The Moisture-Excluding Effectiveness of Finishes on Wood Surfaces

United States

Department of

Agriculture

Forest Service

Forest

Products

Laboratory

Research

Paper

FPL 462

The Moisture-

Excluding

Effectiveness

of Finishes

on Wood Surfaces

William C. Feist

James K. Little

Jill M. Wennesheimer

Abstract

Permeability to water vapor is one of the more important

properties affecting the performance of coatings and other

wood finishes. Often, one of the main purposes of finishing

wood is to restrict moisture movement from the

surroundings. We evaluated the moisture-excluding

effectiveness (MEE) of 91 finishes on ponderosa pine

sapwood, using the Forest Products Laboratory method in

which finished and unfinished wood specimens in equilibrium

with 30 percent relative humidity (RH) at 80 °F are weighed

before and after exposure to 90 percent RH at 80 °F.

Finishes with the best MEE were pigmented, nonaqueous

(solvent-borne) finishes. Two-component epoxy paint

systems had MEE values greater than 85 percent after

14 days when three coats were put on the wood. Molten

paraffin wax and a sheathing grade, two-component epoxy

material with no solvent were the very best finishes found in

this study for controlling moisture vapor movement into

wood. The MEE is a direct function of the number of coats

of finish applied to the wood (film thickness) and the length

of time of exposure to a particular humidity. Only 11 finishes

were found to retard moisture vapor movement into wood

with any degree of success over the relatively short time of

14 days, and then only when two or three coats were

applied.

These studies include evaluations of MEE by finish type,

number of coats, substrate type, sample size, and time of

exposure, and describe the effect on MEE of repeated

adsorption/desorption cycles.

This paper should be useful to builders, architects, wood

furniture manufacturers, those who make wood finish

formulations, and anyone else interested in controlling water

vapor movement into or out of wood. The information will

benefit those who need to select wood finishes with specific

moisture-excluding effectiveness.

Keywords: Water vapor, moisture exclusion, paints, coatings,

wood finishes, adsorption, desorption.

December 1985

Feist, William C.; Little, James K.; Wennesheimer, Jill M. The moisture

excluding effectiveness of finishes on wood surfaces. Res. Pap. FPL 462

Madison, WI: U.S. Department of Agriculture, Forest Service, Forest

Products Laboratory; 1985. 38 p.

A limited number of free copies of this publication are available to the public

from the Forest Products Laboratory, One Gifford Pinchot Drive, Madison, WI

53705-2398. Laboratory publications are sent to over 1,000 libraries in the

United States and elsewhere.

The Laboratory is maintained in cooperation with the University of Wisconsin.

United States

Department of

Agriculture

Forest Service

Forest

Products

Laboratory

Errata

The Moisture-

Excluding

Effectiveness

of Finishes

on Wood Surfaces

Research

Paper

FPL 462 Errata

January 1986

The two figures on page 37 are reversed.

That is, the figure shown with the

caption labeled "Figure 11"

should be with the caption labeled "Figure 12,"

and the figure shown with the caption labeled “Figure 12" should be with the

caption labeled "Figure 11."

Feist, William C.; Little, James K.; Wennesheimer,

Jill M. The moisture-excluding effectiveness of

finishes on wood surfaces. Res. Pap. FPL 462.

Madison, WI: U.S. Department of Agriculture, Forest

Service, Forest Products Laboratory, 1985. 38 p.

Contents

Earlier Work

....................................

Evaluation Method

..............................

Moisture-Excluding Effectiveness (MEE)

.........

Finishes

......................................

MEE of Finishes on Standard Wood Specimens

Results and Discussion

..........................

MEE of Wood Finishes

........................

MEE and Finish Characteristics

.................

MEE and Film Thickness (number of coats of

finish)

......................................

MEE of Aluminum Flake-Pigmented Varnish

.....

MEE of Combination Finishes

..................

MEE of Paraffin Wax Treatments

...............

MEE for Brush vs. Dip Application

..............

MEE of Wood Finishes as a Function of Repeated

Adsorption/Desorption Cycles

................

MEE and the Role of the Substrate

.............

Test Procedures, Materials, Methods

............

Specimens

.................................

Finishes

...................................

Equipment

.................................

Methods

...................................

Tests

......................................

Summary and Conclusions

......................

Literature Cited

................................

Appendix A-Description of Wood Finishes Used in

These MEE Studies

...........................

Appendix S-Composition of Finishes

.............

Appendix C-Adjustment of MEE Values for Wood

Density

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page

The Moisture-

Excluding

Effectiveness

of Finishes

on Wood Surfaces

William C. Feist, Supervisory Research Chemist

James K. Little, Mathematical Statistician

Jill M. Wennesheimer, Physical Science Technician

Forest Products Laboratory, Madison, WI

The primary objective of the work reported here was to

measure the moisture-excluding effectiveness of a wide

range of commercially available surface treatments and

finishes on wood. We studied the important variables of film

thickness, wood species and substrates (plywood,

hardboard, flakeboard, etc.), and time. Our studies were

restricted to the measurement of the effectiveness of

finishes on wood against water vapor between 30 and

90 percent relative humidity (RH) at 80 °F.

Data in the tables and Appendices of this report have been

condensed as far as possible for the convenience of the

reader. Complete data and descriptions of the finishes used

can be obtained by contacting the U.S. Department of

Commerce, National Technical Information Service (NTIS),

5285 Port Royal Road, Springfield, VA 22161 (phone: (703)

487-4650 or (703) 487-4700 (RUSH)). The information is

listed with NTIS by the same authors under the title: ‘The

moisture-excluding effectiveness of finishes on wood

surfaces-support data”(8

Water is a prime factor in the environment of many finished

products and permeability to water (as liquid or vapor) is

therefore a property of major importance in most finish

applications, protective or decorative. With metal substrates,

the usual aim in applying a coating (finish) is to protect

against wet corrosion; with masonry and other porous

inorganic structures, coatings are applied to minimize water

seepage and “breathing” of moisture vapor; with wood

substrates, the protective function of coatings and other

finishes is to minimize deterioration of the underlying

structure by the combined action of moisture, sunlight,

oxygen and microorganisms.

Italicized numbers in parentheses refer to literature cited at the end of this

report.

Moisture affects the shrinking and swelling of wood (

22,25

);

a coating on wood therefore serves a desirable function if it

controls moisture sorption by the wood. Although paints and

other coatings (hereafter called finishes) have many other

important functions and properties, an understanding of the

moisture-excluding effectiveness (MEE) of finishes and its

measurement is essential because of the importance of

dimensional stability to many wood applications, both

indoors and outdoors.

Earlier Work

Evaluation Method

It has long been recognized that coatings of paint and other

finishes can prevent rapid changes in the moisture content

(MC) of wood (

3,4,12,22,26,27

). The ability of finishes to

protect wood against weathering depends partly upon the

property of moisture exclusion (

6,7,11,15,18,19,23

Moisture exclusion is also a valuable property of many

interior finishes for wood (

1,10,14,26

Michaels (14) has shown that in homogeneous polymers

(such as paint resins), water permeability is governed by the

concurrent processes of water adsorption and diffusion,

which are determined by the polarity of the polymer, and

such polymer structural features as chain-stiffness,

crystallinity, and cross-link density. Low-polarity polymers of

high crystallinity, stiffness, and/or cross-linkage are the best

water-barriers. Fillers and pigments added to the polymer

films (i.e. paints) can significantly improve water-barrier

properties when the particles are properly dispersed and of

suitable size and shape. With latex-derived coatings,

moisture permeability is often high and is affected by the

composition and concentration of surfactants and stabilizers

in the latex.

Various methods are used to evaluate the moisture

permeability of finishes. Some methods use isolated films

(

1,5,10,14,28-30

). Others apply the finish directly to wood

(

3,4,12,15-21,23,24

). The Forest Products Laboratory

(FPL) method used in early investigations of moisture-proof

finishes (3,4,12,25 ) measured the moisture (water vapor)

gain of wood protected on all sides by the finish and

exposed to a controlled atmosphere of high humidity.

We used the FPL method in the studies reported in this

paper because it represents a “real world” situation under

non-steady-state conditions. The finishes to be evaluated

were applied by brush to selected specimens of clear

ponderosa pine (Pinus ponderosa) sapwood measuring

3 x 5 x 5/8 in. (tangential x longitudinal x radial

dimensions), and having carefully rounded edges and

corners (fig. 1). Three replicates were used. The specimens

were conditioned to 30 percent RH at 80 °F, finished with

the appropriate material, dried, and reconditioned at

30 percent RH and 80 °F to equilibrium. The specimens

were then weighed and exposed to 90 percent RH and

80 °F for various time periods. Each finished specimen was

accompanied by an end-matched, uncoated control

conditioned and handled in exactly the same way. All

specimens were weighed at appropriate intervals (1,7,14,

21 days and longer as needed). Complete experimental

details and methods are shown in the Test Procedures,

Materials, Methods section at the end of this publication.

Additional studies were conducted on other wood substrates.

Moisture-Excluding

Effectiveness (MEE)

The amount of moisture vapor passing through the finishes

and adsorbed by the wood was determined from the gain in

weight of the sample between 30 and 90 percent RH after

different intervals of exposure. The MEE against water vapor

for the various finishes was calculated by comparison with

the weight of moisture adsorbed by the uncoated panels

(3,4):

MEE =

U-C

x 100

where U = weight of moisture adsorbed by uncoated wood

C = weight of moisture adsorbed by finished wood.

Measurements are readily made by this method with a good

degree of accuracy. Furthermore, the experimental

conditions are such that the results express the combined

effect of the permeability of the finish and the adsorption

characteristics of the wood. However, even with careful

selection of the wood specimens, the possibility of natural

variation in wood structure and adsorption characteristics is

not removed. Variations in the adsorption characteristics of

the wood may not seriously affect the amounts of moisture

adsorbed by the finished specimens, but they do seriously

affect the MEE values. This is because MEE values are

based on the amount of moisture adsorbed by the uncoated

wood, and different bare wood specimens (the controls)

adsorb different amounts of moisture. This aspect of the test

is discussed in detail in the section entitled Effect of Density

of the Wood Specimen.

A further difficulty is that the conditions of the test for MEE

are empirical. As the wood approaches its fiber saturation

point the rate of diffusion of moisture through the finish

decreases (20,22), so that the longer the exposure to high

humidity, the lower is the apparent MEE of the finish. The

countervailing advantages of the test are that it does

represent service conditions of finished wood and does

MEE of Finishes on

reflect the degree of interaction between the substrate and

the finish and any stresses placed on the finish as the wood

Standard Wood Specimens

expands during wetting.

The MEE of the finishes selected for this study were

Finishes

determined for one, two, and three coats of each finish on

ponderosa pine clear sapwood between 30 and 90 percent

RH at 80 °F (table 1 and (

)). Values were determined after

For these studies we selected a wide range of commercially

1, 7, and 14 days. The test for MEE was continued beyond

available finishes and several laboratory prepared finishes

14 days until the observed value fell below 50 percent.

whose descriptions and compositions are summarized in

Since MEE changes with time, all MEE values are discussed

Appendices A and B, and given in detail in (

). For

in terms of the day of measurement (e.g. MEE

, stands for

convenience we grouped the selected finishes under a

the measurement at 14 days) and, unless specified

descriptive classification based on the suggested use and

otherwise, are the average of three determinations.

composition of the finishes (interior/exterior/combination,

pigmented/unpigmented and aqueous/nonaqueous) as set

out in Appendix A. The finishes were applied by brush. From

one to six coats were applied to the conditioned wood

surface. Several finishes were used in combinations.

Figure 1.—Ponderosa pine test specimen pairs in exposure rack.

---------------------------

Table 1.—Moisture–excluding effectiveness (MEE

) of finishes on ponderosa pine sapwood (after t days exposure at 90 pct

relative humidity, average of three replicates)

Number

Finish

Coverage Wood MEE

for—

coats 1 coat

2 coats 3 coats

density

t=1

t=7 t=14 t=21

t=28

t=35

t=60

-----------

/gal ----------- Lb/ft

Pct

---------------------------

EXTERIOR, UNPIGMENTED, NONAQUEOUS FINISHES

1 303

— — 24.8 48

–0 — — —

—

304

494 — 24.6 90

66 46 — —

—

303

517

510 24.9 94

6 6 58 47

——

————

408

— — 24.9

–0 –2

—

——

1 436

—

— 23.0

–0

–1

2 433

679

—

23.4 46

–1

——

3 408

729

868 21.5 78

——

320

— — 22.9 59

13 3

——

2 317

327 — 22.9 81

——

—

3 287

324

345

22.7 88

——

COMBINATION EXTERIOR/INTERIOR, UNPIGMENTED, NONAQUEOUS FINISHES

302

——

24.9

————

325

265

—

24.6 90

————

317

255

261

24.5 94

————

1 256

—

26.1

54 40

———

247 444

—

26.2

79 74

3 346

422

387 27.7

1 552

—

23.9

–1

—

———

2 546

791

—

23.5

—

———

3 526

789

948

23.8

51 31

—

———

1 529

—

22.2 34

–1

—

2 547

778

22.7 46

–1

580

750

844 22.8

52 6

—

325

——

24.5

————

2 349

497

—

24.6

53 33

————

354 537

400

25.4 95

63 52

———

355

—

25.2 28

–0

—

———

357

497

—

25.6

—

341

473 438 25.6 85

—

392

—

22.4

–4

–5

—

423

579

—

22.5 22 – 2

–4

—

3 532

603 872

24.1

33 2

–0

—

561

—

25.3 7 –1 –1

—

2 546

803

—

25.6

———

—

514

649

768 25.6

0–1

———

—

1 505

—

24.5 55

10 2

———

—

507

635

—

24.5 83

43 23

—

———

3 497

580 559 25.2 89 64 4 4

—

———

554

—

26.2 48

—

———

2 573

867

—

25.9 80 36 15

—

3 539 884

776 26.3 87 53

—

526

—

25.3 60 24

—

547

741

—

25.2 87 56

—

575

744 1016

24.5

89 63 44

—

———

1 518

—

24.7

—

———

2 545

626

—

24.6 84 46 27

—

570

635

936 24.5 88 58

—

—— —

Page 1 of 6 pages

-----------

---------------------------

Table 1.—Moisture–excluding effectiveness (MEE

) of finishes on ponderosa pine sapwood (after t days exposure at 90 pct

relative humidity, average of three replicates)–con.

Number

Finish

Coverage Wood

MEE

for–

coats

1 coat

2 coats

3 coats

density

t=1 t=7 t=14 t=21 t=28 t=35

t=60

/gal -----------

Lb/ft

Pct

---------------------------

INTERIOR, UNPIGMENTED, NONAQUEOUS FINISHES

—

22.6

-1 -1 - 1

—

——

382

—

23.3 40

—

390

451

—

23.1

70 22

—

380 460 392 23.4 79 37

—

——

616

——

23.5 52

—

630

781

—

24.0

38 18

—

635

708

922 24.8 87

———

—

512

——

25.7 35

——

532

706

—

26.3

78 39

—

522

708

758

25.3

86 53

—

——

—

573

——

22.4 77

—

——

608 639

—

23.0

—

627 626

647

22.4

—

457

—

25.7

—

——

—

474 580

—

25.6 86 47

25 —

—

416

651 631

25.7 91 67

—

405

—

23.4

—

469 509

—

23.6

84 43

—

422

515

480 24.4 91

42 —

——

—

552

——

26.0

—

———

505

683

—

22.5

87 54

——

—

487

695

729 22.4 92

—

576

——

22.5

-1

—

564

713

—

22.5 80

—

—— —

569 774

821

23.2

— ——

—

550

—

25.6

—

554

640

—

25.8

87 53

—

535

699

760 25.6

—

585

——

25.8

—

——

594 724

—

26.2 79

—

——

—

598

671

587

26.6

—

596

—

22.2

—

644 752

—

22.1

—

664 824

819

22.5

64 43

—

——

—

26.9

-1

—

26.9

-1 1

1—

——

—

26.5

—

330

—

21.8

46 27

416

——

23.1

-4

-6

——

—

423 472

—

23.7

-1

-5

—

477

466

673

24.4

-1

-2

—

—— —

—

22.3 6

-2

-2 —

—

——

—

23.5

-2 -3

—

23.9

-0

-1

———

—

503

—

24.4

—

——

—

507 688

—

24.0

—

483 734 802

24.3 85 52

—

644

—

27.3 66 20

—

——

—

597

831

26.4

85 49

—

632 846

734 25.9

90 63

—

——

—

360

24.4

7—

——

—

336

422

25.2

—

——

—

379 405

378 25.2

—

Page 2 of 6 pages

Table 1.—Moisture–excluding effectiveness (MEE

) of finishes on ponderosa pine sapwood (after t days exposure at 90 pct

relative humidity, average of three replicates)–con.

Number

Finish

Coverage

Wood

MEE

for–

coats

1 coat 2 coats

3 coats

density

t=1 t=7

t=14 t=21 t=28 t=35 t=60

-----------Ft

/gal-----------

LB/ft

----------------------------Pct----------------------------

INTERIOR, UNPIGMENTED, AQUEOUS FINISHES

36 1

485

— — 23.3

-1

—

——

36 2

503

607 — 23.5

——

—

36 3

509

660

750 24.4

68 24

—

——

500

——

23.1

38 2 - 1

——

—

37 2 468

690

— 23.3

12 5

——

—

37 3 514 687 702 23.3

—

38 1 622

—

— 24.6

-1

—

38 2 624 862 — 27.4

—

38 3

570

783

795

24.1

70 22

——

—

39 1 463 — — 22.2

-1 -1 -1

—

39 2

466

597 — 21.6

5-3-3

——

—

39 3 456

640

811

21.0

-2 -4

—

EXTERIOR, PIGMENTED, NONAQUEOUS FINISHES

1 545

——

23.4 88 55 32

—

2 492

545

—

23.6

68 60

40 3

505 504

618 24.1 98

84 78 72

41 1 498

——

23.3 91 64 43

——

—

519

621

—

24.0 9 6 85 72

63 55

—

499

652

594

23.6

98 90

42 1

529

——

23.5

90 60 39

—

42 2

534

598

—

23.6

97 85 74

63 55

—

515

571

554

24.0 98

84 77

43 1 523

—

23.5

—

514

591

—

24.2 97 87

476 577

713

24.3 98

44 1 494

——

25.1

—

44 2 480 755

—

25.3

79 62

—

44 3 459

686

827

25.3

96 86

—

45 1

691

——

26.8

22 1 - 0

—

677

831

—

24.5

76 28

—

597 797

1047

24.9 8 9 57

—

46 1 665

——

27.6 61

16 6

—

670

791

—

26.8

—

658

677

815

26.7

90 66

—

47 1 682

—

26.9

79 38

—

47 2 662

699

—

27.0

——

—

625

791

849 27.3 93

57 44

—

544 —

—

24.9

——

—

48 2 496

807

—

25.1

89 62

——

—

48 3 516 630

842

25.2

53 45

—

49 1 444

——

24.6 66

——

—

50 1 616

—

26.3

——

—

50 2 678

1066

—

25.2 70

——

—

614

1085

1061

25.5 76

——

—

519

—

25.2 69 20

——

594

1132

—

25.3

55 32

——

—

3 597

1078 1090 25.8

69 49

——

—

1 649

—

25.5

-1

——

—

620

468

—

25.4 87 52 28

——

5 2 3 612

505 554 25.6

93 70 48

—

53 1

651

——

24.5 82 39

—

53 2

621

648

—

24.1

93 70 48

—

——

53 3

612

588

645

23.7 95 80 64

——

Page 3 of 6 pages

-----------

---------------------

Table 1.—Moisture–excluding effectiveness (MEE

) of finishes on ponderosa pine sapwood (after t days exposure at 90 pct

relative humidity, average of three replicates)-con.

Number

Finish

Coverage Wood MEE

for–

coats 1 coat 2 coats 3 coats

density

t=1 t=7 t=14

t=21 t=28 t=35 t=60

/gal -----------

Lb/ft

Pct ---------------------

EXTERIOR, PIGMENTED, NONAQUEOUS FINISHES—con.

———

576 — — 25.7

—

561

639

—

26.0

59 36

—

3 536

599 647 26.0

91 69 48 35

—

622

—

25.5

–1

—

———

662

961

—

24.7

–2

–3 —

———

670

991

987

24.8

–0 —

———

——

—

659

——

27.1

—

647

878

—

26.7

84 48

—

3 626 868 820 27.0 90

64 42 —

——

—

586

——

25.1

72 23 8

—

553

671

— 25.0 86

29 —

——

—

592 654

771 25.1

—

——

—

557

— — 23.7 85 46

—

——

—

2 527

659

—

23.4

70 50

—

——

—

487

702

621

23.4

95 78

——

COMBINATION EXTERIOR/INTERIOR, PIGMENTED, NONAQUEOUS FINISHES

—

——

439

23.4

82 37

—

397

480

22.9

93 69

49 36

—

444 474

443

22.6 94 76 59

—

——

—

23.9

—

—— —

24.0

—

60 3

—— —

24.0

96 85

—

542

—

21.1

—

538

705

—

21.2

———

—

——

—

492

708

728

21.2

96 75 52 30

521

——

24.4 80 31

—

——

—

534 766

23.7

———

—

62 3 562 699 888

23.4

92 63

436

—

25.3

——

—

429 532

26.2

67 60 40

63 3 426 467 498

25.6

82 76

—

353

—

25.8

—

342 353

26.0

76 69

333

351

387

26.2

490

——

21.5

——

—

493

877

21.8

477 914 945

21.9

96 80 65 52

—

1 428

——

22.0

———

—

2 412 835

—

22.1

———

3 434

757 980

22.5 96 83

58 48

——

1 429

——

23.6

50 38

———

413 672

—

24.0

96 83 70 60

—

397 608 534

24.2

1 466

—

24.1

78 35

————

—

————

—

———

2 456

610

24.5

88 58

——

—

68 3 458 639

728

24.7

—

435

—

21.7

93 64

—

69 2

411

383

21.6

98 88

52 30

3 400 377 554

21.2

98 90

—— ——

—

70 1

371

23.7 86 45 23

———

—

70 2

357

469

23.8

65 43

———

399

495 502

24.5

93 72

Page 4 of 6 pages

-----------

---------------------------

Table 1.—Moisture–excluding effectiveness (MEE

) of finishes on ponderosa pine sapwood (after t days exposure at 90 pct

relative humidity, average of three replicates)—con.

Number

Finish

Coverage Wood MEE

for—

coats 1 coat 2 coats 3 coats

density

t=1

t=7 t=14 t=21

t=28

t=35

t=60

-----------

/gal

Lb/ft

Pct

---------------------------

COMBINATION EXTERIOR/INTERIOR, PIGMENTED, NONAQUEOUS FINISHES—con.

1 510

—

22.7 80 35

—

458 884

—

22.5

97 87

58 50

—

513 915

938

22.9

98 91

74 68

541

—

24.4

72 27

————

572 757

—

24.3

48 27

————

566

715

1082 24.2

58 58

—

——

330

——

25.9

91 62

————

343

396

—

25.1

———

370

358

378 25.2 96 82

—

74 1

460

——

22.9 79

—

74 2

461

659

—

23.3 89 61

—

74 3

454

666

589 23.2 92 70

—

388

—

23.6

88 51

————

2 385

475

—

23.7

75 56 42

———

382

449 510 22.6 96 82 67 55

—

INTERIOR, PIGMENTED, NONAQUEOUS FINISHES

341

——

25.2

–1

–2

————

303

378 — 25.5 25

–1

—

290

383 375 25.8

–1

—

——

407

— — 25.8 83

45 25

—

2 468

666 — 26.2

—

77 3

396

599

567 25.7 94

—

——

1 365

— — 22.4 92

—

——

2 357

506

— 22.4 95

52 44

—

3 362

523 493

22.1

55 38

EXTERIOR, PIGMENTED, AQUEOUS FINISHES

1 248

— — 24.6 43

—

——

218

359 — 24.4 67

——— —

236

289

382

25.3 72

20 4

—

——

1 495

— — 24.8 52

—

——

471

609 — 24.4 77

—

——

478

528

541

24.5

—

——

1 446

——

23.1

1–1

—

——

445

697 —

22.9

—

——

421

660

663

23.4 59

——— —

1 428

——

23.3

–0

——— —

2 423

710

—

23.8

–3

——

—

454

654 686 22.8 60

—

—— —

455

——

26.5

–1

—

2 460

544 —

26.0

–0

——— —

497

551

549 26.2

–0

—

415

——

23.4

—— —

—

459

599 —

23.3

—— —

—

456

615 686

22.6

—— —

COMBINATION EXTERIOR/INTERIOR, PIGMENTED, AQUEOUS FINISHES

——— —

—— —

322 — — 25.5

—

—— —

328 426 — 25.9

—

—— —

85 3

303 427 419 25.6

—

—— —

424 — — 27.3 29

–1 —

—— —

415

602

—

27.1 38

—

—— —

86 3

374

635

599 27.3 44

1—

Page 5 of 6 pages

-----------

---------------------------

Table 1.—Moisture–excluding effectiveness (MEE

) of finishes on ponderosa pine sapwood (after t days exposure at 90 pct

relative humidity, average of three replicates)—con.

Number

Finish

Coverage Wood

MEE

for—

coats

1 coat 2 coats 3 coats

density

t=1 t=7

t=14 t=21 t=28 t=35 t=60

-----------Ft

/gal

Lb/ft

Pct---------------------------

INTERIOR, PIGMENTED, AQUEOUS FINISHES

87 1 408

——

24.4 23

–0

—

—— —

87 2

422

620 —

24.3

5–1

—

—— —

87 3 396

567 578

25.4 49 8

–0

—

—— —

465

——

24.8 40

—

—— —

88 2

442

532 —

25.1

—

—— —

88 3 439

541

570

25.6 58

—

—— —

406

——

25.7 36

–1

—

—— —

433

596

—

25.9 44

–0

—

—— —

413

533

611

26.1

—

—— —

90 1 314 — —

22.7 78

—

90 2

312

439

—

21.0

47 27

—

—— —

90 3 304

441

413

21.5 88

—

—— —

355

——

24.5 5

–5

—

——

—

315

440

—

24.2

–6

—

—— —

364

423

464

23.8

–3

—

—— —

For complete data, see (8).

Page 6 of 6 pages

Results and Discussion

The protection afforded by finishes excluding moisture from

wood depends on a great number of variables (

4, 23, 25

Among them are finish film thickness, type and amount of

pigment, chemical composition and amount of the vehicle,

volume ratio of pigment to vehicle, vapor-pressure gradient

across the film, and length of exposure period. Under

outdoor conditions, the age of the weathered finish is very

important (

12, 23

). We investigated particularly the effects of

varying film thickness (number of coats), length of exposure

time, and chemical composition of the finish system.

Additional studies included the effect of sample size, role of

cycling humidity, effect of substrates (different woods and

wood panel products), and brush versus dip application of

the finish.

MEE of Wood Finishes

The data in table 1 show the wide range of MEE values

found on ponderosa pine at 90 percent RH for the

91 finishes in the study. MEE always decreases with

exposure time and always increases with greater film

thickness (i.e. two coats of a finish are better than one, and

three are better than two). The effects of finish type, number

of coats, and exposure time on MEE are shown in figures 2

and 3 for several of the finishes evaluated (see also section

on MEE and Film Thickness). The sheathing epoxy finish

(finish 6, fig. 2) was very effective and had a relatively high

MEE

of 74 and 81 percent when the finish was applied in

two or three coats. This is an unusual finish in that it is

composed of virtually 100 percent solids (Appendix B) while

the other finishes range from 20 to 80 percent solids.

In contrast to the epoxy sheathing compound (finish 6, fig. 2)

a latex house paint (finish 80) and a nitrocellulose lacquer

(finish 18, both in fig. 3) had low MEE values even with

three coats. These permeable filmforming finishes do have a

degree of protection against moisture vapor after 1 day at

90 percent RH, but the MEE falls very rapidly after that.

MEE and Finish Characteristics

As shown in Appendix A, this study included 16 aqueous (or

water-borne) and 75 nonaqueous (solvent-borne) finishes;

52 were pigmented (opaque) and 39 were unpigmented

(transparent). The finishes were also described as exterior

grade, interior grade, or combination exterior/interior grade.

These characteristics were important in determining the

MEE of the finish when applied to wood surfaces.

The finishes were ranked from highest to lowest MEE for

each day of measurement by the number of coats applied.

For convenience, only the first 30 finishes in each ranking at

each period of exposure are considered here. From table 2

it can be seen that the dominant characteristics of the

30 finishes with highest MEE are nonaqueous, pigmented,

and either exterior grade or combination exterior/interior

grade.

Table 2.—Distribution of finish characteristics for the 30 finishes with highest moisture-excluding effectiveness

Number of finishes

Duration

of test

Days

Number

coats

Combination

Pigmented +

Aqueous +

exterior/

+ Exterior +

lnterior

unpigmented

nonaqueous

interior

10 +

28 + 2 1 + 29

11 +

27 + 3 0 + 30

11 +

26 + 4 0 + 30

9 + 4 27 + 3 1 + 29

11 +

27 + 3 0 + 30

11 +

26 + 4 0 + 30

28 + 2 1 + 29

+10+

26 + 4 0 + 30

+10+

25 + 5 0 + 30

Figure 2.—Moisture-excluding effectiveness (MEE) of wood finishes on ponderosa

pine sapwood at 90 percent relative humidity and 80 °F as a function of time.

Number of coats is displayed at right of cuves. A. Finish 6: Two-component

sheathing epoxy; B. Finish 64: Two-component epoxy/polyamide paint: C. Finish

43: Aluminum-pigmented varnish; D. Finish 60: Pigmented flat shellac; E. Finish

44: Two-component polyurethane gloss paint: F. Finish 58: Tall maleic alkyd/soya

alkyd flat primer paint.

Figure 3.—Moisture-excluding effectiveness (MEE) of wood finishes on ponderosa

pine sapwood at 90 percent relative humidity and 80 °F as a function of time.

Number of coats is displayed at right of curves. A. Finish 77: Soya alkyd gloss

enamel; B. Finish 54: Soya/silicone alkyd gloss enamel: C. Finish 51:

Semitransparent linseed oil-based stain; D. Finish 5: Modified butyl-acrylo-styro

epoxy finish; E. Finish 80: Acrylic latex flat house paint; F. Finish 18: Nitrocellulose

semi-gloss lacquer.

In table 3 a ranking of finishes with MEE

70 percent

regardless of the number of coats of finish shows a

predominance of pigmented, nonaqueous, exterior or

combination exterior/interior finishes. The best finishes

(finishes 30 and 6) were special systems in that they did not

contain any solvent. Also, finish 30 (paraffin wax) was

applied by dipping or brushing the melted material. This

special finish is discussed in a separate section.

The individual conventional finishes with the best MEE

were the two-component epoxies (finishes 63, 64 and 69),

aluminum-pigmented varnishes (finishes 40-43), an

aluminum paint (finish 71) and a soya-tung alkyd satin

enamel paint (finish 67). The two-component epoxy/

polyamide paints or enamels in gloss, or satin finish had a

higher MEE

than the two-component polyurethane (finish

73). Generally, two-component epoxies were far better than

two-component polyurethanes whether pigmented or

unpigmented (table 1).

Of the 91 finishes used in this study, 87 were applied and

evaluated as 1, 2, or 3 coats. The overall effectiveness of

the 87 finishes against water vapor is illustrated in table 4

where the MEE values have been separated into four

ranges. After 1 day at 90 percent RH, 31 of the 87 finishes

had an MEE of 75 to 100 percent: with three coats, 63

finishes had MEE

of 75 to 100 percent. After 14 days there

were no finishes with an MEE of 75 to 100 percent when

one brush coat was applied, 7 with two coats, and only 11

with three brush coats of the finish. Thus, only 11 out of 87

commercial finishes were found to retard moisture vapor

movement into wood between 30 and 90 percent RH with

any degree of success over the relatively short time of

14 days and only when two or three coats were applied.

On the low end of the MEE ranges, 69 finishes had an

MEE

of 0 to 24 percent when only one coat was applied

(table 4). Twenty-eight finishes were in this MEE range even

with three coats.

Table 3.—Ranking of finishes with moisture-excluding effectiveness MEE

70 percent after

14 days exposure at so percent relative humidity. All finishes were nonaqueous.

Number

Finish

Finish characteristics

MEE

Standard

error

coats

Pct

1 Combination exterior/interior, unpigmented 95 3.7

Combination exterior/interior, unpigmented

0.7

Combination exterior/interior, unpigmented

1.4

Combination exterior/interior, pigmented 88 0.6

Combination exterior/interior, pigmented 87 0.3

3 Exterior, pigmented

1.1

Exterior, pigmented

1.1

Exterior, pigmented

1.7

Combination exterior/interior, pigmented

2.0

2 Combination exterior/interior, pigmented 82 0.5

3 Combination exterior/interior, pigmented

2.6

3 Extetior, pigmented

1.0

Combination exterior/interior, pigmented 80

1.4

2 Combination exterior/interior, pigmented 79

1.5

Interior, pigmented 78 2.3

Combination exterior/interior, pigmented 78 2.7

2 Exterior, pigmented

77 0.8

Combination exterior/interior, pigmented 76 2.9

2 Exterior, pigmented

1.7

2 Exterior, pigmented

2.1

3 Exterior, pigmented

1.8

Combination exterior/interior, pigmented 73 2.0

2 Exterior, pigmented

3.8

2 Interior, pigmented

3.2

Combination exterior/interior, pigmented

70 1.7

Combination exterior/interior, pigmented

70 3.9

Combination exterior/interior, pigmented

1.1

Mean of three observations.

Melted paraffin wax, dip applied.

Table 4.—Finishes in four ranges of moisture–excluding ef-

fectiveness (MEE) as a function of coat number and time al

90 percent relative humidity and 80 °F

Number

Number of finishes

in MEE ranges of—

Duration

75-100

50-74

25-49 0-24

of test

coats

pct

pct pct pct

31 21

260

14 16

324 36 5 22

14 1

27 42

30 28

Data from table 1. Eighty-seven finishes were applied to ponderosa

pine in 1, 2, and 3 coats.

Figure 4.—Moisture-excluding effectiveness (MEE) of

one to six coats of an aluminum flake-pigmented

varnish (finish 43) on ponderosa pine sapwood at

90 percent relative humidity and 80 °F as a function

of time. Number of coats is displayed to right of

curves.

MEE and Film Thickness

(Number of Coats of Finish)

The amount of finish applied to the wood surface is very

important to MEE values, as shown in figures 2 and 3.

Generally, the first and second coats are the most important

in the overall MEE of the finish. The first coat serves to seal

the wood but it is generally impossible to produce a totally

defect-free uniform film over the wood. Also some

penetration of resins or oils undoubtedly occurs (3,5,10

The second coat covers any defects of the first coat, and

doubles the film thickness. Each succeeding coat increases

MEE but, compared to the MEE produced by the first and

second coats, later increments are relatively small and

uniform even up to six coats (fig. 4). These small but

uniform increases result from the fact that film thickness is

doubled for the second coat but is increased only by

one-third for the third, by one-fourth for the fourth and so on.

We investigated the effect of increasing film thickness

(number of coats) up to six coats of finish for eight of the

finishes described earlier (table 5). The results for aluminum

flake-pigmented varnish (finish 43) illustrate the effect of six

coats of a finish with an excellent MEE. At the opposite end,

the acrylic latex house paint (finish 84) was quite porous to

moisture vapor and had MEE

of only 11 percent after six

coats of finish were applied (table 5). The butadiene/styrene

latex flat wall paint (finish 90) had an MEE

of 20 percent

for one coat and increased only to 45 percent for six coats.

This paint is recommended by the manufacturer as a vapor

barrier paint for interior walls (plaster, sheet rock, etc.).

The low values of MEE

for latex finishes stand in contrast

to those of the shellac-, varnish-, or paint-based finishes that

we evaluated. A white shellac (alcohol solvent) (finish 23)

with an MEE

of 73 percent for six coats was less effective

than a pigmented flat shellac (also alcohol solvent)

(finish 60) which had MEE

of 83 percent. For each coat

applied the MEE increase was greater for the white shellac

than for the pigmented shellac. This greater increase in MEE

with each successive finish coat for a nonpigmented versus

pigmented finish was also observed with the gloss urethane

varnish (finish 13) and the aluminum flake-pigmented

varnish (finish 43). Increases in MEE for the paints

(finishes 67 and 77) were similar to those for the pigmented

varnish and shellac. Browne (

) has done an extensive

study on the variations of MEE for a linseed oil paint

according to the nature of the pigment. In general,

pigmented finishes have much higher MEE than

unpigmented finishes for any specific resin system.

--------------------------------------------

Table 5.—Coverage and moisture-excluding effectiveness (MEE

) of six coats of finish on ponderosa pine sapwood (after

t days exposure at 90 percent relative humidity, average of three replicates)

Number

Wood

Coverage MEE

Finish

coats

density

1 coat 2 coats 3 coats 4 coats 5 coats 6 coats

t=1 t=7 t=14 t=21 t=28 t=35 t=60 t=90

Lb/ft

1 24.5 505

——

24.5

507

635

—

25.2 497

580

559

4 25.1

501

643

719

5 23.5

493

675

680

23.5

490

619

656

23.4

405

—

2 23.6

469

509

—

24.4 422 515

480

23.9 465 568

526

24.3

455

560

513

23 6 25.8

508

573

575

1 23.5

523

—

24.2

514

591

—

24.3 476

577

713

25.3

503

583

658

25.4

508

564

632

25.1

480 572

627

1 23.9

—— —

24.0

—— —

24.0

—— —

26.0 338

401

444

25.6 336

385

398

26.4

330

376 393

23.6 429

——

24.0 413 672

—

24.2 397 608 534

25.3 427

517

492

67 5

25.4 438

464

487

6 26.0 406

475

456

25.8 407

——

26.2 468 666

—

25.7

396

599

567

25.5

452

535

629

25.6

413 514

586

24.0

407 485

539

/gal

----------------------

——

—

———

774

—

812

784

816

802

793

—

562

—

599

641

637

704

537

——

—

——

—

——

—

623

—

668

743

714 679

691

—

348

383

447

—

385

412

481

—

———

——

—

592

—

459

449

—

476

490

465

———

——

—

——

—

663

—

674 518

—

559

555

732

23.4

415

———

—

23.3

459 599

——

—

22.6

456

615 686

—

26.4

415

589 657 677

—

26.5 386

659 536 608

666

—

84 6

26.6

395

585 623

570

566

582

22.7

314

——

—

90 2

21.0 312

439

——

—

21.5 304

441

413

——

—

23.9

454

662

656 683

—

23.9

474

645

654 702 688

—

26.3

466

621

665 668 706

665

Pct

----------------------

—————

55 10 2

—————

83 43 23

—————

89 64 44

———

51 39

—

———

72 57

———

93 76 62 50 42

65 11

—— — — —

—————

84 43 20

—————

91 63 42

———

93 75 58

45 —

———

94 81 67

——

95 85 73 64 55 49

92 61 41

—————

97 87 77 68 60

—

98 91 84

78 72

98 93 87

56 42

98 94 89

79 75 62 49

98 95 90

86 82 79 67 55

—————

91 67 44

———

95 81 65

54 45

——

96 85 73

56 50

—

96 88 79

64 57 36

—

97 89 81

74 67

—

97 90 83

93 69 50 38

—

———

96 83 70 60 50

——

97 89 80 72 65

—

98 92 85

79 74

53 40

98 93 88 83

98 94 89 85

—— — — —

83 45 25

—— — — —

—

94 76 59 46

—

———

95 80 65 52 43

—

96 84 72 61 53 46

—

96 85 74

56 49

—

—— —

—

—— —

—

11 5

—— —

—

45 11

—— —

—

61 19 1 0

—— —

—

63 19 10

—— —

—

65 20 11

78 37 20

—— —

—

86 47 27

——

—

88 55 33

——

—

90 57 36

—— —

—

90 60 39

—— —

—

92 66 45

—— —

—

-------------------

We investigated the dependence of MEE

on the amount of

material deposited on the surface of the specimen. Grouping

finishes according to the three finish characteristics

described earlier, we have six groups of nonaqueous

finishes with average three-coat MEE

50 percent:

Number of

finishes in

Characteristics

the group

Unpigmented

Exterior

Combination exterior/interior

Interior

Pigmented

Exterior

Combination exterior/interior

Interior 2

To represent the amount of material deposited on the wood

specimen, we calculated the total number of gallons applied

per square foot and multiplied by the percent solids that

each finish contained.

Scatter plots, for one, two, and three coats, showed that the

relation between MEE

and material deposited was

strongest for one coat and comparable for two and three

coats. The exterior grade pigmented finishes showed the

most consistent positive relationship between MEE

and

material deposited, but clusters of finish types (finish 41, 43,

44 near one point, 47 and 48 near another) prevent a simple

inference.

We calculated the correlation between MEE

and the

amount of material deposited for one, two, and three coats

as follows, using data for all finishes that had three-coat

MEE

50 percent:

Number of coats

Correlation

0.592

0.402

0.405

MEE of Aluminum

Flake-Pigmented Varnish

The effectiveness of aluminum flake pigments in varnish or

paint systems has been recognized for a long time

(

3,4,12,25,27

). We investigated the effect of adding several

different aluminum flake pigments in paste form (2 lb/gal) to

an exterior/interior grade polyurethane gloss varnish (finish

13, table 6). The MEE of the aluminum flake-pigmented

varnishes (finishes 41-43) was highest for those prepared

from flake pigments with the highest nonvolatile content and

those with the highest leafing content (finishes 40 and 43,

table 1). The role of film thickness in MEE was discussed

earlier (fig. 4) for finish 43.

Table 6.—Aluminum leafing pigments used In preparing

aluminum-pigmented varnishes

Aluminum leafing pigment property

Finish

Retention on

Nonvolatiles

Leafing

Bulking

325-mesh screen content

value

Pct

-------------------

Gal/lb

40 0.35

73.1

0.084

0.5

65.0

60-65

0.082

42 0.2

65.0

65-70 0.084

43 4.0

67.0

70-80 0.081

Pigment added to a gloss urethane varnish (finish 13) to give a

final composition of 2 pounds aluminum paste per gallon of varnish.

See Appendices A and B for description of finishes.

MEE of Combination Finishes

Paints and other finishes are often applied to wood surfaces

as combinations of finishes, e.g. a sealer, primer, and a

topcoat. Several different finish combination systems were

evaluated for MEE (table 7). An all-acrylic latex house paint

system (finishes 79 and 80) was found to have higher MEE

for the combination system than for the individual finishes

and slightly higher MEE

but the MEE

was virtually

unchanged. The MEE

for the combination primer/topcoat

system (three coats) was essentially the same as the MEE

for three coats of either finish (finishes 79 and 80, table 1).

As noted earlier this observation is true of latex systems in

general.

When the first coat (primer paint) was alkyd-based

(finish 56) with a reasonably good MEE, the application of

two coats of acrylic latex topcoat (finish 84) over the alkyd

primer paint did not have any measurable effect on MEE.

Thus, permeable latex paints have no effect on the MEE of

less permeable alkyd finishes when the permeable paint is

applied over the less permeable finish.

Table 7.—Moisture-excluding effectiveness (MEE

) of combination wood finishes after t days exposure at 90 percent relative

humidity (average of three replicates)

1st Coat

2nd Coat

3rd Coat

4th Coat 5th Coat Mee

for—

Finish

Coverage Finish Coverage Finish Coverage Finish Coverage Finish Coverage t=1 t=7 t=14

/gal Ft

/gal

/gal Ft

/gal

-------

Pct

-------

248

— — —

— — — — —

–1

495

— —

— — — — — —

221

651

— — — — — —

471

80 609

— — — — — —

77 28

231

638

80 724

— — — —

557

————

—

— —

—

415

—————— — —

30 3

453 84

612

———— — —

459 84 598

———— — —

483

84 602 84

637

— — — —

441

——

———— — —

–2

444

——

———— — —

441

49 538

———— — —

669

791

———— —

—

434 49 490 46

968

46 603

— —

662 47 699

—————

—

454

49 486

851

716

—

496

48 807

—

——

—

467 49

535 48

721

48 685

— —

564

25 713 80 37

——

—— —— —— —

–2 –2

—— —— —

589 25 783 32

545

593

——————

84 46 27

—

—— —— —— —

689

16 976

—

—— —

379 35 405

378

————

360 35 394 35

491

—

86 54

See Appendices A and B for description of finishes used.

Finishes 2 and 46-49 are recommended by the manufacturer

for marine uses, particularly on boats. They are intended to

be used as combination finish systems. The

phenol-formaldehyde/linseed-tung wood sealer (finish 2) had

a very low MEE for a one-coat flnish. The soya-linseed alkyd

flat undercoat paint (finish 49) was only somewhat better

(table 7). In combination, however, the MEE was better than

the sum of the two finishes alone. When topcoat paints were

applied over the sealer/primer combination (finishes 46-48)

MEE was significantly increased and the four-coat system

had MEE

’s of 82, 87, and 85 percent, respectively. These

topcoats were soya alkyd or alkyd marine enamels

(Appendix A).

A paste wood wax (finish 32) does not provide protection

against water vapor but does help improve the MEE of a

two-coat phenolic/tung wood floor sealer (finish 25) (table 7).

When two coats of a spray furniture polish (finish 29) were

applied over either two coats of a polyurethane gloss varnish

(finish 16) or three coats of a nitrocellulose/alkyd lacquer,

the MEE

was reduced slightly as were the other MEE

values. The spray furniture polish by itself did not affect

MEE at all even with two-coat application.

MEE of Paraffin Wax Treatments

Paraffin wax (finish 30) gave especially interesting results

when the finish was melted and applied on the wood surface

either by brushing or by dipping. Molten paraffin wax brush

treatment gave the highest MEE

of all the 91 finishes

investigated (table 1). MEE declined fairly quickly with time

as was true for most one-coat finishes and MEE

for one

coat was 69 percent, still the highest MEE of all the

one-coat finishes in the study. The results for a molten

paraffin wax dip were even more impressive (fig. 5).

MEE

was 100 percent and MEE

, 95 percent; MEE

was

still 70 percent. This method of applying paraffin wax gave

the best results of any of the finishes investigated,

regardless of the number of coats (tables 1 and 5). The

presence of a near-perfect hydrophobic barrier on the wood

surfaces produced very high MEE values (table 8).

Table 6.—Moisture-excluding effectiveness (MEE

) of brush- and dip-applied finishes on ponderosa pine sapwood after

t days exposure at 90 percent relative humidity (average of three replicates)

Finish

Application

Number

coats

Coverage

1 Coat 2 Coats 3 Coats

MEE

for—

t=1 t=7 t=14 t=21

t=28 t=35 t=60

------------ Ft

/gal ------------

--------------------------Pct--------------------------

Brush

1 505 — —

2————

Brush 2

507 635 —

83 43

23 — — — —

13 Brush 3 497 580 559

64 44 — — — —

Dip

188

——

44 24 — — — —

13 Dip

2 214 268 —

46 — — — —

Dip 3

220

289

292

63 51 42 — —

Brush

567

94 76 59 46 — — —

407

468

—

———

—

666

—

Brush 3 396 599

77 Dip

199

——

76 59 46 — — —

30 Brush

1 330 — —

97 83 69 60 52 46

Dip

1 448 — —

100 97

95 92 90 87 79

MEE for Brush Versus Dip Applications

The results for dipping versus brushing of molten paraffin

wax prompted additional studies on how the two application

methods affect MEE. The study was brief (table 8) and

included one-, two-, and three-dip coats of an exterior grade

polyurethane gloss varnish (finish 13) and a one-dip coat of

a soya alkyd gloss enamel (finish 77). Results are also

shown for the paraffin wax (finish 30), brushed and dipped.

We found that one-dip coat was equal to two-brush coats of

the gloss varnish, but one-dip coat of the paint was equal to

three-brush coats. MEE

for the one-dip coat of paint was

identical to that of three-brush coats on the ponderosa pine

sapwood specimens. This beneficial effect of dipping is no

doubt due to the fact that more finish material is applied to

the wood surface (see coverage values in table 8), and

because dipping for 30 seconds in the finish (whether paint,

varnish, or molten paraffin wax) increases penetration and

provides greater sealing of the end-grain of the wood.

Figure 5.—Moisture-excluding effectiveness of one

coat of paraffin wax (finish 30) applied by dipping or

brushing of molten material on ponderosa pine

sapwood at 90 percent relative humidity and 80 °F as

a function of time. Method is displayed to right of

curves.

MEE of Wood Finishes as a Function of

Repeated Adsorption/Desorption

Cycles

All of the discussion so far has dealt with the

moisture-excluding effectiveness of relatively new or fresh

finishes on wood surfaces. Normally, only 3 to 6 weeks time

elapsed between applying the finish, equilibrating the

finished specimens to 30 percent RH and then starting the

MEE evaluation at 90 percent RH. The question arises, what

happens with longer times? Does MEE change with time?

What happens to MEE when the finished specimen goes

through repeated cycles between 30 and 90 percent RH?

The effects of outdoor weathering on MEE have been

addressed by earlier workers (3,4,12,19,23

). We wanted to

look at the effects of adsorption/desorption cycles between

30 and 90 percent RH and the accompanying effects of time

in the test for MEE.

MEE and Noncontinuous

Adsorption/Desorption Cycles

Generally, from 4 to 8 weeks elapsed between the time a

new finish was applied to the wood surface and the time the

MEE test was started at 90 percent RH. This was the time

frame required for the sample to reach equilibrium moisture

content (EMC) at 30 percent RH. This means that the finish

was curing and drying for this duration before the test was

begun. The MEE was then determined and the specimens

stayed in test at 90 percent RH until MEE values fell below

50 percent (table 1). The specimens were then returned to

30 percent RH.

We selected 13 representative finishes with varying degrees

of MEE to study the effects of repeat adsorption/desorption

cycles on MEE (table 9). The first adsorption cycle varied

from 2 to 17 weeks at 90 percent RH followed by a

desorption cycle of 8 to 20 weeks at 30 percent RH to EMC.

The second and third adsorption cycles were stopped after

14 days in test at 90 percent RH and the samples returned

to 30 percent RH for equilibration. The specimens were

brought to EMC at 30 percent RH between the second and

third adsorption cycles just as before the first cycle. Thus,

the minimum time that elapsed between the start of the first

and third adsorption cycles was 18 weeks; maximum

elapsed time was 59 weeks. The finished specimens with

the highest MEE took longer to come to EMC between

cycles than did the finished samples with lower MEE.

Table 9.—Repeated adsorption cycles: effect on the

moisture-excluding effectiveness (MEE

) of finishes on

ponderosa pine sapwood (after t days exposure at

90 percent relative humidity, average of three replicates)

MEE

for—

Finish

Number

Cycle 1 Cycle 2 Cycle 3

of coats

t=1 t=7 t=14 t=1 t=7 t=14 t=1 t=7 t=14

---------------------Pct---------------------

83 43

23 85 49 29

44 91 68 50

90 69 52

91 68 51 91 69 51 92 70 52

93 72 57 92 73 57 93 74 58

93 76 62 93 77 61

94 77 62

42 93

72 55

74 57

93 75

58 94

80 65

82 69

67 96 85

97 87

95 85

73 97 88 79

97 90

97 83

94 77

94 74

41 91

90 65 48

97 87

98 90 83

84 98

98 93 89

–1 40

87 52 28 89 56 30 90 57 31

93 70 48 92 71 47 93 71 47

82 39 16 84 43 19 85 44 21

93 70 48 94 73 51

94 74 54

95 80 64 96 83 67 96 83 69

75 30 12 76 32 16 75 32 16

88 59 36 89 61 41

88 61 41

91 69 48 93 72 51

92 71 53

72 23 8 76 28 12 76 29 12

86 52 29 87 55 34 88 56 34

90 63 41

91 64 45 90 65 45

82 37 18 82 42 19 75 23 6

93 69 49 96 82 67 97 84 71

94 76 59 97 85 74 98 88 78

88 48 25 87 49 28 88 49 27

94 73 54 94 73 54 95 74 54

96 80 65 96 81 66 96 81 67

93 69 50 93 74 56 94 76 59

96 83 70 97 86 76 97 84 77

97 89 80 98 92 85 98 92 87

83 45 25 85 51 31

85 53 33

91 64 43 91 69 50 91 71 53

94 76 59 94 80 64 95 81 68

95 80 65 95 83 71

95 83 71

96 84 72 96 87 77 96 87 77

96 85 74 97 88 80 96 88 80

78 37 20 89 52 31

90 55 32

86 47 27 96 76 56 96 77 59

88 55 33 96 78 60 95 78 60

MEE was found to increase through each new adsorption/

desorption cycle for nearly all the 13 finishes examined

(table 9). Even one-coat finishes generally showed an

increase in MEE in going from one cycle to the next. The

largest increases in MEE were usually observed between

the first and second cycles. This increase in MEE with time

and repeat adsorption/desorption cycles is most likely

caused by the continued curing of the finish and loss of any

last small amounts of solvent that could be trapped in the

finish film. The stresses placed on the film by the swelling

and shrinking of the wood during adsorption and desorption

were not sufficient to create any micro-crazing of the finish.

This means permeability is not affected even for a relatively

brittle finish like shellac (finish 23) and alkyd paints

(finish 52) and MEE is not reduced. Even relatively thick

films caused by applying six coats of finish (finishes 13, 23,

and 77) were found to have slowly increasing MEE with

each adsorption/desorption cycle.

Two of the finishes (a one-component epoxy enamel,

finish 59, and a butadiene/styrene latex primer paint,

finish 90) were found to have significantly higher MEE on

the second adsorption cycle (figs. 6 and 7). These two

finishes apparently undergo quite slow final cure of the resin

in the finish, and the permeability of the finish film decreases

with time and thus MEE increases. The increase in MEE

was much smaller between cycles two and three, as it was

with all the other finishes examined. These results show #at

several weeks to several months may be required before

any finish reaches its maximum MEE.

MEE and Continuous Adsorption/Desorption Cycles

In monitoring noncontinuous cycles we observed that

moisture vapor movement through the finish when the

humidity was increasing from 30 to 90 percent RH

(adsorption) was greater than when decreasing from

90 to 30 percent (desorption) (fig. 8). For example,

100 days were required for a finished specimen to adsorb

approximately 9 g H

O/100 g ovendried (OD) wood at

90 percent RH but even after 170 days at 30 percent RH,

only 7 g H

O/100 g OD wood had been desorbed. Such

behavior would be expected from the sorption hysteresis

observed for wood and other cellulosic materials (22).

Additionally, the vapor diffusion resistance of both wood and

finish depends on the moisture content (MC) and the

moisture gradient. These change with the different cycles

studied.

Since water vapor movement through a finish is slower

during desorption than during adsorption (fig. 8), a

continuous regular cycling of the RH (such as would be

found between summer and winter seasons, for example)

could cause an overall increase in the wood moisture

content if the adsorption/desorption periods were similar.

Three of the finishes used in the noncontinuous adsorption/

desorption studies were selected to demonstrate the effect

of continuous humidity cycles on wood MC. New wood

specimens were coated with one, two, or three coats of a

Figure 6.—Moisture-excluding effectiveness (MEE) of a one-component n epoxy

enamel (finish 59) on ponderosa pine sapwood at 90 percent relative humidity (RH)

and 80 °F as a function of time and for repeat adsorption cycles with equilibration

to 30 percent RH and 80 °F between each cycle. Number of coats is displayed to

right of curves.

Figure 7.—Moisture-excluding effectiveness (MEE) of a butadiene/styrene latex

primer paint (finish 90) on ponderosa pine sapwood at 90 percent relative humidity

(RH) and 80 °F as a function of time and for repeat adsorption cycles with

equilibration to 30 percent RH and 80 °F between each cycle. Number of coats is

displayed to right of curves.

white shellac (finish 23), an aluminum-pigmented metal and

masonry paint (finish 65), or a soya alkyd gloss enamel

paint (finish 77). The coated wood specimens and their

uncoated controls were conditioned to 30 percent RH for

30 days and then exposed to alternating cycles of 14 days

at 90 percent and 14 days at 30 percent RH for three

complete cycles (28 days per cycle, 84 days total).

The MC changes found for wood finished with the aluminum

paint have been chosen to illustrate the results of the test

(fig. 9). The uncoated wood was very near equilibrium after

14 days at either 30 or 90 percent RH and the second and

third adsorption/desorption cycles caused only slightly more

moisture in the specimens than the first. With one coat of

finish however, the MC at the end of each 14-day period at

either 30 or 90 percent RH was higher than that found for

the previous cycle at the same RH. The effect was similar

with three coats of finish except that the amount of moisture

in the specimen was less because the finish had a higher

MEE. These results show that the MC of a piece of wood

finished with a high MEE finish slowly increases as RH

increases and decreases over time periods that are not long

enough for the piece to come to EMC. The times to reach

EMC are very long for finishes with high MEE (figs. 7-9).

Figure 8.—Change in moisture content of ponderosa

pine sapwood finished with three coats of aluminum

paint (finish 65) when exposed to 90 percent and

30 percent relative humidity at 80 °F compared to

unfinished wood.

Figure 9.—Change in moisture content of ponderosa pine sapwood finished with

one and three coats of aluminum paint (finish 65) when exposed to alternating

cycles of 30 percent and 90 percent relative humidity at 80 °F.

MEE and the Role of the Substrate

Different Wood Species and

Composite Wood Specimens

All of the results reported thus far have been for finishes on

ponderosa pine sapwood. The MEE of any finish depends,

in part, on the type of substrate to which the finish is

applied. We investigated nine solid or composite wood

substrates to determine the contribution of the substrate to

MEE. A polyurethane gloss varnish (finish 13), soya-tung

alkyd enamel (finish 67) and molten paraffin wax (finish 30,

dip application) were chosen as finishes for this study. For

convenience, only the MEE

results are discussed

(table 10).

The substrate and the related amount of finish (coverage)

applied to the substrate have a significant effect on MEE for

a given finish. This is best illustrated by comparing the

results for western redcedar and hard maple. The varnish

was applied in nearly equal amounts to these two solid

wood substrates but the MEE on hard maple was greater

than that on western redcedar by a factor of 10 for one-coat

application. This large dfference decreased as more coats

were applied but three-coat MEE

on western redcedar was

48 percent compared to 78 percent on hard maple. This low

MEE on western redcedar is probably related to the high oil

content of this species compared to the other species

investigated. Similar results were found for the enamel but

the differences in MEE were not as large.

Table 10.—Moisture-excluding effectiveness (MEE

) of three finishes on different substrates after 14 days at 90 percent

relative humidity

Coverage

Wood substrate

MEE

coat

2 coats

3 coats

1st 2nd 1st

2nd 3rd

1 coat 2 coats

3 coats

----------------------------Ft

/gal----------------------------

FINISH

Ponderosa pine

484

458

510

473

563 534

Southern pine

574

543

726 577

724

888

Red oak

387

369

397

363

419

377

Douglas-fir plywood

412

374

498

378 504

464

Flakeboard

450

456 671

481

611

880

Western redcedar

358 339

390

313

356

435

Hard maple

361

357

381 422

376 427

Particleboard

307

317

469

294 459

419

Hardboard

314

324

424

307

450

438

FINISH

Ponderosa pine

411

382

463

366

379

444

Southern pine

Red oak

560

498

512

513

845

711

497

483

853

760

776

830

Douglas-fir plywood

Flakeboard

342

486

344

498

525

787

348

449

525

712

471

819

Western redcedar

564

507

851

547

787

917

Hard maple

Particleboard

585

328

548

316

797

468

513

319

815

474

747

431

Hardboard

391

347 523

360

513

498

-----------------Pct-----------------

65 77

64 72

60 74

76 82

All values are averages of three replicates. For complete details see (8).

Red oak is a ring porous hardwood (25) and it was

expected that the large vessels would be difficult to seal.

Thus, one-coat MEE

on red oak for the varnish was

38 percent compared to 57 percent on the diffuse porous

hard maple at nearly equal coverages but three-coat MEE

was nearly the same for both species (table 10). Even

though more enamel was applied to the red oak than to the

hard maple (i.e. coverage was lower for red oak), all MEE

values (one, two, or three coats) for red oak were lower than

that for hard maple.

The MEE of the varnish (finish 13) on the four composite

wood products (Douglas-fir plywood; Douglas-fir flakeboard

(

13); particleboard; and hardboard) was influenced by the

surface texture of the substrate especially for one and two

coats of finish. The smooth-surface hardboard and

Douglas-fir plywood had the highest one-coat MEE and the

rough-surface flakeboard and particleboard, the lowest.

Differences in MEE were less for two and especially for

three coats. Similar trends were found for the enamel

(finish 67) but the differences for this pigmented finish were

much less than for the unpigmented varnish.

The MEE

of one coat of paraffin wax (finish 30) applied

by dipping, was affected by the general structure of the

specimen. The results were as follows:

Ponderosa pine

Southern pine

Red oak

Douglas-fir plywood

Flakeboard

Western redcedar

Hard maple

Particleboard

Hardboard 89

Composite wood products that had surface and/or edge

irregularities (Douglas-fir plywood, flakeboard, particleboard)

had the lowest MEE; smooth specimens the highest

(hardboard, southern pine, hard maple, western redcedar).

The MEE of red oak was lower than that for hard maple,

reflecting the difficulty of sealing the large red oak vessels.

The western redcedar could be effectively sealed with

paraffin wax and the oils and other extractives in western

redcedar apparently did not disrupt the ability of the wax to

prevent the penetration and adsorption of water into the

wood. The MEE value for ponderosa pine was unexpectedly

low (62 pct) as compared to the value of 95 percent found

earlier for different samples (table 6).

Different Size Specimens

The size (3 x 5 x 5/8 in.) and shape of the specimens used

in these studies were chosen to represent a typical

predominantly flat-grained surface and for convenience.

Since water is adsorbed more rapidly through the end grain

of the wood (22,25) the size and shape of the wood

specimen are expected to have an effect on MEE, as is the

amount or ratio of end grain to lateral surface (tangential

and radial). Ease of application and the size and shape of

edges must also be important factors in determining the

MEE of any finish.

We did only one brief study on shape and size of

specimens. The soya-tung alkyd enamel (finish 87) was

applied in one, two, or three brush coats to ponderosa pine,

southern pine, red oak, Douglas-fir plywood, and Douglas-fir

flakeboard wood specimens. Specimen sizes were 3 x 5

x 5/8 in. and 6 x 10 x 5/8 in. (tangential x longitudinal

x radial). This means the flat grain (tangential) surface was

increased by a factor of four while the end grain and vertical

grain (radial) surfaces were increased by a factor of two in

going from the small to the large specimens.

The results of this study (table 11) show that the MEE is

affected most by the amount of finish (coverage) applied to

the wood surface. The greater the amount applied (the lower

the coverage in ft

/gal), the greater the MEE. We found we

could not consistently apply equal amounts of finish to the

surfaces by brushing. In one case (ponderosa pine) the

finish was applied at lower coverage for the small (3 x 5 in.)

versus the large (6 x 10 in.) specimens. For three others

(southern pine, red oak, flakeboard) the opposite was true.

Only in the Douglas-fir plywood specimens were relatively

equal quantities of finish applied to the surfaces. In this case

the MEE

found for the finish was only slightly higher for

the large specimens than for the small specimens (for

convenience only MEE

values are shown in table 11, other

MEE values were similar).

The overriding effects on MEE of coverage or amount of

material applied to the wood surface are illustrated in

figure 10. The MEE for each substrate for one, two, or three

coats of finish is plotted against the total coverage of the

finish. Each set (one, two, or three coats) shows a fairly

close relationship between MEE and total or cumulative

coverage (individual coverage values were added together

for two- and three-coat applications). Assuming a linear

relationship between MEE and coverage, regression

analysis gave squared correlation coefficient (R

) values for

one, two, and three coats of 61, 48, and 47 percent,

respectively.

Table 11.—Effect of specimen size and finish coverage on the moisture-excluding effectiveness (MEE

) of finish 67 on dif-

ferent wood substrates after 14 days exposure at 90 percent relative humidity

Coverage

Wood subtrate

Specimen

Size

1 coat

2 coats

3 coats

1st

2nd 1st 2nd 3rd

MEE

1 coat 2 coats 3 coats

In.

----------------------Ft

/gal-----------------------

--------------Pct--------------

3 x 5

429

413

672

397

608

534

56 77

6 x 10

575

586

667

575

628

759

3 x 5 560 512 845

497 853 776

6 x 10 419

417

495 390

472

493

73 85

3 x 5

498

513

711

483

760

830 52

70 76

6 x 10

415 403

437

422

436

453 60 80 86

3 x 5

342 344

525

348 525

471

74 83

6 x 10

331

345

424

325

443

510

77 87 90

3 x 5

486 498

787

449

712

819

6 x 10

297 316

396

327

386 437

83 86

Ponderosa pine

Southern pine

Red oak

Douglas-fir plywood

Flakeboard

Figure 10.—Linear regression lines for

moisture-excluding effectiveness (MEE) of 1, 2, and

3 coats of a soya-tong alkyd enamel (finish 67) as a

function of coverage. For each coat number, each

point derives from a different substrate.

Test Procedures, Materials, Methods

Effect of Density of the Wood Specimen

In general, all other factors being constant, the higher the

density of a wood specimen, the higher will be the nominal

MEE of the finish applied to it. This dependence is roughly

linear but changes with time from treatment–the slope gets

steeper with the lapse of time. The dependence of MEE on

density also weakens as the number of coats increases.

We investigated the relationship between wood density and

MEE with scatter plots. Using a simple statistical model, we

made an adjustment for the density of the wood specimens.

Details of this adjustment are given in Appendix C.

However,

when we ranked the MEE values adjusted for the density of

the wood specimens used in this study, we found only minor

differences from the ranking of unadjusted values. For

example, ranked by either type of MEE, the same finishes

fell in categories excellent (>75 pct MEE) and good (50 to

75 pct MEE). There was one switch from fair (25 to 50 pct

MEE) using unadjusted MEE’s to poor (<25 pct MEE) using

adjusted MEE’s. To reduce the effect of density our

experimental procedure called for three wood specimens as

replicates, one each of low, medium, and high density, for

each finish coat combination. This procedure seems to have

been adequate.

Variability of MEE

The variation in MEE comes from several sources. Variation

in density and variation in finishing are two major sources

easy to identify. The contribution of density varies from finish

to finish and, within finish, from one coat to two coats to

three coats.

The finishes that show relatively little variance may be

particularly impermeable to water. This seems likely for three

coats of finishes 6, 63, and 64; for these, wood density

should have little effect on MEE (see the right-hand column

of table 3) The other identifiable component of MEE

variability, coverage (equivalently, volume of finish applied),

can decrease MEE variability as well as increase it. From

plots and summary statistics, it seems that the small

variation in coverage seen in this study usually has a small

impact on MEE when only one substrate is considered. (In a

few cases, notably finish 41, the relatively large amount of

finish applied to the least dense specimens for one coat and

three coats may reduce the variability that one might expect

from considering only the effects of density.)

The MEE of modern surface finishes (paints, lacquers,

varnishes, etc.) and intrasurface coatings (water repellents

and wood sealers) was evaluated on clear ponderosa pine

(

Pinus ponderosa) sapwood. This substrate was used to

determine the rate of vapor transmittance for all the finishes

and combination of finishes used in these studies, and was

the reference medium for extensions to other substrates. A

diversified selection of commercially available finishes was

evaluated.

Specimens

The degree of protection obtained from a finish as a vapor

retarder is governed by the size and species of the

specimen (3,4,12). The effects on MEE of nine wood

substrates and two sample sizes were evaluated. All solid

wood specimens were cut from flat-grained lumber of

differing densities. A representative pair from each of three

density ranges (high, medium, and low) was used to

evaluate each coating.

Standard Specimens

Ponderosa pine sapwood was chosen as a representative,

commercially available softwood substrate. The 3-inch

(tangential) by 5-inch (longitudinal) by 5/8-inch (radial) clear

specimens had the corners, ends and edges rounded to a

1/4-inch radius by routing and then sanding. This shape

decreased application difficulties and damage due to

handling of the finished surfaces (fig. 1).

Specimens from Other Substrates

Nine wood substrates were used to evaluate differences

between species of a few selected finishes. The specimens

were cut from the following substrates. (Note exceptions to

the standard ponderosa pine dimensions for some of the

wood composite products):

1. Douglas-fir plywood.–Commercial, exterior-grade

Douglas-fir (

Pseudotsuga menziesii

) heartwood plywood

(3 x 5 x 5/8 in.).

2. Hardboard.–Commercial, sanded, exterior-grade

hardboard (3 x 5 x 3/8 in.).

3. Particleboard.–Commercial, exterior-grade, softwood,

particleboard (3 x 5 x 5/8 in.).

4. Flakeboard.–Structural flakeboard manufactured from

Douglas-fir at the Forest Products Laboratory, Madison,

WI (

13) (3 x 5 x 1/2 in.).

5. Southern pine.-Southern pine (Pinus sp.) sapwood

(3 x 5 x 5/8 in.).

6. Western redcedar.–Western redcedar (Thuja plicata)

heartwood (5 x 5 x 5/8 in.).

7. Ponderosa pine.–Ponderosa pine (Pinus ponderosa

)

sapwood (3 x 5 x 5/8 in.).

8. Red oak.–(Quercus sp.) heartwood (3 x 5 x 5/8 in.).

9. Hard maple.-Hard maple (Acer sp.) heartwood

(3 x 5 x 5/8 in.).

Large Specimens

Southern pine, red oak, Douglas-fir plywood, flakeboard, and

ponderosa pine as described above were cut into 6-inch

tangential by 10-inch longitudinal specimens with respective

radial measurements. The corners, edges, and ends were

rounded to a 1/4-inch radius by routing and sanding.

Finishes

A wide variety of wood finishes, both commercially available

and laboratory prepared, were evaluated in this study.

Appendix A and (8) list the individual finishes with

classification by manufacturer according to use. Composition

details are shown in both Appendices and in (8).

Combinations of these individual finishes were also

evaluated (table 7).

Equipment

Belt Sander

An 18-inch belt sander equipped with 100-grit cloth belts

was used for the preliminary sanding of all specimens.

Exposure Racks

To achieve optimum vapor transmittance, it was necessary

to use exposure racks in which the specimens were able to

hang freely. The racks were made of 5/8-inch plywood to fit

the appropriate size specimen. Each rack held

18 specimens (fig. 1). Three pairs of specimens

(pair = finished plus unfinished, end-matched control) for

each one-, two-, and three-coat applications of that finish

were hung 5/8-inch apart on screw eyes.

Relative Humidity Rooms

Specimens were conditioned, tested, and weighed in three,

walk-in, humidity-temperature rooms; 30 percent RH/80 °F;

90 percent RH/80 °F; and 44 percent RH/72 °F.

Balance

An electronic balance with a 1,200-gram capacity was used.

This balance was maintained in the 44 percent RH/72 °F

room where the specimens were weighed to the nearest

0.01 gram. Specimens were kept in polyethylene

bags while being transported from the 30 and 90 percent RH

rooms to the 44 percent RH weighing room.

Methods

Specimen Preparation and Selection

Specimens cut from both lumber and sheets of wood

composite products were prepared in the same manner.

Each board or sheet was abrasive planed with 50- then

80-grit paper, ripped longitudinally into 3-inch or 6-inch strips

that were then cut tangentially into their respective lengths,

5 or 10 inches. The comers, edges, and ends of all

specimens were routed to a 1/4-inch radius using a steel

carbide router. A 3/32-inch hole was drilled in the center top

of each specimen. Final sanding of the specimens began

with one quick pass of the face surface and rounding of the

routed edges and ends of each specimen on a belt sander.

The smoothing of edges, ends and corners was done by

hand using 120-grit paper. All specimens were vacuumed

using an industrial-size vacuum with a brush attachment,

tied into bundles according to board designation, and with

end grain exposed preconditioned at 30 percent RH/80 °F

for a minimum of 4 weeks. The average conditioned weight

per specimen per board was recorded.

For each finish in test, three pairs of end-matched

specimens were selected from boards with different

densities for each of one-, two-, and three-coat applications.

This yielded a total of 18 specimens per finish tested. The

three pairs were selected with as large a weight difference

as possible to guarantee that each would represent a

different range of density being tested (high, medium, or

low).

Finish Application

An aluminum hook was screwed into the predrilled hole at

the top of each preconditioned specimen following selection

for test. One specimen from each pair was labeled as the

control and placed on an exposure rack for further

conditioning. The other member of the pair was labeled by

attaching a cardboard tag to the hook, wrapped in a plastic

bag, and taken to the laboratory for finishing.

A 1-1/2-inch nylon or natural bristle brush was used to seal

the specimen completely with the finish while holding it by

the hook. All ends and edges were coated first and the

excess material worked into the face surfaces. Uniform

brushing was maintained. Wet weights applied (±0.01 g)

were determined by weighing the brush and finish container

before and after application. A small selection of finishes

was also applied using a 30-second dip. Excess finish was

allowed to drip off the specimen and back into the dipping

receptacle. Approximate wet weights applied were again

recorded. The average finish coverage par coat for three

specimens in ft

/gal was calculated (table 1).

The finished specimens were air dried in a laboratory hood

while hanging on an exposure rack for a minimum of

24 hours before applying the next coat or returning them to

30 percent RH/80 °F.

Conditioning, Cycling, and Recycling

Preliminary testing was done to determine the type and

extent of exposure to be used in this study. Prior to test, the

finished specimens and their matching controls were brought

to EMC at 30 percent RH/80 °F. The three pairs of

specimens per number of coats for a finish (total of six

specimens) were always tested simultaneously. Preferably,

one-, two-, and three-coat applications of each finish (total of

18 specimens) were tested together. This procedure

eliminated the slight variability in humidity room conditions

(±1 pct) for comparing the effectiveness between coats. To

minimize additional error, the specimens were individually

wrapped in plastic freezer bags for transportation from

humidity room to humidity room, thereby preventing

exposure to the uncontrollable ambient conditions. The

exposure racks used in test were transferred from one

humidity room to another along with the bagged specimens.

As the same specimens were recycled, identical techniques

were used.

Weightng Techniques

Weights recorded in this study were taken from an electronic

balance in a 44 percent RH/72 °F room. Exposure of the

specimens to these conditions was brief. Weights were

taken at days 1, 7, and 14 from the time the specimens

entered each humidity.

Summary and Conclusions

Tests

Original and Noncontinuous Cycles

The standard and the large specimens of every substrate

were all exposed to 90 percent RH/80 °F until MEE was

50 percent or less while in the adsorption phase of the first

test cycle. This procedure was designated the original test

cycle. The specimens were weighed after 1, 7, and 14 days

without exception and every 7 days thereafter while at

90 percent RH/80 °F until MEE reached 50 percent. Once

this MEE was reached, the specimens were returned to

30 percent RH/80 °F for re-equilibrating.

Various finishes that performed moderately well to very well

on the standard specimens in the original test cycle (table 9)

were selected and recycled through two more adsorption/

desorption cycles. In each cycle, the specimens were first

equilibrated at 30 percent RH/80 °F, then transferred to

90 percent RH/80 °F for 14 days of adsorption and

immediately returned to the 30 percent RH/80 °F for 14 days

of desorption, followed by re-equilibrating. This form of

recycling, equilibrating-testing-equilibrating, is referred to as

noncontinuous cycling of specimens.

Continuous Cycles

To monitor the effects of continual humidity changes, three

new pairs of duplicate standard specimens for each of one,

two, and three coats were prepared using one of the

following finishes: an unpigmentad white shellac (finish 23),

a pigmented enamel (finish 77), or an aluminum-pigmented

paint (finish 65). The specimens were tested for three

continuous cycles of 14 days of adsorption at 90 percent

RH/80 °F followed by 14 days of desorption at 30 percent

RH/80 °F, followed immediately with the next adsorption

cycle at 90 percent RH. The specimens began the cycling

after being brought to EMC at 30 percent RH/80 °F.

This report presents results of investigations made to

increase knowledge of the degree to which wood can be

stabilized in MC with protective finishes. Wood finishes are

generally regarded as inefficient stabilizing devices because

they merely retard the rate of change in MC without

changing the equilibrium that will finally be reached under

any given conditions of temperature and humidity. Our work

demonstrates that some effective modern finishes limit the

changes in MC of wood to a fraction of the changes that

take place in unprotected wood.

These studies deal with wood while it is undergoing changes

in MC. They are not directly concerned with the equilibrium

MC of wood in an environment of constant dampness or

dryness because finishes have no effect on such

equilibrium. Our interest was in how long it takes the wood,

with and without finishes, to change from the initial to the

final equilibrium MC or to change some fraction of the

difference between the two equilibriums, when the dry wood

was exposed to damp air. In other words, we are concerned

with variable state not with steady-state conditions. Under

variable-state conditions, some factors that are relatively

unimportant under steady-state conditions become

significant; size of test specimens is an example; the

permeability of the finish taken by itself is not enough to

determine the results obtained.

The moisture-excluding effectiveness described in our

studies is a measure of the moisture gain (from water vapor)

of wood protected on all sides by the finish, as compared to

the moisture gain of unfinished wood, when both pieces of

wood are exposed to a controlled atmosphere of 90 percent

RH/80 °F for a given time after being brought to equilibrium

at 30 percent RH/80 °F. For example, compared with the

unprotected wood, three coats of an aluminum

flake-pigmented varnish on ponderosa pine wood were

84 percent effective in controlling moisture vapor movement

after 14 days; after 60 days, the MEE was still 51 percent. In

contrast, one coat of the pigmented varnish was only

41 percent effective after 14 days while a three-coat

penetrating water-repellent finish was only 11 percent

effective after 14 days.

The most effective moisture-resistant finishes found for wood

were two finishes not usually considered for use on wood.

The first, an epoxy sheathing compound (two-component) is

essentially an adhesive consisting of 100 percent solids.

Most of the commercial finishes used in our studies had

some solvent or dispersant ranging from 20 to 50 percent of

the weight of the finish. The second finish was molten

paraffin wax applied to the wood by dipping (brushing was

also used but was not as effective). One coat of the epoxy

finish gave an MEE of 54 percent after 14 days; three coats

gave an MEE of 91 percent after 14 days, and 70 percent

after 60 days. The paraffin wax finish gave an MEE of

95 percent after 14 days and 79 percent after 60 days.

When brushed on, the paraffin wax effectiveness was

69 percent after 14 days and 27 percent after 60 days.

With the exception of the two finishes described above, both

of which were unpigmented, the wood finishes found most

effective in excluding moisture were pigmented products with

nonaqueous solvents or dispersants in their compositions

such as mineral spirits, alcohol, or turpentine. The MEE of a

transparent finish was improved markedly by the addition of

pigments. For example, a polyurethane varnish

(one-component) was improved by adding 2 lb/gal of an

aluminum flake pigment; its three-coat MEE, originally

44 percent after 14 days, rose to 84 percent after the

addition. Corresponding MEE values for an unpigmented

and pigmented shellac were 42 and 73 percent.

The most effective of the commercially available pigmented

finishes were two-component epoxy enamel paints,

aluminum flake-pigmented varnishes (already mentioned), a

soya-tung alkyd enamel, and a soya-linseed alkyd enamel.

These studies illustrate the MEE of many commercially

available finishes for wood ranging from transparent,

penetrating-type finishes (sealers, waxes, oils) to pigmented

film-forming ones (paints, sealer shellacs, primers). The

effects of cycling humidity conditions and different wood

substrates are illustrated. Time effects and coating thickness

are discussed in detail.

It is clearly shown that any use of paints and other finishes

as a moisture vapor retarder for wood must take exposure

conditions into account. Good moisture barriers may almost

completely insulate wood from short-cycle humidity

variations, and at the same time be ineffective against

long-term seasonal cycles. Time is an extremely important

factor in determining not only the extent but also the

character of the response of wood to humidity change.

Few seem to realize the effect that finishes may have on

wood warping. In many kinds of goods, the front surface of a

wood panel is finished with a coating very effective against

moisture movement while no attention is paid to the back or

even the edges. When moisture changes take place, the

gain or loss of moisture is much greater on the unprotected

side than on the finished side and, if the changes occur

rapidly, warping is almost sure to result. Such difficulties

could be avoided by applying to the back and edge of the

wood any coating that will balance the front coating in

moisture resistance. This practice would not only reduce the

rate of change in MC but permit better equalization of

adsorbed moisture. The results of these studies should help

in the selection of finishes, whether for the similarity of their

moisture vapor resistance or for very high

moisture-excluding effectiveness.

Acknowledgments

Literature Cited

Edward A. Mraz (retired) and Peter G. Sotos, physical

science technicians at Forest Products Laboratory, gave

helpful suggestions and technical assistance in this study.

Appendix A

Description of Wood Finishes Used in

These MEE Studies

The information in table Al is as supplied by the manufac-

turer on the container label. A complete description of fin-

ishes used in these studies is available (8).

Table A1—Finish numbers and description

Finish Description

EXTERIOR, UNPIGMENTED, NONAQUEOUS

(4 FINISHES TESTED)

Two-component, polyurethane, clear gloss finish

Phenol-formaldehyde/linseed-tung wood sealer

Water repellent (9)

Polymeric roof coating

COMBINATION EXTERIOR/INTERlOR, UNPIGMENTED,

NONAQUEOUS (12 FINISHES TESTED)

Modified butyl-acrylo-styro epoxy finish

Two-component sheathing epoxy (adhesive)

Polyurethane gloss varnish

Tung oil

Two-component polyurethane sealer

Linseed oil

Linseed oil in mineral spirits

Polyurethane gloss varnish

Soya alkyd/phenolic/tung gloss varnish (spar)

Polyurethane gloss varnish

INTERIOR, UNPIGMENTED, NONAQUEOUS

(19 FINISHES TESTED)

Lemon oil furniture polish with silicone

Nitrocellulose semigloss lacquer

Polyurethane satin varnish

Soya alkyd/maleic/China wood oil gloss wood finish

Linseed alkyd/maleic/China wood oil satin wood finish

Orange shellac

White shellac

Epoxy gloss varnish

Phenolic/tung floor sealer

Linseed/phenolic/tung floor sealer

Soya alkyd gloss wood finish

Alkyd satin wood finish

Spray furniture polish (lemon creme, wax, and silicone)

Paraffin wax

Wallpaper sealer

Brazilian Carnuba paste wax

Soya epoxy gloss floor and trim sealer

Polyurethane satin varnish

Nitrocellulose/alkyd lacquer

Table A1—Finish numbers and description—con.

Finish

Description

EXTERIOR, UNPIGMENTED, AQUEOUS

(NO FINISHES TESTED)

COMBINATION EXTERIOR/INTERIOR, UNPIGMENTED,

AQUEOUS (NO FINISHES TESTED)

INTERIOR, UNPIGMENTED, AQUEOUS

(4 FINISHES TESTED)

Acrylic gloss varnish

Acrylic satin varnish

Alkyd varnish

Acrylic gloss wood finish

EXTERIOR, PIGMENTED, NONAQUEOUS

(19 FINISHES TESTED)

Aluminum-pigmented polyurethane gloss varnish-l

Aluminum-pigmented polyurethane gloss varnish-2

Aluminum-pigmented polyurethane gloss varnish-3

Aluminum-pigmented polyurethane gloss varnish-4

Two-component polyurethane gloss paint

Soya alkyd flat marine enamel

Soya alkyd gloss marine enamel

Soya alkyd semigloss marine enamel

Alkyd (monopoxy) gloss enamel

Soya-linseed alkyd flat undercoat paint

Semitransparent linseed oil-based stain (2)

Semitransparent linseed oil-based stain

Soya alkyd/linseed flat paint

Soya alkyd flat primer paint

Soya/silicone alkyd gloss enamel

Semitransparent linseed oil-based stain

Solid color linseed oil-based stain

Tall alkyd/soya alkyd gloss house paint

Tall maleic alkyd/soya alkyd flat primer paint

COMBINATION EXTERIOR/INTERIOR, PIGMENTED,

NONAQUEOUS (17 FINISHES TESTED)

Epoxy resin gloss paint

Pigmented Hat shellac

Sealer-primer flat finish

Phenolic alkyd floor and deck gloss enamel

Two-component epoxy/polyamide gloss paint

Two-component epoxy/polyamide satin paint

Aluminum-pigmented ester gum/vegetable oil metal and ma-

sonry paint

Aluminum-pigmented petroleum resin utility paint

Soya-tung alkyd satin enamel

Soya-linseed alkyd gloss floor and deck enamel

Two-component epoxy/polyamide gloss enamel

Pigmented sealer-primer flat shellac

Linseed-phenolic/menhaden-phenolic aluminum paint

Linseed-menhaden alkyd gloss paint

Two-component polyurethane gloss enamel

Soya alkyd gloss enamel

Primer/sealer paint

Appendix B

Composition of Finishes

Table A1—Finish numbers and description—con.

This listing provides quantitative measures of the

Composition of the 91 finishes investigated in this study,

Finish Description

which are grouped according to their finish characteristics.

INTERIOR, PIGMENTED, NONAQUEOUS (3 FINISHES TESTED)

Table B1.—Composition of finishes

76 Soya alkyd flat paint

77 Soya alkyd gloss enamel

Solids

Soya-linseed alkyd semigloss enamel

Total

Finish

Surface

Finish

Manu-

pigment

resin

oil

density

Meas-

fac-

EXTERIOR, PIGMENTED, AQUEOUS (6 FINISHES TESTED)

Acrylic latex flat primer paint

Acrylic latex Hat house paint-1

Acrylic latex flat house paint-2

Acrylic latex flat house paint-3

Acrylic latex solid color stain

Acrylic latex/soya alkyd flat house paint

COMBINATION EXTERIOR/INTERIOR, PIGMENTED, AQUEOUS

(2 FINISHES TESTED)

Acrylic latex/epoxy ester concrete floor paint

Pigmented acrylic shellac primer

INTERIOR, PIGMEMED, AQUEOUS (5 FINISHES TESTED)

Vinyl acetate-acrylic latex enamel undercoat

Acrylic latex satin enamel

Acrylic latex flat enamel

Butadiene-styrene latex flat primer paint

Vinyl acrylic latex flat wall paint

ured

turer

Lb/gal

------------------Pct------------------

EXTERIOR, UNPIGMENTED, NONAQUEOUS

Gloss

8.22

43.6 34.7 0.0 34.7

0.0

Satin

7.14

20.3

22.5

0.0

10.4

12.1

—

6.71 12.6 10.9

0.0 9.8

0.0

— 7.52 31.5 38.4 6.9 26.2 0.0

COMBINATION EXTERIOR/INTERIOR, UNPIGMENTED, NONAQUEOUS

Flat

8.39

29.0

—

0.0

29.0

0.0

Goss

8.80

99.0

—

0.0 99.0

0.0

Gloss

7.51

50.7

47.5

0.0

47.5

0.0

—

7.79 99.9

99.9 0.0

0.0

99.9

Gloss 8.24

46.2

40.0

0.0

40.0

0.0

Gloss

7.98 40.7

40.0

0.0

40.0

0.0

— 7.78 99.9

99.9 0.0

0.0 99.9

—

7.11 52.1 50.0

0.0 0.0 50.0

Gloss 7.63 53.5 49.5 0.0 49.0 0.0

Gloss 7.55 59.0 54.3 0.0 47.6 6.7

Gloss 7.37 55.3 45.6 0.0 45.3 0.0

Gloss 7.52

54.1

52.0 0.0 52.0 0.0

INTERIOR, UNPIGMENTED, NONAQUEOUS

—

———

0.0

—

Satin 7.59

19.0

21.5

0.0 21.5 0.0

Satin

7.74 51.3 44.8

0.0

44.8 0.0

20 Gloss

7.43 43.9 40.3

0.0

38.5

1.8

Satin

7.75

44.7

39.1

8.0

28.4 2.7

22 —

7.54

31.0 30.7 0.0

30.7 0.0

23 —

7.52

30.0 30.7 0.0 30.7 0.0

GIoss

7.18 42.7 40.8

0.0 39.6

0.0

25 — 7.04 37.2 33.6 0.0

13.3

20.3

26 — 7.32 52.0 48.3

0.0 41.1 7.2

27 Gloss

7.50

43.1

41.5

0.0

41.5 0.0

Satin 7.20 33.7 29.5

7.6

21.9 0.0

—

4.5 —

0.0

——

—

30 — — 99.9 0.0 0.0 0.0

—

8.47

19.0

—

0.0

——

—

———

32 — 0.0

Gloss 7.02

30.4 31.5

0.0

30.0 0.0

Satin

7.41

46.7 45.0

2.0

43.0 0.0

35 GIoss

7.60

25.5 31.0

0.0

25.4 2.5

Table B1.—Composition of finishes

—con.

Table B1.—Composition of finishes

—con.

Solids

Finish

Surface

Finish

Manu-

Total

Total Total

Finish

Surface

Finish

Manu-

Total

density

Meas-

fac-

pigment

resin

oil

ured

turer

Lb/gal

---------------------Pct---------------------

INTERIOR, UNPIGMENTED, AQUEOUS

36 Gloss 8.63 36.7 28.2 0.0 28.2 0.0

37 Satin 8.75 38.0 29.5

1.5

28.0 0.0

38 Gloss

9.33 44.4 44.0 4.0 40.0 0.0

39 Gloss

8.88 39.6 37.0 0.0 33.0 0.0

EXTERIOR, PIGMENTED NONAQUEOUS

Gloss

8.32 63.3 54.0

13.8 40.1

0.0

Gloss

8.29 63.8 54.0

13.5

39.2 0.0

Gloss 8.28 62.6 54.0

13.5

39.2 0.0

Gloss 8.30 64.5 54.0

13.6

39.5

0.0

44 Gloss

10.32 63.5 52.1

42.1

29.4 0.0

Flat

10.58

66.4 61.4

42.1 19.1

0.0

Gloss

9.66 69.2 65.7

28.6

36.7

0.0

47 Satin

10.73

71.8

70.2

40.5

29.4 0.0

Gloss

9.74 63.2

61.1

30.0

29.2 0.0

Flat

11.74

75.3 70.9 52.7

16.6 0.0

— 7.96 75.5 75.7

8.4

0.0 61.0

— 7.90 77.9 — —

——

Flat

10.57

73.0

62.1

42.2

13.5

6.4

Flat

10.61

70.0 63.7 44.0

19.7

0.0

Gloss 9.49

65.1

62.5 25.6

35.1

0.0

—

6.91

20.5 — —

— 9.48 59.6 — —

——

Gloss 9.10

67.3 65.0 26.0 38.0

0.0

Flat

11.40 78.5 74.0

45.0

28.0 0.0

COMBINATION EXTERIOR/INTERIOR, PIGMENTED, NONAQUEOUS

Gloss 9.69 43.0

41.1 18.1

23.0 0.0

Flat

9.92

53.7

53.2 31.7

21.5 0.0

Flat 10.14

63.3 71.2 40.0

30.0

0.0

62 Gloss

7.88

57.0

83.4 24.0

59.4 0.0

63 Gloss

10.27

64.1

61.0 28.3 32.7 0.0

64 Satin

10.66

65.6 65.0 28.3

32.7

0.0

Satin 8.09 61.0

59.2 19.3

4.8

35.1

66 Gloss 7.97

56.0 57.4

14.4

43.0 0.0

67 Satin

12.36

79.3 73.9

50.8

23.1

0.0

Gloss

9.11

66.8 56.4 24.9 31.5 0.0

Gloss

10.25

63.0

58.4

24.1

34.3

0.0

70 Flat 9.90 52.8

53.4

34.0

19.4

0.0

Gloss

7.86

47.0 49.3

13.0

36.3

0.0

Gloss

9.23

60.0 60.9 26.7

34.2 0.0

Gloss

10.10

63.7

—— —

0.0

Gloss

9.45

67.3 61.5 26.9 34.6 0.0

Flat

10.30

55.6

——

density

Meas-

fac-

pigment

resin oil

ured

turer

Lb/gal

---------------------Pct---------------------

INTERIOR, PIGMENTED, NONAQUEOUS

Flat

12.46

72.2 69.2 61.5 7.7 0.0

Gloss

10.39 71.1

69.5 34.6 34.9 0.0

Satin

10.74

68.3 69.0 46.2 22.8 0.0

EXTERIOR, PIGMENTED, AQUEOUS

Flat 9.76 52.0

51.0

21.0 28.0 0.0

Flat

10.93

53.0

50.0

29.0

19.0

0.0

flat

11.02

58.7

57.2

35.6 21.6 0.0

Flat

11.28

59.1

57.5

37.4

20.1

0.0

83 —

10.43 45.1

—

———

Flat 11.27 52.2

50.0 33.0

17.0

0.0

COMBINATION EXTERIOR/INTERIOR, PIGMENTED AQUEOUS

Satin

10.76

56.0 49.4 27.9 21.5 0.0

86 Flat

10.65 52.3

50.0 29.6

20.4

0.0

INTERIOR. PIGMENTED, AQUEOUS

Flat

11.20

51.4 55.7 41.3

14.4

0.0

Satin

10.60

51.1

49.2 26.6 22.6 0.0

Flat

11.03

53.5 48.0

28.2 19.8

0.0

90 Flat

10.67

56.0 57.0 31.9

23.1

0.0

Flat

11.09

51.0 49.5 38.0

11.5

0.0

For names of finishes and additional composition information see(8 ).

Satin = Semigloss.

Appendix C

Adjustment of MEE

Values for Wood Density

If an investigator wishes to compare finish performance

precisely yet knows that the effect of wood density has not

been controlled, then one way to take account of density is

by an analysis of covariance. A simpler approach, in the

same spirit as analysis of covariance, involves inserting an

emperical factor in the formula for MEE. For example, the

small ponderosa pine specimens all have the same volume.

We inserted a correction factor in the change-in-weight ratio

that appears in the MEE formula. The unadjusted MEE

(percentage) at time t is given by:

MEE

= (1 - ∆W

/∆

) x 100

where

∆

is the change in weight of the finished specimen

between time zero and time t and

∆

is the corresponding

change in weight of the matched control specimen. If

∆

and A = W

UTo

ref

where W

UTo

is the

equilibrium weight of the control specimen at time zero and

ref

is some reference weight for the uncoated specimens,

then adjusted MEE (percentage) is given by:

MEE

adj·t

= (1 - R

·A) x 100

For a reference weight, we used W

Uto

, the average at time t

of the 852 uncoated specimens in this part of the study.

Figures 11 and 12 are scatter plots respectively of the

unadjusted MEE

values plotted against wood density

(three replicate specimens) for each of 9 finishes.

Comparison of the figures shows that adjustment eliminates

much of the dependence of MEE on density.

A simple linear relationship (with average = a and slope = b)

seems to characterize the dependence of ordinary MEE on

density when we exclude the poor finishes. We used this

equation to relate MEE

to density d for k coats:

MEE

tkij

tki

+ b

tokij

-d

tok

)

where i is finish number and j is replicate number (1, 2, or

3). Then MEE

tkij

is MEE

with k coats of finish i for the j-th

replicate; a

tki

is the average MEE

with k coats of finish i

when the corresponding density is at its average value; b

the slope of MEE

on “centered” density at time t with k

coats; d

tokij

the density at time zero of the control

specimen before k coats of the i-th finish are applied to the

j-th replicate matching treated specimen; d

tok

is the average

density at time zero over all finishes and replicates having k

coats. Specimens with any good finish of k coats at time t

fall on one of a set of parallel lines when MEE is plotted

versus wood density. The particular finish of the specimen

determines on which line the specimen falls.

Table Cl shows the results of fitting the above equation to

data from good finishes. (The table gives our definition of

the “good” finishes used in fitting the equation. Thus, a good

l-coat finish has MEE

at least 25 pct for all three

replicates.) As the equation indicates, we fitted a separate

regression for each time and coat combination, for time = 1,

7, and 14 days and coats = 1, 2, and 3. The number of

observations used depends on the number of good finishes

for each coat, and thus varies with coat number. For

example, for 1 coat, we have 13 good finishes, and three

replicates for each finish; thus we have n = 39 observations

for our regression for t = 1 day, t = 7 days, and t = 14 days.

Each yields the same picture: 13 noncoincident parallel

levels. With these 13 levels plus one common slope we

have p = 14 parameters. Because p/n is about 1/3 in each

of the regressions, we report an R

that is adjusted for the

relatively large number of parameters.

The table shows two patterns involving the b

defined in the

above equation. First, for any particular number of coats k,

the slope b

gets steeper as time increases from t = 1 to

t = 14 days. This pattern follows from the definition of MEE

for any particular good finish, the ratio R

tends to be larger

for low-density specimens and smaller for high-density

specimens. We see this tendency in R

because the

denominator W

increases linearly with density while the

numerator W

remains relatively unaffected by density.

Second, for any particular time t, the common slope gets

smaller as the number of coats increases. This pattern

makes sense because the thicker the finish layer, the less

effect the wood specimen has as a sponge to draw in

water-poor sponges and good sponges alike will be

hindered by a thick coating.

The values of R

reported in table Cl should be viewed with

a bit of caution. If there were no relationship between wood

density and MEE but very good finishes were applied to

high-density specimens and merely good finishes were

applied to low-density specimens, one could still obtain a

high R

. We examined plots and regressions of MEE versus

density for each finish and for days 1, 7, and 14. The

general comments made above about the common slope

regression models appear valid in light of these individual

finish results.

Figure 11.—Moisture-excluding effectiveness (MEE) of 9 finishes on ponderosa

pine sapwood (specimens of low, medium, and high density) after 14 days at

90 percent relative humidity and 80 °F.

Figure 12.—Adjusted moisture-excluding effectiveness (MEE) of 9 finishes on

ponderosa pine sapwood (specimens of tow, medium, and high density) after

14 days at 90 percent relative humidity and 80 °F.

Table C1.—Regression of moisture-excluding effectiveness (MEE) on wood density for three

thicknesses of coat and three times of exposure

1 Coat

Slope = b

(standard R

adj

2 Coats

3 Coats

Slope = b

(standard

adj

(standard R

error)

Duration

of test

error)

Days

0.34

1.94

0.14

0.45

0.10

0.36

(0.158)

(0.025)

(0.022)

1.80 80 3.92

1.00

1.53

0.69

97 0.98

(0.320)

(0.084)

(0.058)

2.46

4.34

1.67

2.45

(0.355)

1.13

1.50

(0.135)

(0.010)

1 coat: 39 observations, 3 replicates of 13 good finishes (MEE >

25 pct for every replicate, excluding finishes 1 and 6)

2 coats: 60 observations, 3 replicates of 20 good finishes (MEE >

37 pct for every replicate, excluding finish 6)

3 coats: 63 observations, 3 replicates of 21 good finishes (MEE >

50 pct for way replicate, excluding finishes 1 and 6)

Adjusted correlation coefficient (R

) is related to ordinary R

by this equation:

adj

= (n - 1)R

/(n - p) - (p - 1)/(n - p)

where n = number of observations in the regression and p = number of parameters fitted.

is the square root of the mean square error of regression and gives a summary measure of variability of regression. Values of s within the same column

are directly comparable.

3.0-12/85

The Forest Products

Laboratory (USDA Forest

Service) has served as the

national center for wood

utilization research since

1910. The Laboratory, on the

University of Wisconsin-

Madison campus, has

achieved worldwide

recognition for its

contribution to the knowledge

and better use of wood.

Early research at the

Laboratory helped establish

U.S. industries that produce

pulp and paper, lumber,

structural beams, plywood,

particleboard and wood

furniture, and other wood

products. Studies now in

progress provide a basis for

more effective management

and use of our timber

resource by answering critical

questions on its basic

characteristics and on its

conversion for use in a variety

of consumer applications.

Unanswered questions remain

and new ones will arise

because of changes in the

timber resource and

increased use of wood

products. As we approach the

21st Century, scientists at the

Forest Products Laboratory

will continue to meet the

challenge posed by these

questions.