expansion joints in concrete tanks

huangyhg

expansion joints in concrete tanks?
does anybody ever use expansion joints in large concrete tanks? i'm trying to decide whether there is a need to use expansion joints. the tank is rectangular with dimensions 150ft x 300ft.
find a job or post a job opening
it depends on the code that you use.  some codes suggest a limit of (say) 30 m.
i tend to design tanks (for water & wastewater treatment plants) with no joints (other than required construction joints).
for reservoirs (sloping walls on soil), i usually design 150 mm slabs up to 200 m x 200 m with only construction joints but the reinforcement is continuous thru all joints & for a 150 slab, would be n12-150 ew, 50 top cover.  concrete about 32 or 40 grade.
reservoirs do need joints at all locations of restraint - inlet pipes, outlet pipes, scour pipes, overflows, & all junctions such as sloping wall to sloping wall at the corners, & sloping wall to floor.  it all depends on the restrain & any possibility of differential settlement.
do a google of this site (top of this page) for other answers - this subject has been done before.
my basic requirement is that if the reinf is sufficent to distribute the cracks, why do you require joints?  all of my joints are control joints (articulation or shrinkage) but (very rarely - if at all) are joints expansion joints.  check movement due to thermal (first three days of heat rise), long term thermal, shrinkage of the cement paste, & swelling due to water absorbtion) & sum the effects, & you will find that it is very rare to require expansion joints.
i also cannot understand the requirement for partial joints (only a % of reinf thru joint).
what is the configuration of your tank - is it concrete on soil or is it free standing rect tank?
no expansion joints are needed are required for constant temperature liquids.
i hate to disagree with y'all, but a 300 ft. long structure seems a little long without an expansion joint.  i get nervous about anything over 200 ft.  while the temperature of the liquid might be constant, the tank has to be built, and might sit around for a long time before it's filled.  a 40 degree (f) temperature delta results in almost an inch change.
while i'll admit that expansion joints are a nuisance to build, i don't want to remove them from my arsenal just for that.  we've had a lot installed without problems. looking at it in a crass, selfish manner, if you don't put one in and need it, it's your fault, but if it's not built right, it's the contractor's fault.
jedclampett,
how do you install an expansion joint in a tank wall?  do you model the tank as a free standing retaining wall?  how do you transfer the shear between the two walls on either side of the expansion joint without compromising the watertightness?  what does the detail look like? it scares me to put an expansion joint in a tank wall because i am afraid the waterstop won't hold.  i've been trying to come up with these answers and i don't know if my ideas are correct. i don't have anyone in the office that has done tanks with a wall that long.
vincentpa, all good questions.  i'll try to answer them in order:
1) installation isn't that hard.  since the reinforcing stops 2 inches from the face of the joint, the forms don't have to accomodate the lapped bars.  they do, however, have to be split to allow the centerbulb of the waterstop to project through.
2) you model the wall as a cantilever.  for a wall long enough to need an expansion joint, two way action doesn't apply.
3) there is no shear transfer.  since the wall is only supported at the base, there's no shear.  some engineers like to use smooth dowels and plastic pipe to help alignment, but i think it's just a nuisance.
4) as far as a detail, if you look in the old aci 350-89, there's some illustrations.  or "borrow" a detail from one of the big projects in your area.  it's basically a waterstop, some expansion joint material and a good caulk to seal the detail from dirt.
the key is to locate the waterstop in a place where differential movement is not expected.  if both sides of the joint move the same distance, the waterstop won't tear.
jedclampett-
some questions on the use of an expansion joint:
does the joint extend through the base mat, or is it only in the wall?  do you design for any bending, tension and shear at the corner where the walls meet?  
the expansion joint needs to continue through the slab and any intermediate walls.  don't forget that any conduits embedded in the walls or slabs, pipes and even guardrails need expansion fittings where they cross the joint.
for external corners and any other cross walls where there might only be liquid on one side, i use the "moments and reactions for rectangular plates" case for the maximum width over depth (3:2) and design the corner for the resultant shear and moment.  but besides that i design the wall as a cantilever.
jedclampett,
you write that two-way action doesn't apply for a long wall.  but when you look at pca rectangular tanks for the case for a maximum length to height ratio, the my is very large. i have also modeled this in staad and the my was very large.  they seemed to agree.  can you explain the discrepancy in what you are writing and what the computer and book give?  maybe i am misuderstanding you.  i've heard a couple of old timers say this also. i've been looking for an explanation of this for some time and i have come to trust your advice, especially for tank design.
jedclampett-
i guess you mean figure 30 from moodys for a plate fixed on one side and free on the other.  i've always wondered why they provided that table.
for a tank with an expansion joint through the base mat, how do you resolve the force in the mat perpendicular to the joint?  i normally think of a wall having a shear at the base that is carried in tension through the mat to the opposite wall.  is it carried through friction to the ground like a retaining wall?
miecz, if you use figure 30, you're too conservative.  use figure 4 for a/b of 3/2 and you'll have a good approximation.
vincentpa, for the pca rectangular tank for case 3 and b/a of 4.0, the moments are large, but not nearly as large as the moment at the base (or the pure cantilever moment).  these are the moments you should use at the corner.  for a lot of cases, you'll have to add bars at the corner.  i usually end up with minimum flexural reinforcing (.0033)horizontally for the majority of the wall and doubling it at the corners.  the reason for the large moment is that the corner acts as a support for the wall and prevents the cantilver action.  you end up with two way action for some distance.  a computer might give you the exact results, but us old-timers figured out a way to approximate it without the computer.