Advanced

Water Absorption in Two-Layer Masonry Systems - Properties, profiles and predictions

Johansson, Peter LU (2005) In Report TVBM 1024
Abstract (Swedish)
Popular Abstract in Swedish

Många beständighetsproblem uppträder i gränsytor mellan material. Exempel är lossfrysning av puts på tegel, avflagning av färg, sprickbildning hos ytskikt, vidhäftningsbrott mellan reparationsmaterial och underlag. I nästan alla fall är nedbrytningen kopplad till alltför höga fuktnivåer i gränsytan. Flera beständighetsproblem, t.ex. frostnedbrytning, uppstår enbart om materialen kan ta upp vatten genom mer eller mindre långvarig kapillärsugning. Det har varit mycket svårt att teoretiskt förutse vilka fuktnivåer som kommer att uppstå i dessa fall eftersom man saknat materialdata för beräkning av fukt vid mycket höga fuktnivåer. Dessutom saknas experimentella metoder att bestämma erforderliga... (More)
Popular Abstract in Swedish

Många beständighetsproblem uppträder i gränsytor mellan material. Exempel är lossfrysning av puts på tegel, avflagning av färg, sprickbildning hos ytskikt, vidhäftningsbrott mellan reparationsmaterial och underlag. I nästan alla fall är nedbrytningen kopplad till alltför höga fuktnivåer i gränsytan. Flera beständighetsproblem, t.ex. frostnedbrytning, uppstår enbart om materialen kan ta upp vatten genom mer eller mindre långvarig kapillärsugning. Det har varit mycket svårt att teoretiskt förutse vilka fuktnivåer som kommer att uppstå i dessa fall eftersom man saknat materialdata för beräkning av fukt vid mycket höga fuktnivåer. Dessutom saknas experimentella metoder att bestämma erforderliga materialdata liksom verifierade metoder för beräkning av fukttransport vid höga fuktnivåer. Inte heller har det varit känt hur olika materialegenskaper såsom jämviktsfuktkurvor och porstorleksfördelningar hos angränsande material påverkar fuktförhållandena i gränsytan mellan materialen.



Målet med projektet var att utveckla metoder för bestämning av relevanta materialegenskaper och att tillämpa dessa metoder på ett antal material, främst cementbruk, kalkcementbruk och kalksandsten. Målet var dessutom att undersöka möjligheter att teoretiskt beräkna fukttillståndet i gränsytor genom att jämföra teoretiskt uppmätta fuktprofiler med beräknade profiler baserade på uppmätta materialegenskaper.



Jämviktsfuktkurvor bestämdes över hela fuktområdet, från fullständig vattenmättnad till fullständig uttorkning. Inom det hygroskopiska området (0 till 98% relativ fuktighet, rf) bestämdes sorptionsisotermen, dvs. sambandet mellan jämviktsfukthalt och rf. Såväl desorptionsisotermen som absorptionsisotermen bestämdes med hjälp av en automatisk sorptionsvåg i vilken omgivande rf ändras i steg enligt ett förprogrammerat schema. Inom det överhygroskopiska området (rf>98%) bestämdes den kapillära jämviktsfuktkurvan, dvs. sambandet mellan fukthalt och kapillärt vattentryck. Pressure plate-tekniken användes. Det vattenmättade provet placeras på ett finporöst vattenmättat keramiskt filter och innesluts i en tryckkammare. Luftrycket i denna ändras stegvis varvid vatten efter varje tryckökning trycks ut ur provet tills en resterande vattenhalt i provet som motsvarar det aktuella trycket uppnåtts. Fukttransportkoefficienten bestämdes över hela fuktområdet på prover som konditionerats till olika nivå och sedan utsatts för ensidig kapillärsugning. Inflödet av vatten mättes. Transportdata utvärderades med Bolzmanntransformation.



Två tekniker för bestämning av fuktprofiler användes; (1) Nuclear Magnetic Resonance (NMR) med vars hjälp den totala vattenhalten i ett smalt ?fönster? vinkelrätt mot fuktflödet kan registreras, (2) ?spräck-och-torktekniken? vid vilken långa kapillärsugande prover efter en förutbestämd uppsugningstid momentant spräcks upp i ett antal skivor som vägs och torkas. Fuktprofiler under ensidig bestämdes för enskilda material och materialkombinationer. Dessa avsågs simulera vanliga fasader av puts på murverk. Såväl tjocklek hos yttre putsskiktet som dess egenskaper varierades. Experimentella materialdata användes i ett existerande fuktberäkningsprogram för att teoretiskt simulera fuktförhållandena hos de experimentella kapillärförsöken. Samma randvillkor användes vid beräkning som vid försök.



Såväl desorptions- som absorptionsisotermer kunde bestämmas med hög precision med sorptionsvågen. Pressure-platemetoden att bestämma jämviktsfuktkurvor inom överhygroskopiskt område fungerade utmärkt för desorption (uttorkning). Däremot kunde man inte ens efter stora ansträngningar bestämma absorptionskurvan. Vid de teoretiska beräkningarna användes därför desorptionskurvan trots att försöken var absorptionsförsök.



Försöksmetoden att bestämma fukttransportkoefficienter fungerade väl för höga fuktnivåer. Däremot visade sig metoden ge mindre trovärdiga resultat vid lägre fuktnivåer. Resultatspridningen kunde reduceras kraftigt genom att fukttillståndet uttrycktes som porfyllnadsgrad. Genom detta kunde effekten av variationer hos provkroppar reduceras. NMR-tekniken för bestämning av fuktprofiler under pågående kapillärsugning gav hög upplösning och är icke-destruktiv, men är dyrbar och visade sig vara känslig för järn i provet. Spräck-och-tork-tekniken visade sig vara enkel och tillförlitlig men har lägre upplösning och är förstörande. Mätningarna på materialkombinationen puts på underlag visade att fuktförhållandet i puts och gränszon berodde mer på fuktinträngningen och fuktnivån i underlaget än på putslagrets tjocklek.



Datorberäkningarna för materialkombinationer gav god överensstämmelse med försöksdata. Samstämmigheten var sämre när den tillämpades på enskilda material. Analys visade att detta sannolikt beror på felaktigheter hos uppmätta fukttransportkoefficienter vid lägre fuktnivåer.



Det använda fuktberäkningsprogrammet JAM-2, som främst utvecklats för fukttransport inom hygroskopiskt område, visade sig även kunna användas för beräkning av fukttransport vid mycket höga fuktnivåer. Beräkningar av fuktprofiler vid kapillärsugning visade god överensstämmelse med uppmätta fuktfördelningar vid. Överensstämmelsen var bättre för kombinationer av olika cementbruk och kalksandsten än för materialen testade enskilda.



De mätmetoder som användes för att bestämma erforderliga materialkoefficienter, gällande vid höga fuktnivåer, visade sig fungera väl. Metodik måste dock utvecklas för bestämning av uppfuktningskurvor i överhygroskopiskt område. Detta krävs för att man skall förbättra noggrannheten i beräkningarna och för att man även skall kunna beräkna fukttransport vid varierande yttre fuktbelastning. Den utvecklade mät- och beräkningstekniken bör kunna användas för utveckling av materialkombinationer med hög beständighet. (Less)
Abstract
The main impetus for this research was the desire to verify models for determining transient moisture profiles in the capillary range. For the evaluations, transient moisture profiles were both simulated and measured, after which the results of the two methods were compared. The study was performed for on specimens exposed to continuous water absorption, using specimens of both single and combined materials. The materials used for the study were two mortars and a lime silica brick. For testing, we chose materials with moisture properties corresponding to those of common inorganic façade materials.



As input for the simulations, data such as sorption isotherms and moisture diffusivities were determined for the studied... (More)
The main impetus for this research was the desire to verify models for determining transient moisture profiles in the capillary range. For the evaluations, transient moisture profiles were both simulated and measured, after which the results of the two methods were compared. The study was performed for on specimens exposed to continuous water absorption, using specimens of both single and combined materials. The materials used for the study were two mortars and a lime silica brick. For testing, we chose materials with moisture properties corresponding to those of common inorganic façade materials.



As input for the simulations, data such as sorption isotherms and moisture diffusivities were determined for the studied materials. Sorption isotherms were experimentally determined over the complete moisture range, two separate methods having to be used for the hygroscopic and the capillary parts of the moisture range. In the hygroscopic moisture range, both absorption and desorption isotherms could be determined, but in the capillary moisture range only desorption isotherms could be determined. To allow measurement of absorption isotherms in the capillary moisture range, the measuring equipment had to be modified; however, despite the modifications, reliable absorption measurements were still not possible. The desorption isotherms determined by the two different methods generally agreed well for all tested materials.



Moisture diffusivities were determined by evaluating a series of water absorption tests using a Boltzmann transformation principle. To determine moisture diffusivities over the complete moisture range, the water absorption tests were performed on specimens conditioned to different initial moisture contents, ranging from completely dry to the vacuum saturation point. The variation in the results was significantly reduced when the water content was expressed as the degree of vacuum saturation; thus, moisture diffusivity was determined as a function of moisture content, expressed as a degree of vacuum saturation.



Different methods for measuring the transient moisture profiles were considered. The moisture profile measurements were supposed to be performed on specimens of both single and combined materials. Two of the most interesting methods were the nuclear magnetic resonance (NMR) technique and the slice and dry method. The NMR technique displayed excellent precision in terms of both moisture content measurement and the spatial resolution. However, the method was severely limited, in that even an insignificant amount of iron in the specimen disturbed the measurements. Another limitation was restricted access to the measuring equipment.



For the larger-scale measurements of the moisture profiles, the slice and dry technique was chosen. This method offered high and conclusive precision in terms of moisture content measurement, and only relatively simple laboratory equipment was needed. One drawback of the method was that the specimens had to be sliced for testing, meaning that a particular specimen could not be tested repeatedly. Other drawbacks were limited spatial resolution and the destruction of the specimen.



Moisture profiles were measured after a continuous wetting phase, for both the specimens of three single materials and the two material combinations. The combinations were intended to simulate a façade consisting of a layer of mortar on masonry: the outer layer of mortar was exposed to continuous wetting, and the moisture profiles were then measured for the material combination.



Both the thickness of the outer mortar layer and the properties of the mortar were varied. The measurements showed that the moisture properties of the outer layer of mortar depended more significantly on both the moisture penetration depth and moisture levels in the underlying material than on mortar thickness.



Moisture profile simulations were performed using off-the-shelf software, using as inputs the absorption isotherms and moisture diffusivities determined in this work. The simulations were intended to reflect the larger-scale experimental portions of this research. To this end, the boundary conditions and wetting durations used as inputs corresponded to the conditions of the larger-scale moisture profile measurements.



The simulated and measured results for the combined materials largely agreed with each other, in terms of both moisture levels and depth of moisture front. However, the simulated moisture profiles of the two mortars as single materials significantly diverged from the measured values. The simulated profiles lacked a clear front in the moisture levels and moreover were overestimated after the front. Alternative simulations were carried out in which the deviation between the simulations and the measurements was clarified, and was shown to depend on errors in the technique for evaluating diffusivities at low moisture levels.



Comparison between the simulated and measured moisture profiles indicated that moisture profiles can be predicted with considerable precision in the high moisture range. However, by determining additional material data pertaining to moisture diffusivity in the low moisture range, the precision of the predictions can be improved. A primary conclusion of the research is that moisture profiles can be predicted for combined materials using off-the-shelf software. (Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • PhD Hjorslev Hansen, Morten, SBI, Danmark
organization
publishing date
type
Thesis
publication status
published
subject
keywords
moisture-diffusivities, masonry, measuremets, moisture-profiles, Material technology, simulations, sorption-isotherms, Materiallära, materialteknik
in
Report TVBM 1024
pages
151 pages
publisher
Division of Building Materials, LTH, Lund University
defense location
Room V:B, V-building John Ericssons väg 1 Lund Institute of Technology
defense date
2005-12-09 13:15
external identifiers
  • other:ISRN: LUTVDG/TVBM--05/1024--SE(1-151)
ISSN
0348-7911
ISBN
91-628-6680-X
language
English
LU publication?
yes
id
d42bee14-5ba9-433f-a345-b8f5e859dd58 (old id 24750)
date added to LUP
2007-06-01 08:22:57
date last changed
2016-09-19 08:44:58
@phdthesis{d42bee14-5ba9-433f-a345-b8f5e859dd58,
  abstract     = {The main impetus for this research was the desire to verify models for determining transient moisture profiles in the capillary range. For the evaluations, transient moisture profiles were both simulated and measured, after which the results of the two methods were compared. The study was performed for on specimens exposed to continuous water absorption, using specimens of both single and combined materials. The materials used for the study were two mortars and a lime silica brick. For testing, we chose materials with moisture properties corresponding to those of common inorganic façade materials.<br/><br>
<br/><br>
As input for the simulations, data such as sorption isotherms and moisture diffusivities were determined for the studied materials. Sorption isotherms were experimentally determined over the complete moisture range, two separate methods having to be used for the hygroscopic and the capillary parts of the moisture range. In the hygroscopic moisture range, both absorption and desorption isotherms could be determined, but in the capillary moisture range only desorption isotherms could be determined. To allow measurement of absorption isotherms in the capillary moisture range, the measuring equipment had to be modified; however, despite the modifications, reliable absorption measurements were still not possible. The desorption isotherms determined by the two different methods generally agreed well for all tested materials.<br/><br>
<br/><br>
Moisture diffusivities were determined by evaluating a series of water absorption tests using a Boltzmann transformation principle. To determine moisture diffusivities over the complete moisture range, the water absorption tests were performed on specimens conditioned to different initial moisture contents, ranging from completely dry to the vacuum saturation point. The variation in the results was significantly reduced when the water content was expressed as the degree of vacuum saturation; thus, moisture diffusivity was determined as a function of moisture content, expressed as a degree of vacuum saturation.<br/><br>
<br/><br>
Different methods for measuring the transient moisture profiles were considered. The moisture profile measurements were supposed to be performed on specimens of both single and combined materials. Two of the most interesting methods were the nuclear magnetic resonance (NMR) technique and the slice and dry method. The NMR technique displayed excellent precision in terms of both moisture content measurement and the spatial resolution. However, the method was severely limited, in that even an insignificant amount of iron in the specimen disturbed the measurements. Another limitation was restricted access to the measuring equipment.<br/><br>
<br/><br>
For the larger-scale measurements of the moisture profiles, the slice and dry technique was chosen. This method offered high and conclusive precision in terms of moisture content measurement, and only relatively simple laboratory equipment was needed. One drawback of the method was that the specimens had to be sliced for testing, meaning that a particular specimen could not be tested repeatedly. Other drawbacks were limited spatial resolution and the destruction of the specimen.<br/><br>
<br/><br>
Moisture profiles were measured after a continuous wetting phase, for both the specimens of three single materials and the two material combinations. The combinations were intended to simulate a façade consisting of a layer of mortar on masonry: the outer layer of mortar was exposed to continuous wetting, and the moisture profiles were then measured for the material combination.<br/><br>
<br/><br>
Both the thickness of the outer mortar layer and the properties of the mortar were varied. The measurements showed that the moisture properties of the outer layer of mortar depended more significantly on both the moisture penetration depth and moisture levels in the underlying material than on mortar thickness.<br/><br>
<br/><br>
Moisture profile simulations were performed using off-the-shelf software, using as inputs the absorption isotherms and moisture diffusivities determined in this work. The simulations were intended to reflect the larger-scale experimental portions of this research. To this end, the boundary conditions and wetting durations used as inputs corresponded to the conditions of the larger-scale moisture profile measurements.<br/><br>
<br/><br>
The simulated and measured results for the combined materials largely agreed with each other, in terms of both moisture levels and depth of moisture front. However, the simulated moisture profiles of the two mortars as single materials significantly diverged from the measured values. The simulated profiles lacked a clear front in the moisture levels and moreover were overestimated after the front. Alternative simulations were carried out in which the deviation between the simulations and the measurements was clarified, and was shown to depend on errors in the technique for evaluating diffusivities at low moisture levels.<br/><br>
<br/><br>
Comparison between the simulated and measured moisture profiles indicated that moisture profiles can be predicted with considerable precision in the high moisture range. However, by determining additional material data pertaining to moisture diffusivity in the low moisture range, the precision of the predictions can be improved. A primary conclusion of the research is that moisture profiles can be predicted for combined materials using off-the-shelf software.},
  author       = {Johansson, Peter},
  isbn         = {91-628-6680-X},
  issn         = {0348-7911},
  keyword      = {moisture-diffusivities,masonry,measuremets,moisture-profiles,Material technology,simulations,sorption-isotherms,Materiallära,materialteknik},
  language     = {eng},
  pages        = {151},
  publisher    = {Division of Building Materials, LTH, Lund University},
  school       = {Lund University},
  series       = {Report TVBM 1024},
  title        = {Water Absorption in Two-Layer Masonry Systems - Properties, profiles and predictions},
  year         = {2005},
}