Discontinuous stiffening of existing floor joist systems 1

[BT71]

Greetings,

I received a call yesterday (and I'm going to have a look later today) from a gentleman who has paid one firm to stiffen his existing joists underneath a single story kitchen by "doubling" them. And then paid another firm to add better subfloor and then ceramic tile over the top.

Tiles are now cracking.

As he described it to me on the phone, the doubling of the joists in an attempt to stiffen them was more or less on the central two-thirds of the span and perhaps not well fastened (his intent was to cover the full span and anchor thoroughly). I believe the joists are 2x10 on 16 inch centers, but my question for the forum is [blue]"Does anyone have a model, analysis, or formula for the stiffening effects of this discontinuous (in section) beam?[/blue] I just downloaded the design suite posted recently but it is more of a pure beam simulator. Suppose this home has 18 foot long joists resting on block foundation over a crawl space and the length of the doubler is only 14 feet and the location is centered -- what does the floor strain profile look like?

The owner of this property is of the opinion he wants me to add in the missing ends of the "doubling" and tie them together with some wood or metal shear plates. I'm not so sure what to do if his functional goal is to stiffen the floor to the extent the strain is below some maximum I need to think about and also that changes in strain are below some value.

I expect the first two steps are to make sure his doubling covers the area in question and then to make sure the two are well fastened together.

Approaches welcome from mathematicians and practitioners alike!

 

[JAE]

Structural response:

You can model the joist in a standard 2-dimensional frame analysis program. The joist would have three sections, one in the middle with an Area/Moment of Inertia value representative of the built-up section, and the two outside sections representing the original 2x joist.

The model would then give you a load/deflection result that could be compared with the actual in-place condition. The joist was presumably built-up prior to the tile being placed on it.

The problem here: The partial length pieces probably didn't do a whole lot of stiffening for two reasons,

First, the partial length only would slightly help the deflection.

Second, the nailing of the add-on pieces are not totally rigid, thus, you get slight slippage between the two pieces and an even more flexible condition than that modelled in my first paragraphs above.

By adding more "pieces" to the ends, you really aren't going to accomplish much. The add-on pieces probably should have been full length, or, if the installation of full length is difficult, at least extend over 90% of the length.

Again, nailing is an issue, - glued and screwed would be better for the partial length fixes. Full length/full bearing add-ons would not need any nailing as they would simply path the load to the supports.

Even with extra 2x joist, the stiffness of the system may still be too flexible to avoid cracking in tile. Many times you can add a flexible neoprene or rubber substrate under the tile to soften the movement - also, lots of flexible joints between tile at an appropriate spacing is helpful in compartmentalizing the tile and minimizing cracking. These flexible grout joints could be filled with a semi-rigid epoxy tinted to match the grout color.

 

[TFL]

if you must stiffen it you can look into adding 2 layers of 3/4" plywood to the bottom of the joist this would form a bunch of box beams. you need to layers to provide a lap at the joints to gt something equivalent to a continuous bottom chord.
it seams like alot of work to be for little reward. the existing cracking will not go away becasue you stiffen the floor.

 

[BT71]

Thanks for the replies, I did go and find/use a more standard FEA system to make some pictures that helped me explain to the owner his hoped-for effect and his likely outcome. And the box beam suggestion helped me illustrate the level of work he'd perform if more stiffness was truly his goal/solution.

After a couple hours in his crawl space, the first description wasn't entirely accurate. Whoever doubled up his joists did a fair job covering the entire length (other than not getting up on the sill boards at either end). Where they failed was in fastening the two together. I counted maybe 12-15 nails on an 11 foot span ! So, I described what needed to happen there to make that lumber have some effect, but in the end I think I talked him out of it because now that his tile is down, that amount of stress might bring a lot of it up.

I also ran a little stiffness experiment for the owner by putting a dial indicator at mid-span on several joists and having he, his wife, and some neighbors walk in and out of the room to stand over the mid-point, just to see the deflection, and it was running about 0.020 inches on an 11 foot span with about 600 pounds load. That was on the longer spans, the shorter ones were proportional but equally stiff.

So, we concluded his doubling was ineffective but that stiffness (or rather adding more stiffness) was not the likely cause or solution.

I did find one anomaly in the original construction. This home had three sections of 8-11 foot joist span, resting on perimeter foundation or on internal masonry cross beam that was well supported. Above one of the internal beams, I did find a run of maybe six joists where neither side coming in really crossed the full length of the 2x6 sill board, and then (the surprising part) they were nailed with three fasteners just 1 inch in from the end. So the overlap was something like 2-3 inches with one row of fasteners.

I advised him to leave the whole room alone and see if additional tiles or grout break, and if/when then go with the flexible joint. The owner mentioned that above the area of no joist overlap a tile had broken and the tile tradesman had mentioned during the replacement that his concrete sheet sub-floor had cracked as well. So maybe the only action to take is to glue and screw some plates over those weaker joints. I proposed some 2x8 pieces about 24 inches in length as filler on one side of the joint, and then to bridge the join with about 48 inches of 2x8, with heavy glue on all internal surfaces and then lag or bolt on a close pattern to draw the plates together until the glue bonds sufficiently.

I doubt any further actions will be taken until and unless additional tiles break, but thanks again for the replies above, they were helpful.

 

[boo1]

Consider renailing the doubler joists. That will decrease joist stress and deflection.

Was there bridge blocking at mid span?

Tile cracking from reflection cracks are common.

 

[BT71]

Hello Boo1,

Well . . . he doesn't want to nail additionally as his tile man has told him it will void all tile warranty due to impact and stress. From my inspection, many doubled joists are slightly cupped or angled with a noticeable gap where he may need to lag or bolt anyway, to draw together and bind . . . he doesn't want to do that for now either.

Curious about the blocking. I forgot to mention that the entire house has none! While I can understand a little on an 8 foot span, I don't understand so much as we get to 11-12 feet. I did mention the possibility of adding blocking thorughout but again not willing due to remarks from the tile installer.

 

[boo1]

What observed seems less than adequate to effect a real connection. It also appears to be a particularly ineffective approach when adding a beam to share load as part of a repair. I usually tend to specify at least two rows of 16d nails at 16" oc for attaching 1 1/2" members together. This at least helps to equally share the load between members. Consider attaching with wood or lag screws and adding mid span blocking.

Are the ends solid blocked?

 

[BT71]

Hello boo1,

Agreed. As observed, it's weak.

The terminal ends are blocked, but none of the intermediate joints or overlaps is blocked. I've made the recommendation, but need to wait until the owner is comfortable or ready to risk the further stress/disruption of the tile in favor of strengthening/fixing the structure.

 

[JAE]

I would think that adding solid blocking between the joists would do a lot to minimize differential deflections between the joists - Granted that the tile man says NO to nailing - but what about screwing the blocking in? There is no impact with screws.

 

[boo1]

JAE that what I tried to recommend in my Aug 17 post, last sentance

 

[JAE]

Sorry boo1 - you're right - I need to slow down and READ more carefully.

 

[BT71]

Hi boo1, JAE,

Thanks a lot, I know what to do now, just need to wait for permission. I really appreciate your patience in reading and responding to the trickle of information.

 

[Tmoose]

I'm curious to know more about the location and direction of the cracks.

18 foot span is a little long for 12X10s 16" OC if the framing lumber is HEM-FIR common around here. To my mind it might be hard to crack tile or grout floors atop a floor joist system even with moderately excessive vertical deflection. Of course if the tile was part of a ceiling system attached to the >>underside<< of the floor system it would be easy to crack, just as fear of cracking the drywall ceiling finish is part of the codes that chops 2 or 3 feet off the allowable span to limit floor deflections to span/240 or 360.

Adding a series of mid span strong backs in the crawl space might be an easier way to put questions of stiffness/deflection to rest, but with the added danger of un-bending the floor, which certainly would put the tiles in tension.

Some tile specs are specific that the floor sheathing must be over 1 inch thick. The modern standard of 3/4 inch plywood sheathing plus 1/2 inch plywood gets the top of the tile even with typical hardwood floor, but also adds considerable local "surface" stiffness to reduce 2 way bending that would subject the tiles to tension loading or excessive local bending.

 

[BT71]

Hello Tmoose,

Longest span is maybe 11 feet 6 inches. Shortest was just under 8 feet.

Prior to the tile work, the owner had subbed some work to stiffen and cover the floor, doubling the joists (which we've seen was ineffective) and then putting down 1/2 inch cement board on top of his 3/4 subfloor.

Three areas of cracking, cracks running perpendicular to the length of the joists. Two series in grout and one in tile.

The tile cracking was above the area of no/minimal overlap in the original joist system as described above. I expect we'll see that one again soon.

The grout cracking was mid-span on one, and on the other one I think the fit to the wall was way too tight based on the repair actions of the tile man (second hand story)and I doubt that one will be back.

I had mentioned the strong backs during the deflection measuring visit, as this same owner had added 4 of them near a fireplace he was going to re-work with stone, but again, he didn't want to risk bending the floor by adding more now that the tile was down.

As of today, no actions, so I think we're in wait and see mode, unless and until another tile breaks.

 

[boo1]

You added more info now, with stone fireplace can be very heavy. The crack direction indicated eccessive deflection has occured. Have you checked the floor for level? Was the stone material or other heave material loaded over or on the tile?

When I place tile over wood subfloors we use increase the number of joist, add cement board, use elastomeric thin set and add latex modifiers to the grout. This reduces cracking......

Look at Simpson's new lag bolt screws for fasteners.

 

[BT71]

Hello boo1,

Floor is level. No material was loaded over the tile. In fact, the stone for the fireplace isn't inside this home yet, it's out in his garage. I would also say the fireplace is very well supported and has 8 masonry columns around it's perimeter at this point and should not interact with the tile floor.

This owner did try to do the right things by adding joists and the cement board sub-floor, he's just wound up with a problem or two he didn't know about. I'm still in agreement with you that blocking at mid-span and bridging the one area of weak overlap in the joists will provide the most help once the owner decides to take the plunge.

 

[boo1]

With the following inputs:
Width: 1.50 #2 Douglas Fir - Floor Loads
Depth: 9.25
Area 13.88 in2 Cd 1.00
Sx 21.39 in3 Cr 1.15
Ix 98.93 in4 Cf 1.10
Fb 900 psi Ch 1.50
Fv 95 psi F'b 1139 psi
E 1,600 ksi F'v 143 psi

Live (psf) Dead (psf)
Total Defl. L / 360 40 15
Live Defl. L / 480

Single Joists @ 16" o.c. allowable span is 14.73'
Double @ 16" o.c. allowable span is 18.56'