New steel beam supports beside existing p.t. bea,s 2

I assume this is the same structure as we discussed in thread507-445965. In that thread, the slab was 4" thick but the spacing of beams was not given. In this thread, the slab is 4.5" thick and the spacing of beams is given as 13'-3" c/c or 11'-3" clear. It appears that the slab thickness is marginal at best.

If all of the existing tendons fail, I would agree that a steel member is required each side of the existing beams for the reasons stated in the OP. A beam on one side only will not ensure prevention of collapse.

If the new members are exposed steel beams or trusses, measures are needed to prevent corrosion.

As an alternate to the above, it may be prudent to consider an ongoing inspection program to detect and replace tendons which have corroded. It seems likely that, if corrosion is to occur, it will occur at midspan of the exterior spans. These tendons may be examined by means of an inspection port at the bottom of the existing beams.



BA

TO BARetired. Good to hear from you again. Yes it is the same structure that was described in earlier posts here. The slab is complicated--it was designed in different years by different engineers for different additions to the garage. It generally specified that the specified slab thickness is the minimum, and that the soffit is to be maintained flat and the top surface slopes up from the low points at drains to provide drainage slopes. The slab thickness minimum is 4.5" in some bays and 5" min in other bays. I am fairly sure that after accounting for some added thickness for the slopes the slab meets the Code minimum thickness for continuous both end slabs, of span/28 = 11.33x12/28= 4.58", and generally has #5@12 top & #5@12 bottom, although this too can vary. The slab is generally specified as 4.5" thick in some locations, and as 5" thick in other locations. If I had been designing it, I would have designed it 5" minimum thick throughout.

Anyway, I think the slab is ok. My question was directed primarily to whether one new steel beam one side of each existing p.t. conc beam is adequate, if all the tendons fail? Would you agree that it is not adequate (because all of the bottom bars in the slab stop short of one side of the proposed new steel beam), even if the slab is 5" or even 6.5" thick when surface drainage slopes are included, and that there should be a new steel beam at both sides of each existing p.t. beam?

The new steel beams would be hot dip galvanized for corrosion protection, which with proper maintenance to repair leakage promptly when it occurs, should have a 30 year life of more in my opinion.

Monitoring wire breaks by acoustic monitoring over the past 4 or so years does not indicate any particular pattern of wire breakage. My experience of almost 35 years of dealing with corroding tendons is that they can and do corrode anywhere along there length, depending on where water gets trapped along the sheath. This is water that can enters during construction at sheath tear/damage location before concrete is placed; as well as after construction, thru end anchors etc., etc. Although low points can be locations of accumulation of water, it is not necessarily so, depending where the grease forms a dam to stop the flow of water.

Threading new tendons in is problematic because in many cases they found that hey cannot get the tendon thru to the other end, probably because the tendons tend to cross over themselves, etc. Even if tendons could be installed, the new tendons would corrode and fail in as little as 8 years if the sheath is not totally cleaned of moisture and corrosive de-icing salts etc. That cannot be reliably be done.

There are inspection ports at midspans, but these have likely done as much harm as good, in my opinion, because they allow a fresh supply of oxygen to continuously reach the tendons, as well a warm moist air which condenses on the relatively cold tendon steel.

Th client does not want the disruption and bother of periodically having to try to replace tendons. It is very expensive and disruptive.

The so-called drying and greasing rehab is also useless, as we have fund, in my opinion.

I agree that if all tendons fail, new beams are required on both sides of the existing beams because otherwise, with a beam on only one side, the bottom slab steel would not be continuous between supports.

Also, if all tendons fail and a steel beam is used on one side only, the span of the slab increases from 11.25' to 12.75' assuming a six inch wide flange. Then the minimum slab thickness is L/28 = 5.46" for an interior span and possibly more for an exterior span depending on the spacing of the existing beams in that location.

I agree that replacement of tendons is difficult and disruptive. I mentioned it because it seems to be common practice for remedying unbonded post-tensioned floors.

BA

"Unbonded post-tensioned floors" is the main problem, but I am sure you know that.

I agree that the steel beam solution should have one on each side.

to BARetired:
I see your point about the clear span between the new steel beams increasing from 11.25' to 12.75'. Good point. Perhaps though the "stiffening effect" of the existing beam (i.e. the slab thickness is the beam depth for the width of the beam) will help reduce any additional deflection caused by the extended span, although the moment will increase so the amount of rebar might not be sufficient. Actually, we normally design our slabs based on centre-to-centre span of steel beams, so on that basis the span increases from 11.25' to 13.3' which would be a 40% increase in moment.
Our engineer is arguing that the existing concrete beam, even after all its tendons break, still has stiffness to it and can be considered the support for the slab with respect to bottom bar length i,e, that the existing concrete beam and new steel beam act in conjunction as a single support . I don't agree with him.

to ajk1:
Well, I guess it's a matter of engineering judgment. The existing beam certainly must add some stiffness to the slab. In any event, there is one very good reason for providing support on both sides of the beams which we have already discussed, namely the discontinuity of the bottom slab steel.

Hokie66 raises a good point when he says "Unbonded post-tensioned floors" is the main problem, but I am sure you know that. Some of us may know that, but the Post-Tensioning Institute apparently does not know that as it continues to recommend unbonded post-tensioned floors and roofs. Plastic sheaths should be inspected for cuts or tears, then wrapped with tape to prevent the ingress of water but despite all precautions, some water may find its way into the sheaths.

It seems unlikely that water would find its way into all of the sheaths, so the loss of ALL of the tendons seems remote. One idea that may be worth considering is to provide a steel beam half way between existing beams designed to carry a 6'-0" tributary width of slab. In that way, the load going to the existing beams is reduced considerably. It would not be reduced by a factor of 2 because the steel will deflect more than the existing beams. That is okay because it reduces the tendency to form a crack in the slab over each of the new steel beams.

BA

Thanks. I wrote a long response about unbonded p.t. but it seems that the system people deleted it. So to make it short, the problem is mainly due to the old "push-thru" type sheath, with its annular space where water could collect, not so much the extruded type sheath in use today where there is no annular space. There are other problems but they too can be addressed. So far as I know, the extruded sheath tendon has now been in use for many years in the U.S. I would also require full time inspection of the tendon placing, watertight connection between sheath and anchor, and watertight end caps at the anchors. P.T. design is not something to be taken lightly, and I would use it only on very large garages where there was the commitment and budget to ensure it is done right. So in summary I agree with your sentiments.

Your idea about putting a steel beam at midspan of slab in lieu of steel beam beside concrete beam is interesting. Thanks.

Would you agree that if the "engineering judgement" is that the existing beam acts together with the new steel beam to form the support (even after all the tendons have failed), then the existing bottom rebar meets the code requirement about the need for continuity of a portion of the bottom steel extending into the supports?
However my opinion is that once all the tendons have failed, the beam will extensively crack and lose stiffness (the bottom reinf is only 2#5 bars, or in some beams 2#6 bars, so when 13 tendons fail I expect a significant effect on the beam. Top rebar at centre column is only 4#9. Beam span is 53'-6" and beam is 24" x 24").

I was not aware of the old push-thru type of sheath, but your thread507-417723 contains some discussion and photos of it. Also, the type of grease used is an important factor in preventing corrosion of tendons and the photos show "good grease" and "bad grease". There have been instances where grease has emulsified inside the more recent type of extruded sheathing, so the chemical composition of the grease is an important consideration.

ajk1 said:

Would you agree that if the "engineering judgement" is that the existing beam acts together with the new steel beam to form the support (even after all the tendons have failed), then the existing bottom rebar meets the code requirement about the need for continuity of a portion of the bottom steel extending into the supports?

Click to expand...

I would question that judgment. The existing beam will stiffen the slab for a length of 2'-0" but if all tendons failed, it cannot be considered any kind of support for the slab beyond.



BA

I agree with you.

ajk1 said:

...(the bottom reinf is only 2#5 bars, or in some beams 2#6 bars, so when 13 tendons fail I expect a significant effect on the beam. Top rebar at centre column is only 4#9. Beam span is 53'-6" and beam is 24" x 24").

Click to expand...

Wow, 2#5 or 2#6 is a ridiculous small amount of bottom mild steel reinforcing for a 50+ foot span beams, with UNbonded PT.

And 4-1/2" depth RC slabs is equally ridiculous, IMO.

ajk1 said:

Would you agree that if the "engineering judgement" is that the existing beam acts together with the new steel beam to form the support (even after all the tendons have failed), then the existing bottom rebar meets the code requirement about the need for continuity of a portion of the bottom steel extending into the supports?

Click to expand...

There is a version of this that I could get behind:

1) We say that the steel beam is providing all of the stiffness and strength for vertical deformation in the system.
2) We say that the defunct PT beam is providing all of the stiffness and strength for torsional deformation in the system.
3) 1+2 = you have not violate the bottom bar support detailing requirements in my opinion.

I suspect that you'd struggle to get the torsional performance that you'd need out of the beam at such a long span. And, even if you managed it, this isn't a concept that really feels warm and fuzzy to me. It smacks of desperation which is rarely a good sign.

With external PT off of the table, I still kinda feel as though the right answer here is the addition of the new, underside CIP beam depth and new reinforcement that you proposed earlier. If you can scare bats away with spike strips or something, I like it even better with the new concrete 6" wider than the beam on both sides (ease of concrete pour and consolidation). I realize that others have expressed some valid concerns with that system but, with some creativity and determination, I feel that stuff could probably be overcome.

to Kootk - the deepening of the conc beam was my idea and I developed it extensively, until I realized that there would be wide flexural cracks in the negative moment regions where the existing top rebar is relatively small reinforcemet once all the p.t. tendons have failed. So I abandoned this option for that reason. I wish I saw a way around this, but I do not.

I do not see how 1+2 means that the Code requirements do not need to apply, namely that a portion of the bottom bars must extend into the support.

Moreover I do not see how a 24x24" T-beam spanning 53'-6" with only 2#6 as bottom reinf, and only "open" hoops as shear reinforcement (4#9 top bars at interior support) can be shown to have adequate torsion resistance. However even it it were, it would not negate the code requirement to extend one-fourth the positive moment reinforcement into the supporting member. Note that there must be an 8"± gap between the conc beam face and the centre of the new steel beam in order to install the new steel end connection.

To Ingenuity:

The beam is post-tensioned with 23 0.60" dia tendons and the rebars mentioned. o I don't think the rebars are ridiculously small unless they fail to meet the minimum amount of bonded rebar. I will check that.

To Ingenuity: The code minimum thickness for slab continuous both ends, is span/28 = 11.33x12/28= 4.85". Add to this the thickening of the slab for drainage slopes, and you get an even thicker slab. So the slab meets Code generally and has performed well for 40 years, though has been repaired due to rebar corrosion. I would have preferred a little thicker slab, but I don't see anything ridiculous or code violation about what was used.

ajk1 said:

The beam is post-tensioned with 23 0.60" dia tendons and the rebars mentioned.

Click to expand...

23 x 0.6" dia tendons is RIDICULOUS. Surely this is a typo!

If it was 23 x 0.6" tendons you would have a 'balanced' load of 200% of SW...or more than 100% of [SW+ADL+LL]. P/A would be 1000 psi. Uplift at midspan at transfer would have been more than 1.5". Excessive compressive stresses at transfer. Quite a 'laundry list' of issues, IMO.

I have not designed to CAN code for more than 25+ years, but this design would not meet ACI min. bonded non-prestressed reinforcing steel requirements - even using older ACI codes dating back several decades - you would need more than 2 times that provided.

ajk1 said:

I do not see how 1+2 means that the Code requirements do not need to apply, namely that a portion of the bottom bars must extend into the support.

Click to expand...

Because 1 + 2 = the beam is held in place and can therefore continue to serve as a support for the slab on both sides.

ajk1 said:

Moreover I do not see how a 24x24" T-beam spanning 53'-6" with only 2#6 as bottom reinf, and only "open" hoops as shear reinforcement (4#9 top bars at interior support) can be shown to have adequate torsion resistance.

Click to expand...

Precisely as I anticipated. But you never know for sure until you run the numbers.

KootK said:

I suspect that you'd struggle to get the torsional performance that you'd need out of the beam at such a long span

Click to expand...

ajk1 said:

until I realized that there would be wide flexural cracks in the negative moment regions where the existing top rebar is relatively small reinforcemet once all the p.t. tendons have failed.

Click to expand...

Unfortunately, most reinforcement schemes will likely cause some degree of this problem. With the steel beams, those will be pin-ish at the supports and will therefore attempt to create a slope discontinuity and cracking at that location. It'll take a pretty stiff reinforcement strategy to meaningfully avoid that outcome. Seems to me that, of the available non-PT options, extending the beam depth probably is one of the stiffer solutions. That said, I don't know just what you've got up your sleeve for alternates. Of course, if you've got 23 tendons in the bottom of the existing beam, trying to install dowels may not be fun. Maybe they're strategically bundled.

I believe the number of tendons is 13, not 23.

BA

BAretired is correct. The number of existing tendons is 13, in 24" wide existing beam and they are in 2 layers (i.e about 7 tendons per layer).

I think the only way to determine with confidence to what extent the existing beam is "held in place", or acts more as a lump of concrete hanging from the slab soffit, is to run a finite element analysis (bear in mind that we need at least 8" clear between the vertical face of the existing concrete beam and the nearest edge of the proposed steel beam flange, in order to have access to install the end connections for the steel beam). Maybe we should leave this topic now, because I would likely have to provide more info to properly describe the situation to you. I appreciate the thoughts and comments that you have provided.

Another Question

What should be the appropriate limitation for the deflection of the propose steel beams, bearing in mind that this is an existing parking garage and the steel beams are to pick up all the load in the event all existing 13 tendons fail? The clear span of the beam is 53'-6". Should we limit the total load deflection of the proposed steel beams to:

a) L/240= 2.68"
b) L/280= 2.25"
c) some other limit, and if so, what?

Note that if a) or b) is appropriate, that will govern the steel beam design, not the moment.
Let us assume for the moment, for simplicity in providing an answer to this question, that we install a steel beam each side of each existing p.t. concrete beam (i.e. 2 steel beams per existing p.t. conc beam).
In general there are no brittle finishes like ceilings or masonry walls.

As to alternatives, I think external p.t. is probably the most practical and perhaps the most economical although it probably means anchoring the new tendons on the mall side of the garage supporting girder, where they will be visible in the open atrium type multistorey mall and need some sort of architectural covering. I do not yet know if external p.t. might cause excessive tension in the bottom of the existing conc beam at the support...that is the next thing I will calculate. If that is the case, some existing tendons could be cut, but I would prefer not to cut them.