Sunday, November 20, 2011

RE: Concrete Floor Finishing

 
When we design a metal composite deck, we often assume simple span for the beams/purlins.  The beams at the girders assume simple span (generally).  But concrete will try to resist bending unless there is a definitive joint (usually there is not a defined joint in the slab).   Increase the reinforcing in the areas over the girders and over the beams where there will be negative bending (whether we have assumed it in the design or not). 
 
Keep in mind that the cracks will form early, but will not be wide enough to notice for many weeks.  The rate of shrinkage is not linear with time.  Cracks noticed in the first month will only get worse. 
 
See below.

Regards, Harold Sprague
 

From: h.d.richardson@shaw.ca
To: seaint@seaint.org
Subject: Concrete Floor Finishing
Date: Sun, 20 Nov 2011 17:18:49 -0700

Fellow engineers,
 
        I have an architect client specializing in high-end housing.  His client, the owner, is having a residence built that is intended to have an "industrial" look, namely exposed structure consisting of exposed metal deck supported on wide flange beams and topped with a 5" concrete slab.  The problem is finishing the floor.
 
        The 5" (from bottom of deck) is intended to contain in-floor heating (3/4" o.d. water pipes) plus any reinforcing required.  The owner wants to grind off and polish the floor surface.  About 1/16" (or 1mm if you prefer metric) is to be cut off for this purpose.  The desire is to have a perfectly finished floor; any cracks or other blemishes are exceedingly undesirable.
 
        The building is nearly closed in; the steel beams and 1.5" High-Bond" steel floor deck spanning 5 feet c/c are in place; the concrete will not be placed until after Christmas due to low temperature possibilities.  We have already had overnight temperatures down to -18 degrees C or 0 degrees F; the architect wants a fully heated building before any concrete floors are placed.
 
        Given that I can guarantee that the concrete WILL crack and that I can not even strongly suggest that cracking WILL NOT occur how can I best keep the owner happy?
Ideas that have already been discussed include:
1.) Add sufficient top steel over the beams (or over all) and calculate crack width.  Here I am open to suggestions as to what crack width I should accept.
Increaseing cover may increase the crack width.  I would try to minimize the cover.  If you use closer spaced reinforcing, you will minimize the apparent cracking.  I would suggest using at least the ACI minimum for temperature steel (as opposed to the SDI).  I would use WWR, and I would require bolsters.   
 
2.) Cast the floor in two pours creating a topping.  Increasing the overall thickness to achieve this end is possible.
This is a legitimate alternative, but cracks in the first placement tend to reflect in the second.  I would advise against it.
 
3.) We can add a shrinkage reducing agent (to one or both layers if we use the topping rout).
Shrinkage reducing agents require experience and provision for the slight expansion that will take place as the concrete sets. 
 
4.) The idea of providing shoring at midspan to ensure a shored composite concrete construction effect has been discussed
I am gussing that you are using a composite deck.  Using a heavier deck does not cost that much and will reduce the deflection at the deck midspan. 
 
5.) We could use fibre reinforcing, although I am inclined to reject this option, at least for the top layer because the grinding could expose the fibres with bad results.
No.  I would advise against it. 
 
6.) With an owner willing to spring (-+)$25.00 per square foot for grinding and polishing the floor I don't think cost is a serious object.
 
I would suggest the following:
A.  Develop a mix design with a uniformly graded aggregate (no gap graded aggregate).  Jim Shilstone may be able to offer more definitive advice.  Use a good polycarboxilate super plasticizer to minimize the water and minimize the total cement content.  Avoid using flyash.
 
B.  Use a relatively heavy WWR placed on bolsters so that the height of steel is accurate.
 
C.  Have the contractor use a Copperhead Laser screed so that the placement level will be accurate.  A 2x4 screed is not a good way of placing concrete.  The Copperhead can be placed on structural decks.  These can be rented. 
 
D.  Use a good stained concrete finish.  Bass Pro Shops uses stained concrete and the quality is very good.  I think they could use a better mix to limit the cracks, but it is very good. 
 
E.  Use a conveyor to place the concrete as opposed to an overhead pump.  If pushed, you can use a horizontal line pump. 
 
F.  Assemble a team meeting with the pumper, ready mix supplier, finisher, admixutre technical rep, concrete stain contractor, and anyone else involved.  Make sure everyone is aware of your performance requrements, and take suggestions seriously. 
 
G.  Carl Ytterberg (sp) came out with a series of articles in Concrete International a while back where he discussed the various aspects of concrete shrinkage and which compontent in the mix contributed the most to shrinkage.  The aggregates do not shrink.  The cement paste shrinks.  Therefor do what you need to reduce the amount of cement.  Uniformly graded aggregate will reduce the amount of cement required by reducing the total surface area of the aggregate. 
 
H.  The best way to eliminate joints is to use a post-tensioned slab.  But this will present another whole list of issues. 
 
        Thank you in advance for any suggestions you might make.
 
Regards,
 
H. Daryl Richardson