Need a book? Engineering books recommendations...

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

RE: API 650 App E near fault

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Steve,
Thank you for your very helpful explanation and opinion. 
Bob

-----Original Message-----
From:	Steve Meier [SMTP:meier(--nospam--at)tankindustry.com]
Sent:	Thursday, July 29, 1999 7:59 AM
To:	SEAINT
Subject:	API 650 App E near fault

Bob,

Assuming I understand your problem......There are a number of factors to
consider.

1.  I assume the 0.14 g is based on total mass of the tank which is close to
my quick check.  This corresponds to an 0.24 g on the impulsive mass and
about  5% g for the convective mass based on a PGA of 0.4g. Even tho the
convective mass for this tank configuration is a little over 50% of the
total mass, the convective shear is low because the sloshing period is over
6 secs.

2.  The impulsive shear in the API eqns is independent of period and is a
constant 0.6 x the PGA.  Therefore, if you are concerned about higher PGA
than standard values,  the new impulsive load can be scaled up.  Be careful
however,  I have seen ridiculous accelerations that are unrealistic and
exceed the capacity of the soil to transfer the load into the tank.  In the
case you are looking at,  it appears the impulsive load would be increased
by 30% (i.e.  Na) which is a PGA of 0.52g or maximum of 1.3g on the response
spectra.  The convective load would increase by 60%, but this is a small
number.  Overall the load would be about be about 18%g vs the 14%g if my
quick numbers are correct.

3.  To ratio the ASD to an strength level load, simply reduce the Rw by a
factor of 1.4 (or multiply the ASD shear by 1.4).   API does not give a Rw,
but the eqns are based on an Rw of about 4, which gives an R value of about
3.

4.  IMHO it it rare that a welded steel tank needs to be designed for values
in excess of the API or AWWA requirements unless it is unusual
circumstances.  The performance of steel tanks when designed by these
standards has been very good, and there are literally thousands of tanks
designed to API and AWWA criteria only. IMO, it is clear that UBC intends
that the modified repsonse spectra curve be used in near fault zones;
however, what is not clear is the acceptance criteria for nonbuilding
strucutres when these higher values are used.  I recommend you also consider
what the approriate design criteria will be (is soil structure interaction
permited?, is an increase in buckling stresses permitted, can higher
stresses for hydrodynamic hoop stresses be used, will an adjusted annular
plate dimensions and stresses for anchorage be allowed, etc).  I do not
think it is always approriate to go beyond the API loads AND then impose API
stresses and other criteria without considering approriate adjustments.
IMO, It is likely that the increase from 14% g to 18%g for this tank will
have only minor impact on the design even if the API rules are followed so
this is proabaly a moot point for this case.  It is probabaly appropriate to
design for the API level loads and then check for problems at the higher
load level.  In fairness, I have seen many engineers specify these higher
loads and retain the API design criteria, but that is up to you.

Hope this helps.


Stephen W. Meier PE, SE
Tank Industry Consultants, Inc
    7740 West New York St                       684 West Boughton, Suite 101
    Indianapolis, IN 46214                          Bolingbrook, IL 60440
    317-271-3100                                       630-226-0745
    317-271-3300 fax                                  630-226-0802   Fax
    800-539-7102 pager
Meier(--nospam--at)TankIndustry.com