Reinforcing Existing Wood Floor Joists 4

[bookowski]

I've looked through some of the past discussions on this topic but didn't find any consensus.

Dealing with renovations of existing rowhouses, single span joists bearing wall to bearing wall. Existing joists are usually 2.75" x 8" to 10" deep spanning around 20ft. Typically the ceilings are down so full access to the joists but they don't want to remove the flooring or existing joist, they also want to avoid new wall pockets. I'm left with reinforcing the existing joists with new members that can go close to the ends of the existing members, but do not have bearing at their ends. Two questions:

1. I've specified this in the past with typical details that have been handed down. I started thinking about the calcs today and I'm not sure if I'm thinking straight. Assume that a new LVL of the same depth is provided at the side of each existing member. The existing floor probably won't directly load the new LVL due to existing deformations/unevenness, so the floor load goes through the existing member. For arguments sake - if I was to through bolt at the center (vertically) of the two members there should be no horizontal shear between the two. However, the members will be forced to have deformation compatibility correct? It seems to me that the majority of the bolting (or nailing/glueing) does not do much, except at the end of the new member. As long as the connection between the two members at the ends can transfer the shear out of the LVL reinforcing back into the existing member I should be ok correct?

2. I get a lot of requests to do the above fix but using a cold formed metal stud attached to the existing joist (lighter, easier to maneuver into an exist. bldg etc). Has anyone done this and can recommend for or against it?

 

[msquared48]

bookowski:

I take it that mechanical/plumbing/electrical can be worked around so that additional joists can be inserted?

Also, sistering is usually done here, but with bridging, does not necessarily have to be done at every existing joist, depending on the spacing that you eventually need for the modifications.

As for the end bearing, if you slide the joists into the joist space flat and then rotate it into a vertical position, you can get more bearing area at the end of the joist. If there is not enough, then put blocking in the joist space flush with the outside face of the bearing wall, or other convenient location, and hang one end from the blociing, and bear the other end on the wall. The blocking can be hung off each existing joist. Shear on the existing joists would hacve to be checked. This scenario could be rotateed for the next joist, and so on.

Just a couple of ideas...

As for the metal joists, I have never done it, but you would have to check your local fire requirements. Sounds like it would work though.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[bookowski]

These are usually gut renovations, so all new elec./mech. - the joists and flooring usually do remain though.

This is for joists going into brick party walls, I left that out - so there is no sliding in flat and flipping up. Without excavating new pockets at each joist end the new reinforcement needs to stop short, so there is no bearing.

 

[msquared48]

Then the blocking between the existing joists, with adequate shear and bearing at the ends of the existing joists, would work. (I think)

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[RFreund]

Your reasoning sounds correct. The members would be loaded according to their bending (EI) stiffness and most of the connections will be to x-fer the shear at the end of the members.

EIT

http://www.HowToEngineer.com

 

[woodman88]

Not sure why you are doing this, but if these are supporting a 40 psf ll and 10 to 20 psf dl. Than if the add-ons are connected without any live load on the floor the add-on will work to resist the live load forces. I would prefer to have the joists jacked up for the dead load deflection amount before connecting the add-on.

As for using bolts, as the bolt holes are drilled larger than the bolt size there will be more deflection required before the add-on starts working. I would use Simpson SDS's or pre-drilled lag screws to avoid this.

As for the placement of the SDS's/lag screws, I prefer to spaced them along the joist for 1/2 the total plf on the combined joist. With clusters at the ends. I do this as I think it is a better to make the add-on and joist act more like a composite joist rather than two separate joists connected at three points.

If using a cold formed metal stud for the add-on, you to have compatible deflections between them. So if the metal stud deflects 1/4” under x plf and the wood joist deflects 1/4” under y plf. Then the composite joist supports (x+y) plf.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[dhengr]

Bookowski:
Make some engineering judgement about the max. allowable end shear and bearing on the existing joists. This will set the max. uniform load on an existing joist, without modifying the bearing or end pockets. Then, if you have reserve shear and bearing, you want to increase the section modulus for bending and the moment of inertial for deflection. You have to jack each existing joist up when you apply any sisters so that when you release the jacks the two joined members do act in unison, not just that the added member takes any new loading. The sistering connection should be glued and screwed or nailed, maybe with some shear transfer bolts, or hangers, near the ends of the new members.

What kind of floor sheathing do you have? Would you add a second layer of plywood sheathing on the top to allow you to reinforce every other joist and average the two for your new strength. Try an experiment, using a fairly light aircraft cable with one or two turnbuckles, fix the ends up near the ends of an existing joist and near the top of that joist. Harp the cable down, under two bars under the bottom edge of that joist, at 7' and 13'. Study or test this arrangement for the load and deflection you need.

You should know much more about the existing party walls, since their cap’y. and stability are an important part of this problem. You may not need a great deal of additional bearing strength or area. So, assuming a 1.5" wide sister member, fab a 1.5" long tee section (the width of the sister member) with a 8" high stem to mate with the end of the sister member, with the bottom of this tee extending 2 or 3" into the wall and extending 6 or 8" out onto the bottom edge of the sister member. This tee is screwed or nailed to the end and bottom of the sister member, and goes into a saw cut kerf (or mortar joint) in the brick wall. Note also, that the reason for fire cutting joists is to allow them to fall out of a wall without pulling the wall down too.

 

[BAretired]

The existing joists seem much too shallow to span 20' but they may be adequate for shear. It would be desirable to make the sister joist a little deeper than the existing whether it is wood or steel. It would be prudent to jack the existing joists up to relieve dead load deflection prior to attaching new joists.

BA

 

[XR250]

I would sister each with dimensional lumber on EACH side instead of an LVL on ONE side.
That would keep your loading symmetric and you would not need to add blocking to resolve the twist.
It will also be stiffer.
Also, around here, LVL's are four times as expensive as dimensional lumber - so you would be saving your client money as well.
Just nail the crap out them and don't worry about bolting.
I agree with others about jacking out the dead load deflection.

 

[bookowski]

My main question was about the transfer of the load and the connection design between the two members. I didn't mean to imply any specifics - i.e. bolts vs. nails etc. I was just checking if the logic sounded reasonable - that the most critical connection point is at the end, and at this point I should be checking to transfer all shear back into the original joist. The intermediate connectors only need minimal capacity - maybe 1/2 the plf as woodman suggested? This is what I've done in the past but I've never really thought too much about the calcs - was worried I am missing something.

Several suggested jacking. These are typical floor joists, usually @18" o.c. or so, and everything is stripped off of them except the original subfloor. So there is minimal DL in place. I'm ok with leaving this small DL in place and taking all added DL + LL with both members.

Doesn't sound like anyone has any objection to doing this with cold formed joists? I transform the CF into wood and then ratio the load based on the EI. Without getting into the argument of why this is done here - it's a situation that has to be repeated many times, basically reinforcing almost all existing floor joists. So the contractors prefer to cart around cold formed instead of wood. It feels like it's not quite right - but I can't find any argument against it.

Anyway - sounds like I wasn't missing anything big - thanks for the input.

 

[BAretired]

The idea of using a cold formed metal stud adjacent to each joist seems perfectly workable. The new joist would transfer its main load to the existing joist within a few inches of the wall. Any nailing within the depth 'd' would not increase design shear for the existing joist. The nailing throughout the member would be nominal, just enough to carry the uniform load (roughly one half of the total).

If the existing joist is adequate in shear without an additional joist, there is another method which might be worth considering, namely to apply a 'U' shaped plate to the underside of the existing joist with the idea of increasing its moment of inertia. The steel and timber would act as a composite member, so that fasteners would have to be designed to handle shear flow. Glulam rivets could be driven through the walls of the 'U' into the existing joist. It would be very easy to provide one jack under the steel plate at midspan and jack the joist up a bit.

If there are any obstructions near the wall, the 'U' plate could be cut shorter as it would not be required for the full length of the existing joist.

BA

 

[XR250]

Can't imagine why a contractor would rather use CFS than wood.
Nothing faster then a nail gun and wood.
The only benefit I can see is that you could get a stiffer section for the same depth.
CFS is not cheap compared to wood and would have to be pre-drilled as well. Also, you would need to add blocking due to the unsymmetric loading - not fun with CFS.
I have done this numerous times in old buildings by sistering each face of the joist - works fine and is cheap and fast.

 

[woodman88]

To clarify, the transfer of 1/2 the plf may not be enough unless the x < y in my above statement. To be correct you need to transfer x/(x+y) to the metal stud. Or just do the full plf.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[structSU10]

You said they do not want to pocket into the wall - Is there a way to put a ledger board agaisnt the wall for the new members to bear on?

 

[DBuker]

Bookowski,

If I understand you correctly, you are sistering joists the same way that I usually do in similar situations.
1. check shear of original joist to ensure that it can support the full load at the support.
2. sister new joist alongside existing joist, size connection along length of span to transfer portion of the load that then new joist is expected to carry based on relative stiffness (I usually overdo this connection because nails and screws are relatively cheap).
3. provide a cluster of connectors at the ends of the new member to transfer the reaction of the new member back into the existing member. This is the connection that is usually tricky.

In this situation you may worry about the eccentricity caused by having the sistered joist on only one side but we're talking about a wood framed floor where the top (compression) edge is braced by the diaphragm so I don't see a reason to be concerned.

There are other ways to strengthen joists if you have room. For instance, you could lay a 2x6 flat along the bottom edge of the existing joists to make a T shape. In that case you would do a shear flow analysis to size the connection between the flange and web. It's no different than sizing welds in a built up steel section.

With regard to using metal joists with wood. Technically, I don't see a problem, but I always get a little uncomfortable with mixing dissimilar materials. That's why I think it's a good idea to sister LVL's with old joists. Presumably, the old joists will be fairly dimensionally stable relative to new green lumber so there's an advantage to using dimensionally stable manufactured lumber.

 

[BAretired]

Are the existing joists LVL material?

BA

 

[bookowski]

BA - No, the existing are not LVL, they're floor joists of old rowhouses - ususally around 100yrs old.

DBuker - Thanks, that's how I've been doing it but I just wanted to make sure I wasn't missing something. The end cluster nailing doesn't seem too bad from what I'm calc'ing - why do you say that's the tricky part?

I agree that the metal joists seem strange. That's why I was asking, I couldn't figure out a good reason to say no. For whatever reason contractors around here sometimes prefer using cf to wood.

 

[DBuker]

BAretired,
I should have read your last post closer. Your recommendation to sister something along the bottom of the joists to increase the section properties is what I was getting at with the idea of adding a 2x6 to form a T-section.

I figure LVL material and old, well seasoned sawn lumber are both wood products with low moisture content so they should perform similarly. In reality, LVL, LSL, and PSL will all respond differently to changes in moisture and they'll respond differently than the existing sawn joists. What I've experienced is contractors installing green sawn lumber adjacent to existing old framing and connecting with lag bolts staggered along the length. It's not uncommon for the new lumber to shrink and split. This could be resolved by specifying kiln dried lumber but how often do the contractors read the specs that close?

 

[XR250]

@ Dubuker:

Why would you use an LVL on one side versus dimensional lumber on both sides? The dimensional lumber is cheaper, has no symmetry issues and will be more effective as you have more stiffness than a single LVL.

As far as making a "T" section, the amount of nails req'd really starts to get ridiculous near the ends.

 

[DBuker]

ExcelEngineering,

I have used LVL's in this scenario because LVL's and the existing joists are both dimensionally stable. Green lumber from the lumber yard, even if it is kiln dried, is not necessarily dimensionally stable. When you sister a joist along side an existing joist you typically attach the two together with some sort of staggered pattern. Even if you don't specify a staggered pattern, The contractor is likely to do it that way. So, you run the risk that the new lumber will shrink and split.

I'm not saying you have to do it that way, or that one option is always superior to another. I'm just saying that shrinkage compatibility shouldn't be ignored.