
method, namely subjecting the panels to a climate of
100% humidity and a temperature of 70~ prior to
testing, are shown in Fig. 6. The strength before and after
degradation was measured by carrying out tensile tests
on specimens measuring approximately 120 mm x 120 mm
with a core thickness of 100 mm and with the faces intact.

Materials and Structures

37

Shear strength of the adhesive bond [MPa]


4
3,5

2,5

zinc-coated steel
with a primer layer

1,5
1

50

copper with a
primer layer

t
..............................

1O0

Number of ageing cycles


Fig. 4 Strength of adhesive bond before and after ageing.

degradation factor
1.0 ~
0.8-

good durability

0.6
0./,
0.2

poor durability

0
time

Fig. 5 Range of durability of available panels.

Fig. 6 Degradation of strength of different mineral wool panels.

The procedure for carrying out tensile tests is described


in more detail elsewhere [I]. It can be seen that the
degradation factor varied between 0.37 and 0.80 for
different types of mineral wool which are available on
the market.
As a result of the studies described above, a mineral
wool panel for commercial manufacture as a fire-resistant
wall was developed. This was tested in a conventional
cyclic ageing regime by cycling between -30~ and
+ 85~ while maintaining a moisture content of 0.4 wt%.
It was also subjected to the recommended ageing method.
The results of this study are shown in Fig. 7.
It can be seen that conventional ageing gave a
degradation factor of about 0.8, whereas the recommended
method gave the lower value of 0.6. These figures may

38

Berner, Davies, Helenius and Heselius

Relative strength
250

2O0

.x.
~ 150
C

P, loo
E

50

9 0.5

...............................................................................................................

50

100

I I

150

Number of temperature cycles: T = -30 ... +85~


Fig. 7 Strength reduction of developed panel: (o) conventional ageing, (o) accelerated ageing with 70~ and 100% RH.
Relative strength

kPa

1,5

| 60

1,4 I

1,3
1,2
1,1

.........................................................................................................

i 50
40

19

.......................................................................................................

30

0,0 i ...........i! ............... .......


0,6
0,5

50 cycles

.......

20

0}4

................................................................................................................

0,3
0,2

.........................................................................................................................

11

~" ~" ~ - ,

10

......................................................................................................................

0
0

i
500

r
T
i
1.000
1.500
2.000
Water content [g/m 2]

I
2.500

0
3.000

Fig. 8 Strength reduction of developed panel after more severe ageing.


be considered to be representative of good durability and
can be used as a guideline when choosing suitable
materials for structural sandwich panels.
In an even more severe test, the chosen panel was aged
by keeping it in steam for up to 9 days with a 50~
temperature gradient. This resulted in a vei'y high
moisture content but even so, as shown in Fig. 8, after
50 temperature cycles between - 2 0 ~ and 70~ the
measured degradation was similar to that observed in the
previous tests.

5. CONCLUSIONS
1. Although it may not be widely appreciated, unless
careful choice of materials is made, certain types of
sandwich elements, notably those with mineral wool
cores, may degrade significantly with time.
2. A suitable ageing regime in order to investigate this
important possibility is to prepare test specimens by
storing them for 24 h under conditions of 100% humidity
and a temperature of 70~

Materials and Structures


3. Wide variations in the degradation factor have been
observed in commercially available panels.
4. Panels for structural applications should be chosen
with care, paying particular attention to the adhesives
and the surface preparation used. Equally important in
panels with mineral wool cores is that material suitable
for structural use should be chosen, and that particular
attention should be paid to the quantity and quality of
the binder.

39
REFERENCES

1. 'Preliminary European Recommendations for Sandwich


Panels: Part 1: Design', European Convention for
Constructional Steelwork, Publication No. 66
(1991).
2. Tiainen, T. and Hiekkanan, I., 'Service life prediction of
PAROC building elements in Southern Finland climate',
Report No. 54/1991 (Tampere University of Technology
Institute of Materials Science, 1991).

Further information regarding the tests described in this


paper can be found elsewhere [2].

RESUME
Durabilit6 d'~16ments sandwich de structures
Les kIbments sandwich de structures prksentent de maniOre
9~nbrale deux parois mktalliques fines et une paroi centrale
l~g~re. La partie centrale peut ~tre soit en polyurbthane
ou en polyisocyanurate injectks in situ, ou bien elle peut
~tre formke soit de mousse de plastique rigide, soit de
panneaux de laine minbrale. C'est notamment dans ee
dernier casque peuvent survenir des problOmes de durabilitb
pour lesquels l'industrie n'a pas trouvk de solutions adaptbes.
Dans cet article, on examine des cycles d' essai appropriks
pour l'examen de la durabilitb de l'adhkrence partie
centrale-parois. On considOre aussi, dans le cas de
panneaux avec une pattie centrale formic de plaques de

laine mindrale, la dOgradation bventuelle du mat~riau de la


pattie centrale. Les rdsultats des essais ddmontrent que les
laines minbrales disponibles dans le commerce ne prOsentent
pas une durabilitO uniforme. II convient de prOter une
attention particuliOre d ce facteur dans le choix du matdriau
central de panneaux sandwich de structures.
Bien que la recherche dOcrite ici air OtO mende en pensant
g~ un matOriau central en laine minbrale, on estime que les
procOdures d'essai sont Ogalement applicables ~ d'autres
matdriaux. On est en train de les codifier dans les
recommandations europOennes pour les panneaux sandwich,
avec recommandations additionnelles pour les panneaux
avec matOriau central en laine minOrale, qui ont Otb publiOes
en collaboration par le ECCS (European Convention for
Constructional Steelwork) et par le CIB.

You might also like