Building Addition w/ Existing on Post Tensioned Slab on Grade 2

[MJC6125]

I'm working on a relatively small addition/remodel to an existing one story building. The foundation is a post tensioned slab on grade. I've never designed a post tensioned slab on grade foundation, so I was wondering if anyone has some references that they know of to get familiar with this foundation type? I do see some books from the Post-Tensioning Institute, but I think they all have to be purchased, so I wanted to ask the question before purchasing anything.

More specific to this project, how do you usually go about doing an addition on a building with this foundation type? I'm pretty sure the site has expansive soils. I assume you can't really do a new post-tensioned slab on grade adjacent to the existing one, so if we do have expansive soils are you looking at doing some sort of structural slab on drilled or helical piers and grade beams?

If I'm opening up what use to be an exterior wall by adding a beam with posts at each end, how exactly do you check the post point loads on the post tensioned slab on grade? There is a turndown slab edge along the exterior walls. Is that slab edge just acting as a trench footing with the bearing area matching the width of the turndown or is there more to it than that?

 

[KootK]

1) I've never actually done PT SOG myself. I almost did one last year and did a fair bit of research in the process which I'll share with you. So, if you get a response from anybody else, be sure to weight that more heavily that what I'll offer up. My own thread on the subject can be found here: Link. The warranty stuff was particularly enlightening for me. Heuvo and JoelTXCive are both old hands at this stuff; hopefully one or both of them surfaces here.

2) Here's the sum total of the PT SOG freebies that I have in my library:

Link
Link
Link
Link
Link
Link
Link
Link

3) I don't envy you this task. My impression of PT SOG design as it's one of those "do some smart stuff and hope for the best" kind of design situations. Obviously, that's not too conducive to a clean addition assignment. I think that it does point to your path forward though: the exercising of copious amounts of judgment.

I'm pretty sure the site has expansive soils.

4) Yup.

More specific to this project, how do you usually go about doing an addition on a building with this foundation type?

5) Great question. If your addition were large-ish, I feel that a reasonable solution may in fact be a new PT slab shear connected to the existing one. My understanding of PT SOG is that it's one of the few foundation systems that does not really attempt to prevent settlement. Rather, it uses the stiffness of the mat to minimize differential settlement. As though your superstructure were mounted on a barge floating in poorly behaved waters.

I'd do the shear connection at the interface because I feel that outright isolation between the buildings is impractical and allowing differential settlement at the interface could cause serviceability issues. One thing that you'd kind of have going for you with this approach is that, as I understand it, foundation movement is mostly about moisture changes that happen at the perimeter (clip below). Where the new and old foundations abut, this would presumably no longer be the perimeter so hopefully you wouldn't see much swelling there.

If your addition is smallish, then it will probably be impractical to attempt the PT SOG. In that case I probably would go with piles but keep in mind that, even though the new building will be staying put, the existing building may be moving around. Maybe you can devise some appropriate details at the interface for this and make a go of it. With the piled solution, at least you've only got one side and you've mostly eliminated the initial settlement that one would expect with strip footings etc.

I recommend posting a quick sketch here of the rough layout of the addition next to the existing SOG. Folks may come out of the woodwork with some better answers once they have more of a sense for your situation.

There is a turndown slab edge along the exterior walls. Is that slab edge just acting as a trench footing with the bearing area matching the width of the turndown or is there more to it than that?

6) I'm not positive but I suspect that turn down is really more of an edge distribution member than a trench footing per se in many applications.

If I'm opening up what use to be an exterior wall by adding a beam with posts at each end, how exactly do you check the post point loads on the post tensioned slab on grade?

7) With some difficulty. Hopefully your post loads are small enough that you could use the edge beam things to distribute the loads out to resemble the more uniform condition for which the slab was likely designed. Again, more detail will get you better answers. A plan layout, some loads, and the proportions of your edge beam would go a fair ways. You might be able to scab a new beam on to the existing downturn and force the distribution that you want to an extent.

 

[bhiggins]

In central Texas, where horrible expansive soils are fairly prevalent, post tensioned slabs on grade are designed using the PTI method. These are stiffened (ribbed) slabs and the depth and spacing of the ribs is determined from the PTI equations to resist differential slab movement and to ensure tensile stresses in the slab do not exceed a threshold. At the minimum, ribs are typically 24" deep at 15 ft spacing each direction. Depths of ribs will increase (sometimes as much as 36"-48") and spacing of ribs will decrease depending on how expansive the soils are and your loadings. Bearing capacities are checked by assuming all loads transfer from the slab to the ribs then to the soils.

Basically every project requires a geotech report. The geotech provides the 4 critical PTI design values, em (centerlift and edge lift), ym (centerlift and edge lift), and allowable bearing capacity. The slab is designed using a uniform slab load (PSF) and a uniform perimeter line load (PLF). All of these values are plugged into the PTI equations and the beam depths, spacing, and number of beam and slab tendons are adjusted until you get acceptable values for differential deflections and stresses. I would recommend to look at this spreadsheet if you want to get familiar with the equations and workflow of designing a post tensioned slab.

We typically assume the slab can handle light point or line loads without checking anything. We try to align ribs with moderate point loads and will create a footing and use rebar for larger loads. There is definitely a lot of judgement involved because the PTI method assumes simplified uniform and perimeter loads and there is not much guidance for other scenarios.

An addition will typically designed as a conventional (rebar) stiffened slab or post tensioned stiffened slab. If the existing slab is performing well you can usually match the beam depth and rib spacing to expect similar performance. The new slab would have a perimeter grade beam which would be doweled into the existing perimeter beam. I would avoid using different foundation types due to differential movement which is likely to occur. It is important to note that anytime you drill into a post tension slab or do any trenching, the tendons must be located by scanning the slab and the sides of grade beams. It is an absolute pain to repair severed tendons which is a can of worms you absolutely want to avoid.

Some of these slabs are designed supported on drilled piers and the beams slab are suspended over voidboxes so you should also check for voidboxes under the beams. This would indicate that the soils are especially bad.

 

[KootK]

Great contribution bhiggins. Some questions for you:

1) In your area, where the ribs are used, is it then the case that the ribs do function as trench footings rather than distribution members as I hypothesized?

2) We seem to agree that a PT SOG would be ideal for the addition. However, if the addition is very small, I worry that would neuter the advantages of PT (losses too large relative to cable length). If that's the case (small addition), do you think there would be merit in continuing with a new ribbed SOG but just not post-tensioning that?

 

[MJC6125]

Attached is a plan view showing the size of the addition. It is sort of centrally located in the existing footprint of the building as you can see.

Thanks a lot for both answers so far. Super helpful.

I haven't sized the new point loads that used to be a distributed load quite yet, but that is something I will look into and use judgement.

I too am curious if a ribbed mild reinforced slab-on-grade foundation is the best approach given the footprint of my addition?

 

[bhiggins]

The first function of the ribs is to increase the moment of inertia of the entire foundation which is governed by the PTI design values (em, ym) and loadings.

The PTI bearing capacity check is all of the dead and live loads of the structure plus weight of the foundation divided by the rib bearing area. This assumes all of the loads are distributed evenly throughout the entire foundation. This rarely ever controls and is honestly never really checked.

It is common to check the ribs like a trench footing for bearing capacity, but most clays have a 3,000 PSF - 4,000 PSF bearing capacity so this rarely controls for most residential structures. For point loads I have checked the ribs as a distribution member and put rebar in the beam if I want to avoid using a footing. Heavy point loads (at a steel column for example) will usually align with a rib and dig out a square footing to meet bearing capacity as a belt and suspenders solution. Designing the ribs like this is beyond the PTI method so this is where engineering judgement kicks in.

These slabs can definitely be designed as a stiffened slab with rebar instead of post-tensioning. You can use the WRI method which is perfectly fine but is a little too empirical for my tastes. I personally use the PTI method for stiffened rebar slabs which is still valid for determining moments and deflections if you follow certain assumptions. A slab less than 15 ft in dimension should definitely be reinforced with rebar. It is also common to have post tensioned stiffened slabs where a portion is reinforced with rebar because some of the tendons are too short to work. Sometimes you have to mix and match depending on the conditions that arise.

 

[KootK]

Thanks for the plan sketch.

Lengthwise, it sounds as though you're on the cusp of what would make sense for PT. Unfortunately, you don't have a whole lot of access to do the stressing without using the fancier details required to get that done. Given that, and the relatively small scale of the thing, I vote for a mildly reinforced, ribbed slab even if the cost of that decision is a deeper system than it would have otherwise been. And I'd definitely be looking to shear dowel into the existing foundation.

 

[hokie66]

Not that I know much about the PTI unbonded slab on ground method, but I vote with KootK on this. The addition is confined and relatively small, ruling out stressing in my opinion.

 

[JoelTXCive]

You guys are doing great here! I have a lot of experience building residential post-tensioned slabs from my pre-engineering days. I do not have experience designing them. All the design work I do is conventionally reinforced.

I have designed a lot of slabs with the WRI method, and am in agreement with bhiggens that it is a very empirical method. The more you use it, you start to realize that you always get the same answers out of the design equations.

Another comment to add: One of the goals and underlying assumptions of post tensioned (or WRI) slabs is that you have a subgrade with consistent material properties; and consistent moisture underneath it. You don't want one side of the slab dry and the other soaked. The perimeter beam depth is always a minimum of 24" so as to reduce the impact of rainwater from getting under the slab and creating and inconsistent moisture content.

Here is another document to add to your collection: "Peer reviewed Commentary on the PTI method." Houston has a pretty active professional group called the foundation performance association (

http://www.foundationperformance.org).

It is made up of PE's and contractors. They formed a committee and put together this document. It's pretty good.

The big question on your proposed add-on is: Do you dowel-in the new slab? Or, let it float free? I'm inclined to say that you want to dowel in, but you have to worry about hitting an existing cable in the perimeter beam. To help aid this decision, I would attempt to verify if you have a single cable system (top only), or do you have a second draped cable. You can go to the corner of the slab and see how many grout pockets are there to figure out. I'll try and find a photo.

 

[JoelTXCive]

@MJC6125 Has the flooring been decided yet?

If not, then I would suggest NOT running tile over the transition between old and new slab. You can have tile in the new add-on, but running it over the transition between old and new slab is just looking for trouble. I would advocate wood (with a little membrane under the transition), or carpet. Nobody ever wants carpet though; so that leaves wood.

 

[JoelTXCive]

Here's a photo pre-pour....