Initial and effective prestress 2

[saadpervez1]

It is commonly believed that effective prestress will always be less than initial prestress because of the different losses that occur after concrete has hardened. I am working on a pre-tensioned beam design and using the software Concise Beam for this purpose. The results show an INCREASE in from initial prestress and shows effective prestress is greater than initial prestress. How is this possible? What can be the reason behind this?

 

[Ingenuity]

It is not possible. Either you have input some data incorrectly (sign conventions?) or the software algorithms are junk.

I am not familiar or know of "Concise Beam" software - are you a registered user of the software, or is it shareware? Is there user support available?

Seems like this may be your first time doing a pretensioned beam design. I would encourage you to do it by hand calculations to gain some experience and 'feel' for the numbers and concepts. There are lots of textbooks that go into detailed calcs and theory on this subject.

 

[pikku14]

I'm not familiar with the software but is there a detailed output? or does it just spit out a few key answers?

Likely to be an error in inputting some of the design parameters. Check loads, properties, creep & shrinkage parameters etc.

 

[saadpervez1]

No the software gives a very detailed output. the problem i'm solving is a precast beam with a 9 metre span. I am attaching some graphics with this post.

 

[saadpervez1]

and the graphical output

 

[Ingenuity]

saadpervez1,

On the text data that you provided, below the Prestressing Steel Tendon data table it states:

"Calculated Losses: Initial = 2.8%, Final = -0.1%"

You can NOT have a negative final loss. You need to check your input data and amend the final losses so the magnitude (and sign) is greater than your initial losses.

 

[saadpervez1]

Another thing I would like to share, when I decrease the span to say 6 metres, the results are OK. A messed up algorithm most likely?

 

[Ingenuity]

In the data input of the software does the user input DIRECT % values for INITIAL and FINAL losses, or are they based upon software calculated values from strand, concrete properties, time of loading etc etc?

I am not sure what you mean when you state that the results are "OK" is you reduce the span to 6m. What results are you referring?

 

[saadpervez1]

No, these values are calculated by the software.

By OK I mean that the effective prestress becomes < than initial prestress and the % loss is a positive value

 

[cooperDBM]

It is possible if you have a large superimposed dead load relative to your level of prestress. The SDL is added after the initial prestress and increases the tension in the tendon towards mid-span (see your graph). Similar to the ultimate condition where under full factored loading the tension in the prestress far exceeds the initial prestress.

You should check with their technical support to confirm this.

 

[rapt]

Normally, external load effects are not considered in determining effective prestress. They have an effect on creep as the creep loss is determined under the permanent load condition, but only the creep effect is normally included in the effective prestress force calculation, not the full permanent load effect.

If CooperDBM is correct, it is a matter of terminology. His is implying they are reporting the stresses under the permanent load condition not the Effective Long Term Prestress.

 

[Ingenuity]

Besides same unconventional terminology re "effective prestress force" CONCISE BEAM v4.59 does some very strange things when you apply very large superimposed loads (after transfer). I did a trial run of a pretensioned beam with 5x12.7 strands, span 9.0m, 400x1000 section, and using program calc losses it output:

Pj(jacking) = 642.5 kN ==> seems right at 70% of UTS
Pi(transfer) = 825.1 kN, ===> this is 30% HIGHER than jacking!!!
Pe(effective) = 1046.4 kN @ x = 4.500 m, ===> this is more than the UTS of 5 strands!!!

 

[cooperDBM]

Detailed loss methods do account for the explicit regain of elastic shortening loss due to superimposed sustained loads (i.e. ACI committee 209). Simplified empirical methods may not because the the creep coefficients are taken from test beam measurements than implicitly include the regain. With any method putting in unrealistic loading won't give you a good answer. Garbage in, garbage out.

 

[saadpervez1]

Thank you all for these splendid answers, I have taken note of all this and will be contacting the software vendor for a detailed explanation. Will share their response here for all of you. Thanks again, you're awesome