nails or screws?

[astidham]

I'm getting close to decking the floor in with 3/4 T&G, what will be best for this installation? Nails or screws?

[John Raabe]

I used ring shank nails but others may have additional options. Construction adhesive on the joist tops also reduces squeaks.

[astidham]

Thanks John

[Jeff922]

Whenever I'm deciding what fastener to use, I try to think of the forces the connection will be under (like shear, tensil, compression, compound, etc.)  A screw has great holding strength but low shear strength, a nail has fantastic shear strength, but won't hold as well as a screw.  Maybe I just stated the obvious, but I see people misuse screws a lot.

[glenn kangiser]

Screws won't meet code for a lot of things.  They are brittle and break in expansion/contraction conditions.  I still use them a lot where that or breaking is not an issue.

[MountainDon]

Ring shank nails hold extremely well. They will not walk back out after time like a common nail can and does. (I've seen that with my own eyes). Plus they are approved if sized properly. You can also get them to fit air nailers. Watch overdriving though; best to go light on the air and finish with a hammer. Never screws for anything structural (those special Simpson ones excepted and then only if on the approved use list). 

[zion-diy]

All the houses I've helped build have had the floor decking nailed. that said, my own house is built entirely with screws. the only nails I have used is some of the trim.  Built in 2005. been through the super tuesday F-4 tornado, ice storms, and lots of other weather extremes. All is well so far.

[John Raabe]

Here is an interesting product that is a self-tapping hardened screw that can replace lags screw for things like a ledgerboard for a deck. It seems to be certified for most codes.

http://www.deckbuilderoutlet.com/ledgerlok.html



PS - Understanding the allure of nails - they fail gracefully and can bend and thus absorb short term loads without snapping like a screw would. So for things where flexibility is important (framing, sheathing) nails are generally considered better.

[Jeff922]

I've often wondered about the friction between mating surfaces in a connection and how much significance it has in a connection.  This would change with the amount of surface area obviously.  I read some time ago that not providing the proper expansion gaps on sheathing edges can be a problem as the panels swell resulting in a loss of friction between the studs and the sheathing (swelling edge against edge results in nails being pushed out slightly).  The walls racking resistance was decreased by 11% (I think). 
What I'm getting at is, if a screw increases the friction between surfaces (more so than a nail), it stands to reason that the connection would have to rely less on the connecter itself.  Therefore, perhaps there isn't that much overall difference between a nail and a screw in large surface area connections.  (Yes, I have a tendency to over-analyze  )

[Don_P]

Not me  .

These are all called "dowel type fasteners" in the Nat'l Design Spec for Wood Construction.
The two forces connectors are rated as resisting are withdrawal and shear.

Withdrawal is the ability to pull the fastener out lengthwise, deformed and threaded fasteners shine here. I've taken a class at one of the state university testing labs. I've also built and used a hydraulic joint destruction machine for a furniture company. I talked to one of the professors about ring shank and spiral shanked nails. I've seen rings walk out with moisture cycling and not spirals (both galvy hand driven nails). He had taught in class that the ring is a higher withdrawal strength nail. I quizzed him later, a ring has a higher initial withdrawal resistance but once it "pops" it withdraws the same way a smooth fastener does. A spiral has an initial ultimate strength that is between a smooth and ring but after that pop it withdraws at a higher strength than the others. My contention is that the roofers and masons take care of the initial pop.
Screws are often specced for floors, the NDS publishes screw strength values...if you know the properties of the steel used.

Shear is the ability of the connection to resist a racking type side load. All things being equal that is about connector diameter and density of the wood after adequate depth (10 diameters) is reached. By the time you get a screw with a head that is hard enough to drive well it is often brittle when loaded sideways. The screw design specs in the NDS are specifying the same steel for screws as for nails. A common black drywall screw is easy to drive, great in withdrawal, but lousy in shear.

To my knowledge there is nothing "given" for friction but if a joint is gapped the connection is weaker.
The awc connections calc has some info, their publications list has a number of possibly related papers;
http://www.awc.org/Publications/download.html
I'd bet the help desk at the APA could answer a question like this as well.

[Jeff922]

Very interesting Don P!   There is another factor also involved in wood connections that is of little significance in building construction but interesting to ponder anyway:  "Compression Shrinkage".  Some of you may be aware that you can make a board thinner without actually cutting it.  Take a piece of wood and put two pieces of angle iron on each side and then clamp it together with a few clamps.  Now wait about a year.   The board will be thinner.  The reason is that wood cells (shaped like straws along the grain) will actually change their shape when under pressure. The wood expands with moisture content (think humidity) and with nowhere to go (because of the clamps) the cells are forced to change their shape.  It's important to note that the change is permanent and the wood will not return to its original size unclamped.

Now think of what this means with any kind of mechanical fastener.  Over time ALL screws will loosen because of compression shrinkage.  This is why all-wood connections in furniture construction stand the test of time.  A mortise and tenon will move TOGETHER and therefore compression shrinkage will be minimized.  When I design and build a piece of furniture, I try to only use wood and glue for my connections - screws/lags/bolts really only make disposable furniture.

[John Raabe]

Interesting discussion.  I'm learning stuff!

Where would split ring connectors factor into this? They are sometimes used to strengthen the mating connection between beam and truss connections.

http://www.portlandbolt.com/products/others/splitrings.html

[davidj]

Here is an interesting product that is a self-tapping hardened screw that can replace lags screw for things like a ledgerboard for a deck. It seems to be certified for most codes.

http://www.deckbuilderoutlet.com/ledgerlok.html

Aren't these pretty much the same as the Simpson SDS screws but just another brand?  I used SDS screws for all of my ledgers - 5 minutes with an impact driver and you can hang several trucks-worth of weight off it.

[Don_P]

Yes they are in that same family of hardened screws with proprietary strength tables published by the manufacturer... look for an esr number on packaging for this type of connector.

Split rings distribute the load across a larger cross section of wood and so have higher allowable loads per connector. One rule of thumb, never use a bolt larger than 1" in wood, step to another method. A hundred nails spread around is stronger than a single bolt crushing wood. We want to stay in the elastic, recoverable, region of the wood's strength not in the plastic zone. We load it, it deforms, unload it and it returns entirely. This is drifting towards what Jeff was talking about. Wood is unique in that it can "creep" when loaded with a moderate load for a long period, this is not recoverable. A floor joist designed just for strength and not for deflection would be one example of a joist that would probably end up with a permanent sag over time.

I removed a ~150 year old poplar floor one time that had pretty healthy gaps in the wide tongue and groove planks. They had over a century's worth of smashed dirt and debris in the cracks. Dirt had filtered in seasonally and then been packed each summer by humidity. The edges of the wood had the compression set Jeff describes.

Shear plates are a similar high shear connection
http://www.portlandbolt.com/products/others/shearplates.html

The NDS has tables for both in various configurations and in different densities of wood. The "Wood Handbook" a free download at the US Forest Products Laboratory website has a good section on these connectors.

You aren't fooling around if you want them to fit and do their job.

Metal or wood sideplates with bolts, knifeplates, timber rivets are more ways to make high shear capacity joints.  Also read carefully the comments on tensile loading perpendicular to grain, go carefully if this is a solution to the loads in a high tie.

[NM_Shooter]

What about lag bolts for securing a ledger board? 

I'm already thinking about my deck, and how to hang a ledger board on the rim joist to fasten the deck joists.  Seems like the ledger board requires some fasteners in tension to keep it in place, as well as to handle the shear forces. 

[John Raabe]

Lags, or through bolts are the most common way of making this important connection.

If you are a member of PlanHelp.com this detail is based on the some recent "best practice" research.

[Don_P]

The designer on a house I'm helping with specified Simpson HDU4-SDS2.5 with anchor bolt SSTB14 to go in the place of the anchor bolt and up the back of the girder at the pier locations. That helps with uplift numbers in that connection.

Bolts are superior to lags when possible. Don't spread rows of bolts or lags more than 5" in lumber, it causes splitting as it dries.

I've seen the pictures of a high row of bolts in a line along a ledger. The thin strip and bolts were on the wall, the ledger had split lengthwise. The deck and lower 2/3 of the ledger, well connected by the joist hangers, was on the ground. A number of folks ended up in the hospital. Watch where the hangers connect and where your bolts are connecting the ledger to the building. Well, thinking about it that is a fine example the tension perp failure I was mentioning at the end of my last post.

[John Raabe]

Design loads on decks are usually half again as high as the main floor loads - 60 psf vs 40 psf for the room where you have the piano!

It isn't always intuitive but decks can and do get very heavily loaded, and, of course, that is when they will fail. Think Marti gras party.

[Jeff922]

You know you stuff Don P.  Very informative.  It seems to me that the craft of reading a wood's grain and carefully selecting lumber for its intended job is another casualty of mainstream, disposable, mcmansion building practices.  As Don pointed out, a larger connector isn't always better, and loading a piece of wood with connectors compromises the wood's strength.  Granted, most lumber these days doesn't have a grain worth reading, but it is, in my opinion, another example of how the building trade has been dumbed down by vaious market forces.

[Don_P]

well, we delegated reading the wood to another specialist, the lumber grader. It is a mixed blessing. Most people look at a piece of lumber and think of it as a uniform material. He sorts it into broad catagories for those people. They cannot visualize their way into it, where was it in the log, which way are the knots going, what is the grain doing... how does this relate to strength and stability.

Here's something for folks to try next time you're in the lumber store. Pick up a stick, tell me where the heart is. Now you can also tell me where the bark is. That should be happening as you approach any board.

With that known, effectively all knots (branches) come from the heart and extend radially out to the bark. How are the knots you are seeing oriented within the piece? How much volume do they occupy and where? How does the grain move around them, is it continuous or did the saw cut through curving grain making it short and apt to snap?

Before we go too far in bemoaning our times though, in my forest there are many ring porous species, oaks, locust, sassafras, ash, hickory... They grow faster in todays forest than in an older climax forest. They are also much stronger because of that, less vessel per unit volume. Hickory handle manufacturers limit the number of growth rings per inch knowing that tight ringed woods like hickory are actually weaker... although more dimensionally stable, kinda counterintuitive.

Before we sidetrack too far, we've got folks making and using their own lumber... as related to nails, a nail put into green lumber that subsequently dries is under normal conditions FAR weaker than one that is driven into and through seasoned materials. With the exception of those hardened ring and spiral shanked nails, they hold and do not lose strength.

And a quote from an old timer "there's a difference between sawing wood and making lumber". By the time we get good at it you can tell me if you are walking up on the crown or the roots  .

[rick91351]

Interesting Don I did not know that about hickory.  Idaho is hardly the hickory capital of the USA.  But we do have Idaho Locust but the large stands have mostly disappeared to due to  locusts borers who do not know the difference between a Black Locust and Idaho Locust much less seem to care, and of course the encroachments of developments.