SixkHz
Structural
- Oct 16, 2008
- 20
I was wondering if I could get some input on this model I’ve been having trouble with.
Some background on the project (refer to attached sketch):
The structure I’m analyzing is a system of stainless steel beams and hangers for “cassettes” (basically boxes made up of HSS sections that holds water filtration membranes), that are suspended in rectangular concrete water treatment tanks. The beams consist of HSS8x4x0.25 sections, and the hangers consist of bent plate forming a 4”x3”x0.25” channel.
The beams are anchored to the concrete tank walls at either end, and have brackets (c/w pins) along their length upon which the hangers are supported. The hangers have a 1” thick top plate with an oversized slotted hole that is supported eccentrically on the beam bracket. A double-split shaft collar is added to the pin to prevent the cassette from popping off. The hangers are connected to the cassette below with a 0.75” thick plate and a single 0.75” diameter bolt, with double nut and washers. There are three different hanger lengths depending on the tank configuration: 29”, 41”, and 60”.
This is an existing design that the client has been using for years, but it was designed according to UBC 97, and they now need it to comply with IBC 2006. It has to be designed for both gravity and seismic loads. The worst case is when the tank is empty, as the hanging weight of the cassette is around 9000 lbs, compared to only 1000 lbs when the tank is full (due to buoyancy). They’ve asked for the design to be checked for a range of Sds values (0.25g to 2.50g), as this system is used in water treatment plants all over the US. The idea is to keep the same design, and just increase the member sizes depending on the Sds value.
Regarding my model, I used dimensionless rigid members down and out from the centerline of the beam to try to mimic the bracket plate with the pin. I also used a rigid member for the top plate of the hanger. At the connection of the top plate to the bracket, I’ve put in a full moment release, to account for the oversized hole and the fact that the ¾” diameter pin really can’t take any significant moment. In order for my model to be stable, I have to provide a fixed connection from the hanger bottom plate to the cassette below. The problem with this is my moments are quite large (especially for the 60” long hanger), and the connection consists of only a single bolt. I did an FEA for the bottom plate, and it’s overstressed even at low Sds values. The bolt is not working for a lot of cases either, even if I go high-strength.
Ideally, I want to just consider the bottom plate connection as pinned, and add in cross-bracing to take the lateral seismic forces. Unfortunately, the client does not want to make any changes to the design, because any additional steel results in significant costs (it’s all stainless steel, and there can be hundreds of these systems in a given treatment plant). They don’t want to change the design because they’ve been doing it this way for years and the likelihood of a seismic event occurring when the tanks are empty is quite low.
I’m struggling with how this system was designed for seismic in the first place, there doesn’t seem to be much lateral stiffness at all. Am I missing something? Any input is appreciated.
Thanks,
Matt
Sidenote: I used a static seismic analysis. I determined my seismic loads from ASCE 7-05 Chapter 13, Eqn 13.3-1. I used ap = 1.0 and Rp = 3.0, z/h = 0.84, Ip = 1.0, so my load ranged from 0.09W (Sds = 0.25g) to 0.89W (Sds = 2.50g)
Some background on the project (refer to attached sketch):
The structure I’m analyzing is a system of stainless steel beams and hangers for “cassettes” (basically boxes made up of HSS sections that holds water filtration membranes), that are suspended in rectangular concrete water treatment tanks. The beams consist of HSS8x4x0.25 sections, and the hangers consist of bent plate forming a 4”x3”x0.25” channel.
The beams are anchored to the concrete tank walls at either end, and have brackets (c/w pins) along their length upon which the hangers are supported. The hangers have a 1” thick top plate with an oversized slotted hole that is supported eccentrically on the beam bracket. A double-split shaft collar is added to the pin to prevent the cassette from popping off. The hangers are connected to the cassette below with a 0.75” thick plate and a single 0.75” diameter bolt, with double nut and washers. There are three different hanger lengths depending on the tank configuration: 29”, 41”, and 60”.
This is an existing design that the client has been using for years, but it was designed according to UBC 97, and they now need it to comply with IBC 2006. It has to be designed for both gravity and seismic loads. The worst case is when the tank is empty, as the hanging weight of the cassette is around 9000 lbs, compared to only 1000 lbs when the tank is full (due to buoyancy). They’ve asked for the design to be checked for a range of Sds values (0.25g to 2.50g), as this system is used in water treatment plants all over the US. The idea is to keep the same design, and just increase the member sizes depending on the Sds value.
Regarding my model, I used dimensionless rigid members down and out from the centerline of the beam to try to mimic the bracket plate with the pin. I also used a rigid member for the top plate of the hanger. At the connection of the top plate to the bracket, I’ve put in a full moment release, to account for the oversized hole and the fact that the ¾” diameter pin really can’t take any significant moment. In order for my model to be stable, I have to provide a fixed connection from the hanger bottom plate to the cassette below. The problem with this is my moments are quite large (especially for the 60” long hanger), and the connection consists of only a single bolt. I did an FEA for the bottom plate, and it’s overstressed even at low Sds values. The bolt is not working for a lot of cases either, even if I go high-strength.
Ideally, I want to just consider the bottom plate connection as pinned, and add in cross-bracing to take the lateral seismic forces. Unfortunately, the client does not want to make any changes to the design, because any additional steel results in significant costs (it’s all stainless steel, and there can be hundreds of these systems in a given treatment plant). They don’t want to change the design because they’ve been doing it this way for years and the likelihood of a seismic event occurring when the tanks are empty is quite low.
I’m struggling with how this system was designed for seismic in the first place, there doesn’t seem to be much lateral stiffness at all. Am I missing something? Any input is appreciated.
Thanks,
Matt
Sidenote: I used a static seismic analysis. I determined my seismic loads from ASCE 7-05 Chapter 13, Eqn 13.3-1. I used ap = 1.0 and Rp = 3.0, z/h = 0.84, Ip = 1.0, so my load ranged from 0.09W (Sds = 0.25g) to 0.89W (Sds = 2.50g)
