P/S Bridge Girders in CONSPAN 1

[asaldana]

I am analyzing a 2-Girder 90 foot span with train loading. The worst-case beam seems to work as I increase the tendons everywhere except for allowable stresses in the bottom flange from about 0.7L to Midspan. If I allow CONSPAN LA to "Auto Debond" the stress is decreased, but not enough. Any suggestion to manually input debonding lengths (maybe a good methodology) so that the span may work with only 2 girders with a reasonable amount of tendons.

Thanks, Any input is greatly appreciated - even if it is on p/s beam design with CONSPAN - am new at it.

ASaldana

 

[purdue86]

I have designed single voided box girders for railroad loading in Conspan, although it was a few years back. Currently, I use a spreadsheet that I developed for designing prestressed railroad spans.

You didn't specify what kind of railroad loading you are using. Please indicate what kind of "train" loading you are designing for.

Purdue86

 

[asaldana]

I am applying a Light Rail (LRT) Loading.

 

[71corvette]

Try adding strands to the top (you'll likley need to then add more to the bottom, but it often helps), or alternately increasing f'c and f'ci or changing the inital pull on the prestressing strands. Try debonding manually, sometimes CONSPAN doesn't always yield the best result...

Is this LFD or LRFD? If LFD, are the stresses you're talking about at Final 1, 2, 3, or at release? What type of beam section? I think we need more specifics to offer meaningful suggestions.

 

[asaldana]

It is an AASHTO Type IV Beam. The stresses that I am failing at are at Final 1. I need some meaningful guidance to debonding manually, I have figured that CONSPAN does not yield the best results. I am not that familiar with CONSPAN and not too experienced with designing prestress concrete beams. I am already at f'ci = 7000 psi (at release) and f'c = 8000 psi (at final). I was not sure if it is okay by code to bump that up to make the beam work. The controlling live load is a derailment load that is almost rediculous. It seems that I could be allowed some overstressing for that particular load, at least initially. Thanks for your input. And any other suggestions or guidance is appreciated.

ASaldana

 

[rapt]

I may be missing something here, it being late at night, but if the problem is at the bottom between .3L and .7L then it is full load that is causing the problem, not transfer and debonding at the ends will not help anything. You do not have enough PT in the bottom, maybe because extra in the bottom will cause insoluble problems at transfer.

As you are the "Design Engineer" on this, you should not be relying on software to design it in lieu of your lack of understanding of the design methodology. Software should be used to augment you design ability, not mke up for a lack of undersyanding of it.

Learn how to do the calculations by hand before using software and then you will be able to see what the problem is and the solution. You are being paid to be the "engineer" on this, the software is not. Software is a tool to help experienced engineers do the calculations quickly, not to "create" experienced engineers.

 

[asaldana]

I have done p/s beam designs by hand, I have never encountered a loading condition of this magnitude nor have I had much experience with running CONSPAN to know enough about what the program is spitting out. I really needed some guidance, I should say, for debonding the strands once the compressive stresses at release were being exceeded. I, personally, have had no previous experience in relying on debonding to help push a beam to work under the applied loads. I have resolved my beam design issues that were previously presented, but any further constructive guidance on debonding are welcome and appreciated.

ASaldana

 

[71corvette]

8000 psi for f'c is quite high, I wouldn't want to go any higher. You'll already be paying a premium for the f'ci of 7000 psi, It will mean the beams need to sit longer in the casting bed before the strands can be released (time is money).

If its a derailment that controls, I would consider puting that load in the extreme event category if you haven't already. This would allows you to increase your allowable stresses. Anyone else have thoughts on this? Have you checked AREMA for guidance?

I do agree with Rapt, at final 1 if your beam is failing I doubt debonding will help. Sounds like you need more PS steel in the bottom.

The PCI bridge design guide has some very good design examples to follow. You need to be careful with Conspan, I've found the input can be a little tricky sometimes. As they say garbage in, garbage out...

 

[rapt]

You miss the point.

Debonding is only useful in relieving transfer problems at the ends of beams.

If debonding is reducing the stress at midspan in a determinate member, throw the software away because either it is wrong or you are using it wrongly. debonding will only affect the results at the cross-sections where the strand is debonded and you can only debond at the end of the strands.

As I said, do the calculations for yourself so that you can learn the different effects of changes you make (such as debonding) so you know how it should be solved. Your program obviously does not know and you need to know to be able to realise this and to use any software properly!

 

[71corvette]

So, how did you get the beam to work? I would think that debonding some of the top strands may have helped you out a little...

 

[asaldana]

I ended up getting more information on the governing code from the light rail officials. Apparantly, when running the derailment load you are allowed a lower load factor for the live load with impact, instead of magnifying the allowable. I also put straight strands in the top, which helps because the resulting stresses are still very high. I did not feel comfortable increasing the f'c at release to over 7000 psi. That has been done previously, but I did not agree with going that route to make the beam work. Thanks so much for your valuable input.

ASaldana

 

[KSRAO]

May be too late response for this particular question. But thought will be helpful for future reference.

Debonding can be done in the middle too. Not necessarily at the beam ends. Also, CONSPAN does allow debonding both the at ends as well as in the middle of the girders.

PCI manual is of a great help.

-Rao.

 

[rapt]

KSRAO,

Debonding is mnormally done on the ends of the bottom strands to allievate transfer stress problems at points where the P.e is high but the self weight moment is low. This is simply done by sleving or greasing the end of the strand so that it does not bond to the concrete. The strand effectively starts at the inner end of the debonding as far as the beam is concerned. The debonded part is simply there because the strand cannot be cut shorter in a pretensioned member due to the stressing requirements.

Can you describe how the debonding is done in the middle of a beam?