DIY Home Improvement Forum banner

Nail Shear Strength Test

32K views 32 replies 13 participants last post by  Duckweather 
#1 ·
Recently there was a thread regarding the virtues of nails vs screws....with the common theme being 'shear strenght'.

Me, being the engineering technogeek I am, I'm not one to accept the a blanket statement of "Nails are better" without something to back it up...

So...I did a test.

I took 2 2x4's and nailed them together with 16d nails. (edit...ok....wrong pic...this is the screws....but the nails were in the same spot)



I left the other ends about 12" apart. The intent is to take this end, set it on a scale and measure how much force it takes to bring the ends together.



So I rotated the test assembly and put the end on the scale....



The weight of the stud was enough to make it drop down to the position shown. The only measurement on the scale was the raw weight of the wood.

The nails did not give....the wood did.

So...I repeated the test using #8 3" screws.

Basically....alsmost exactly the same results.....

Initial conclusion....all the talk about shear strength....bogus....the softness of the wood negates the shear strength of the nails or screws..I don't think the average DF1 stud is strong enough to put a nail or screw into shear failure.....

To add to that...when I was taking the test pieces apart...the nailed pieces came apart easy...just lift one board, pulled apart with no problem....the screws? Damage to the wood...and significant more effort to pull apart....at least 3-4x more force.

With that said......

I can not imagine trying to do a framing job using just screws......the time to use screws would be an easy 5-10x longer over an air nailer...

Given the way typical framing is done....I see no advantage to screws over nails....few of the forces are in tension vs shear....in tension, screws would be significantly better....but that is why we have so many Simpson hardware choices....
 
See less See more
3
#5 ·
THIS is how the test should be done!

HOWEVER you also need to test it over time, those gold screws will hold up ok when new but after 5 years they'll snap with barely a tap. I've demo'd countless decks and most of the decking is screwed down with those gold screws or the newer decks have the coated deck screws, you can snap them easily with barely a swing from the hammer.
 
#3 ·
While the wood gave and softened the stress on the nails and screws, adding the Simpson ties adds a new dimension of stress, the metal and the wood now have a different stress factor other than just the wood and will produce much more shear force.
 
#6 ·
the shear test that is best is to drive a srew into a piece of wood leaving some of it sticking out, do the same with a nail... next,with a hammer from the sides tap on the screw from two different directions. the screw will maybe bend just a little bit and then snap off. Now do that with a nail. you will have to get that nail really hot with friction and movement before it will snap off.

what you are showing with the soft wood giving before the screw should always be considered when using joist hangers on a structure, especially anything up to a triple hanger and over. We have had to add structural plywood in some cases because the wood was to soft to attach the hanger to effectively.
 
#8 ·
the shear test that is best is to drive a srew into a piece of wood leaving some of it sticking out, do the same with a nail... next,with a hammer from the sides tap on the screw from two different directions. the screw will maybe bend just a little bit and then snap off. Now do that with a nail. you will have to get that nail really hot with friction and movement before it will snap off.

what you are showing with the soft wood giving before the screw should always be considered when using joist hangers on a structure, especially anything up to a triple hanger and over. We have had to add structural plywood in some cases because the wood was to soft to attach the hanger to effectively.

What you are describing is not shear....that is bending.

By nature, a screw is a much stiffer metal....and takes more force to bend than a nail...while nails are a soft ductal metal by comparison.....

That one YouTube video where a guy is smacking the nail and screw from the side bending it and 'trying' to demonstrate shear....is not demonstrating shear....that is bending...and screws will break before a nail does....

The intent of the test was to demonstrate shear......and I think it shows that the wood is more of a factor than the fastner.

I'm not saying to use screws....as already noted...nails are a lot cheaper and faster...and given the application...work fine....

Kwikfish....I don't see the difference in your test vs mine other than a shorter stud....can you post a skectch of what you have in mind? I'll try it.

With that said....I will be pulling up a bunch of nails soon.....

When I put down my floor on the 2nd story I used 10d nails thru the 1 1/8" T&G plywood per the drawings....along with glue....I have a bunch of them working their way out....those that are...I'll replace with screws....which won't come out....(as easy)
 
#9 ·
I don't think this test is really measuring sheer either. Since you are rotating the board around an axis point (actually 2 different points), you are measuring sheer and torsion.

For pure sheer, you would need to secure a nail or screw into a board which is secured to the floor, and then pull the other board horizontally until the fastener reaches its breaking point (the wood would probably break before the fastener). That would be pure sheer.
 
#10 ·
Oh, it's shear....one nail has shear forces in one direction...the other nail, the opposite direction. The rotation is only changing the direction of the shear....which is actually a pretty small change.

If you draw it....it will be a bit more obvious.....but it is shear.....all I'm doing is using the mechanical advantage of the stud vs setting up the test sample in a press along wth a strain gauge.....not something I keep around the house....
 
#14 ·
Its a great idea its just by using the ends of the board its going to split as soon as you apply the side pressure.

Try using a 2x6 and make the axis about 3 ft up the board you will get a better reading.
Good point....

I think I'll try it again....one test at the end....another 12" inboard.....I think anything over 12" is not going to show any real difference....

I also assume that the moisture content of the wood would be a significant variable....

"wood would" Now that that 10 times....
 
#17 ·
Nice test, I applaud your enthusiasm. The topic of nail versus screw has been a fairly common theme on this forum for at least two years. I have written several times on the topic, no need to repeat previous posts here. Unfortunately the discussion often bogs down because of differing terminology, and lack of understanding of the actual mechanics of stress on fasteners, which is surprisingly complicated. For those interested in what I consider an outstanding study of the strength of nails, screws, staples, lag screws, bolts and similar types of fasteners, see

http://www.fpl.fs.fed.us/documnts/fplgtr/fplgtr190/chapter_08.pdf

which examines in great detail the strength of various types of wood fasteners, including of course screws and nails, in shear, tension, and compressive loading. The paper also discusses the strength of the wood versus the strength of the fasteners. Under certain conditions, the fasteners will fail before the wood, but you have to read the paper to understand when this may occur, it isn't simple.

As a note, it is essentially impossible to put a nail or screw into pure bending under any realistic field conditions due to the method of use. A nail or screw is almost always driven approximately perpendicular to the board (toenailing is the most obvious exception). Movement of the board due to sliding, or rotation of the board due to bending, is felt by the nail or screw as shear, not as bending. If you do not fasten the nail or screw, but say drive it partially into the wood, leaving say an inch or two sticking out, it is certainly possible to put the nail or screw into bending by using a plyers or similar device to grip the end, then exert force perpendicular to the nail. This effectively treats the nail or screw as a cantilever beam, which experiences bending. However, this type of loading is almost impossible to achieve in a realistic framing scenario. By the way, all Simpson and similar brackets are specifically designed to put the nails or structural screws into pure shear.

Another point to note is that banging on a nail or screw that is sticking up actually most likely fails the fastener due to fatigue, which is a complex failure mechanism of most metals whereby repeated cyclical stress causes weakening of the metal, and ultimately failure. This is a serious problem in rotating equipment, and metals subject to repeated cyclical stresses approaching their normal failure strength, but is not typically a problem in normal wood framed construction, since the loads on the fasteners are typically static or near static, and the rare case of cyclical high loading due to wind or impact almost never approaches the required number of cycles to cause fatigue failure of the fastener. Bending a nail or screw back and forth is an example of very high stress cyclical loading which certainly will fail a steel fastener relatively quickly, but has little or nothing to do with actual loads encountered by real world structures.

For those who do not wish to read the FPL article, let me summarize a couple of important conclusions. The shear strength of any metal fastener is essentially the same if the root diameter of the fastener is the same and the same metal is used in fabrication. It is a common error to compare a nail and screw by measuring the outside thread diameter of the screw, as this is always greater than the root diameter, and the shear strength of the screw is always governed by the root diameter. I believe that much of the urban legend that screws are weaker than nails is due to improper measurement of the diameter of the screw.

The other common mistake is comparing a screw made of one type of metal with a nail made of another type of metal. Unfortunately most wood screws are NOT made for structural use, rather they are designed for drywall installation, furniture, deck board installation, or the like. A drywall screw should never be compared to a framing nail in shear, since their purpose is totally different, and the drywall screw is likely to be made of a completely different type of metal than the framing nail. The framing nail is generally made from mild steel, which is a good choice in shear, while the drywall screw (and most other wood screws) are made of metal that is typically of lower strength than mild steel. A perfectly reasonable comparison would be a Simpson framing nail versus a Simpson framing screw of the same root diameter, and not surprising they have very close shear strength, as documented by the FPL study. Structural screws are only available from select manufacturers, they are typically more expensive than an equivalent nail, and are generally slower to install, hence they are rarely used in framing, except where direct pullout (tension) is critical, as in installation of roof sheathing.
 
#18 ·
Thanks Dan, good informative info, how many times have you found a deck screw sheared off simply in a deck board? I've found on numerous occasions that a screw has sheared off just holding decking down.

Something to keep in mind is that most homeowners and DIYers shop at home depot or lowes, I bet 8 times out 10 if you grab some deck hangers and wander over to the screw aisle (the ones with deck screws and drywall screws) and ask an employee if you can use these screws with this hanger you'll get a yes.
 
#20 ·
Daniel...thanks for the outstanding write up. I book marked your technical reference....even read part of it....I was surprised at the impact wood moisture had....

You covered the fatigue issue related to side bending forces.....but what about impact forces...specifically, the change in grain growth that most likely takes place when a nail is hammered in with multiple hammer hits? Case in point....in the aerospace industry, when using rivets, there are different types based on how they are going to be installed. The average rivet that manually buck is designed for only so many hits.....if you buck it too much, you actually make the rivet more brittle....hence, more prone to premature failure. Other rivets are set with a press....one shot....done....if you go hitting it with a rivet gun you will actually make it weaker...

Hence....a nail sunk with a nail gun would in my opinion be less prone to fatigue failure....where as, someone driving a nail in with about 200 hits....I would expect a lot of the ductal properties to be shot....

Is that accurate?
 
#21 ·
Fatigue is typically caused by cyclical bending stress where the maximum stress applied to the element is generally well below the failure point of the metal. After a relatively predictable number of cycles (the number depends on the metal or alloy), the metal weakens and can fail catastrophically. The aerospace industry was one of the earliest to notice the problem, which can occur due to cyclical pressurization/depressurization of the aircraft body. Several well documented catastrophic failures of aircraft parts such as engines and doors were due to metal fatigue.

Banging on a nail is a little different. When you first drive a nail, it behaves like a slender column. A nail which is 1/8 inch diameter and 3 inches long has a slenderness ratio of 24, which means the nail can buckle easily. Buckling is made much worse if the nail is struck eccentrically, which is often the case with novice carpenters. If the nail begins to bend a little, the eccentricity gets worse, making the buckling problem worse. Everyone has probably noticed this, if you miss with the first hit, and bend the nail a little, it is very hard to drive it home without bending it to the breaking point. The bending part is the nail buckling, which is not a metal fatigue issue, but rather due to the slenderness problem.

Another issue with nailing is that many light taps are more likely to bend the nail than a few heavy blows centered on the nail. This is due to the mechanics of impact between the hammer and nail, the harder you swing, the more force is transferred into the nail versus the hammer, and the nail accelerates rapidly into the wood and lacks the time to buckle. With light taps, the nail barely moves, and there is a much higher probability of bending the nail in buckling due to an off center blow, since you need so many blows to drive it home, and the nail remains in the most vulnerable position (the start of the nailing) for a large number of hits. This is also not related to fatigue. It takes typically hundreds or thousands of cycles for fatigue to occur.

If you bend a nail beyond its elastic limit, you get permanent deformation of the nail due to inelastic bending. This is not due to fatigue, simply due to the fact that every metal has an elastic limit, if you bend it beyond that limit it becomes plastic and will not return to original shape. If you do force a bent nail back to straight condition, the nail is in a permanently weakened condition because it was bent plastically, and the metal sustains strain weakening. This is related to fatigue, but is not strictly fatigue, since fatigue normally applies only to deformation in the elastic zone. Plastic deformation of metal is very complicated, and some metals have good tolerance for cold bending, while others (certain copper alloys) exhibit strain hardening under deformation, and become extremely brittle if bent beyond their elastic limit.
 
#22 ·
I like your replies Daniel! Not that it means much more than a hole in a snowbank but a physicist at the Army Corps of Engineers told me a nail also compresses, (shortens starting at the top), as the hammer strikes. As it completes the travel at each blow it rebounds to almoast the original length at the point. Is it getting to the point where we will need micrometers to measure our studs, and hammers with dial indicators to drive nails. I think I will go back to mortise and tenon with pegs.
 
#23 ·
I have seen those threads and could not understand whay anybody would suggest the shear strength of a screw was greater than the shear strength of a similar sized nail. A screw likely has an advantage in pulling out, which translates then into bending and tension situations, but not shear.
Screws plus PL Premium is my prefered way of holding wood together if I definately don't want it coming apart
 
#25 ·
We had a trimmer put up a set of pull down stairs with 3 inch screws, after a few pulling down the stairs they actually fell out of the ceiling.

We use 16D Nails now because of that.
Also try to pass inspection in my town with screws in your joist hangers.

Screws have their place as do nails.
 
#26 ·
Here's my 2 cents.

Comparing nails to screws, one isn't necessarily better than the other. There are three elements at play in a wood connection, shear, bending and pullout resistance. Initial failure of steel fastener connections in wood is always in terms of shear failure of the WOOD not the steel. The larger the diameter of the fastener the more surface area available to bear on the wood. When of the same length, a 16d nail with 5/32" (0.156") diameter has more bearing surface than a #8 screw (nom. 5/32" dia.) because the screws open threads reduce its available bearing surface. Thus, the fastener with the lesser bearing surface will shear the wood first. If a larger screw, say #11 (0.20" dia.) having equal available bearing surface is compared to the 16d nail, the wood shear failure point will be essentially equal. After the wood shears, the moment arm on the fastener moves further from the connection surface (greater bending moment) and the fastener begins to fail in bending. As fastener bending increases, pullout tension increases until the pullout resistance limit is reached and the connection completely fails. Here the screw is stronger because of its greater pullout resistance.

Rick
 
#28 ·
Arkitexas, I suggest you read the following useful document that discusses in detail the design of wood connections:

http://www.huduser.org/Publications/pdf/res2000_4.pdf

If you read this document, you will note that all fasteners (nails, screws, bolts, lag screws, dowels, pins) are designed essentially the same way, and they ALL fail in either shear or direct pullout. All bending loads on the wood elements resolved into shear forces on the fasteners, either single or double shear depending on the connection detail. If you think about it, you will realize that there are no practical joints used that allow fasteners to be loaded in bending. Direct pullout, as I noted in my previous post, occurs occasionally, specifically on roof sheathing due to wind uplift, and in other less common situations.

You are also mistaken in your belief that the wood always fails first. Bolts in particular may fail before the wood, especially if they are loaded in double shear. The design engineer must ALWAYS check both the wood and the fasteners to see which is more critical in failure, assuming that the wood fails first would cause you to skip the critical step of checking the fasteners in shear.

One circumstance in particular is very common, specifically horizontal shear at the neutral axis when building composite beams from individual pieces of wood nailed together. This is a case where the nailing must be checked for horizontal shear capacity.

You also noted that as the wood at a connection progressively fails due to shear force from fasteners, the moment arm on the fasteners increases, and the fasteners begin to fail in bending. This is incorrect. The total moment that a connection must resist is developed by shearing forces on the individual fasteners, not by bending forces on the fasteners. If the wood begins to fail in shear, the remaining fasteners carry additional shear load, which may well cause either failure of the fastener or failure of the wood, depending on joint design, which can and does lead to progressive failure of the joint, however this is not due to the individual fasteners failing in bending. Fasteners in wood always fail in shear, except for direct pullout failure. You will note in the reference document that there is no discussion about designing fasteners for bending moment OF THE FASTENER, since that type of failure cannot occur under an realistic scenario. The beams and joists can and do fail in bending, however you need to be careful to distinguish between failure of a structural member in bending versus failure of a fastener in bending, they are completely different problem.
 
#29 ·
How do you ever have time to learn all that Daniel?? You must have been the person Merlin was talking about in the Once and Future King. I think I got the grasp of most everything you say except "moment". I am sure it doesn't mean time which is what I need more of to look it up. Until then I will stick to straight framing, which I think most wanna be engineers should do, and leave the critical stuff to you.
 
#31 ·
How do you ever have time to learn all that Daniel?? You must have been the person Merlin was talking about in the Once and Future King. I think I got the grasp of most everything you say except "moment". I am sure it doesn't mean time which is what I need more of to look it up. Until then I will stick to straight framing, which I think most wanna be engineers should do, and leave the critical stuff to you.
Daniel...please correct me if I'm wrong....but this is how I remember it from my statics class........some 30 or so years ago....

Think of a torque wrench....the force you exert on the socket is the 'moment'...which is a function of the force on the wrench and how far away it is....
 
#30 ·
One thing's for sure, I always appreciate Daniel's thorough explanations where he doesn't try to wow us using complicated language and terms for those of us unfamiliar w/ a lot of engineering concepts and language. :thumbup: It's always a pleasure reading his explanations. Thanks as always!
 
#32 ·
Moment and torque are identical. They are computed by multiplying the force on the arm by the distance from the arm to the pivot point, so if I have a 2 foot long torque wrench, and I apply 100 lbs of force on the end of the wrench, the moment (or torque) is 200 foot pounds.

The distance often must be computed between the centroid of the applied force and the centroid of the loaded point. For example, your hand may be four inches wide, if you push down on the end of the wrench the moment is computed based on the distance from the center of your hand to the center of the bolt.

In the case of multiple bolts, things get considerably more complicated, since each bolt is likely to be a different distance from the applied force, so each bolt sees different load. The key point is that each bolt sees a shear force, not a moment, so the actual force on a fastener is always resolved into a shear force except in the case of direct compression (hammer blow) or direct tension (pullout). From the point of view of the bolt, it doesn't know or care if the bolt load arises from direct shear of the framing member, moment on the framing member, or some combination of the two, the bolt always feels shear. Whether or not the bolt fails is entirely due to the magnitude of the shear, and the strength of the bolt in shear. This applies equally to nails, screws, lag screws etc.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top