Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: Testing Epoxied Anchors

[Subject Prev][Subject Next][Thread Prev][Thread Next]

I thought I had sent this answer before I left the list (I was traveling)
but found that it somehow didn't get through when I returned to the list
this week.  If your situation has resolved itself, please disregard the

Your contractor has put you in a tough spot.  This is an issue that has no
easy answer.  Since the anchors were not inspected at the time of
installation, you do not know whether the embedment depth was maintained,
whether the holes were cleaned, etc.  Proof loading after the fact can only
determine whether you have a certain minimum level of bond strength; it
cannot verify the anchor for the design load unless a) you accept a lower
factor of safety, or b) you test a certain number of the anchors to failure.

Typically, the value that you are using for design represents a mean
ultimate number divided by 4.  This is based on tests in a lab, far from an
edge, i.e., without edge distance or spacing reduction.  When you test an
anchor in the field, you must consider what the ultimate capacity of each
particular anchor would be given it's proximity to edges or other loaded
anchors.  Then again, you must consider whether the level of proof loading
that you are applying will damage the anchorage.  Microcracking and bond
deterioration will begin well before the ultimate load is reached, and may
not be visible at the surface.

I assume testing to failure is not an option.  Therefore, what you are left
with is proof loading to a conservative value, and dividing your proof load
value by a reasonable factor of safety to arrive at your new, reduced design
load.  You could decide that a FS = 4 is no longer appropriate since you
have data relating to your specific concrete, installation conditions, etc.
I wouldn't go below 2.5, which, if you proof load to 1.6 x the allowable,
would result in a verified capacity equal to about 60% of what you designed
for.  Again, remember that if you test an anchor in a corner condition, say,
to 1.6 x the allowable without edge or spacing effects, you could end up
damaging or even failing the anchorage, so the proof load must be reduced
according to where the anchor is located.  Increasing the allowable by 1.33
is just another way of accepting a lower FS (in this case, 3).  1/3
increases for earthquake loading are one of the oddities of the code,
especially when applied to anchors.

I presume that 1.6 comes from twice the OSHPD allowable of 80% of ICBO's
numbers, not a very scientific way to proceed.  You might want to see if the
manufacturer of your anchor system has data that might support a higher
number for a safe proof load, e.g., given that you will be performing a
confined test (no support spacing).

That's my take.  I'd be interested to see if someone else has a less
draconian approach to offer.


J. Silva, SE
Hilti, Inc.

> ----------
> From: 	Larry Hauer[SMTP:LRHauer(--nospam--at)]
> Reply To: 	seaint(--nospam--at)
> Sent: 	Tuesday, November 16, 1999 8:05 AM
> To: 	INTERNET:seaint(--nospam--at)
> Subject: 	Testing Epoxied Anchors
> I have a project where the contractor failed to follow the plans and use
> special inspection for the epoxied anchor bolts. Can someone refer me to,
> or give me the procedure for testing the anchors. So far, I have: Test 50%
> to 1.6 times the catalogue allowable tensioin value, if any anchor fails
> test all. Does this seem reaonable or meet some code requirement? Also,
> whatever value I test them to, if they are used for seismic purposes,
> should I also multiply by 1.33?
> Thanks in Advance,
> Larry Hauer S.E.