API 650 Uplift Force Calculation

[nvivekan]

Hi ,


I am in need of calculating the Uplift force calculation for a API 650 tank (with Base plate fixed with Shell ).
When I refer the Table 5.21 for the uplift force calculation , I noticed that the calculations are based on the Ringwall type foundation.
But the foundation type for our Tank is Slab base / Block Foundation.

Is there any change in calculation for Uplift force?

Because as per Clause 5.12.13 , The foundation should provide adequate counterbalancing weight to resist the design uplift loads.

Please guide me , if anything wrong.

Regards,
Vivek.

[sigilli]

The uplift force in table 5-21 refers the force at the interface between the tank proper and its foundation and is used to design its anchorage (if required). The type of foundation is mostly irrelevant to this calculation (aside from the soil type which affects the seismic forces).

Remember that the uplift is load on the foundation. Your foundation design should include all other relevant loads (weights, liquid pressure, snow, overturning moments from wind and seismic).

[mariog]

I guess you refer to Addendum 3 of last API 650.
Indeed, now the uplift is specified to be calculated with "Mrw = seismic moment in (N-m) (see Appendix E)".
Before it was "Ms" seismic moment (in my interpretation, in the previous editions it wasn't the intention to impose there Ms- seismic on slab, rather it was a result of uncorrelated work of two API 650 committees).

The impulsive and convective action of liquid on wall and bottom were considered as supplementary pressures and the math integration leads to Mrw (seismic acting on wall) and Ms (acting on wall and bottom).
This is not so evident in API650. This approach was considered by Housner in 60’s. He initiated studies regarding seismic action on tank's shell and bottom. Still exists in some contemporaneous works- see the figure in

[link Point to another website Only the registered members can access]
So the explanation is the bottom transfers a part of seismic load to slab, when there is a slab.
A slab will receive Mrw by shell- annular plates- anchors and the rest by bottom (Ms- Mrw)
When there is no slab, (Ms- Mrw) is transferred to soil which is inside the ring wall, not to the ring wall.

You can conclude that anchors are always (slab or ring wall) loaded by Mrw.
But when you have a slab, all Ms is transmitted to slab: Mrw through anchors and (Ms- Mrw) by modifying the hydrostatic pressure on bottom.

Best regards.

[sigilli]

mariog,
I guess I am not that involved with the ins and outs of API 650. I just meant to say that although the uplift force does account for all the loads (including the uplift caused by overturning), that force is there to design the anchors, not the slab. Yes, the anchors will transfer that uplift to the slab and the slab needs to be designed for it, but the slab also needs to be designed for other cases, such as bearing or unbalanced loads caused by overturning moments.

[mariog]

nvivekan,

For me was not so clear what your question refers to.

You said "When I refer the Table 5.21 for the uplift force calculation , I noticed that the calculations are based on the Ringwall type foundation".

No, not true; as sigilli said "The type of foundation is mostly irrelevant to this [uplift] calculation".

But it is true that in the last Addendum 3, you may be confused by Mrw that seems to refer to "ringwall" type. In fact, that formula is valid for both slab and ringwall, as I tried to explain previously.

You may proceed in two ways.

1. There are a lot of things in API 650 Table 5-21 that you may clarified yourself, because they might be tricky without deeply understanding of the kernel.
Here I do not second sigilli's opinion"I guess I am not that involved with the ins and outs of API 650."
I guess you must understand deeply what is if you want to do some good quality engineering.
Otherwise just do not try to do more than you can find out in API!

I'll give you some ideas:


- P*D^2 × 785 is in fact P*D^2 × (PI)/4*1000 [N] with [P]=kPa and gives an math term which participate to the uplift by route Pressure underneath Roof-->Shell-->Anchor chairs-->Anchor-->Slab.

In the same time P*D^2 × 785 is acting to slab by route Pressure above liquid level-->liquid-->bottom--->Slab, and this fact is not so evident under API rules!


Overall, these terms are "compensated" on slab, locally they give a shear force in slab.

Note that in operational cases P is replaced by 0.4P (or more) as operational pressure.

- Or you may note that 8000*10*(th[mm]/1000)*D^2 ×(PI)/4 is the roof weight [N], calculated with 8000 kg/m3 density and g=10 m/s2
and this term is presented as following
0.08*th[mm]*D^2 ×(PI)/4*1000=
= 0.08*(th[mm]*D^2 ×785
This term diminishes uplift and is included in net uplift; there is no need to count it twice!

Case by case, the uplift may be affected by the roof structure that loads the upper shell. Note this is not included in API uplift calculation and must be considered separatelly.

- Hydrostatic product term does not appear in 5-21 as it is not affecting uplift. Of course Hydrostatic is a slab load by pressure on bottom

- Or you may note that Mrw is linearly distributed on the shell-annular plates as 4Mrw/D/(PI*D)*cos(theta) where theta is a polar-angular variable.
On theta=0 the "peak" as a force calculated for a unit of length is 4Mrw/D, on opposite side we say it is -4Mrw/D.

- note that Ms-Mrw does not appear into uplift but this seismic moment is exerted on slab as a pressure distribution that affect the hydrostatic pressure. The API 650 does not give you enough information, you may calculate it back considering it as having a distribution as detailed in IITK-GSDMA/EQ08.pdf. Not an easy task, even you assume a linear distribution over the bottom surface!

So the work should start by understand the loads considered by API and think how the loads are transferred to slab, case by case.
And you need a deep analysis: as you can see some terms must be consider twice, some of them are included in uplift and you need not to double them, others are "peak" terms, others terms are missing explicitly, etc
This is because API does not cover the subject you need.

This approach may be very useful if you want to accurate count the slab with FEA (considering soil modeled as "springs")
You may note that there would be a lot of math here, not covered by API 650 requirements and making an engineer very unhappy!



2. You may ignore all the "subtle things" above remembered.

As a normal engineer, you should consider the 5-21 uplift as a "peak" value.
Simply, your slab must resist to this local "uplift" force.

Note that 5.12.13 gives indication for ringwall only, for slab case you must engineer something by Civil knowledge, since is not covered by API!

And....
Your slab must resist also to "global" uplift forces AND other loads, as sigilli explained.

How?

Not so simple: if you want to evaluate the global effect considering uplift, you would note that considering wind and seismic loads, there isn't the same same uplift for all anchors.
This is not accurate described in API. To solve it, you must consider 1st approach- analyze all loads as they are!

But you still should consider API 650/ E.6.2.3 Foundation and check the overturning stability ratio for mechanically-anchored tank, as E.6.2.3-1 requires.
These requirements can be very conservative and you may ignore the subtle things, as they are not really API requirements...
Note that Appendix E refers to seismic, so you can find anything there for case "wind". You may try to do yourself, by similitude and improving Figure 5-27 by including your slab weight there....but again.... remember the result is not an API requirement!

Consider also E.6.2.2.2 as giving the peak load for seismic events (on maximum compressed side).




Just hope this will help you!

Best regards.