Playing an intrinsic part in the construction industry, concrete anchors – whether they are cast-in-place or post-installed – enable the connection of a steel member to a concrete substrate and similar to construction products such as concrete and steel, need regulation, assessment, and design methods for them to be used in any serious form by engineers. This article focuses specifically on post-installed bonded and mechanical anchors.

When the Eurocodes were first developed, the design of cast-in-place anchors and anchor channels, as well as post-installed anchors, was not established well enough to warrant inclusion within the EN design standards. Under the European framework, the EN design standards are developed by a key organisation, CEN (European Committee for Standardisation), with the other being EOTA  (European Organisation for Technical Assessment), which develops European Assessment Documents (EADs) that define the criteria to assess the performance of various products. EOTA  was responsible for the development of the first assessment and design provisions for post-installed anchors: ETAG 001. In particular, Annex C of ETAG 001 dealt with the design under static and quasi-static loading conditions for mechanical anchors, with Annex E added later to include assessment of these anchors under seismic conditions. Later, EOTA published two vital documents that included design provisions for bonded anchors under static / quasi-static loading (TR 029) and under seismic loading for both bonded and mechanical anchor types (TR 045).

Technical Differences and Similarities

For the benefit of the design engineer, all these design guidelines are now merged under the new design document EN 1992-4, which is now an enforceable design standard in countries under the EN umbrella, with adoption nearly complete. The assessment aspect of each anchor remains unchanged under the EADs and must be conducted in independent and accredited Technical Assessment Bodies (TABs) such as CSTB, DIBt, among others. The ETAs (European Technical Assessments) published by these bodies is then used to design with EN 1992-4, which is now referenced in the EADs. An ETA is an alternative for construction products not covered by a harmonised performance standard and provides information on a product’s performance based on assessment criteria outlined in a common EADs for a specific application, such as connecting a steel member to a concrete substrate.

However, this does not imply that the design philosophy has changed. At its core, EN 1992-4 carries over many of the design provisions of ETAG 001: Annex C, TR 029, and TR 045 and, as such, differences are limited.

Design under various loading conditions

The design to EN 1992-4 also unifies the design provisions of post-installed anchors to various loading conditions such as static, seismic, fatigue, and fire, which was previously spread across various individual Technical Reports (TRs). These Reports will still exist to form the basis of further inclusions and updates, however, EN 1992-4 is now considered as “state-of-the-art” for design purposes.

Baseplate & fastener configurations

Similar to ETAG 001, the design provisions of EN 1992-4 still use the fastener design theory while assuming a perfectly rigid baseplate, with the same partial safety factor concept and derivation of forces on the anchors applicable as before.

The range of fastener configurations remains the same as before, with seven configurations applicable to anchors far from the edge () and four configurations for anchors close to an edge.

Concrete Strength Classes & Condition

Anchor qualification from C20/25–C50/60 in ETAG 001 to C12/15–C90/105 in the EAD 330232 (mechanical) & 330499 (bonded) means that the anchors can now be tested and designed in a wider concrete strength class. To harmonise with the rest of EN 1992, design provisions of EN 1992-4 now use the cylinder strength of concrete rather than the cube strength used in ETAG 001: Annex C and TR 029.

One key introduction in Section 4.7 of EN 1992-4 refers to determining whether the concrete’s condition is “cracked” or “uncracked”, with the standard recommending designers to consider the state of concrete as “cracked” for the duration of its service life. However, engineers also may assume the concrete as “uncracked” if the anchor is located in uncracked concrete along its entire anchorage depth under a characteristic load combination at serviceability conditions. Moreover, this also needs a detailed stress analysis to be conducted.

Differences to Main Characteristic Resistances

Concrete Cone

As a consequence of the change to the cylinder strength being used in EN 1992-4, higher k factors are now used to determine the characteristic resistances in concrete cone, pull-out, and splitting in tension, as well as pry-out and edge failure in shear. However, this does not fully compensate for the lower input of the concrete’s strength class and as an example of the largest reduction, the characteristic resistance () of the concrete cone failure in cracked concrete of class C35/45, for an anchor embedded at an effective embedment (hef) of 100mm, sees a decrease of 5.7% when designing to EN 1992-4. Here, k1 is 7.7 and 11.0 for cracked and uncracked concrete, respectively, where previously this value was 7.2 and 10.1.

This also changes the equation to calculate the group effect of closely spaced bonded anchors represented by , where the expression is also modified by adding  to the denominator like so:

EOTA TR029

EN 1992-4

Where

Bending Moment

Under Section 7.2.1.4 (7) of EN 1992-4, if the baseplate is subject to a bending moment, an additional factor is considered for the equation for the resistance of concrete cone failure in tension, which accounts for the compression force CEd resulting from bending. Conservatively,  may be considered as 1.0.

Sustained Load

The biggest addition to EN 1992-4 is of the sustained load factor, , which quantifies the reduction due to sustained loads. Failures in the past due to the effects of creep and sustained loads led to large safety concerns over bonded anchors subjected to these loads.  is now included in an individual bonded anchor’s ETA and combines with the αsus value – ratio of the sustained actions to the total actions at the Ultimate Limit State – to provide the sustained load reduction factor. If an anchor’s ETA does not mention a value of , the default recommendation is 0.6, meaning that maximum reduction in the tensile capacity in this failure mode is 40%. Under ETAG 001 Part 5, creep behaviour was included in the qualification criteria and was a pass or fail sustained load test and was included in the ETA; however, there was effect on the design resistance of the bonded fastener.

Supplementary Reinforcement

The consideration of supplementary reinforcement is new to EN 1992-4 and was not previously a consideration in ETAG 001 Annex C and EOTA TR029. Supplementary reinforcement can be designed through Section 7.2.1.9 for tension and 7.2.2.6 for shear and does not require verification of the concrete cone and edge failure. However, there is a 25% reduction to the pry-out resistance under Section 7.2.2.4 (2) if this reinforcement is used. If taking advantage of the supplementary reinforcement, there is an additional check for combined tension & shear, in addition to those of concrete carried over from the previous guidelines. With this, engineers may plan for any potential post-installed fastening or use any margin in the existing reinforcement for unplanned fastenings, provided the reinforcement detailing in Figure 7.2 (for tension) and Figure 7.10 (for shear) are adhered to, reproduced below.

Conditions to omit splitting failure & changes to Ψh,sp

The splitting failure resistance mode, NRk,sp, sees changes to the factor Ψh,sp that takes into account the influence of the member thickness on the splitting resistance.

Steel failure in shear without lever arm

Wherever grout is required under the fixture, a larger fastener diameter is required due to reduction in the shear resistance and, in certain cases, to provide a larger resistance for steel where the baseplate requires grouting. Under ETAG 001 Annex C & TR 029, shear loads act on an anchor without a lever arm if:

1. The baseplate is metallic, and the anchorage area is fixed directly to the concrete substrate without an intermediate layer, or with a levelling layer of mortar of a compressive strength higher than 30 MPa, all while having a thickness less than or equal to half the anchorage diameter.
2. The baseplate is in contact with the anchor over its entire thickness.

This carries over in EN 1992-4, which lists additional scenarios that satisfy the condition “without lever arm” in uncracked concrete:

1. At least two anchors in group;
2. No tension/moment on the baseplate;
3. Fastener spacing in the direction of the shear force exceeds 10d and in both directions if inclined shear forces are present
4. Grout thickness is no thicker than 40mm and 5dnom
5. The grout’s strength is not less than 30 MPa and is applied to a rough concrete surface

 Concrete Edge Failure (Section 7.2.2.5)

Concrete edge breakout sees the biggest changes under the new EN 1992-4 with several factors impacted. With reference to the image above, omitting edge failure checks in both directions is no longer possible in EN 1992-4 as was the case previously in ETAG 001 Annex C when the edge distances c1 & c2 are the larger of 10hef or 60dnom; however, the scope of this verification in EN 1992-4 is limited to:

1. C1 > max (10hef, 60𝑑);
2. C1 ≤ min (10hef, 60𝑑), without lever arm, and the thickness of the baseplate is larger than 0.25hef

If shear acts away from one edge, the factor  accounting for the angle αV between the applied load, VSd, and the direction perpendicular to the free edge has also changed. Under ETAG 001 Annex C, when αV > 90°, only the component of shear load parallel to the edge impacts the anchor and the component acting away from the edge can be ignored for edge failure verification. This statement is no longer present in EN 1992-4 and edge failure parallel to the shear direction also can occur. Moreover, the factor  now leads to a lower value of concrete edge resistance when compared to ETAG 001 Annex C:

The largest impact comes from the length of the fastener acting in shear: lf. In EN 1992-4, this now dictates the value of the basic edge resistance rather than hef and lf is limited to 12dnom for dnom ≤ 24mm and the larger of 8dnom or 300mm for dnom > 24mm. In the ETA for bonded fastener, lf ≠ hef, so the actual embedment for a fastener loaded in shear is limited to 12dnom. This limit stems from the fact that most tests of bonded fasteners in shear are conducted at 12dnom, as referenced in the fib Bulletin 58.

The factor accounting for edge reinforcement (not to be confused with supplementary reinforcement) in ETAG 001 Annex C previously accounted for three conditions of edge reinforcement: (1) No edge reinforcement where ; (2) Straight edge reinforcement where ; and (3) Dense edge reinforcement . In EN 1992-4, the second condition has been omitted.

To conclude, EN 1992-4 now harmonises with the rest of the Eurocode series and represents the new state-of-the-art standard, unifying in one document the design of cast-in anchors and anchor channels, as well as for post-installed bonded and mechanical anchors, under various loading conditions such as static, seismic, fatigue, and fire. The few major technical changes mentioned above do not render previous designs to ETAG 001 Annex C obsolete. As with any new design standard, engineers should not worry about their previous fastenings if they were designed correctly. The change from a “guideline” to a “standard” represents a positive shift towards greater awareness, acceptance, and obligation of anchor design.