MOISTURE CONTENT OF HOUSE FRAMING Dr. William C. Hopkins

MOISTURE CONTENT OF HOUSE FRAMING

Dr. William C. Hopkins

The dimensional changes in wood which occur as a result of changes in moisture content areprobably the cause of more dissatisfaction with wood than any other single characteristic or defect ofthe wood. Items which are free to move, such as windows, doors, and drawers swell and become tightor shrink and become loose as the relative humidity of the surrounding atmosphere rises or falls. Woodwhich is fixed in place such as studs, joists, flooring, paneling, and door and window casings mayshrink and swell sufficiently to cause opening of the joints, cracking of walls and splitting or bucklingof paneling.

It is theoretically possible to overcome or prevent most of this shrinking and swelling by the addi-tion to the wood of certain chemicals (1). However, the materials which will accomplish such dimen-sional stabilization are expensive and at present their use is limited to such high value items as gunstocks, wood-carvings, etc. The only practical method of minimizing changes in dimension is to havethe wood at a moisture content, when installed, which closely approximates the average which willprevail in the particular location. The fact that this method can be very effective is illustrated by thebehavior of the paneling in some of the offices in the School of Forestry at Louisiana State University.The moisture content of this material on these walls fluctuates between 8 and 11 percent. There isone wall, however, which has a steam radiator against it. The moisture content of the paneling onthis wall ranges from a low of 2 to .3 percent during the winter to a high of 11 percent during the summer.The paneling on all walls was installed at a moisture content of 9 to 10 percent. The three wallsaway from the heat have never given any trouble (Figure I), whereas the wall above the heater hasbeen subjected to so much drying that it has shrunk badly, resulting in cracking and splitting which area constant eyesore (Figure II). Admittedly the source of the difficulty in this case is in the designwhich subjected the wood to severe conditions, but it serves to illustrate the point.

The United States Forest Products Laboratory has studied the moisture content of wood in dwellingsin representative cities throughout the nation and has developed a series of average moisture contentvalues for various regions (2). This information is of great value but it is believed that more precisevalues would be better in cases where it is desirable to keep shrinking and swelling to an absolute mini-mum. There is a need for information on the moisture content of wood in restricted areas and inspecific building components. A study designed to obtain this type of information in the Baton Rouge,Louisiana area was initiated by the School of Forestry at Louisiana State University several years ago.The one-year results on three types of buildings have been previously reported (3). This report willdeal with the results over a period of 26 months in a single residence.

STUDY AREA

The data which are reported herein were collected on structural components of a new residentialbuilding belonging to the Department of Horticulture of the Louisiana Agriculture Experiment Station.This building is located on the campus of Louisiana State University at Baton Rouge, Louisiana. Aver-age annual rainfall here is 58 inches, which is reasonably well distributed throughout the year. Thelowest rainfall occurs in October, the highest in July. The frost-free season averages 270 days in length,beginning about March 1, ordinarily. The area is characterized by a generally high relative humiditywhich is the result of the proximity of numerous large bodies of water and the Gulf of Mexico.

Temperatures at Baton Rouge are generally mild during the winter and warm during the summer,the monthly averages ranging from a low of 52.0 degrees F. in January to a high of 81.1 degrees F.during July. The high in 1961 was 93 degrees F. ; the low was 21 degrees F. A total of 273 dayselapsed between the last freeze in the spring and the first freeze in the fall (4).

-85-

•

FIGURE I. HACKBERRY PANELING TWELVE FEET FROM HEATER

FIGURE II. HACKBERRY PANELING ADJACENT TO HEATER

The building involved in this study is a typical slab-on-ground brick veneer residence with apolyethylene vapor barrier between the gravel fill and the slab. All framing material is southern pineand the bottom plate is separated from the slab by one thickness of heavy, tar-impregnated buildingpaper. The walls are standard dry-wall construction and do not contain a vapor barrier. This residencewas completed in July, 1959 and has been occupied continuously since that time. It is not air conditioned.

METHODS OF STUDY

Moisture content values were determined with a Delmhorst Model RC-1 moisture meter, whichwas fitted with a special cable with banana plugs on one end. At each location where knowledge ofthe moisture content was desired, two stainless steel screws were embedded in the wood to a depth of1/4-inch. The screws were placed a distance of one inch apart along the grain, insulated wires wereattached to these screws with spring connectors, and were led from the screws to a panel where theywere connected to a pair of jacks. Each pair of jacks was numbered according to the location of thescrews.

Readings were taken every other day throughout the year at the same time each day. The samemoisture meter was used throughout the study. At the beginning of the investigation and every threemonths thereafter, this meter was calibrated against oven-dried samples. The readings were adjustedaccordingly and recorded as the true moisture content.

The original and all subsequent calibrations were accomplished by the use of 10 sets of end-matched samples of Southern pine 2 by 4 by 12 inches in size. The ends of these samples were coatedwith an end sealer to slow down any end-wise adsorption of moisture as much as possible. One pieceof each set of the samples was fitted with screws and wired to a pair of jacks located outside of theconstant temperature-humidity cabinet. The other was weighed, and the original moisture content andcalculated oven-dry weight were computed from a third end-matched sample 1 by 2 by 4 inches insize. The two larger samples were then placed in the humidity cabinet and the controls set for the de-sired relative humidity. A constant dry-bulb temperature of 90

o F. was used. Each day the non-wired

sample was removed from the cabinet and weighed, and moisture content was calculated. At the sametime a reading was taken on the matched sample with the moisture meter. When the weight of thenon-wired sample stabilized, the relative humidity was changed. The moisture meter readings werethen plotted over the calculated moisture content values of the matched samples, and the balancedcurve was used as a basis for correcting the moisture meter readings.

DISCUSSION OF RESULTS

The monthly average moisture content values at three locations in this building are shown inFigure III. This graph reveals some interesting facts and leads to some interesting conclusions. In thefirst place the total range of moisture content over a 26-month period is only six percent. When welook at individual locations we find that the range in the bottom plate, which is separated from theconcrete slab by only one thickness of building paper, is only five percent. Although the high shownon the graph for the bottom plate is 14 percent, we should really consider 13 percent as the high be-cause the 14 percent reading was obtained only one month after construction was finished and the woodand slab were still drying out. The range in the studs and rafters is even less than in the bottom plate.Likewise, the average moisture content is less, with the rafters having the lowest moisture content.

A glance at the curves will reveal that the seasonal low moisture content is during the wintermonths and the seasonal high is during the summer months. This coincides exactly with the pattern ofthe relative humidity and rainfall in the Baton Rouge area. Heating during the winter months probablyhas some effect on the moisture content, too.

-86-

A 00

• • •0 0 .0

cdA r•4

BOT TOM PLA TE

__ STUD

RAFTER■III •

VA 111■11111 111110/11■

NM1111111111111

AIME\\1311111111111WA MINIM •

.1111111111011111111111111111111111111

IIIV

1111.11110111011M111.11

FIGURE III. MO NT I1TX AVERAGE MOISTURE CONTENT AT THREE LOCATIONS

9

8 8

7

•

ao4-10,

•4.3

A 8

" 1 1Mid-point of StudsBottom Plates Rafters

140

120

100

z60

A 140

20

MOISTURE CONTENT IN PERCENT

FIRE IV. FREQUENCY OF OCCURRENCE OF VARIOUS MOISTURE CONTENTS, THREE LOCATIONS

Number of days after installation of screws

FIGURE V. COMPARISON OF MC READINGS FROM PINS AND SCREWS

It is also interesting to note that the fluctuation of moisture content in the rafters is less than ineither the studs or the bottom plate. Ventilation in the attic is better than in the wall, and we havefound, in another study, that the temperature of the rafters always exceeds that of both the studs and the

bottom plate.

The number of days a particular moisture content occurred at a particular location is shown inFigure IV. The maximum moisture content ever attained in any location was 15 percent. It reachedthis level in the bottom plates only 24 days out of 316 or 7.6 percent of the time. At the same locationthe moisture content ranged from 9 to 12 percent 248 days or 78.5 percent of the time. At the mid-point of the length of the studs the maximum of 15 percent moisture content was reached only 5 timesor 1.6 percent of the time. At this location a moisture content between 8 and 11 percent was found 277times or 87.6 percent of the time. In the rafters the moisture content was between 8 and 10 percent94.0 percent of the time. It attained the maximum of 15 percent only 3 days or 0.95 percent of the

time.

More important, perhaps, than the data obtained on this one house is the experience gained inattempting to measure the moisture content of a member in place. On the basis of this experience Iam inclined toward the belief that some procedure such as that used by Bois (4) will give more preciseresults. One of the shortcomings of the meter in this study was that it measured the moisture contentat the depth at which the maximum moisture content occurred. Thus, if relative humidity was risingor if there was condensation on the surface of the wood, we measured moisture content at the surface.During periods of drying we measured moisture content at a depth of 1/4-inch. This difficulty can beovercome by the use of the insulated pins now available, but then the measurement is valid only forthe depth at the points of the pins or screws, and, depending upon the thickness of the lumber, thismay or may not be representative of the moisture content of the piece as a whole.

It was found also that there are some differences in readings obtained with permanently installedscrews and with the usual pins which are inserted into the wood at the time the moisture content is de-termined. At the time of installation the two readings are comparable. Thereafter, for a period ofabout six days there is a gradual divergence of the two, with the moisture content determined from thescrews being somewhat less. In this work we found a fairly constant difference of about one percent

(Figure V).

CONCLUSIONS

The results of this study lead to certain conclusions with reaged to the moisture content of struc-tural members in the ordinary residential building in the Baton Rouge, Louisiana area.

1. There are some differences in the moisture content of structural members at various locations,with the bottom plate generally having the highest moisture content and the rafters having the lowest.

2. The average moisutre content of structural members is considerably lower than that set up byF. H. A. as a minimum standard (5). It is so low, in fact, that only kiln-dried material should be used.

3. Framing material for use in the Baton Rouge area should be at a moisture content of 8 to 11percent at the time of installation.

4. Difficulties attendant upon drying and shrinking of the structural members might be avoided bythe use of portable heaters within the house for several days after enclosure but prior to installation ofthe interior wall.

5. The use of a moisture meter and permanently installed electrodes for this type of study isfeasible, but considerable care must be taken to ensure valid results.

-87-

LITERATURE CITED

1 H. L Mitchell and E. S. Iverson, 1961. Seasoning Green Wood Carvings With PolyethyleneGlycol-1000. Forest Products Journal XI, No. 1, January, 1961.

2. Peck, Edward C., 1961. Moisture Content of Wood in Use. U. S. Forest Products LaboratoryReport No. 1655.

3. W. C. Hopkins, 1960. The Moisture Content of Building Structural Members. Forest ProductsJournal X, No. 10, October, 1960.

4. U. S. Department of Commerce, Weather Bureau, 1961. Climatological Data, Louisiana-Annual Summary, 1961. Volume 66, No. 13.

5. Paul Bois, 1959. Wood Moisture Control in Homes, Seasonal Variation in the Southeast.Forest Products Journal IX, No. 9, September, 1959.

6. U. S. Federal Housing Administration, November, 1958. Minimum Property Standards.F. H. A. No. 300, Section 705-1,3, November, 1958.

-88-