txeng91
Structural
- Sep 5, 2016
- 180
My firm works on a lot of foundation designs in expansive soils for residential and small commercial structures. Up to this point we have always used the BRAB methodology to design our slabs. I am currently reviewing our design process and working on incorporating the WRI method into our design as well. I’ve read through the WRI design document (TF700-R-07) multiple times and have a pretty good grasp on the methodology. I’m in the process of combing through the BRAB manual, and while I’m having a hard time understating all of the calculations they go through, I believe the methodology is more or less the same. Some of the numbers I’ve run show that the BRAB method appears to be more conservative than WRI in many cases, which I haven’t exactly figured out why.
From the research I have done, I have a few concerns on both methods:
1. The design moments are based on a rigid support at the start of the cantilever distance. While I haven’t done a finite element analysis with full composite action of the slab and beams, I did do a first order analysis by modeling a waffle pattern of beams with compressions spring supports at the suppprted interior area (using 100 pci subgrade modulus), and beam moments came out significantly higher than what you obtain with the design equations. Maybe this is factored into the cantilever length or they are counting on the slab supplying additional bearing area, but it they don’t factor in the settlement of the soil when considering moment or deflection.
2. The analysis only considers 1 direction at a time. Obviously you could have a shrink or swell condition around the entire structure, which would increase the moment on the stiffening beams closest to where the cantilever condition begins. The model I ran confirmed this.
3. Perimeter beams are considered to provide equal contribution to the slab stiffness as interior beams. Doesn’t make sense to me, the perimeter beams should have half the tributary load as interior beams. They calculate the moment for the entire slab and apply the same load to each beam. The model I ran showed much higher loads on the interior beams then the perimeter beams. The part that gets me is that they spread the load evenly across all beams then tell you to up the beam reinforcement where you have beam spacings exceed the average beam spacing.
4. The loadings on the stiffening beams are too arbitrary for the application we are using them for. Obviously this isn’t the fault of the design guides, they are made for small, cookie cutter residential houses and they explicitly say something to that affect. Most of our designs our on custom houses and a lot of the time the framing is not engineered so the load paths aren’t always clear to us. Using a generic uniform load may be fine when you have bearing lines every 10’ or so and the loads are well distributed on the slab. But if you have framing members with long spans it’s probably not adequate.
Anyway, I can solve issues 3 and 4 by modifying the WRI analysis. Solving issues 1 and 2 would be rather difficult without running a computer model of each slab, which we certainly don’t have the project budgets to do.
Since there is limited information available on designing foundations in expansive soils, I wanted to reach out and get anyone else’s opinion who may be in the same boat. If you have experience in this, I would like to know what methodologies you are using and if you have come up with any modifications to address the concerns I noted above or anything else. I apologize for the long post and thanks in advance to anyone who actually read all of this
From the research I have done, I have a few concerns on both methods:
1. The design moments are based on a rigid support at the start of the cantilever distance. While I haven’t done a finite element analysis with full composite action of the slab and beams, I did do a first order analysis by modeling a waffle pattern of beams with compressions spring supports at the suppprted interior area (using 100 pci subgrade modulus), and beam moments came out significantly higher than what you obtain with the design equations. Maybe this is factored into the cantilever length or they are counting on the slab supplying additional bearing area, but it they don’t factor in the settlement of the soil when considering moment or deflection.
2. The analysis only considers 1 direction at a time. Obviously you could have a shrink or swell condition around the entire structure, which would increase the moment on the stiffening beams closest to where the cantilever condition begins. The model I ran confirmed this.
3. Perimeter beams are considered to provide equal contribution to the slab stiffness as interior beams. Doesn’t make sense to me, the perimeter beams should have half the tributary load as interior beams. They calculate the moment for the entire slab and apply the same load to each beam. The model I ran showed much higher loads on the interior beams then the perimeter beams. The part that gets me is that they spread the load evenly across all beams then tell you to up the beam reinforcement where you have beam spacings exceed the average beam spacing.
4. The loadings on the stiffening beams are too arbitrary for the application we are using them for. Obviously this isn’t the fault of the design guides, they are made for small, cookie cutter residential houses and they explicitly say something to that affect. Most of our designs our on custom houses and a lot of the time the framing is not engineered so the load paths aren’t always clear to us. Using a generic uniform load may be fine when you have bearing lines every 10’ or so and the loads are well distributed on the slab. But if you have framing members with long spans it’s probably not adequate.
Anyway, I can solve issues 3 and 4 by modifying the WRI analysis. Solving issues 1 and 2 would be rather difficult without running a computer model of each slab, which we certainly don’t have the project budgets to do.
Since there is limited information available on designing foundations in expansive soils, I wanted to reach out and get anyone else’s opinion who may be in the same boat. If you have experience in this, I would like to know what methodologies you are using and if you have come up with any modifications to address the concerns I noted above or anything else. I apologize for the long post and thanks in advance to anyone who actually read all of this
