Joist blocking doesnt help deflection?! Edumacate me please

[Itsjustdirt]

I'm being told by a couple experienced floor guys that adding blocking (2x6 blocking between 2x6 joists) will NOT decrease deflection of the floor.

Can someone explain how having blocking every 2 feet (excessive, I know) DOESN'T spread the point load out over more nearby joists? When one joist deflects, the blocking is also trying to make the neighboring joists deflect, as well as their neighboring joists. How does that not help?

My in field "bounce test" sure says blocking works great but that's far from scientific.

What do you guys say?

Also, is there an actual test I can do to test deflection in the field? IE: Place 200 lbs in center of room and measure deflection?

 

[Jeff95TA]

The amount that the blocking will help spread the load depends on how thick/stiff your floor sheathing is since the sheathing is already helping to distribute the load. But since the sheathing is laying in the weak bending direction, I would think the blocking would at least help somewhat.

 

[lakeroadster]

Let's think about your bounce test.

You're point loading the floor (where your foot / feet lands on the floor). The sheathing spreads that load across multiple joists. Depending on how far apart the joists are spaced, the blocking helps reduce the span the sheathing covers. The reduction in flex is likely only the sheathing span reduction, not joist deflection.

 

[egdede]

I believe the (almost unbelievable) consensus finally was that a simple 2x4 on the bottom of the joist is surprisingly effective:

https://www.garagejournal.com/forum/showthread.php?t=88914

 

[My Old Tools]

Blocking prevents the joists from twisting and laying down. Watch a 1x6 on edge as it is loaded. It will bow in the middle and start rolling over. It is strongest straight up and down. So you are correct, blocking does make a floor stiffer and stronger by bracing the joists from twisting and by spreading the load. The sheathing, or sub-floor also helps but will not prevent a joist from bowing and rolling over on its own. And yes, a strap across the bottom is effective combined with the sheathing on top as it approaches a torsion box design.

A simple test is setting a stack of plywood mid span of a floor that is not yet sheathed. Watch how the joists react.

 

[wssix99]

Blocking prevents the joists from twisting and laying down. Watch a 1x6 on edge as it is loaded. It will bow in the middle and start rolling over. It is strongest straight up and down. So you are correct, blocking does make a floor stiffer and stronger by bracing the joists from twisting and by spreading the load. The sheathing, or sub-floor also helps but will not prevent a joist from bowing and rolling over on its own. And yes, a strap across the bottom is effective combined with the sheathing on top as it approaches a torsion box design.

This is it.

I'm being told by a couple experienced floor guys that adding blocking (2x6 blocking between 2x6 joists) will NOT decrease deflection of the floor.

They are not understanding that increasing the longitudinal stiffness (with stiffeners or blocking) of wood joists also increases strength. However; the last time I looked at the numbers involved, I recall that this effect is much much greater for deeper joists vs. something as small as a 2X6. (The design assumption/calculation for a 2X6 may be that over it's safe span, which is short, blocking may not add any material strength.)

 

[ericlar80]

I'm being told by a couple experienced floor guys that adding blocking (2x6 blocking between 2x6 joists) will NOT decrease deflection of the floor.

Can someone explain how having blocking every 2 feet (excessive, I know) DOESN'T spread the point load out over more nearby joists? When one joist deflects, the blocking is also trying to make the neighboring joists deflect, as well as their neighboring joists. How does that not help?

My in field "bounce test" sure says blocking works great but that's far from scientific.

What do you guys say?

Also, is there an actual test I can do to test deflection in the field? IE: Place 200 lbs in center of room and measure deflection?

You are both right. The blocking will absolutely reduce deflection and bounce when used since it helps transfer a point load into adjacent joists. However, when calculating the deflection, you cannot take credit for additional load carrying capacity since the calculation usually assumes an evenly distributed load. There is no modifier to the math for using blocking or not.

 

[Itsjustdirt]

Thanks! So basically..... It helps in real life but an engineer can't calculate exactly how much it helps, therefore it doesn't help.

 

[wssix99]

Thanks! So basically..... It helps in real life but an engineer can't calculate exactly how much it helps, therefore it doesn't help.

Real life engineers can and do calculate this.

ericlar80 is correct in that designers don't "get" anything back for adding blocking.

Instead, designers subtract load capacity of a structure when blocking isn't used through a multiplier called the Beam Stability Factor: http://www.wikiengineer.com/Structural/BeamStabilityFactor

(Full disclosure - my recollection above was incorrect. 2X6's as beams perform really poorly.)

Generally, designers (in order to get the full capacity of a beam) will specify required blocking to get the stiffness needed for a 1.0 beam stability factor. If one is in a situation where blocking "helps", there could be some other issues going on. Perhaps sufficient blocking was not used originally, the decking may not be attached or glued properly, etc.

 

[QwikKotaTx]

Thanks! So basically..... It helps in real life but an engineer can't calculate exactly how much it helps, therefore it doesn't help.

Generally we ignore minor increases in strength such as you are talking about to remain very conservative in the load capabilities. It is always best to add to your safety factory rather than cut down. Aerospace is a different animal and I have had to point out safety factors of less than 1.25 to project managers. It's really odd when they try to argue about it with you.

 

[bullnerd]

First thing I thought of is why are their 2x6's used for floor joists?

Is the plywood attached yet or is this new construction?

 

[ericlar80]

Real life engineers can and do calculate this.

Engineers can, but they don’t for residential construction. It’s mostly very conservative, using basic equations that are simplistic models, and not taking credit for nuanced differences.

Aerospace engineers (like myself in medical devices), have a reason to optimize the design for weight, performance, and cost. Structural engineers for residences will just have you overbuild and not worry about it; this keeps both the engineering costs and liabilities low.

 

[ericlar80]

Thanks! So basically..... It helps in real life but an engineer can't calculate exactly how much it helps, therefore it doesn't help.

They can, it just becomes very esoteric, and costly to do the engineering. It’s easier, cheaper, and more obvious to just put more joists in, or make them larger.

 

[wssix99]

Engineers can, but they don’t for residential construction. It’s mostly very conservative, using basic equations that are simplistic models, and not taking credit for nuanced differences.

Not everyone lives in a McMansion derived from standard tables.

Particularly with wood being a carbon neutral material, engineering for residential structures is still very much alive. However, as bullnerd points out, 2X6's for floor joists are kind of nuts. I can't even start to imagine where that idea came from. (If it's suspended. I actually have a 2X6 raised floor, but it's fully supported on top of another floor - and I have the hell blocked out of it!)

 

[ericlar80]

Not everyone lives in a McMansion derived from standard tables.

You're right. The reason I said "most" was because most houses are not built anything like what you show.

 

[Itsjustdirt]

2X6's for floor joists are kind of nuts. I can't even start to imagine where that idea came from.

Yup! 1950s built on side of a hill with a basement. All the floors (ceiling to basement) are 2x6's, 12" on center with many spanning as long as 12 feet! The floors had a lil bounce. We are already limited on room in the basement so sistering anything taller wasn't an option. The best solution I could come up with was sistering 2.0 Microlam LVLs 1.75x5.5 and then blocking every 2 feet, topped with Advantech 3/4" subfloor (limited on height of surrounding floors). The floor is very noticeably stiffer now. Hopefully its stiff enough for tile.

 

[Superbec]

Hopefully its stiff enough for tile.

if you don't know about uncoupling membranes you should research them a bit , Schluter Ditra is one , and I think Ardex makes a free floating one

 

[bullnerd]

Glue the subfloor down?

 

[matt_i]

My answer would be that the subflooring already is transferring the load to adjacent joists so the side-blocking isn't doing much more to assist.

The idea mentioned above of adding a 2x4 to the bottom edge of a 2x6 would help, 2x6s sistered would be the most effective. Even if they can't be weaseled in to have bearing on both ends, it will still be effective.

Suppose your original joist is 6 feet long and has 3" of bearing on each end. Then you sister a joist that just slips in at 5'6" long but has no bearing on either end....sister via typical array of 16d nails, clinched and construction adhesive in a box-X formation. This will help tremendously.

 

[Itsjustdirt]

if you don't know about uncoupling membranes you should research them a bit , Schluter Ditra is one , and I think Ardex makes a free floating one

Oh yeah! I forgot to mention the Schluter Ditra. I use that on EVERYTHING that gets tile (except showers)

 

[Viper98912]

but an engineer can't calculate exactly how much it helps

 

[johninct]

2" x 6" always bounce. Technically, the extra blocking is increasing the dead load, so the deflection should be increased. If deflection is a problem for you, you have big problems.

 

[gearhead1]

It can be calculated, but you need FEA software to do it. That’s because it would take a very long time doing the matrices to figure it out without a computer.

It could be measured also. Build it, the use extensometers/LVDTs or something to measure the deflection. Then add the blocking, and load it with the same load and take the measurements again.

When a long beam (floor joist) is supported on the ends and loaded, the beam twists. So, blocking is installed to prevent the twist.

Take a wooden yard stick and place flat between 2 chairs 35” apart. Now take the yardstick and rotate it vertical so it looks like a floor joist. Holding it vertical with your fist in the center, push down and watch what the ends do, they twist. Blocking prevents that twist.

As previously mentioned, the blocking helps a little bit to transfer load from joist to joist, but the subfloor also does that, but that’s not the primary purpose of the blocking.

Simpson makes metal rafter ties that get nailed on that prevent the twist. They’re used because they’re faster to install, pull them out of the box and put on. Blocking requires someone to cut each piece.

 

[gearhead1]

Here are the joist ties.

https://www.homedepot.com/p/Simpson...VTNbACh1b5wVrEAQYAiABEgKLsvD_BwE&gclsrc=aw.ds

 

[jkuro]

Sister up another 2x6, glued and screwed, to your existing joists. Then add some blocking.

 

[firebirdparts]

Thanks! So basically..... It helps in real life but an engineer can't calculate exactly how much it helps, therefore it doesn't help.

There is an additional factor. When you say blocking "obviously" transfers load sideways, you are assuming that it can't just wilt under load.

With wood joinery, you don't presume to be able to hold it together in that situation with the sort of fasteners we use. To fail entirely in real life, all the blocking has to do is spread the joists out a little bit and slip on a nail in tension, and a lot of people wouldn't even nail it that way to start with. If that happens, it is carrying nothing. If you don't understand this, I sympathize, but I can't really explain it very well. Wood is soft, so we just don't assume we can control it in that situation to the degree required. If it was steel, yes.

*So* if you combined blocking with plywood subfloor you'd have a noticeable effect from the blocking above the effect from the plywood. If you put plywood top and bottom, then you're getting closer to something that an engineer would be willing to consider mathematically. in that structure the blocking joinery is no longer required in tension for it to work.

 

[WNYflyer]

- for a floor subjected to a uniform psf load over the entire floor, blocking has no effect since all the floor joist will essentially deflect the same since they are equally loaded.

- For point loads, blocking will distribute portions of the point load to adjacent floor joist thus reducing deflections to some unknown quantity. For typical wood construction no engineer or architect who knows what they are doing would take this fact into account when sizing the floor joist.

- yes an engineer/architect could use a column-beam computer program to figure deflections for a floor joist system subjected to point loads while utilizing blocking. I doubt the numbers the program spits out can be relied upon though due to the unknown rigidity of connections, non homogeneous nature of lumber, etc. Plus typically nobody is going to pay you to expend the effort unless it is some kind of research project.

- Some of the old codes state blocking or cross bridging shall be installed at 8' max o.c. or something like that. I highly doubt there was any truly engineering basis for this criteria and was simply passed down through the years as "its always been done that way" or is "good practice". You would be surprised at some of the criteria in prescriptive building codes that have no true engineering basis but have been brought forward through the years because history shows they work and are good practice. For a normal residential wood framed floor current residential codes don't require blocking/bridging unless floor joist are greater than or equal to I believe a nominal 2x12. Of course some blocking maybe required for other reasons.

 

[firebirdparts]

I think blocking on 8' centers by code was totally to keep the joists upright. Just my opinion, though.

 

[matt_i]

When a long beam (floor joist) is supported on the ends and loaded, the beam twists. So, blocking is installed to prevent the twist.

I snipped this out because it deserves more discussion...technically while the beam is stiff in the vertical direction when its aligned in that usual way, its floppy in the other perpendicular direction (across the 1.5" width). When the loading is not perfectly aligned with the vertical axis of the beam (very possible due to the natural bowing, twisting, dog-leg nature of dead wood) the resistance to deflection shifts more and more to the floppy axis. Worth mentioning is wood is not a homogeneous-ish substance like steel, there are many changes and imperfections in the grain structure on any given board. A couple generations back when they cut the old growth forests this was less of an issue but prevalent with the sustainable wood of today.

Tying the top flanges together via the floor underlayment (ply, osb) helps tremendously as it at least constrains one surface. The side-blocking or X-bracing helps constrain the rest of the cross-section.

Another factor is that an open-cross section like a channel loaded vertically with orientation like C is also a way to automatically get twist, a closed cross-section like solids, tubes or I-beams do not inherently twist in the same way.

 

[WNYflyer]

I think blocking on 8' centers by code was totally to keep the joists upright. Just my opinion, though.

Probably correct that it is at least one if not the reason for the bridging. Probably good practice when guys are walking the joists or laying down materials when no sheeting is down to stabilize the joist during during construction.

They have similar OSHA code requirements for horizontal/cross bridging etc. when erecting steel roof joist like you see above you in Home Depot, Lowes, etc. The OSHA requirements are to stabilize the joist during erection and make it much safer for the iron workers walking and setting the steel and laying down materials.

 

[rust in the eye]

Anything that spreads the load is going to decrease deflection.
Where did your flooring guy earn his engineering degree?

 

[wssix99]

- for a floor subjected to a uniform psf load over the entire floor, blocking has no effect since all the floor joist will essentially deflect the same since they are equally loaded.

This is not correct. As wood floors are loaded, the members twist and weaken compared to what they can hold oriented fully in-plane with the loads.

- For point loads, blocking will distribute portions of the point load to adjacent floor joist thus reducing deflections to some unknown quantity.

No loads are transferred by wood blocking because they are pin joints and don't resist moments. This is not the case with some steel and concrete floors, which can be engineered to behave in this way.

 

[wssix99]

All the floors (ceiling to basement) are 2x6's, 12" on center with many spanning as long as 12 feet! The floors had a lil bounce. We are already limited on room in the basement so sistering anything taller wasn't an option. The best solution I could come up with was sistering 2.0 Microlam LVLs 1.75x5.5 and then blocking every 2 feet, topped with Advantech 3/4" subfloor (limited on height of surrounding floors). The floor is very noticeably stiffer now. Hopefully its stiff enough for tile.

All of this will help. If you are going to invest in renovations for a bathroom and more (I assume that is the case.) I would highly recommend investing in a structural engineer to do the calculations. They can calculate the stiffness of the resulting floor and give you assurance that your tile won't crack and/or you won't be able to feel the bounce. This analysis should cost you a few hundred dollars and is great insurance when you are putting thousands (and possibly a grumpy spouse) at risk.

If you only have 5.5" to work with, the answer is just "more." The engineer can calculate how much more. If you had to, and could tolerate making a few holes in the exterior walls for installation, you could even do steel and be rock solid. (That requires a lot more engineering and $$$, though.)