| 12th Romanian International Conference on Chemistry and Chemical Engineering.- Bucharest, Romania, University "Politehnica" of Bucharest, 13-15.09.2001 |
The Role of Primers in Plastering Systems on Concrete Surfaces
H.-B. Fischer, B. Vtorov, J. Stark
F.A. Finger-Institut für Baustoffkunde; Bauhaus-Universität Weimar, BR Deutschland
Keywords: primer, plastering system (plaster), concrete surfaces (concrete)
Abstract
A drastic reduction in time frames between the manufacturing process of concrete units and the rendering phase (including prior priming) does not allow the concrete to dry well. This fact is also underlined by changes in concrete technology (denser concrete and denser concrete surfaces).
The tests showed that the reduction of drying time (storage time) had a significant influence on the bonding properties of gypsum plaster on concrete surfaces. In such cases it is absolutely necessary to use an appropriate primer no matter what the processing temperature (2 C to 20 C) might be.
The influence of different types of concrete surface coarseness (in the micro-area), and of varying primer quality, is significantly lower.
Introduction
Indoor plasters have the primary task of covering substrates (walls and ceilings) for thermal, hygienic and aesthetic purposes.
Gypsum plasters are remarkable for a variety of distinct advantages. Their most significant advantage is that they are able to create a pleasant room atmosphere. Their physical properties allow for the regulation of room climate, and their low thermal conductivity leads to a sense of cosiness. They are quick and easy to process; they dry easily. Apart from these factors, gypsum plasters also contribute to the improvement of fire protection qualities.
Gypsum plasters can be applied on almost all kinds of substrates (background surfaces):
Since the 1960s, their application has generally been based on dry mortars, and is carried out with the help of machines.
The adhesive mechanism of plasters on surfaces can be described as follows (cf. KHODA /23/ and OHNEMÜLLER /11/):
Adhesion of fresh plaster based on the negative pressure effect
- during roughcasting of plaster, air is dispelled and a zone of low pressure that enhances adhesion is created
Adhesion of fresh and hardened plaster based on interlocking
- surface coarseness of the substrate leads to interlocking with the plaster
- the crystals emerging from the hydration process in the interface layer also continue to partly develop in the pores of the substrate, thus resulting in additional mechanical interlocking
Adhesion of hardened plaster based on chemical bonding
- crystal in-growth between plaster and substrate requires similar or chemically related plastering surfaces; this, however, is a seldom occurrence .
Despite diverse and long-standing know-how /1-23/, fractures between gypsum plasters and concrete surfaces sometimes occurred. Possible reasons cited were as follows : (/03, 04, 05, 07, 08, 09, 10, 12, 13, 16, 20, 21, 23 /):
- Changes of structural gypsum crystal ( i.e. change in crystal habits)
A study of pertinent scientific literature shows that highly differentiated individual factors influence gypsum bonding on concrete surfaces. Contradictory data cannot be explained otherwise.
For over 20 years now, primers have been used in order to avoid problems, and to improve bonding qualities. The bonding process of gypsum plasters on concrete surfaces has improved, and the rate of damage has thus decreased /21-23/.
New innovations in pre-fabrication technology, especially in concrete compacting, have led to increasingly smoother and denser surfaces in the last few years. Fast construction methods have led to a radical shrinking of the construction process until the plastering phase. This often results in only a partial drying of the concrete. Satisfactory bonding of the gypsum on the concrete surface, therefore, is much more difficult to achieve.
Particularly difficult plastering surfaces are pre-fabricated concrete units with light aggregates. These have now been manufactured for around 20 years. Inflated light aggregates cause retarded drying and very high moisture levels of approximately 15 %.
Recently there have been some cases in which gypsum plasters did not adhere well on smooth concrete surfaces in ceiling areas. Similar problems have occurred with light concrete units.
The scope of this study has been to examine some of the factors described in the cited literature, and to clarify their influences on bonding characteristics in present conditions. The following points were examined within the framework of this study:
The storage period of the concrete specimens, and thus the moisture conditions of the substrate, were varied for this purpose. Apart from that, the influence of additional moisturisation (on the back) of the concrete specimens and surface coarseness were also determined.
Materials and Methods
Preparation of specimens
Factory-made concrete plates (from WITTMER + KLEE, Waghäusel) with dimensions of 1000 mm x 700 mm were used for the study of the influence of various primers. These specimens were freed of surface residues and stored horizontally with the smooth surface facing upwards. Plaster application was carried out immediately on 6 plates (5 days after manufacturing), and 35 days later on the rest of the 6 plates (intended partial carbonation, superficial drying).
The client put eleven easily available primers (synthetic emulsions based on styrene acrylate and vinyl acetate polymer) at our disposal. These were prepared according to the in-structions provided (partly diluted with water, and homogenised through stirring). They were then applied to the smooth side of the concrete plates (area of application 500 mm x 500 mm, respectively) with rollers. A comparative surface (reference test) was left untreated.
One day after the primers had been applied, gypsum mortar made from specially supplied manual plaster of low quality with a water/plaster ratio of approximately 0.52 was mixed (2 measures of ready-made gypsum plaster, 1 measure of stucco, in order to determine the influence of the primers exactly). This mortar was applied manually at a thickness of approximately 8 mm on the horizontal specimens. Processing time was around 20 minutes.
The freshly rendered concrete plates were turned after 24 hours (with the plaster facing downwards) in order to achieve similar conditions as in practice (moisture diffusion moving downwards, as in the case of ceilings). Until measurements were recorded, the specimens were stored in stacks (with the help of wooden planks) in roofed enclosures at a minimum distance of 10 cm to the floor or to other specimens . Temperature and relative humidity of the storage area were similar to natural conditions (as in spring). They were between 10 C and 33 C, and 30 and 95 % relative humidity The specimens were turned again with the rendered side on top to measure adhesive strength.
Disc-shaped concrete units (ordinary concrete) with a thickness of 18 mm and a diameter of 300 mm were used to determine the temperature-based influence of certain selected primers. These were specially customised in a laboratory belonging to a building materials manufacturer (Heidelberger Maxit). Various degrees of surface coarseness in the micro-area were achieved by varying the synthetic sheeting of the mould.
Application of primer and plaster was carried out either immediately after delivery, or after a storage of 4 weeks at 20 C and 50 % relative humidity. Storage of the specimens led to a reduction of moisture in the concrete units from 7.6 % (as supplied, in sealed foil) to 6.7 % after one day, to an average of 3.8 % after 28 days.
Application of primers (Rikombi-Kontakt, Betokontakt, and a reference test without priming) and plasters (manual plaster IP 27 and spray plaster MP 75) was carried out as described above. The concrete plates were initially stored for 3 days (sealed in foil) at the desired temperatures (2 C, 10 C, 20 C ), and the primer was then applied. The primer coat was allowed to dry for 24 hours, after which the plaster was applied with a primary setting phase of 3 days at the respective temperatures. The plaster was then cut open along the sides after initial setting. For testing purposes, the specimens were stored in normal conditions (20 C and 50 to 65 % relative humidity) only after this point in time.
Test methods
Tests for adhesive strength were carried out in accordance with the German Industrial Norm 18555 (DIN 18555), "Test of mortars with mineral bonding agents; hard mortar; determination of adhesive strength". The test surface was drilled open (snap ring groove), and lightly abraded with sand paper. The test stamp was glued on the test surface with a synthetic resin-based glue. Breaking load was determined after one hour with the adhesion test system HP 850 at a load velocity of 25 N/s. The determined adhesive strength was a mean value of 6 or 7 tests. Individual values that showed high disparity to the mean value (freak values) were not taken into consideration.
The structure of the gypsum plaster was analysed with an environmental scanning electron microscope (XL-30 ESEM) fitted with an integrated energy-dispersive x-ray micro-analysis system. Phase analysis was carried out with an x-ray diffractometer (Siemens D 5000). Surface coarseness was charted with a laser scanning microscope with a micro-focus sensor (UBM Meßtechnik GmbH). The moisture of the specimens was determined in the usual mould, with drying up to weight constancy at 105 C (concrete) and 40 C (other building materials containing calcium sulphate).
Results and Discussion
Comparison of various primers
Generally speaking, almost all the primers fulfilled expectations that were attributed to them in the gypsum plaster/concrete system. It can be assumed that a gypsum plaster possesses sufficient bonding characteristics with the concrete when adhesive strength is measured at 0.2 N/mm2. However, a limiting value for guidelines and norms does not exist. It also became clear that a primer must definitely be used in the conditions chosen by us.
In a majority of the specimens, fracturing occurred in the interface between gypsum plaster and primer (adhesion fracture). In the case of some primers with high adhesive strength (primers 2 and 7), the fracture occurred partly in the interface between primer and mortar, and partly in the mortar itself (adhesion-cohesion fracture).
Plaster adhesion of non-stored concrete units was very high with some primers (2, 7 and 11). A lack of adhesion during the application of gypsum plaster without primer, as also was the case with primer 9, could be determined. Primers 1 and 8 showed low adhesive values.
It can be assumed that a five-week storage leads to superficial drying and partial carbonation of the surface. Due to these reasons, adhesion characteristics of the plasters changed. All specimens that originally showed low or average adhesive qualities increased their adhesive strength significantly. Only the reference test (specimen without primer) showed absolutely no adhesive characteristics even in this case.
Table 1
Adhesive strength of gypsum plasters on concrete depending on type of primer
|
Primer no. |
Application of primer without storage |
Application of primer after 35-day storage |
|
Adhesive strength N/mm2 |
Adhesive strength N/mm2 |
|
|
0 |
0 | 0 |
|
1 |
0,19 |
0,10 |
|
2 |
1,37 |
1,06 |
|
3 |
0,58 |
0,77 |
|
4 |
0,54 |
0,70 |
|
5 |
0,36 |
0,80 |
|
6 |
0,41 |
0,77 |
|
7 |
1,35 |
1,10 |
|
8 |
0,19 |
0,66 |
|
9 |
0 |
0,16 |
|
10 |
0,34 |
0,30 |
|
11 |
1,23 |
1,19 |
The test results could prove the differences in quality of the 10 primers used. We therefore recommend that users get appropriate information from manufacturers as a form of reassurance. Sufficient adhesive strength (0.2 N/mm2) should be achieved by primers in the concrete/primer/gypsum plaster interface. In order to determine the efficiency of a primer it is not only necessary to determine adhesive strength - fracture patterns are equally important.
Influence of the application temperature of primer
A comparison of the data in Fig. 1 clearly shows that in "normal" conditions (storage of concrete one month prior to application, temperature 20 C, 50 to 65 % relative humidity) sufficiently high adhesive strengths can be achieved one month after plaster application. This is almost completely independent of the type of primer, the type of plaster and the combinations used (Fig. 1: top).
A reduction in the application temperature of primers and plasters (2 C and 10 C) did not lead to any non-permitted low adhesive strength values in the variants observed (lower than 0.2 N/mm2). On the contrary, a significant improvement in bonding in the gypsum/concrete interface could be determined. This could also be observed particularly when a primer had not been used.
When application of primer and plaster took place immediately after the concrete sur-faces were manufactured (without storage), a completely different pattern emerged (Fig. 2). At temperatures of 20 C and 10 C adhesive characteristics for all variants (primer + gypsum plaster) corresponded to the (good and sufficient) adhesive qualities that were determined with a prior storage of one month. Sufficient adhesive strength could not be determined when no primer was used (with the exception of MP 75 at 10 C).
At extremely low temperatures (2 C), adhesive qualities of the primer Rikombi-Kontakt remained almost unchanged. In the same conditions, application of Betokontakt led to a decrease in adhesive strength; however the required value of approximately 0.2 N/mm2 was achieved. Even at this temperature adhesive qualities were completely lost when no primer was applied.
Influence of thorough back moisturisation of concrete
No significant influence on the adhesive strength could be determined when the back of a concrete plate with 18 mm thickness was subsequently moisturised with a 28-day old plaster consisting of a watery solution (water and saturated Ca(OH)2 solution ) for 28 days (Fig. 3: centre, bottom). As a result of moisturisation, the moisture content of the concrete plate increased from 3.8 % to only approximately 7.8 %. A thorough moisturisation of the gypsum plaster could not be observed. On the contrary, the moisture content of the plaster was reduced to 0.28 % as a result of the ongoing process of drying at 20 C and approximately 50 % relative humidity (at the beginning of moisturisation it was 0.45 %). In our opinion the manufactured concrete was very dense and therefore water impermeable (at least during the 28-day moisturisation period).
Figure 1
Adhesive strength (= Haftzugfestigkeit) in the gypsum plaster/primer/concrete system depending on application temperature of primer (= Haftbrücke) and plaster (= Putzmörtel)
top: 20 C centre: 10 C bottom: 2 C
Manufacturing and storage conditions: Storage of concrete plates 1 month
Average moisture of concrete: 3.8 %
Temperature impact : primer: 1 day
plaster: 3 days
subsequent storage (20 C, 50 % relative humidity ): 25 d
Figure 2
Adhesive strength (= Haftzugfestigkeit) in the gypsum plaster/primer/concrete system depending on application temperature of primer (= Haftbrücke) and plaster (= Putzmörtel)
top: 20 C centre: 10 C bottom: 2 C
Manufacturing and storage conditions: Storage of concrete plates: none
Average moisture of concrete: 7.6 %
Temperature impact: primer: 1 day
plaster: 3 days
subsequent storage (20 C, 50 % relative humidity): 25 d
Figure 3
Adhesive strength (= Haftzugfestigkeit) in the gypsum plaster/primer/concrete system depending on thorough back moisturisation of concrete
top: no moisturisation centre: moisturisation with water bottom: moisturisation with Ca(OH)2
Manufacturing and storage conditions : storage of concrete plates: 1 month
Application temperature of primer and plaster: 20 C
Subsequent storage (20 C, 50 % relative humidity): 28 d
Subsequent back moisturisation: 28 d
Influence of subsequent carbonation of the rendered plates
It was further determined how the adhesive qualities of the rendered plates were influenced by the carbonation process. Selected specimens chosen for this purpose were stored in a CO2 chamber (3 % CO2) after a 28-day setting period of the plaster (plaster application without storage at 20 C) additionally at 28 d, 20 C and 95 % relative humidity. Illustration 4 shows that changes can occur in such conditions; these changes do not have any significant influence. Apart from that, during this comparison of results it is important to pay attention to the fact that the plaster moisture levels of the subsequently carbonated specimens were initially higher at 0.7 % than in the reference tests (0.45 %). A drying of the plates (further storage at 20 C and 50 % relative humidity) did not lead to any changes in adhesive strength (deviation 3 %). It therefore becomes clear that low moisture content in the plaster (< 0,7) does not influence plaster strength.
Figure 4
Adhesive strength (= Haftzugfestigkeit) in the gypsum plaster/primer/concrete system depending on subsequent carbonation of rendered plates
Manufacturing and storage conditions: storage of concrete plates: 1 month
Application temperature of primer and plaster: 20 C
Subsequent storage (20 C, 50 % relative humidity): 28 d
Carbonation (3% CO2 at 20 C and 95 % relative humidity): 28 d
primer: Betokontakt mortar: IP 27
Influence of micro-coarseness
Various degrees of surface coarseness in the micro-area of the concrete were achieved by varying the synthetic sheeting in the mould. These differences were recorded with a laser scanning microscope.
Table 2
Technically determined values of coarseness (laser microscope UBM) on various plaster substrates
|
Substrate |
Recorded depth of coarseness acc. to DIN 4758 [μm] |
|
Concrete, without storage, no primer, ordinary synthetic sheeting |
8,3 |
|
Concrete, without storage, no primer, smooth synthetic sheeting |
5,1 |
|
Concrete, without storage, no primer, coarse synthetic sheeting |
43,1 |
|
Concrete, without storage, thin primer coat, ordinary synthetic sheeting |
16,4 |
|
Concrete, without storage, single primer coat, ordinary synthetic sheeting |
26,7 |
|
Concrete, without storage, double primer coat, ordinary synthetic sheeting |
29,6 |
Note: primer: Betokontakt (Knauf)
No significant influences were recorded in comparisons of adhesive strength in all variants of plaster/primer on concrete with varying coarseness in "normal" conditions (storage for 1 month prior to plaster application at 20 C and 50 to 65 % relative humidity).
(Cf. Fig. 5).
Environmental Scanning Electron Microscope (ESEM)
Tests of various specimens with ESEM show that dihydrate crystals of gypsum plaster grow into or through the primer in order to achieve good bonding characteristics (Fig. 6). Apart from that, a denseness of the immediate interface between gypsum plaster and primer or concrete can be observed. Primer-based differences can also be observed in this process (Fig. 7). However, dense interfaces could also be seen in cases where primer had not been used (Fig. 8). On the one hand, moisture transition (substrate absorbs moisture) with ion movement leads to full crystal formation of the gypsum crystals in the substrate (Fig. 6):
Figure 5
Adhesive strength (= Haftzugfestigkeit) in the gypsum plaster/primer/ concrete system depending on micro-coarseness of the concrete surface :
top: smooth sheeting centre: ordinary sheeting bottom: coarse sheeting
Manufacturing and storage conditions: Storage of concrete plates: 1 month
Application temperature of primer and plaster: 20 C
Subsequent storage (20 C, 50 % relative humidity): 28 d
Figure 6
Depiction of the gypsum plaster/primer interface with ESEM
Top: In-grown gypsum crystals in primer (Rikombi-Kontakt)
Centre: full crystallisation on primer facing substrate (Rikombi-Kontakt)
Bottom: In-grown gypsum crystals (Betokontakt) with detached surface
Figure 7
Depiction (ESEM) of a section of gypsum plaster with primer (substrate removed!)
Top: Transition area containing primer became denser
Gypsum/primer 10 to 20 μm (primer 1 Betokontakt)
Bottom: Clear denseness of transition area containing primer
Gypsum/primer 15 to 60 μm (primer 2 Quarzgrund)
Figure 8
Totally detached gypsum plaster from light concrete unit (claim for damage Bad Homburg), depicted with ESEM
Top: gypsum structure in centre of plaster (unchanged)
Bottom: gypsum structure in the plaster/concrete interface (carried out without primer!)
à Dense structure, change in crystal habits, detachment of gypsum crystals
centre), which in turn causes interlocking of the gypsum plaster with the substrate. On the other hand, however, the subsequent drying process causes a continual moisture transition in the opposite direction, i.e. from the concrete to the gypsum plaster. This leads to a detachment process with a change in crystal habits in the interface crystals. Fig. 8 depicts the differences as compared with unchanged structures. It can be assumed that both the detachment process and the change in crystal habits have a negative influence on the adhesive properties of gypsum plaster on concrete.
Conclusions
On the one hand, dihydrate crystals of the plaster must grow into the substrate in order to achieve sufficient adhesive properties. This requires a certain capillarity of the substrate, and a sufficient moisture gradient between plaster and concrete. On the other hand, detachment and changes in crystal habits caused by subsequent drying must be kept at a minimum so that the gypsum/concrete bonding is not weakened. For this reason the substrate must possess sufficient capillarity; it should not, however, be too moist.
Damages at a later stage cannot be ruled out when time frames between the manufacturing process of concrete walls and the rendering phase are extremely short.
All tests on this topic were commissioned by the Federal Association of Gypsum and Gypsum Board Industry e.V., Germany (Bundesverband der Gips- und Gipsbauplattenindustrie e.V.). The authors are grateful for the interesting assignment and fruitful discussions. We would also like to take this opportunity to thank Mr. Forg (of Heidelberger Maxit) and Professor Dr. Hummel (of Gebrüder Knauf Westdeutsche Gipswerke, Iphofen) for their support.
References
/01/ Ludwig, U.; Kuhlmann, J.:
Untersuchungen an Putzgipsen
In.: Tonindustrie-Zeitung. 98 (1974) 1. S. 1 4
/02/ Lucas, G.
Beton als Putzgrund
In.: Tonindustrie-Zeitung. 98 (1974) 1. S. 5 7
/03/ Poch, W.
Gipse zum Verputzen von Beton
In.: Tonindustrie-Zeitung. 98 (1974) 1. S. 7 - 10
/04/ Haerter, M.
Untersuchungen der Grenzflächenvorgänge zwischen Gipsputzen und verschiedenen Untergründen mit dem Rasterelektronenmikroskop
In.: Tonindustrie-Zeitung. 98 (1974) 1. S. 13 - 15
/05/ Lange, J.
Einfluss von Methylcellulose auf die Haftung von Gipsputzen
In.: Tonindustrie-Zeitung. 98 (1974) 1. S. 15 16
/06/ Kollmann, H.; Strübel, G.; Trost, F.
Reaktionsmechanismen zur Bildung von Treibkernen in Kalk-Gips-Putzen durch Ettringit und Thaumasit
In: Zement-Kalk-Gips. 30 (1977) 5. S. 224 228
/07/ Hurbanic, M.; Schicht, R.; Scholze H.
Haftung von Gips
In.: Tonindustrie-Zeitung. 101 (1977) 1. S. 232 - 235
/08/ Engelke, H.
Veränderung mechanischer Eigenschaften von Gipsbaustoffen bei und nach Durchfeuchtung
In: Zement-Kalk-Gips. 32 (1979) 11. S. 560 568
/09/ Grunau, E.B.
Verhindern des Abfallens von Deckenputzen
In: Stuckgewerbe (1980) 1. S. 17 19
/10/ Ostrowski, C.
Abfallen der Gipsputze von Betonunterlagen
In: baustoffindustrie. (1980) 3. S. 8 - 12
/11/ Scholze, H.; Hurbanic, M.; Engelke, H.
Einfluss des Gefüges auf Haftung und Oberfläche von Gipsputzen
In: Zement-Kalk-Gips. 34 (1981) 6. S. 318 328
/12/ Manns, W.; Wisotzky, T.; Zimbelmann, R.
Einflüsse auf das Haften von Innenputzen auf Beton
In: Schriftenreihe des Otto-Graf-Institutes. 77 (1983). S. 1 27
/13/ Grün, I.
Gipsputz an Betondecken
In: Bau. (1983) 5. S. 367 371
/14/ Moisset, J.
Prüfungen der Haftung von Gipsputzen auf Baustoffen
In.: Tonindustrie-Zeitung. 108 (1984) 5. S. 289 291
/15/ Aeppli, E.
Einfluss der Betonfeuchte auf die Ablösung von Gipsputzen Feuchtemessung in Beton
In.: Tonindustrie-Zeitung. 108 (1984) 6. S. 383 - 386
/16/ Khoda, F.
Innenputz aus Gips Sichere Haftung auf Beton-Untergründen
In.: Tonindustrie-Zeitung. (1985) 9. S. 694 697
/17/ Moldan, D.
Ursachen für mangelhaftes Haften von Maschinenputzgips auf Beton
RWTH Aachen, Dissertation. 1986. 197 S.
/18/ Khoda, F.
Richtige Applikation von Gipsputzen an Betondecken (S. 59 63)
In : WTA Seminar Schadensfreier Putz. 1986. 167 S.
/19/ Koslowski, Th.; Ludwig, U.
Mögliche Ursachen mangelnder Haftung von Gipsputzen an Betondecken (S. 64 92)
In : WTA Seminar Schadensfreier Putz. 1986. 167 S.
/20/ Böttcher, R.
Schäden an Gipsputzdecken
In: Der Stuckateur (1988) 6. S. 6 9
/21/ Schütz, P.
Notwendigkeit + Wirkung + Prüfparameter einer Haftbrücke für Gipsputz auf Betonuntergrund
In: BMI, Universität Innsbruck. 1992. S. 37 44
/22/ Khoda, F.
Haftmechanismus der Gipsputze auf ihrem Untergrund
In: BMI, Universität Innsbruck. 1992. S. 45 48
/23/ Khoda, F.
Innenputze aus Gips
In: Stuck Putz Trockenbau (1992) 2. S. 10 13
