Designing a 3M VHB Adhesive Hook: From a 1 kg Load to Creep and Edge Peeling

Abstract

A stick-on hook holds a pair of jeans immediately after installation, then falls off later. That observation started this design study. This article examines a 3D-printed hook using VHB tape 20 mm wide and 1.1 mm thick, with a maximum clothing load of 1 kg. The initial target was a base 45 mm wide and 60 mm tall. Multiple tape pieces could be used, but the roll’s full width had to be retained.

Preliminary calculations show that the small base falls well short of the chosen static-load area guideline. A 45 × 170 mm concept increases the bonded area while keeping the hook projection short to reduce the bending moment.

Figure 1. Concept sketch from the design discussion: a 45 × 170 mm base with eight 20 × 42 mm tape pieces. The AI-generated image communicates the intended shape.

1. The problem and what we know

The initial observation was a hook detaching after carrying a hanging load for some time, with signs resembling downward slip. This suggests creep as a possible explanation, but does not establish the failure mechanism. The separated surfaces, the point where detachment began, and displacement before failure all need to be examined.

Parameter	Value used in this article
Mass of jeans	Up to 1.00 kg
Hook mass	Assumed 0.10 kg; the actual prototype must be weighed
Total mass for preliminary calculations	1.10 kg
Tape	20 mm wide, 1.1 mm thick; product grade not yet identified
Initial base	Maximum width 45 mm; height 60 mm
Tape-cutting constraint	Cross-cutting allowed; retain the full 20 mm roll width
Illustrative moment arm	15 mm from the bond plane to the load’s line of action
Scope	Hook on a vertical surface; sustained hanging load, with loading and unloading

A thickness of 1.1 mm is not enough to identify the VHB grade. This article uses the 5952 documentation for general principles and a preliminary guideline [1]. Final tape selection still requires the correct grade, substrate materials, and operating conditions. The additional dimensions and calculations below are part of this article’s design analysis.

2. Two simultaneous effects: shear force and bending moment

For total mass m and gravitational acceleration g = 9.81 m/s², the downward force is W = m·g. If A is the effective bonded area, the average shear stress is τ = W/A. “Average” matters here: local loading also depends on base stiffness, tape placement, and contact quality.

The jeans hang at an offset e from the bond plane, also producing a moment M = W·e. This moment tends to pull the top edge away from the substrate and press the lower region into it. Once the top edge opens, the remaining bonded area decreases and failure may progress.

Figure 2. A simple force-equilibrium model. Orange indicates the tape; teal indicates the base and hook. The diagram explains the loading tendency, not the actual stress distribution.

The distance to minimize is the offset perpendicular to the wall. Moving the hook up or down while keeping e unchanged leaves the moment from the vertical force unchanged. Vertical position can change how the base bends and transfers load; it does not eliminate W·e.

For an ideal rigid base with a uniform elastic bond over a rectangle of width b and height H, the edge normal stress due to the moment is approximately σmax ≈ 6M/(b·H²). H is the bonded height; the distance h between the tensile and compressive resultants is derived in Section 7. This model helps compare geometries. Segmented VHB tape and a flexible base require consideration of the actual bond distribution.

3. Why is a pull test lasting a few seconds not enough?

3M describes VHB as viscoelastic: its response depends on loading rate, and creep can occur under sustained loading. The 5952 documentation gives a preliminary rule of approximately 57 cm² of tape per kilogram of static load to limit creep [1].

Here, creep means displacement that grows while the load remains constant. For an adhesive hook, both downward slip and opening at the top edge should be tracked. A strong pull by hand mainly reveals behavior during that particular action; it does not tell us the displacement after a month.

The 57 cm²/kg guideline is an area-screening tool. A product load rating also requires evaluation of the substrate, geometry, temperature, installation process, and test results. In particular, bending moment and edge peeling are not resolved by dividing W by A.

4. Checking the 45 × 60 mm base

Even with full coverage of the rear surface, the geometric area is only:

A_max = 45 × 60 = 2,700 mm² = 27 cm².

The reference area for the 1 kg jeans alone is 57 cm². Assuming a 0.10 kg hook raises that reference to:

A_reference = 57 × 1.10 = 62.7 cm².

The ideal 27 cm² base provides only about 43% of this reference area. A practical layout with two 20 × 55 mm strips and top and bottom margins provides 22 cm². Its average shear stress is 10.791 N / 0.0022 m² ≈ 4.91 kPa.

Design conclusion: the selected guideline does not support assigning a sustained 1 kg load rating to the 45 × 60 mm configuration. This does not prove that every bond this small will fail immediately. It shows that this design is already short of area at the screening stage, even before edge peeling and installation variation are considered.

5. A taller 45 × 170 mm base concept

The concept from the discussion retains the 45 mm width and increases the height to 170 mm. A short J-shaped hook holds a belt loop; two longitudinal ribs stiffen the base, and a triangular rib supports the hook root. The aim is to keep the load close to the substrate and limit deformation of the printed part.

Eight pieces, each retaining the full 20 mm roll width and cut to a length of 42 mm, provide:

A = 8 × 20 × 42 = 6,720 mm² = 67.2 cm².

Place each 42 mm edge across the base. Vertically, the eight pieces occupy 160 mm, seven 1 mm gaps occupy 7 mm, and two 1.5 mm margins occupy 3 mm, totaling 170 mm. The rear face must be flat and retain enough bonding area after corner rounding and dimensional tolerances are considered.

Figure 3. Nominal dimensioned layouts. Two vertical strips are an alternative that reduces cutting and the number of joints. No test data yet establishes which arrangement is more durable.

Assuming m = 1.10 kg and e = 15 mm:

Quantity	Calculation	Result
Downward force	1.10 × 9.81	10.79 N
Average shear stress	10.791 / 0.00672	1.61 kPa
Bending moment	10.791 × 0.015	0.162 N·m
Area relative to the guideline	67.2 / 62.7	1.072×

For illustration, the entire mass is assigned the same 15 mm moment arm. Once CAD and an actual prototype are available, use M = m_jeans·g·e_jeans + m_hook·g·e_hook. Each moment arm must include the actual offset through the tape, base, and the point where the belt loop bears on the hook.

The 67.2 cm² area is only about 7.2% above the 62.7 cm² guideline. This is a nominal area margin, not a product safety factor of 1.072. Uncovered areas, rounded corners, misalignment, or poor contact can consume that margin.

6. How compact can the design be?

Two strips retaining the full 20 mm width provide a combined tape width of 40 mm. Reaching the 6,270 mm² reference area requires a bonded length of at least 6,270/40 = 156.75 mm. Adding a total of 5 mm for the two end margins gives a geometric height of about 161.75 mm. This is a calculated limit for this layout, not a validated product dimension.

Figure 4. Calculated nominal bonded areas. A bar above the reference line only indicates that the area-screening criterion has been met.

Base and tape layout	Area	Margin relative to 62.7 cm²
45 × 60 mm; two 20 × 55 mm strips	22.0 cm²	64.9% short
45 × 165 mm; two 20 × 160 mm strips	64.0 cm²	2.1%
45 × 170 mm; two 20 × 165 mm strips	66.0 cm²	5.3%
45 × 170 mm; eight 20 × 42 mm pieces	67.2 cm²	7.2%

The 165 mm version saves 5 mm but uses almost all of the calculated margin. Among the options considered, 170 mm is therefore a more suitable starting point for initial testing. If a 60 mm height is mandatory, the load path or mounting method needs to change, for example to an over-door hook or a mechanical fastener carrying the main load. Each alternative requires its own redesign.

Two vertical strips have fewer joints, while eight horizontal pieces use a 42 mm width. Changing strip orientation alone does not change W/A when area is unchanged. Real differences come from tape distribution, base stiffness, application pressure, and bond edges. These require comparative testing rather than an assumption that one orientation is always better.

7. How large must the separating force T be before edge peeling begins?

Three quantities need to be distinguished: the tensile resultant T in N, normal stress σ in Pa, and peel strength, often reported in N/cm. They describe different aspects of the bond. Even a small T can be concentrated in a narrow region, so T alone does not determine when peeling begins.

7.1. From bending moment to stress distribution

Assume a rigid base, a bond layer of uniform thickness and normal stiffness, small deformation, a fully intact bond, and no external force perpendicular to the wall. Let H be the height of the continuous bonded region and b its width. Capital H distinguishes it from h, the distance between the lines of action of the tensile and compressive resultants.

Set y = 0 at the center of the bonded region. A small base rotation produces an opening that varies linearly with y; in the elastic model, stress is also linear: σ(y) = c·y. Moment equilibrium gives M = ∫σ(y)·y·dA = c·bH³/12, hence:

σ(y) = 12M·y/(bH³); σmax = 6M/(bH²).

Integrating the tensile stress over the upper half gives T = ∫₀ᴴ⁄² σ(y)·b·dy. The compressive resultant in the lower half has the same magnitude, C = T. The centroids of the two triangular stress distributions are at y = +H/3 and −H/3, so:

T = C = σmax·bH/4 = 3M/(2H); h = 2H/3; M = T·h.

Here, h is the distance between the two resultants, not the distance from the J-hook to the top edge. Using h = H instead of 2H/3 underestimates T relative to this triangular-distribution model.

Figure 5. Left: the area under the positive stress distribution, multiplied by width b, gives the resultant T. Right: with W, e, and b fixed, doubling H halves T and reduces σmax to one quarter. These relationships apply only while the model assumptions remain valid.

To stay consistent with the earlier calculations, use a total mass of 1.10 kg, e = 15 mm, b = 42 mm, and M = 0.161865 N·m. A continuous bond 60 mm tall gives T ≈ 4.05 N and σmax ≈ 6.42 kPa; at 160 mm tall, T ≈ 1.52 N and σmax ≈ 0.903 kPa. Counting only the 1.00 kg jeans, as in the discussion example, gives T ≈ 1.38 N and σmax ≈ 0.821 kPa for H = 160 mm. The difference comes from the mass included in the calculation.

For eight tape pieces separated by gaps, the total 8 × 20 = 160 mm cannot be treated exactly as a continuous bonded height. A more general expression under the same rigid-base assumption is σ(y) = M·(y − ȳ)/I_A, where I_A = ∫(y − ȳ)²dA is integrated only over areas containing tape. The eight-piece layout in Figure 3 spans 167 mm, with I_A ≈ 15.782 × 10⁶ mm⁴. Integrating each piece gives T ≈ 1.45 N and an outer-edge stress of approximately 0.856 kPa.

7.2. Why not use the datasheet’s peel value as the allowable T?

The September 2024 English-US 5952 technical data sheet reports a typical 90° peel strength of 39 N/cm, tested on stainless steel with an aluminum backing at 23 °C after 72 hours, at a rate of 304 mm/min. It also reports normal tensile strength of 480 kPa and overlap shear strength of 550 kPa from different tests [3].

Multiplying 39 N/cm by a 2 cm peel-specimen width gives 78 N for that conversion. This is not the hook’s T limit: a flexible peel specimen is pulled at a specified angle and rate, whereas the hook base transfers a moment into a bond under sustained load. The roll width also does not necessarily equal the width of the advancing peel front in the actual layout.

Likewise, 480/0.903 cannot be called the hook’s safety factor. The numerator is a material-test result; the denominator is a nominal stress from a rigid-base model. They do not yet represent the same failure mechanism, duration, or conditions.

7.3. Assessing debonding initiation and propagation

A stress-based assessment needs allowable values appropriate to the actual tape, substrate, temperature, load duration, and combined tensile–shear loading. Once part of the bond has separated, the possibility of that region growing must also be considered. 3M’s modeling guidance distinguishes damage initiation from damage evolution, with fracture energy Gc as one parameter in a cohesive-zone model [4].

The equations above therefore help compare geometries and identify areas to investigate. Calculating a debonding threshold requires additional material and joint data, plus prototype validation. There is not yet enough information to specify a universal maximum T for this design.

8. Does placing the J-hook higher or lower change the tensile and compressive forces?

8.1. Rigid base: the moment and nominal distribution remain unchanged

Let yJ be the load application height measured from the bottom of the bonded region. In side view, the position vector has components (e, yJ), and the load is (0, −W). Their cross product gives a moment of magnitude W·e. The yJ term multiplies a horizontal force of zero, so it contributes nothing.

Thus, moving only the hook’s vertical position while holding W, e, the bonded region, and bond stiffness constant gives the same σ(y), T, and C in the rigid-base model. A lower hook does not automatically create a longer lever arm resisting peeling. A user pulling the jeans outward introduces a horizontal force, while a changing contact point may alter e; those are different load cases.

8.2. Flexible base: where the moment enters the base matters

Overall equilibrium still requires the bond to resist W·e. However, a flexible base can bend locally near the hook root. The regions of tape in tension and compression then depend on where the moment is introduced. An extended section of base may carry very little load if deformation is concentrated near the hook.

To illustrate this, Figure 6 uses an Euler–Bernoulli beam on a linear Winkler spring foundation, whose reaction is proportional to displacement [5]. Both beam ends are free; springs carry tension and compression; and a concentrated moment M is applied at yJ/H = 0.2, 0.5, and 0.8. This supplementary elastic model was calculated for the article and does not use identified VHB 5952 material parameters.

Figure 6. The vertical axis is position along the bonded region; the orange dotted line marks the hook position. The horizontal axis is normalized stress: positive for tension and negative for compression. Dashed curves represent a rigid base; solid curves represent one flexible-base case. All three cases have the same total moment.

For the rigid-base model, the three dashed curves coincide. In the flexible model, the tensile and compressive regions move with the hook, and the peak locations and values change. In this example, a high hook puts the top edge under strong tension, while a low hook moves the main tensile region close to the area above the hook root and increases compression at the bottom. These results depend on relative stiffness and boundary conditions; they do not establish one optimum location for every 3D-printed hook.

The illustrative parameter is λ = KH⁴/(EI) = 10⁴, where EI is the base’s flexural rigidity and K is foundation stiffness per unit length. In this one-dimensional model, the distributed reaction has magnitude q = K·u, and normal stress is σ = q/b. λ is a dimensionless value chosen to illustrate local bending, not a measured parameter of the hook. T* = TH/M and C* = CH/M sum all tensile or compressive regions; the asterisk denotes normalization. The model does not solve in-plane shear, creep, or debonding propagation.

The solution uses beam elements with a continuous elastic foundation. Normal-force and moment equilibrium were checked, followed by comparison of 80- and 160-element meshes. The maximum stress-curve difference was below 0.001% for the illustrated cases. This checks the numerical solution; it is not experimental validation of the model.

For the current design, bring the load close to the wall, stiffen the load path from the hook root into the base, then compare hook positions on specimens made with the same material and bonding procedure. Extra base height is fully useful only when that part of the base actually transfers load.

9. Design conclusions

Three direct and manageable variables are tape area, moment arm, and hook stiffness. The 45 × 60 mm layout lacks area under the chosen preliminary guideline. The 45 × 170 mm concept with eight 20 × 42 mm pieces provides 67.2 cm², enough to proceed to prototyping, but with only a small nominal margin.

The next step is to identify the tape grade, substrate material, actual mass, and geometry, then test both tape layouts under the same conditions. Only then is there a basis for marking a working load on the product. The current design’s value lies in its checkable reasoning: from requirements and force equilibrium through area and geometry to testing.

The additional analysis shows why resultant T must be distinguished from edge stress and datasheet peel strength. With a rigid base, moving the J-hook vertically leaves the nominal distribution unchanged if e is constant. With a flexible base, the moment application point changes local loading. Hook position therefore has to be assessed together with base stiffness.

References and scope of evidence

[1] 3M, “3M VHB Tape 5952 — Technical Data Sheet,” English edition, October 2022, Design Considerations, pp. 5–6. Source for viscoelastic behavior and the static-load guideline of 57 cm²/kg. [Source]

[2] 3M, “Surface Preparation for 3M VHB Tape Applications,” General Procedure. Source for surface preparation, application pressure, and bond-strength buildup over time. [Source]

[3] 3M, “3M VHB Tape 5952,” English-US, revised September 2024, p. 2: Typical Performance Characteristics. Test conditions and values for 90° peel, normal tension, and overlap shear. [Source]

[4] 3M, “Introduction to Modeling Structural Adhesives,” Sections 4–5: cohesive-zone modeling and model validation. Used as methodological guidance; it does not provide a hook-specific debonding threshold. [Source]

[5] NPTEL, “11 Beam on Elastic Foundation,” Sections 11.2 and 11.4: elastic-foundation models and concentrated-moment loading. Figure 6 is this article’s own numerical solution for a finite beam, not a figure copied from the reference. [Source]

Analysis of edge-separating forces and J-hook position added on September 28, 2026. The additional plots are results from illustrative models, not product load-test data.