Replies

1. 
   StanleyNiemiec | Jul 24, 2002 06:55am | #1

   Frank:

   Shrinkage between green and oven dry moisture content --

   Radial  --  4.6%

   Tangential  --  7.2%

   Volumetric  --  10.2%

   Coefficients for dimensional change due to shrinkage or swelling within moisture content levels of 6 to 14 percent:

   Radial  --  0.00158

   Tangential  --  0.00252

   Change in dimension is calculated by multiplying the appropriate coefficient by the change in moisture content (MC initial minus MC final), and then multiplying that value by the width of the piece of wood.   Thus dimensional change will depend on grain orientation, the variation in moisture content and the initial width of the piece

2. 
   StanleyNiemiec | Jul 24, 2002 12:56pm | #2

   Frank:

   In anticipation of your next question  --  What is the maximum MC the pieces will equilibrate to in your location?

   I seem to remember that you are in ridge runner country!  Anyway the USDA Forest Service publication "Storage of Lumber" (Ag Handbook # 531) gives the following EMC values for wood stored outside in Virginia:

   Spring  --  12.9%

   Summer  --  13.9%

   Fall  --  14.0%

   Winter  --  12.2%

   If the wood was kiln dried, the adsorption of vaporous moisture will follow a hysterisis curve that at first will have a lesser moisture content because of slower re-adsorption.  However, you can expect that over time, eventually, the wood will reach these values.

   There may be other issues like whether your particular site is wetter or drier, but the values above are a good starting point.  Obviously, you will want to use the highest value and you may want to fudge a bit and maybe go slightly higher to be safe especially when you are making you calculations for estimating expansion.

   Rather than attempting to calculate a dimensional change value for each board, I think you might get a better perspective if you calculate the dimensional change (expansion) on both a per foot basis and an overall expansion for the entire distance of the porch.  Those numbers might give you a better indication of the movement you can expect.  And if you have ever laid wood floor, you will know that the floor is oriented so that expansion is minimized (by having the longitudinal axis of the boards parallel to the longitudinal axis of the room).

   Rather than being concerned about coating the back surface (unless it is somehow exposed to the elements) I would recommend you coat the end grain.  For your situation, the addition of paraffin to the oil would greatly enhance the shedding of liquid water and (if concentrated enough) would act as an effective end coating.

   Oils as finishes tend to have a low MEE number (Moisture Excluding Effectiveness).  Tests were conducted at the Forest Products Lab (USDA Forest Service - Madison WI) for most standard wood coatings.  The tests measured the adsorption of vaporous water over various intervals (1, 7, and 14 days).  I do not have the study in front of me, but my recollection of the data comes up with a MEE number for 14 days for boiled linseed oil of <2%.  That means that the coating is only 2% better than raw wood is retarding/reducing moisture adsorption/loss.

   And with regard to oil coatings, the wood itself is decay resistant but the oil is not.  As such, it could be a potential food source for surface bacteria, molds, mildew and fungi.  If you are applying an oil, make sure that it has some additive (as a wood preservative) that will act as a mildicide.

1. 
   Biscardi | Jul 24, 2002 02:12pm | #3

   Stanley,

   Thankyou for your excellent reply.

   Lets see if Idid this correctly-

   assuming the wood is about7% and will increase to 14 % sittting outside

   assuming a 30 foot (660 inch) long and 8 foot wide deck

   .00158 x (14-7) x360 = 3.9 inch of movement down the length

   and .00158 x (14-7) x 96 = 1 inch across the width

   ?sound right?

   Thanks

   Frank

1. 
   StanleyNiemiec | Jul 24, 2002 09:30pm | #4

   Nope!

   For all intents and purposes, longitudinal expansion and contraction is nil.  Remember that cellulose is a long chain glucose polymer with a degree of polymerization about 25,000.  Water does not get between the monomer units (links) of the cellulose but rather it does get in between two parallel, side by side cellulose polymers, which is the cause of the expansion/contraction.

   Within each wood cell (as a vessel for conducting water from roots to stem) there are various cell wall layer (Primary, S1, S2 and S3).  Each of these cell wall layers has a different fibril angle (and thus different angle for the cellulose orientation.  The S1 layer may be 15 degrees to the left and the S2 might be 15 degrees to the right.  This is entirely different than spiral grain (which shows that spiralling through the tree).  As such the cell wall layers appear to look somewhat like the wrapped layers of insulation that are used for some wires.  This crossing of fibril angles imparts incredible strength and also longitudinal stability.

   The two coefficients deal with grain orientation/sawing pattern.  If a board is quartersawn so that the width of the piece is parallel to a straight line between the center and the outside edge (as a radius line or spoke might be), it is considered radial.  Conversely, if the width of the piece is oriented to a line perpendicular to a radius line, it is considered tangential (as if it were a tangent to some imaginary circle).  Tangentially oriented pieces are considered flat sawn.

   The difference between radial and tangential shrinkage is primarily due to cellular forms that are used to transport fluids and nutrients horizontally.  A real good visual image of this are the medullary ray flakes that are prominent in Qsawn white oaks.  Those flake act as a restraining mechanism.  While less prominent these medullary rays exist in (many/most/all???  -- I'd need to check some references to be sure) hardwood species like cherry, beech and sycamore.  Jon might know better as he tends to be more the wood anatomist while I am more a wood technologist.

1. 
   Biscardi | Jul 24, 2002 11:11pm | #5

   Stanley,

   So I guess you are talking more about the tendency to get cupping or bowing then the expansion acros the width or length?

   Thanks,

   Frank

2. 
   jonsherryl | Jul 24, 2002 11:31pm | #6

   Frank, you and Stanley have me mesmerized with all this higher math. I was able to keep up with Stanley's average seasonal EMCs for your region, but I don't have the foggiest idea what you two are calculating beyond that point.

   The average shrinkage stats, green to ovendry, for black locust are as follows: Volumetric = 10.2%, Radial = 4.6% and Tangential = 7.2%. T/R ratio = 1.57 : 1.

   Longitudial shrinkage (shrinkage parallel to the grain) is so small that it is normally ignored when calculating a wood's propensity to shrink.

   Shrinkage begins as a wood's moisture content declines below the Fiber Saturation Point (FSP.) For most woods, the FSP falls somewhere between 26% and 30% MC and shrinkage occurs rather uniformly with loss in moisture content (until the wood reaches about 3%MC, when it begins to become too rigid to collapse much further...but since wood is seldom used when in the completely desiccated "ovendried" condition, this isn't something the wood worker needs to be concerned about.)...

   So, for practical purposes, the rule of thumb is that wood experiences about 1/25th of it's total (maximum potential) shrinkage for each 1% decline in its Moisture Content once the FSP is reached. Given this 1/25th relationship and the above shrinkage data, you should be able to estimate black locust's rate of shrinkage (or expansion) form whatever MC it has when first put in use to any Equalibrium Moisture Content (EMC) it is likely to reach in your climate zone.

   The problem is, in climates where homes use central heating, the real EMC range the wood will experience, if used in an interior project, is much wider than the EMCs quoted for that climate zone. Also, it's important to take into consideration the difference between tangential and radial shrinkage, in that flatsawn stock typically shrinks in width about twice as much as does quartersawn stock. Fortunately, black locust has a rather low T/R ratio. However, when using run-of-the-mill standard lumber where the grain orientation varies form board to board, to play it safe, you should use it's tangential shrinkage in calculating the probable maximum movement. Hope this helps.

   Edited 7/25/2002 4:10:10 PM ET by Jon Arno

3. 
   HowardAcheson | Jul 25, 2002 12:13am | #7

   Let's make it easy.  Go to http://www.woodbin.com and click on the "Shrinkulator".  Input the species, width, starting moisture content or relative humidity and the change in either and it will calculate the expansion/contraction.

1. 
   Biscardi | Jul 25, 2002 01:40am | #8

   Guys,

   Thanks for all the replies. The wood was milled in 1999, kilned and then put in an outside shed. I will check the MC, but I bet it is close to 12% so the maximum change in MC should only be about 2% in my area, using this on a covered portch. I think it will move only a little.

   Thanks

    Frank

   According to Shrinkolator 1.9 inch change over a 30 foot portch for 2% change in MC. So if I use 3 inch boards there should be .008 inch shrinkage per joint.

   Edited 7/25/2002 3:33:55 AM ET by BISCARDI

1. 
   StanleyNiemiec | Jul 25, 2002 04:46pm | #9

   Frank:

   Cupping is a very definite manifestation of unequal expansion and contraction due to grain orientation (radial vs. tangential).  I generally attribute bowing to be more associated with either weird sawing patterns, odd grain angles or growth stresses from the tree.

   When you lay floor, you never really have to worry too much about shrinkage other than the fact that it produces unsightly openings between boards.  Expansion is a whole "nuther" beast  --  too great of expansion can cause a floor to buckle.

   The values I gave you for EMC values in Virginia apply only to moisture adsorption from humidity in the air.  Somehow you are going to have to consider the likelihood of wind blown rain getting on the wood.  Were this to happen, the MC of the wood could increase significantly  --  not totally but as a 'spot' phenomenon.  If one side is wetted, you will get swelling on that side only, which could contribute to warping of individual boards, or, if a sufficient area becomes surface soaked, buckling.

   I did mention the layout of strip flooring; that procedure may not be appropriate for a porch.  For a porch, I believe that many are installed so that joints run perpendicular to the side of the house.  This is done, I guess, to facilitate water draining (downhill) within the T&G joint; for if the joint was to run parallel to the side of the house, water could get trapped.  Running strips perpendicular to the side of the house however creates a very wide surface that equates to a very large total shrinkage/swelling potential.