Wood Screws Breaking - Not Sure Why

[bostonman]

I built a fence, the vertical posts are 4x4s (pressure treated/ground contact) and between them are three 2x4s (top/middle/bottom) which I call "runners" (the horizontal pieces to hold the panels), and the panels are non-PT 5/8th thick.

Attached are the labels for the wood used and the screws.

I installed each panel with six screws (two on the top/middle/bottom) approximately 1.5" from the edges. In the last few days, after about four-weeks of installing the panels, I'm seeing the screws breaking (about eight so far). They are breaking at the point where they enter the 2x4", so I'm able to unscrew the broken piece (which is about 5/8th long - the part that is in the panel) and the rest remain inside the 2x4".

The screw box states they are made for all types of treated wood, I'm not over-tightening them, the panels aren't under mechanical stress, so I'm uncertain why these are snapping.

Does anyone have ideas on why these would be breaking?

Update: I forgot to add that I looked at these under the microscope. They don't appear to have pits due to chemicals eating them.

[jpanhalt]

A #2 size for that application seems too skinny.  That's about the size used for drywall and they do tend to break.  I would use a #9 or #10 and probably longer.  I usually get TimberLok brand; although, the store brands may work too.  I consider a fence structural.

[bostonman]

I would use a #9 or #10 and probably longer.

Would stress due to being too thin cause them to snap after weeks without a load? I would have expected them to snap while screwing them in if that were the case.

As for being longer, they are going into 1.5" side of the 2x4"s, so I'm limited to this length due to wood depth and size options.

[jpanhalt]

Treated wood is usually quite wet and changes after installation due to drying.  I wouldn't say they were without stress.

[wraper]

Could be due to thermal expansion/contraction of what they're holding. Or due to wood moisture change as already said.

[coppercone2]

check it with a strait edge

[IanB]

I believe deck screws are quite brittle, so they snap easily.

Also, you need to consider shear stress (sideways stress). If the panels expand due to moisture at a different rate than the runners, then the screws would be under significant stress due to differential expansion.

Have you built your fence according to a recommended design pattern? For example, when I think of "fence panel", then I think of a complete, self supporting panel that goes between the posts. Typically, panels are fixed directly to the posts, and do not have separate runners.

An alternative design is without panels, where you install the runners, and then fix vertical boards to the runners.

[bostonman]

Just to elaborate, this wood has been outside under tarps and slowly removed for a few months as the fence was being built.

The 2"x4"s were fully installed for about a week or two before the panels began being installed. Most of the wood (in my opinion) was acclimated and the outside temperature has remained within 20 or so degrees (we are in the winter months in the Boston area).

Stress seems the obvious issue as if the screws were a cheap grade of chemical protection against treated wood, then I'd assume it would take longer before the chemicals ate through them.

[bostonman]

An alternative design is without panels, where you install the runners, and then fix vertical boards to the runners.

.

This is what I did, correct?

[IanB]

This is what I did, correct?

I see. Now when the boards expand and contract, they will push the horizontal runners up and down. Are the runners free to float in the 4x4s, or are they fixed tightly? If they are fixed tightly so they cannot move, then when the boards expand, something will have to give. If the runners cannot give, then maybe the screws will break instead?

[wraper]

Boards look packed with no gap at all. Might be due to that as the only direction left to expand is outwards.

[bostonman]

The panels aren't spaced; they are laid against each other.

The first panel is against the inner side of the 4x4, six screws (two top/middle/bottom), and then the next panel is pushed against the first, etc... sixteen panels per section.

[Kim Christensen]

A #2 size for that application seems too skinny.  That's about the size used for drywall and they do tend to break.  I would use a #9 or #10 and probably longer.  I usually get TimberLok brand; although, the store brands may work too.  I consider a fence structural.

They are actually #8 screws... The #2 on the box is the Philips screwdriver size.

Bostonman has the boards installed with no space between them. They will expand once out in the rain for months at a time. You need a slight gap between each board to allow for expansion for when they get wetter after installation. With dry wood, 1/4" should do it. If already very wet, you could space them less since they'll shrink in the summer.

[Muttley Snickers]

One of my best mates is a fencer and what you call runners we call rails. As others have mentioned the fence will take some time to settle so expect a bit of movement.

Also, can I suggest you grab an off-cut of the rail and place a test screw using the same technique you placed screws in previously, same depth, speed, torque etc. After installing the screw check its head immediately with your finger to see how warm or hot it got, if it's hot to the touch then you may have altered its temper. Dishwashing liquid or soap is always your best friend when putting screws in timber.   

[Bud]

You can try to pre-drill the runners for a few screws to see if those screws will experience same behaviour. If they also snap later then it is shear stress.

[vk6zgo]

One of my best mates is a fencer and what you call runners we call rails. As others have mentioned the fence will take some time to settle so expect a bit of movement.

Also, can I suggest you grab an off-cut of the rail and place a test screw using the same technique you placed screws in previously, same depth, speed, torque etc. After installing the screw check its head immediately with your finger to see how warm or hot it got, if it's hot to the touch then you may have altered its temper. Dishwashing liquid or soap is always your best friend when putting screws in timber.

If it was Jarrah, at the least, I would rub the screws on a wet cake of soap, or preferably, drill smaller pilot holes, first.

[RoGeorge]

If all else fail, try iron nails instead of screws.

[mikerj]

That style of woodscrew can be quite brittle, IME it takes relatively little bending stress to fracture them.  I'm with RoGeorge, nails are superior in this kind of application where significant movement from wind and moisture is inevitable.

[vk6zgo]

If all else fail, try iron nails instead of screws.

Even better, nuts & bolts.

[Marco]

Maybe you just got unlucky and got a really well dried batch and it's expanding, usually it will just shrink a little over time but expansion with no gap will be trouble.

[jpanhalt]

Wood moves perpendicular to the grain most with changes in weather (temperature and humidity).  Storing it under a tarp might be equated to seasoning, but it doesn't change that fact.*  Even with kiln dried wood seasoned indoors for years, one still gets movement with weather changes.  Ask anyone who installs hardwood floors.  Movement with packed boards can sums those effects. Wind is also a greater factor with packed boards. 

Observations:
Screws are breaking.
You used #2 screws.
The boards are closely packed.
Treated wood was not dried after treatment (perhaps irrelevant for swelling but not for the twisting and warping).

Assumption:
Wood is not stressed.

Conclusion:
Fence was constructed without compensation for swelling during wet months.  The wood is swelling/contracting, warping and twisting.
The screws used initially were too small. (I only use #2 for indoor trim that may need to be removed or toys.)

Recommendation:
Use at least #8, but I recommend #9 or #10 (coarse).
Consider ripping about 1/8" to 1/4" off every 4th board or so.  Pressure exerted by wood swelling when wet can be used to split stone.
I do not think the entire fence needs to be rebuilt.  If you move the boards, you will have some screws only slightly offset from the original and that is weak.  Reinstalling larger screws in the same holes will be stronger.

*As a general rule, wood stickered and stored in a dry barn, air dries at about 1" per year.   That is barely adequate and requires good airflow around the wood.  Wood simply stacked under a tarp does not qualify as seasoned or air dried.

Edit:  While nuts and bolts may be stronger, they typically leave the nut and a portion of the bolt exposed.  That snags people and clothes.

[Marco]

If the wood was still relatively high humidity and was drying, it wouldn't really be a problem. It would create its own gaps.

[jpanhalt]

That's why I added "perhaps irrelevant."  It's warping and twisting, but shrinking can also produce stress.  Fall in our NE, as elsewhere in the US, is a wet season.  Despite being "wet" compared to dried wood by definition, treated wood still expands when put in the rain.

[bostonman]

Just as another note, I pre-drilled all the holes before installing the screws. Not a very large hole, but enough to reduce the chances of the wood splitting when installing the screws.

The winter months are upon us, so unfortunately all I can do is wait to see if, and how many more screws break. This is a bit of a scary situation as the fence wasn't completed and already had screws breaking.

I should have called the "runners" rails instead, seems like the correct term.

[jpanhalt]

Bundle up.  Here in the Western Reserve we are 50°F today.  I suspect the screws will be harder to remove next year than they are today.

In the 1980's, I worked outside in Minnesota at temperatures as low as 10°F.  That's when I learned about 3M Thinsulate coveralls with nylon(?), fire resistant outer and inner layers.  Today's Carhartt and Dickie are OK, but more bulky.  I am not suggesting that you go to that extreme, but East coast cold should not be too much of an impediment.  Besides, it can be fun.

[Kleinstein]

When the humidity changes wood does swell / contract quite a bit in the direction perpendicular to the fibers. For the part shown the point to worry most would the wood getting wet and expanding with no space between. This can shear off the srews. One should have space between the vertical wood. Also larger predrilled holes could have helped a little.

[Poroit]

G'day bostonman,

I agree with other members .....your screws are two small for the job and are breaking due to thermal expansion/contraction.

From the charts I read a #2 has only a 3/32" (2.38mm) shank diameter.

[Kleinstein]

With wood it is not the thermal expansion, but the humidity effect that cause the large dimensional changes. Especially the expansion when getting wet can cause large forces. In ancient times they used wedges from dry hard wood to split stones.

[mikerj]

G'day bostonman,

I agree with other members .....your screws are two small for the job and are breaking due to thermal expansion/contraction.

From the charts I read a #2 has only a 3/32" (2.38mm) shank diameter.

As Kim Christensen previously noted these are NOT #2 screws.  The #2 number of the front of the box is the size of the Pozi drive required.  The actual screws are #8, though I don't know why this is not made clear on the box as it would normally show #8 x 1-5/8"

[xrunner]

I installed each panel with six screws (two on the top/middle/bottom) approximately 1.5" from the edges. In the last few days, after about four-weeks of installing the panels, I'm seeing the screws breaking (about eight so far). They are breaking at the point where they enter the 2x4", so I'm able to unscrew the broken piece (which is about 5/8th long - the part that is in the panel) and the rest remain inside the 2x4".

My opinion follows -

You have defective screws

I've built a whole lot of outdoor projects with screws like that - fences, decks, fence gates, other types of outdoor projects ... I've never had a single screw snap or break over decades. I've taken out old screws and they were a bit rusted but still pretty strong. The screws you used snapped? A panel attached with six? Ain't going to snap a screw by expanding or contracting. No way IMHO. The wood will split before that happens

Unless ... The screws are defective.

Try some experiments on a spare piece of wood. Screw in one of those screws and leave part of it exposed, then twist on it with a vise grip. It better be pretty damn hard to bend it and break it, although of course it can be done. But a panel with six screws in it should not be close to breaking a single screw - no way.

That's my opinion here.

[bostonman]

Last night, prior to reading the latest post, I performed experiments.

First of all, I had six screws from one panel I removed. Long story short, the new panel had a bad area that I overlooked when installed, and then it just looked horrible being next to all the other nice panels, so I wanted to replace it. The panel was only about two-three weeks old; hence the new screws had equal age on them. Upon removing the six screws, two snapped.

I took these six screws and set them aside to look at the two broken ones under a microscope. They were porous (I was expecting solid metal), but the interesting part: they had very thin traces/lines of red/maroon color. One line looked like it went across the diameter of the screw. Thought maybe it was wood debris from when it snapped as I removed it, so I cleaned it with brake cleaner, but the red/maroon color was still there.

I also took a pick and scrapped the metal. It became very shiny whereas the other parts I didn't scrap were dull.

I took another one of the six, this time one that didn't snap when I removed it, and bent it with two pliers to test its strength. Also note, the screw wasn't bent at all indicating some sort of mechanical stress, it looked as straight as it did new. The screw immediately broke with very little force (with enough force, I could have probably broke it with my fingers). Upon looking at it under the microscope, it had a dot of red/maroon on the outside edge. Looked kind of as if the coating chipped and it was starting to "rust".

Although this information may not be relevant to the issue, it is odd that I'm seeing discoloring, but wanted to mention it.

As for tests, I took brand new screws out of the box and performed a series of tests. I tried the bend test (putting one end in a vice and bending with pliers). They bent almost 45 degrees before snapping whereas the one I removed from the panel broke much sooner. I took another and screwed it into wood and removed it to test whether heat affected it. I screwed and unscrewed the same screw easily twelve times without pausing to let it cool, attempted to bend it in the vice until it snapped, and it still took about the same 45 degrees before snapping.

I took another screw, screwed it about 3/4 of the way into wood, placed the wood against one jaw of the vice and the screw against the other jaw, and tightened the vice to put excessive force on the screw. It bent, but didn't break. I then flipped the wood and bent the screw back without it breaking. Using the same (now warped) screw, I removed it and screwed it in and out another ten-twelve times in the wood without any issues.

None of this concluded anything, but does seem 'heat' due to screwing didn't compromise the strength of the screw; the wood on the fence was also outside in <50 degrees F weather, so it would have pulled heat out of the screw as it turned. Bending the screw doesn't seem to weaken it, although, if the wood expanded/contracted numerous times, thus bending the screw back and forth, then it would snap.

Also note: the three rails (the horizontal 2x4s) are just resting in the saddles and the panels are against each other, but they weren't installed with excessive force to press them against each other. I'd stand one, adjust the height so it's even, place a clamp on it and the 2x4 (i.e. rail) to hold in place, drill, screw, move onto the next panel. So the fence is "tight", but technically has room to move.

I'll attach some of the microscope pictures, but I couldn't get a very clear picture due to holding my phone into the lens of the microscope.

One of the two pictures shows the lines of red/maroon and the other shows the small dot of red/maroon on the outside (this was the one that snapped easy when I bent it with two pliers).

[xrunner]

I still hold you have defective screws. Look, you are a careful person from what I've read on the forum. There are many MANY laborers who put up fences just like you are and they don't give a sh*t about the screws they are given, how tight the screws go in, or the wood panel's condition. They just blast screws in, get the job done, and go home. This is not a normal thing you are experiencing and it's because of the screws.

[Kleinstein]

The screws may get a bit hot when driving in hard wood, butt his should not be enough to alter the steel - it would need a temperature that would char the wood.

Normally the screw should be pretty hard steel and not at all easy to bend. It is not unusual to have screws that breake instead of bending.  So the describtion sounds a bit like bad screws, though nothing really quantitative.

When wood gets wet it can produce significan force when under compression and the horizonal wood can withstant quite some tensile force in the fiber direction with much less swelling in that direction. This can very well breake srews, even if they are good. The swelling can also be quite a bit and the inilial gaps may not be enough.  If the wet wood is the issue, the crews to break should be more on the outside and less in the center of the sections.

[metrologist]

OT but related. Fence posts. The professionals seem to always pour the concrete a couple inches below the soil level so the post rots, then they get to come back and replace posts, or get a second fence job. I always the finish the concrete a couple inches above the soil with a slope away from the post down to the soil.

[scoooter]

Most of my experience comes from working on wooden cabins up in the mountains (where they sit in the snow for about 8 months out of the year: lots of thermal effects).

Couple things I've been told and learned:
- Screws are brittle - something about the cold process of making them makes them far more brittle and really aren't the best in shear
- Nails don't hold - nails are formed differently (hotter temperatures), making them more ductile and higher strength, but obviously since they don't have threads they aren't good with grip strength

So if you're building a fence, make sure you use your screws in grip and your nails in shear. As your wood expands and contracts (humidity, temperature, age), these stresses will change.

You could also just buy higher rated screws and be fine. Most of the other replies are good advice too.

https://www.artofmanliness.com/skills/manly-know-how/nails-or-screws/

[coppercone2]

I thought of something else, what if its like a animal ?

I hope we are not trying to solve thermal mechanical conditions when there is a bear occurring.


Is that what happened on the Nostromo? Wow that is alot of rust, I better stand over this hole in the ceiling while reading my textbook on galvanic compatibility and corrosion.

"I know! There must have been trace amounts of zinc in the  *transcription failure*"

[bostonman]

I thought of something else, what if its like a animal ?

We don't have large animals in the area such as bears. Usually just squirrels, rabbits, etc...  Also, most fences in the area are wood and have been standing for several decades. In fact, one neighbor has had their fence for probably forty-years and only one post has rotted; the rest of the posts are still dense.

[CatalinaWOW]

I have had similar experiences with this type of screws.  The grain structure in your broken screws indicates a pretty hard steel.  I know that this is a characteristic of many types of screws intended for power driving.  The most extreme example being dry wall screws.  Also associated with brittleness, inability to withstand much bending.  This doesn't seem consistent with your tests on the screws.  Perhaps bending them beyond 45 or whatever it takes to break them would give some clues in the grain structure.  Bending them back and forth to break could work harden them, invalidating the results.

It might be a quality control problem as suggested previously, but one of a specific character.  Not every screw in the batch you bought being bad, but some fraction of them.  I know some steels are very sensitive to the heat treatment process, and that an amazingly small temperature difference can make a huge difference in properties.  So screws at one corner of the tray might end up over hard and brittle while others were fine.

In my experience, both in outdoor wood fences and in an aviary like structure, some fraction of the screws have failed after a short service life.  Others are carrying on for some years now with no problems.  I have replaced the failed screws with similar screws (in some cases from the original box) and had a small number of failures in the replacements.

I have no explanation for the following, but construction screws typically have some form of coating on them.  The two types I see in local stores are a gold coating similar to the screws you are using, and the other is a grey coating.  I don't have any real numbers, but feel that they grey coated ones have a lower failure rate.

[coppercone2]

the less fasteners you need the better quality they need to be, I am almost certain with the cheap ones they expect there is a 'backup' and that its over engineered.

When you use like a kiln properly, it does NOT look 'efficient'. Very easy to overload it.

[bostonman]

The two types I see in local stores are a gold coating similar to the screws you are using

Mine are black.

I'm not doubting any of the feedback, but just thinking out loud. Seems hard to believe the numerous boxes on the shelf and thousands of screws produced at the factory probably per day resulted in approximately 1% failure from two boxes (I needed 960 screws and a single box didn't have that many); also failed in (being conservative) six-weeks. I can see if maybe after a year or two, but, if these broke this quickly, I fear what will happen during the serious cold winter months.

Initially I was using the yellow colored screws from Home Depot to mount the hangers (the hangers are screwed into the 4x4 posts). After the first one or two, I found they didn't grab well and would spin (edit: spin after they were fully threaded in). This resulted in questioning the threads and whether maybe they weren't tall enough to grab enough wood.

Long story short, this led to using these 1-5/8" phillips head black screws for the panels (screwed into the 2x4 rails).

For testing, currently I have five or six screws threaded in about 3/4" into a scrap piece of a 2x4 used for the rails. Also, I have a scrap piece of panel I had cut screwed into another scrap piece of 2x4 that was used for the rail. All this wood has experienced the same conditions and were unneeded lengths cut from the wood used to build the fence.

All the "test" pieces are sitting on the ground and have experienced rain/snow along with single digit temperatures and temperature deltas of approximately 30-40 degrees F. The one with the five or six screws are face down so the screws are sitting in the dirt/grass and the other has the panel flat on the ground with the 2x4 facing upwards. The rain/snow is going directly between the panel and 2x4, and, if it sits frozen, should put a good amount of tensile force on the screws.

I'm leaving them outside, maybe the entire winter, but will check them (along with the fence) periodically. If the test piece with the panel (or the other with just screws) break, then it will reduce the potential factors that are causing these screws to break.

[IanB]

After the first one or two, I found they didn't grab well and would spin (edit: spin after they were fully threaded in).

That's crazy:

1. Are you using a drill-driver, or an impact driver?
2. Are you adjusting the torque control to lower the torque to an appropriate setting?

Spinning screws in wood is almost unheard of, it suggests way too much torque. Using too much torque on screws can easily twist the heads off.

[CatalinaWOW]

After the first one or two, I found they didn't grab well and would spin (edit: spin after they were fully threaded in).

That's crazy:

1. Are you using a drill-driver, or an impact driver?
2. Are you adjusting the torque control to lower the torque to an appropriate setting?

Spinning screws in wood is almost unheard of, it suggests way too much torque. Using too much torque on screws can easily twist the heads off.

In addition to Ian's completely correct comments the black finish are almost universally dry wall screws, extremely hard and totally unsuitable for load bearing applications. 

The pictures of the box you originally posted showed suitability for decks and other applications along with the golden color.   I have never seen a "construction screw" with a black finish, but can't deny the possibility.

[bostonman]

Damn, I typed black, but I typed too quickly without thinking. Mine are gray.

The yellow (or gold) Deckmate from Home Depot I mentioned that turned, much like all the screws, were screwed in with a regular drill on a very low setting.

I've experienced this before, mainly (I believe) with short screws. If I remember correctly, I screwed a hanger into a 4x4 with a 1-1/4" Deckmate. The minute it went in, I could feel it lacked a good grip. After I took a scrap piece of 4x4 and tested several screws, screwing them in slowly, regular speed, etc... and found only some tightened.

My thought is the coated screws reduce the height of the threads. With the softer wood produced now (or maybe softer due to being wet from pressure treating) they don't have much material to grab and result in spinning.

[Monkeh]

Spinning screws in wood is almost unheard of, it suggests way too much torque.

It is actually quite easy for smaller screws with very sharp threads to simply cut all the fibres in fast grown timber and spin.

[IanB]

Spinning screws in wood is almost unheard of, it suggests way too much torque.

It is actually quite easy for smaller screws with very sharp threads to simply cut all the fibres in fast grown timber and spin.

While there are always exceptions, I don't believe appropriately torqued deck screws should do that. Even if modern lumber is tending towards balsawood these days.

[bostonman]

I'll also add that in most cases I wasn't depending on the drill to begin grinding when it reached the torque setting.

In most cases I set the drill to a low setting, maybe 2, and I'd stop before the screw was all the way in to avoid the possibility of snapping the screw. Then I'd ease it to the final resting spot.

Most times I noticed what while tightened the screw at a setting of 2 may cause the drill to stop with the screw halfway on another, so I'd have to change the setting to 3 or 4.

So I wasn't hammering away letting the drill go full blast until the screw was fully seated. It was more or less a combination of the drill torque setting and me "feeling" the amount of torque on the screw.

[Monkeh]

When you've got a piddly little 4x30 screw in wet softwood, you can easily strip the hole out with a screwdriver. You'll not be snapping these screws with a drill on any clutch setting unless they're grossly defective.

I send screws from 3.5x30 to 6x150 with an impact. You may be a tad overcautious.

As for the screws snapping by themselves, that's just brittle screws and wood moving. More screws, larger screws, or nails is the solution. I recommend nails for the panels, and the 'runners' (rails) should be fixed with either really big nails, or in US terms #10x4" screws (or bigger).