Refined Beam Model for AFM Cantilevers: Analysis of Elastic Fixing

Supervisor: Ustinov K.B.

Reviewer: Goldstein R.V.

Submitted: July 2012

---

1. Abstract

Standard AFM force spectroscopy models assume a perfectly rigid clamp ($U(0)=0, U'(0)=0$). However, real-world cantilevers are attached to a substrate (base) that deforms under load, introducing significant compliance errors. This project refines the standard beam model by introducing an Elastic Fixing Coefficient ($\beta$) that accounts for the rotation and displacement at the cantilever root. Our 3D FEM analysis reveals that for typical materials, this effect is non-negligible and can introduce errors in stiffness calibration if ignored.

Figure 1: Working principle of Atomic Force Microscopy showing the laser deflection detection method.

2. Theoretical Background

Standard vs. Refined Model

Standard Beam Theory: Assumes the base is infinitely rigid.

Refined Model: Models the base as an elastic half-space. The boundary condition at the root becomes:

$$ \theta_{root} = \frac{M_{root}}{k_{\theta}} $$

Where $k_{\theta}$ is the rotational stiffness of the clamp.

Figure 2: Simplified beam model with fixed end (A) and free end (B).

Governing Equations

For a cantilever under a point load $f$ at length $L$:

Rigid Clamping:

$$ U_{rigid}(L) = \frac{f L^3}{3 E I} $$

Elastic Clamping (Refined):

$$ U_{elastic}(L) = \frac{f L^3}{3 E I}\left(1 + 3\beta\frac{H}{L}\right) $$

Elastic Fixing Coefficient ($\beta$)

The thesis defines $\beta$ via the deflection ratio derived from the above equations:

$$ \frac{U_{elastic}}{U_{rigid}} = 1 + 3\beta\frac{H}{L} $$

where:

• $H$: Cantilever thickness
• $L$: Cantilever length
• $I$: Area moment of inertia ($I = \frac{W H^3}{12}$)

If the clamp is modeled as a rotational spring with stiffness $k_{\theta}$,

$$ \theta_{root} = \frac{M_{root}}{k_{\theta}}, $$

then the same correction can be written in the standard form

$$ \frac{U_{elastic}}{U_{rigid}} = 1 + \frac{3 E I}{k_{\theta} L}. $$

Equating the two expressions gives an explicit link between $\beta$ and $k_{\theta}$:

$$ \frac{3 E I}{k_{\theta} L} = 3\beta\frac{H}{L} ;;\Rightarrow;; \beta = \frac{E I}{k_{\theta} H}. $$

The FEM results in this repository are used to estimate $k_{\theta}$ (and therefore $\beta$) as a function of base modulus and geometry.

3. Methodology

3.1 Finite Element Simulation (ANSYS APDL)

A parametric 3D FEM model was developed in ANSYS to simulate the bending of a rectangular cantilever on a massive elastic base.

Figure 3: 3D geometric model of the cantilever defined by length ($L$), width ($W$), and thickness ($H$).

• Geometry: Parametric sweep varying Base Modulus ($E_b$), Width ($W$), and Thickness ($H$).
• Meshing: Refined mesh at the stress concentration zone (cantilever root) to capture local deformation gradients.
• Loading: Point force applied at the free end.
• Output: Tip displacement ($U_z$), Root Moment ($M_{root}$), and Root Rotation ($\theta_{root}$).

3.2 Data Analysis

Results were post-processed using Wolfram Mathematica to extract:

1. Rotational Stiffness ($k_{\theta}$): Calculated as $M_{root} / \theta_{root}$.
2. Normalized Stiffness ($R$): Dimensionless parameter $R = k_{\theta} / (E I / L)$ to isolate geometric effects.

4. Results & Discussion

4.1 Effect of Base Modulus ($E_b$)

The stiffness of the clamp is strongly dependent on the Young's Modulus of the base. As expected, a softer base leads to higher root rotation and lower effective stiffness.

Figure 4: Log-log plot showing the increase in Rotational Stiffness as the Base Modulus increases.

4.2 Normalized Stiffness Analysis

To compare different geometries, we normalize the rotational stiffness. The results show that even for very stiff bases, the clamp is not perfectly rigid ($R$ does not approach infinity), confirming the necessity of the elastic fixing model.

Figure 5: Normalized Stiffness saturates at a finite value, indicating the inherent compliance of the 3D connection.

4.3 Geometric Dependencies

The analysis reveals that the cantilever thickness ($H$) plays a dominant role in the elastic fixing effect. Thicker cantilevers (larger $H/L$ ratio) induce greater local deformation in the base relative to their own bending, leading to a more pronounced deviation from the rigid theory.

Figure 6: 3D Surface plot illustrating the sensitivity of Rotational Stiffness to geometric parameters for a soft base.

4.4 Stress Analysis

The simulation also evaluates the maximum Von Mises stress at the cantilever root. This serves as a validity check for the beam model assumptions and helps in assessing the structural integrity of the clamp under load.

Figure 7: 3D Surface plot of Maximum Stress vs Geometry.

Animation 1: Progressive bending (deformation) of the cantilever under load.

Animation 2: Evolution of Von Mises stress distribution during bending.

5. Key Conclusions

1. Validity of Elastic Fixing: The simulations confirm that the "rigid clamp" assumption is an idealization. Real clamps exhibit finite rotational stiffness.
2. Magnitude of Error: For soft bases ($E_{base} \approx 0.1 E_{cant}$), the deflection can be ~0.6% larger than theoretical predictions, corresponding to an elastic fixing coefficient $\beta \approx 1.0$ in our 3D model.
3. Consistency with Theory: These results align with the thesis predictions ($\beta_{2D} \approx 2.0$), with the difference attributable to 3D constraints (Poisson effects) which stiffen the base slightly compared to 2D plane strain models.

6. Usage

Prerequisites

• ANSYS Mechanical APDL 13
• Wolfram Mathematica 9

7. License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.