Joist stiffening/sistering with partition wall in the way

[Kodi Crescent]

I'm considering sistering some joists to strengthen the floor in an attic bathroom that is about to receive tile a tile shower and a quartz countertop.

The joists are 2x10, 16" oc, with a 12' span.  This would be a very simple job, had I thought about this before I built the bathroom walls.  But, I didn't think of this, and now I have a partition wall built above the subfloor about mid span.

I plan on cutting the subfloor open on the inside and outside of the bathroom so I can access the bearing points.  I'm concerned that I can't quite slide the joist through because of the partition wall in the way.  Is there a way to deal with this?  Thanks!

 

[Peter_C]

Maybe use 2x8's? They should provide enough strength for tile, and even stone. Not sure I understand how the wall comes into play without a photo but it doesn't sound good. You need to do 2/3 of the joist when sistering. Obviously edge to edge is best, especially if you can get the ends supported. Glue them with PL400 or another type of glue and either screw them together or nail them. I like to clamp them first to pull them together.

Use this load calculator.http://www.johnbridge.com/vbulletin/deflecto.pl

 

[Daver]

Kodi,

Cheese and Holmz made some interesting contributions to your previous query regarding floor loading.
Removing subfloor for joist cavity access and subsequently replacing it ("patching in") will not yield as stiff a subfloor as pulling up the existing flooring for the bathroom and then replacing it after the sistering work is done.
Your attic floor will be stiffer if you take up the whole bathroom floor and corresponding (covering the rest of the 12' span of those floor joists) floor area outside the bathroom/on the other side of your partition wall, and then replace all of it after the sistering is complete.
This is definitely preferable to just pulling up the area over the two cavities and then repairing them.
Also cut out the subfloor underneath the partition wall that's above the sister cavities. Removing these 3/1/2"x14 1/2" sections of subfloor will make life easier when you want to roll your 2x10s up to mate with the existing joist.

As for the sistering of the two joists;
I'd get (4) 2x10x8' joists, as knot and defect free as you can buy. Be sure to inspect them at point of purchase for twist, warp, crown, etc. as these aspects will make the job more difficult. I'd also double-check their height (the "10" dimension) against your existing attic joists before install. Rip as necessary to remove any crown or small discrepancy if the new joists are marginally taller than the existing.
After your subfloor is up (and any bridging/blocking removed), it's helpful to inspect the existing joist thoroughly with a flashlight. Note any fasteners from the ceiling below, etc., that would provide obstructions to mounting your sister joists to existing. Remove/repair as necessary. Check both sides of each joist*.

From your previous thread, it sounds like there's more room from the bathroom side of the partition wall to feed sister joists into the cavity.
Take an 8' 2x10 and lay it down and feed it through.
Take a 2nd 8' 2x10 and drop it into the next cavity.
At this point, you should have a new 2x10 on either side of one of your joists to be sistered.
One 2x10 should be butted against the bearing wall in the bathroom, the other should be butted against the other bearing wall, 12' away.
Roll/stand them both up so that they are parallel to the existing joist. Clamp in place.
Drill paired holes the length of the assembly;
Make 2 holes, top hole is 1/4 to 1/3 down from top edge of joist, bottom hole is 1/4 to 1/3 up from bottom edge of joist.
1st pair of holes is 6" from bearing wall. Each subsequent pair is 12"-16" from there.
Hole diameter for 'tails' (the 4' of sister, starting at the bearing end) is 5/16". The tails will be lag-screwed in with 5/16" x 3" lag screws.
Hole diameter for center (where you have a 'triple': existing joist flanked by 4' of sister on each side) is 5/8". The center will use carriage bolts, washers, and nuts. If you can get nuts with integrated nylock washers, that's great insurance against them loosening. 6" carriage bolts should suffice.
When you've done all of your drilling, UNCLAMP.
Lay the sisters back down on their faces. Spread construction adhesive on the face of each sister. Tip up in place. Be sure to push one carriage bolt through in the middle and run a pin (anything to be sure the lag holes in the tail are aligned) through the tail.
Clamp up. When you're satisfied that the sisters are where you want, that they don't have any crown above the existing floor joist, and that all looks good, screw the assembly together.
At this point, you have one joist sistered. It's not the same as putting in a 2x10x12', bearing on both ends, but in this instance it's arguably the next best thing.
You now have one joist sistered.
Repeat this process for the second joist.
Now you're ready for subfloor.
Treat yourself to a cold beverage.

[wink]
Dave

 

[Kodi Crescent]

Thank you! 

I think I'm going to play it safe and hire a structural engineer to come in and evaluate.  He may tell me I don't have to do anything.  We'll see.

 

[Cheese]

I'd get (4) 2x10x8' joists, as knot and defect free as you can buy. Be sure to inspect them at point of purchase for twist, warp, crown, etc. as these aspects will make the job more difficult. I'd also double-check their height (the "10" dimension) against your existing attic joists before install. Rip as necessary to remove any crown or small discrepancy if the new joists are marginally taller than the existing.

I agree with everything [member=23629]Daver[/member] suggests except I'd go directly to LVL's or LSL's (engineered lumber) and that way you don't have to worry about knots, warp, twist or crown. Just measure them for height (width) and install just as Daver suggested. They're a little bit more expensive but you're not using a ton of them. The only real downside is that you may need to special order them.

FWIW...I've had 4ea 12' 2x4 LVL's sitting on 3 small wood blocks on the cement floor of an unheated, unattached, damp floor garage for the last 9 months and when I used them last week to frame in a wall they were all just as straight as when they were purchased.

 

[bobfog]

Can I ask why you feel they need strengthening?

2x10, 16" OC and a 12' span seems perfectly adequate to me.

 

[Daver]

[member=8843]Kodi Crescent[/member]

Please post engineer's assessment when you receive it. I'd be interested to read it as, I'm sure, others would be.
Always money well-spent to engage the services of an expert.

[member=44099]Cheese[/member] Good call on the laminated veneer/laminated strand.

[smile]
Dave

 

[Michael Kellough]

Tight blocking adds strength and is pretty easy to add.

 

[Kodi Crescent]

[member=23629]Daver[/member]

I had the engineer come out.  In his report he said that the structure was adequate for what I was proposing, although the joists would deflect a 1/4", and perhaps a bit more over time due to seasonal shrinkage.  Even though the joists would deflect, he did not believe the deflection would be sufficient to crack the tile, and it would not be sufficient to crack drywall.

He stated that IF I wanted to do something, I could sister the two joists, or I could install blocking with the adjacent joists at 24" intervals.  I could have the installer use lighter weight materials, which may cost more, but would reduce the deflection. 

I've taken a look at what it would take to do either of those recommendations, if I'm concerned about the deflection.  I'm not sure what to think about that yet.

If I were to do something, the thing that makes the most sense is to open the drywall in the ceiling under them and do the blocking.  I'm not sure I can easily get to the joists next to each load supporting joist to block effectively.  There are wires and other things in the way that may prevent sistering, or make sistering much more difficult.  I've asked the contractor for a quote using lighter weight materials.  I may pursue that, or just have the install done the traditional way and wait and see what happens.  Something may happen, or nothing may happen.  I'm still indecisive.

 

[Michael Kellough]

Another option is to add steel angle to the top and bottom sides of the joists. That might allow you to work around the existing utility lines. Stiffening the joists locally will transfer the load so it might be an issue for the wall tile instead of the floor tile.

The traditional way to add tile to floor that wasn't designed for it is to use small (1" square) tile that is virtually impossible to break.

Now you have Ditra which allows use of larger tile.

 

[Daver]

Sounds like you got some good feedback from the engineer. Well worth the cost, IMO.
Hopefully the contractor can fill in more of the puzzle with respect to the use of lighter materials.
It's always a judgment call but the more information you have, the easier the decision becomes.

Dave

 

[Holmz]

Think of your 16" OC joists as the core of a honeycomb cored composite structure.
In that structure all the the loads are tensile and compression in the face sheets and the core is in shear.

For you structure this means that the long axis of the floor has the joists going the right way, but orthogonal the joist are not helping shear.
It would be better to have sections going 90 degrees to the joists.
Then if you had a face-sheet on each side the floor would be a composite structure. Of course the line from the shower and toilet you want to end up somewhere, as well as some ports for servicing the drains.

Ideally one could mock up a 12 foot by 16" (+2) section with a face-sheet on one side and test for deflection, then add the opposing face-sheet and test that.
Then mock up the cross ways deflection as 16" pieces (even leaving out the dividers that represent the 12' pieces), and test that.

Basically a bottom face-sheet will be in tension and will add a lot more to the stiffness than more joists for the same poundage of wood.

Another option is to add steel angle to the top and bottom sides of the joists. That might allow you to work around the existing utility lines. Stiffening the joists locally will transfer the load so it might be an issue for the wall tile instead of the floor tile.

The traditional way to add tile to floor that wasn't designed for it is to use small (1" square) tile that is virtually impossible to break.

Now you have Ditra which allows use of larger tile.

This scheme could also be tested for deflection.

 

[rizzoa13]

If your worried about the joists deflecting and not being able to get joists sistered in the tight space you could get some flat stock steel, maybe 1/4" thick and laminate each side of the joists. Even a 4" wide piece of steel on both sides, through bolted through the joists and the steel on the opposite side would really do a lot to keep the joists deflecting.

You'd have to open up the ceiling below but the relatively thin steel would be a lot easier to get in around wires and pipes. You could then add some 2x blocking between joists as another assurance if you absolutely can't sleep at night.

Quick story- I placed a huge old cast iron claw foot tub in my upstairs bathroom in the middle of a large floor span. I had to cross a lot of joists with the drain line so needed to start at the top of the joists to keep a good pitch and end up where I needed to be. That's not kosher by any means to notch so much out of the joist. So I got some large 5/8 plate steel and made some saddles for the joist. Welded them up and bolted it on either side of the notches. Mortised the top of the joist down 5/8 so it'd sit flush and put them on all the joists. No movement even with a full and I haven't gotten a tile crack or drywall crack in 5 years.

Long story short- steel is your answer.

 

[Cheese]

...although the joists would deflect a 1/4", and perhaps a bit more over time due to seasonal shrinkage.  Even though the joists would deflect, he did not believe the deflection would be sufficient to crack the tile, and it would not be sufficient to crack drywall.

The tiles may not crack but if there is 1/4" deflection, the grout will probably crack and that starts the process in the wrong direction. I'm renovating a bathroom now that has that exact problem, not because of joist deflection but because of subfloor deflection. I couldn't understand why every 6-9 months I had to reattach small ceramic tiles to the floor until I ripped out the floor and found that the subfloor was only 1/2" thick, consequently, the subfloor was flexing between the joist cavities, it doesn't take much.

 

[Michael Kellough]

Cheese brings up an important additional but separate concern. The joists could have zero deflection but if the subfloor is inadequate...

A 1/2" subfloor cab easily sag 1/8" underfoot over just 16". The engineer predicted 1/4" deflection over 144". Use Ditra to decouple the tile from the subfloor and I doubt there will be a problem.

 

[Tinker]

I had the same problem (sot of) when i redid our own bathroom.  A few years before, I had worked a deal with one of my contractors where he would redo our bathroom and I built a fireplace and chimney for a vacation home for himself in Vermont.  I stayed in his trailer he kept on his property in VT.  He lived close by my place so he just comuted between his home and mine.  I had given him specs to abide by.  #1 was to put adequate insulating in wall behind the tub (against outside all).  I was sure there was no, or very little insulation as all the tiles were falling of the wall.  #2 The floor tiles were coming lose from the plywood.  I wanted the plywood built up with 3/4" over top of the 1/2" CD that had originally been installed.  I did a fine job on his chimney and expecrted same for our bathroom.

I suspected some problems when i discovered the flor had not been built up with 3/4" ply.  Soon, the tiles started falling off the wall around the tub.  I decided to investigate and discovered NO INSULATION behind the outside walls.  The floor ply had been replaced.  He went from 1/2" CD to 1/2" AC.  I ripped everything out, redesigned and moved the tub to inside wall.  Reframed utside wall from 2x4 to 2x6 and added full thickness insulation.  I did not add 3/4" ply, but put down galvanized wire lathe and 1-1/2" of concrete over the entire floor.  The inside of the wall and around behind where the old tub had been I added wire lathe and 1" of cement plaster with fiber added to both the floor concrete and the cement plaster.  The tiles did not crack or pull away.  Evntuall, the boss decided she wanted a different look so, since I was way too busy with outside work to fool around with it, we hired a contractor friend who was working with our son.  They tore out the concrete floor and added 3/4" ply where I had concrete.  I would not let them remove the wire lathe but sandwiched between the 1/2" and the new 3/4" ply. We added wire lathe over the 3/4" ply and put electric heat in the floor with troweled in cement plaster surounding the wire. The plastered walls, I would not allow to be removed.  We ended up with narrow sleepers glued (flexible construction adhesive) to the plaster and ash panneling nailed and glued to the sleepers. 

The whole thing has held together for 6 years now and no cracking anywhere. The floor framing was/is 2x8 with all sorts of pipes and wiring involved.  It would have been impossible to sister any of the joists.If any cracking appears, I told the builter we will add firing strps to underside of the joists.  Bridging is there, but it is sheet metal "X" bracing.  I don't know if that will prove to be edequate.  So far, I am satisfied we did all that was necessary. 
Tinker

 

[Cheese]

Use Ditra to decouple the tile from the subfloor and I doubt there will be a problem.

I added a 3/8" underlayment layer plus am installing Ditra Heat, I expect that will take care of my issues. [cool]

 

[TooManytoys]

Hey guys.

You guys don't know me as I never really post over here, a little on the Mafell forum, but I'm more automotive in my travels.

Three years back I participated in a floor stiffening thread over at Garage Journal and put together a table discussing this same thing and alternate ways to do this.

The table:

[attachimg=1]

For glue attachment you have to be wary of the type of adhesive used.  The typical PL type adhesives have too much flexibility and will not transfer the loads as well as structural glues.

The full thread is here:
http://www.garagejournal.com/forum/showthread.php?t=88914

As I believe as discussed in the thread, attaching steel to the bottom 1 1/2" of the joist puts the steel in tension where it provides the best resistance to bending for the assembly, although often that is not the common way done in carpentry.

Anyway, while I'm not a mechanical engineer but those on GJ did review it and I had my former director who has a MIT master in ME also agreed with the table.  I've done this twice (option 5) and measured the deflection of the floor before and after (wife - "Explain again why you want to have another party?) and doing the wood I-Beam works well.

I was thinking it might be easier to fit in 2x4s rather then full size joists.

 

[Brice Burrell]

TooManytoys, thanks for the table.

I have to admit I've used PL adhesive for laminating framing members. [embarassed]

 

[TooManytoys]

I have too.  The term "creep" is used for a number of traits of adhesives.  What I didn't think of before that thread years ago was how an adhesive may move over time when two members are in constant tension, lessening the load transfer.  I had contacted Loctite and a few other manufacturers and it was disconcerting once they understood the conversation was truly of a structural nature, how they backed off about their products.

Now in reality we're not talking about true structural qualities with catastrophic failure, we are trying to lessen deflection.  Sistering a like joist handles that well and the adhesive is not that of an important part of the process.  Building structures in the table are more so, but again it's deflection, not a full on structural component.  In the end it may be more of an technical point then an in-service practical point.