Allowable floor deflections versus workmanship standares. 1

[BEFORT]

I have a question regarding an apparent conflict between the requirements of the UBC and published workmanship standards.

I will first set up the scenerio. A rectangler building 40'x 20' with CMU walls all around. A single wood beam bisects the 40' dimension. Wood joists bear on a wood ledger each end and the bisecting wood beam. A wood ledger runs parrallel to the wood joists at the side walls.

Now the issue: Table 1604.3 of the 2003 UBC allows for the following deflections: L/360 for LL deflection (20x12/360= .67") and L/240 for DL+LL deflection =1"
The workmanship standards state "unevenness of floors should not exceed 3/16" in 48". Floors should be level within 1/2" in any 12'".

Now the conflict: If I were to design the floor using the UBC requirements, the allowable deflection could be as much as 1". However, per the standards, anything more then 10'/12'x.5"=.42" would be unacceptable (unlevel floor). Also, at mid span, at the side walls the wood joist could theorectically deflect 1" while the bolted wood ledger would not, it may even be higher due to placement tolerance and any crown in the ledger. This condition resulting in an uneven floor per the standards (1"+ versus 3/16"). To make matters worse, the ledger could theoretically be cut to match the profile of the deflected wood joist but depending if the live load was realized or not, could ossilate between being within standards to being outside of the standards.

The issue: I could meet all the design requirements of the Building code, the contractor build the project per the drawings and yet still have some major problems if the workmanship standards were strickly enforced. I have been doing this for a long time and I have always used the code values and have never considered the apparently much more restrictive workmanship standards.

Your thoughts please.

Wayne Befort SE AZ36611

 

[Lion06]

I interpret "unevenness" as a sort of measure of the floor being level when constructed, not deflection during service. Using the unevenness criteria, you wouldn't simply use your deflection because the beam has also deflected 12" away from it. So the real floor unevenness for that 12" section is the differential deflection, not the total deflection.
I don't think the workmanship standards you reference have anything to do with deflection.

 

[COEngineeer]

yeah.. I agree with structural EIT. Probably has a lot to do with concrete floor.

 

[BEFORT]

Thanks for the input, but I'm not sure I understand what you mean. This is a wood framed second floor. What would the conrete floor have to do with it?

The standards I have referred are State Construction Standards. Uneveness, is different then level. Unevenuss tries to take into account the joists being set at different elevations thusly producing a wavey affect.

The bottom line, an Engineer prepares the construction documents per code, the builder constructs per plans, the owner moves in and loads up the floor and the floor deflects more then what is allowed per State mandated construction standards. In this case, in regards to level, the joist will deflect 1" at the center and only .42" is allowed per the State Construction standards.

Should I start using more restrictive deflection criteria then what is allowed per code?

Thanks in advance and I really appreciate the consideration.

 

[BEFORT]

Ouch. No wonder you were confused. The workmanship standard is 1/2" in 12 feet for level.

 

[Lion06]

Befort-
I am not quite following you. Your workmanship standards are as you state, for "how the builder places the beams".
Even if you make the leap that they are referencing deflection somehow with this standard, your total deflection is happening over half the span (10'). The value of 1/2" is per 12" or 3/16" per 48". One way you get 5" of deflection is allowed (10')*(.5"/ft), the other way you get 9.5/16" -> (10')*(3/16"/ft).
I am still not seeing this as a deflection issue though, more a "workmanship standard" for how the contractor should build it.

 

[Lion06]

Ok, I just saw your last post and that certainly seems more reasonable, but I still think it is setting tolerances and possibly to preclude the contractor from using wood that is warped beyond this amount.

 

[BEFORT]

O.K. I'll get more specific. This is real world stuff. I designed the above project, the owner moves in and puts a water bed right in the middle of the room. The owner says his floor is not level. The Regestrar of Contractors inspects the floor by placing a 12' level not at the center of the room (if the had it would have been level) but places one end at the end of the joist and the other at the center of the span and measures a difference in elevation of 5/8" and says it's no good. Keep in mind, the elevation at each end of the joist is identical. What he measured is the sag in the joists which meets the code but doesn't meet the workmanship standard of 1/2".

Does this make sense now?

Thanks once again for your input.

 

[Lion06]

That certainly makes sense - Is the workmanship standard part of UBC or is it something seperate?
I just don't see floor uneveness or levelness as deflection. Levelness evenness, IMO, is the state of the floor while not under load; while deflection is the state of the floor under load.
I would venture to guess that if you start designing for the deflection of L/576 (1/2" per 12'), you will get a lot of complaints from owners and contractors about increased cost due to beefier framing.
I would like to hear other's opinions on this.

 

[BEFORT]

Great, it sounds like your getting it.

Now the second condition. Unevenuss of the floor. The floor is perfectly even across the length of the room (it has to unless the wood beam on one end and the bearing ledger on the other end or the joist itself had waves in it) However, the first joist adjacent to the wall (2' away) has deflected 5/8". The parrallel ledger was placed by running a straight chaulk line from the end of the wood beam to the end of the bearing ledger. Therefore, at the center of the joist the plywood rises 5/8"+- from the joist to the wall. Thusly exceeding the 3/16" in 48" required per the workmanship standards by a substantail margin. I guess the GC could have cut the top of the ledger to match the profile of the joist but thats not practical.

Your statements regarding workmanship standards have been very helpfull. I'm beginning to believe that the Regestra is taking a somewhat simplistic approach at determing allowable construction tolerances. I suspect a better method would be to determine the actural elevation at each end of the wood beam and each end of the bearing ledger and in no case should this elevation deviatate by more then 20/12x.5"=.83" (20'x20'room). Does this make sense?

 

[Lion06]

It sounds like the person doing this check on your building should have done that BEFORE the water bed was put in place!! Of course you are going to have relatively significant deflection placing a water bed in the middle of a 20' span.
What caused the 5/8" deflection before the plywood was placed?

 

[DonPhillips]

I am not familiar with statewide workmanship requirements but it appears I would condering beefing up the structure to reduce the deflection to 1/2" to avoid fingerpointing after the fact. The additional costs should be explained to the owner during design development and if the owner agrees, you can proceed with the heavier structure. If not, at least it was discussed, and the meeting notes should help deflect the finger from your way.

It is similar to tight floor standards in warehouses using high-bay equipment that require very flat floors. The customer's specifications should be what you are designing to if it exceeds the minimum code requirements.

Don Phillips

http://worthingtonengineering.com

 

[BEFORT]

Structural EIT, very good point. However, the owners don't notice the floors until after they move in and after the live load is placed. Thats when the inspection would be performed.

I do not understand your second question. Most of the deflection is due to the Live Load application.

In any event, if the GC has placed the beams and ledger at the correct elevation (+- the applicable tolerances) then there shouldn't be any issues. I will inform the Regestra that they are measuring deflections not constuction standards/tolerances.

In regards to Don Phillips, your comments are well taken. In the long run it would appear that neither the GC or myself have not followed our requirements but it may take a while to convince everyone.

I appreciate all your input. I'm a first time user and have been working for myself for 15 years so this is a great opportunity to bounce off ideas.

If anyone else has any input, please still respond.

 

[youngstructural]

I agree completely with StructuralEIT:

- How even a floor system has been achieved (ie: how flat the floor is) is to be measured unloaded.

- Floor deflection is to be measured loaded, and would be satisfied in your scenario (ie: 1" allowable, 5/8" acheived).

Floor evenness is a measure of the quality of the workmanship, particularly problematic in large area concrete floors. In this timber floor example, the contractor would be expected to check the joists, place them crown up, and reject any which are too warped (as per StructuralEIT's suggestion). By and large the floor deflection is a stiffness issue and will be a measure of whether or not the engineer has given proper consideration to serviceability (SLS) concerns.

So after all of that, I agree with StructuralEIT.

Regards,

YS


Have I missed something here?

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...