+91 98335 30200  ·  sales@tcreng.com

Insights · construction

Pull-off, not pull-out: what an ASTM D4541 test on plaster and punning actually measures

2026-11-21 · 10 min read · By TCR Newsroom

Macro of a plastered wall with a smooth punning skim coat and three small circular pull-off test locations, two showing the plaster beneath where the bonded dolly came away and one with the steel dolly still in place

Article

TCR Engineering carries out pull-off adhesion testing to ASTM D4541 on plaster, punning, mortar and other finishing products, both at its Mahape laboratory in Navi Mumbai and on walls at site. In the laboratory the method is on NABL certificate NABLT0726MH18640; at site it is not. One recent enquiry called it a pull-out test, which is a different test.

  • The test asked for: a builder wanted a "pull out test" for punning on an ALC panel and on RCC, and a "pull off test" for cement plaster, one point each on an RCC column, an ALC panel and cement plaster.
  • The test quoted: a pull-off test to ASTM D4541 at site for all three, with one point taken as three dollies.
  • Pull-off versus pull-out: a pull-off test pulls a bonded dolly off a finish; a pull-out test pulls an embedded bar or anchor out of the material around it.
  • On the NABL scope: ASTM D4541 in the laboratory, for finishing products, gypsum, tile adhesive, mortar, plaster and paint coatings, listed in the annexure as "Pull Out Test".
  • Outside the NABL scope: the same method on a wall at site, reported as standard testing, as the quotation stated.

Two tests that share a name in practice

A pull-off test measures how well a layer holds to what is beneath it; a pull-out test, how well an embedded item is held by the material around it.

Pull-offPull-out
What is loadedA finish, coating or render, through a dolly bonded to its surfaceAn embedded bar, anchor, fixing or cast-in insert
Direction of loadPerpendicular to the surface, in direct tensionAlong the axis of the embedded item
What it answersDoes the layer hold to its substrate, and which layer is weakestDoes the material grip the item, and at what load does it slip or break out
Typical standardsASTM D4541, ASTM D7234, EN 1015-12IS 2770 (Part 1) for bond of bars in concrete; ASTM C900 for pullout strength of hardened concrete
ResultStress at failure, with the failure locationLoad at slip or failure, or a bond stress

The confusion has a respectable origin: the annexure to TCR's own NABL certificate lists the ASTM D4541 test as "Pull Out Test". A request for a "pull out test on punning" is, in nearly every case, a request for a pull-off test.

The genuine pull-out test construction projects meet most often is the bond test on a reinforcing bar cast into a concrete cube, described in the post on the pull-out test on TMT bars to IS 2770. It belongs with TMT rebar testing and shares nothing with a dolly on a wall except the word.

How a pull-off test works on a plastered or punned wall

A pull-off test isolates a small area of finish, bonds a loading fixture to it and pulls it off under steadily increasing tension. On a wall the sequence is:

  1. Preparation. Loose dust is removed, and the report states which surface the dolly was bonded to.
  2. Bonding. A metal dolly of known face area is fixed with a structural adhesive, usually a two-part epoxy, and left until the adhesive reaches the cure its manufacturer states.
  3. Scoring. The finish around the dolly is cut through to the substrate, so the pull acts only on the area under the dolly. Whether the area was scored is recorded.
  4. Loading. A portable adhesion tester pulls the dolly perpendicular to the surface until something breaks.
  5. Result. Failure load divided by loaded area gives the pull-off strength in MPa (N/mm²), and the location of the break is recorded.

Each step can spoil the result. An adhesive pulled before it has cured fails in the glue. A shallow score leaves the dolly pulling on finish outside its own area; a careless one cracks a brittle skim coat before loading. On porous punning, a thin adhesive can soak in and stiffen the layer, or run down the cut and bridge it to the substrate.

Reading the failure mode

ASTM D4541 requires the nature of the failure to be reported, and a pull-off number without its failure mode cannot be judged.

Where the break occursCalledWhat the number means
At the interface between finish and substrateAdhesive failureThe bond strength of that interface, which is what most specifiers assume every pull-off measures
Within the finish layerCohesive failure of the finishThe tensile strength of the finish; the bond is at least as strong
Within the substrate, with a plug of ALC, concrete or plaster on the dollyCohesive failure of the substrateThe near-surface tensile strength of the substrate; finish and bond are at least as strong
In the adhesive, or between adhesive and dollyGlue failureNothing about the finish; the test did not reach the system

Mixed failures are common and are recorded as proportions of the dolly area. A glue failure is not a low result; it is no result, and only shows that the finish withstood at least that stress.

Why punning on ALC and plaster on RCC give different numbers

A pull-off test breaks the weakest layer in the stack, whatever the specifier intended to measure. The three points in this enquiry are three different stacks.

Punning on an ALC panel. Autoclaved lightweight concrete is porous and of low density, and its near-surface tensile strength is low compared with dense concrete. A well-bonded skim coat will often pull a thin layer of the panel away with it. The result then measures the ALC, and the correct reading is that the bond exceeded the substrate; rejecting the punning on that number rejects the wrong material. A dry, unprimed panel can also draw water out of a fresh skim coat before it sets, a bond defect the test finds as adhesive failure at the interface.

Punning or plaster on RCC. A cast concrete face is dense and strong, so the risk runs the other way: a smooth formwork face, residual release agent or laitance can leave the finish poorly keyed, and the break comes at the interface. On RCC, an adhesive failure at low stress is a real finding about surface preparation.

Cement plaster. A plaster coat is thicker and more variable than a skim coat, and plaster tested young can fail cohesively at a stress it would exceed later, so age and curing belong in the enquiry.

"Many enquiries ask us for a pull-out test when what they need is a pull-off test, and that is understandable, because even accreditation documents use the two words loosely. The harder part comes after the pull. On punning over an ALC panel, the dolly very often comes away with a layer of the panel on it. That is not a weak bond, it is a weak substrate, and the report has to say so plainly. A number without its failure mode is not a result anyone should accept or reject a wall on."

Parul Hariya, Head of Civil and Building Materials Testing, TCR Engineering Services Pvt. Ltd.

Parul Hariya heads civil and building materials testing at TCR Engineering, where pull-off testing sits within building materials testing.

Why one point is three dollies

Pull-off results on cementitious finishes scatter widely: local skim thickness, a void, a sand grain at the interface or a scoring micro-crack each changes the stress at which one dolly fails. TCR's quotation for this enquiry defined one point as three dollies, which gives a mean, a range and three failure modes. If two break in the substrate and one at the interface at a lower stress, the interface failure is the finding. If all three break in the glue, the point has not been tested. Three points on three elements is a spot check, not a survey; a specification meant to accept or reject finishes should state points per area.

ASTM D4541, ASTM D7234 and EN 1015-12

ASTM D4541, the method on TCR's NABL scope, covers pull-off strength of coatings using portable adhesion testers. It was written with metal substrates in view and recognises that results can differ between types of tester and between substrates. Two other standards address these substrates directly:

  • ASTM D7234 covers pull-off adhesion strength of coatings on concrete using portable testers. A break within the concrete is a common outcome and is reported with the value.
  • EN 1015-12 covers the adhesive strength of hardened rendering and plastering mortars on substrates, and is the method European specifications cite for plaster bond.

All three share one principle but differ in fixture size, preparation and reporting, so a result under one is not automatically comparable with an acceptance value written under another. The specification should name one method.

The accreditation boundary: laboratory yes, site no

NABL accredits specific tests at specific locations, and the annexure to certificate NABLT0726MH18640 lists permanent laboratory testing and site testing separately. ASTM D4541 on finishing products and plaster is in the laboratory section; the site section carries no pull-off or ASTM D4541 item.

TestWhereOn NABL scope
Pull-off strength, ASTM D4541, on finishing product, gypsum, tile adhesive, mortar, plaster or paint coatingMahape laboratoryYes
Pull-off strength, ASTM D4541, on plaster or punning on an RCC column, ALC panel or wallAt siteNo
Compressive strength of autoclaved cellular aerated (AAC) blocks, IS 6441 (Part 5)Mahape laboratoryYes
Density of autoclaved cellular aerated (AAC) blocks, IS 6441 (Part 1)Mahape laboratoryYes
Bond pull-out of reinforcing bars, IS 2770 (Part 1)LaboratoryNo

The site test uses the same method and apparatus and is reported as standard testing. A client who needs an accredited result has one route: the finish system applied to substrate panels and tested in the laboratory. A panel shows what the system can achieve; a site test shows what was achieved on that wall, and a workmanship dispute is usually settled by the second. The full annexure is on the quality and accreditation page.

What sets the pass value

ASTM D4541 sets the method, not the acceptance value. That comes from the project specification, the manufacturer's declared value for a proprietary product applied as instructed, or a product standard the specification references.

A usable criterion also deals with failure mode. "Minimum pull-off strength X, or cohesive failure within the substrate at any value" works. "Minimum pull-off strength X" on its own will sooner or later fail a skim coat on ALC because the ALC broke.

What an enquiry should carry

  1. The system: substrate (RCC, ALC panel, blockwork, plaster) and each layer over it, with product names where proprietary.
  2. The question: bond of the punning, bond of the plaster, or the whole finish system.
  3. Age and curing: when each layer was applied and how it was cured.
  4. Points: number of locations, which elements, and dollies per point.
  5. Location and access: floor, height, access, and whether the finish may be patched afterwards.
  6. The acceptance value and its source: specification clause, manufacturer's declaration or product standard, and the method it assumes.
  7. Accreditation: whether an accredited result is needed, in which case laboratory panels should be considered alongside site points.

Pull-off testing is part of TCR's civil testing practice for construction and infrastructure projects; related posts cover AAC block testing and tile adhesive testing to EN 12004. Schedule and site visits are confirmed in the quotation.

Frequently asked questions

What is the difference between a pull-off test and a pull-out test?

A pull-off test bonds a dolly to a finish or coating and pulls it perpendicular to the surface, measuring how well the layer holds to its substrate. A pull-out test extracts an embedded item, such as a reinforcing bar, an anchor or a cast-in insert, from the material around it. They use different specimens, standards and equipment.

Is ASTM D4541 pull-off testing on TCR Engineering's NABL scope?

In the laboratory, yes. The annexure to certificate NABLT0726MH18640 lists ASTM D4541 for finishing products, gypsum, tile adhesive, mortar, plaster and paint coatings under permanent laboratory testing. The site testing section carries no pull-off item, so a pull-off test on a wall at site is reported as standard testing.

Why does a pull-off test on punning over an ALC panel often break inside the panel?

ALC is a porous, low-density substrate with a low tensile strength near its surface. When the punning is well bonded, the weakest layer is often the panel itself, and the dolly comes away carrying a thin layer of ALC. The result then measures the substrate, and shows that the bond was at least that strong.

Why does one pull-off test point use three dollies?

Pull-off results on cementitious finishes scatter because of local thickness, voids, sand grains and scoring damage. Three dollies at one point give a mean, a range and three failure modes, so a single unrepresentative pull does not decide the result. The number of points across a building is a separate decision for the specification.

What acceptance value applies to a pull-off test on plaster or punning?

ASTM D4541 sets the test method but no pass value. The acceptance value comes from the project specification, the manufacturer's declared value for the product, or a product standard the specification references. A good criterion also accepts cohesive failure within the substrate at any stress, since that shows the finish bonded well.

Bring us your hardest integrity problem.

Request a Quote
Chat on WhatsApp