DIY Home Improvement Forum banner

OSB Stronger than Plywood?

50K views 15 replies 8 participants last post by  ToolSeeker 
#1 ·
I posted a while back about putting ceramic in my kitchen, 16' span, 2x12 floor joists, 16" on center, 5/8 plywood sub floor. I received some excellent advice and learned that my deflection was L/444.

http://www.diychatroom.com/showthread.php?t=22526


So I've decided to laminate to the sides of my floor joist to cut down on the deflection. I was going to use 3/4 plywood but then did some checking, as most I-Joists that I have seen use OSB webs. It looks to me like 1/2" OSB actually has a stronger shear rating than 3/4 plywood. Shear is the only area that OSB beats plywood and high shear strength is what you are after for a joist application. Here's a link to plywood and OSB specs.

http://www.canply.org/english/products/comparison/comparison_all.htm

What do you think?
 
#2 · (Edited)
Weathermaker:

In a nutshell, NEITHER OSB nor plywood bonded to the sides of your joist will double the strength of your floor joists. You'd be much better off sistering your joists than you would be to laminate them with either OSB or plywood. The reason why is that it's tension at the bottom of the joist that gives the joist strength to resist bending. Adding more wood to the bottom of your joists means there's more material there that will carry the tensile load of a beam that wants to bend. And, to maximize the tensile strength of wood, you want all the wood cells oriented along the lines of those tensile forces. In OSB and plywood, a very large percentage of the wood cells are going in the wrong direction to carry a tensile load. You want them going in the same direction as those in the joists.

You just have to imagine hanging a heavy weight from a hole in a board. If the wood grain in the board is going up and down, the wood grain is going in the right direction to carry a tensile force. If you turn that board so that the wood grain is horizontal, you have a much better chance of breaking that board as it will more easily break along the wood grain through the hole. Consequently, adding 3/4 inch of OSB or plywood on each side of a joist DOES NOT double the strength of the joist. To do that, you need to use lumber on each side of the joist so that the wood grain is going in the direction where it's most able to resist a tensile load. In plywood, almost half the plies have the wood grain going in the direction LEAST capable of carrying a tensile load. In OSB, the wood grains are going all over the place, so OSB wouldn't be able to carry a tensile load as well as lumber either. If you want to increase the strength of your joists by adding meat to the sides of them, then you need to add LUMBER, which has all the wood cells going in the same direction, and add that to the sides of your joists. Attaching OSB or plywood to the sides of your joists won't increase the strength of the joists nearly as much as adding lumber, and it's all cuz of the direction of the wood grain.


Consider the following:

For a beam with a uniformly distributed load the deflection in the middle of the beam is:

a) according to this web site it's (5/384)(wl**3/EI)
http://www.engineersedge.com/beam_bending/beam_bending1.htm

b) according to this web site it's (5/384)(wl**4/EI)
http://www.uoregon.edu/~struct/courseware/461/461_lectures/461_lecture40/461_lecture40.html

Where w is the load per foot of length, l is the length, E is the modulus of the material the beam is made of and I is the moment of inertia.

For a rectangular beam, I is bh3/12, and you can confirm that at this web site:
http://mysite.du.edu/~jcalvert/tech/beam.htm
where it says:

"Two useful examples, especially for wooden beams, are the rectangular beam of height h and width b, and the circular beam of diameter d. The moment of inertia of area of the rectangular beam about a centroidal axis parallel to the width is I = bh3/12, and for a circular beam, I = πd4/64."

Basically, the parameter "E" is proportional to the strength of the material the beam is made of and the parameter "I" accounts for the shape of the beam. When you turn a rectangular beam like a 2X12 on it's side, the width, "b", now becomes the height "h", and the height "h" now becomes the width "b". Thus, which dimension gets cubed is critical in determining the value of "I" and hence the deflection of the beam. This is why a fir 2X12 can be used as EITHER a very rigid floor joist or a really bouncy diving board. It's all in which dimension gets cubed, and the direction of the applied force.

Now, according to the formulae above, whichever one is right, you can lower the deflection of your floor a little by decreasing the load, "w" on it per linear foot, or a lot by decreasing it's span, "l" since that dimension is cubed or fourthed.
Similarily, you can decrease your floor's deflection by making your floor joists out of a stronger material, "E", like steel instead of wood, or by using a beam whose shape, "I" provides for greater rigidity, like using a rectangular beam instead of a round beam, or ensuring the 2X12 remains upright with blocking instead of installing it in the horizontal orientation.

Now that your house is built, there's not much you can do about any of those parameters, except the amount of weight on the floor and that Moment of Inertia, "I", that reflects how much the SHAPE of the beam promotes greater rigidity. Let's assume you don't want to get rid of your dog, cat and water bed in order to reduce "w", and hence the weight on the floor. We are then stuck with changing the shape of the beam.

Since I = (1/12)(bh**3) = (b*h*h*h)/12

we can immediately see that doubling the width of the beam, "b" which is your gameplan, will double the calculated value of "I", thereby reducing the floor's deflection by half what it was before.

However, we can also immediately see that if we make the beam taller, say by 2 inches, that this will have a more pronounced effect on the strength of the beam because the height of the beam is cubed.

For example, if we have a 2X10 that's 9.5 inches high, and we add 2 inches of wood to the bottom of it, we now have a 2X12 that's 11.5 inches high, and the calculated value of I becomes higher by:

(11.5/9.5)**3 or cubed, or 177 percent, or most of the way to double.

If we were to add two and a half inches of wood to the bottoms of the joists, then the calculated value of I becomes 202 percent larger, or double.

That is, glueing and screwing a fir 2X3 to the bottoms of your fir floor joists will strengthen your floor AS MUCH as sandwiching each joist between 3/4 inch thick lumber.

All that is necessary is:

1. That the glued joint between the new wood and old wood be as strong as the wood or stronger, and

2. That the wood you add be as strong as the wood the joist is made of, or stronger.

But, you do loose 2 or 3 inches of headspace in the basement below.

What I think you're missing in your understanding of engineered wood is that the stresses in the OSB web of the joist are weaker than they are in the lumber flanges at the top and bottom. In order to increase I the most, it's important to add material at the top and bottom flanges of the joists, not to the web.

aside: If you ever watch a science fiction movie made in the 50's, they show space ships with holes in the webs of the I beams that keep the space ship rigid. The reason for cutting those holes in the web, but not the flanges, is because there is less stress in the web of an I beam as there is in the flanges, so the web doesn't need to be as strong as the flanges are. And, by cutting round holes in the flange, you remove the most weight (and also weaken the web the most) in the middle of the web (the "neutral axis") where the stresses in the beam are smallest and you therefore need the least strength. And of course, the purpose in cutting those holes is to reduce the weight of the beam, which is always of paramount importance in putting anything into orbit.

Back to wood beams...

Also, what's important in the calculation is that the "modulus", "E" of the wood you're adding be as high or higher than that of the existing wood. The modulus is not related to shear strength. It's really the slope of a stress strain curve when the wood is in tension, and is therefore a measure of it's tensile rigidity.

So, if it wuz me, I wouldn't use EITHER OSB or plywood because neither would have the same tensile rigidity as your lumber joists because both OSB and plywood have a high percentage of their wood cells going in the wrong direction to carry a tensile force. You need all the wood cells going in the same direction as they are in your floor joists for maximum strength in tension (and therefore bending because it's the tension in the wood at the bottom of the beam that resists bending).

If it were me, and I wanted to laminate the joists, I would use 1X6 lumber; two pieces on each side of the joist with the joint between them in the middle of the joist's height, (which is the "neutral axis" where stress is lowest). But, it would be easier to sister your joists to achieve the same increase in strength. Alternatively, I would glue and screw wood to the bottoms of your floor joists using LePage's PL Premium construction adhesive to ensure the bond was as strong or stronger than the wood. You could also use LePage's PL Premium to glue square cross section steel to the bottoms of your joists to also increase the strength and rigidity of your beams, but it's been too long for me to remember how to calculate the increase in strength.

That's what I think. Let's see what others think.

PS: You can confirm everything I'm proposing by printing off this response and getting anyone at your local university department of mechanical or civil engineering to read it and comment. If you corner a student into reading it, make sure he's at least 2nd year. They don't take this stuff in first year. I graduated from the U. of Manitoba with a B. Sc. degree in Mechanical Engineering in 1978, so I remember seeing this stuff, but haven't used it in 30 years.
 
#3 ·
Holy crap Nester, and people say I spend too much time here.....:laughing:

Even if all you wrote is true, where did you get the idea that weathermaker wants/needs to double the deflection rating of his joists? He doesn't. His joists are stiff enough already. All he wants to do is stiffen things up a little more, I believe.

As we know, sometimes it is very difficult if not impossible to sister joists. Laminating and screwing ply or OSB to the joists will improve the structure to a degree. How much? No way to tell. Maybe instead of laminating he could sister the center 2/3's of the span, and/or sister with smaller joists if obstructions prevent using the full size.

Jaz
 
#4 ·
Holy crap Nester, and people say I spend too much time here.....
I haven't got a life.

Even if all you wrote is true, where did you get the idea that weathermaker wants/needs to double the deflection rating of his joists? He doesn't. His joists are stiff enough already. All he wants to do is stiffen things up a little more, I believe.
It is true. All you have to do is read it and follow through it to convince yourself it's true.

Anything he does will "stiffen things up a little more". But, if he understands the technical aspects of what he's doing, he'll get more bang for his buck, and more bang for the time he spends strengthening his joists.

Laminating and screwing ply or OSB to the joists will improve the structure to a degree. How much? No way to tell.
If he gets the modulus "E" for plywood and OSB, any second year mechanical or civil engineering student could calculate how much of an improvement that laminating would result in.

Maybe instead of laminating he could sister the center 2/3's of the span, and/or sister with smaller joists if obstructions prevent using the full size.
Maybe.
 
#5 · (Edited)
Nestor, you spent some time on that reply, interesting reading.

JazMan, correct I was just looking to stiffen things up a bit.

----------------------------------------------------------------------

It would be just about be impossible to sister in another 2x12. I don't really want to drop my ceiling either. I'm just looking to decrease my deflection a bit as I believe I'm currently near the maximum.

I had a look at I-Joists and the factors effecting strength are flange size, web depth and web thickness. I'm not going to get really wordy, but just food for discussion. When the OSB is laminated to the existing 2x12 does the 2x12 not then become the flanges? If the center 7" were cut out of the 2x12, is it not then similar to an I joist, from an applied stress perspective. Obviously taking into account that the engineered I-joist has the OSB web bonded together in a continuous length and the flanges are bonded near perfectly to the OSB.

If the laminated OSB were to give me a nominal drop in deflection, for the sake of conversation say from L444 to L390, would it make it worth the effort of installing it?

Thanks,

Ken
 
#6 ·
Ken,

I'll let Nester answer the engineering part. I only require the proper joists and spec how tile is to be set. I use a span chart for correct framing.:thumbsup:

BTW, L390 is more deflection than L444. To decrease deflection you want a higher #. For example L360 is max allowable deflection for ceramic, while L720 is half the deflection (1/360 vs 1/720), and is the max allowable deflection for natural stone tiles. It's kinda like gear ratios, the higher the number, the lower the gear.:whistling2:

Jaz
 
#8 ·
I had a look at I-Joists and the factors effecting strength are flange size, web depth and web thickness.
Every dimension affects the strength and deflection of a beam, but the most important ones are the flange size, web depth (which is the "h" in the calculation of I) and the distance between supports. The web thickness (which is "b" in the calculation of I) plays a much less important role in determining how strong the beam is and how much it will deflect under a given load.

When the OSB is laminated to the existing 2x12 does the 2x12 not then become the flanges?
No. What happens if you laminate both sides of your 2X12 with OSB is that you now have to calculate the composite value of "I" by calculating the contribution of each piece of wood; one 2X12 and two pieces of OSB.
Since the OSB is weaker in tension than the lumber, it won't contribute as much strength as lumber of equal dimensions. Also, if the OSB pieces aren't as high as the 2X12, they will contribute even less strength. You don't need to add material over the whole length of the joist. The stresses are greatest in the middle of the span, so adding material there is going to make the most difference.

So, simply making the web wider with OSB is not going to contribute much strength to the joist. If you used 1X6's instead of OSB, and put two on each side of your joist, then the resulting beam would give you about the same strength as had you sistered the joist. Ditto for plywood, as every second ply in your plywood is going to have the grain going in the wrong direction to carry a tensile force.

If the center 7" were cut out of the 2x12, is it not then similar to an I joist, from an applied stress perspective. Obviously taking into account that the engineered I-joist has the OSB web bonded together in a continuous length and the flanges are bonded near perfectly to the OSB.
If you made a 2X12 thinner in the middle 7 inches, say by routering out a half inch on both sides, you wouldn't weaken the joist significantly. It's the material at the top and bottom of the joist that is under the most stress when the beam bends, and therefore contributes the most force to straighten out the beam and prevent it from bending. The material in the middle of the joist is under the least stress and therefore contributes the least to the tensile and compressive forces trying to prevent the beam from bending.
What you're talking about is similar to my analogy about cutting holes in the I-beams of the "space ships" of 1950's sci-fi movies. The middle of an I beam contributes the least strength, so by removing material from the web, you lower the weight MORE than you lower it's strength.


You could also get a much greater increase in strength and rigidity if you added a stronger material to the bottoms of those beams rather than laminate them. For example, if you glued and screwed a 1 1/2 inch wide by 1 inch thick steel bar to the bottoms of the joists, you would get a much greater increase in strength than if you glued and screwed the same amount of wood to the bottoms of your joists. If you'd asked me 30 years ago how much more strength, I could have worked it out for you.

Similarily, if you were to use an inexpensive hardwood like ash, you could get a greater increase in strength than by using a softwood like fir.
 
#13 ·
Aren't most of the tile mastics today made with a latex base, and therefore able to deal with some deflection? To add to that incredible post by nestor, those T joists depend on the bottom 2 x going into tension under load, not so much the web.

It seems to me that steel banding at the bottom of the joist would do more than adding dead weight, no?
 
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