Seismic calculations carried out by TANK seem to be inconsistent with Code requirement. For sites where only Peak Ground Acceleration is defined and non ASCE methods apply, terms as SDs, Ss, So do not apply as defined for ASCE methods.Tank internally works out value of SDS and assigns a value for Vertical acceleration AV which is the maximum (0.14xSDS) for ASCE regions as stated in Code. Sections of Code clearly indicate that AV can be nil or any other value if so specified by Purchaser. Tank does not address this option the offshoot of which is that one ends up unnecessarily specifying thicker bottom courses or anchors, both of which are costly propositions.Also while several calculation errors in Ver 3.1 as Ac, Wrs seem to be corrected in Ver3.3 ( I have just run a sample at the dealer, not conducted in-depth study), others as shell compressive, compressive allowable and Hoop stress errors still remain.These can easily be cross checked by trying out the example problems in the Code.Hopefully these will be addressed before long if full benefit of use of software is to be realised.

Regards,

I've worked through the Code examples and TANK produces what the Code gives (exclusive of the inconsistencies and errors). Send your hand calculations in to TechSupport and we'll go throught them.

Dear Richard,

Thanks for your mail.

Problem 5 has Av=0. I have been unable to get this condition anywhere in TANK as TANK makes some internal calculations and assigns a non zero value to Av. Kindly elucidate how this can be brought about as this has great impact on stsability and stresses.


Pl find herebelow output of Problem 6. I have tweaked the input to get the relevent parameters in line with the example values.D=100ft, H=40ft, G=0.7. Only Ai is slightly off. The Code result is around 15000psi while in this output the result exceeds 38000psi, hence declared "unsafe" while not so. My hand calculations agree with Code results. These are in a state of dribble and shall be mailed to your tech support after tidying.



TANK, API-650/653 Analysis COADE, Inc. @1994-2008 Page
Licensed To: MOTT MACDONALD CONSULTANTS PVT LTD I
Job: APItrlAY Date: 03/03/11 Time: 11:45:17
SEISMIC EVALUATION RESULTS - Appendix E
API-650 11th Edition, June 2007

Site-Specific Ground Motion
Design Fluid Weight (lb.) 0.13715E+08
Sp -design level peak accel for nonASCE 0.00000
Ss -MCE at period of 0.2 seconds 1.0300
S0 -MCE at period of 0.0 seconds 0.28000
S1 -MCD at period of 1.0 seconds 0.42000
SDS-design spectral accel parameter 0.89000
FA -acceleration based site coefficient 1.0880
FV -velocity based site coefficient 1.5800
TS - FvS1 / FaSs 0.59216
TC -convective sloshing period (sec) 6.0928
Ac -convective spectral accel parameter 0.54015E-01
Ai -impulsive spectral accel parameter 0.25429
Wc -effective convective fluid weight(lb. ) 0.70914E+07
Wi -effective impulsive fluid weight (lb. ) 0.61704E+07
Vc -convective liquid base shear (lb. ) 0.38304E+06
Vi -impulsive liquid base shear (lb. ) 0.16533E+07
V -total design base shear (lb. ) 0.16971E+07
VS -shear resistance (lb. ) 0.53387E+07
Xc -ring wall convective moment arm (ft. ) 22.956
Xi -ring wall impulsive moment arm (ft. ) 15.000
XCS-slab convective moment arm (ft. ) 35.380
XIS-slab impulsive moment arm (ft. ) 39.449
WS -shell+appurtenances weight (lb. ) 0.24383E+06
Wrs-roof, framing 10% snow weight (lb. ) 0.00000
Mrw-ringwall overturning moment (ft.lb. ) 0.26170E+08
Ms -slab overturning moment (ft.lb. ) 0.64451E+08
AV -vertical acceleration parameter 0.12460
Ge -effective specific gravity 0.66511
wa -resisting annulus force (lb./in.) 147.04
wt -tank + roof weight at shell base (lb./in.) 64.678
J -the anchorage ratio 1.0460
L -reqd min annular plate projection (ft. ) 0.00000
Wab-Minimum anchorage resistance (lb./in.) 0.00000
N -number of anchor bolts required 0



TANK, API-650/653 Analysis COADE, Inc. @1994-2008 Page
Licensed To: MOTT MACDONALD CONSULTANTS PVT LTD I
Job: APItrlAY Date: 03/03/11 Time: 11:45:17
Pab-anchor seismic design load (lb. ) 0.00000
Sc -shell compressive stress (lb./sq.in.) 866.12
Sa -shell allowable stress (lb./sq.in.) 4506.7
Height of sloshing wave (ft. ) 2.4419
Required Freeboard (ft. ) 1.7093
Course Hoop Stress Hoop Allowable
(lb./sq.in.) (lb./sq.in.)
1 38370. 26600.
2 27450. 26600.
3 19294. 26600.
4 30193. 26600.
5 18420. 26600.
6 8467.9 26600.
7 9942.9 26600.
Seismic Evaluation Summary.
Seismic shell stress check passed.
Base shear within limits.
Hoop stress exceeds allowable.



Regards,

Dear Richard,

Further to the calculation in Ver 3.1 I posted earlier I got the same file processed in Ver 3.3, tweaking the input to get the relevent parameters in line with the example values.D=100ft, H=40ft, G=0.7.The value of ts was taken as 0.355 in line with the value calculated by TANK in design mode. Since stress induced is in inverse proportion to thickness, the Code result would work out to 24479 psi, very close to allowable limit.
TANK gives result of 20179 psi for the bottom course.

Manual calculations give following values, disregarding the accuracy lost in conversion from FPS to SI units and are comparable to Code results

Ni = 1288, Nc=162.5 Nh= 7098 Hoop Stress=24500 psi

TANK gives values of

Ni = 1563, Nc=163.69 Nh= 7277

I believe Ni needs a review

TANK, API-650/653 Analysis COADE, Inc. @1994-2010
Licensed To: MOTT MACDONALD CONSULTANTS PVT LTD ID: 24153
Job: APItrl_33 Date: 03/03/11 Time: 14:58:23 Ver: 3.30
18
SEISMIC EVALUATION RESULTS - Appendix E
API-650 11th Edition, Addendum 2, Nov. 2009
Site-Specific Ground Motion
Design Fluid Weight (lb.) 0.13715E+08
Sp -design level peak accel for nonASCE 0.00000
Ss -MCE at period of 0.2 seconds 0.90000
S0 -MCE at period of 0.0 seconds 0.36000
S1 -MCD at period of 1.0 seconds 0.42000
SDS-design spectral accel parameter 0.89200
FA -acceleration based site coefficient 1.1400
FV -velocity based site coefficient 1.5800
TS - FvS1 / FaSs 0.64678
TC -convective sloshing period (sec) 6.0928
Ac -convective spectral accel parameter 0.54730E-01
Ai -impulsive spectral accel parameter 0.25486
Wc -effective convective fluid weight(lb. ) 0.70914E+07
Wi -effective impulsive fluid weight (lb. ) 0.61704E+07
Vc -convective liquid base shear (lb. ) 0.38811E+06
Vi -impulsive liquid base shear (lb. ) 0.16831E+07
V -total design base shear (lb. ) 0.17272E+07
VS -shear resistance (lb. ) 0.53769E+07
Xc -ring wall convective moment arm (ft. ) 22.956
Xi -ring wall impulsive moment arm (ft. ) 15.000
XCS-slab convective moment arm (ft. ) 35.380
XIS-slab impulsive moment arm (ft. ) 39.449
WS -shell+appurtenances weight (lb. ) 0.24383E+06
Wrs-roof, framing 10% snow weight (lb. ) 0.10210E+06
Mrw-ringwall overturning moment (ft.lb. ) 0.27336E+08
Ms -slab overturning moment (ft.lb. ) 0.65741E+08
AV -vertical acceleration parameter 0.12488
Ge -effective specific gravity 0.66503
wa -resisting annulus force (lb./in.) 147.03
wt -tank + roof weight at shell base (lb./in.) 91.761
J -the anchorage ratio 0.97265
L -reqd min annular plate projection (ft. ) 0.00000
Wab-Minimum anchorage resistance (lb./in.) 0.00000
N -number of anchor bolts required 0
Pab-anchor seismic design load (lb. ) 0.00000
Sc -shell compressive stress (lb./sq.in.) 924.21
Sa -shell allowable stress (lb./sq.in.) 4506.7
Height of sloshing wave (ft. ) 3.5931
Required Freeboard (ft. ) 2.5152
Course Hoop Stress Hoop Allowable
(lb./sq.in.) (lb./sq.in.)
1 20179. 26600.
2 20735. 26600.
3 21358. 26600.
4 21445. 26600.
5 16004. 26600.
6 10363. 26600.
7 4723.0 26600.
Seismic Evaluation Summary.
Seismic shell stress check passed.
Base shear within limits.
Hoop stress within allowable.


Regards

Please send the .tki file so we can work with your entire model.

Dear Linus,

I'm not able to clarify your points, but I would highlight some aspects.

First I think you made two calculations having different seismic spectra: S0= 0.28g and S0=0.36g.

Second, regarding Dynamic Liquid Hoop Forces/ stress evaluated by API/650 E.6.1.4, you may note that there are consistent differences between API 650 Basic 11th edition and API 650/ Addendum1 and Addendum 2.

Comparing the formulas, you may note that in API 650 Basic 11th, the calculated values were 4 times bigger than in corrected formulas- Add 1 and 2.
More specific, today USC formulas coefficients have been changed from 18 to 4.5, from 11 to 2.77, from 5.54 to 1.39 and from 3.9 to 0.98.

Hope these remarks would help you.
Regards,

Dear Mariog

Thanks for your interest.

As mentioned by me I have "tweaked" the input to match in value the parameters as given in Code Problem 6. So, Si are unimportant. Ai,Ac,Av decide the Hoop stress and I have manipulated the inputs to achieve in TANK as near values as possible as given in Code example. Interstingly this has required different values of So in the two versions as rightly pointed out by you.

I agree Code constants have changed in Addendum 1 and Addendum 2.However with all the patches and updates our IT man has been installing I always thought the version 3.1 at my disposal was having the latest update. Moreover unlike PVelite, Codecalc and other software TANK leaves you in the dark about the equations and values considered. I have therefore attached calculations in ver3.3 which undoubtedly complies with Addendum 2 to show the discrepancy in the results between Tank and Code calculations.
My main issue is also with Av which in some projects I handle is nil or some other Purchaser specified value. I am unable to process such requirement in TANK as TANK assigns its own value to Av.


Regards,

Well, at least for versions enhancement list, this is clear explained in

http://www.coade.com/products/tank

Version 3.30 Changes (09/10)
Description:

* Incorporated changes to address API - 650 11th Edition Addendum 2.
* Incorporated changes to address API - 653 4th Edition Addendum 1
* Added a scratchpad to perform the area computations of Figure F.2.

Version 3.20 Changes (09/09)
Description:

* Implemented changes to address API - 650 11th Edition Addendum 1.
* Implemented changes to address API - 653 4th Edition.

Version 3.10 Changes (09/08)
Description:

* Incorporated changes to address API - 650 11th Edition.
* Incorporated metric bolt sizes.
* Added colorized output.
* Relocated\Examples and \Systemdirectories to %allusersprofile%, consistent with Microsoft recommendations.

Addendum 2 came out in Nov 2009. I do not think Ver 3.10 would become obsolete within months of its purchase.The patches and updates installed by dealer would take care of enhancements.

Also Tank Forum Home Page states "Please note, the current edition of TANK is Version 3.20, Build 100715."

The second run by me is in Ver 3.3. Hopefully what I wish to bring out would be addressed.


Regards,

Dear Linus,

I do not comment your expectations on version 3.10.
However, just looking to the release date of version 3.10 (09/2008) your Client would ask you whether you comply or not with the contractual API 650 edition.

About your Tank 3.30 calculation, first I understand is based on thickness 0.355in and you've considered tuned seismic spectrum as giving Ai, Ac values near the values found in example #6 from EC/ API 650 Addendum 2 . In addition, you've considered AV -vertical acceleration parameter 0.12488g as in Example #6.
I understand your question is related to the fact in example #6 the total hoop stress is 15,449 psi @thickness of the bottom shell course of 0.5625 in.
Your claim is that when thickness of the bottom shell course is 0.355in in the same seismic calculation, the hoop stress should be 15,449*0.5625/0.355=24,479 psi,
instead you can see other value in your 3.30 Tank report based on thickness 0.355in.
I hope I understood your question.

My opinion: as a Tank user involved in this discussion or as an engineer looking to your Tank posted results, I have no evidence which is the thickness considered by your seismic Tank calculations and I have no chance to understand why a particular value of thickness has been considered in your seismic calculation. So I can say nothing about your claim.

I cannot see other solution than sending your calculation to Coade; for sure they would find an explanation based on an analysis of your Tank file.

And just as a personal remark, for me is hard to understand why you are following a complicated procedure to check Tank results; I think it is enough to input 0.5625in as thickness with CA=0 and to check the Tank 3.30 hoop stress is or is not 15,449 psi (or nearby value).

My best regards.

Dear Mariog,

You understand my question very well.

As I mentioned in earlier post, TANK is not very user friendly and you have to decipher the input values in the end result. Another problem I have been facing is that the new software is not fully installed as during the trial run there were issues. The software works in "Design Mode" but crashes in "Analysis Mode". Hence even though I input 0.5625in as thickness and CA=0, the software calculates the required value which is 0.35in in this case and uses this value for all further calculations irrespective of the input. I am compelled to use a seemingly convoluted approach to establish my point. My dealer has apparently logged this issue with Coade and hopefully this will be sorted. I am also therefore not able fine tune, create a straight example and send file to Coade.Maybe I shall send my initial file today.

I would like to have your views on how to have Av=0. Here we generally have only Peak Ground Acceleration as input and at Purchaser's option Av may be nil or some other value.

Regards,

I second your thoughts on the issue of Av.
For non-ASCE calculations, Av would have any value (and often that value is zero) depending on the local seismic code/ Project specification.

My best regards.

Maybe somebody being near Tank 3.30 would help you running a simple model.

D=100ft, H=40ft, G=0.7
one course H=40ft, thickness 0.5625in, CA=0

open roof
any good trail value for bottom thk
no wind

seismic parameters as posted above
i.e.
Sp -design level peak accel for nonASCE 0.00000
Ss -MCE at period of 0.2 seconds 0.90000
S0 -MCE at period of 0.0 seconds 0.36000
S1 -MCD at period of 1.0 seconds 0.42000
SDS-design spectral accel parameter 0.89200
FA -acceleration based site coefficient 1.1400
FV -velocity based site coefficient 1.5800

Self-anchored Rwi=3.5, Rwc=2, I=1

AV -vertical acceleration parameter 0.12488
Ac -convective spectral accel parameter 0.54730E-01
Ai -impulsive spectral accel parameter 0.25486

The goal is to check
if calculation gives Course Hoop Stress 15,449 psi?

This clarification would end Linus' incertitude on seismic calculation.
Thank you.

Send the ".tki" files in to Tech Support so we can work with your data.

Good morning Richard,

Could you explain me what height "H" used TANK program to calculate Xi?

I have a tank with the following dimensions:

D = 150 feet
H = 48 feet
H Liquid = 45,932 feet

According to API 650 standard, When D / H is Greater than or equal to 1.3333, the height Xi is Determined by Equation E.6.1.2.1-1:
Xi = 0.375H

But in the TANK report shows:

Xi-ring impulsive moment arm wall (ft.) 18,093

Which does not coincide with the result that should give
Xi = 0.375 * 45,932 = 17.2245 feet

Could you help me please,

Elaine Urbina
Venezuela

The value of Hliquid can be increased due to internal pressure. This can happen if:

a) you activate this via the configuration, or
b) Appendix F.7 is determined to be necessary.