FEA analysis on Roto Moulded PE underground Rainwater Tanks 1

[Grill]

Hi All,

Im just about to embark on designing some underground cylindrical rainwater tanks roto moulded using Polyethylene. The tanks will be backfilled with pea gravel. I have tried to do as much research as i can on loading of undeground tanks as this is a brand new area for me.

I have found an earth pressure equation which takes into consideration the angle that the backfill is at in relation to the side wall of the tank. Was planning to split the faces of the tank up in equal heights and work out the pressure for each height increment. I would then just apply thsi normally?

I dont know if this is the right approach. Could you even point me to some examples of where this sort of analysis has been done?

Tank wall thick nom 10 mm. Will use shell elements, large displacments and Cosmos Advanced Prof.

Thanks in advance

Grill

 

[GBor]

Your approach sounds reasonable, but you would have to err on the conservative side and apply the pressure at the lowest point in a given segmented surface. In other words, if the face that you created runs from 1 foot under the surface to 2 feet under the surface and the calculated pressure is higher at 2 feet under the surface, you would have to apply that to the entire section from 1 foot to 2 feet, but that wouldn't be that big of deal.

Could you use a hydrostatic pressure loading and just apply a really high fluid density that somewhat matches the curve that would be generated by your other equation? Can you publish the equation that you found?

 

[Grill]

Yeah i take on board what you are saying about the difference in height but im planning to do it in 100 mm increments as the tank will only be 1200 mm in diameter. The difference in the values should be minimal.

I dont know how to scan and attach but it is

Ph=Koyh

Ko an K1 i cant type out teh equation but they take into consideration the angle of the material against the side of the tank.

Grill

 

[GBor]

Looks a lot like rho * g * h to me...I would think you could graph the pressure equation for your situation and calculate a curve fit for a given density so that your pressure could be gradually increasing. I agree that it probably won't make much of a difference, but if stress becomes and issue, may be worth considering.

 

[namklof]

If the water table is higher than the bottom of your tank, your approach is ok, with the load being only due to the water. The soil load is much more complicated. With a PE material, your tank is flexible. If installed in compacted pea gravel or crushed stone, the soil is much stiffner than your tank. If your tank lies horizontally like a pipe and the tank has a couple feet of soil on top, the soil can act like an arch. In this case, only a small fraction of the soil load goes into the tank. You also need to consider live loads (do vehciles drive over it). Thus, you should analyze the soil as well as the tank. Getting a good material model for soil can be hard, but for compacted pea gravel, it can be modeled as linear elastic. You should probably work with a geotechnical engineer on this. Another factor to consider is creep in the PE. A PE pipe after installation gradually relaxes shedding more of its load to the soil. However, hydrostatic loads due to the water table can not be shed to somewhere else. Pubished values for yield strength are usually "short term" values. You need to use "long term" values for hydrostatic loads. Hope this helps. Plastics need to have realtively low tensile stresses to avoid fracture failures.

 

[ThomasH]

Grill:

I agree with namklof that the soli pressure is a lot more complicated then the equation you give. The basis for the equation you give is a hole with vertical walls supporting a surrounding soil with a horisontal ground level.

Then the pressure can be calculated as: Pressure = constant * depth * density of soil * g

Since your tank is cylindrical the pressure constant will depend on the inclination of the supporting wall. And since the tank is flexible the pressure will also depend on the deformation.

There are ways to simplify this but I you might need help from somebody with geotechnical experience to do it.

GBor mentioned fluids and there are similaritys but one huge difference. Fluids typically don't have any "inner" strengt, soils have. Ever heard of a pile of water?
"Hydrostatic" can be conservative but also unconservative depending on deformation.

I hope I haven't confused you to much but part of the problem is the flexible tank. If it was rigid it would be simpler.

Good Luck

Thomas

 

[GBor]

Thomas,

I think you got my point, but just to make it clear, I was simply trying to make it easier to apply a smooth load. Even the equation you provided, my point was to multiply the "constant" times the soil density (also a constant, I think) to give a modified density. Then, the pressure really is rho * g * h (depth), which is what most processors use to apply hydrostatic pressure.

 

[ThomasH]

GBor,

I got your point. And you are correct, there are similaritys between soil pressure and hydrostatic pressure. In some respects they are the same for a vertical rigid surface. But for inclined nonrigid surfaces they can differ significantly.

My point, and I think you agree, when you analyze this you have to understand the mechanics / physics of the problem.

Regards

Thomas

 

[GBor]

Absolutely.

 

[Grill]

Thanks for all the insight guys. I have done an example calc and posted it on here. Is this the right way to go?

From further discussion my two loading conditons are as follows.

(1) Concrete backfill with a high water table - I have to analyse the tank when it is being installed due to weight of wet concrete on it. Also once set i will analyse with Hydrostatic pressure

(2) Dry site with Pea gravel backfill - Again this is a soil loading problem.

I got the equation from the Roark book which is is Formulas for Stress and Strain. Im trying to find the correct page.

Hope this makes sense

Sean

 

[Grill]

P.S. Im fully aware of the joys of creep in PE!! It is the bane of my life. We use a reduced modulus approach which is extrapolated from a creep curve depending on the average stress teh product is seeing.

I believe there is some method of simulating creep within Cosmos Works ADV Prof that i have but i havent worked out how to do it yet.

Grill