Effective adhesive systems

and optimal bonding

parameters for hybrid CLT

Blake Larkin

Oregon State University

Materials Science and Wood Science & Engineering

Co-authors: Lech Muszynski, Arijit Sinha, Andre Barbosa, Rakesh Gupta, and Vahid Mahdavifar

Presentation Overview

• Brief intro to cross laminated timber (CLT)

• Hypothesis

• Approach

• Screening study goals

• Preliminary results

• Further Research

Cross Laminated Timber (CLT)

What is

CLT ?

http://www.prosalesmagazine.com/products/lumber/x-marks-the-opportunity_o

Layers: Typically 3, 5, or 7

Area: Up to ~ 10 X 60 ft2

http://www.internationaltradenews.com/klh_massivholz_gmbh/portrait/ http://www.structurearchives.org/article.aspx?articleID=1474

"KLH: Sustainability." KLH UK. Web. 30 Oct. 2014. <http://www.klhuk.com/sustainability.aspx>.

Our Hypothesis is that…

…CLT panels with hybrid layups (layers arranged from

high- and low-grade lamellas or composed of mixed

species) can meet the critical performance parameters

specified in the ANSI PRG 320 2012

and

that adhesive systems alternative to polyurethane (PUR)

can be successfully utilized in hybrid CLT products.

Approach

• Screen for species/grade/adhesive

combinations that pass bond integrity

criteria (varying pressing parameters)

Test viable combinations for:

• engineering performance criteria

• connector performance

Goal of Screening Study

1. Check the compatibility between standard sampling scheme and the inherent variability of the bond integrity characteristics in the product

2. Adjust the standard sampling scheme (if needed)

3. Screen the combinations using the adequate sampling scheme

Screening Study

Testing 2 x 2 foot panels with 3 layers

Qualification Standard: ANSI/APA PRG 320-2012

• Bond line shear

• Delamination

Screening Study Test VariablesPressure (MPa) [psi]

0.69 [100]

0.40 [58.0]

0.10 [14.5]

Core LayerLodgepole Pine (Low)

Douglas-fir (Low)Hemlock (Low)

Lodgepole Pine (High)

Adhesive

Types

PRF

PUR

Face LayersDouglas-fir (High)

Lodgepole Pine (High)

4 panels per

combination

Bond Line Shear Test

Modified ASTM 1037

• Modified for 90 o grain orientation between layers

• 2x1.5 inch bond area

• 1 inch thick layers

Focus:• Shear strength• Percent wood failure

Percent Wood Failure

1) Initial Image 2) Red Spectrum

Goal: Automatic optical assessment of the % WF

3) Thresholding

Shear Test Results:

assessing variability

A7 B7 C7

A6 B6 C6

A5 B5 C5

A4 B4 C4

A3 B3 C3

A2 B2 C2

A1 B1 C1

Shear Test Results:

assessing variability

A7 B7 C7

A6 B6 C6

A5 B5 C5

A4 B4 C4

A3 B3 C3

A2 B2 C2

A1 B1 C1

A7 B7 C7

A6 B6 C6

A5 B5 C5

A4 B4 C4

A3 B3 C3

A2 B2 C2

A1 B1 C1Strength

(psi)

Strength

(MPa)

Wood

Failure

3 sample

Average553 ±179 3.81 ±1.24 70 ±13.3%

Variance 32% 32% 19%

Panel

Average518 ±140 3.57 ±0.97 66 ±15%

Variance 27% 27% 23%

Average variance in Industry Panel

StressWood Failure

30% 14%

Delamination Test AITC Test T110-2007: Cyclic

Delamination Test

1) Vacuum Soak

30 minutes

@ 25±5 mm Hg

2) Pressure Soak

2 hours

@ 75±5 psi [0.517±0.0344 MPa]

3) Oven Dry to

original weight ± 15%

Within 10-15 hrs

@ 160±5 oF[71±2.7 oC]

4) Measure

Delamination

If <5%, Pass

If >5%, step 5

5) Repeat steps 1-4

with failed specimens

If delamination <10%,

Pass

3”3”

h

Common Modes of Sample Delamination

Mixture of Delamination

and Wood Failure

Delamination only in

early wood.

Delamination Test Results:

assessing variabilityA B C

7 7.95% 0.00% 4.14%

6 0.00% 2.15% 2.65%

5 2.15% 2.73% 0.00%

4 0.41% 0.00% 0.00%

3 0.00% 0.00% 3.16%

2 0.00% 0.00% 3.97%

1 0.00% 0.00% 3.33%

Key

Pass after 2 cycles

Failed to dry within

tolerance

A7 B7 C7

A6 B6 C6

A5 B5 C5

A4 B4 C4

A3 B3 C3

A2 B2 C2

A1 B1 C1

Preliminary Conclusions

• Large variation in shear of shear block strength and

wood failure within a single panel

• Optical measurements offer semi-automatic method to

measure WF% while reducing user input, but shallow

wood failure is hard to categorize.

Future Work

• Investigate variation between billets produced within the

same test variables

• Determine reference rolling shear values for all species

involved in the study

• Develop and Test PUR optical WF% assessment

• Develop automated optical method for assessment of

delamination

Acknowledgments• USDA NIFA Forest Products Research Grant Program (proj. ref.

#2013-05978)

• Swanson group and Interfor for lumber

• Momentive and Henkel for adhesives

• OSU Lab manager: Milo Clauson

• Oregon State University faculty and graduate students

http://commons.wikimedia.org/wiki/File:USDA_NIFA_Twitter_Logo.jpg

http://www.ute1stop.org/home/ute/smartlist_127

http://www.interfor.com/careershttp://www.momentive.com/careers/

http://www.henkel.com/contact-us?pageID=184414

Preliminary Shear Test

Results

A7 B7 C7

A6 B6 C6

A5 B5 C5

A4 B4 C4

A3 B3 C3

A2 B2 C2

A1 B1 C1

A7 B7 C7

A6 B6 C6

A5 B5 C5

A4 B4 C4

A3 B3 C3

A2 B2 C2

A1 B1 C1

ASTM D905: Precision and BiasPrecision has two components:

repeatability at a given test site and

reproducibility between test sites.

The precision of this test method is affected by many factors including, but not limited to:

5. the quality of the bonded joint,

6. the condition of the shear tool,

7. the precision on the testing machine,

8. the operator.

1. the wood species,

2. the boards selected,

3. grain direction,

4. growth ring orientation,

When the specimen fails primarily in the wood, the normal variability of the wood strength affects the precision. The coefficient of variation of shear strength parallel to the grain is 14 % for a given species of wood.

ASTM D905: Precision and BiasPrecision has two components:

repeatability at a given test site and

reproducibility between test sites.

13.1.2 Repeatability at a Given Site:In a study in which all the above factors, except board & growth ring orientation, were controlled, the COVs ranged from:

8 to 11 % for casein bonded specimens,

4 to 12 % for two polyvinyl acetates,

22 to 27 % for an elastomer-based, and

14 to 21 % for PRF adhesive bonded specimens.

In a study in which all the above factors were closely controlled and the specimens were bonded with PRF adhesive, the COVs ranged from:

2.4 up to 13.0 %, with most values in the range of 3 to 6%.

13.1.3 Reproducibility from site to site:

The reproducibility of this test method has not been established.