Product Knowledge

What Happens If an Undercut Anchor Is Installed Without Undercutting? Rework Criteria and a 3-Step Remedy

Views: 0Posted on: 2026-10-21

The conclusion first:If undercutting is forgotten, the anchor is reduced tolittle more than a threaded rod inserted into a straight hole— with no conical cavity for the expansion segments to engage, no mechanical lock is formed at all, and the load capacity is usually less than 20% of the design value with very high scatter. Its typical manifestation is not 'gradual loosening' butthe entire anchor pulls out during the pull-out test, and the expansion segments remain in their contracted state. Determination relies on four inspection points:whether the torque can reach the specified value, whether the exposed length deviates, whether pull-out values are dispersed, and whether the hole contains an undercut cavity. Remediation is classified into three levels: Level 1, re-expand in place; Level 2, relocate and re-drill; Level 3, report to the designer for strengthening.

Scope of application:This article addresses the identification and rework treatment of omitted hole-bottom undercutting during installation of undercut (mold-undercut) anchors, applicable to post-installed anchoring scenarios such as curtain walls, equipment foundations, and supports and hangers. Self-undercutting anchors complete undercutting and installation simultaneously, so there is no separate step to omit; their anomaly assessment differs from that described here.

Installation of an undercut anchor involves two separate steps: first drill a straight hole, then use aspecial undercutting drill bitto cut a reverse-cone cavity at the bottom of the hole, and finally insert the anchor so that the expansion segments open and engage in the cavity. The second step is separate and requires changing tools, which makes itthe step most easily skipped in the entire installation process— especially when the schedule is tight, crews change, or undercut anchors are treated like ordinary expansion anchors.

The problem is that if this step is skipped, the anchor can still be 'installed' and hammered fully home, with almost no visible abnormality. The issue only surfaces at pull-out testing or even later. This article breaks down 'how to detect it' and 'how to handle it once detected' into actionable steps.

I. How much load capacity remains when undercutting is forgotten

1.1 The undercut cavity is the sole load-bearing fulcrum

The load capacity of an undercut anchor comes from a single source:the expansion segments open and wedge against the inclined surface of the reverse-cone cavity at the hole bottom, forming a mechanical lock that transfers tensile load to the concrete through this lock. It relies neither on radial expansion stress (the mechanism of ordinary expansion anchors) nor on adhesive bonding (the mechanism of chemical anchors). Throughout the entire load path, the undercut cavity is the only fulcrum.

Therefore, once no cavity has been cut at the hole bottom, when the expansion segments openthere is nothing for them to bite into, and they can only press against the straight hole wall. At this point, the anchor's 'anchoring force' is reduced to hole-wall friction and the frictional resistance generated by the normal force of the expansion segments against the wall — an order of magnitude completely different from the design value, highly dependent on hole-wall smoothness, concrete surface condition, and hammering force during installation, and virtually impossible to calculate.

1.2 On-site differences between proper undercutting and omitted undercutting

Observation itemProper undercuttingOmitted undercutting
Hammering feel during installationA distinct 'biting' jolt when the expansion segments open, followed by sinking to the specified depthSmooth hammering all the way to the bottom, or premature bottoming out that cannot be driven further, with no sense of engagement
Torque behaviorReaches the specified torque and holds steadily, with clear feedback from the wrenchNever builds up torque, or suddenly spins freely, with the torque value failing to rise
Protruding lengthMeets the design and matches adjacent anchors in the same batchToo long (not fully seated) or too short (bottomed out), inconsistent with adjacent anchors
Pull-out behaviorReaches the design value, with steel failure or concrete cone failure as the failure modeFull pull-out of the anchor, with the expansion segments still contracted and virtually no cone failure on the hole wall
Relative load capacity100% (per certified design value)Empirically, usuallyless than 20%, with very high dispersion

The last row in the table is an empirical judgment, not a code value—because a missing undercut is itself a non-design state with no corresponding calculation model. Its purpose is to help you build an intuitive sense of magnitude:This is not a matter of being off by a little—it is an order-of-magnitude difference.。

2. On-site determination: four actionable inspection points

2.1 Torque re-check—the fastest preliminary screening method

Re-tighten with a calibrated torque wrench to the specified torque. An anchor with a properly formed undercut, after reaching the torque value,will hold firmly and stop rotating; an anchor with a missing undercut will either keep turning along with the wrench (with no fulcrum, the threads carry the entire anchor around in the hole) or suddenly spin free and slip. Both are clear warning signs, and the test costs very little, making it suitable for 100% inspection.

One caveat: torque re-checking can only screen out anchors that are clearly not keyed in; it cannot prove the reverse—that an anchor is definitely keyed in. An anchor with adequate torque but an incomplete undercut cavity may still hold the torque by friction. So it is a preliminary screen, not a final verdict.

2.2 Exposed-length comparison—the most intuitive corroborating evidence

Anchors from the same batch, same specification, and same base material should have essentially consistent exposed lengths after proper installation. Spot-measure an area with a ruler or vernier caliper,and pick out those that are noticeably too long or too short. A length that is too long usually means the anchor was not seated deep enough (the expansion segments did not bite, and the anchor is hanging in the hole), while one that is too short often means it bottomed out against the end of the hole. This method needs no instruments, but it does require comparison samples, so compare laterally within the same area rather than looking at absolute values alone.

2.3 Dispersion in pull-out sampling—the most direct evidence

Pull-out testing is the most direct criterion, but its value lies not only in whether a single value passes but also inthe degree of dispersion within the same batch of data. For anchors installed under normal conditions, pull-out values from the same batch should cluster within a narrow range; a distribution with a few very high and a few very low values usually signals a consistency problem in installation, and a missing undercut is one common cause.

Even more important is the failure mode. Failure that reaches the design value and occurs in the steel or in the concrete cone is normal;an anchor that pulls out entirely before reaching the design value, with the expansion segments still contracted after extraction,can essentially be judged directly as a missing undercut. Such failures are brittle and give no warning.

2.4 In-hole inspection—the step that most needs to be done before installation

Using a borescope or a dedicated hole gauge, check,before installation,whether the hole bottom has an undercut cavity, whether the cavity is complete, and whether the hole depth is sufficient. This is the only inspection point that can stop a problem before it is installed wrong, at a cost far below rework afterward. For critical locations, oversized panels, and high-rise curtain walls, it is recommended to include in-hole inspection at the hole bottom as a mandatory item in concealed-work acceptance, with spot checks at a set ratio and image records kept.

Inspection pointTimingReliabilityLimitation
Torque re-checkAfter installationMediumCan only screen out obvious anomalies; cannot prove compliance
Exposed-length comparisonAfter installationMediumOnly meaningful with lateral comparison within the same batch and type
Pull-out samplingAfter installationHighIt is sampling-based and cannot cover every anchor
In-hole inspectionBefore installationHighRequires tools such as a borescope/hole gauge; inspecting each hole is time-consuming

3. Rework classification: determine the level first, then decide the action

A missing undercut occurs in two situations: one isbefore the connector has been installed and no load has been applied; the other iswhen adapters have already been installed and the curtain wall panels are hung. The remediation cost of the two differs by an order of magnitude, so classify before taking action—do not simply tear everything out from the start.

LevelDetermination criteriaRemediation approach
Level 1Installed but with no connection element fitted and no load applied; hole wall intact, no edge spalling, and hole depth sufficientWithdraw in place → re-cut the hole bottom → reinstall with a new anchor
Level 2Connection element fitted or load applied, but hole wall intact and space available nearby for relocationRelocate and re-drill (original hole abandoned and sealed), or a strengthening scheme approved by the designer
Level 3Edge spalling of the hole wall, cracking of the base material, insufficient hole depth, no space for relocation nearby, or proximity to rebarStop work, report to the designer, and use reinforcing steel plates, a larger size, or a design change

3.1 Step 1: Stop work and isolate and number the anchors

As soon as the first omitted undercutting is found, assume thatthis is a batch-wide problem, not an isolated mistake— because a missed step is usually caused by the same crew, the same time period, and the same tool. Three actions are required: immediately stop further installation in the area, number each installed anchor individually, and perform torque verification plus a protruding-length survey on all anchors in the batch, flagging any abnormalities.

This step cannot be skipped. Many projects go wrong precisely because 'only the anchor picked for pull-out testing was addressed'; the remaining missed anchors start causing problems two years later, when rework means dismantling the curtain wall.

3.2 Step 2: Re-cutting in place (only if conditions permit)

Level 1 cases can be handled in place: first withdraw the anchor with a special tool, then use anundercutting drill bitto re-cut a complete reverse-cone cavity at the bottom of the original hole, clean the hole thoroughly, and install anewanchor. Three prerequisites must be met simultaneously — the hole wall has not been damaged by withdrawing the anchor, the hole depth still meets the requirement, and the undercutting drill bit can reach the hole bottom smoothly.

A common mistake here isreinstalling the withdrawn old anchor. The expansion segments were already opened and plastically deformed during the first installation, and the cutting edges may have dulled or chipped; a second installation cannot ensure proper opening, let alone proper engagement with the cavity. The old anchor must be scrapped.

3.3 Step 3: Relocate and re-drill, or change the system

Level 2 and 3 cases generally cannot be re-cut in place and require relocation and re-drilling. Relocation has two hard constraints: first,edge distance and spacing, the distances from the new hole to the member edge and to adjacent anchors must satisfy product parameters and code requirements, and holes must not be crowded together merely to avoid the original hole; second,existing rebar, the relocated hole must still avoid existing structural rebar, so detection before drilling is required. The original hole must be filled solid with non-shrink mortar or a special sealing material; empty holes must not be left.

Remedial approachApplicable conditionsAdvantagesLimitations
Re-cutting in placeLevel 1, hole wall intact and hole depth sufficientPosition unchanged, no impact on the completed layoutOld anchor must be scrapped; withdrawal may damage the hole wall
Relocation and re-drillingLevel 2, with room for relocation and acceptable edge distances and spacingsHighest reliability, equivalent to redoing the installationConstrained by edge distances, spacings, and rebar; original hole must be sealed
Increasing the sizeHole position cannot be moved and edge distance still meets the requirements for a larger sizeCan re-drill at a larger diameter in the original positionMust be verified by the designer for edge distance and base material thickness; not to be decided on site
Strengthening / design changeLevel 3, base material already damaged or no space availableResolves problems at the base material levelRequires a design solution, with the greatest impact on schedule and cost

IV. Five actions that must never be taken

4.1 Increasing torque to 'tighten it a bit more'

Torque serves to open the expansion segments, provided there is a cavity at the hole bottom for them to engage. Without a cavity, increasing torque has only two outcomes: it cannot be turned (bottoming out) or it spins freely (rotating the entire anchor). Neither produces anchoring force, and both wear the hole wall loose and damage the threads and sleeve, making subsequent in-place re-cutting more difficult.

4.2 Injecting chemical adhesive into the hole to 'bond' it

Combining a mechanical anchor with chemical adhesive has no certification basis or design model to support it. A more practical problem is that the anchor is already in the hole, so the adhesive cannot be fully injected and the gap between hole wall and threaded rod cannot be guaranteed; with the expansion segments unopened, the thin annular layer formed after curing becomes a weak plane instead. This practice is common on site but amounts to 'appearing to fix it while actually not fixing it' and must be prohibited.

4.3 Addressing only the anchor picked for testing

Omitted undercutting is a process defect caused by 'the whole batch was not done,' not by 'this one was forgotten.' Fixing only the sampled anchor leaves the hazard in the structure. The correct approach is to verify torque and protruding length for every anchor in the batch and todouble the sampling forpull-out testing.

4.4 Reinstalling with old anchors

The expansion segments have been plastically deformed and the cutting edges may be chipped, so a second installation cannot guarantee opening travel or engagement. Rework must use new anchors; there are no exceptions and no room for a 'looks fine' judgment.

4.5 Relying on subsequent pull-out testing to 'verify there is no problem'

Pull-out testing is sampling with limited coverage. Three passing anchors in an area do not prove the two-hundredth also passes, especially in cases of omitted undercutting where 'some were missed and some were not.' Pull-out testing can detect problems but cannot replace process control; the real line of defense lies before and during installation.

V. How to eliminate the problem through process design: three practical control points

This type of problem cannot be solved by 'reminding workers to pay attention'; it requires process design. The three control points are low in cost but direct in effect:

Control pointSpecific measureWhy it works
Separating the stepsAssign undercutting as a separate step with separate acceptance; do not combine it with drilling for measurementWhen combined, undercutting is easily treated as a 'quick extra step' and omitted
Distinguishing the toolsColor-code undercutting and straight-hole drill bits and store them in separate boxes; the undercutting bit should have a depth scaleOn site, it is immediately apparent 'whether the undercutting bit has been used'
Pre-installation inspectionSample the hole-bottom cavity with a hole gauge/borescope at a set ratio and keep image records as part of concealed-work acceptanceThe only checkpoint that can stop the problem before incorrect installation

VI. Daring's products and supporting tools

Daring supplies undercutting drill bits, installation sleeves, hole gauges, and torque wrenches to accompany its undercut anchors. Parameters and certifications are as follows:

ItemParameter
Product categoryUndercut (mold-undercut) mechanical anchors
Size rangeCommonly up to M16; for larger sizes the self-undercutting series is recommended (up to M20)
MaterialGrade 5.8 carbon steel, grade 8.8 carbon steel, 304 stainless steel, 316 stainless steel
Surface treatmentElectroplated zinc, hot-dip galvanizing (≥50μm)
Applicable base materialCracked/uncracked concrete C20–C60, with no radial expansion stress
Certification and testingEU ETA certification, Class S anchor certification, seismic test under a rare earthquake of intensity 9, 2 million cycle fatigue test, salt spray test
Supporting toolsSpecial undercutting drill bit, installation sleeve, torque wrench, hole gauge, borescope

For rework scenarios, Daring can provide on-site technical assessment support: based on torque verification, protruding-length surveys, and pull-out sampling, we help classify the rework level and recommend either in-place re-cutting or relocation and re-drilling. Daring also supplies inspection tools such as hole gauges, hole probes, and borescopes, moving inspection of the hole-bottom cavity to before installation.

VII. Frequently asked questions

Q1If undercutting was forgotten but hammering felt like it reached the bottom, will there be a problem?
Yes. 'Hammering to the bottom' only means the anchor touched the bottom of the hole, not that the expansion segments engaged in a cavity. Without an undercut cavity, the opened expansion segments can only press against the straight hole wall, bearing load by friction, and the load capacity is usually less than 20% of the design value. Judgment should be based on whether the torque reaches its target stably, whether the protruding length matches the rest of the batch, and whether the pull-out failure mode is full anchor pull-out.
Q2Can chemical adhesive be injected directly into the hole without withdrawing the anchor?
Not recommended, and it should not be treated as a routine remedy. Combining mechanical anchors with chemical adhesive lacks certification basis and a design model; the anchor is already in the hole, so the adhesive cannot be fully injected, and the thin annular adhesive layer formed with the expansion segments unopened is a weak plane instead. If a change of load-bearing system is truly required, it must be submitted to the designer for confirmation and handled as a design change.
Q3Can omitted undercutting be compensated by increasing torque?
No. Torque drives the expansion segments open, provided there is a cavity at the hole bottom to engage. Without a cavity, the anchor either cannot be turned or spins freely as a whole; neither produces anchoring force, and both wear the hole wall loose and damage the threads, making later in-place re-cutting more difficult.
Q4If several hundred anchors have been installed, is testing only a few sufficient?
No. Omitted undercutting is a process defect, usually a batch problem involving the same crew and time period. Work in the area should stop, torque verification and a protruding-length survey should be performed on every anchor in the batch, and pull-out sampling for that batch should then be doubled; it is not enough to address only the sampled anchor.
Q5After in-place re-cutting, can the original anchor be reused?
No. The expansion segments were plastically deformed during the first installation and the cutting edges may have dulled or chipped, so a second installation cannot guarantee opening travel or engagement with the cavity. Rework must use a new anchor and the old one must be scrapped.
Q6How can omitted undercutting be prevented through process design?
Three control points: first, assign undercutting as a separate step with separate acceptance, not combined with drilling for measurement; second, color-code and separate the undercutting and straight-hole drill bits and fit a depth scale so that use is immediately apparent on site; third, before installation, sample the hole-bottom cavity with a hole gauge or borescope at a set ratio and include image records in concealed-work acceptance.

Returning to the question in the title: what happens if undercutting is forgotten? The answer isthe anchor degrades from 'mechanical lock load bearing' to 'pure friction,' losing an order of magnitude of load capacity, with brittle failure and no warning. What makes it formidable is not difficulty but concealment — it can be installed, hammered home, and look normal, with only pull-out testing revealing the problem. The response logic is therefore to classify first and act second: Level 1, re-cut in place and replace with a new anchor; Level 2, relocate and re-drill; Level 3, report to the designer for strengthening. The real solution lies in process design: make undercutting a separate step, make the drill bits identifiable, and move hole-bottom inspection to before installation. Daring undercut anchors are supplied with special undercutting drill bits, hole gauges, and borescopes, and come with ETA certification, Class S certification, and seismic test reports for a rare earthquake of intensity 9. For on-site rework assessment support, please consult the Daring technical team.

References

JGJ 145-2013 Technical Specification for Post-Installed Anchoring in Concrete Structures — requirements for anchor reaming, installation, and inspection

JG/T 160-2017 Mechanical Anchors for Concrete — classification, performance classes, and test methods for undercut anchors

GB 50367-2013 Design Code for Strengthening Concrete Structures — detailing requirements for post-installed anchoring in load-bearing structures

GB 50550-2010 Code for Acceptance of Construction Quality of Structural Strengthening Works — on-site inspection of anchoring quality

Daring website — Reaming: the key to undercut anchor installation

Daring website — The five most common mistakes in undercut anchor installation: no reaming, no hole cleaning, and incorrect torque

Daring website — How to choose installation tools for undercut anchors? Differences between special undercutting bits and ordinary drill bits

Daring website — What to do if the drilling depth for undercut anchors is insufficient? Graded solutions and the relocation and re-drilling process

Daring website — An in-depth explanation of the undercutting principle of undercut anchors and the mechanism for achieving zero expansion stress

Daring website — How to perform an on-site pull-out test on undercut anchors? Sampling frequency and acceptance criteria explained

Daring website — How is the effective embedment depth of undercut mechanical anchors determined? Understanding the code requirements

Daring website — What happens if installation torque on mechanical anchors is too high? Common installation mistakes and their consequences

Back to Daring Home  Back to Top