Prestressed column 5

[OAP]

Hi folks,
Let's assume we have a long column (10.0m height ) and cross section of column is 0.3*0.3 m. which we can't have the lateral restraint to reduce the slenderness ratio due to the architectural concept. This column are spacing 6.00 m. and have to support roof truss span of 25.0 m. Do anybody have an idea that if we prestress this column, it's gonna help the slender problem or not?
Thanx
OAP

 

[ishvaaag]

Normally the specificaitons for the design of concrete sections are stated such way that you get tensile side failures. P-delta in-member buckling once considered should be showing the same tensile side failures hence from the way the code imposes to treat the sections.

The axial loads of so slender colums are likely to be small, there being the moments, or flexure, what is going to dominate te behaviour at failure. Hence a moderate increase in compression is unlikely to force the compression side fragile failure, and contrarily will retard the tensile side failure.

Other way said, with some prestress, your section is to stay more integral and with more stiffnes along a bigger range of solicitations, what opposes to buckling (Inertia factor in the Euler equation).

However, if your prestress is excessive, buckling you won't help, because if geometrically central straightens the member, but you may get end squashing, hence fragile compression side failure; yet it is unlikely you go near such prestress.

So in my view and prior to some verification by calculation it is likely that a moderate level of prestress on a given member reinforced with some passive rebar will allow it take a somewhat higher load.

 

[flamby]

While agreeing with Ishvaaag's reasoning, I can add that this might be an interesting topic for research and testing. But before such theories are tested and proven, we engineers cant take them to the field as the codes dont allow.

 

[ishvaaag]

Well, it is clear that the reasoning was in answer to the shown interest. Obviously one must check the column as is (prestressed or not) in accord with the relevant applicable parts of the compulsory code.

 

[dlew]

OAP,

Prestressing would help a free-standing column subject to significant moments. I doubt it would help much, if any, in a regular column.

The prestressing does not change the properties of the column section, thus, the buckling load would be the same than a regular concrete column of the same dimensions.

Prestressing does not increase the capacity of the column to compressive loads.

Prestressing will increase creeping in the column, and might introduce some additional stresses in the beams supported by that column.

The column you described seems to be very close to the upper limit for slenderness given by ACI 318-89, para. 10.11.4.1. Maybe increasing slightly the column size it would comply with the code.

Good Luck!

AEF

 

[OAP]

Thanks all,
I have further question as in case of slender column as I mention above, if I do the non-linear structural analysis and the result shown me that the magnitude of the bending moment don't have any different from linear analysis, let say 3% diff. Can I use the bending moment from the non-linear analysis to design the column section without using the moment magnifier? If I can, what are things which I will concern?
Thanks again
OAP

 

[ishvaaag]

You can directly use the moment so obtained if the material and geometrical nonlinearities have been contemplated.

The material one it seems from your description you have, then the question is if you have the moment values from a P-Delta analysis at the factored level consistent with the final (nonlinear) sectional properties in use.

 

You ask if you prestress the column would it help with the slenderness problem?
Now if you prestress it,you are applying the prestress at every point on the column.So what is the effective length?
->"0"("ZERO&quot
Right?
Hence slenderness will not at all come in the way as the effective length is reduced to "0".I suppose prestressing is a most ideal means of solving the slenderness problem unless proved uneconomical

 

[ishvaaag]

Visitor, I think moderate prestress may help moderately solicited moderately slender columns columns in that moment of inertia is bigger along the range of solicitations. However prestress itself is not the only solicitation, and when in combination with other solicitations -say a compression load applied at ends- buckling appears.

Hence prestress does not solve the slenderness problem, is just something more to consider in the problem.

Furthermore, in a extremely slender column or wall, the perfect centering of the load becomes critical, since any eccentricity will severy affect the initial stress status...this then MAY add to the buckling induced stresses to cause the failure before than if there was not prestress. In this case extremely tight tolerances would be required, and even then the lack of homogeneity of concrete could cause similar behaviour to that described. Hence my clause in 1st paragraph "for moderately slender columns".

 

Ishvaag,
What i am trying to point out(or thinking) is not the intensity of prestress or the quantum of slenderness.I am trying to say since the prestress is going to act on every point on the column , this almost nullifies the effective length of the column(effective length be reduced to "zero&quot,then how will slenderness come into effect?

 

[ishvaaag]

The phenomena of buckling would mean more or less failure upon frank departure of initial shape. If you prestress a slender column well centered in the ends and with significant eccentricity at the center of the column, prestress itself upon growth of the prestress force to the required amount will make te column snap, buckling. By the way it is so that in prestressed beams gravity forces are counteracted.

Now, for your perfecty centered prestress, prestress becomes uniform, hence prestress ITSELF won't be making the member more prone to buckling that was prior to prestress.

Yet the purpose of any structural member is to sustain loads, for columns mainly compressive loads. These will need be checked according to the science of construction to ascertain if they are safe, if contemplated presently according to what specified in the mandatory code. These checks one way or other need to address the modes of failure, as one colleague put here, squashing or buckling, failure in any case.

Now imagine a hollow HSS section prestressed perfectly centered. You can readily see that if you compress such column from both ends such member, or the same made part of some structure, it has the ability to buckle and in fact will if the distance between restraint points does not ensure strength enough for the force applied, or if you want, if one squashing or local mode of failure does not prevails.

Respect your statement, what you can attest is not that the buckling length is zero, but that there's no trend in the perfectly central prestress to buckle out of perfect equality between the section's stresses.... something by the way also understood for any buckling, to later make the assumption of that in real structures some initial imperfection there is.

The buckling length needs yet be determined for the MEMBER in appropriate way (as long as we proceed in our customary memeber by member checks, there are other seminal promising procedures through P-Delta and other Advanced -mostly steel- Structural design).

 

[Ingenuity]

You need to be careful about the concept of buckling under prestress application versus externally applied loads.

Under the action of the prestressing force application, with internal tendons that are in CLOSE contact with the element being prestressed it is not possible to buckle the member whilst prestressing it. The reason is that, as long as the tendons and concrete are in close contact, they will "deflect" together, and any lateral movement of the concrete will be followed by a corresponding movement of the tendon, but since the tendon is in a field of applied tension this will counteract the tendency for the memebr to buckle. Again, this is under the application of the prestress force and assuming NO external loadings. If there is a significant space between the internal tendon and the concrete, such that lateral movement can occur before engaging the tendon, then buckling effects can arise.

As an analogy, If you take say an "S" shaped (or any generally curved, in single, reverse or mulitple curvature) concrete element under the effects on concentric INTERNAL prestress, during the prestressing force application the curved shapes will NOT tendon to straighten NOR buckle. IF the prestress was eccentric there would be a deflection of the shape (due to the Pxe), BUT again, no tendency to buckle.

If the member is EXTERNALLY prestressed, then the above is NOT true.

To demostrate this concept to students, I have taken small rigid styrene foam blocks (say 1" in size and angled on opposite ends) with a concentric hole. Assemble say 10 blocks to form a curved shape (C, S, other) and thread an elastic band in the center. Use a match stick to grip one end (dead-fixed end) then pull on the other end (live end). The "prestress" provides uniform P/A and there is no tendency to straighten NOR buckle because the INTERNAL tendon (under tension) will balance any column action effects. You can keep appplying tension to the elastic band and the block memeber will not produce and global opening of the joints etc and no buckling.

BUT, under the application of an EXTERNAL load to an already PRESTRESSED ELEMENT the member can buckle, as per other axially loaded structural members. The only real advantage to prestressing a column is the delay in the onset of cracking, and the increased moment capacity (so you can argue you have extra stiffness since less cracked, so a net gain in global slenderness effects). There will be a decrease in the axial load capacity depending on the level of axial prestress applied.

 

[ishvaaag]

Ingenuity, it is only a matter of labeling and eccentricity and force to buckle or distroy a beam postensioned from the ends, precisely out ouf the straightening attempt imparted to the tendons, it is due to this that there are limits to the tensile stresses in the prestressing stage there. I came to name such example in order to indicate that prestress itself if eccentrical can itself be a contributor to buckling upon the imparted bending stresses, and that the likelihood of the eccentricity being significant to it increases with the member being slender.

 

[Ingenuity]

ishvaaag,

I think the tensile limits imposed at tranfer of prestress are to control cracking and not much to do with buckling.

The prestress can only be a "contributor" to buckling in so far as eccentric prestress will laterally "deflect" the member and this initial deflection, under the action of an EXTERNAL axial load, may induce more bending, which will inturn introduce secondary moments associated with buckling (P-delta). You also have to consider the external moments and their sign (direction) to correctly evaluate the buckling behaviour.

For a prestressed column, is would not be practical to provide eccentric prestress, and all of the appplication i have experience with are CONCENTRIC prestress.

BUT at the time when the prestressing is taking place, provided the tendon is internal and "in contact" with the concrete it will not buckle.

As an example, if you have a 100m long, 100mm thick precast panel with concentric prestress via a duct and a P-T tendon and you stress it over this 100m (very slender with high l/r) will it buckle? I say it will not due to the internal tendon in close contact. BUT, if you apply the prestress via external jacking or external tendons and no tendon internal, of the concrete then it will buckle because this is analgous to an EXTERNAL LOAD.

Sorry if I appear to be repeating myself, or if I misinterpreted you response.

 

[ishvaaag]

In a 100 mmm thick 100 m long member, the likelihood of some unwanted eccentricity becomes big, and only the fact of that prestress is limited in amount prevents this eccentricity causing the straightening effect caused by the prestress (or the bending action induced if in contact) to ruin the weak member. And yes, limits of the amount of the induced stresses are to prevent cracking, that is, some kind of ruin. So it is a contributor internal or external, as long on eccentricity bending action is induced, and this bending action adds to that in buckling happening, that one considers is feasible to happen in any direction. So central prestress will more likely benefitial where the relative unwanted estochastically eccentricity of the tendon is less, that is, in less slender columns.

Respect that eccentrical prestress is not seen for columns, that is general for precast, but to some extent common in portal frames over highways, and even on whole buildings prestressed of the back of a sustaining core or structure are made to help stability by direct counteraction of prevailing cantilevering bending forces.

 

[v2]

I just read these masseges.

I strongly agree with ingenuity.
and strongly disagree with ishvaaag

What we are talking about here is totally two seperate issues. and these are external loading vs. internal loading.

In general, buckling load limit can not be increased significantly by prestressing the column, other than improving the moment capacity over the expence of the compression capacity.

The external load carrying capacity of a column just before its buckling failure is greatly influenced by the columns cross sectional geometric properties.

 

[Ingenuity]

Okay, lets take a similar example of a 100m long, 100mm thick precast panel with ECCENTRIC prestress at the ends, and also drapes within the elements. Lets forget about local effects that the tendon profile may have on spalling stresses due to curvature etc. Under the action of the prestressing application (not EXTERNAL loadings) with internal tendons that are in close contact with the concrete, will the member buckle?

I say it will NOT for reasons I have stated above. Any other opinions?

 

[Taro]

ACI 318 Section 18.2.5 addresses this issue. It says "the possibility of buckling in a member between points where the concree and prestressing tendons are in contact and of buckling in thin webs and flanges shall be considered".

The commentary says "if the tendon is in complete contact with the member being prestressed, or is an unbonded tendon in a duct not excessively larger than the tendon, it is not possible to buckle the member under the prestressing force being introduced".

 

[ishvaaag]

v2

"In general, buckling load limit can not be increased significantly by prestressing the column, other than improving the moment capacity over the expence of the compression capacity"

I haven't said otherwise. It is only the fact of having the compressive forces where tensile action is expected to develop at the factored level what can be of some benefit, be it from buckling or other solicitation.

Ingenuity

If you have draped tendons in the bench for fabrication, concreted -internal, close contact- and the eccentrical prestress surpasses the flexural capacity of the section, when you release the tie-downs the beam wil snap, cracked under a movement out of the straight initial alignment anc caused by an eccentrical compressive force, internal prestress. Not exactly buckling, but quite close to. It is feasible to ruin a member by excessive internal eccentrical prestress alone.

 

[Ingenuity]

ishvaaag,

I agree it is possible "to ruin a member by excessive internal eccentrical prestress alone" but the failure is compressive via concrete crushing not buckling.

Back in the 1980's i was involved with a university research and testing project that tested to failure prestressed beams with eccentric prestress under the application of the internal prestressing. We were validating a ultimate strength approach to the strength at transfer of PSC beams - not "stresses" since a true limit state code does not talk about "permissible stresses".

As we incremented the prestressing force, crack developed, then the cracks increased substantially, then as the prestressing force was increased the area of concrete in compression was decreasing, whilst the compressive force acting on it was increasing. As the prestress was increased further, the concrete crushed at midspan, very dramatically. I have a cool photo that we took with a fast shutter speed and a trip switch that shows the beam basically exploding upwards.

What is also interesting is that if you use a permissible stress approach to compressive stresses at transfer (eg 0.60 fci to ACI 318) and compare it with a strength approach at transfer, the 0.60fci limits are UN conservative! In the real world, it is not a big concern because the typical amounts of prestress (P/A) we usually apply to concrete are low compared fci and fc.

Taro,
thanks for the ACI reference - it is the first time I have read section 18.2. Thanks.