Tuesday, August 15, 2006

How to Retrofit a Poorly Constructed House

Banda Aceh, Aceh, Indonesia

So the "Nameless NGO" (or should I refer to them as "the NGO we do not speak of"?) wants to retrofit the houses rather than demolishing them, even if it costs more to do so, which I believe will.

So how do you retrofit a poorly constructed confined masonry house? Let's start with listing what's wrong.

Connections

The connections in the reinforced columns to ring beams and the ring beams to each other at most, the rebar only hooks, and doesn't overlap. I have even seen where the rebar didn't even reach the column (see photo below), but I hope this is more uncommon than I think. Solution? I see two.



First, sandwich the beam/beam or beam/column/beam connections with steel plates on both sides and bolt through the concrete. Problem? It's going to cost a lot, steel and welders are not cheap, even out here, and this is also assuming the concrete is okay (many places you can break it away with your hands). Second, you'll have to remove a lot of masonry in order to get the plates connected to the columns and beams. And third, you got esthetics, it'd look horrible, like some Borg house and the home owners won't want to live in it.

Second solution, remove the concrete exposing the rebar at the connection and remove it as far back as needed in order to get a lap splice. We're using 50cm splices on 10mm diameter bars. So I envision removing the concrete at least 50cm from the joint along the beams and column, then place a 90degree bent bar so it'll lap with the longitudinal reinforcing bars in each perpendicular direction. You'll have to do this for each of the longitudinal bars. So you'll have four coming in from each of three directions. In a way it's "bridging the gap". I see this as the cheapest and esthetically the best solution, assuming we can fit all the bars in the given space, which it may.

By the way, if any of these ideas sound crazy or wrong, please let me know. That is why I'm posting this.

Masonry

You have poorly built masonry infill. A number of bricks pop out with just the force of your hand. Also, many of the bricks are weak due to not being fully fired and the overall workmanship of the masonry is poor. Solutions? Again, I see two.

First, wrap the masonry wall with webbing or mesh of either carbon fiber or welded wire mesh embedded in plaster. At the ends of the walls, you'll have to wrap it around the columns. Questions? How do you wrap the webbing around the column when you have another masonry infill on the opposite or adjacent side of the column without removing that masonry? How do you wrap around the bottom, or the plinth beam? Do you need to? Carbon fiber is expensive, even in the States, so let's knock that one off the solutions. And lastly, what would prevent the mesh from "delaminating" from the masonry?

Second solution; build a stud wall on the internal face of the wall out of timber or light gage steel and fasten the stud wall to the masonry wall via bolts with plate washers spaced out on a grid pattern. Problems? You loose a lot of floor space on an already small house. It also requires a timber covering or drywall to cover up the stud framing. You want to stay away from timber if you can, because most likely it'd be coming from an illegal forest.


There are other issues with the houses, gable wall, terrace beams and foundation but we either have answers to those or require further research.

So I just got back from a meeting with the "nameless NGO" and they were very understanding and willing to do what is needed to get these houses seismically safer. They agreed the workmanship was very poor. What Elizabeth and I need to do is narrow down the options and come up with some details and procedures showing how to implement all of this. They, the "nameless NGO", will in turn give the same options to the contractor and tell them, retrofit it this way or tear it down and rebuild it, either way, you're not going to get paid until something is done.

1 comment:

Taij said...

For the connections, I think chipping out the rebar and adding dowels is your best option. In addition to all of the problems you list with the steel plates the plates would have to be specially fabricated and welded together in the field to develop the loads around the corners. From what I've seen the quality of the welding there is suspect at best. Plus the contractor would be better able to add the dowels than the plates. I remember that they agreed to do this for the columns where the rebar was too short, so it would be an easy carry-over for them.

For the walls, both options you propose would help mitigate the out-of-plane problem but you'll still have an in-plane shear problem, which is bad news since these walls are shear walls. Carbon fiber would be great, especially since this assembly has been tested and proven effective in URM applications, but it would be very costly and not sustainable. If the homeowner wants to add an opening later it's toast.

You might consider another option that is similar to what I've done to retrofit URM walls and in a way is a combination of the two options you propose. You can add concrete instead of plaster to either side of the wall (though the inside would be prefered since you don't have to worry about water penetration) with wire mesh inside. The aggregate would have to be smaller than what is used for the beams and columns but they could perhaps shift that out like what is done for the sand. The concrete should be placed tight to the columns and beams as well as the masonry. To create a good bond between the concrete and brick you can either roughen the brick surface (which given the quality of the bricks should be easy to do) or use the through bolts with plate washers. On URM retrofits I normally use 4" of concrete, which would translate to 10 cm. If you can convince them to use that much that would be great, but my guess is that would be too thick for the Nameless NGO to accept. Something on the order of 5-6 cm may be sufficent. You might even be able to use 5 cm of concrete on both sides of the wall.

To be honest, though, I think the "Nameless NGO" would be better off taking down the walls and rebuilding them.