Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: Coefficient of Friction

[Subject Prev][Subject Next][Thread Prev][Thread Next]

Large diameter tanks with small height to diameter ratios (around 0.5) do not have to be anchored because the sloshing action of the liquid prevents the whole tank to be lifted at the same time.  The failure mode of these tanks is to have an “elephant foot” bulge on the download side and/or a tension rip on the opposite side.  These things can be prevented by proper design of the shell thickness.  This relationship was defined in a paper by Mitchell and Wozniak after studying tank failures in the 1964 Alaskan and other earthquakes and incorporated in API  650 App E and AWWA D100.  The main object was to encourage building tanks with greater diameters in relation to their heights where that is possible and economical.

Tanks, reservoirs or vessels with supported bottoms must be anchored.


Richard Hess, S.E.


From: Harold Sprague [mailto:spraguehope(--nospam--at)]
Sent: Sunday, August 29, 2010 9:32 AM
To: seaint(--nospam--at)
Subject: RE: Coefficient of Friction


Unanchored tanks are allowed per Section 15.7.  For those who design large tanks, it is almost impossible to resist lateral loads due to earthquake events.  When you run the numbers, it becomes obvious that anchoring large tanks is almost impossible.  That is why there are provisions to contain the liquid in very large events.   There are a lot of rules depending on the material contained in the tank.  The coefficient of friction is not well defined for FRP tanks.  FRP tanks do not have much ductility and must be designed accordingly.  But the bottom line is that there are provisions to design unanchored tanks per ASCE 7, Section 15.7.   

Regards, Harold Sprague


Date: Sat, 28 Aug 2010 16:51:17 -0700
Subject: Re: Coefficient of Friction
From: paul.blomberg(--nospam--at)
To: seaint(--nospam--at)

This area (Wyoming) is relatively low seismically (SDC B).  I'm working in the tanks section of ASCE 7 (I don't have it with me but I think it is section 15.7).  An unanchored or self anchored tank is an acceptable approach as long as the numbers work out.   There are many tanks small and large throughout the US that are not anchored. 


This tank is indoors so wind is not a factor.  When outdoors, the tank is calc'ed for wind loads with the tank empty to check sliding and overturning.  This tank is not considered a component so the ASCE 7 friction disallowance does not come into play. 


The problem I have is the friction factors that I can find are for a steel tank on concrete.  I have a fiberglass or plastic tank on concrete.  If it were a steel tank it would work out easily but I need to adjust for the difference in friction between steel and fiberglass.  For reference, the tank is 12' diameter by 17' tall.  It was not placarded as an ASME, API, AWWA, etc. tank so I'm not using one of those standards.



On Sat, Aug 28, 2010 at 1:17 PM, sgeconsulting <sgordin(--nospam--at)> wrote:

The code implicitly disallows the use of friction to resist seismic forces, hence the use of "self-anchored" tanks does not appear a great idea. Besides, such tanks will be quite susceptible to wind when empty.


That said, we routinely use friction between the footings and the soil to resist seismic forces. How does that work from the code and just rational standpoints?

SGE Mobile


On Aug 28, 2010, at 8:29, "Richard Hess" <RLHess(--nospam--at)> wrote:


That will not work if it is in earthquake country.  The vertical forces will dance it all over the place.


Richard Hess, S.E.