A month or two ago I asked a question about experiences comparing Titebond 3 (T3) against Titebond 2 extended (T2E) glues. Feedback wasn't as strong directly comparing these as I'd hoped. So I did some experiments myself.
I've been making small pieced candy bowls (5-6") and cookie plates (8-9") by face grain gluing several pieces of different species (see avatar picture). For the last several months I've been using Titebond 3, but when used on joints near the bottom of bowls that have shallow angled glue joints with light colored wood the Titebond 3 adds a dark color that isn't appealing. Consequently I've started using Titebond 2 Extended because of it's clear/light color as well. My question was if this mattered for durability.
For those that like answers up front:
1) Wood thickness at the glue joint is the most important factor. As long as the joints are ~3/16" or greater there should be no problem; 1/8" is probably OK; less than that doesn't hold up as well.
2) Titebond 3 performs marginally better that Titebond 2E at very thin thicknesses
3) Wood species matters, again at the margins; but less than the T3 vs T2E difference for what I tested (did NOT look at oily exotics etc - YMMV)
Methodolgy:
I glued up separate blocks of oak and sapelle, four laminations each, using both T3 and T2E for a total of 4 blocks each a little over 7" long. Then I sawed each block across the laminations to thicknesses of 1/16", 1/8", 3/16", 1/4", 3/8" and 1/2". This provided 6 thicknesses each in four groups (2x2 matrix) of oak or sapelle, and T3 or T2E glue. Each sample contained 3 glue lines a little over 7" long.
After waiting a week for the glue to fully cure, I subjected them to a worse stress than I hope anyone does with the bowls I make. I put them through our dishwasher (top rack) on a regular cycle (3hr15min) including heat dry. Soap was used about half the time as I snuck them in with our dishes when my wife wasn't watching
. I ran them through this cycle *10 TIMES* checking for delaminations between each cycle.
Results:
1) NO delamination failures observed through 10 cycles for either glue, either wood species, for all thicknesses 3/16" and greater.
2) NO delamination failures observed using Titebond 3 on oak, regardless of thickness 1/16" up.
3) 1/16" thicknesses of both oak and sapelle using Titebond 2 Extended Time had some delamination failure after one or two cycles, and 50% or more of the 21" combined glue line (3x7") had failed by 10 cycles. The 1/16" T2E sapelle sample was discontinued after 3 cycles when 2 of the 3 glue joints completely failed.
4) Sapelle 1/16" samples performed poorly with both glues.
5) Sapelle 1/8" samples started evidencing some delamination in mid-cycles with both T3 and T2E demonstrating <10% delamination on 1 of the 3 joints after 10 cycles.
6) Oak 1/8" T2E sample evidenced about 15% delamination on 1 of 3 joints after 2 cycles, this remained constant through 10 cycles.
Additional confirmation:
Glued up straight blocks don't stress the glue joint as much as rounded bowls will with various expansion and contractions working against each other. Believe it or not I had a few 'funnels' laying around so I put four of those into our dishwasher torture chamber as well, two T3 and two T2E. These were not as well controlled for wood species and thicknesses but casual observation confirmed the findings from above.
Three of the bowls (two T2E and one T3) with many 1/16" joints (and rim lamimations with ~3/16" joints) were discontinued after 1 dishwasher cycle because of many 1/16" joint failures on each; in all cases the 3/16" rim joints were still solid.
One of the T3 bowls with glue joints between 1/8" and 3/16" survived 10 dishwasher cycles with only 2 of the ~25 joints delaminating. Additionally a small plate glued with T2E, joints ~1/8" -1/4", survived 10 dishwasher cycles although the edges of half of the glue joints were separating (plate joints were straight across and had a little more stress than the controlled experiment above but not nearly the stresses of the bowl joints; different woods were used on this).
Additionally, all of the bowls and plate subjected to the dishwasher were sealed with Watco. After the first dishwasher cycle, that finish was gone.
Conclusions:
1) The most important thing I learned from this was not to let wall thicknesses get too small. Since starting to see the results a couple weeks ago I've been careful to target minimum thicknesses to around 3/16" and try not to allow any glue joints to be thinner than 1/8".
2) My initial goal was to better understand T3 vs T2E for my pieced bowls. Based on these results I will use those two glues interchangably for my bowls. There are some reasons to choose T3, but the difference is insignificant for most of my purposes.
3) Don't put wooden bowls in a dishwasher
(Frankly, I was pleasantly surprised how well both the T3 and T2E joints survived this stress)
I've been making small pieced candy bowls (5-6") and cookie plates (8-9") by face grain gluing several pieces of different species (see avatar picture). For the last several months I've been using Titebond 3, but when used on joints near the bottom of bowls that have shallow angled glue joints with light colored wood the Titebond 3 adds a dark color that isn't appealing. Consequently I've started using Titebond 2 Extended because of it's clear/light color as well. My question was if this mattered for durability.
For those that like answers up front:
1) Wood thickness at the glue joint is the most important factor. As long as the joints are ~3/16" or greater there should be no problem; 1/8" is probably OK; less than that doesn't hold up as well.
2) Titebond 3 performs marginally better that Titebond 2E at very thin thicknesses
3) Wood species matters, again at the margins; but less than the T3 vs T2E difference for what I tested (did NOT look at oily exotics etc - YMMV)
Methodolgy:
I glued up separate blocks of oak and sapelle, four laminations each, using both T3 and T2E for a total of 4 blocks each a little over 7" long. Then I sawed each block across the laminations to thicknesses of 1/16", 1/8", 3/16", 1/4", 3/8" and 1/2". This provided 6 thicknesses each in four groups (2x2 matrix) of oak or sapelle, and T3 or T2E glue. Each sample contained 3 glue lines a little over 7" long.
After waiting a week for the glue to fully cure, I subjected them to a worse stress than I hope anyone does with the bowls I make. I put them through our dishwasher (top rack) on a regular cycle (3hr15min) including heat dry. Soap was used about half the time as I snuck them in with our dishes when my wife wasn't watching
. I ran them through this cycle *10 TIMES* checking for delaminations between each cycle.Results:
1) NO delamination failures observed through 10 cycles for either glue, either wood species, for all thicknesses 3/16" and greater.
2) NO delamination failures observed using Titebond 3 on oak, regardless of thickness 1/16" up.
3) 1/16" thicknesses of both oak and sapelle using Titebond 2 Extended Time had some delamination failure after one or two cycles, and 50% or more of the 21" combined glue line (3x7") had failed by 10 cycles. The 1/16" T2E sapelle sample was discontinued after 3 cycles when 2 of the 3 glue joints completely failed.
4) Sapelle 1/16" samples performed poorly with both glues.
5) Sapelle 1/8" samples started evidencing some delamination in mid-cycles with both T3 and T2E demonstrating <10% delamination on 1 of the 3 joints after 10 cycles.
6) Oak 1/8" T2E sample evidenced about 15% delamination on 1 of 3 joints after 2 cycles, this remained constant through 10 cycles.
Additional confirmation:
Glued up straight blocks don't stress the glue joint as much as rounded bowls will with various expansion and contractions working against each other. Believe it or not I had a few 'funnels' laying around so I put four of those into our dishwasher torture chamber as well, two T3 and two T2E. These were not as well controlled for wood species and thicknesses but casual observation confirmed the findings from above.
Three of the bowls (two T2E and one T3) with many 1/16" joints (and rim lamimations with ~3/16" joints) were discontinued after 1 dishwasher cycle because of many 1/16" joint failures on each; in all cases the 3/16" rim joints were still solid.
One of the T3 bowls with glue joints between 1/8" and 3/16" survived 10 dishwasher cycles with only 2 of the ~25 joints delaminating. Additionally a small plate glued with T2E, joints ~1/8" -1/4", survived 10 dishwasher cycles although the edges of half of the glue joints were separating (plate joints were straight across and had a little more stress than the controlled experiment above but not nearly the stresses of the bowl joints; different woods were used on this).
Additionally, all of the bowls and plate subjected to the dishwasher were sealed with Watco. After the first dishwasher cycle, that finish was gone.
Conclusions:
1) The most important thing I learned from this was not to let wall thicknesses get too small. Since starting to see the results a couple weeks ago I've been careful to target minimum thicknesses to around 3/16" and try not to allow any glue joints to be thinner than 1/8".
2) My initial goal was to better understand T3 vs T2E for my pieced bowls. Based on these results I will use those two glues interchangably for my bowls. There are some reasons to choose T3, but the difference is insignificant for most of my purposes.
3) Don't put wooden bowls in a dishwasher
(Frankly, I was pleasantly surprised how well both the T3 and T2E joints survived this stress)
