Column Reinforcing Lap Splice Question (ACI 318) 2

[Hokie93]

I am seeking opinions as to whether the provisions of ACI 318-11 12.14.2.3 can be applied to a column member that is solely in compression. ACI 318-11 12.14.2.3 states:

Bars spliced by nonontact lap splices in flexural members shall not be spaced transversely farther apart than the smaller of one-fifth the required lap splice length, and 6".

I have a situation where the dowel bars at the base of a cast-in-place concrete column need to be offset from the column longitudinal bars by approximately 3 inches. The fact that 12.14.2.3 specifically indicates noncontact lap splices are permitted in "flexural members" is giving me pause. I would not typically classify a column subject solely to compression as being a flexural member. I checked ACI 318 back to the 1989 version and the requirement is unchanged.

 

[manstrom]

If the base of the wall is assumed fixed (like a retaining walls), then I would pay close attention to this. If the wall is pinned and the base, I don't think it matters.

 

[DaveAtkins]

I believe that requirement has to do with preventing a diagonal or step tension failure between bars which are spaced too far apart. So I don't see how it applies to a column in compression.

But are you asking if lap splices are even permitted in columns?

DaveAtkins

 

[Hokie93]

Lap splices in columns are definitely permitted by ACI 318. My question is whether noncontact lap splices in columns are permitted by ACI. The Code language is just not clear to me though from a mechanics standpoint, I would think that noncontact lap splices in columns would be permitted.

 

[jayrod12]

I would think if they allow it in flexural members then it would definitely be permitted in columns. And since you say the column is compression only I'm assuming it was designed as a pinned base in which case as manstrom noted, I don't think it matters.

 

[KootK]

I think that this issue is of some consequence, even in a compression only situation.

1) the compression lap mechanism is more or less the same as the tension lap mechanism. If bars are spaced too far appart. If The concrete mini-struts between knurls get too steep, a diagonal tension failure of some sort is initiated.

2) Hokie's column may not work in bearing without the bars running through to the footing. If so, the compression lap matters.

I can think of no rational reason why 14.12.2.3 should not apply. KootK approved.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[precast78]

Does it matter how far it is from the longit. steel in the column if the dowel is fully developed into the column?

 

[KootK]

Yup. All development length is good for is passing force from one bar to the next.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[KootK]

The dowels could go anywhere I think. Once we're beyond the scope of a compression lap, however, I think that we'll find ourselves in strut and tie territory. That will yield a generous lap and some localized ties. Were I in Hokie's shoes and truly concerned, I'd add the ties as a matter of course.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[TXStructural]

Non-contact laps and lap splices are THE SAME under ACI 318. The restrictions on tension/flexural laps that you note do not apply to non-contact compression laps. The commentary of 318-11 says:
"
R12.16 — Splices of deformed bars in compression
Bond research has been primarily related to bars in tension.
Bond behavior of compression bars is not complicated by
the problem of transverse tension cracking and thus
compression splices do not require provisions as strict as
those specified for tension splices. The minimum lengths
for column splices contained originally in the 1956 Code
have been carried forward in later Codes, and extended to
compression bars in beams and to higher strength steels. No
changes have been made in the provisions for compression
splices since the 1971 Code.
"

 

[KootK]

Yes, based on a strict reading of ACI, one could come to two,contradictory conclusions:

1) Because there is no guidance provided for non-contact compression splices, anything goes or;
2) Because there is no guidance provided for non contact compression splices, they're simply not allowed.

Neither of these interpretations seems rational to me. I agree that the compression case is better than the tension case but, at the same time, it's not as though you can lap one bar to another bar down the street just because it's in compression. I still contend that:

1) The non-contact tension lap provisions are a conservative and simple way to address the compression lap case and;
2) When non-contact tension lap requirements cannot be satisfied, strut and tie analysis makes sense.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[TehMightyEngineer]

What about non-ACI codes, do they offer any guidance on this issue?

Maine EIT, Civil/Structural.

 

[jayrod12]

CSA only has stuff about laps in flexural and tension members. And it says the transverse spacing between lapped bars must be less than 1/5 of the lap length or 6".

 

[TXStructural]

It is extremely common to have non-contact laps of longitudinal bars in columns. In Texas (and many other places where there is not high seismicity), column bars are lapped at every level in a structural frame. The cage for the level above is simply slipped over the bars extending from the level below, and a bar-for-bar lap is not always made. Frequently, there are different bar counts on these cages.

This is standard practice and is permitted by ACI 318. The theory is that you are developing each bar into the concrete, and the bars above connect to the bars below through the concrete because they are fully developed. (NOTE that you are not permitted to reduce splice length for excess reinforcement the way you can for development length.)

Now, if you ask about MINIMUM spacing of bars to allow consolidation of concrete between bars, we can have a discussion about bond and development. (ACI 318 says that for CONTACT laps, the clear spacing must be maintained between splices, but is silent on spacing of non-contact splices except to have a maximum spacing.

 

[KootK]

This is standard practice and is permitted by ACI 318

Can you provide a clause reference for that TX? I'd like to check it out.

The theory is that you are developing each bar into the concrete, and the bars above connect to the bars below through the concrete because they are fully developed.

This worries me. It strikes me as a fundamental misunderstanding of what development length is. Transfer of force laterally needs to be addressed in some fashion, either by prescriptive limits or explicit design. To quote myself from four threads up:

it's not as though you can lap one bar to another bar down the street just because it's in compression.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[TXStructural]

Contact laps have the same issues with lateral restraint from splitting, the bars do not interlock any better when in side-by-side contact than when separated by a short distance. All column bars are inside of ties or spirals, which should provide adequate confinement where the concrete does not.

ACI does not differentiate contact and noncontact lap splices EXCEPT for a single provision related to bar spacing in flexural members, as we have been discussing:

(ACI 318-11) 12.14.2.3 — Bars spliced by noncontact lap splices
in flexural members shall not be spaced transversely
farther apart than the smaller of one-fifth the required
lap splice length, and 6 in.

Knowing how ACI committees operate and how tight ACI TAC is with wording, I feel fairly confident that had they intended for all laps to be contact unless noted otherwise, it would be clearly stated. I do find it odd that ACI 318 does not really define "lap splice" any place I can find. CRSI does define lap splices as both contact and noncontact without preference (except to say that it is easier to maintain bar position when they are wired together.)

http://www.crsi.org/index.cfm/steel/lap

It is my assertion that noncontact lap splices are analogous to developing two not-too-distant bars into a mass of concrete. This corresponds well to the way we understand rebar bond and development to occur in concrete. For example, while development length and lap splice length are not the same:

(ACI 318-11) 12.15.3 — When bars of different size are lap spliced
in tension, splice length shall be the larger of ld of
larger bar and tension lap splice length of smaller bar.

Which says to develop the larger bar into the concrete AND meet the lap splice requirements for the smaller bar.

ACI 318-14 (draft) commentary says:

R25.5.5 — Lap splice lengths of deformed bars in compression
Bond research has been primarily related to bars in tension.
Bond behavior of compression bars is not complicated by the
problem of transverse tension cracking and thus compression
splices do not require provisions as strict as those specified
for tension splices. Lap splice requirements particular to
columns are provided in Chapter 10.
[There are no relevant provisions in 318-14 chapter 10, which is the new "Columns" section.]

And 318-11 (and -14) has:
[in relation to the flexural member provisions]
R12.14.2.3 — If individual bars in noncontact lap splices
are too widely spaced, an unreinforced section is created.
Forcing a potential crack to follow a zigzag line (5-to-1
slope) is considered a minimum precaution. The 6 in.
maximum spacing is added because most research available
on the lap splicing of deformed bars was conducted with
reinforcement within this spacing.

The May-June 1996 ACI Structural Journal had an article ("Bond Strength of Noncontact Tension Lap Splices") that presents test results showing that non-contact laps up to 5 db (30% of Ld) actually perform marginally better than contact splices in tension. As said in ACI 318-14 (draft) paragraph R25.5.5 (above), flexural conditions are generally considered less ideal for splices, so I expect a non-contact lap in a compression to perform better than in tension or flexure. To me, this would indicate that unrestricted noncontact laps are permitted in compression members. At the very least, they would be no more restricted than those in tension or flexural members.

I hope this explains my position a bit more thoroughly.

 

[KootK]

That was great TX. Thanks for putting all that together. That does clarify your opinion and I agree with every word.

I don't think that anyone, including Hokie, was making the argument that contact column splices are necessary. The question being debated is how far can lapped bars in compression be separated before it is too far?

I firmly believe that there is such a thing as too far. Just because two neighbouring bars are developed in compression doesn't mean that the concrete between them will pass muster. Since 318 appears to give no explicit guidance at all for maximum separation between lapped compression bars, what is a designer to take as the limit? I think that it should be the tension lap provisions.

Most columns contain enough vertical reinforcement that's almost impossible to violate the 6"/0.2Ld limits of course.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[TXStructural]

"Most columns contain enough vertical reinforcement that's almost impossible to violate the 6"/0.2Ld limits of course."

I think this is probably true.

There is also the issue of what rule to follow about corners in ties. Which bars in a noncontact lap are in corners of ties? One more thing to examine, since I have seen nothing that covers this.

 

[RUStructural]

In the absence of any guidance on this I would use the tension lap limits on non contact splices. WashDOT did some research on non-contact splices for bridge column to drilled shaft connections. It's all for tension laps but, its well done and worth a read. Parts of it have been incorporated into the most recent AASHTO LRFD code (6th edition).

[URL unfurl="true"]http://www.wsdot.wa.gov/research/reports/fullreports/417.1.pdf
[/url]