Bulletin of the University of Minnesota ENGINEERING EXPERIMENT STATION fRANK B. RowLEY, Director BULLETIN NO. 17 METHODS OF MOISTURE CONTROL ~ND THEIR APPLICATION TO BUILDING CONSTRUCTION fRANK B. RowLEY, M.E. Professor of Mechanical Engineering . AxEL B. ALGREN, M.S. in M.E. Assistant Professor of Mechanical Engineering and CLARENCE E. LUND, M.S. (M.E.) Mechanical Engineer, Engineering Experiment Station Vol. XLIII No. 28 April 1 0 1940 MINNEAPOLIS Entered at the post office in Minneapolis as second-class matter, Minneapolis, Minne- sota .. Accepted for mailinq at special rate of postaqe provided for in section 1103, Act of October 3, 1917, authorized July 12, 1918. TABLE OF CONTENTS Page PART !-General Acknowledgment ..................................................................... ..................................................................................... v Introduction ..................................................................................... ............................................................................ 1 Analysis of problems ....................................................................................................... ....................................... 2 Theory of vapor transmission through materials........ ......................... ................................... 5 PART II-Equipment and Instruments Test apparatus ............................................................ ................................................................................. 15 Cold room ................. ...................................................... ....................................... ................................................. 15 Cooling equipment ............................................................ . ................................................ . Air conditioning equipment ....................... . ..................................................... . Control equipment ................................................ . ............................................................................ . Instruments .............................................. . .................................................................................... . Moisture control methods . . .................................................................................................. . Wet and dry bulb temperature apparatus. . ........................................................................ . 15 15 20 21 21 21 Special humidity controller ................... ........................... ......................................... 21 Dew-point temperature apparatus ......................... ......................................... 22 Test units . ..................... . ....... H............................................................ ............................. 23 Wall construction ....................... . Attic construction .................................... . Vapor barrier test apparatus . PART III-Test Results 23 23 25 Vapor barriers .... ......................................... ......................... ............................................................. 29 Test procedure ........................ ....... ............................................................ 29 Barriers tested in combination with walls .............................. .. .................................... 29 Barriers tested by special test apparatus....... ...... ...................... 34 Practical application of barriers......... ......................................................... 36 Rate of condensation with variable temperatures..................................................................... 36 Rate of condensation with variable relative humidity......................................................... 38 Water vapor within a wall......................... ..... ................................. ...................................... 39 Wood absorption blocks ................... ................................. ............................. 40 Dew-point temperature measurements..... . ............................ 41 Vapor transmittance coefficients........ ........................... .... ................................ ............... 43 Coefficients by two different test methods........................... .................................................... 43 Vapor transmittance coefficients for mineral wool insulation and plaster 45 Vapor transmittance coefficients for paint .............. . ...................................... 47 Vapor transmittance coefficients for plaster and plaster base materials 48 Wet plaster and vapor barrier..... ........................ ............................................................ 49 Ventilation and condensation within a structure.. . . ..... .............................................. 50 Ventilation of walls by diffusion................. .......................... ......................... 51 Ventilation of walls by convection currents.......................................................................... 51 Ventilation of attic spaces ................... :... ............................................................... 51 Conclusions ................................................ H.. ......................... .....................................................• 56 TABLES Table Page I. Condensation on inner surface of sheathing for different types of mem- brane vapor barriers placed between metal lath and studs ............................. 30 II. Condensation on inner surface of sheathing for test walls with different types of finishes on interior surfaces of plaster...................................................... 31 Addenda for Table III.............................................................. 32 III. Vapor transmission for different materials by special test method 32 IV. Relative vapor transmission for different surface coatings by special test method ...................................................................................................................................................... 34 Addenda for Table IV........................................................................................................................... 35 V. Calculated temperatures and vapor pressures through insulated wall for different outside air temperatures-other conditions remaining constant .................................................................................................................................................................. 38 VI. Dry bulb and dew-point temperatures by test at various vertical planes in an uninsulated and an insulated wall...... ............................................. 42 VII. Vapor transmittance coefficients for ~-inch plaster on metal lath and 3% inches of mineral wool............................................................................................................... 44 VIII. Vapor transmittance coefficients for mineral wool pad 2;.-8 inches thick and for ~ inch of plaster applied to metal lath......... ............................ 46 IX. Vapor transmittance coefficients for different types of vehicles and paints 47 X. Vapor transmittance values for plaster in combination with plaster base material ............................................................................................................................. 48 XI. Air temperatures and humidity conditions for wet plaster tests........................ 49 XII. Attic ventilation ........................ ............................................................ ................................................... 53 ILLUSTRATIONS Figure Page 1. Wall built of nonhygroscopic, homogeneous material per~able to water vapor .................................................................................................................................... :.............................. 9 2. Wall built of nonhygroscopic, homogeneous material permeable to water vapor .............................................................................................................................................................................. 9 3. Wall built of hygroscopic, homogeneous material impermeable to water vapor ................................................................................................................................................................................. 9 4. Wall built of nonhygroscopic, homogeneous material permeable to water vapor .................................................. ........................................................................................................ 11 5. Wall built of nonhygroscopic, homogeneous material permeable to water vapor .......................................................... ..................................................................................................................... 11 6. Frame wall without insulation showing normal temperature and vapor pressure gradient through walL.................................................. ......................................... .......... 12 7. Frame wall with insulation showing normal temperature and vapor pres- sure gradients as established through wall........................ ................................................. 13 L__.__ IV CONTENTS Figure Page 8. View of cold room showing test houses in place ........ . 16 9. Sectional view through cold room and cooling unit.. 16 10. Interior view of cooling unit while under construction..... .............................................. 17 11. View of cooling unit while under construction .................................................. . 17 12. View of completed cooling unit and one end of control room............ 18 13. Ammonia compressor, control room, and defrosting fan for cooling equipment ..................................................................... .................................................................... 18 14. Air conditioning unit for test houses.............. ........................................................................ 19 15. Sectional view of air conditioning unit with distribution ducts to test houses in cold room......................................................... ....................................... ................................ 19 16. Control and thermocouple panel for test equipment ................................................ . 17. Sectional view of test houses showing details of construction .. 20 24 18. Type of wall construction used to facilitate inspection.. ................................. 24 19. Test Jlo~se used for wet plaster test and also subsequent tests on wall ven- ttlatton .......................................................................................................................................... ............................... 25 20. Construction details of attic for attic ventilation tests .. 21. View of set-up for ventilated and unventilated attics 22. Line drawing showing details of construction for special vapor barrier 26 26 test apparatus ......................................................................................................................................................... 27 23. View of apparatus for making tests on vapor barriers .... 27 24. Per cent moisture absorbed by wood absorption blocks placed at various planes in walls........................................................ ................................ ............................................................. 40 25. Dry bulb and dew-point temperatures in the interior sections of an unin- sulated wall .................................................................... ............................ ........................................................... 41 26. Dry bulb and dew-point temperatures in the interior sections of an insulated wall ...................................................... . ............................................................................................... 41 · ACKNOWLEDGMENT The authors wish to acknowledge the co-operation and support of the National Mineral Wool Association, which by its generous contributions has made this program possible, and also the work of the Technical Committee of the Association consisting of Jan S. Irvine, chairman, W. W. Cullin, E. W. Mcl\,Iullen, C. L. Neumeister, and E. R. \Villiams, who have contributed liberally of their time and technical ability in formulating the research program and considering methods of procedure. The authors and the Technical Committee acknowledge the co-operation given by Mr. L. V. Teesdale of the Forest Products Laboratory, Depart- ment of Agriculture, lVIadison, Wisconsin, and also the assistance of Mr. Robert Lander in the design of special instruments. Methods of Moisture Control and Their Application to Building Construction PART I-GENERAL INTRODUCTION Experience in low temperature storage rooms and in many industrial buildings where high indoor relative humidities prevail (such as cream- eries and laundries) has demonstrated that certain precautionary meas- ures are necessary with respect to moisture control. Specifically, it has been found that in such cases insulation is necessary to prevent surface condensation. It has also been found in building construction of the types mentioned that moisture barriers must be properly employed to protect the insulation and other constituent portions of the construction against undesirable effects of excessive moisture. In recent years it has become common practice to increase indoor humidities in residences for purposes of better health, greater comfort, and protection of buildings and furnishings. Thus, the control of mois- ture in residential construction has become quite as important as the control of moisture in industrial construction, and accordingly this prob- lem has had the consideration of architects and residential builders. While the principles of proper moisture control which have been estab- lished in certain industrial fields obviously apply in residential construc- tion, the conditions in residences vary sufficiently from those which pre- vail in industry to justify specific investigations to develop the most practical means of adjusting residential building practice and operating methods to the new conditions which prevail. In the early part of 1937 the National Mineral Wool Association in co-operation with the University of Minnesota started a compre- hensive research program, the objects of which were: first, to determine the extent of condensation within modern buildings and with different methods of operation ; second, to measure the effectiveness of different types of construction; and third, to find the limiting conditions under which various types of building designs may be used without excessive condensation and still retain the benefits of air conditioning and good building construction. It was decided to approach the problem in a practical way and to make tests, the results of which would apply to full-scale building con- struction under temperature and humidity conditions to be expected in severe climates. To accomplish this, a large-size test room was con- structed and provided with cooling equipment of sufficient capacity to maintain inside air temperature at 25 o F. below zero. Small-size test houses were constructed and placed within this room, with provision for maintaining the air temperatures and humidities within these houses 2 METHODS OF MOISTURE CONTROL at any desired level. The walls of the test houses were so constructed that they could be taken apart and examined from time to time through- out a test period to determine the physical conditions of their interiors, and to make comparisons between various types of structures. The details of the test apparatus will be described later in this bulletin. In addition to the results presented in this bulletin three papers have been published as follows: "Condensation within Walls," by F. B. Rowley, A. B. Algren, and C. E. Lund. Journal of the American Society of Heating and Ventilating Engineers, 10 :49-60. January, 1938. "Condensation of Moisture and Its Relation to Building Construction and Operation," by F. B. Rowley, A. B. Algren, and C. E. Lund. Journal of the Ameri- can Society of Heating and Ventilating Engineers, 11:41-49. January, 1939. "A Theory Covering the Transfer of Vapor through Materials," by Frank B. Rowley. J mtrnal of the American Society of Heating and Ventilating Engineers, 11 :452-65. July, 1939. The next step in the program is to construct a large-scale test house within the cold room and to apply to full-scale building construc- tion those principles which have been proven satisfactory in the smaller test houses. ANALYSIS OF PROBLEMS In practical ventilating problems all air contains moisture in the form of water vapor, and_ we speak of it as saturated or partly saturated, depending upon whether the maximum, or only a part of the maximum amount of moisture, is present. The amount of vapor which may be contained in a given volume of air increases as its temperature is in- creased. Consequently, if air which is partly saturated is cooled, the percentage of saturation will be increased until it reaches 100 per cent, after which further cooling will condense out some of the vapor. The · properties of the water vapor are not changed by the presence of the air. The maximum amount of vapor that can be contained in a given space at a given temperature, the pressure exerted by the vapor, the dew- point temperature, etc., are all the same, regardless of whether or not air is present. Water vapor exerts a pressure which depends upon its temperature and percentage of saturation. Water at 80° F. temperature will boil at a pressure of 1.03 inches of mercury, and this is the vapor pressure of saturated vapor at 80° F. If the vapor at 80° F. is only SO per cent saturated, then the vapor pressure is likewise SO per cent that of saturation, or .S 16 inch of mercury. The saturation temperature corresponding to .S16 inch of mercury is S9.8° F. Therefore, if a vapor which is SO per cent saturated at 80° F. is cooled to S9.8° F., it will become saturated and further cooling will cause condensation. When air is mixed with the vapor, the pressure exerted by the air is added to that exerted by the vapor, but the vapor pressure is the one that must be considered in connection with the properties of the vapor or its action in connection with the condensation problem in a building. ANALYSIS OF PROBLEMS 3 Condensation may take place on the interior surfaces of the windows or walls of a building if the temperatures of these surfaces are below the dew-point temperature of the vapor within the building. For most types of building construction now in use sufficient data are available for calculating the minimum inside surface temperatures which may be .expected for given outside and inside air temperatures. These surface temperatures, together with data available on the properties of water vapor, give a basis for determining the temperature and humidity con- ditions which are likely to cause surface condensation. The vapor within a building is not confined to contact with the interior surfaces of the glass and walls. It penetrates int,o the interior parts of the walls and will pass • through some types of materials very readily. As the vapor passes out- wards through the walls of a house located in a cold climate it will come in contact with colder materials, and a condition may be set up in which the dew-point temperature of the vapor in a given part of the wall is above the temperature of the material in contact with this vapor. Under these conditions condensation may take place, and free moisture or frost may be formed in the wall, depending upon the temperature. The conditions which cause condensation ''vithin a wall are no dif- ferent from those which cause condensation on the surfaces of the wall. In both cases the temperature of the material is below the dew-point temperature of the vapor in contact with it. For most types of con- struction it is possible to calculate the temperatures which may be expected throughout the wall with the same degree of accuracy as for the interior surface temperatures. There is, however, at present no method by which the vapor density within the wall may be calculated with any degree of certainty, even tho the vapor conditions on both sides of the wall are known. The laws governing the transmission of vapor through materials, and the vapor transmitting properties of the various materials and combinations of materials used in building construction have not as yet been investigated. The motive force which causes vapor flow is usually considered to be the difference in vapor pressure along the path of flow, and the rate of vapor flow through any part of the wall is considered to be directly proportional to the drop in vapor pressure, and inversely proportional to the vapor resistance of the material along the path of flow. Thus, high vapor pressures within a building as compared with those on the outside will give a high potential rate of vapor flow. In general there will be a vapor pressure gradient through the wall from the high vapor pressure to the low vapor pressure side. The vapor density at various sections of the wall will depend upon vapor resistance throughout the walls. Thus, if the interior section has a· high resistance, and the exterior section a low resistance, to the passage of vapor, there will be a tendency toward a low vapor density within the wall ; whereas if reverse conditions are true there will be a tendency for high vapor densities 4 METHODS OF MOISTURE CONTROL within the wall. For any given condition of temperatures and vapor pressures on the two sides of a wall, the vapor pressure and temperature gradient through the wall will be governed by the type of material selected and its arrangement in the wall. If the vapor pressures within all sections of the wall are below the saturated vapor pressures which correspond to the temperatures for the given sections there can be no condensation within the wall. There is, however, no relation between the heat-transmitting and vapor-transmitting properties of materials. During recent years several changes have taken place which empha- size the importance of proper moisture control in residential construe- • tion. Part of these changes come under the heading of air conditioning, , and part under the heading of better building construction. The essen- tial change under air conditioning is that of maintaining higher relative humidities which are so often advocated for buildings located in cold climates. Those changes which may be classified under better building construction are the prevention of air leakage by the addition of weather strips and tighter building construction, the prevention of vapor passage through the exterior surface of the wall by better building papers, and the addition of insulation to the exterior walls of the building. For several years past the low relative humidities found in the average residence and public building located in cold climates have been a subject for much discussion, and a great deal has been said about the possible effect of these low relative humidities upon health, furniture, and the interior finishings of buildings. As a result of these discussions, together with the development of the science of air conditioning and air conditioning appliances, many devices have been installed in homes and public buildings for the express purpose of increasing the relative humidities carried in cold weather. :Many of these devices have been installed without control, and have often been of a type which operate to excess in the coldest weather. Excessive humidities thus are supplied to buildings equipped with these humidifiers during the period when there is the greatest possibility of condensation trouble. The loss of heat caused by the leakage of air through the exterior walls of buildings is another factor which has received a great deal of attention in recent years. Weather strips have been developed and added to doors and windows, and there has been a tendency toward tighter building construction. The reduction of air filtration through the exterior walls of a building due to better construction has resulted in higher inside relative humidities due to the build-up of moisture from normal living processes. The widespread use of insulation in recent years has probably been one of the most influential factors for better building construction in cold climates. The addition of insulation to a wall retards the flow of heat through the wall, changes the temperature gradient, and reduces the possibility of condensation on the warm side of the wall. THEORY OF VAPOR TRANSMISSION THROUGH MATERIALS 5 There is no logical argument against the use of moderate humidities for better conditions of comfort and health, nor against the addition of weather strips and other means of reducing the air leakage between the inside and outside of the building. Much less could any argument be applied against the addition of insulation in cold climates as a factor in saving heat. Each of these are essential elements for a well-con- structed building and a comfortable place in which to live. Under certain conditions each may be a factor in the condensation problem, but the elimination of any one would not be a logical solution. The conden- sation problem requires investigation on a practical and scientific basis-a practical basis to the extent of finding what may be expected from present practice in building construction and operation with prac- tical methods of eliminating trouble, and a scientific basis to the extent of investigating the laws which govern the transmission of vapor through materials, and of determining the vapor-transmission properties of various materials and combinations of materials which are practical for building construction. THEORY OF VAPOR TRANSMISSION THROUGH MATERIALS The theory relating to the transfer of vapor through materials has not been fully developed, altho it has often been assumed that the laws governing vapor transmission are similar in form to those governing the flow of heat through the walls of buildings, and that coefficients of vapor transmittance may be developed for materials, or combinations of materials, which may be applied in the same manner as the coefficients of heat transmission are applied. According to this theory the difference in vapor pressures between the two parts of a structure is the motive force which causes the flow of vapor, 'and the amount of vapor trans- mitted is directly proportional to 'the difference in vapor pressure, and inversely proportional to the vapor resistance of the material between the two parts of a wall. The overall vapor resistance of a built-up wall section is then equal to the sum of the vapor resistances of its component parts. A vapor transmittance coefficient would be defined as the quantity of vapor transmitted per unit of time per unit cross-sectional area per . unit difference in vapor pressure along the path of transmittance, and there would be coefficients for surfaces, air spaces, homogeneous ma- terials, and combinations of materials. This theory appears to satisfy many of the practical conditions, but there are specific differences between heat and vapor, and the methods by which they may be transmitted, which make it doubtful that it could be applied in all cases. Heat is a form of energy having no physical properties, and unless it is changed from sensible to latent, or vice versa, it will be transferred through a wall without change of state. Vapor is a substance with physical properties, and in the course of its transfer through a structure may change its state several times. Heat may be transmitted by radia- 6 METHODS OF MOISTURE CONTROL tion, conduction, and convection. Vapor may be transmitted by molec- ular diffusion and convection, and the condensed vapor may be trans- mitted by capillarity or other means. The method by which water vapor travels from one point to another depends partly upon the material through which it travels and partly upon the temperature along its path. In so far as its transfer through building construction is concerned, the materials through which it travels may be classified as air and solids. The conditions of air which seem to be of greatest importance are temperature, pressure, and air movement. The properties of the solids which seem to have the greatest significance are permeability to gases and power to absorb water vapor from the air. These two properties of solid materials, together with temperature, largely determine the state of the moisture as it travels through them. If the temperature of air is above the dew-point temperature of vapor the vapor may be transmitted through the air by turbulence and by molecular diffusion. The transfer of vapor by convection currents is similar to the transfer of heat by convection currents. The resistance of the flow of vapor through air by molecular diffusion is a function of density and temperature of the air through which it passes. For a given temperature and density of air vapor mixture the rate of vapor travel by molecular diffusion between two points may be considered as inversely proportional to the distance between the points. Thus under ordinary . conditions it may be substantially correct to assume that the laws govern- ing the rate of vapor travel through ·air by either convection currents or molecular diffusion are similar in form to those governing the flow of heat through air by conduction and convection. When vapor passes through a solid material there are at least three types of materials which seem to be of importance: first, those materials which are permeable to air or gas and which will not absorb water vapor; second, those materials which are impermeable to gas but which will absorb water vapor; and third, those materials which are permeable to gas and which will also absorb water vapor. Temperature is also an important factor. If a nonhygroscopic, homogeneous material is permeable to water vapor and its temperature at all parts is above the dew-point tempera- ture of the vapor in contact with those parts there will be no conden- sation. If the temperatures and static pressures on the two sides of the material are balanced, any vapor transmittance should be by molec- ular action, and the rate should be directly proportional to the vapor pressure drop along the path. If there is a temperature gradient along the path of vapor flow and the material is of a porous nature, then the rate of vapor travel may be increased by convection currents. This additional rate of transfer due to convection currents will lower the vapor pressure in the material, which would normally have been estab- THEORY OF VAPOR TRANSMISSION THROUGH MATERIALS 7 lished by the vapor pressure conditions on the two sides of the material. The temperature gradient through a wall establishes the maximum vapor pressures which can be carried at any point without condensation. If there are no convection currents possible, then condensation may occur at any point in the wall where the normal vapor pressure line as established by diffusion alone, crosses the maximum vapor pressure line. If, however, there is a temperature gradient through the wall and the porosity of the material is such that convection currents are possible, the rate of vapor passage through the material will be increased and the actual vapor pressure within the material may be far below that indicated by the normal vapor pressure line due to molecular diffusion alone. Under these conditions condensation will not occur in the porous material. This point is well illustrated in the average frame wall with fill insulation between the studs. The increased rate of vapor flow through the insulation due to convection currents reduces the actual vapor pressure line through that part of the wall below the line which would normally be established by molecular diffusion, and condensation does not occur in the loose material, but may under certain conditions occur on the surface of the sheathing, lining the cold side of the material. This fact is discussed later and has been proved by tests as well as by practical installations. If a material which is impermeable to water vapor is hygroscopic it will absorb water vapor and establish a moisture equilibrium when in contact with air vapor mixtures. The percentage of moisture absorbed by the material will depend somewhat on the temperature, but largely upon the relative humidity of the air with which it is in contact. The transmission of moisture through materials of this character will be by capillarity or some similar process and will not depend upon vapor pressure difference. Under certain conditions it appears that vapor may be absorbed from a mixture of low vapor pressures, transmitted through the material, and delivered to a mixture of high vapor pres- sures. If the temperature of either surface of the material is below the dew-point temperature of the vapor in contact with it, some of the vapor will be condensed. Free moisture or water may disturb the normal moisture equilibrium which would have been established by the water vapor in contact with the material and thus change the rate, and per- haps direction, of moisture flow through the material. If a material is hygroscopic and permeable to water vapor, it may transmit water vapor and free moisture, the net result depending upon several factors. For instance, a condition might be set up in which vapor would be transmitted in one direction due to vapor pressure dif- ferences and the permeability of the material, and moisture would be transmitted in the opposite direction due to the hygroscopic properties of the material and the relative percentages of vapor saturation on the two sides of the material. The rate of vapor transmission due to the 8 METHODS OF MOISTURE CONTROL hygroscopic properties of a material is very low and the net result for any practical wall would be in the direction of the vapor pressure drop. The above principles are illustrated in Figures 1, 2, and 3. In order to illustrate the above theory the walls· of Figures 1 to 5, inclusive, are assumed to be constructed of materials having ideal prop- erties for the theory under consideration. The wall of Figure 1 is built of a homogeneous, nonhygroscopic material which is permeable to water vapor. The material will transmit the vapor by diffusion but is not suf- ficiently porous to transmit the vapor by convection. The air in contact with the left-hand surface of the wall is at 70° F. and 40 per cent relative humidity, and that in contact with the right-hand side is at oo F. and 40 per cent relative humidity. The temperature gradient line through the wall is plotted to the temperature scale shown at the left, and the vapor pressure lines are plotted to the vapor pressure scale shown at the right. In each instance a reasonable drop in temperature and vapor pressure is indicated between the surface of the material and vapor in contact with this surface. Under the conditions given, the Jine A-B, plotted to the temperature scale, represents the temperature gradient through the wall. The curved line C-D, plotted to the vapor pressure scale, represents the maximum vapor pressure that it would be possible to carry at any point within the wall without condensation. In other words, any point on this curve gives the vapor pressure at saturation for the temperature of the material in the wall at that point. The straight line E-F is the normal vapor pressure gradient established in the wall due to the vapor pressures on each side of the wall. In this case the line E-F does not cross the curved line C-D, indicating that the temperatures at all points in the wall are above the dew-point temperature of vapor at corresponding points. Vapor should thus pass through the wall from E to F without condensing within the wall. The wall shown in Figure 2 is identical in every respect to that shown in Figure 1, and the conditions as to surrounding temperatures and vapor pressures are the same with the exception that the relative humidity of the air in contact with the left-hand surface of the wall has beet?- raised from 40 to 60 per cent, thus increasing the pressure of the vapor in contact with this surface. In this case the vapor pressure line E-F which would be established through the wall due to vapor pressure conditions in contact with the two surfaces, will be much steeper than that for Figure 1, and as shown will cross the limiting vapor pres- sure line C-D at some point X within the wall. This means that the temperature of the material at X will be below the normal dew-point temperature of the vapor at this point and condensation might be ex- pected. Since the temperature of the material in the wall at points im- mediately beyond X is such that the maximum vapor pressures are below those which would normally be established, the rate of vapor travel will be accelerated and new vapor pressure gradients will be estab- ln$it:le Atr Out .side Air ln3t(/eAir Ovl3t'de Air :limp 70° Temp. 0° li>mp. 70° l;mp. 0° RH. .t/0% R.H. 40% R.ti. 60% RH. 40% 80 .9 80 .9 ~ ~ ~ 70 .8 ~ TO .8 ~ ~ ~ ~ .? ~ 60 .? ~ .6 ~ ~ ~ 50 .6 ~ .5 G ~40 .5 ~ ~ ~ 30 -.;;:: "' I .4 I ~ ~ ~ . .J ~ ~ zo . .J ~ E § ~ ~ 10 .2 ~ 10 .2 ~ ~ ~ ~ 0 ./ ~ 0 .I ~ §- ~ 0 ~ 0 FIGURE 1. WALL BUILT OF NON- FIGURE 2. WALL BUILT OF NON- HYGROSCOPIC, HOMOGENEOUS HYGROSCOPIC, HOMOGENEOUS MA- MATERIAL PERMEABLE TERIAL PERMEABLE TO WATER VAPOR WATER VAPOR /n5tfleAir Wall Bvtlt of' a Ovfside Air 7emp.80° Jemp. 30° Rtt. 50% ltomo?eneoos R.H. 70% V.P.-0~1 11 dnd li~?/7JXOpic V.P.-O.!!f:' Ualedo/ 80 16 A .8 ~ !.: ' .? ~ ~ 70~ 15 .6 ~ ~ 0~ 14 ~ ~ . .,. c !:).. .50~ 1.3 .; ~ ~ ~ ~ .4 I 1 ~0·~ /2 ~ ~ JO~ II .J ~ ' ~ ~ .~:eo~ 10 .2 ~ !:: ~ ~ 10 ~ .I ~0~ 8 0 ~ C ond 0-J/q/)or P~s~on-~ ol ..5or/dct? cY 11&11 FIGURE 3. WALL BUILT OF HYGRO- SCOPIC, HOMOGENEOUS MATERIAL IMPERMEABLE TO WATER VAPOR TO 10 METHODS OF MOISTURE CONTROL lished through the wall. If the straight line E-Y is drawn tangent to tbe curve C-D at YJ it seems probable that E-Y would represent the vapor pressure gradient established in the wall up to the point Y. From this point on, the maximum vapor pressure gradient for each suc- cessive point would be established by a tangent to the curved line C-DJ and when the slope of this tangent is less than the slope of the line E-FJ the rate of vapor travel would be less than the normal rate. The rate at which moisture enters the wall would thus be increased and the rate at which it leaves would be decreased from the normal, and there would be an accumulation of moisture within the wall from the point Y to the outer surface. Since in this case the interior temperatures of the wall might be below 32° F., frost or ice would be formed. The wall of Figure 3 is assumed to be built of a homogeneous, hygro- scopic material which is impermeable to water vapor. The air on the left-hand side is at goo F. and SO per cent relative humidity, and that on the right-hand side is at 30° F. and 70 per cent relative humidity. The normal temperature gradient through the wall is shown by the straight line A-BJ plotted to the temperature scale at the left. The pressure of the water vapor in contact with the two surfaces is shown by the two short heavy lines C, D, and plotted to the vapor pressure scale on the right. This represents the vapor pressure difference on the two sides of the wall, but, since the material is impermeable to vapor, the vapor cannot pass through as such and therefore there is no connecting vapor pressure line between C and D. Materials of this nature absorb water directly from the vapor with which they are in contact. The percentage of water which they will absorb does not depend upon the absolute vapor pressure, but does depend on the per cent of saturation or, in this case, the relative humidity of the air. While wood is not an impermeable material, it is hygroscopic and its hygroscopic properties may be used to illustrate this principle. From data published by the United States Forest Products Lab- oratory, the moisture equilibrium content of wood in contact with goo F. and SO per cent relative humidity air would be approximately 9 per cent by weight and, when in contact with air at 30° F. and 70 per cent relative humidity, it would be slightly over 13 per cent by weight. Thus the line E-F shows the probable moisture gradient through the wall which would be established by the air conditions shown. From this it is evident that the moisture travel through this wall would be from right to left, or from the side in contact with low vapor pres- sure air to that in contact with high vapor pressure air. This is con- trary to the usual assumption that moisture travels through a wall in direct proportion to the vapor pressure drop between the two sides of the wall. It should be remembered, however, that in this case the vapor does not pass through the wall, but is condensed and absorbed by one surface of the wall, transferred through as absorbed moisture, and THEORY OF VAPOR TRANSMISSION THROUGH MATERIALS 11 evaporated from the other surface. The direction of vapor travel is determined by the moisture equilibrium conditions for the two sides of the material and not by the absolute vapor pressure differences. As previously pointed out the rate of moisture travel by this process would be very low, and while wood has been taken as a hygroscopic material to illustrate the principle, it should be remembered that when wood is used in a wall the sheets are not continuous and vapor will normally travel through the openings by diffusion or convection at a much more rapid rate than it could travel by the hygroscopic action of the wood. Furthermore, wood is not entirely impermeable to water vapor. /n5tl:le Air Temp 70" Rtf. 60"/o Oulside Air Inside Air Temp. oo Temp. 70° R.H. 40% R./i.60% Vqpor 8t7rnf!r (JO 4.: 10 ~ 60 ~.50 ~ I ~ ~ )0 i 20 ~ 10 () FIGURE 4. WALL BUILT OF NON- HYGROSCOPIC, HOMOGENEOUS MA- TERIAL PERMEABLE TO WATER VAPOR FIGURE 5. WALL BUILT OF NON- HYGROSCOPIC, HOMOGENEOUS MA- TERIAL PERMEABLE TO WATER VAPOR In most practical cases where the vapor pressures are different on the two sides of a building wall, a vapor pressure gradient will be estab- lished through the wall, and the vapor will travel in the direction of vapor pressure drop. The slope of the vapor pressure line for different sections of a wall will depend upon the resistance to vapor passage at various parts of the wall, and to the temperature of the material. Figures 4 and 5 represent two walls which are built of nonhygroscopic, homo- geneous materials which are permeable to water vapor but not sufficiently porous to allow convection currents to be set up within the wall. Both are subjected to air on the warm side at 70° F. and 60 per cent relative humidity, and air on the cold side at oo F. and 40 per cent relative 12 METHODS OF MOISTURE CONTROL humidity. The inside or warm surface of the wall shown in Figure 4 is lined with a material which has a high resistance to vapor penetration, and for the wall shown in Figure 5 this vapor-resisting material has been improperly applied to the outer or cold surface of the wall. The normal vapor pressure gradient which would be established by the vapor pres- sures on the two sides of the walls is shown by the solid lines E-F. Since ln~idf' Air Temp. 70"E R. ft. 40% I ~ t 80 70 6() !0 /0 ~ ~ () ~ -I() -Z() £ tJubiti~ Air Temp-20"1: R.ft. 40% L() .9 ~ ~ .8 ~ ~ .7 ~ .6 ~ .~ ~ ~ ' .4 ~ .3 ~ .z ct .I ~ 0 ~ FIGURE 6. FRAME WALL WITHOUT INSULATION SHOWING NORMAL TEMPERATURE AND VAPOR PRESSURE GRADIENT THROUGH WALL the wall of Figure 4 has a high vapor resistance on its warm surface the vapor pressure within the wall is comparatively low, but in the wall of Figure 5 with the vapor resistance on the cold surface the vapor pressure within the wall is high. If the temperatures throughout both walls were at all points above the dew-point temperatures of the vapor at these points in the vvall there would be no condensation. For the temperature conditions shown, however, the high vapor pressures within the wall of Figure 5 are above the maximum allowable vapor pressures throughout the greater part of the wall, and moisture will be formed. In the wall of Figure 4 the vapor pressures are below the condensation point throughout the wall and no vapor will condense. From the dis- cussion of the condition shown in Figure 2 it is evident that the solid line E-F of Figure 5 does not represent the true vapor pressure gradient established in the wall. The actual line cannot be above the curved line C-D. THEORY OF VAPOR TRANSMISSION THROUGH MATERIALS 13 Figures 6 and 7 represent two typical frame walls which are identical with the exception that fill insulation has been added to Wall No. 7. In both cases the air on the warm side is at 70° F. and 40 per cent relative humidity, and that on the cold side is at -20° F. with 40 per cent relative humidity. The temperature gradient through each wall is shown by the line A-B, and the maximum vapor pressure cor- /n~/o'~ Air Oubide Air Temp. 70°r Temp. -20°/:: RH. 40% Rli. 40% 80 70 /.0 60 .9 ~ 4.: .JO .8 ~ ~ 40 ·T ~ ~ JO .6 ~ 'S zo .~ ~ I 10 .-/- ~ 0 e .J ~ ~ -10• .2 ~ ~ -zo .I ~ ~ 0 FIGURE 7. FRAME WALL WITH INSULATION SHOW. ING NORMAL TEMPERATURE AND VAPOR PRESSURE GRADIENTS AS ESTABLISHED THROUGH WALL responding to the temperature at any point in the wall is shown by the line of dashes C-D. The probable vapor pressure gradient for each wall as established by the vapor pressure conditions on the two sides of the wall is shown by the line E-F. These lines are shown as curved lines through the central section of the wall as it is probable that the vapor is carried through this section partly by convection. The addition of insulation to Wall No. 7 reduced the heat loss through this wall to approximately one third of that for Wall No. 6. This caused a reduction in the temperature of the inner surface of the sh~athing from approximately 26° F. for Wall No. 6 to -6° F. for Wall No. 7, thus reducing the maximum vapor pressure that can be carried at the inner sheathing surface without condensation. In Wall No. 6 the normal vapor pressure line as established by the conditions on the two sides of the wall does not cross the maximum vapor pressure 14 METHODS OF MOISTURE CONTROL line as established by the temperature gradient through the wall, and under these conditions condensation will not occur on the warm surface of the sheathing, but may occur between the sheathing and siding. In the Wall of Figure 7 the point H on the inner surface of the sheathing represents the normal vapor pressure which would be required at this surface to give a uniform rate of vapor flow throughout the wall. The point G which is below H represents the maximum vapor pressure which can be carried at the sheathing line without condensation. The point G is established by the temperature of the sheathing and will be the governing point in determining the rate of vapor flow to and from the sheathing under the given air conditions. In order to have a uniform continuous rate of vapor flow through the wall the vapor pressure drop from inside air to sheathing line should be no greater than that between E and H, and the drop from sheathing line to outside air should be at least equal to that from H to F. Since the point G is below H, the vapor pressure drop, and therefore the vapo.r transmission from inside air to sheathing line, has been increased above the normal and that from sheathing line to outside air has been decreased below normal. Obviously there will be an accumulation of moisture at the sheathing line and the accumulation will continue until a correct relation .is established between the vapor pressure drop for the inner and outer sections of the wall. There are several possible methods for making this adjustment. First, if the air tempe.ratures and type of construc- tion are to remain the same, the relative humidity of ·the inside air may be reduced until the rate of vapor travel through the inner section is no greater than the possible rate through the outer section under the tem- perature conditions. Second, if the vapor pressures of the two sides of the walls and the type of construction are to remain the same, conditions may be improved by raising the air temperatures on the warm side of the wall. Third, if the temperatures and humidities are to remain the same, vapor resistance may be added to the warm section of the wall, or some type of ventilation may be applied to the cold section. The effectiveness of the various methods will be discussed in Part III of this bulletin. PART II-EQUIPJ\tiENT AND INSTRUMENTS TEST APPARATUS Cold room.-The large cold room in which test houses were built is 30 feet square, 25 feet high, interior dimensions. The photograph (Fig. 8) shows an interior view with six small test houses in place, and the drawing (Fig. 9) shows a vertical cross-sectional view with cooling coils and fan in basement below. The exterior walls, floor, and ceiling are constructed with 6-inch studs spaced 16 inches on center with the space between filled with mineral wool. The interior surfaces of the walls and ceiling are constructed with Vz-inch insulating board without additional finish. The exterior finish consists of a vapor-proof paper thoroly sealed over the studs, cemented at the joints, and covered with Vz-inch insulating board. The floor joists are sealed on top and bottom with vapor-proof paper and the space between is filled with mineral wool. The object of this construction was to prevent the leak- age of any outside air into the insulation and to allow some ventilation between the insulation and the cold interior air. Two doors are pro- vided in one wall, one of them 6 feet by 7 feet 6 inches, used to move large pieces of equipment into and out of the room, and the other a small service door, 3 feet 6 inches by 7 feet. The floor is provided with 9 openings, each 10 inches square, through which air ducts can be passed from the basement below to supply conditioned air to the test units. C ooZing equipment.-The large test room is cooled by circulating the air through direct expansion ammonia cooling coils. There are two coils, each 10 rows deep, 18 inches wide, and 34.5 inches high. Each coil is provided with face and by-pass dampers. The coils are placed in separate parallel compartments and provided with shut-off doors on each side so that either coil may be operated independently of the other. Figure 10 shows a side view, and Figure 11 a top view of the partially completed cooling unit assembly, taken during the course of construction. Figure 12 shows in the foreground the complete assembly of cooling coils with air ducts leading to test room. Air is circulated through the cooling coils by a 6,000 c.f.m. fan located on the discharge side of the unit as shown in Figures 9 and 10. The cooled air is carried to the top of the large test room and distributed through outlets in a horizontal duct at ceiling, from whence it circulates through the room and is then taken by a horizontal return duct, placed at the ceiling of the opposite side from the supply duct, and carried back to the cooling unit. The direct expansion cooling coils are supplied with ammonia from a 25-ton ammonia machine shown in the left foreground of Figure 13. Air conditioning equipment.-Conditioned air is supplied to each of the test units in the large cold room through openings provided in the floor directly under the test unit from an air conditioning plant located FIGURE 8. VIEW OF COLD ROOM SHOWING TEST HOUSES IN PLACE l..........r-- ,- Ret,rn Sliding Door COLD ROOM 61 0" x 7'- 6" 6,ii.:£Zi13:ii~ Test House "\f ~I~ Supply f---- OJ Fan :FIGURE 9. SECTIONAL VIEW THROUGH COLD ROOM AND COOLING UNIT I L FIGURE 10. INTERIOR VIEW OF COOLING UNIT WHILE UNDER CONSTRUCTION FIGURE 11. VIEW OF COOLING UNIT WHILE UNDER CONSTRUCTION FIGURE 12. VIEW OF COMPLETED COOLING UNIT AND ONE END OF CONTROL ROOM FIGURE 13. AMMONIA COMPRESSOR, CONTROL ROOM, AND DEFROSTING FAN FOR COOLING EQUIPMENT FIGURE 14. AIR CONDITIONING UNIT FOR TEST HOUSES FIGURE 15. SECTIONAL VIEW OF AIR CONDITIONING UNIT WITH DISTRIBU- TION DUCTS TO TEST HOUSES IN COLD ROOM 20 METHODS OF MOISTURE CONTROL in the basement, and shown in the photograph of Figure 14 and draw- ing of Figure 15. The air conditioning plant controls the humidity and distributes the air to the test houses from whence it is collected and returned to the air conditioner. The temperature of the air is controlled independently for each test house by electric heating elements which are placed in the supply duct for the particular unit, and controlled by a thermostat placed in the corresponding return air duct. Electrical heaters are arranged so that a part of the load may be carried by a con- stant supply and a part by a thermostatically controlled supply. FIGURE 16. CONTROL AND THERMOCOUPLE PANEL FOR TEST EQUIPMENT Control equiptnent.-The air temperature in the cold room is regu- lated partially by the operation of the ammonia compressor, but finally by face and by-pass dampers at the entrance to the cooling coils, which are controlled by a thermostat in the return air duct. The humidity for the test houses is controlled by a humidistat which is placed in the air discharge line from the air conditioning unit and which controls the water supply to a spray nozzle. The temperature of the air to the test units is controlled as previously described. All the control instru- ments are operated from a control panel shown in Figure 16 and located in the control room shown at the right in Figure 12. At the top of the INSTRUMENTS 21 control panel there are nine switches with pilot lights for the heating elements for the nine test stations located in the cold room. At the right central section is shown a switch and pilot light for the humidity control of the air conditioning unit, and at the central and left section of the panel are shown the controls for the motor which operate the face and by-pass dampers for the control of the air temperature in the cold room. On the lower part of the board there are switches for sixty thermocouples which are wired directly to corresponding panels on the inside walls of the large cold room. Thus the thermocouples may be connected directly from the test houses to the panels within the cold room without the necessity of running wires from the cold room to the control board for each set-up. INSTRUMENTS Moisture cont1'ol methods.-In practically all of the tests the air has been supplied to the test houses at a given temperature and humidity, and it has been important throughout the research program to accurately measure• and control the amount of water vapor in the air. In some cases it has been necessary to measure not only the moisture content of the inside air, but also that of the air within the walls of the test houses. In general the wet and dry bulb temperature method has been used as a standard of measuring the moisture content of the air, but due to the nature of the problem, the dew-point temperature method was better adapted for measuring the moisture conditions within the walls of the building. In all cases the humidity controls have been operated by some form of humidistat sensitive to the relative humidity of the air. In some instances it was necessary to build special test ap- paratus to get the desired measurements. Wet and dry bulb temperature apparatus.-The wet and dry bulb apparatus used consisted of two thermocouples each mounted in a glass tube with the beaded end extending through into a solid brass tip of the same diameter as the glass tube. The brass tip of the wet bulb thermometer served to give a good thermal contact between the thermo- couple and wicking material. Water was fed to the wick of the wet bulb thermometer from a constant level water reservoir, which in turn was kept filled by an inverted water bottle. In operation the wet and dry bulb part of the apparatus was placed inside an air duct with an air velocity of at least 800 feet per minute over the thermal elements. For accurate wet bulb readings it was found desirable either to wash or renew the wick every forty-eight hours. Special humidity controller.-The humidity control apparatus devised for the test consisted of a hygroscopic tape about 10 inches long sus- pended at each end and deflected at the center with a spring tension. The amount of deflection at the center was controlled by the length of the tape, which in turn was controlled by the initial setting of the instru- 22 METHODS OF MOISTURE CONTROL ment and the relative humidity of the air. The deflection of the center ·· section of the tape, which varied with the humidity in the air, operated a make-and-break mechanism which controlled the circuit to a relay through a vacuum tube. Dew-point temperature apparat~ts.-:Measuring the dew-point tem- peratures within the test walls proved to be a difficult problem. The wet and dry bulb method was considered, but it was impractical to get circulation of the air over the wet bulb thermometer without seriously disturbing the air within the test wall and thus making the readings of doubtful value. The dew-point temperature method was considered next, and the first apparatus built was one irt which a small stream of air from the test wall was conducted over a polished mirror, the tem- perature of which was controlled by a stream of carbon-dioxide gas. The temperature of the mirror was reduced until condensation appeared, and this temperature was taken as the dew-point temperature of the air. While this method proved fairly sensitive it still disturbed the air within the test walls and left some doubt as to the accuracy of the readings. The instrument as finally developed consisted of a glass tu.be about ~-inch internal diameter, covered on the exterior surface for a distance of approximately 4 inches with a silver and copper plate. A band of the plating was completely removed around the central section of the tube for a length of about Vs inch. The ends of the copper-plated section thus separated were connected through a vacuum tube circuit to a milliammeter, and a copper constantan thermocouple was soldered to one section of the copper-plated tube. In using this instrument the tube was anchored in the wall at the point where the dew-point tem- perature was to be measured, and its temperature was controlled by passing carbon-dioxide gas through the tube. As the temperature of the tube was gradually reduced the dew-point temperature of the sur- rounding vapor was reached and moisture was condensed on the exposed surface of the glass. The moisture deposit reduced the electrical re- sistance across the gap between the two sections of metal plating and the deflection of the ammeter was appreciably increased. " When the tube is washed in distilled water before the test, dew-point temperatures may be obtained which will check within less than 1 o F. as compared with wet and dry bulb temperature readings, providing the temperature of the tube is above 32° F. If the temperature of the tube is below 32° F., the conditions may be improved by dipping the tube into a 5 per cent salt solution and letting the salt precipitate and dry on the bare section of the tube. The thin salt coating thus formed does not materially affect the electrical resistance across the gap, but it does pre- vent the formation of frost at low temperatures. Since there is prac- tically no movement of the air around the tube when it is placed in a wall, it is necessary to cool it very gradually in order not to pass through the dew-point temperature during the cooling process. TEST UNITS 23 TEST UNITS Wall construction.-The test units were all of wood frame construc- tion, and the walls were built in sections so that they could be taken apart and inspected during a test without disturbing the continuity of the test. During the first tests 4-foot cubical test units were used. Later these were increased to 8 feet in height. Figures 17 and 18 show typical wall construction for all test units. Figure 17 is a sectional view of one of the smaller test units. The main frame is built of 3}1 x 3}1-inch wood framing members. The top and each wall section is built with 2 x 4 studs spaced 16 inches on the centers with metal lath and plaster on the inside surface and 8-inch pine shiplap, building paper, and 6-inch redwood siding on the outside surface. Figure 18 shows the type of wall construction used. The sheathing, building paper, and lap siding are applied to nai,ling strips which fit snugly into grooves made in the out- side surface of the studs. The assembled outside wall section is applied with screws and may be removed without disturbing any other part of the test set-up. Small removable panels are provided in th~ sheathing and may be removed for the purpose of determining the accumulation of moisture or frost in the sheathing during a test period. In some of the test panels removable aluminum sheets were applied to the surface of the sheathing between the stud sections. The plaster was applied to the metal lath in three coats : scratch, brown, and finish coats, making a total thickness of approximately % of an inch. Various types of vapor proof paper were placed under the metal lath and different surface fin- ishes were used on the plaster, as described in the test results. Figure 19 shows. the assembled view of an 8-foot high test house. The details of wall construction are the same as those shown in Figures 17 and 18, with the exception that the removable sheathing panels shown in Figure 18 include sections of the siding as well as the sheathing. Attic construction.-In the preliminary investigation covering the ventilation of attics three of the small-sized test units were rebuilt to provide attic spaces over the ceilings. The roof construction, as shown in Figure 20, was built with removable units similar to the outside sec- tions of the walls for the small test houses. In some instances aluminum condensation panels were provided for the space on the under side of the roof boards between the rafters. The ceilings dividing the attic spaces from the rooms below were constructed with metal lath and plaster applied to 2 x 4 joists spaced 16 inches on center. In most cases 3% inches of mineral wool was applied between the ceiling joists with no vapor barrier underneath. This provided attic spaces which were typical, without vapor barriers. Three types of attics were used : the first, completely enclosed, with- out ventilation; the second, provided with an adjustable opening in each gable, as shown in Figure 20 ; and the third provided with mechanical J~~ (X) I() FIGURE 17. SECTIONAL VIEW OF TEST HOUSES SHOWING DETAILS OF CONSTRUCTION 112" x 6" Redwood Siding Sui lding Paper 1" x a" Shiplap Three center strips of siding are removable for weighing 12" horizontal strip of building paper removable 2" x 2" Angle Iron FIGURE 18. TYPE OF WALL CONSTRUCTION USED TO FACILITATE INSPECTION TEST UNITS 25 ventilation by which the amount of air supplied to . the attic could be measured. An observation opening was provided in one of the gables of each test house for inspection of the interior surfaces during a test period. FIGURE 19. TEST HOUSE USED FOR WET PLAS· TER TEST AND ALSO SUBSEQUENT TESTS ON WALL VENTILATION Figure 21 shows the three types of test attics assembled in the cold room, together with the pump and meter used to furnish the air to the mechanically ventilated attic. Vapor barrier test apparatus.-Two different test methods were de- vised for evaluating vapor barriers. In the first, the barriers were built into the walls of a test house, and in the second, the vapor-resisting properties of the barriers were compared by the use of a special test apparatus. The general method of constructing the walls for the first method of test is shown in Figure 18. The barriers were applied either under the metal lath or on the surface of the plaster, depending upon the type of barrier used. Small removable sheathing panels were pro- FIGURE 20. CONSTRUCTION DETAILS OF ATTIC FOR ATTIC VENTILATION TESTS FIGURE 21. VIEW OF SET-UP FOR VENTILATED AND UNVENTILATED ATTICS FIGURE 22. LINE DRAWING SHOWING DETAILS OF CONSTRUCTION FOR SPECIAL VAPOR BARRIER TEST APPARATUS FIGURE 23. VIEW OF APPARATUS FOR MAKING TESTS ON VAPOR BARRIERS 28 METHODS OF MOISTURE CONTROL vided, as shown, for the purpose of measuring the accumulation of frost on the surface of, or moisture within, the sheathing. In addition to the removable sheathing panels, sheets of aluminum were applied to some of the test walls to completely cover the surface of the sheathing between a given stud space. The aluminum panels were removed and weighed at various periods during the test, ana the rate of frost accumulation on the surface was used as a measure of the vapor transmitted through the barrier for the given test conditions. In using this test method it was necessary to maintain outside air temperatures low enough to keep the interior surface of the sheathing below 32° F. The aluminum panel prevented some of the natural ventilation of the wall to the exterior, and in some cases the accumulation of frost on the aluminum panel was slightly greater than that on the removable sheathing panel for the same type of construction. In many tests, however, these differences were negligible. For the second test method, the four walls of a small test house were each divided into four parts and special test panels were constructed to fit each of these divisions. The construction of one of these units is shown in the drawing of Figure 22 and the photograph of Figure 23. Each of the special test panels was made up of a mineral wool insulatir1g pad 2}i inches thick held in place by a coarse-mesh wire screen on each side of the pad. The insulating pads were carefully made up to get uniform density and thickness of material for all of the panels. The pads were lined on the warm side with a barrier to be tested and on the cold side with a light weight sheet of aluminum. A clearance of approxi- mately 3/16 inch was allowed on each side of the insulation. The vapor barrier and aluminum panel were both sealed into the test frame to pre- vent leakage of water vapor, and the assembled test panel was sealed into an opening provided in a wall of the test house. PART III-TEST RESULTS VAPOR BARRIERS Test procedure.-From the nature of the vapor condensation problem in buildings it is evident that one of the best solutions in so far as con- struction of the building is concerned is to use a type of construction which will prevent vapor from penetrating the warm side of the wall. Certain types of building materials are known to have a high resistance, and others a low resistance, to the passage of vapor. There is, however, no standard method of evaluating the vapor-resisting properties of build- ing materials, and in most cases very little information is available as to the relative merits of these materials when applied to the walls of a building. All materials have some resistance to the passage of vapor, but in this bulletin the term vapor barrier has been applied to those ma- terials which have a relatively high resistance to the passage of vapor. In studying the vapor-resisting properties of materials, two different test methods were used. In the first method, the materials were used in the construction of a full scale wall and in the second, they were placed in a special test panel. In all cases where the barriers were ap- plied to a standard test wall, the wall was built of 2 x 4 studs spaced 16 inches on centers, finished on the inside with metal lath and plaster and on the outside with 8-inch Ponderosa pine shiplap sheathing, asphalt- saturated building paper, and 6-inch redwood siding covered with three coats of white lead paint. The finished plaster was approximately ~ inch thick and applied in three coats : a scratch coat, a brown coat, and a finish coat. The scratch and brown coats were made by mixing one part of Gypsum plaster with two parts of plaster sand, and the finish coat was a thin coat of white lime plaster. Two types of barriers were investigated by both test methods ; first, those barriers which were built in the form of a self-supporting mem- brane, and second, those barriers such as paints which were used as a surface coating. When the membrane barriers were tested in combina- tion with the complete wall they were applied in a continuous sheet underneath the metal lath. When they were tested with the special test apparatus they were applied as a continuous sheet over the warm surface of the special test panel. . When the surface coating type of barrier was tested by either method it was applied as a surface finish over the plaster. In the standard wall it was applied over the interior surface of the wall, and in the special test apparatus it was applied over the warm surface of specially prepared plaster samples. The plaster samples were all built after the same specifications for both types of tests. Barriers tested in combination with zvalls.-The results from tests for several types of membrane barriers placed underneath the metal lath of the test walls are shown in Table I, and the results for several TABLE I CONDENSATION ON INNER SURFACE OF SHEATHING FOR DIFFERENT TYPES OF MEMBRANE VAPOR BARRIERS PLACED BETWEEN METAL LATH AND STUDS Inside Outside Surface Condensation on Sheathing, Grams* Vapor Barriers Air Temperature per Square Foot per 24 Hours Temperature, of Sheathing, Degrees Degrees Test Test Average of Fahrenheit Fahrenheit number results tests None -19.5 -0.3 4 2.16 -19.5 -1.5 4 2.26 -19.3 1.3 8 2.02 2.15 None 9.9 23.0 7 1.02 10.1 22.8 20 1.55 9.9 22.5 7 1.65 1.41 rdges lapped -19.5 -4.1 4 0.09 Asphalt impregnated and surface coated glossy l not sealed -19.5 -2.0 8 0.05 0.07 sheathing paper. Weight of paper tested, 51.9 9.9 20.5 7 0.00 0.00 pounds per roll of 500 square feet Edges lapped -19.5 -3.5 4 0.02 0.02 and sealed 9.9 19.7 7 0.00 0.00 30-30-30 duplex. 2 sheets 30 lb. kraft paper 1 { Edges lapped -19.5 -3.2 4 0.31 -19.3 0.7 8 0.20 0.25 cemented together with layer of asphalt, equal not sealed 9.9 20.9 7 0.00 0.00 in weight to one layer of paper. Weight of paper tested, 16.1 pounds per roll of 500 square Edges lapped -19.5 -3.0 4 0.19 0.19 feet and sealed 9.9 20.7 7 0.00 0.00 Rag felt paper saturated with asphalt. Weight of Edges lapped -19.5 -2.2 4 0.52 0.52 paper 73.0 pounds per roll of 500 square feet not sealed 9.9 21.3 7 0.18 0.18 Duplex crepe paper. Weight of paper tested, Edges lapped -19.5 -1.8 4 0.09 0.09 40.7 pounds per 500 square feet and sealed 9.9 21.6 7 0.00 0.00 NoTE: Inside air conditions 70° F., 40 per cent relative humidity. All walls constructed with 2 x 4 studs spaced 16 inches on centers; metal lath and plaster on interior surface; 8-inch Ponderosa pine shiplap, building paper, and 6-inch redwood siding on outside; 3% inches of mineral wood be- tween studs; vapor barriers as specified. * 1 gram=.0022 pound. VAPOR BARRIERS 31 types of barriers in the form of surface coatings placed on the inner surface of the plaster for the test walls are shown in Table II. The general description of the membrane barriers is given in Table I, and the specifications for the surface coatings are given as an addenda for Table II. The results for these tests are given as the rate of moisture condensation on the warm side of the sheathing in grams per square foot per twenty-four hours. From these results it is evident that the vapor passes ·rather freely through the untreated plaster of the ordinary TABLE II CONDENSATION ON INNER SURFACE OF SHEATHING FOR TEST WALLS WITH DIFFERENT TYPES OF FINISHES ON INTERIOR SURFACES OF PLASTER Inside Outside Surface Condensation AirTem· Tempera- on Sheathing, Finishes on Inside Surface No. of perature, ture of Grams* per of Plaster Tests Degrees Sheathing, Square Foot Fahrenheit Degrees per 24 Hours Fahrenheit Unfinished 3 -19.5 -0.2 2.15 3 +10.0 22.8 1.40 2 coats seal coat paintb 1 -19.5 0.8 0.20 1 +10.0 23.6 0.00 2 coats white flat painta -19.5 -1.5 0.24 +10.0 20.6 0.00 2 coats aluminum paintc 1 -19.5 0.5 0.25 1 +10.0 23.6 0.00 1 coat asphalt applied hot 1 -19.5 1.1 0.13 1 +10.0 23.3 0.00 1 coat seal coat paintb } { l -19.5 1.8 0.23 2 coats white flat paint• +10.0 22.5 0.00 1 coat seal coat paint } { ~ -19.5 3.9 0.16 2 coats aluminum paint +10.0 24.3 0.00 coat glue sized with } { l -19.5 -0.9 2.11 plain wall paper +10.0 22.2 1.11 coat glue size with dull } { l surface-treated can- -19.5 0.4 0.64 vas wall covering +10.0 22.3 0.10 1 coat glue size with glossy } { l -19.5 0.7 0.47 surface-treated canvas wall covering +10.0 21.7 0.01 1 coat glue sized with } { i -19.5 -0.2 0.32 duplex crepe paper +10.0 21.8 0.00 NoTE.-Inside air conditions 70° F. and 40 per cent relative humidity .. All. walls con- structed with 2 x 4 studs spaced 16 inches on centers; metal lath and \)laster on mtenor surface; 8-inch Ponderosa pine shiplap, building pap.er, and 6-i!lch redwood sidmg on outside; 3 o/8 inches of mineral wool between studs; surface fimsh as specified. * 1 gram=.0022 pound. See Addenda for Table II on page 32. 32 METHODS OF MOISTURE CONTROL ADDENDA FOR TABLE II a White flat interior paint specifications: Pigment ........................................................................................................................................................... . Vehicle .............................................................................................................................................................. . Pigme~t c.ompositiol! . T1tan1um calc1um p1gment .......................................................................................................... . Calci urn carbonate ................................................................................................................................ . Comp~siti-;m of ti~anium calcium pigment T1tan1um ox1de ...................................................................................................................................... . Calcium sulphate .......................................................................................... , ........................................ . Vehicle composition Vegetable oil and resin .................................................................................................................. .. Volatile thinner and drier ............................................................................................................. . bSeal coat specifications for priming coat on plaster surfaces: Pigment ........................................................................................................................................................... . Vehicle ................................................................................................................................................... . Pigme!lt <:ompositi~n . T1tamum calc1um p1gment .................................................................................................. . Comp~siti-;m of tit.anium calcium pigment f1tamum ox1de ..................................................................................................................................... .. Calcium sulphate ........................................................................................................................... . Vehicle composition Resins ................................................................................................................................................................ .. Vegetable oil ........................................................................................................................................... .. Drier and thinner ............................................................................................. ., .................................. .. e AI umin urn paint specifications: Aluminum paste ..................................................................................................................................... .. Vehicle ............................................................................................................................................................. .. Composition of paste Pure aluminum powder ................................................................................................................. .. Volatile liquid ....................................................................................................................................... . Vehicle Water-resisting spar varnish IIGJue size specifications: A cold water size containing animal glue Mixture: 11 pints of water to 1 pound of size N OTE.-Outside white lead paint specifications Pigment by weight ................................................................................................... . Vehicle by weight ..................................................................................................... . Composition of pigment A pure white lead paste Composition of vehicle Linseed oil .................................................................................................................................................. . Drier .................................................................................................................................................................... . 62.3 per cent 37.7 per cent 78.0 per cent 22.0 per cent 30.0 per cent 70.0 per cent 34.3 per cent 65.7 per cent 40.0 per cent 60.0 per cent 100.0 per cent 30.0 per cent 70.0 per cent 16.7 per cent 28.8 per cent 54.5 per cent 1.75 lb. 7.40 lb. or 1 gal. 65.0 per cent 35.0 per cent 70.0 per cent 30.0 per cent 90.0 per cent 10.0 per cent wall, but that there are many membrane barriers, and also surface coat- ings, which are effective in preventing the passage of vapor through to the inner section of the wall. TABLE III VAPOR TRANSMISSION FOR DIFFERENT MATERIALS BY SPECIAL TEST METHOD Inside Air, 70° F. and 40 Per Cent Relative Humidity. Outside Air, -10° F. Lab. No. 1 21 22 57 58 64 65 Type of Material Insulating pad only .......................................................... . Asphalt impregnated and surface coated glossy sheathing paper 51.9 pounds per 500 square feet... ............ 69.0 pounds per 500 square feet.. 31.0 pounds per 500 square feet 48.6 pounds per 500 square feet 34.0 pounds per 500 square feet. 47.5 pounds per 500 square feet 35.2 pounds per 500 square feet * 1 gram=.0022 pound. No. Tests 17 8 4 3 2 1 1 2 Condensation, Grams* per Square Foot per 24 Hours Maximum Minimum Average 13.4 10.7 12.2 0.22 0.04 0.14 0.30 0.08 0.15 0.21 .011 0.15 0.14 0.12 0.13 0.12 0.25 0.18 0.14 0.16 VAPOR BARRIERS 33 TABLE 111-C ontinued Lab. Type of Material No. Condensation, Grams* per Square Foot per 24 Hours No. Tests Maximum Minimum Average Duplex paper 30-30-30 2 16.1 pounds per 500 square feet .............. 4 0.34 0.11 0.22 17 17.2 pounds per 500 square feet .............. 6 0.42 0.14 0.22 18 16.5 pounds per 500 square feet .............. 6 0.50 0.17 0.38 Duplex paper 30-60-30 16 21.9 pounds per 500 square feet .............. 6 0.40 0.14 0.22 Duplex paper 12 40.7 pounds per 500 square feet.. ............ 7 0.30 0.02 0.13 52 29.6 pounds per 500 square feet .............. 3 0.17 0.14 0.15 53 19.5 pounds per 500 square feet ............. 2 0.24 0.19 0.22 67 20.7 pounds per 500 square feet .............. 2 0.19 0.17 0.18 60 20.3 pounds per 500 square feet (1 side metal coated) ........................ 2 0.07 0.08 0.08 62 25.8 pounds per 500 square feet (both sides metal coated) ............... 2 0.11 0.03 0.07 Duplex paper reinforced 8 29.6 pounds per 500 square feet .............. 5 0.28 0.06 0.19 51 29.6 pounds per 500 square feet .............. 2 0.26 0.19 0.22 9 32.0 pounds per 500 square feet .............. 6 0.32 0.09 0.17 13 63.0 pounds per 500 square feet (1 side metal coated) ........................ 3 0.13 0.04 0.08 Insulation back-up paper 4 15.3 pounds per 500 square feet .............. 6 0.46 0.11 0.29 24 Approximately 15.0 lbs. per 500 sq. ft. 2 0.35 0.16 0.25 25 Approximately 15.0 lbs. per 500 sq. ft. 3 0.27 0.13 0.19 20 Approximately 15.0 lbs. per 500 sq. ft. •2 0.32 0.24 0.28 26 Approximately 15.0 lbs. per 500 sq. ft. 2 0.33 0.27 0.30 68 13.9 pounds per 500 square feet .............. 1 0.23 69 15.6 pounds per 500 square feet.. ............ 1 0.23 23 14.3 pounds per 500 square feet (1 side metal coated) ........................ 4 0.23 0.06 0.12 59 49.5 pounds per 500 square feet (1 side metal coated) ........................ 1 0.06 19 Paraffin-coated kraft paper, 13.2 pounds per 500 square feet 3 0.32 0.27 0.29 3 Asphalt saturated rag felt building paper, 73.0 pounds per 500 sq. ft. 6 2.39 1.04 1.70 63 Red rosin paper, 24.8 pounds per 500 square feet.. .............................. : ........ 1 9.36 56 50-pound brown kraft paper ................... , .... 1 9.44 40-pound black kraft paper, 6.94 pounds per 500 square feet ........... 4 10.83 9.66 10.22 40-pound black kraft paper, 1 side coated with asphalt, 24.9 pounds per 500 square feet ................................ 4 0.29 0.18 0.22 ~-inch plaster on wood lath, ~-inch grounds ·································· 2.21 ~-inch plaster on metal lath, ~-inch grounds .................................... 4.19 ~-inch plaster on metal lath, 20 3.79 ~-inch grounds ....................................... 4.46 3.44 l-inch plaster on metal lath, l-inch grounds .......................................... 1 3.24 13/16-inch clear pine .......................................... 1 0.43 * 1 gram=.0022 pound. 34 METHODS OF MOISTURE CONTROL Barriers tested by spec-ial test apparatus.-Both the membrane and surface coating types of barriers were tested by the special test appara- tus as previously described. The test results for the membrane barriers are given in Table III and those for the surface coating barriers are given in Table IV. The general description of the membrane barriers is included in Table III, and that for the surface coatings is given in the Addenda for Table IV. In all tests the results are in grams of moisture condensed per square foot per 24 hours on the aluminum panel which was used to line the cold side of the test panel. As will be noted, the relative values for two barriers tested by two different methods are the same. However, the absolute test values for a barrier tested by the two different methods may not be the same. The reason for this difference in test values for the same barrier tested by two different methods is that the net vapor pressure drop across the test material is different in the two methods, even tho the air conditions are the same for both sides of the test specimen. When a vapor barrier is TABLE IV RELATIVE VAPOR TRANSMISSION FOR DIFFERENT SURFACE COATINGS B.Y SPECIAL TEST METHOD All Surface Finish Applied with Brush on Surface of ~ Inch of Plaster. Inside Air, 70°F., 40 Per Cent Relative Humidity. Outside Air, -10° F. Surface Coating No. General Description None ~-inch plaster on metal lath (3.79 grams) ...... V-1 P-1 V-2 P-2 V-5 P-5 V-6 P-6 V-7 P-7 V-8 P-8 V-10 P-10 V-13 P-13 Linseed rosin varnish .......................................................... .. White flat interior paint... .................................................... .. Rosin Perrilla oil .................................................................... .. White flat interior paint... .............................................. .. Varnish ............................................................................................... . White seal coat paint ........................................................... . Phenol formaldehyde varnish ..................................... .. High grade aluminum paint ......................................... .. Varnish (glycerol phthalate) ..................................... .. White enamel .................................................................... . Bronzing liquid .......................................................................... . Aluminum bronze ....................................................................... . 90 per cent linseed oil and 10 per cent drier and thinner ............................................................................. . A linseed oil paint used on both interior and exterior surfaces ........................................................... . 50 per. cent lin;;~ed-tung oil and 50 per cent !llmeral. sp1r~ts .......... ; ............... ; ......................................... . Semt-gloss mtenor wh1te pamt ............................... .. * 1 gram=.0022 pound. Condensation-Grams* per Squar.e Foot per 24 Hours Two Three coats coats 0.82 0.83 0.45 0.58 0.63 0.63 0.55 0.61 0.77 0.49 1.16 1.04 0.93 1.05 1.19 0.80 0.45 0.62 0.33 0.44 0.44 0.41 0.41 0.34 0.54 0.39 0.84 0.45 0.92 0.50 0.97 0.60 See Addenda for Table IV on page 35. No. from Table IV V-1 V-2 V-5 VAPOR BARRIERS ADDENDA FOR TABLE IV DESCRIPTION OF PAINTS AND VEHICLES Vehicle' No. from Table IV Paint 35 Per Cent Per Cent Linseed rosin varnish Linseed oil ......................................... . N:~~r:~::::~~~~:~~~::::::::::::::::::::::::::::::::: Rosin Perrilla oil Perrilla oil ......................................... . ~;fe~ra~::::~~~~:~~~::::::::::::::::::::::::::::::::: Varnish Rosin ..................................................... . Kr~~e~':t0s~ir~~~ :::::::::::::::::::::::::::::: Drier ..................................................... . 60.4 3.6 35.0 1.0 33.0 12.0 54.0 1.0 16.7 28.8 53.5 1.0 P-1 P-2 P-5 White flat interior paint Pigment ............................................... . Vehicle (V-1) ............................. . Pigment composition: frl~~~l~~e cii;~id~····::::::::::::::::::::: Magnesium silicate White flat Interior paint Pigment ..................................... . Vehicle (V-2) .................... . Pigment composition: Titanium calcium dioxide .. Zinc oxide ...................................... . Magnesium silicate ................. . White seal coat paint Pigment ............................................ . Vehicle (V-5) ............................. . Pig~en~ comp