If the bolts were designed with plates on the bottom and you didn't put them in, you have a real problem.  A lot of capacity comes from the plate on the bottom.  Get with the structural engineer of record to devise a fix for this.  If you are tight on space at the base plate, this is one where you might be tearing out concrete, since you cannot achieve much resistance to uplift with just embedment of a bolt that was designed for a plate at the bottom.  Since you mention a Special Inspector, can I assume you are in Florida?  If so, this will prevent you from obtaining a Certificate of Occupancy, if not resolve to the satisfaction of the Engineer of Record.

Good luck!

Commentary in the AISC manual (LRFD Vol II, page 8-92)indicates that the size of the "head" at the end of the bolt isn't that critical.  In fact, it states that "Marsh and Burdette (1985) showed that the head of the anchor rod usually provides sufficient anchorage and the use of an additional washer or plate does not add significantly to the anchorage."

Take for instance, a typical headed anchor.  It has only a nominal head (no big plate washer) but it is enough to engage the concrete and develop the shear cone mechanism that anchor bolt capcities are based on.  If there is at least a nut or head at the end, there should be enough to provide your capacity, assuming the embedment and other checks have been performed.  If you use a nut, it should be tack welded to avoid turning of the bolt during tightening operations above.

You might also try AISC technical personnel if you need further substantiation.

Jae...I agree.  Am making the assumption of nothing at end of bolt, leaving only a development length capacity from relatively smooth (though threaded) rod.  ANYTHING at the bottom is help, but NOTHING is a concern.

JAE is correct that a standard head or nut will develop the concrete cone adequately.  Furthermore, there is some evidence that providing large plates at the ends of anchor rods may actually decrease the pullout capacity of anchor groups.  This occurs because the plates spread out the failure surface so that instead of individual cones around each anchor, the cones intersect and the concrete failure surface decreases.  Basically, the plates form a plane of weakness in the concrete.

Make sure that there is in fact a head or nut at the end of the rod, though.  If not, you will be drilling or chipping.

Options include:

1. Check the calculation for any redundancy.  Can the design still work?

2. Test the anchors to something like 1.5 x working load.

3. Can you drill in longer anchor bars or extra anchors bars?

4. Breaking out?  The last resort.

nb

I may take some heat for this but...In reply to Newbuilder's comments I would hope that checking the design for the as-built condition/redundancy is the last resort. There is plenty of time for the contractor/engineer to propose and employ an appropriate fix.  A fix such that the component behaves as it was designed to.

Q..well said.

To continually compensate for contractor error by re-analysis and compromise in an effort of expedience and cooperation implies that our original analyses are flawed and that we just use the premise "when in doubt, make it stout!".  

We all realize that construction is difficult and sometimes screw-ups occur, but the responsibility lies with the contractor and the liability and pressure should not then be reverted to the designer.

Ron, Qshake,

Speaking as a person who has been on all sides of the fence I believe
that completing the Client's project as a team is the main objective.  It portrays
a professional image to our employers.

Engineers should accept that mistakes occur both in the design office and
on the site.  A reputable contractor should be able to assist the Engineer
by offering solutions to construction errors.

A viable solution is to review the design and should not be dismissed
if a solution is found in a short time.  Designers will know whether their
designs can take a little more etc.  I know a few occasions
where this has helped the designer escape from awkward situations.

I am sure that all parties acknowledge their contractual liabilities and the Client will
be even more assured if the team he has employed has worked through their problems together.

I strongly dislike the "it's your problem attitude" which goes against the
modern partnering concept.  The old adversarial system is for dinosaurs.

Let's use our brains before reaching for our Contracts chaps!

NB

Newbuilder, I have stated that the contractor/engineer should resolve this issue to the satisfaction of all involved.  (See my post)  I do however, take issue with your statement that a design (one story or ten) should be reviewed to determine the redundancy of that particular detail.  It is a rare occasion where the engineer can re-analyze a design more economically or efficiently than a "fix" can be resolved between contractor and engineer.  Thus I firmly believe going back to the design should be a last resort.  

In some cases, going back to the original design to analyze the effects of the modified structure is unavoidable (your spread footing is 6' lower than plan elevation as rock was not where it was anticipated, so several columns are 6 feet longer etc).

I am not against partnering nor do I believe that the contractor alone should bear the sole responsibility of originating the appropriate fix but I do beleive that the problem must be approached with good sense.

Sorry to disagree with Qshaker but going back to the design is the first thing to be done. When I design steel structures I always rationalise holding down bolt sizes - I only use 1 diameter and length of holding down bolt on a contract - and rationalise baseplate sizes to mimimise the numbers of templates required. This reduces the risks of the wrong bolt size or pattern being set into the wrong foundation.
It is so easy for the engineer in a nice warm dry office with Autocad to draw 50 different baseplates with different diameters and lengths of bolts in each, but a different matter on a cold wet building site to put the right set in the right base.
With the steelwork manufacturing methods employed in the UK it is no more expensive to make 50 baseplates identical ( even if 49 are bigger than really required ) than to make 50 different. This is because labour is expensive and steel comparatively cheap - the cost of the extra steel in making all the baseplates the same is outweighed by the supervision and other labour costs in matching different sized baseplates to different columns in the fabrication shop and similar costs on site.

In practice this means that there is a degree of redundancy which can be "dipped into" when problems occurr on site. Back check the design first, identify the areas of concern and design an appropriate solution.

If the steelwork erector puts some temporary "X" bracing in the frame then erection can proceed ignoring the base fixity of the columns whilst a solution to the problem is devised - time is money and delays in getting the main frame up can be very expensive.

In situations like this I have retro-fitted resin anchor holding down bolts using high tensile allthread bar. Use a mix-in-the-nozzle resin with low shrinkage and some moisture tolerance whilst liquid and proof load-test every one after curing. I do this after the main frame has been lined, levelled, plumbed and the baseplates grouted. This gives the foundation concrete time to cure and reach its full strength.

Mike, I don't argue that redundancy should be built into the details or that redundancy is there perhaps because of the inherent simplicity of that particular design.  And I will be the first to let anyone know that when it comes to practical building issues I don't optimize such that each beam or base plate is sized for the load right out of the computer.

But when it comes to the issue of sorting through all of that information a year or two later with all of the correspondance, legalities, and proper protocal etc., it is, in my opinion, to provide a "fix" to the satisfaction of the contractor, engineer, and owner.

Regarding your commentary, let me point out that while it one thing to acknowledge that mistakes happen in the field, it is quite another to make excuses for them.  Contractors and or inspectors are professionals just like us, and are expected to read the plans and build/insure that what has been detailed is built properly.

grr...

I recommend  pull out tests, if the results are adequate to take the loads (plus load factors) what's the problem?